2014-02-25 12:48  mkrack

	* [r13638] cp2k_info.F: Add sub-release counter

2014-02-25 10:24  mkrack

	* [r13635] cp2k_info.F: Store the revision number of the release
	  branch

2014-02-25 09:40  mkrack

	* [r13634] Copy the CP2k trunk version (SVN revision 13632) to the
	  release branch

2014-02-24 07:14  vondele

	* [r13632] change cholesky default from reduce to restore, which is
	  faster in some cases.

2014-02-23 12:50  vondele

	* [r13629] fix potential overflow

2014-02-23 12:47  vondele

	* [r13628] prettify

2014-02-21 16:31  ibethune

	* [r13625] Add a __MKL flag to protect an mkl header include.

2014-02-21 15:50  agloess

	* [r13624] Add configuration for DBCSR library.

2014-02-21 14:59  ibethune

	* [r13622] Make MKL check more specific

2014-02-21 12:28  mkrack

	* [r13619] Prettify

2014-02-21 12:04  jhutter

	* [r13618] Use Jacobi diagonalization: Improve numerical stability
	  of atomic code
	  Reset all regtests of atomic code
	  Symmetrize relativistic correction matrix

2014-02-20 16:44  agloess

	* [r13616] Remove unused variables and prettify.

2014-02-20 16:43  agloess

	* [r13615] Add initialization of DBCSR library and prettify.

2014-02-20 11:13  fschiff

	* [r13613] leak fix

2014-02-19 13:38  jhutter

	* [r13612] vdW functionals. Don't call for virial if energy_only
	  flag is set.

2014-02-16 10:28  vondele

	* [r13610] improve format and output

2014-02-16 08:39  vondele

	* [r13609] Bugfix: don't use size of deallocated variable
	  (Vittorio).

2014-02-13 17:50  oschuett

	* [r13608] Fix: NAG preprocessor does not support #error directive

2014-02-13 16:56  mkrack

	* [r13607] Prettify

2014-02-13 10:25  agloess

	* [r13606] Fix correct variable type

2014-02-13 09:41  agloess

	* [r13605] Remove unused variables

2014-02-13 08:09  agloess

	* [r13604] Bugfix symmetric-transposed matrix multiplication

2014-02-12 17:29  oschuett

	* [r13603] Remove OBJECTS_ARCHITECTURE, rely on compiler-flags

2014-02-12 15:33  oschuett

	* [r13602] remove unused directory src/elpa_lib

2014-02-12 15:09  oschuett

	* [r13601] remove unused file

2014-02-12 14:22  oschuett

	* [r13600] Change extension of included files to .f90

2014-02-12 12:59  fschiff

	* [r13599] fix for DFTB EMD. dftb overlap matrix now directly
	  computable

2014-02-12 10:34  ibethune

	* [r13597] Add support for regions when using Langevin MD,
	  following the approach
	  of Kantorovich et al, 2008. 4 new regtests added. (Lianheng Tong)

2014-02-11 15:15  oschuett

	* [r13595] remove unused variables

2014-02-11 15:09  mkrack

	* [r13594] Generic description independent of the actual compile
	  and link flags

2014-02-11 14:33  oschuett

	* [r13593] Split dirctory cuda_tools into src/cuda and
	  src/dbcsr_lib/cuda

2014-02-11 13:50  oschuett

	* [r13591] prettify: Fix continuation lines interrupted by empty
	  lines.

2014-02-11 12:47  oschuett

	* [r13590] Fix compilation with OpenMP

2014-02-10 17:12  tlaino

	* [r13589] bug fix

2014-02-10 13:05  jhutter

	* [r13579] Deallocate matrix before recalculation (memory leak)

2014-02-08 16:51  tlaino

	* [r13577] fix

2014-02-07 19:13  oschuett

	* [r13576] Rename directory tools/metadyn into src/metadyn_tools

2014-02-07 17:55  oschuett

	* [r13574] Change extension of included files to .f90

2014-02-07 16:42  jhutter

	* [r13573] Fix memory leak, OpenMP parallel statement, KG regtest

2014-02-07 16:41  jhutter

	* [r13572] Fix parallel bug for XRD interface

2014-02-07 16:35  oschuett

	* [r13571] Change extension of included files to .f90

2014-02-07 16:30  fschiff

	* [r13570] DFTB should work now with RTP/EMD

2014-02-07 16:23  oschuett

	* [r13569] Remove unused files.

2014-02-07 16:02  oschuett

	* [r13568] Add se_fock_matrix.o to OBJECTDEFS

2014-02-07 15:55  oschuett

	* [r13567] Add se_fock_matrix_coulomb.o to OBJECTDEFS

2014-02-07 15:23  oschuett

	* [r13566] Move directory machine into src

2014-02-07 15:17  oschuett

	* [r13565] Remove testing-program from module file min_heap.F

2014-02-07 15:16  oschuett

	* [r13564] Remove dbcsr_example_*.o from OBJECTDEFS

2014-02-07 15:10  oschuett

	* [r13563] Prettify

2014-02-07 14:51  oschuett

	* [r13562] Remove #ifdef-block from USE-statements

2014-02-07 14:03  jhutter

	* [r13561] New approximation for KG method. Nonadditive kinetic
	  energy modeled by
	  a local atomic potential. Regtests for the method and the
	  generation of
	  the potential (atom code) added.

2014-02-06 13:17  fschiff

	* [r13559] bug fix for scaling and squaring for Taylor exp

2014-02-06 13:00  delben

	* [r13558] Improved communication scheme in RI-RPA and
	  RI-Laplace-MP2 including OMP
	  parallelization

2014-02-06 12:25  mkrack

	* [r13557] Bug fix for smear (init HOMO value)

2014-02-06 08:10  mkrack

	* [r13556] Prettify

2014-02-05 18:53  marci73

	* [r13555] bug fix

2014-02-05 18:27  marci73

	* [r13554] 2d periodic not orthorhombic

2014-02-05 16:08  fschiff

	* [r13550] rtp_mos_type, encapsulation, clean-up, bug-fix (Samuel
	  Andermatt)

2014-02-05 15:19  mkrack

	* [r13549] Prettify

2014-02-05 13:05  fschiff

	* [r13548] enable MO extrapolation for exponential methods

2014-02-05 12:05  fschiff

	* [r13547] more refactoring and making ETRS and EM extrapolation
	  consistent

2014-02-05 11:50  marci73

	* [r13546] ipi interface

2014-02-05 08:09  vondele

	* [r13544] some more OMP

2014-02-04 14:56  fschiff

	* [r13543] refactoring and cleaning RTP/EMD step 2

2014-02-04 08:45  oschuett

	* [r13540] MP2: Dellocate DBCSR memory-pools prior to large hybrid
	  dgemm/pdgemm calls

2014-02-03 15:46  oschuett

	* [r13539] make cp_result independent from input_val_types

2014-02-03 15:45  oschuett

	* [r13538] make cp_blacs_env independent from input_constants

2014-02-03 15:45  oschuett

	* [r13537] make cuda_memory independent from input_section_types

2014-02-03 15:44  oschuett

	* [r13536] make distribution_1d_types independent from
	  input_section_types

2014-02-03 15:43  oschuett

	* [r13535] move some routines from cp_dbcsr_operations and
	  cp_dbcsr_interface into new module cp_dbcsr_cp2k_link

2014-02-03 15:42  oschuett

	* [r13534] make realspace_grid_cube independent from
	  atomic_kind_types and particle_list_types

2014-02-03 15:41  oschuett

	* [r13533] make pw-modules independent from input

2014-02-03 15:39  oschuett

	* [r13532] make realspace_grid_types independent from
	  input_constants

2014-02-03 15:38  oschuett

	* [r13531] make realspace_grid_types independent from
	  input_section_types

2014-02-03 15:36  oschuett

	* [r13530] make pw_spline_utils independent from
	  input_section_types and cp_output_handling

2014-02-03 15:34  oschuett

	* [r13529] make cp_ma_interface independent of cp_output_handling

2014-02-03 15:33  oschuett

	* [r13528] make cp_log_handling and cp_ma_interface independent of
	  cp2k_info

2014-02-03 15:31  oschuett

	* [r13527] make pw-modules independent of cell_type

2014-02-03 14:01  mkrack

	* [r13526] Inprove numerical handling for orthorhombic cells

2014-02-03 13:39  mkrack

	* [r13525] Prettify

2014-02-03 12:51  fschiff

	* [r13524] leak fix

2014-02-03 10:39  mkrack

	* [r13523] Prettify

2014-02-03 10:33  vondele

	* [r13522] bug fix threaded case

2014-02-03 10:33  fschiff

	* [r13521] RTP refactoring. Remove separate CN routines as it is
	  nothing more than 1st order EM/PADE

2014-02-03 10:17  vondele

	* [r13520] Improve docs, always output cell orthorhombic

2014-02-03 09:34  fschiff

	* [r13519] beginning of RTP refactoring and cleaning. Removed EXP
	  via diagonalization, added convinience function to perform a
	  complex gemm using fm matrices (Samuel Andermatt)

2014-02-03 09:20  oschuett

	* [r13518] Input Description Change (Iain Bethune)

2014-02-03 09:15  oschuett

	* [r13517] Input Description Change (Maximilien Levesque)

2014-02-03 09:13  oschuett

	* [r13516] Input Description Change

2014-02-01 10:26  vondele

	* [r13515] correct and adjust dbcsr statistics report

2014-02-01 08:50  vondele

	* [r13513] init variable

2014-02-01 00:03  mkrack

	* [r13512] Fix if clause and relax threshold value

2014-02-01 00:01  mkrack

	* [r13511] Prettify

2014-01-31 15:58  marci73

	* [r13510] proper location of vacuum
	  

2014-01-31 15:28  marci73

	* [r13509] mp_sum energy%surf_dipole
	  
	  
	  -This line, and those below, will be ignored--
	  
	  M surface_dipole.F

2014-01-31 14:43  mkrack

	* [r13508] Add DCD dump option for an aligned cell which allows a
	  reconstruction of scaled coordinates from the DCD dump only

2014-01-31 07:09  vondele

	* [r13507] prettify

2014-01-30 22:52  marci73

	* [r13506] reference

2014-01-30 22:51  marci73

	* [r13505] dipole correction for surface supercell calculations (no
	  forces yet)

2014-01-27 15:27  vondele

	* [r13501] workaround: fix huge bound

2014-01-27 11:58  vondele

	* [r13500] fix a few potential overflows

2014-01-24 16:38  delben

	* [r13498] Never free HFX buffer in case numerical RI-MP2 stress
	  tensor is
	  required.

2014-01-24 14:29  delben

	* [r13497] RI-MP2 stress tensor (Analytical and Numerical)

2014-01-20 10:23  mkrack

	* [r13494] Fix typos

2014-01-18 01:25  mkrack

	* [r13493] Fix bottleneck to speedup FIST runs (Joost)

2014-01-17 12:07  vondele

	* [r13492] 2014 a little more

2014-01-17 11:30  vondele

	* [r13491] add reference

2014-01-16 11:38  oschuett

	* [r13489] Timing-report: fix mem-leak

2014-01-16 07:58  marci73

	* [r13488] i-Pi interface (Ceriotti)

2014-01-15 17:34  oschuett

	* [r13487] Fix crashes with missing dipoles in mixed_energy_forces
	  (Samuel Andermatt)

2014-01-15 17:17  oschuett

	* [r13486] DBCSR: Fix statistics

2014-01-15 16:45  oschuett

	* [r13485] Timing-report: Also show routines not called on first
	  MPI-rank

2014-01-14 15:43  oschuett

	* [r13483] GLBOPT: Fix length unit of trajectories (Samuel
	  Andermatt)

2014-01-14 12:34  delben

	* [r13481] Change the status of the restart file in tmc
	  print_restart_file from NEW
	  to REPLACE

2014-01-14 11:40  delben

	* [r13480] An independent atomic_pair_list has been created for HFX
	  forces calculation,
	  this fixes problem if the HFX energy has to be computed a second
	  time
	  after the HFX forces are computed (for example in RI-MP2
	  gradients).

2014-01-13 09:50  oschuett

	* [r13478] Input Description Change

2014-01-13 09:48  oschuett

	* [r13477] Input Description Change (Samuel Andermatt)

2014-01-10 06:37  mkrack

	* [r13476] Stop all processes to avoid hanging runs

2014-01-08 15:26  oschuett

	* [r13475] SWARM: Add timer-regions, speedup restarts

2014-01-08 10:40  oschuett

	* [r13474] GEO_OPT: Pipe status message through print-key.

2014-01-08 10:23  oschuett

	* [r13473] Remove cp_lbfgs_optimizer__nametype1_.template, because
	  it has only one instance.

2014-01-07 15:32  oschuett

	* [r13472] SWARM: Improve Doxygen documention

2014-01-07 15:28  oschuett

	* [r13471] GLBOPT: Output found minima in separate trajectories

2014-01-07 11:04  oschuett

	* [r13470] refactoring cp_subsys_release()

2014-01-06 15:51  delben

	* [r13469] Part of the mp2_ri_gpw module OMP parallelized
	  (gradients part)

2014-01-04 12:25  marci73

	* [r13467] inerface for i-PI, path-integral driver of Michele
	  Ceriotti

2014-01-01 10:56  tlaino

	* [r13466] 2014

2013-12-23 15:31  mkrack

	* [r13463] Fix unit descriptions; add force unit kind

2013-12-18 17:51  oschuett

	* [r13461] Input Description Change

2013-12-18 16:51  mkrack

	* [r13460] Prettify

2013-12-18 16:25  marci73

	* [r13458] fixed problem in x-ray emission calculations

2013-12-18 14:59  mkrack

	* [r13457] Fix usage for the _DEFAULT_KEYWORD_ entries which are
	  not automatically handled.

2013-12-18 14:19  vondele

	* [r13456] Beta version of RI-MP2 gradients (Mauro).

2013-12-18 12:39  mkrack

	* [r13454] Replace "<" where needed

2013-12-17 16:26  mkrack

	* [r13451] Keep case here and handle it at the XSLT level

2013-12-17 15:30  vondele

	* [r13449] fix lone_keyword default for USE_MODEL_HESSIAN

2013-12-16 10:29  oschuett

	* [r13447] Input Description Change (Samuel Andermatt)

2013-12-10 19:22  oschuett

	* [r13438] GLBOPT: Fix history lookup

2013-12-09 19:40  vondele

	* [r13433] omp 3.1 workaround

2013-12-09 14:41  oschuett

	* [r13428] Add regtest for dipole calculation with linear scaling
	  code (Samuel Andermatt)

2013-12-09 10:56  tlaino

	* [r13427] wrap and unwrap (for adaptive QMMM) - patch04

2013-12-09 08:28  tlaino

	* [r13426] adding the "ignore_outside_box" argument to few calls -
	  patch03

2013-12-07 12:55  vondele

	* [r13425] fix possible modification of look indices.

2013-12-06 20:41  vondele

	* [r13424] correct a couple of places that use single precision
	  instead of double precision

2013-12-06 18:34  tlaino

	* [r13423] patch 02: new QMMM centering syntax

2013-12-06 16:47  fschiff

	* [r13421] fix missing dp for 2 reals

2013-12-06 14:51  vondele

	* [r13420] take typical clock resolution into account

2013-12-06 13:25  tlaino

	* [r13418] patch01: do_force_mixing refactoring + bug fix
	  (restoring correct behaviour of QMMM walls - broken long time
	  ago)

2013-12-06 12:54  mandes

	* [r13417] TMC restart file corrected:
	  the frequency handling and error output is improved

2013-12-06 10:15  oschuett

	* [r13416] GLBOPT: Avoid division by zero

2013-12-06 07:08  vondele

	* [r13414] by default report resident memory

2013-12-06 06:59  tlaino

	* [r13413] wrong logic : assertion never executed

2013-12-05 16:11  tlaino

	* [r13411] remove INTENT(OUT) attribute for POINTERs - check
	  ASSOCIATED() -> fails

2013-12-05 14:49  vondele

	* [r13410] add a m_flush to the HFX output

2013-12-05 14:48  vondele

	* [r13409] prettify

2013-12-05 14:47  vondele

	* [r13408] prettify

2013-12-05 13:54  mandes

	* [r13406] TMC: add check for molecule information when molecule
	  move selected

2013-12-05 10:25  oschuett

	* [r13404] Allow calculation of dipoles with linear scaling code
	  (Samuel Andermatt)

2013-12-05 10:02  oschuett

	* [r13403] Input Description Change

2013-12-05 09:31  tlaino

	* [r13400] initialize pointers

2013-12-05 07:20  vondele

	* [r13399] add almo reference

2013-12-05 07:20  vondele

	* [r13398] prettify

2013-12-04 14:30  marci73

	* [r13396] deallocate

2013-12-04 11:04  tlaino

	* [r13395] do not start a write statement with a comma
	  do not use X descriptor without an integer specification
	  do not use "imaginary" abort() calls

2013-12-03 14:56  ibethune

	* [r13390] OpenMP for build_core_ppnl() (Fiona Reid, PRACE)

2013-12-03 10:39  marci73

	* [r13387] ZMP method: read external density

2013-12-03 09:13  marci73

	* [r13384] fix raman output

2013-12-03 08:06  marci73

	* [r13383] remove dbg comments

2013-12-03 07:54  marci73

	* [r13382] linear response : raman output (Sandra)

2013-12-03 07:48  marci73

	* [r13380] ZMP method, second part (Barborini, Varsano, Guidoni)

2013-12-02 19:24  vondele

	* [r13379] fix a corner case where roundoff would lead to a wrong
	  number of cells being employed, depending on compiler
	  optimization.

2013-12-02 12:43  mandes

	* [r13378] TMC: volume move selectable rescaling atomwise or
	  moleculewise (using
	  the center of mass)

2013-11-29 14:07  oschuett

	* [r13376] SWARM: fix leak

2013-11-29 13:28  vondele

	* [r13375] fix data race

2013-11-29 10:53  ibethune

	* [r13373] Fix format descriptors to keep ifort happy (Fiona Reid,
	  PRACE)

2013-11-28 20:54  mkrack

	* [r13371] Provide more info after allocation failure

2013-11-28 18:01  oschuett

	* [r13369] SWARM: fix leaks

2013-11-28 14:28  oschuett

	* [r13368] SWARM: Explicitly allocate lhs

2013-11-28 11:42  oschuett

	* [r13367] SWARM: Replace ALLOCATABLE arguments with POINTERs

2013-11-28 09:53  mkrack

	* [r13366] Avoid division by zero

2013-11-28 09:37  oschuett

	* [r13365] SWARM: Replace TRANSFER with own conversion routines

2013-11-27 15:55  oschuett

	* [r13363] Bibliography: Collect citations from all ranks

2013-11-27 15:18  oschuett

	* [r13362] GLBOPT: Add citation for Goedecker2004 also to input
	  section.

2013-11-27 15:11  oschuett

	* [r13361] GLBOPT: Add regtests

2013-11-27 13:30  oschuett

	* [r13360] GLBOPT: Add citation for Goedecker2004

2013-11-27 13:03  oschuett

	* [r13359] GLBOPT: Add keywords for controlling comparison of
	  minima.

2013-11-27 12:12  oschuett

	* [r13358] SWARM: fix format descriptors

2013-11-27 12:05  oschuett

	* [r13357] Set year in copyright banners to 2013

2013-11-27 11:45  oschuett

	* [r13356] SWARM: Convert some ALLOCATABLEs to POINTERs to make g95
	  happy

2013-11-26 18:02  oschuett

	* [r13355] New Feature: Parallel global geometry optimization based
	  on novel swarm-framework.

2013-11-26 17:23  vondele

	* [r13354] typo fix last commit

2013-11-26 17:21  vondele

	* [r13353] fix a data race on getting the matrix id

2013-11-26 17:20  vondele

	* [r13352] fix a data race, requires OMP 3.1

2013-11-26 17:07  vondele

	* [r13351] fix OMP data race on init

2013-11-26 16:25  oschuett

	* [r13349] sort OBJECTDEFS

2013-11-26 15:34  oschuett

	* [r13348] MD: Add generic external control mechanism

2013-11-26 15:08  oschuett

	* [r13347] TIMINGS: Allow printing of regions with 0% cpu-time

2013-11-26 14:59  oschuett

	* [r13346] GEO_OPT: Make reset of STEP_START_VAL optional

2013-11-26 14:50  oschuett

	* [r13345] BFGS: Add safeguard for choose_eigv_solver()

2013-11-25 12:26  mkrack

	* [r13343] Optionally be also permisse for the stress tensor
	  debugging

2013-11-23 05:56  mkrack

	* [r13342] Fix logic to avoid printout of misleading warnings

2013-11-22 20:43  mkrack

	* [r13341] Prettify

2013-11-22 16:40  marci73

	* [r13340] ZMP exchange-correlation potential (first part) :
	  Varsano, Barborini,
	  Guidoni

2013-11-22 16:36  marci73

	* [r13339] info control variable

2013-11-22 16:12  vondele

	* [r13337] add a QS respa testcase and some references.

2013-11-20 19:37  mkrack

	* [r13334] Revision of cp2k_debug

2013-11-20 18:09  mkrack

	* [r13333] Prettify

2013-11-20 15:13  oschuett

	* [r13332] Input Description Change (Iain Bethune)

2013-11-19 15:07  marci73

	* [r13326] ELPA is out, it has to be linked as external library
	  input keyword to choose the diagonalization library
	  syevd of scalapack is default, if the preferred libary is not
	  available, syevd is used

2013-11-14 17:38  vondele

	* [r13324] avoid a timeset in clear0 as long as it can be called on
	  an atomic block basis

2013-11-14 16:08  dorotheagolze

	* [r13323] Add citation for image charges

2013-11-14 09:04  mkrack

	* [r13322] Fix docu

2013-11-13 20:40  ibethune

	* [r13321] OpenMP for build_core_ppl. (Fiona Reid, Adam Carter,
	  PRACE)

2013-11-13 20:06  vondele

	* [r13320] turn warning into fatal on verify failures

2013-11-13 13:43  mkrack

	* [r13319] Fix typo

2013-11-11 13:53  oschuett

	* [r13316] DBCSR: Transfer only filled part of stack-buffers to GPU

2013-11-11 11:29  oschuett

	* [r13313] DBCSR: Fix flag __HAS_NO_CUDA_STREAM_PRIORITIES

2013-11-11 09:39  oschuett

	* [r13311] prettify

2013-11-11 09:19  oschuett

	* [r13310] Kim-Gordon: Decrease total number of colors via local
	  color swaps (Samuel Andermatt)

2013-11-10 20:08  vondele

	* [r13309] remove unnecessary fill guess

2013-11-10 16:22  vondele

	* [r13308] remove write-only variable

2013-11-10 16:15  vondele

	* [r13307] flush

2013-11-09 07:09  mkrack

	* [r13304] Workaround for g95

2013-11-08 20:26  vondele

	* [r13303] fix typo

2013-11-08 14:50  oschuett

	* [r13298] harmonize template with its instances

2013-11-08 12:19  oschuett

	* [r13294] DBCSR: Use cuda priority streams to improve overlapping
	  of MPI-communication with GPU-computation

2013-11-08 07:30  vondele

	* [r13292] one more timeset, replace a forall.

2013-11-07 17:11  oschuett

	* [r13291] Kim-Gordon: Speedup of dsatur coloring algorithm (Samuel
	  Andermatt)

2013-11-07 15:51  vondele

	* [r13290] add default_resize_factor for merging.

2013-11-06 14:35  tlaino

	* [r13288] allow different units for K (MORSE potential)

2013-11-04 11:04  marci73

	* [r13284] stress_tensor print key from medium to high print level

2013-11-03 15:56  vondele

	* [r13282] use msglen also in the serial case.

2013-11-02 18:32  vondele

	* [r13281] improved size guess for work matrix creation in the
	  threaded case

2013-11-02 14:38  vondele

	* [r13280] fix memleak

2013-11-02 11:50  vondele

	* [r13279] added timeset/stop

2013-11-02 10:38  vondele

	* [r13278] correctly compute reallocation with a growth factor

2013-11-02 07:58  vondele

	* [r13277] fix libtest, respect keepsparse input keyword

2013-11-01 16:57  oschuett

	* [r13276] DBCSR: Add some more timer-regions

2013-11-01 16:01  oschuett

	* [r13274] DBCSR: Add oversize_factor to memtype, centralize
	  resize_factor

2013-11-01 15:29  oschuett

	* [r13273] DBCSR: Simplify cp_dbcsr_print_config()

2013-11-01 14:44  oschuett

	* [r13272] prettify

2013-11-01 14:42  oschuett

	* [r13271] DBCSR: Added null-pointer checks to dbcsr_cuda_streams

2013-11-01 14:42  oschuett

	* [r13270] DBCSR: Tune size_guess, move GPU-transfer up (Joost)

2013-11-01 14:41  oschuett

	* [r13269] DBCSR: Add mempool for cannon working matrices
	  (conserves size-guess for next multiplication)

2013-11-01 14:39  oschuett

	* [r13268] DBCSR: Clear product_wm lazily

2013-11-01 14:38  oschuett

	* [r13267] DBCSR: Add smart default value for comm_thread_load

2013-11-01 14:37  oschuett

	* [r13266] DBCSR: Rename keyword n_stack_buffers to cuda/n_buffers

2013-11-01 14:35  oschuett

	* [r13263] DBCSR: Make multrec_limit an input keyword

2013-11-01 14:34  oschuett

	* [r13262] DBCSR: Name threads in cuda nvtx

2013-11-01 13:24  oschuett

	* [r13260] DBCSR: New cuda kernels based on transposed b-blocks.

2013-11-01 13:03  oschuett

	* [r13259] prettify

2013-11-01 06:53  marci73

	* [r13258] linear response as post scf not yet ready

2013-10-31 08:39  oschuett

	* [r13257] DBCSR: Add marketing-flops to statistics

2013-10-30 13:01  marci73

	* [r13255]
	  linear response : polarizability (Sandra Luber); some rewriting
	  of
	  linres modules; linear response calculations as post scf, rather
	  than as
	  run_type, need still some testing

2013-10-30 11:17  mandes

	* [r13254] TMC analysis: add missing varianbles in restart file
	  handling

2013-10-30 09:47  mandes

	* [r13253] Communication parameters for qs external comminication
	  of intermediate SCF energies are handled using the result
	  container.
	  Therefore, a testing routine for the results is implemented.
	  The procedure is used in TMC.

2013-10-29 16:06  mandes

	* [r13252] TMC memleak bug fix (using the correct error handler)

2013-10-29 14:16  vondele

	* [r13251] speedup of cp_dbcsr_multiply_local_d (Alfio)

2013-10-29 13:31  oschuett

	* [r13250] prettify

2013-10-28 16:07  vondele

	* [r13249] do not abort if not converged or on touch exit

2013-10-28 14:10  mandes

	* [r13248] TMC analysis remove the output file prefix underscore
	  for an empty prfix

2013-10-28 08:29  mandes

	* [r13247] TMC prettify the print move routines, avoiding memcopy
	  to same address

2013-10-26 13:21  vondele

	* [r13246] move allocation of matrix_w a little further down

2013-10-26 12:53  vondele

	* [r13245] almo: assert .NOT.calc_forces

2013-10-25 17:35  vondele

	* [r13244] remove hard-coded almo_max_av_neighbors (Rustam)

2013-10-25 13:58  mandes

	* [r13243] TMC analysis add a output file prefix, which can be
	  specified by the
	  user

2013-10-24 15:11  vondele

	* [r13241] add citation to Tran2013.

2013-10-24 14:40  marci73

	* [r13240] warnings related to right keywords to calculate and
	  print the stress tensor

2013-10-24 10:29  oschuett

	* [r13238] Use ACHAR(10) instead of NEW_LINE('A')

2013-10-24 10:15  mandes

	* [r13237] TMC: - the output of the number of requested Markov
	  chain elements is
	  corrected
	  - the analysis output file unit is corrected
	  - the missing headers in the master and analysis file are added
	  - the file description is added for the case of using (-o option)

2013-10-24 09:27  oschuett

	* [r13236] Input Description Change (Samuel Andermatt)

2013-10-24 09:24  oschuett

	* [r13235] Add newline-character and <pre> html-tag to
	  string_utilities.

2013-10-24 08:41  oschuett

	* [r13233] Remove length-limit on input-descriptions

2013-10-24 07:52  vondele

	* [r13232] add citation

2013-10-24 07:22  vondele

	* [r13231] add refs for energy decomposition

2013-10-23 14:44  mandes

	* [r13230] TMC: move type specification update
	  - now the user can decide which move types are
	  performed within the NMC

2013-10-22 16:16  vondele

	* [r13228] bugfix atprop energy

2013-10-22 16:05  tlaino

	* [r13227] do not mix different kinds.

2013-10-22 14:49  mandes

	* [r13226] TMC: atom swap move implemented
	  - the atom types to swap can be selected
	  in the input file

2013-10-22 10:29  oschuett

	* [r13225] typos

2013-10-22 10:11  tlaino

	* [r13224] allow the usage of absolute_position in the XYZ colvar

2013-10-21 08:19  oschuett

	* [r13223] typos (Samuel Andermatt)

2013-10-18 17:25  oschuett

	* [r13222] prettify

2013-10-18 13:14  mandes

	* [r13221] TMC changed molecule moves, now for variable molecule
	  sizes
	  - the proton reordering searches for molecules with 3 atoms to
	  create a
	  chain

2013-10-17 10:42  vondele

	* [r13219] Add ALMO methods (Rustam Khaliullin)

2013-10-16 16:45  oschuett

	* [r13218] Remove last implicit saves from eip_silicon.F

2013-10-16 16:28  oschuett

	* [r13217] Make eip_silicon.F IMPLICIT NONE

2013-10-16 13:04  vondele

	* [r13216] prettify

2013-10-16 12:23  mandes

	* [r13214] TMC file handling changed to open_file and close_file
	  provided by
	  cp_files

2013-10-16 12:17  oschuett

	* [r13213] Input Description Change (Nguyen Huu Chuong)

2013-10-16 10:32  oschuett

	* [r13212] Input Description Change

2013-10-15 14:01  mandes

	* [r13210] TMC move specification changes in the input file
	  now selecting each move with:
	  &MOVE_TYPE {THE_MOVE_TYPE}
	  SIZE {THE_MOVE_SIZE}
	  PROB {THE_PROBABILITY_TO_PERFORM_MOVE}
	  INIT_ACC_PROB {INITIAL_GUESS_FOR_ACCEPTING_MOVE}
	  &END

2013-10-15 10:12  oschuett

	* [r13209] prettify

2013-10-15 10:00  oschuett

	* [r13207] remove more implicit saves

2013-10-15 09:30  oschuett

	* [r13205] trim tmc filenames

2013-10-15 08:44  oschuett

	* [r13204] remove implicit saves

2013-10-14 14:54  mandes

	* [r13203] TMC update: added the TMC task type for an ideal gas,
	  completely neglecting energies
	  Ananlysis: adding the variances for average volumes and cell
	  sizes

2013-10-14 11:20  dorotheagolze

	* [r13202] improve comments for siepmann pot.

2013-10-11 12:46  oschuett

	* [r13201] harmonize template with instance

2013-10-11 12:36  oschuett

	* [r13200] instantiate templates

2013-10-11 12:28  oschuett

	* [r13199] Fix race-condition in prettify.py

2013-10-11 11:45  oschuett

	* [r13198] Bugfix of previous workaround

2013-10-10 20:18  ibethune

	* [r13197] Workaround for ifort 14.0 compiler bug - match
	  caller/callee array dimension info.

2013-10-07 16:15  tlaino

	* [r13196] fixed version of check

2013-10-07 15:38  jhutter

	* [r13195] SCF flag converged: consistent with inner/outer SCF

2013-10-07 15:30  jhutter

	* [r13194] Small cosmetic

2013-10-07 15:28  tlaino

	* [r13193] temporarily reverting to 13156 - remove the check -

2013-10-07 10:20  tlaino

	* [r13192] improve check

2013-10-07 08:39  oschuett

	* [r13191] Input Description Change

2013-10-04 10:30  ibethune

	* [r13190] Fix an out-of-bounds string access when building the
	  HTML manual

2013-10-02 10:47  vondele

	* [r13183] take only referenced data into account

2013-10-01 12:22  jhutter

	* [r13181] Remove SCP_DFT method: code and regtests

2013-09-30 06:55  vondele

	* [r13178] slightly improved flush parameter

2013-09-24 15:47  oschuett

	* [r13163] DBCSR: Allocate padding for new libsmm kernels

2013-09-24 11:36  oschuett

	* [r13162] DBCSR: Add some more timers

2013-09-24 11:34  oschuett

	* [r13161] DBCSR: Fix dbcsr_memtype_setup

2013-09-23 08:59  vondele

	* [r13158] make MAX_ELEMENTS_PER_BLOCK a config parameter

2013-09-23 06:08  tlaino

	* [r13157] add check

2013-09-23 06:08  tlaino

	* [r13156] prettify

2013-09-22 11:16  vondele

	* [r13155] use size specific stacks by default

2013-09-20 16:06  ibethune

	* [r13154] Allow for thread-safe execution of FFTs using Intel MKL
	  11.1+,
	  fail safe for older versions.

2013-09-19 09:11  jhutter

	* [r13153] Remove non-standard function call SIZEOF

2013-09-19 07:28  jhutter

	* [r13152] SCPTB: add some new functionalities

2013-09-18 12:39  tlaino

	* [r13151] bug fix

2013-09-18 08:28  vondele

	* [r13150] do not hard code kind types, do not convert to single
	  precision, do not cause integer overflow

2013-09-16 17:07  tlaino

	* [r13149] restore NAG compilation

2013-09-16 15:33  oschuett

	* [r13148] DBCSR: correct bit position in previous fix

2013-09-16 15:04  oschuett

	* [r13147] DBCSR: Fix int4 overflow due to sign-bit.

2013-09-16 14:25  tlaino

	* [r13146] bug fix for adaptive QMMM + allow distribution 1d to be
	  based on a previous distribution (Noam)

2013-09-16 14:10  tlaino

	* [r13145] pretty

2013-09-16 12:47  tlaino

	* [r13144] fix some exotic fortran and restoring NAG compilation

2013-09-13 08:49  vondele

	* [r13139] one more timing region.

2013-09-13 08:49  vondele

	* [r13138] prettify

2013-09-12 16:43  oschuett

	* [r13137] DBCSR: Add timer-region to check comm_thread_load

2013-09-12 14:18  oschuett

	* [r13136] Callgraph: Include power consumption

2013-09-12 08:59  vondele

	* [r13135] use kJ instead

2013-09-11 17:23  vondele

	* [r13134] add an energy report

2013-09-11 12:10  oschuett

	* [r13133] DBCSR: Remove misplaced reports

2013-09-10 14:27  oschuett

	* [r13131] DBCSR: Collect statistics from all MPI ranks

2013-09-10 13:44  oschuett

	* [r13130] DBCSR: Workaround for bug in ifort 12.1 (Noam Bernstein)

2013-09-10 11:11  mandes

	* [r13129] TMC bug fix in message interface.
	  - wrong position of parenthesis (not observed by gfortran
	  compiler)

2013-09-05 14:26  mkrack

	* [r13126] Stop MP2 only if MP2 is indeed active

2013-09-04 12:48  mandes

	* [r13125] TMC output added
	  - printing the average time for computing a Markov chain element

2013-09-04 09:38  delben

	* [r13124] Bug fixed when using non-local vdW potential with
	  specified CUTOFF in
	  combination with CELL_REF

2013-09-03 13:30  mandes

	* [r13123] TMC analysis classical dipole handling bug fix
	  - in the case of absent dipole file (using only classical
	  dipoles),
	  the pointer handling was incorrect

2013-09-02 17:16  mandes

	* [r13122] TMC analysis density calculation bug fix
	  - the subinterval volume was calculated wrong (not for cubic
	  cells)

2013-08-30 11:08  delben

	* [r13119] Deallocate RI basis at the end of optimization.

2013-08-29 15:40  delben

	* [r13115] RI-MP2 auxiliary basis optimization.

2013-08-26 09:11  vondele

	* [r13096] remove some uncalled routines

2013-08-26 08:24  oschuett

	* [r13095] Input Description Change (Ari Paavo Seitsonen)

2013-08-21 15:13  mkrack

	* [r13093] Prettify

2013-08-21 11:54  ibethune

	* [r13091] Reorder argument declaration to keep ifort happy (Fiona
	  Reid, PRACE)

2013-08-19 15:28  oschuett

	* [r13089] DBCSR: Make comm_thread_load-fix more userfriendly

2013-08-19 14:29  oschuett

	* [r13088] DBCSR: Fix comm_thread_load==0

2013-08-16 11:26  jhutter

	* [r13087] SCPTB: some updates on forces

2013-08-16 11:14  oschuett

	* [r13086] libsmm: Improve build process; Add vectorized interface
	  -D__HAS_smm_vec (Alfio Lazzaro).

2013-08-16 09:53  mkrack

	* [r13085] Suppress the trailing (redundant) DBCSR timing output

2013-08-15 11:57  fschiff

	* [r13084] complete refactoring and slight restructuring of qs_scf

2013-08-14 18:33  mkrack

	* [r13083] Add sup tag

2013-08-14 15:58  mkrack

	* [r13081] Enable the use of HTML tags and entities in the CP2K
	  input descriptions

2013-08-14 15:06  jhutter

	* [r13080] DFT-D2/3 improved information for missing referenz
	  functionals.

2013-08-13 17:05  oschuett

	* [r13079] Various micro-optimizations to speed-up md_output()

2013-08-12 12:14  oschuett

	* [r13078] Velocity softening: Add missing docu

2013-08-12 12:06  oschuett

	* [r13077] Add velocity softening used by Minima-Hopping

2013-08-12 11:50  oschuett

	* [r13076] Restructure velocity initialization

2013-08-12 09:28  oschuett

	* [r13075] Input Description Change (Ralph Koitz)

2013-08-12 09:23  mkrack

	* [r13074] Prettify

2013-08-09 13:32  ibethune

	* [r13072] Correctly overlap (threaded) local data movement with
	  MPI_Alltoall (Fiona Reid, PRACE)

2013-08-08 12:54  oschuett

	* [r13070] DBCSR: Change cuda default settings

2013-08-07 18:06  oschuett

	* [r13068] DBCSR: Clarify input description

2013-08-07 17:38  oschuett

	* [r13067] DBCSR: Improve binning of stacks (Joost)

2013-08-07 17:37  oschuett

	* [r13066] DBCSR: Add libcusmm with new cuda kernels (Peter
	  Messmer, Nikolay Markovskiy)

2013-08-07 12:52  oschuett

	* [r13065] DBCSR: Change default-settings for cpu/gpu balancing

2013-08-07 12:51  oschuett

	* [r13064] DBCSR: Make binning-parameters available as input

2013-08-07 12:50  oschuett

	* [r13063] DBCSR: sort statistics

2013-08-07 12:49  oschuett

	* [r13062] DBCSR: Restore cuda as default driver if available.

2013-08-06 17:14  oschuett

	* [r13061] Remove unused variable

2013-08-06 17:08  oschuett

	* [r13060] DBCSR: Add better control over CPU/GPU load-balancing

2013-08-06 11:54  jhutter

	* [r13059] SCPTB: some bug fixes

2013-08-05 15:23  vondele

	* [r13058] some more frequent timings

2013-08-05 14:39  oschuett

	* [r13057] Fix: init and config of DBCSR

2013-08-05 14:06  vondele

	* [r13056] remove timings for some commonly called subroutines

2013-08-04 21:39  mkrack

	* [r13055] Prettify

2013-08-02 09:51  jhutter

	* [r13054] SCPTB updates: enforce MIC and SPME stress correction

2013-07-31 14:42  mandes

	* [r13053] TMC update:
	  - the cell size is communicated to all workers, hence we avoid
	  wrong
	  cell sized in approx input file

2013-07-31 12:35  fschiff

	* [r13052] fix uninitialized variable

2013-07-30 13:34  vondele

	* [r13051] fix potential overflow

2013-07-30 08:37  jhutter

	* [r13050] Bug fix scZORA(MP): access of not initialized data (Ole)

2013-07-29 20:09  oschuett

	* [r13049] Convert dbcsr_type%initialized into LOGICAL

2013-07-29 08:20  vondele

	* [r13048] add reference

2013-07-26 18:01  vondele

	* [r13047] revert revert ?

2013-07-26 17:43  vondele

	* [r13046] revert commit for the timing being, as it exposes a
	  pre-existing tricky bug.

2013-07-26 04:58  vondele

	* [r13044] more g95 workarounds

2013-07-26 04:41  vondele

	* [r13043] make explicit a default save

2013-07-25 19:57  vondele

	* [r13042] another g95 workaround

2013-07-25 14:05  oschuett

	* [r13041] Fix g95 compilation.

2013-07-25 13:49  oschuett

	* [r13040] Improve DBCSR cuda performance through asynchronicity,
	  new kernels by Peter Messmer.

2013-07-24 11:55  jhutter

	* [r13039] Relativistic corrections for atomic blocks:
	  Methods implemented: DKH(2), DKH(3), ZORA(MP), scZORA(MP)
	  6 new regtests in tests/QS/regtest-rel

2013-07-19 14:10  mandes

	* [r13038] TMC analysis file opening bug fix

2013-07-17 13:06  tlaino

	* [r13036] allow the printing of QMMM dipole

2013-07-16 16:27  jhutter

	* [r13034] Atom code: implemented scaled ZORA with model potential

2013-07-15 16:26  mandes

	* [r13033] TMC PT bug fix:
	  - checked too much for existing pointers

2013-07-15 11:20  oschuett

	* [r13032] Input Description Change (Andreas Gloess)

2013-07-15 07:56  jhutter

	* [r13031] SCPTB: Many bug fixes, still far from full
	  functionality.

2013-07-10 13:02  mkrack

	* [r13030] sort

2013-07-10 12:06  mkrack

	* [r13029] sort

2013-07-10 11:46  mkrack

	* [r13028] Remove empty file; sort

2013-07-10 11:19  mkrack

	* [r13027] sort

2013-07-09 11:58  mandes

	* [r13026] bug fix:
	  correct the missed line break

2013-07-09 09:01  mandes

	* [r13025] correct the TMC input manual description

2013-07-09 07:04  mandes

	* [r13024] TMC: file open corrected
	  and some output corrections

2013-07-05 11:34  mkrack

	* [r13023] Prettify

2013-07-04 15:24  jhutter

	* [r13022] Small change in Copyright text

2013-07-04 12:33  jhutter

	* [r13021] Bug fix for vdW non-local functionals: reduced
	  cutoff/large number of
	  CPU case, disable block distribution (not allowed for subgrids)

2013-07-04 08:53  ibethune

	* [r13020] Take extra care to ensure orthorhombic cells are
	  strictly orthorhombic on platforms with non-standard
	  floating-point arithmetic

2013-07-02 12:13  vondele

	* [r13019] fix an undefined pointer (Vittorio)

2013-07-02 08:07  mandes

	* [r13018] TMC bug fix
	  - file opening corrected
	  - REAL type usage corrected

2013-07-01 18:25  mkrack

	* [r13017] Fix (outdated) timer routine call

2013-07-01 13:19  marci73

	* [r13016] get rid of the leftover of a removed section

2013-06-29 10:15  mkrack

	* [r13015] Fix memory leaks (workaround); prettify

2013-06-27 14:09  ibethune

	* [r13014] More robust occupation calculation to handle the case
	  where real(a)/real(a) != 1 on some architectures (e.g. Intel MIC)

2013-06-27 13:15  mandes

	* [r13013] TMC bug fix
	  - correct deallocation problems (memory leaks)

2013-06-27 09:27  ibethune

	* [r13012] Minor bugfixes (format string & typo)

2013-06-27 09:16  mandes

	* [r13011] TMC bug fix
	  - starting global tree element reference corrected

2013-06-27 06:05  mandes

	* [r13010] TMC bug fix
	  - correct the tmc atom deallocation

2013-06-26 21:53  mandes

	* [r13009] TMC deallocation bug fix

2013-06-26 19:21  mandes

	* [r13008] TMC corrected deallocation
	  (reduce memory leaks)

2013-06-26 13:19  ibethune

	* [r13007] Remove a duplicate USE (module is included via
	  cp_common_uses.h)

2013-06-26 12:29  mandes

	* [r13005] Tree Monte Carlo (TMC) integration in the trunk
	  with source and regtest files

2013-06-26 09:47  mkrack

	* [r13003] Further OpenMP tuning of the Fist nonbonded interaction
	  loop

2013-06-25 14:10  mkrack

	* [r13002] Prettify

2013-06-25 13:53  mkrack

	* [r13001] Tune OpenMP scalability by using atomic instead of
	  critical

2013-06-25 11:30  jhutter

	* [r13000] Change of license

2013-06-25 07:02  mmcgrathcp2k

	* [r12998] Updating MC input, part 1

2013-06-24 13:49  mkrack

	* [r12996] OpenMP parallelisation of Fist nonbonded interactions;
	  reset various regtests due to numerical noise

2013-06-20 10:04  ibethune

	* [r12995] Remove a superfluous variable in 3D FFT

2013-06-19 12:27  mkrack

	* [r12993] Separate elpa more cleanly

2013-06-19 11:50  mkrack

	* [r12992] Prettify; extent release doc

2013-06-18 15:40  vondele

	* [r12989] a first go at OMP-ing cp_dbcsr_multiply_local_d, needs
	  improvements

2013-06-18 11:54  vondele

	* [r12986] OMP parallel dbcsr_filter

2013-06-18 06:50  vondele

	* [r12985] properly deallocate kg_env if not used. Initialize
	  converged flag for LBFGS

2013-06-14 22:13  oschuett

	* [r12984] Input Description Change (Matt Watkins)

2013-06-13 13:35  oschuett

	* [r12978] Input Description Change

2013-06-13 13:35  mkrack

	* [r12977] Increment CP2K development version number

2013-06-12 14:29  fschiff

	* [r12967] slight change in optimization parameters improving curvy
	  steps and bfgs

2013-06-12 10:11  vondele

	* [r12966] avoid potentially quadratic loop (Hossein).

2013-06-12 08:05  delben

	* [r12965] This modification avoid out of bound when using Meta GGA
	  functional with
	  GAPW.

2013-06-12 07:05  vondele

	* [r12963] use int_8 and precompute things (Hossein).

2013-06-11 17:21  vondele

	* [r12962] workaround for a bug.

2013-06-10 12:22  oschuett

	* [r12956] Input Description Change (Matthias Krack)

2013-06-07 12:59  vondele

	* [r12953] use proper return type for scnrm2 (Vittorio)

2013-06-06 13:32  fschiff

	* [r12952] adding approximate energy to optimization criterium

2013-06-06 12:43  fschiff

	* [r12951] more refactoring, adjusting RTP to a more unified call
	  structure

2013-06-06 09:22  mkrack

	* [r12950] Prettify; allow only for integer numbers of alpha and
	  beta electrons with smearing for fixed_mag_mom = -1

2013-06-06 08:17  fschiff

	* [r12949] refactoring of qs_energy_scf

2013-06-05 13:37  fschiff

	* [r12946] bugfixes and refactoring ofr basis optimization

2013-06-03 11:46  vondele

	* [r12944] add homo-lumo computation after curvy

2013-06-02 15:48  vondele

	* [r12943] fix format

2013-06-02 15:42  vondele

	* [r12942] refactor compute_homo_lumo out of the trs4 routine

2013-06-02 14:24  vondele

	* [r12941] further reordering to save mem

2013-06-02 13:33  vondele

	* [r12940] reorder some allocates to save memory

2013-06-02 10:15  vondele

	* [r12938] cleanup, move Mulliken charges to computed MO free
	  results for LS usage

2013-06-02 09:28  vondele

	* [r12937] fix format

2013-06-02 07:19  vondele

	* [r12936] some cleanup of lanczos usage

2013-06-01 17:49  mkrack

	* [r12935] Prettify

2013-06-01 16:08  vondele

	* [r12934] split a large file

2013-06-01 12:11  vondele

	* [r12933] adjust format

2013-06-01 12:10  vondele

	* [r12932] prettify

2013-06-01 09:53  vondele

	* [r12931] output checksum on write

2013-06-01 09:53  vondele

	* [r12930] prettify

2013-05-31 15:24  jhutter

	* [r12929] Add SCPTB option for non-SCF TB

2013-05-31 15:23  jhutter

	* [r12928] Prettify

2013-05-31 14:11  oschuett

	* [r12926] prettify of fft_tools.F

2013-05-31 07:51  marci73

	* [r12922] mtd: optional cutoff of hills tail and slow growth of
	  V_mtd

2013-05-29 13:32  vondele

	* [r12921] fix DOI

2013-05-29 12:40  oschuett

	* [r12920] Improved cuda acceleration of FFTs, renamed flag to
	  __PW_CUDA (Andreas Gloess).

2013-05-27 13:41  oschuett

	* [r12918] fixed implied save of failure variables

2013-05-27 13:24  vondele

	* [r12917] some more implied saves corrected

2013-05-27 12:53  vondele

	* [r12916] add a couple of implicit none

2013-05-26 18:04  vondele

	* [r12914] fix a few incorrect uses of implied save. Either
	  correctly use the parameter attribute, explicit save attribute,
	  or initialize at the beginning of the procedure

2013-05-24 22:36  mkrack

	* [r12912] Prettify; use default_string_length, if appropriate; 1
	  regtest changed as workaround for Intel compiler

2013-05-24 13:23  vondele

	* [r12910] more fixes for large files

2013-05-24 06:38  tlaino

	* [r12908] bug fixes (adaptive qmmm) (Noam)

2013-05-24 06:26  vondele

	* [r12907] fix some problems with binary reads of large files

2013-05-23 19:03  vondele

	* [r12906] fix memory leaks on binary read (Hossein)

2013-05-23 09:33  mkrack

	* [r12905] Do not rely on the * format (fix for the Intel compiler)

2013-05-23 08:12  mkrack

	* [r12903] Prettify

2013-05-23 06:07  vondele

	* [r12902] assorted fixes for binary_io (Hossein)

2013-05-21 20:26  vondele

	* [r12901] check for cube file format limit

2013-05-21 17:31  vondele

	* [r12900] avoid a case of implicit save

2013-05-21 14:49  oschuett

	* [r12898] Bugfix in generation of new f_env_id

2013-05-21 14:46  vondele

	* [r12897] don't rely on Fortran2003 reallocation on assignment

2013-05-21 12:26  fschiff

	* [r12896] bug fix for long BASIS_FILE names

2013-05-21 10:35  vondele

	* [r12895] optionally read/write density matrix in LS calculations

2013-05-19 14:34  vondele

	* [r12894] revert r5896 (2007), which was an incorrect
	  spaghetti-transformation (Vittorio).

2013-05-18 07:18  vondele

	* [r12893] Fix loss of precision, CMPLX yields a single precision
	  result unless specified explicitly (Vittorio)

2013-05-17 12:58  vondele

	* [r12892] fix quadratic behavior

2013-05-17 11:53  vondele

	* [r12891] remove stray write

2013-05-17 11:12  vondele

	* [r12890] some kg coloring improvements

2013-05-15 19:02  vondele

	* [r12889] bug fix: release should nullify the pointer (Vittorio)

2013-05-09 09:47  ibethune

	* [r12888] Fix build break (for once, Intel is more permissive than
	  g95 & gfortran)

2013-05-09 09:28  ibethune

	* [r12887] Abort if Intel MKL's FFTW3 wrapper is used in threaded
	  builds as it is not (currently)
	  thread-safe.

2013-05-08 11:13  ibethune

	* [r12885] Add Linux memory reporting to intel machine file

2013-05-08 10:59  fschiff

	* [r12884] enable stress tensor for NLCC pseudos/ fix bug for admm
	  and nlcc

2013-05-07 15:10  oschuett

	* [r12882] re-enabled support for multiple cuda devices

2013-05-07 07:21  tlaino

	* [r12881] bug fix

2013-05-06 12:22  delben

	* [r12879] Reset auxiliary basis normalization, add a key for
	  calculating the
	  condition number of the (P|Q) matrix for the RI methods.

2013-05-03 09:41  tlaino

	* [r12878] remove (assumed-to-be) dbg statements

2013-05-03 09:40  tlaino

	* [r12877] bug fix (DOF)

2013-05-03 09:26  marci73

	* [r12876] cp2k-plumed interface (Ralph), it works with the version
	  1.3 of plumed
	  see in INSTALL how to compile them together
	  in plumed a couple of files have to be changedi and a pach script
	  has to
	  be added, the developers of plumed
	  have been informed and should soon provide the required
	  modifications in
	  their repository.
	  
	  We are considering to prepare the inteface also for the new
	  version of
	  plumed, which at the moment is still in a beta phase

2013-05-03 07:29  tlaino

	* [r12875] bug fix (NOF for adaptive QMMM)

2013-05-02 15:07  fschiff

	* [r12874] bug fix for set assignments

2013-05-02 14:54  jhutter

	* [r12873] Make multipole calculation for SCP-NDDO more robust for
	  different
	  parameter sets

2013-05-02 14:53  jhutter

	* [r12872] Bug fix for not initialized variable

2013-05-02 14:52  jhutter

	* [r12871] Prettify

2013-05-02 12:39  fschiff

	* [r12869] fix for file path and regtests

2013-05-02 09:29  fschiff

	* [r12868] 1st version of a general basis fitting method

2013-05-02 09:17  tlaino

	* [r12867] bug fix for K (amber FF) (Letif)

2013-05-02 07:13  tlaino

	* [r12866] bug fix for adaptive QM/MM

2013-05-01 09:30  ibethune

	* [r12865] Avoid numerical divergence of replicated calculation
	  (only affects psmp build) by doing calculation on rank 0 only and
	  broadcasting the result

2013-04-29 13:51  ibethune

	* [r12864] And fix another if statement which assumes short-circuit
	  evaluation

2013-04-26 15:25  ibethune

	* [r12862] Expand an IF statement to avoid segfault if the compiler
	  (Intel) chooses to evaluate the second operand even if the first
	  is false.

2013-04-26 14:25  jhutter

	* [r12861] Prettify and reset one test (vdW) for numerics.

2013-04-26 10:55  ibethune

	* [r12860] And some more uninit variables...

2013-04-26 10:31  ibethune

	* [r12859] Fix an uninitialised variable

2013-04-26 10:04  ibethune

	* [r12858] Fix dummy parameter intent

2013-04-26 08:52  jhutter

	* [r12857] Print out the orbital expansion coefficients in
	  the atom program (Toon Verstraelen)

2013-04-23 13:39  jhutter

	* [r12856] DFTB: bug fix for stress tensor in parallel runs

2013-04-23 13:38  jhutter

	* [r12855] Correct recent change, core potential radius for SCPTB

2013-04-23 11:29  jhutter

	* [r12854] Correct calculation of radius of interaction for
	  SCPTB core interaction

2013-04-22 14:14  tlaino

	* [r12852] bug fix P param am1

2013-04-22 12:05  oschuett

	* [r12851] Bugfix: cp_logger_create() ignored
	  template_logger%iter_info

2013-04-18 15:06  ibethune

	* [r12848] Bugfix: Ensure arrays are zeroed correctly when
	  expanding a zero-length array

2013-04-18 12:50  delben

	* [r12847] Improvements in communication RPA module
	  (create_integ_mat)

2013-04-18 08:59  vondele

	* [r12846] only use quadrupoles if computed (Vittorio)

2013-04-18 08:46  vondele

	* [r12845] can't pass a nullified pointer to scp_coulomb and
	  tb_coulomb

2013-04-18 08:39  vondele

	* [r12844] init variable to zero (Vittorio)

2013-04-17 10:45  vondele

	* [r12842] fix leak

2013-04-17 08:30  vondele

	* [r12839] fix restart file for pseudopotentials with special
	  features (e.g. NLCC).

2013-04-16 19:29  vondele

	* [r12838] protect m_flush calls for negative unit numbers (Ted
	  Packwood)

2013-04-12 04:06  mkrack

	* [r12836] Fix g-space mixing in the case of a cell change

2013-04-11 16:09  ibethune

	* [r12834] Avoid increment of uninitialised variables (which are
	  never used in any case)

2013-04-11 14:36  ibethune

	* [r12833] Change interface to hybrid_alltoall to avoid problems
	  when send/recv buf has length 0.

2013-04-11 14:13  jhutter

	* [r12832] SCPTB: New function for short range potential, new
	  parameters (Ralph)

2013-04-11 10:19  ibethune

	* [r12829] Move barrier out of subroutine and fix intents to ensure
	  threads never see deallocated array

2013-04-11 09:54  vondele

	* [r12828] Fix format (Vittorio)

2013-04-10 22:24  vondele

	* [r12827] format update

2013-04-10 10:53  ibethune

	* [r12826] Bugfix: move recursion termination check to avoid
	  passing out-of-bounds array sections

2013-04-10 06:53  delben

	* [r12825] Bug fixed in the creation of the process 2D grid in RPA
	  and Laplace-MP2

2013-04-09 13:31  delben

	* [r12824] Calculation of the exact exchange (EXX) contribution to
	  the total energy
	  after RI-dRPA

2013-04-03 13:36  marci73

	* [r12817] fix uninitialized variable

2013-04-03 12:01  jhutter

	* [r12816] Fix for bug report 5 (Ake Sandgren).
	  All temporary arrays allocated and used with same length.
	  GAPW one center terms.

2013-04-03 09:25  jhutter

	* [r12815] Remove pre-compiler flags __SGL and __FFTSGL. (Andreas
	  Gloess)

2013-03-30 19:20  mkrack

	* [r12809] Bug fix for UKS atomic calculation with CORE keyword;
	  allow to print more output for atomic basis fit runs

2013-03-29 17:23  vondele

	* [r12808] clean up an unused variable

2013-03-29 16:51  vondele

	* [r12807] fix logic, make it a hard failure

2013-03-29 16:21  vondele

	* [r12806] print condition number

2013-03-29 11:25  vondele

	* [r12802] bug fix: don't use uninit variables

2013-03-29 11:25  vondele

	* [r12801] prettify

2013-03-28 20:47  ibethune

	* [r12797] Bugfix: nullify pointers before use. Intel compiler
	  doesn't guarantee these are NULL, with
	  predictibly bad results...

2013-03-28 15:56  ibethune

	* [r12796] Fix compiler warnings about unitialised intent(out)
	  variables

2013-03-28 13:41  jhutter

	* [r12795] SCPTB: Full method energy calculation enabled

2013-03-27 09:53  ibethune

	* [r12794] Added missing RECURSIVE attributes to routines which
	  form part of recursion loops. Strictly required by Fortran
	  standard, gfortran is permissive but required by ifort.

2013-03-26 12:00  ibethune

	* [r12793] ... and remove one further reference to a now deleted
	  file

2013-03-26 11:51  ibethune

	* [r12792] Remove some legacy MKL FFT code

2013-03-25 15:33  marci73

	* [r12791] initialize logocal to false

2013-03-25 12:37  jhutter

	* [r12790] Nullify pointers before use. (Andreas Gloess)

2013-03-25 12:08  oschuett

	* [r12789] BFGS: Reset Hessian to unity when matrix diagonalization
	  failes (very rare case).

2013-03-25 12:05  oschuett

	* [r12788] Added optional info argument to choose_eigv_solver() for
	  external error handling.

2013-03-24 07:55  vondele

	* [r12786] dbcsr_lapack_larnv: use correct intent for iseed.

2013-03-21 16:07  marci73

	* [r12785] linres: orthogonalization of Ap

2013-03-21 14:58  marci73

	* [r12784] bug fix

2013-03-20 16:31  marci73

	* [r12779] LINRES: external control

2013-03-20 11:55  vondele

	* [r12778] prettify with prettify hack

2013-03-20 10:35  tlaino

	* [r12777] restoring non-prettified version (something wrong with
	  the prettifier)

2013-03-20 10:06  tlaino

	* [r12776] pretty

2013-03-20 10:00  tlaino

	* [r12775] bug fix + 1 new reg

2013-03-20 08:38  vondele

	* [r12774] improvements for complex matrices (Martin).

2013-03-19 09:49  fschiff

	* [r12773] fix for extrapolation with S preconditioner

2013-03-14 13:43  jhutter

	* [r12767] Updated reading from parameter file(SCPTB), separate
	  global parameters

2013-03-14 13:42  jhutter

	* [r12766] prettify

2013-03-14 10:20  fschiff

	* [r12763] bug fix for extrapolation and improve robustness for
	  curvy steps

2013-03-13 20:01  tlaino

	* [r12761] allow CV rmsd with 1 frame only

2013-03-13 12:58  vondele

	* [r12759] bug fix forces for large grids

2013-03-12 12:31  jhutter

	* [r12757] Print optional Charges and Dipole moment for DFTB

2013-03-09 15:29  vondele

	* [r12755] improve write format

2013-03-08 17:12  jhutter

	* [r12753] Bug fix for DFTB code. Solves a major problem in the
	  calculation of
	  overlap and Hcore matrices.

2013-03-06 18:41  mkrack

	* [r12750] Enable the use of HTML entity names in the documentation

2013-03-06 15:06  jhutter

	* [r12745] Build up framework for LRIPAW method

2013-03-06 15:00  jhutter

	* [r12744] Delete incorrect description of functional (F. Tran)

2013-03-06 11:55  delben

	* [r12742] Free the hfx containers in the mp2 module without
	  releasing the hfx
	  structure

2013-03-06 08:09  mkrack

	* [r12741] Fix doc

2013-03-05 14:55  mkrack

	* [r12740] Simple constraint for keeping a requested cell symmetry

2013-03-04 12:26  oschuett

	* [r12738] Fixed g95 compilation.

2013-03-04 11:24  vondele

	* [r12737] refactoring of the lower level of the DBCSR multiply
	  (Ole Schuett).

2013-02-27 08:47  vondele

	* [r12734] fix grid spacing for DOS (Jinwoong)

2013-02-26 16:11  ibethune

	* [r12732] replace MPI_COMPLEX16 (optional in MPI) with
	  MPI_DOUBLE_COMPLEX for
	  better compatibility

2013-02-25 17:22  vondele

	* [r12730] force_env_set_cell: bug fix, pass the right error for
	  mixed_force (Marc Van Houteghem and Ward Poelmans).

2013-02-25 13:28  vondele

	* [r12729] use int_8 for the output result

2013-02-23 11:10  vondele

	* [r12728] some enhancements for complex matrices

2013-02-22 20:01  mkrack

	* [r12727] Safeguard for empty input lists (may cause infinite loop
	  and no output);
	  List support for optional atomic position vector output;
	  Position vector output in scaled coordinates

2013-02-22 11:31  mkrack

	* [r12726] Prettify

2013-02-19 16:39  marci73

	* [r12722] bug fix restart localization

2013-02-18 12:36  vondele

	* [r12719] improved nlogn coloring scheme (Martin).

2013-02-18 12:31  vondele

	* [r12718] fix input description (Martin)

2013-02-15 16:46  marci73

	* [r12715] linres restart file

2013-02-13 16:08  marci73

	* [r12707] INPUT CHANGED: print keys related to localization are
	  moved to section LOCALIZE

2013-02-13 15:54  fschiff

	* [r12706] Ole Schuett: bug fix for ifort O2 level

2013-02-08 09:37  dorotheagolze

	* [r12695] Implementation of Siepmann-Sprik potential

2013-02-08 08:56  vondele

	* [r12694] rs_distribute_matrix: rename from distribute_matrix for
	  clarity

2013-02-06 16:36  jhutter

	* [r12687] Improved description (F. Tran)

2013-02-06 15:06  vondele

	* [r12685] remove unused variable / keyword MOLECULAR (Martin)

2013-02-06 14:37  vondele

	* [r12684] prettify

2013-02-06 14:36  vondele

	* [r12683] add flush

2013-02-05 16:44  jhutter

	* [r12680] VV10 Dispersion functional (Fabien Tran)

2013-02-04 17:31  mkrack

	* [r12678] Prettify

2013-01-29 16:12  vondele

	* [r12674] write energy specific density cubes (Cha Jinwoong)

2013-01-29 11:25  vondele

	* [r12672] don't raise a negative number to a real power, use
	  integer instead (Peter).

2013-01-25 08:26  vondele

	* [r12671] add an additional timing call for the packing of send
	  buffers in make_images.

2013-01-24 19:13  vondele

	* [r12670] some more OMP

2013-01-24 18:41  vondele

	* [r12669] some OMP useful for the serial code

2013-01-24 18:40  vondele

	* [r12668] some omp useful for the serial code

2013-01-24 15:16  mkrack

	* [r12667] Prettify

2013-01-23 12:37  jhutter

	* [r12665] Update to libxc 2.0.1 (Fabien Tran)

2013-01-23 09:01  ibethune

	* [r12664] OpenMP fixes for Intel compiler (Alfio Lazzaro)

2013-01-21 19:35  vondele

	* [r12663] instantiate all templates freshly

2013-01-21 19:25  vondele

	* [r12662] adjust all templates to 2013

2013-01-21 19:20  vondele

	* [r12661] removes unnecessary calls to mpi_comm_size (Ole)

2013-01-21 19:14  vondele

	* [r12660] cleanup: remove detailed timing infrastructure (Ole)

2013-01-21 19:12  vondele

	* [r12659] fix

2013-01-21 19:08  vondele

	* [r12658] prettify

2013-01-21 19:07  vondele

	* [r12657] fix to allow prettify

2013-01-21 18:51  vondele

	* [r12656] prettify

2013-01-21 09:22  jhutter

	* [r12655] Update of parametrization for SCPTB (Ralph Koitz)

2013-01-20 10:24  mkrack

	* [r12654] Fix syntax for continued character constants

2013-01-19 17:42  vondele

	* [r12653] allow for building without libxc present

2013-01-19 16:39  jhutter

	* [r12652] Update to libxc version 2.0, reset 2 regtest
	  (introduction of cutoff).
	  Version libxc 1.2 will no longer work!
	  You have to update the ARCH file to __LIBXC2.
	  Fabien Tran

2013-01-19 11:52  vondele

	* [r12651] bug fix (stop threads from launching kernels before data
	  is there)

2013-01-19 10:05  vondele

	* [r12650] OMP parallel checksum

2013-01-19 09:45  vondele

	* [r12649] remove unused routine

2013-01-19 06:20  vondele

	* [r12648] init devprop early and reuse

2013-01-18 20:27  vondele

	* [r12647] bug fix

2013-01-18 18:03  marci73

	* [r12646] numerical stability

2013-01-18 15:43  vondele

	* [r12645] prettify

2013-01-18 14:51  vondele

	* [r12644] refactoring by splitting of dbcsr_internal_operation.F
	  into dbcsr_mm_cannon.F, dbcsr_mm_multrec.F, dbcsr_mm_csr.F,
	  dbcsr_mm_driver.F, and dbcsr_mm_types.F (Ole Schuett).

2013-01-17 15:18  vondele

	* [r12641] small improvement to GPU stack sorting

2013-01-14 13:03  delben

	* [r12637] New matrix redistribution / multiplication scheme in RPA
	  and Laplace
	  energy calculation

2013-01-13 19:29  vondele

	* [r12635] sort stacks in the proper order for GPUs

2013-01-04 08:20  vondele

	* [r12634] if flop is present make it the average over all ranks.

2013-01-03 20:55  vondele

	* [r12633] untranspose_blocks: add OMP

2013-01-03 07:48  vondele

	* [r12632] prettify

2013-01-03 07:45  vondele

	* [r12631] fix one more data race

2013-01-01 23:26  marci73

	* [r12630] add output control for diis of band

2013-01-01 20:23  vondele

	* [r12629] fix data race

2013-01-01 20:04  vondele

	* [r12628] fix data race

2013-01-01 20:03  vondele

	* [r12627] fix data race

2013-01-01 20:03  vondele

	* [r12626] fix data race

2013-01-01 11:16  tlaino

	* [r12623] update to 2013

2012-12-27 16:49  vondele

	* [r12622] add some explicit timings in cuda_process_mm_stack

2012-12-27 16:45  vondele

	* [r12621] fix a more subtle data race

2012-12-27 09:52  vondele

	* [r12620] avoid data race and possibly speedup the code

2012-12-26 20:55  vondele

	* [r12618] more buggy omp code...default(none) needed ?

2012-12-26 20:45  vondele

	* [r12617] fix potential data race

2012-12-26 20:40  vondele

	* [r12616] bug fix: don't imply save on local variables, init
	  instead

2012-12-26 20:35  vondele

	* [r12615] fix some obviously buggy omp

2012-12-26 08:01  vondele

	* [r12614] bug fix sort, initialize, no implied save.

2012-12-25 20:59  vondele

	* [r12612] use atomic update to avoid a data race

2012-12-25 15:12  vondele

	* [r12611] fix freeing of stack_state_dev for threaded/cuda-enabled
	  dbcsr.

2012-12-25 15:09  vondele

	* [r12610] make timeset/timestop more thread safe... unclear what
	  happens if called from nested parallel regions.

2012-12-25 15:02  vondele

	* [r12609] prettify

2012-12-24 10:22  fschiff

	* [r12606] (Raimondas Gavelis) PBEsol implementation

2012-12-22 13:59  vondele

	* [r12605] check stat

2012-12-22 12:25  vondele

	* [r12604] detab

2012-12-22 12:09  vondele

	* [r12603] don't modify loop variables, prettify

2012-12-22 11:11  vondele

	* [r12602] take care not to modify do look vars

2012-12-21 21:20  mkrack

	* [r12601] Workaround for the g95 compiler

2012-12-21 20:18  mkrack

	* [r12597] Bug fix for run type debug; use correct atom index;
	  reset one regtest input

2012-12-21 20:03  mkrack

	* [r12596] Workaround for the g95 compiler which seems to need a
	  local declaration of the PROCEDUREs inspite of their global
	  declaration

2012-12-21 15:45  vondele

	* [r12595] get rid of timings_mp.F, allow for hooking the cp2k
	  timing routines in DBCSR automatically (Ole).

2012-12-21 15:22  vondele

	* [r12593] prettify, remove unused files

2012-12-21 15:18  vondele

	* [r12592] cleanups for kinds, and slight optimization for zero
	  sized sorts

2012-12-20 22:08  mkrack

	* [r12590] * Map core and shell contributions to the right atoms
	  * Print optionally an atomic decomposition of the energy

2012-12-20 20:48  vondele

	* [r12589] add citation

2012-12-20 09:31  ibethune

	* [r12588] Fix kind specification for IBM XL compiler

2012-12-19 20:58  vondele

	* [r12587] fix mem leak (Ole)

2012-12-19 19:09  vondele

	* [r12586] provide a mold to the Null to make things clear, also
	  for g95.

2012-12-19 16:12  vondele

	* [r12585] more of the same. Final ?

2012-12-19 16:01  vondele

	* [r12584] more of the same

2012-12-19 15:52  vondele

	* [r12583] make sure both integer kinds match

2012-12-19 14:41  vondele

	* [r12582] prettify

2012-12-19 14:00  vondele

	* [r12581] improved timing facilty. Can write a CALLGRAPH for use
	  with kcachegrind. Trace single (or a few) subroutines. Added
	  templated lists and dictionaries. (Ole).

2012-12-19 13:09  vondele

	* [r12580] get rid of hard-coded integer KINDS (Ole)

2012-12-19 12:48  delben

	* [r12579] Improvements in the communication scheme in
	  create_integ_mat routine.

2012-12-19 10:41  tlaino

	* [r12578] bug fix (Noam)

2012-12-18 16:48  mkrack

	* [r12576] Fix property atomic energy for core-shell models

2012-12-15 17:35  vondele

	* [r12574] use int_8 to prevent overflow

2012-12-14 22:18  vondele

	* [r12572] faster implementation of the distribution_1d (in the
	  limit of many molecule kinds and many mpi tasks). While the load
	  of all tasks is the same in the new and old implementation, the
	  local molecules might differ, hence some resets of tests

2012-12-14 21:56  vondele

	* [r12571] prettify

2012-12-14 08:29  vondele

	* [r12570] check a few more stat variables

2012-12-13 13:03  fschiff

	* [r12568] bug fix for nlcc LSD forces

2012-12-11 16:46  jhutter

	* [r12566] SCPTB: bug fixes and some new functionality

2012-12-11 10:32  vondele

	* [r12565] introduce a hack to improve PDGEMM performance
	  (-D__PILAENV_BLOCKSIZE=XXXX) if scalapack comes with an
	  non-optimized value. Further enhance the libtest for PDGEMM. With
	  these changes we can obtain peak performance for PDGEMM.

2012-12-10 14:57  mkrack

	* [r12564] CODATA 2010 constants

2012-12-10 11:54  vondele

	* [r12563] change to collective IO

2012-12-10 09:23  vondele

	* [r12562] don't modify a loop index by using it as an
	  intent(inout) subroutine argument

2012-12-07 21:48  mkrack

	* [r12561] Trajectory output in PDB format

2012-12-05 12:47  vondele

	* [r12559] Replace quicksort with mergesort to avoid quadratic
	  behavior in certain cases, e.g. merging of the indices of various
	  threads in dbcsr (Ole Schuett). Many resets needed due to
	  different order of the index result in case the values are equal.

2012-12-05 08:33  mkrack

	* [r12558] Prettify

2012-12-04 14:56  mkrack

	* [r12557] Fix typo

2012-12-04 14:20  jhutter

	* [r12556] Fix memory leak in SCPTB

2012-12-04 11:57  delben

	* [r12555] Improvement in the minimax module (avoid useless
	  allocation)

2012-12-04 11:00  jhutter

	* [r12553] Fix SCPTB input

2012-12-03 19:53  vondele

	* [r12552] use IMPLICIT NONE

2012-12-03 19:23  vondele

	* [r12551] set the default PRIVATE attribute of module variables
	  and procedures

2012-12-03 16:55  jhutter

	* [r12550] Fix parallel code for SCPTB

2012-12-03 10:18  mkrack

	* [r12549] Perform initialisation

2012-12-01 16:11  jhutter

	* [r12548] SCPTB: SCC code, no SCP yet, add regtests

2012-11-30 08:26  mkrack

	* [r12547] Fix uninitialised variables

2012-11-30 08:18  vondele

	* [r12546] add a timer

2012-11-29 16:56  mkrack

	* [r12545] Initialise INTENT(OUT) variables

2012-11-29 14:10  vondele

	* [r12544] binary IO of a DBCSR matrix (Hossein). Uses MPI IO
	  (parallel) or Fortran stream IO (serial).

2012-11-29 08:59  jhutter

	* [r12543] Fix initialization problem, some printing

2012-11-28 16:42  mkrack

	* [r12542] Fix memory leaks due to libint

2012-11-28 12:13  delben

	* [r12541] RI-Laplace-SOS-MP2 with Minimax approximation for the
	  decomposition of
	  the energy denominator

2012-11-28 11:56  vondele

	* [r12540] fix division by 0 in the tddfpt eigensolver, including a
	  change of convergence check.

2012-11-27 16:21  vondele

	* [r12539] some reordering for the computation of the deviation
	  matrix

2012-11-27 12:27  vondele

	* [r12537] add an option to use cudablas in RI-GPW-MP2
	  (__HAVE_CUBLAS)

2012-11-26 16:50  jhutter

	* [r12536] Implementation of nonlocal vdW-Functionals, 10 new
	  regtests
	  This is an adaptation of the implementation from Quantum-Espresso

2012-11-25 20:05  vondele

	* [r12534] cleanup of LS scf loop

2012-11-25 18:37  mkrack

	* [r12533] Fix undefined last energy in the trajectory file

2012-11-24 19:13  mkrack

	* [r12532] * Enable parsing of products of integer and floating
	  point numbers
	  * Allow for multiple products and fractions and their combination
	  * Check new regtest input for examples

2012-11-23 15:39  vondele

	* [r12529] generalize radius calculation

2012-11-23 09:21  fschiff

	* [r12528] adjust the atomic code nlcc to bigdft convention

2012-11-22 16:41  mkrack

	* [r12527] Delete forgotten debug print statement

2012-11-22 13:32  mkrack

	* [r12526] * Update of LBFGS version from 2.4 to 3.0
	  * Improve settings of regtest inputs

2012-11-21 10:17  fschiff

	* [r12524] scaling factor to make CP2K NLCC compatible with bigdft

2012-11-20 13:45  jhutter

	* [r12523] Safeguard for serial jobs and RESTART OFF: no printing
	  of files

2012-11-20 12:31  mkrack

	* [r12522] * Cell optimisation revised: redundant code and
	  dispensable logic removed
	  * Direct cell optimiser works now also for core-shell models
	  * LBFGS and CG_GOLD enabled for cell optimisation
	  * LBFGS seems to perform best for geometry and cell
	  optimisations, i.e. at least as good as BFGS, and it also works
	  for larger systems
	  * 28 new regtest inputs added

2012-11-19 20:43  vondele

	* [r12521] add a DOS calculation for LS (Cha Jinwoong)

2012-11-19 16:57  tlaino

	* [r12520] mom conservation (QMMM-adapt)

2012-11-16 12:46  mkrack

	* [r12519] Prettify

2012-11-16 12:21  fschiff

	* [r12518] finish the correct print key

2012-11-16 12:01  fschiff

	* [r12517] restore printing of VXC, bug fix for wfn mix (mo_coeff_b
	  problem)

2012-11-16 10:50  fschiff

	* [r12516] enable some result printing from ls_scf

2012-11-16 10:14  vondele

	* [r12515] some bibliography updates

2012-11-15 14:57  mkrack

	* [r12513] Prettify

2012-11-15 14:53  vondele

	* [r12512] fix some unused code

2012-11-15 13:38  fschiff

	* [r12511] nullify subsection

2012-11-15 13:02  fschiff

	* [r12509] linear scaling optimizer using curvy steps

2012-11-15 09:10  dorotheagolze

	* [r12508] Adding input option for RESP and bug fixes for compiling
	  with g95

2012-11-13 09:37  marci73

	* [r12507] interface to choose between elpa and syevd

2012-11-13 06:21  tlaino

	* [r12506] bug fix

2012-11-09 19:54  marci73

	* [r12505] restore implicit none and include mpif.h

2012-11-08 14:53  jhutter

	* [r12504] Make Elpa immune to pretifier

2012-11-08 14:17  delben

	* [r12503] RI-MP2 and RI-RPA Open-shell regtest, new blocking
	  communication scheme
	  in RI-MP2, output with performance of dgemm/pdgemm

2012-11-07 18:04  vondele

	* [r12502] cleanup

2012-11-07 18:04  vondele

	* [r12501] fix memory leak

2012-11-07 12:06  vondele

	* [r12499] prettify

2012-11-07 12:05  vondele

	* [r12498] add history based extrapolation for the initial guess in
	  LS.

2012-11-07 09:36  vondele

	* [r12497] use a more general check

2012-11-07 08:34  vondele

	* [r12496] add checks for trace and addition of matrices with
	  different symmetry

2012-11-06 16:24  marci73

	* [r12495] repeat keywords

2012-11-06 13:29  vondele

	* [r12494] remove old stuff

2012-11-02 19:11  vondele

	* [r12493] prettify

2012-11-02 15:20  pousa

	* [r12492] Bugfix: elpa library with omp

2012-11-02 12:12  pousa

	* [r12491] Adding library ELPA for the solution of the eigenvalue
	  problem in cp2k (Christiane & Marcella)

2012-11-01 19:23  vondele

	* [r12488] admm forces with purification none (Matt Watkins)

2012-11-01 15:19  jhutter

	* [r12487] Stubs for the calculation of magnetic dipole moments

2012-11-01 14:52  mkrack

	* [r12486] * Improve spline info
	  * Dump splines also in GROMACS table format

2012-11-01 12:11  delben

	* [r12485] Release auxiliary pw_p_type at the end of the RI
	  integral routine, auxiliary basis norm type changed to 1

2012-10-31 16:24  mkrack

	* [r12484] Prettify

2012-10-31 15:57  ibethune

	* [r12483] Link correct caller routine when user supplies
	  libgrid.a, also added more helpful comments in
	  file header

2012-10-31 13:40  delben

	* [r12482] Independent input key for the auxiliary basis used in
	  MP2 and RPA, regtest for RI-MP2 and RI-RPA

2012-10-31 09:01  pousa

	* [r12481] Affinity support for fft and pw on GPU devices

2012-10-31 08:39  delben

	* [r12480] Resolution of identity direct random phase approximation
	  (RI-dRPA) introduced,
	  Changes in input, introduction of section WF_CORRELATION in XC,
	  In mp2.F the call to cp_fm_power substituted with
	  cp_fm_cholesky_decompose,
	  Improvements in the computation and redistribution of the 2 and 3
	  center integrals in the RI routines

2012-10-30 16:25  mkrack

	* [r12479] * Enable MD runs using a core-shell model and a fixed
	  atom constraint

2012-10-30 09:42  mkrack

	* [r12478] * Enable fixed atom constraint for core-shell models
	  * Add timeset/timestop
	  * NYI for Langevin dynamics with core-shell models

2012-10-28 15:10  vondele

	* [r12476] refine bib parsing

2012-10-27 16:16  vondele

	* [r12475] more options for testing cp_fm_gemm, blacs envs can now
	  be created with custom grids

2012-10-27 15:58  vondele

	* [r12474] prettify

2012-10-27 15:07  vondele

	* [r12473] split matmul / dgemm testing, some output changes for
	  cp_fm_gemm

2012-10-27 13:00  vondele

	* [r12472] add a couple of timeset/timestop

2012-10-26 17:17  vondele

	* [r12471] rename mm_workshare to mm_thread_workshare

2012-10-26 13:05  fschiff

	* [r12470] bug fix for fixed atoms in BFGS parallel and tiny
	  improvement on model hessian

2012-10-25 13:05  pousa

	* [r12469] Printing HWTOPO details: reordering+affinity in the same
	  block

2012-10-23 14:34  vondele

	* [r12468] allow for dumping matrix_vxc (GGA and GPW only)

2012-10-23 14:16  pousa

	* [r12467] Bug fix: round-robin reordering needs HWLOC

2012-10-22 14:59  dorotheagolze

	* [r12465] Adding input and output option to RESP and image charges

2012-10-21 14:00  mkrack

	* [r12464] Add citations for core-shell model

2012-10-21 08:19  mkrack

	* [r12463] Remove executable attribute

2012-10-20 17:32  mkrack

	* [r12462] * Fix memory leaks
	  * change file mode bits

2012-10-20 12:13  mkrack

	* [r12461] * Print also SVN revision number with the -v flag
	  * Detab file

2012-10-20 09:46  mkrack

	* [r12459] Prettify

2012-10-20 08:02  vondele

	* [r12458] slightly generalize the string length

2012-10-20 07:42  mkrack

	* [r12456] Allow for longer description strings in the XML manual

2012-10-20 05:59  tlaino

	* [r12455] bug fix

2012-10-19 08:46  tlaino

	* [r12453] 2nd fix

2012-10-19 08:40  tlaino

	* [r12452] bug fix

2012-10-19 07:44  tlaino

	* [r12451] bug fixes (Noam)

2012-10-19 06:45  mkrack

	* [r12450] Remove tabs, prettify

2012-10-18 13:09  jhutter

	* [r12449] Lines too long for gfortran default settings

2012-10-18 11:11  pousa

	* [r12448] Unique Cray database access per cp2k run and new
	  decision for switching of cannon heuristic

2012-10-18 09:19  tlaino

	* [r12445] adaptive QM/MM and a bunch of other stuff (untested)
	  (Noam & Teo)

2012-10-17 14:45  jhutter

	* [r12444] LIBXC: Add Fortran interface files to the CP2K code (F.
	  Tran)

2012-10-17 14:43  jhutter

	* [r12443] SCPTB: minor changes

2012-10-16 18:38  vondele

	* [r12442] improve manual comments (Martin)

2012-10-15 13:44  ibethune

	* [r12440] Refactor of 3D FFT, better load balance for some cases,
	  and better heuristic forchoosing FFT scheme (Ruyman Reyes, PRACE)

2012-10-15 12:34  jhutter

	* [r12438] Redefinition of libxc interface to kinetic energy
	  functionals (F. Tran)

2012-10-15 10:10  vondele

	* [r12437] integrate libxc for KE functionals in KG (Martin). Some
	  libxc interface corrections still needed.

2012-10-15 09:10  vondele

	* [r12436] Define a truncation radius also for the shortrange
	  operator (Hossein)

2012-10-09 13:12  jhutter

	* [r12433] LIBXC: disable not fully tested functionals.
	  replace one regtest (F. Tran)

2012-10-09 13:05  mkrack

	* [r12432] Optionally, sort the CP2K time report by the routine
	  self time instead of the total time

2012-10-07 13:06  mkrack

	* [r12431] Use compacted integer list format also for the restart
	  file

2012-10-06 18:50  mkrack

	* [r12430] Use adaptive integer format

2012-10-06 16:57  mkrack

	* [r12429] Fix missing declaration in enclosing parallel

2012-10-06 06:47  vondele

	* [r12428] enable subcomm for dbcsr by default, improving
	  make_images scalability

2012-10-05 15:49  mkrack

	* [r12427] Extend format

2012-10-05 12:10  jhutter

	* [r12426] Stop if LIBXC requested in input but not available

2012-10-05 12:03  mkrack

	* [r12425] * Fix (BFGS) geometry optimisation for shell models
	  * 6 regtest inputs added

2012-10-05 10:05  fschiff

	* [r12424] put in safe guard to stabilize model Hessian based BFGS

2012-10-04 11:43  jhutter

	* [r12423] Update Install information for LIBXC

2012-10-04 10:56  ibethune

	* [r12422] Fix array overrun in print statement when number of
	  nshells/ncores .ne. natoms

2012-10-04 06:41  vondele

	* [r12421] add some more info to the libxc keywords

2012-10-03 19:57  mkrack

	* [r12420] Prettify

2012-10-03 13:59  jhutter

	* [r12419] Add biblography for LIBXC (Fabien Tran)

2012-10-03 12:00  jhutter

	* [r12418] Add new routine needed for LIBXC

2012-10-03 11:46  pousa

	* [r12417] Set grid size constraints for hilbert and peano curves

2012-10-03 11:45  jhutter

	* [r12416] Interface to LIBXC (Fabien Tran)
	  To use it you have to change the ARCH file:
	  1) DFLAGS = -D__LIBXC
	  2) FCFLAGS = -I$(LIBXC_DIR)/include
	  3) LIBS = -L$(LIBXC_DIR)/lib -lxc
	  There are 9 new regtests in ../tests/QS/regtest-libxc

2012-10-03 11:39  jhutter

	* [r12415] Small update for SCPTB

2012-10-03 11:23  jhutter

	* [r12414] Remove never used function

2012-10-01 14:09  jhutter

	* [r12413] Bug fix for TPSS (LSD), Fabien Tran
	  Add one new regtest

2012-09-28 14:07  pousa

	* [r12412] Command line option for Network topology - MPI mapping

2012-09-27 09:16  pousa

	* [r12409] Include default strategy for machine_arch input and fix
	  switch counting for Cannon

2012-09-26 15:14  jhutter

	* [r12408] Refactoring

2012-09-26 12:48  jhutter

	* [r12407] UPF PP format: bug fixes

2012-09-26 11:50  vondele

	* [r12406] bug fix for changing subsets and cleanup (Martin)

2012-09-25 09:29  jhutter

	* [r12405] Bug fixes in 2nd derivative LSD for B97 and LYP, i
	  reset 1 regtest (Fabien Tran)

2012-09-25 09:09  jhutter

	* [r12404] Prettify

2012-09-24 15:25  vondele

	* [r12403] fix memory leaks for empty graphs

2012-09-24 15:14  jhutter

	* [r12402] UPF converter: try to get units and multiplication
	  factors right

2012-09-24 14:55  jhutter

	* [r12401] Write pseudopotential in QuantumEspresso UPF format.
	  This includes pseudo atom density and wavefunctions.
	  This is an experimental version.

2012-09-24 12:36  pousa

	* [r12400] Bug fix: network topology graph bound (Joost &
	  Christiane)

2012-09-24 11:48  vondele

	* [r12398] init variables, fix out-of-bounds (Martin)

2012-09-24 11:02  mkrack

	* [r12397] Functional of Engel and Vosko, PRB 47, 13164 (1993)
	  added (contributed by Sebastiano Caravati)

2012-09-24 08:49  pousa

	* [r12396] Support for offline nodes case on the Cray Topology

2012-09-19 17:19  vondele

	* [r12394] introduces EPS_KG_ORB which determines the accuracy of
	  the coloring scheme. (Martin)

2012-09-19 12:20  ibethune

	* [r12393] Bugfix missing OpenMP variable in shared clause

2012-09-18 14:08  jhutter

	* [r12392] Update of the SCPTB method (non-periodic/non-polarizable
	  systems)

2012-09-18 13:18  jhutter

	* [r12391] Prettify

2012-09-18 13:13  jhutter

	* [r12390] Bug fix of ring puckering colvar forces, reset regtests

2012-09-18 06:22  mkrack

	* [r12389] Prettify

2012-09-17 13:57  pousa

	* [r12388] removing unnecessary tmp arrays and adding some output

2012-09-17 06:25  mkrack

	* [r12387] Prettify

2012-09-16 05:25  ibethune

	* [r12386] Remove old, untested FFT interfaces FFTESSL, FFTACML,
	  FFTMKL, FFTW2, FFTSCI.
	  FFTW3, FFTSG and FFTCU remain.

2012-09-14 14:54  pousa

	* [r12384] Fix hops for Torus network topology (Joost & Christiane)

2012-09-14 13:17  vondele

	* [r12383] use dedicated task lists for KG (Martin)

2012-09-14 12:09  ibethune

	* [r12382] Conditionally compile new code only if __FFTW3 is
	  defined

2012-09-14 10:12  pousa

	* [r12381] extending input options for RESP (dorothea)

2012-09-14 09:31  ibethune

	* [r12380] Handle MKL's incomplete FFTW interface cleanly by
	  falling back to a
	  safer variant (Ruyman Reyes, PRACE)

2012-09-13 10:33  ibethune

	* [r12378] Replace GLOBAL%FFTW_ARRAYS_ALIGNED keyword with a
	  preprocessor define __FFTW3_UNALIGNED, since
	  this property is platform dependent

2012-09-12 14:26  marci73

	* [r12377] g95 requires public types

2012-09-12 13:57  ibethune

	* [r12376] Bugfix fftw_destroy_plan()

2012-09-12 08:32  pousa

	* [r12375] Optimization network topology also for Seastar (Joost &
	  Christiane)

2012-09-12 08:28  pousa

	* [r12374] Optimizing the extraction of network topology (Joost &
	  Christiane)

2012-09-11 14:52  ibethune

	* [r12373] Fix use of constants and destroy_plan()

2012-09-11 14:38  mkrack

	* [r12372] Prettify

2012-09-11 12:46  marci73

	* [r12371] new CV : conditioned distance

2012-09-11 10:12  ibethune

	* [r12370] Fix destroy_plan() routine

2012-09-11 09:24  ibethune

	* [r12369] OpenMP-parallelized 3D FFT routines. (Ruyman Reyes,
	  PRACE).

2012-09-11 07:07  vondele

	* [r12368] fix for disconnected graphs

2012-09-11 07:04  vondele

	* [r12367] improve coloring output (Martin)

2012-09-06 12:11  jhutter

	* [r12363] SCPTB code, add xml parser, bug fixes, regtest

2012-09-06 11:48  delben

	* [r12362] Open shell RI-MP2-GPW

2012-09-06 07:39  pousa

	* [r12361] Adding debug option for HWTOPO

2012-09-04 15:58  delben

	* [r12359] Resolution of the identity (RI) MP2-GPW

2012-09-04 12:48  mkrack

	* [r12358] Increment version number of the current development
	  version to 2.4 after the release of the CP2K version 2.3

2012-09-04 10:03  mkrack

	* [r12356] Fix memory leak

2012-09-04 09:59  mkrack

	* [r12355] Prettify

2012-09-03 14:21  jhutter

	* [r12347] Ring puckering colvars (not fully tested), 10 regtests
	  added

2012-09-03 13:14  vondele

	* [r12346] a new implementation of KG, first round (Martin)

2012-08-31 13:27  pousa

	* [r12341] Bug fix: Binding for intel compiler

2012-08-29 16:23  vondele

	* [r12340] add flush

2012-08-28 11:53  mkrack

	* [r12338] * Prevent division by zero
	  * One regtest reset due to numerics

2012-08-27 18:55  vondele

	* [r12337] fix more valgrind warnings

2012-08-27 18:20  vondele

	* [r12336] do not write to a character variable being read

2012-08-27 16:24  vondele

	* [r12335] more uninit values

2012-08-27 15:31  vondele

	* [r12334] fix uninit values

2012-08-27 14:51  vondele

	* [r12333] fine tune TRS4 convergence criterium a bit

2012-08-27 08:01  pousa

	* [r12332] bugfix and adding new keyword (dorothea)

2012-08-24 16:01  vondele

	* [r12331] export dbcsr_maxabs

2012-08-24 09:45  vondele

	* [r12330] clarify doc (Martin)

2012-08-23 13:54  vondele

	* [r12329] remove executable property of files

2012-08-23 13:28  vondele

	* [r12328] cleanup calling qs_vxc_create (Martin)

2012-08-21 13:44  pousa

	* [r12327] optimization of RESP print option (dorothea)

2012-08-21 11:35  pousa

	* [r12325] Network topology support and automatic MPI rank
	  reordering strategies

2012-08-21 07:59  vondele

	* [r12323] further TRS4 refinements (Florian Thoele)

2012-08-21 07:57  vondele

	* [r12322] prettify

2012-08-21 07:56  vondele

	* [r12321] prettify

2012-08-20 14:12  vondele

	* [r12318] integrate_v_rspace argument cleanup (Martin)

2012-08-20 14:06  vondele

	* [r12317] typo fixes (Martin)

2012-08-20 12:37  ibethune

	* [r12316] ...and delete no longer needed files.

2012-08-20 12:01  pousa

	* [r12315] speed up periodic RESP (dorothea)

2012-08-20 11:32  vondele

	* [r12314] fix serial build

2012-08-20 10:55  ibethune

	* [r12313] Autotuning framework for collocate and integrate
	  kernels. Generates a library libgrid.a which can be included in
	  CP2K via the __HAS_LIBGRID macro. (Ruyman Reyes, PRACE)

2012-08-17 08:54  mkrack

	* [r12312] Revert to the previous open file flag

2012-08-16 19:47  mkrack

	* [r12311] Bug fix

2012-08-16 13:44  mkrack

	* [r12309] * Enable binary restart with core-shell potential from a
	  non-core-shell potential run and vice versa.
	  * A change of the ensemble is also possible at the same time.
	  * NOTE: The format of the binary restart file has changed and
	  thus all binary restart files generated with
	  earlier versions of the code cannot be read any longer.
	  * HINT: As a workaround a normal ASCII restart file has to be
	  written with the old CP2K version
	  (discard the keywords SPLIT_RESTART_FILE and
	  BINARY_RESTART_FILE_NAME from the input) and then it has to
	  be read by the new CP2K version with the keywords
	  SPLIT_RESTART_FILE and BINARY_RESTART_FILE_NAME activated
	  again to write a binary restart file in the new format.

2012-08-15 10:18  pousa

	* [r12307] implementation of periodic RESP fitting (dorothea)

2012-08-14 09:01  vondele

	* [r12306] Do not write to the do-loop variable (Vittorio)

2012-08-09 08:39  mkrack

	* [r12304] Bug fix for variable step interval (metadynamics)
	  Improve numerical stability of the regtest inputs

2012-08-08 10:03  tlaino

	* [r12303] bug fix

2012-08-08 09:12  tlaino

	* [r12302] fix units

2012-08-07 09:27  mkrack

	* [r12298] Use the internal CP2K (pseudo)random number generator
	  instead of the default one provided by the actual compiler

2012-08-04 13:36  tlaino

	* [r12297] add PBC to plane_point distance and fix the behavior for
	  lone_keyword
	  (if present is .true. - more logical)

2012-08-02 15:15  jhutter

	* [r12296] Test for FFT compatibility of all grid levels (Noam
	  Bernstein)

2012-07-31 12:48  jhutter

	* [r12295] More refactoring, GAPW basis

2012-07-27 14:38  vondele

	* [r12294] use lanczos estimate for the norm estimation in the sqrt
	  iteration

2012-07-27 12:58  vondele

	* [r12293] take care that we get the right result for the unit
	  matrix, and protect a bit against loss of orthonormality

2012-07-27 08:04  vondele

	* [r12292] small changes for lanczos: comments, conservatively
	  mp_min/mp_max results, use dp real

2012-07-26 11:02  vondele

	* [r12291] A dbcsr based lanczos for extremal eigenvalues (Florian
	  Thoele)

2012-07-26 08:57  vondele

	* [r12290] ADMM MP2

2012-07-25 19:15  vondele

	* [r12289] do not use HFX variables in MP2 if not needed.

2012-07-24 20:33  vondele

	* [r12288] refactoring basis_parameter create/release

2012-07-24 14:47  vondele

	* [r12287] prettify

2012-07-24 14:46  vondele

	* [r12286] added mp_gather_Xm

2012-07-23 08:37  vondele

	* [r12285] correction for the TRS4 algorithm (Florian Thoele)

2012-07-22 05:50  vondele

	* [r12283] stabile sqrt iter and implement higher order variants

2012-07-22 04:36  vondele

	* [r12282] prettify

2012-07-19 09:36  jhutter

	* [r12281] Refactoring: remove unused variables

2012-07-18 16:05  jhutter

	* [r12280] Refactoring: remove 3 center terms

2012-07-18 12:42  jhutter

	* [r12279] Remove loop

2012-07-18 11:40  jhutter

	* [r12278] More refactoring

2012-07-18 09:05  jhutter

	* [r12277] Some refactoring

2012-07-17 13:44  vondele

	* [r12276] flush output

2012-07-17 10:21  vondele

	* [r12275] Add TRS4 purification method (Florian Thoele).

2012-07-13 13:48  jhutter

	* [r12274] Add XC functional option to atomic density routine.
	  This is used in the XRD interface, not in the initial guess.

2012-07-13 08:26  marci73

	* [r12273] bug fix

2012-07-12 14:46  marci73

	* [r12272] bug fix

2012-07-05 14:17  jhutter

	* [r12271] Pretty

2012-07-05 14:15  jhutter

	* [r12270] Output of total density (including core) for XRD
	  interface.

2012-07-05 06:36  marci73

	* [r12269] bug fix

2012-07-05 06:35  marci73

	* [r12268] elf: shift

2012-06-29 11:14  mkrack

	* [r12266] Prettify

2012-06-29 07:48  marci73

	* [r12265] elf

2012-06-27 12:39  marci73

	* [r12263] image charge : change sign for external potential
	  (dorothea)

2012-06-27 06:33  mkrack

	* [r12262] Prettify

2012-06-26 15:06  marci73

	* [r12261] preparation for the XRD interface

2012-06-24 08:52  mkrack

	* [r12260] Fix I/O recursion

2012-06-21 16:23  marci73

	* [r12258] omp: bug fix

2012-06-21 14:59  mkrack

	* [r12257] Add missing kind specifier

2012-06-21 12:21  mkrack

	* [r12256] Prettify

2012-06-19 14:41  marci73

	* [r12254] electron localization function (ELF)

2012-06-18 08:05  marci73

	* [r12253] stm input section

2012-06-13 06:58  delben

	* [r12252] Typo corrected

2012-06-13 06:34  mkrack

	* [r12251] Prettify

2012-06-10 22:02  marci73

	* [r12250] localized DOS

2012-06-10 14:18  marci73

	* [r12249] bug fix

2012-06-10 14:18  marci73

	* [r12248] debug PRINT deleted

2012-06-08 13:24  delben

	* [r12247] Reduce memory usage in second quarter transformation
	  MP2_GPW

2012-06-08 13:13  delben

	* [r12246] Reduce memory usage in second quarter transformation
	  MP2_GPW

2012-06-05 12:46  delben

	* [r12245] Diagonalization of the KS matrix before the MP2 energy
	  calculation, open-shell MP2-GPW

2012-06-01 12:26  delben

	* [r12244] Improve estimation of memory usage in mp2_gpw (Mauro Del
	  Ben)

2012-05-31 15:01  vondele

	* [r12243] more compiler bug workarounds for zero-sized array
	  constructors

2012-05-31 06:50  mkrack

	* [r12242] Fix memory leak

2012-05-30 15:26  vondele

	* [r12241] spell check

2012-05-30 14:39  mkrack

	* [r12240] prettify

2012-05-30 13:48  vondele

	* [r12239] Bug fix in calling pw_env_create/rebuild leading to too
	  much memory being used.

2012-05-30 12:21  delben

	* [r12237] improve comment

2012-05-30 12:06  vondele

	* [r12236] Introduce double hybrids. Improved MP2 GPW for Coulomb
	  only MP2 methods. Reduced memory usage for the coefficient
	  matrix. (Mauro)

2012-05-29 19:30  vondele

	* [r12235] workaround for a compiler bug with zero-sized array
	  constructors

2012-05-29 18:53  vondele

	* [r12234] prettify

2012-05-29 14:18  fschiff

	* [r12233] leak fix, remove debug write

2012-05-29 10:25  fschiff

	* [r12232] new BFGS with model Hessian, improved convergence.
	  Changed BFGS defaults (trust radius, rfo optional) New cell
	  optimizer TYPE DIRECT_CELL_OPT, doing coordinates and cell at the
	  same type

2012-05-28 14:08  ibethune

	* [r12231] Auto-generate optimised subroutines for common cases of
	  xyz_to_vab(). >200% speedup observed using HFX_BASIS for NH3 bulk
	  system. (Ruyman Reyes, PRACE)

2012-05-18 10:22  jhutter

	* [r12229] Refactoring of GAPW 1-center Hartree energy

2012-05-10 14:57  pousa

	* [r12221] added restart option for image charges (Dorothea)

2012-05-10 06:57  vondele

	* [r12220] add a few more D2 scaling parameters

2012-05-03 07:24  vondele

	* [r12219] reduce output for failed cholesky, as this is happening
	  'regularly' in MD with the full_single_inverse preconditioner and
	  not a hard error.

2012-05-03 07:22  vondele

	* [r12218] prettify

2012-04-25 16:37  lwalewski

	* [r12217] PINT new feature: quench velocities for all dof

2012-04-24 06:57  mkrack

	* [r12216] Prettify

2012-04-20 15:59  lwalewski

	* [r12215] PINT feature: correlated Levy walk for bead pos init

2012-04-19 13:00  vondele

	* [r12214] improve memory estimate (Mauro and Joost)

2012-04-18 19:21  vondele

	* [r12213] more cleanups (Andreas)

2012-04-18 10:51  vondele

	* [r12212] Some clenaups (Andreas)

2012-04-17 07:57  uborstnik

	* [r12211] Bugfix: should not call MPI operations in serial mode.

2012-04-16 16:20  vondele

	* [r12209] some refactoring

2012-04-13 13:41  uborstnik

	* [r12208] Types that are used for public derived types must also
	  be public.

2012-04-12 07:52  vondele

	* [r12207] add a couple of intermediate timesets

2012-03-26 14:33  tlaino

	* [r12205] prettify

2012-03-26 14:29  tlaino

	* [r12204] bug fix

2012-03-26 09:02  tlaino

	* [r12203] Preliminary framework for adaptive QM/MM (Noam and Teo)
	  + 1 regtest

2012-03-26 07:25  vondele

	* [r12201] Stack sorting for GPU, improved 23x23 CUDA kernel (Neil
	  and Peter)

2012-03-23 11:17  mkrack

	* [r12200] Prettify

2012-03-22 15:28  marci73

	* [r12199] fix in lanczos diag

2012-03-21 15:43  marci73

	* [r12198] cholesky method inverse: bug fix

2012-03-21 15:36  mkrack

	* [r12197] Source cleaned from trailing spaces, tabs, and special
	  characters
	  OBJECT file lists sorted

2012-03-21 09:00  pousa

	* [r12196] Free GPU memory when compiled with cuda but not using it
	  on computation

2012-03-20 20:43  mkrack

	* [r12195] Prettify

2012-03-20 14:39  lwalewski

	* [r12194] PINT Nose thermostat bugfix reported by Florian
	  Schiffmann

2012-03-19 08:50  mkrack

	* [r12193] Prettify

2012-03-18 17:10  cjmundy

	* [r12192] enabling GA in the semiempirical code

2012-03-18 07:24  marci73

	* [r12191] memory leak fixed

2012-03-16 23:33  marci73

	* [r12190] some memory leaks solved

2012-03-16 15:13  marci73

	* [r12189] bug fix

2012-03-16 13:47  marci73

	* [r12188] bug fix

2012-03-16 11:29  marci73

	* [r12187] memory leak fixed

2012-03-16 08:22  mkrack

	* [r12186] Prettifier does not support any multi-line character
	  strings.
	  Fixed and prettify applied.

2012-03-16 07:58  mkrack

	* [r12185] Fix line continuations (gfortran seems to accept line
	  continuation within a string without a second &, but g95 does
	  not)

2012-03-15 15:17  marci73

	* [r12182] image charge (dorothea)

2012-03-15 15:16  marci73

	* [r12181] image charge

2012-03-15 15:06  uborstnik

	* [r12180] Stream syncs only when needed.

2012-03-15 12:47  uborstnik

	* [r12178] Adds CUDA stream synchronization and enables it before
	  stream-specific memory resizing.

2012-03-15 08:41  pousa

	* [r12176] New size guess for c_clocks (gpu side), reduce number of
	  c_locks resizing

2012-03-11 20:46  vondele

	* [r12175] fix 32bit overflow for large matrices

2012-03-09 10:51  vondele

	* [r12174] update m_memory routine so that resident and total
	  memory can both be
	  measured depending on compile time flags.

2012-03-07 18:07  vondele

	* [r12173] Fix checksums (Neil)

2012-03-07 14:48  pousa

	* [r12172] Parameter stack allocation on host is done only once at
	  dbcsr initialization (Andreas and Christiane)

2012-03-07 09:47  vondele

	* [r12171] fix work matrix size estimate for random matrix, in the
	  parallel case

2012-03-06 21:34  pousa

	* [r12170] Parameter stack allocation on host is done only once at
	  dbcsr initialization (Andreas and Christiane)

2012-03-06 13:24  jhutter

	* [r12169] Correct Intent attribute and write routine.

2012-03-05 11:26  pousa

	* [r12167] Fix warning on arch_lib for Cray compiler

2012-03-02 16:05  vondele

	* [r12164] flush after initial IO

2012-03-02 15:10  vondele

	* [r12163] further improvements to MP2_GPW (Mauro)

2012-03-01 15:01  vondele

	* [r12162] swap irecv/isend

2012-02-29 14:10  vondele

	* [r12161] some changes to allow for better performance
	  measurements (i.e. looping
	  will now lead to a profile that is fully multiplication
	  dominated)

2012-02-29 14:00  vondele

	* [r12160] prettify

2012-02-29 08:04  vondele

	* [r12159] allow for an async send also here

2012-02-28 14:05  pousa

	* [r12158] Avoid uninitialised variable for detailed timing

2012-02-28 11:23  pousa

	* [r12157] OpenMP issue with the intel compiler

2012-02-28 10:02  mkrack

	* [r12156] prettify

2012-02-24 12:57  uborstnik

	* [r12154] Bugfix: Out-of-bounds array access.

2012-02-23 08:18  pousa

	* [r12152] Less general block copy for make_images

2012-02-19 05:40  cjmundy

	* [r12151] Forgot these files. Regtest passed. Examples to come

2012-02-18 17:50  cjmundy

	* [r12150] First shot at CAFES dynamics a la Joost. Not yet
	  general. Needs restart, and better integration with other
	  thermostatting methods. Now only works with NOSE in the GLOBAL
	  and MOLECULE option. The latter is much better for sampling.
	  Would like to add MASSIVE but the mapping is a mess. I have
	  reproduced many of the literature results with this (e.g. ions in
	  external potential, RDFs for concentrated. Again, at this point
	  it doesn't interfere and regtests passed. I would like to make it
	  more clean. Comments welcome

2012-02-17 08:41  uborstnik

	* [r12146] Fixes OpenMP typo.

2012-02-16 19:06  vondele

	* [r12145] Fix an issue for large number of tasks and threads

2012-02-16 07:25  mkrack

	* [r12144] prettify

2012-02-15 19:30  cjmundy

	* [r12141] update of scp and pnnl parameters, some regtests will
	  change

2012-02-14 16:42  jhutter

	* [r12139] First implementation (not complete) of the SCP-TB method

2012-02-14 16:37  vondele

	* [r12138] use point-to-point communication in make_images, and
	  (thus) also subcommunicators by default.

2012-02-14 16:36  vondele

	* [r12137] a speedup for image_calculator

2012-02-10 15:48  vondele

	* [r12136] avoid triggering a record too long error message with
	  the cray compiler.
	  Our default_path_lenght just matches their default recl, which
	  results
	  in errors if we write all blanks.

2012-02-05 22:58  mkrack

	* [r12135] Support for cascade initialization

2012-02-03 10:13  vondele

	* [r12134] use the file name without the path in error messages....
	  avoids long lines and corresponding errors

2012-02-01 19:44  mkrack

	* [r12133] Allow the specification of the threshold value for
	  RELAX_MULTIP

2012-02-01 12:24  jhutter

	* [r12132] Remove unused variables

2012-01-30 16:56  mkrack

	* [r12131] Update for new website manual.cp2k.org

2012-01-26 14:41  mkrack

	* [r12121] Prettify

2012-01-26 13:56  pousa

	* [r12120] Arch_lib bug fix: libnuma support initialization crash.
	  Missing error status.

2012-01-26 12:46  marci73

	* [r12119] comm_split : variable key forgotten

2012-01-26 12:20  marci73

	* [r12118] elpa: only nmo eigenvalues

2012-01-26 10:25  marci73

	* [r12117] interface for ELPA library

2012-01-26 08:29  pousa

	* [r12116] Avoid use of hwloc_obj_snprintf to print core number.
	  Fix bugs in arch_lib/hwloc: print hostname for each MPI.

2012-01-24 13:26  jhutter

	* [r12115] =remove an unused array

2012-01-24 11:18  mkrack

	* [r12114] Ignore comment lines in the external coord section file

2012-01-22 14:45  vondele

	* [r12113] avoid overly long lines

2012-01-22 13:58  vondele

	* [r12112] Fix a bug for the molecular blocking case, introduced at
	  rev 11207, related to the fact that complete_redistribute does
	  not touch blocks that exist in target but not in source.

2012-01-20 10:03  mkrack

	* [r12111] Enable to switch the ensemble also when using a binary
	  restart file.
	  NOTE: old binary restart files cannot be used with this new
	  program version

2012-01-20 08:18  mkrack

	* [r12110] prettify

2012-01-19 11:12  lwalewski

	* [r12109] Centroid velocities initialization in PINT: rescaling to
	  the target temperature, unit support, I/O bugfix, regtest reset
	  (r222-251).

2012-01-04 13:38  pousa

	* [r12108] Bug fix on group mapping, correct the new ids. Add
	  machine name information - printing process and thread

2012-01-02 11:26  tlaino

	* [r12107] update 2012

2011-12-20 11:32  ibethune

	* [r12105] More efficient threading in distribute_matrix + avoid
	  extra copy of
	  local data

2011-12-19 08:10  mkrack

	* [r12104] Prettify

2011-12-16 13:42  jhutter

	* [r12103] Reformatting output

2011-12-16 11:58  jhutter

	* [r12102] Restrict topology info output to cases with at least one
	  instance

2011-12-15 08:22  pousa

	* [r12101] Avoid traversing the Hwloc tree to get shared caches

2011-12-14 12:22  mkrack

	* [r12095] Bug fix

2011-12-14 10:07  marci73

	* [r12094] use the scalapack eigensolver pdsyevr, when available,
	  i.e. with the new version of the package: Scalapack2.0

2011-12-13 10:54  mkrack

	* [r12091] * Update makefile to use SVN instead of the CVS
	  information
	  * The SVN source code revision number is now used instead of the
	  last CVS entry information
	  * The revision number is printed in the headers of the output and
	  restart files as reference
	  * Manual updated correspondingly
	  * Target mrproper added to the Makefile for a complete cleaning

2011-12-10 17:32  mkrack

	* [r12088] Optionally, allow to define valid CP2K units for
	  VARIABLES

2011-12-09 16:14  marci73

	* [r12086] dftd output to file

2011-12-08 15:02  uborstnik

	* [r12084] Bugfix: fixes attribute to work for OpenMP.

2011-12-07 15:58  uborstnik

	* [r12083] Bugfix: CUDA is not always used.

2011-12-07 15:47  uborstnik

	* [r12081] Adds CUDA streams and other improvements.
	  
	  Improves GPU load estimator.
	  Adds support for CUDA streams.
	  Better distinction between accelerator use and thread
	  worksharing.
	  Stack queues now have a queue for each defined size.
	  Two module-level options to control printing.
	  Updates timing to be more relevant.

2011-12-06 07:19  tlaino

	* [r12079] Automatic minor version update to 2.3.44 (r12079)

2011-12-05 04:18  vondele

	* [r12076] Automatic minor version update to 2.3.43

2011-12-04 04:18  vondele

	* [r12073] Automatic minor version update to 2.3.42

2011-12-03 14:28  marcella

	* [r12071] odd number of electrons + smearing does not need to be
	  spin polarized

2011-12-03 04:18  vondele

	* [r12068] Automatic minor version update to 2.3.41

2011-12-02 11:57  urban

	* [r12067] Machine arch: network card affinity (Christiane)

2011-12-02 04:18  vondele

	* [r12064] Automatic minor version update to 2.3.40

2011-12-01 04:18  vondele

	* [r12061] Automatic minor version update to 2.3.39

2011-11-30 09:45  jgh

	* [r12060] Replace intrinsic ERFC by cp2k internal

2011-11-30 09:29  urban

	* [r12059] DBCSR: Use CUDA with OpenMP. Some cleanups.

2011-11-30 04:18  vondele

	* [r12056] Automatic minor version update to 2.3.38

2011-11-29 04:18  vondele

	* [r12052] Automatic minor version update to 2.3.37

2011-11-28 10:03  jgh

	* [r12050] Atom code Slater basis: some bug fixes and 10 new
	  regtests

2011-11-28 04:18  vondele

	* [r12047] Automatic minor version update to 2.3.36

2011-11-27 04:18  vondele

	* [r12044] Automatic minor version update to 2.3.35

2011-11-26 04:18  vondele

	* [r12041] Automatic minor version update to 2.3.34

2011-11-25 08:46  tlaino

	* [r12040] bug fix

2011-11-25 04:18  vondele

	* [r12037] Automatic minor version update to 2.3.33

2011-11-24 16:34  jgh

	* [r12036] Atom code using Slater basis functions (testing pending)

2011-11-24 04:18  vondele

	* [r12033] Automatic minor version update to 2.3.32

2011-11-23 15:44  urban

	* [r12032] Bugfix: string length mismatch.

2011-11-23 13:49  urban

	* [r12030] Removes redundant ma_machine* and ma_message_passing*.

2011-11-23 12:51  urban

	* [r12029] Machine architecture updates (CPR).

2011-11-23 10:40  vondele

	* [r12028] fix memory leak (Martin)

2011-11-23 04:20  vondele

	* [r12024] Automatic minor version update to 2.3.31

2011-11-22 17:28  hforbert

	* [r12023] Bugfix (should fix a uninitialized variable access
	  reported by Joost)

2011-11-22 16:23  vondele

	* [r12022] add delta pulse to RTP (Martin Haeufel)

2011-11-22 04:18  vondele

	* [r12019] Automatic minor version update to 2.3.30

2011-11-21 20:34  urban

	* [r12018] Prevents use of uninitialized variable (UB/Alin Marin
	  Elena).

2011-11-21 15:26  vondele

	* [r12016] prettify

2011-11-21 15:22  vondele

	* [r12015] fix distribution of atoms (Urban)

2011-11-21 12:32  urban

	* [r12014] Bugfix.

2011-11-21 12:20  urban

	* [r12013] Removes unnecessary and undefined variables in
	  assertion.

2011-11-21 04:18  vondele

	* [r12009] Automatic minor version update to 2.3.29

2011-11-20 04:18  vondele

	* [r12006] Automatic minor version update to 2.3.28

2011-11-19 04:18  vondele

	* [r12003] Automatic minor version update to 2.3.27

2011-11-18 21:42  ibethune

	* [r12002] Bugfix uninitialised variable

2011-11-18 18:35  vondele

	* [r12001] needed for a CCE compiled binary

2011-11-18 14:52  urban

	* [r12000] Adds DBCSR tester and standalone testing and performance
	  programs.

2011-11-18 04:18  vondele

	* [r11997] Automatic minor version update to 2.3.26

2011-11-17 16:13  marcella

	* [r11995] improvements in davidson diagonalization

2011-11-17 04:18  vondele

	* [r11992] Automatic minor version update to 2.3.25

2011-11-16 04:18  vondele

	* [r11988] Automatic minor version update to 2.3.24

2011-11-15 16:18  hforbert

	* [r11987] fix small mistake in latest pimd commit

2011-11-15 15:19  urban

	* [r11986] Removes behind-the-scenes clustering and its
	  distribution.

2011-11-15 13:48  hforbert

	* [r11985] some cleaning up of the helium density code, one regtest
	  added

2011-11-15 04:18  vondele

	* [r11982] Automatic minor version update to 2.3.23

2011-11-14 20:38  vondele

	* [r11981] pass energy up

2011-11-14 20:02  vondele

	* [r11980] improve somewhat the output handling

2011-11-14 20:00  vondele

	* [r11979] make prettify happy

2011-11-14 14:16  vondele

	* [r11978] MP2 based on GPW (Mauro and Joost)

2011-11-14 13:37  vondele

	* [r11976] fix comment

2011-11-14 13:33  vondele

	* [r11975] prettify

2011-11-14 04:18  vondele

	* [r11971] Automatic minor version update to 2.3.22

2011-11-13 04:18  vondele

	* [r11968] Automatic minor version update to 2.3.21

2011-11-12 04:18  vondele

	* [r11965] Automatic minor version update to 2.3.20

2011-11-11 21:13  ibethune

	* [r11964] Error message bugfix.

2011-11-11 21:13  ibethune

	* [r11963] Special case finalize for make_images (first part)

2011-11-11 11:36  hforbert

	* [r11962] changed default of new section

2011-11-11 04:18  vondele

	* [r11959] Automatic minor version update to 2.3.19

2011-11-10 15:59  hforbert

	* [r11958] Added (helium) density calculations and various small
	  bugfixes.

2011-11-10 04:18  vondele

	* [r11955] Automatic minor version update to 2.3.18

2011-11-09 21:46  marcella

	* [r11954] bug fix

2011-11-09 04:18  vondele

	* [r11951] Automatic minor version update to 2.3.17

2011-11-08 04:18  vondele

	* [r11948] Automatic minor version update to 2.3.16

2011-11-07 04:18  vondele

	* [r11945] Automatic minor version update to 2.3.15

2011-11-06 19:42  vondele

	* [r11944] silence debug output

2011-11-06 17:44  vondele

	* [r11943] some preparatory cleanups/generalizations

2011-11-06 04:18  vondele

	* [r11940] Automatic minor version update to 2.3.14

2011-11-05 04:18  vondele

	* [r11937] Automatic minor version update to 2.3.13

2011-11-04 04:17  vondele

	* [r11934] Automatic minor version update to 2.3.12

2011-11-03 04:18  vondele

	* [r11931] Automatic minor version update to 2.3.11

2011-11-02 18:18  vondele

	* [r11930] remove a bit of write only variables

2011-11-02 15:36  urban

	* [r11929] Improves DBCSR thread distribution (CPR & UB).

2011-11-02 15:33  urban

	* [r11928] Moves non-DBCSR-specific routines from util to toollib.
	  Removes debugging
	  defines.

2011-11-02 15:10  vondele

	* [r11927] prettify

2011-11-02 14:51  vondele

	* [r11926] default skip_load_balance_distributed to true for runs
	  using more than 1024 MPI tasks

2011-11-02 04:18  vondele

	* [r11923] Automatic minor version update to 2.3.10

2011-11-01 04:17  vondele

	* [r11920] Automatic minor version update to 2.3.9

2011-10-31 15:48  ibethune

	* [r11919] Tree-based OpenMP reduction for rs_grid (Kevin
	  Stratford, PRACE)

2011-10-31 13:28  ibethune

	* [r11918] Fix memory leak

2011-10-31 11:50  ibethune

	* [r11917] More rs_grid refactoring - grids now stored in pw_env.
	  Small speedup by
	  avoiding frequent alloc/dealloc of grids

2011-10-31 04:17  vondele

	* [r11914] Automatic minor version update to 2.3.8

2011-10-30 04:17  vondele

	* [r11911] Automatic minor version update to 2.3.7

2011-10-29 03:17  vondele

	* [r11908] Automatic minor version update to 2.3.6

2011-10-28 03:18  vondele

	* [r11905] Automatic minor version update to 2.3.5

2011-10-27 03:17  vondele

	* [r11902] Automatic minor version update to 2.3.4

2011-10-26 14:53  vondele

	* [r11901] remove debug code

2011-10-26 09:46  ibethune

	* [r11900] Cache lgrids in pw_env_type, avoids huge memory access
	  cost in calculate_rho_elec (~15x faster on 24 threads)

2011-10-26 03:18  vondele

	* [r11897] Automatic minor version update to 2.3.3

2011-10-25 09:59  vondele

	* [r11896] Direct canonical MP2 energies (Mauro Del Ben)

2011-10-25 03:18  vondele

	* [r11893] Automatic minor version update to 2.3.2

2011-10-24 11:00  vondele

	* [r11892] mp_sum higher rank arrays

2011-10-24 03:17  vondele

	* [r11889] Automatic minor version update to 2.3.1

2011-10-23 14:35  vondele

	* [r11887] Change default REL_CUTOFF from 30Ry to 40Ry. Resets >500
	  regtests

2011-10-23 13:41  vondele

	* [r11886] test

2011-10-23 13:04  vondele

	* [r11883] trunk now becomes 2.3.0 / 2.2.X branched

2011-10-23 03:17  vondele

	* [r11878] Automatic minor version update to 2.2.382

2011-10-22 10:54  tlaino

	* [r11877] remove intent from pointer

2011-10-22 08:23  tlaino

	* [r11876] tiny fix + prettify

2011-10-22 03:18  vondele

	* [r11874] Automatic minor version update to 2.2.381

2011-10-21 15:23  ibethune

	* [r11873] Fix for gfortran with Intel-compiled MPI - do not pass
	  array constructors as arguments to MPI (PRACE)

2011-10-21 03:17  vondele

	* [r11871] Automatic minor version update to 2.2.380

2011-10-20 15:13  ibethune

	* [r11870] Removed unneccessary create/destroy of rs_grids in
	  generate_qs_task_list, also saves one extra levels of
	  pointer-chasing

2011-10-20 10:56  urban

	* [r11869] Removes another INTENT attribute of a pointer.

2011-10-20 10:16  urban

	* [r11868] Bugfix: using ISO_C_BINDING is not allowed if
	  __HAS_NO_ISO_C_BINDING is
	  defined.

2011-10-20 03:17  vondele

	* [r11866] Automatic minor version update to 2.2.379

2011-10-19 14:45  marcella

	* [r11865] bugfix [M Kilic]

2011-10-19 03:18  vondele

	* [r11863] Automatic minor version update to 2.2.378

2011-10-18 13:21  tlaino

	* [r11862] refactoring

2011-10-18 09:03  tlaino

	* [r11861] bug fix

2011-10-18 06:33  tlaino

	* [r11860] Temperature Accelerated Monte Carlo (Alin Marin Elena)

2011-10-18 03:18  vondele

	* [r11858] Automatic minor version update to 2.2.377

2011-10-17 06:57  tlaino

	* [r11857] removing dead code

2011-10-17 03:17  vondele

	* [r11852] Automatic minor version update to 2.2.376

2011-10-16 03:17  vondele

	* [r11850] Automatic minor version update to 2.2.375

2011-10-15 03:17  vondele

	* [r11848] Automatic minor version update to 2.2.374

2011-10-14 03:18  vondele

	* [r11845] Automatic minor version update to 2.2.373

2011-10-13 03:17  vondele

	* [r11843] Automatic minor version update to 2.2.372

2011-10-12 03:18  vondele

	* [r11840] Automatic minor version update to 2.2.371

2011-10-11 15:50  vondele

	* [r11839] declare also the subroutine arguments explicitly as
	  shared

2011-10-11 03:18  vondele

	* [r11837] Automatic minor version update to 2.2.370

2011-10-10 03:18  vondele

	* [r11835] Automatic minor version update to 2.2.369

2011-10-09 13:27  vondele

	* [r11834] speedup parsing of trajectory files

2011-10-09 03:18  vondele

	* [r11832] Automatic minor version update to 2.2.368

2011-10-08 03:18  vondele

	* [r11830] Automatic minor version update to 2.2.367

2011-10-07 18:49  vondele

	* [r11829] use MGS to retain the order of the basis functions.

2011-10-07 11:30  urban

	* [r11828] Removes spurious warning (CRP).

2011-10-07 03:18  vondele

	* [r11826] Automatic minor version update to 2.2.366

2011-10-06 08:02  urban

	* [r11825] This is a patch for the GPU support on libnuma branch of
	  the arch_lib library. Now, it is possible to have:
	  - MPI and Thread mapping with both HWLOC and Libnuma
	  - Machine description with both HWLOC and Libnuma
	  - GPU physical locality on the machine with both HWLOC and
	  Libnuma
	  (Christiane Ribeiro Pousa)

2011-10-06 03:17  vondele

	* [r11823] Automatic minor version update to 2.2.365

2011-10-05 03:17  vondele

	* [r11820] Automatic minor version update to 2.2.364

2011-10-04 03:18  vondele

	* [r11818] Automatic minor version update to 2.2.363

2011-10-03 19:01  vondele

	* [r11817] correct matrix multiply in output

2011-10-03 12:24  vondele

	* [r11816] use proper string lengths

2011-10-03 10:03  urban

	* [r11815] Adds CUDA/GPU detection to the machine architecture
	  library (Christiane Ribeiro Pousa).

2011-10-03 03:18  vondele

	* [r11813] Automatic minor version update to 2.2.362

2011-10-02 03:18  vondele

	* [r11811] Automatic minor version update to 2.2.361

2011-10-01 03:17  vondele

	* [r11809] Automatic minor version update to 2.2.360

2011-09-30 03:17  vondele

	* [r11807] Automatic minor version update to 2.2.359

2011-09-29 03:17  vondele

	* [r11804] Automatic minor version update to 2.2.358

2011-09-28 12:32  urban

	* [r11803] Rearranges declaration.

2011-09-28 11:36  urban

	* [r11802] Resolves building issues with machine architecture
	  library.

2011-09-28 03:17  vondele

	* [r11799] Automatic minor version update to 2.2.357

2011-09-27 03:18  vondele

	* [r11797] Automatic minor version update to 2.2.356

2011-09-26 19:50  vondele

	* [r11796] pass results down to f77

2011-09-26 19:46  vondele

	* [r11795] do not apply pbc in the non-periodic case, use the
	  correct label

2011-09-26 13:19  urban

	* [r11794] DBCSR/CUDA work.
	  Allows multiple GPUs per process; Fix for multiple stack
	  binning-simplifies input; CUDA optimizations; DBCSR configuration
	  output after setup;

2011-09-26 03:18  vondele

	* [r11792] Automatic minor version update to 2.2.355

2011-09-25 03:18  vondele

	* [r11790] Automatic minor version update to 2.2.354

2011-09-24 03:17  vondele

	* [r11788] Automatic minor version update to 2.2.353

2011-09-23 15:04  vondele

	* [r11787] use imag(log()) instead of atan to get full range of the
	  berry phase.

2011-09-23 03:17  vondele

	* [r11785] Automatic minor version update to 2.2.352

2011-09-22 11:39  vondele

	* [r11784] improve treatment of pbc in moment calculations

2011-09-22 07:26  urban

	* [r11783] Fixes OpenMP declaration error.

2011-09-22 03:17  vondele

	* [r11781] Automatic minor version update to 2.2.351

2011-09-21 13:01  urban

	* [r11780] Proper CUDA data size allocation in parallel; pretties.

2011-09-21 03:18  vondele

	* [r11778] Automatic minor version update to 2.2.350

2011-09-20 13:10  urban

	* [r11777] Fixes compilation options (Christiane Ribeiro Pousa).

2011-09-20 03:17  vondele

	* [r11775] Automatic minor version update to 2.2.349

2011-09-19 15:30  vondele

	* [r11774] check for MO_DIAG with force calculations

2011-09-19 13:15  urban

	* [r11772] Fixes CUDA memory leak and CUDA out-of-memory problem.

2011-09-19 03:17  vondele

	* [r11770] Automatic minor version update to 2.2.348

2011-09-18 03:17  vondele

	* [r11768] Automatic minor version update to 2.2.347

2011-09-17 17:42  vondele

	* [r11767] prettify

2011-09-17 08:31  vondele

	* [r11766] improve error handling for mistakes in the input in
	  parallel runs

2011-09-17 03:18  vondele

	* [r11764] Automatic minor version update to 2.2.346

2011-09-16 03:17  vondele

	* [r11762] Automatic minor version update to 2.2.345

2011-09-15 14:29  fschiff

	* [r11761] ADMM/RTP now possible. Works with purifcation NONE and
	  CAUCHY. EMD not yet implemented.

2011-09-15 03:18  vondele

	* [r11759] Automatic minor version update to 2.2.344

2011-09-14 03:17  vondele

	* [r11757] Automatic minor version update to 2.2.343

2011-09-13 19:23  vondele

	* [r11756] only check occ on filtering

2011-09-13 19:21  vondele

	* [r11755] add timeset

2011-09-13 03:18  vondele

	* [r11752] Automatic minor version update to 2.2.342

2011-09-12 03:17  vondele

	* [r11750] Automatic minor version update to 2.2.341

2011-09-11 03:17  vondele

	* [r11748] Automatic minor version update to 2.2.340

2011-09-10 03:17  vondele

	* [r11746] Automatic minor version update to 2.2.339

2011-09-09 03:17  vondele

	* [r11744] Automatic minor version update to 2.2.338

2011-09-08 14:14  urban

	* [r11741] Adds missing file?

2011-09-08 14:01  urban

	* [r11714] Adds support for the hwloc and libnuma libraries
	  (Christiane Ribeiro Pousa).
	  Supports thread to core pinning and setting memory policy on NUMA
	  architectures.

2011-09-08 03:18  vondele

	* [r11712] Automatic minor version update to 2.2.337

2011-09-07 03:17  vondele

	* [r11710] Automatic minor version update to 2.2.336

2011-09-06 03:17  vondele

	* [r11708] Automatic minor version update to 2.2.335

2011-09-05 03:18  vondele

	* [r11706] Automatic minor version update to 2.2.334

2011-09-04 03:17  vondele

	* [r11704] Automatic minor version update to 2.2.333

2011-09-03 03:17  vondele

	* [r11702] Automatic minor version update to 2.2.332

2011-09-02 10:16  jgh

	* [r11698] DFTB does not need task_lists

2011-09-02 03:18  vondele

	* [r11696] Automatic minor version update to 2.2.331

2011-09-01 03:17  vondele

	* [r11694] Automatic minor version update to 2.2.330

2011-08-31 13:57  urban

	* [r11693] Bugfix.

2011-08-31 03:17  vondele

	* [r11691] Automatic minor version update to 2.2.329

2011-08-30 14:16  urban

	* [r11689] Rewrite of the core of dbcsr_finalize.

2011-08-30 03:18  vondele

	* [r11687] Automatic minor version update to 2.2.328

2011-08-29 03:18  vondele

	* [r11685] Automatic minor version update to 2.2.327

2011-08-28 03:17  vondele

	* [r11683] Automatic minor version update to 2.2.326

2011-08-27 04:51  vondele

	* [r11682] one more data race fixed

2011-08-27 03:18  vondele

	* [r11680] Automatic minor version update to 2.2.325

2011-08-26 07:02  vondele

	* [r11679] more careful_mod

2011-08-26 06:01  vondele

	* [r11678] fix incorrect barrier (not all threads need to reach
	  this place)

2011-08-26 03:17  vondele

	* [r11676] Automatic minor version update to 2.2.324

2011-08-25 19:11  vondele

	* [r11675] fix some (trivial) data races

2011-08-25 08:38  ibethune

	* [r11674] Addition of OMP for full set of XC functionals and
	  cleanup of existing OMP (Kevin Stratford/PRACE)

2011-08-25 07:12  vondele

	* [r11673] fix more data races

2011-08-25 06:29  vondele

	* [r11672] fix data race

2011-08-25 03:18  vondele

	* [r11670] Automatic minor version update to 2.2.323

2011-08-24 17:59  vondele

	* [r11669] more data races

2011-08-24 13:10  vondele

	* [r11668] fix two more data races

2011-08-24 12:39  vondele

	* [r11667] fix an (innocent) data race

2011-08-24 11:04  vondele

	* [r11666] fix format

2011-08-24 10:17  vondele

	* [r11665] compiler workaround

2011-08-24 03:18  vondele

	* [r11663] Automatic minor version update to 2.2.322

2011-08-23 03:17  vondele

	* [r11661] Automatic minor version update to 2.2.321

2011-08-22 03:17  vondele

	* [r11659] Automatic minor version update to 2.2.320

2011-08-21 03:18  vondele

	* [r11657] Automatic minor version update to 2.2.319

2011-08-20 03:17  vondele

	* [r11655] Automatic minor version update to 2.2.318

2011-08-19 06:38  marcella

	* [r11647] block davidson diagonalization using mostly sparse
	  matrices

2011-08-19 03:17  vondele

	* [r11645] Automatic minor version update to 2.2.317

2011-08-18 14:31  jgh

	* [r11644] Correct typo in GAPW code (enables additional screening)

2011-08-18 03:17  vondele

	* [r11642] Automatic minor version update to 2.2.316

2011-08-15 07:24  vondele

	* [r11637] enable nlcc pseudos with GPW.

2011-08-15 07:07  vondele

	* [r11636] all arguments of MIN have the same kind

2011-08-15 03:18  vondele

	* [r11634] Automatic minor version update to 2.2.315

2011-08-14 03:18  vondele

	* [r11632] Automatic minor version update to 2.2.314

2011-08-13 03:17  vondele

	* [r11630] Automatic minor version update to 2.2.313

2011-08-12 03:18  vondele

	* [r11628] Automatic minor version update to 2.2.312

2011-08-11 14:53  vondele

	* [r11625] add a truncated version of the mix_cl operator

2011-08-11 03:17  vondele

	* [r11622] Automatic minor version update to 2.2.311

2011-08-10 10:24  vondele

	* [r11620] more output

2011-08-10 09:15  urban

	* [r11619] Cleanup: adds new DBCSR index entry for list indexing
	  (direct matrix coordinates). Reinstates support for non-local
	  indexing in the multiplication.

2011-08-10 08:29  ibethune

	* [r11618] And a couple more leaks.

2011-08-10 08:11  ibethune

	* [r11617] Fix memory leaks

2011-08-10 03:17  vondele

	* [r11615] Automatic minor version update to 2.2.310

2011-08-09 14:58  ibethune

	* [r11614] Optionally assign an OpenMP thread to progress
	  communications during multiplication
	  See USE_COMM_THREAD and COMM_THREAD_LOAD in the GLOBAL%DBCSR
	  section

2011-08-09 11:14  jgh

	* [r11611] Fix memory leaks

2011-08-09 08:13  vondele

	* [r11610] output atomic radial property

2011-08-09 07:37  vondele

	* [r11609] more output

2011-08-09 07:35  vondele

	* [r11608] prettify

2011-08-09 03:17  vondele

	* [r11598] Automatic minor version update to 2.2.309

2011-08-08 03:17  vondele

	* [r11596] Automatic minor version update to 2.2.308

2011-08-07 03:18  vondele

	* [r11594] Automatic minor version update to 2.2.307

2011-08-06 03:17  vondele

	* [r11592] Automatic minor version update to 2.2.306

2011-08-05 14:17  urban

	* [r11591] Bugfix for divide-by-zero. May occur when product/left
	  row or product/right column dimensions are not the same.

2011-08-05 13:21  jgh

	* [r11588] Atomic energy for HF, one new regtest

2011-08-05 08:01  jgh

	* [r11587] Prettify, reset further regtests

2011-08-05 03:18  vondele

	* [r11585] Automatic minor version update to 2.2.305

2011-08-04 20:50  marcella

	* [r11584] bug fix

2011-08-04 11:05  jgh

	* [r11581] Atomic properties: improvements and new regtests

2011-08-04 03:17  vondele

	* [r11579] Automatic minor version update to 2.2.304

2011-08-03 03:17  vondele

	* [r11577] Automatic minor version update to 2.2.303

2011-08-02 03:18  vondele

	* [r11575] Automatic minor version update to 2.2.302

2011-08-01 03:18  vondele

	* [r11573] Automatic minor version update to 2.2.301

2011-07-31 03:17  vondele

	* [r11571] Automatic minor version update to 2.2.300

2011-07-30 03:17  vondele

	* [r11569] Automatic minor version update to 2.2.299

2011-07-29 03:18  vondele

	* [r11567] Automatic minor version update to 2.2.298

2011-07-28 03:18  vondele

	* [r11560] Automatic minor version update to 2.2.297

2011-07-27 15:09  jgh

	* [r11559] Atomic properties: output for parallel runs adapted.

2011-07-27 11:04  jgh

	* [r11558] Atomic properties, more features and bug fixes.

2011-07-27 08:55  jgh

	* [r11555] New fiels for atomic properties

2011-07-27 08:36  jgh

	* [r11552] Atomic energy and stress, currently only for force
	  fields, 4 new regtests.
	  reset 47 tests due to numerics

2011-07-27 03:17  vondele

	* [r11550] Automatic minor version update to 2.2.296

2011-07-26 07:26  fschiff

	* [r11547] enable ADMM with RTP and EMD/ bug fix initialization
	  (initial energy behaviour)

2011-07-26 03:18  vondele

	* [r11545] Automatic minor version update to 2.2.295

2011-07-25 03:17  vondele

	* [r11543] Automatic minor version update to 2.2.294

2011-07-24 03:17  vondele

	* [r11541] Automatic minor version update to 2.2.293

2011-07-23 03:18  vondele

	* [r11539] Automatic minor version update to 2.2.292

2011-07-22 03:18  vondele

	* [r11537] Automatic minor version update to 2.2.291

2011-07-21 03:18  vondele

	* [r11535] Automatic minor version update to 2.2.290

2011-07-20 15:24  vondele

	* [r11534] bug fix nlcc der

2011-07-20 08:51  jgh

	* [r11533] Typo, initial guess NLCC pseudopotentials

2011-07-20 03:18  vondele

	* [r11531] Automatic minor version update to 2.2.289

2011-07-19 13:23  urban

	* [r11526] Introduces queues of stacks for better accelerator
	  support. Different can be filled with same-sized multiplications.

2011-07-19 03:18  vondele

	* [r11524] Automatic minor version update to 2.2.288

2011-07-18 15:11  vondele

	* [r11522] fix special case

2011-07-18 13:28  vondele

	* [r11519] Add an option to generate mm splines using a fixed
	  number of points.

2011-07-18 03:17  vondele

	* [r11517] Automatic minor version update to 2.2.287

2011-07-17 03:18  vondele

	* [r11515] Automatic minor version update to 2.2.286

2011-07-16 03:17  vondele

	* [r11513] Automatic minor version update to 2.2.285

2011-07-15 03:18  vondele

	* [r11511] Automatic minor version update to 2.2.284

2011-07-14 10:06  tlaino

	* [r11510] rm useless comment

2011-07-14 03:17  vondele

	* [r11508] Automatic minor version update to 2.2.283

2011-07-13 06:26  vondele

	* [r11507] compiler bug workaround

2011-07-13 03:17  vondele

	* [r11505] Automatic minor version update to 2.2.282

2011-07-12 09:32  vondele

	* [r11504] few more timesets/bug fix

2011-07-12 03:17  vondele

	* [r11502] Automatic minor version update to 2.2.281

2011-07-11 09:12  vondele

	* [r11501] allow for longer paths

2011-07-11 03:18  vondele

	* [r11499] Automatic minor version update to 2.2.280

2011-07-10 03:17  vondele

	* [r11497] Automatic minor version update to 2.2.279

2011-07-09 03:18  vondele

	* [r11495] Automatic minor version update to 2.2.278

2011-07-08 03:17  vondele

	* [r11493] Automatic minor version update to 2.2.277

2011-07-07 03:18  vondele

	* [r11491] Automatic minor version update to 2.2.276

2011-07-06 12:54  urban

	* [r11490] Check for duplicate blocks (to prevent infinite
	  recursion).

2011-07-06 03:18  vondele

	* [r11488] Automatic minor version update to 2.2.275

2011-07-05 03:18  vondele

	* [r11486] Automatic minor version update to 2.2.274

2011-07-04 18:34  vondele

	* [r11485] fix bugs in replicated code

2011-07-04 18:28  vondele

	* [r11484] fix error message length

2011-07-04 03:19  vondele

	* [r11481] Automatic minor version update to 2.2.273

2011-07-03 03:17  vondele

	* [r11479] Automatic minor version update to 2.2.272

2011-07-02 03:17  vondele

	* [r11477] Automatic minor version update to 2.2.271

2011-07-01 03:17  vondele

	* [r11475] Automatic minor version update to 2.2.270

2011-06-30 06:48  vondele

	* [r11474] move inits to a better place

2011-06-30 03:19  vondele

	* [r11472] Automatic minor version update to 2.2.269

2011-06-29 03:17  vondele

	* [r11470] Automatic minor version update to 2.2.268

2011-06-28 03:17  vondele

	* [r11468] Automatic minor version update to 2.2.267

2011-06-27 03:18  vondele

	* [r11466] Automatic minor version update to 2.2.266

2011-06-26 03:17  vondele

	* [r11464] Automatic minor version update to 2.2.265

2011-06-25 03:18  vondele

	* [r11462] Automatic minor version update to 2.2.264

2011-06-24 03:19  vondele

	* [r11460] Automatic minor version update to 2.2.263

2011-06-23 08:24  vondele

	* [r11459] looking for EXIT, add frame_count as a simple way to
	  limit the number of frames

2011-06-23 03:18  vondele

	* [r11457] Automatic minor version update to 2.2.262

2011-06-22 03:17  vondele

	* [r11455] Automatic minor version update to 2.2.261

2011-06-21 10:22  krack

	* [r11451] * Optionally, allow a reset of the U ramping value
	  * Reduce DFT+U output for print level low
	  * 2 regtest inputs added

2011-06-21 07:20  vondele

	* [r11450] workaround for a compiler bug

2011-06-21 03:18  vondele

	* [r11448] Automatic minor version update to 2.2.260

2011-06-20 03:17  vondele

	* [r11446] Automatic minor version update to 2.2.259

2011-06-19 03:18  vondele

	* [r11444] Automatic minor version update to 2.2.258

2011-06-18 03:58  mmcgrath

	* [r11443] Small bug fix that would cause a crash

2011-06-18 03:17  vondele

	* [r11441] Automatic minor version update to 2.2.257

2011-06-17 11:50  urban

	* [r11440] Simplified distribution creatino. More options and lower
	  memory requirement for DBCSR testing.

2011-06-17 10:43  urban

	* [r11439] Fixes row/column mismatch in dense distribution;
	  verifies distribution when creating it.

2011-06-17 09:24  krack

	* [r11438] prettify

2011-06-17 03:17  vondele

	* [r11436] Automatic minor version update to 2.2.256

2011-06-16 17:32  vondele

	* [r11433] fix another bug

2011-06-16 14:02  vondele

	* [r11431] fix wrong default

2011-06-16 09:27  mmcgrath

	* [r11430] Making it easier to do hybrid Monte Carlo by itself

2011-06-16 03:18  vondele

	* [r11428] Automatic minor version update to 2.2.255

2011-06-15 03:18  vondele

	* [r11426] Automatic minor version update to 2.2.254

2011-06-14 03:17  vondele

	* [r11424] Automatic minor version update to 2.2.253

2011-06-13 03:18  vondele

	* [r11422] Automatic minor version update to 2.2.252

2011-06-12 03:18  vondele

	* [r11420] Automatic minor version update to 2.2.251

2011-06-11 03:17  vondele

	* [r11418] Automatic minor version update to 2.2.250

2011-06-10 03:17  vondele

	* [r11416] Automatic minor version update to 2.2.249

2011-06-09 03:17  vondele

	* [r11414] Automatic minor version update to 2.2.248

2011-06-08 03:18  vondele

	* [r11412] Automatic minor version update to 2.2.247

2011-06-07 03:18  vondele

	* [r11410] Automatic minor version update to 2.2.246

2011-06-06 03:18  vondele

	* [r11408] Automatic minor version update to 2.2.245

2011-06-05 03:18  vondele

	* [r11406] Automatic minor version update to 2.2.244

2011-06-04 03:18  vondele

	* [r11404] Automatic minor version update to 2.2.243

2011-06-03 03:18  vondele

	* [r11402] Automatic minor version update to 2.2.242

2011-06-02 03:18  vondele

	* [r11400] Automatic minor version update to 2.2.241

2011-06-01 03:18  vondele

	* [r11398] Automatic minor version update to 2.2.240

2011-05-31 10:44  vondele

	* [r11397] fine-tune filtering

2011-05-31 03:18  vondele

	* [r11395] Automatic minor version update to 2.2.239

2011-05-30 16:42  urban

	* [r11394] Faster data copy when mutable (B-tree based) work
	  matrices are used.

2011-05-30 16:09  urban

	* [r11392] Left unprettified.

2011-05-30 03:18  vondele

	* [r11390] Automatic minor version update to 2.2.238

2011-05-29 03:17  vondele

	* [r11388] Automatic minor version update to 2.2.237

2011-05-28 03:18  vondele

	* [r11386] Automatic minor version update to 2.2.236

2011-05-27 03:17  vondele

	* [r11383] Automatic minor version update to 2.2.235

2011-05-26 16:51  marcella

	* [r11380] missing file

2011-05-26 16:43  marcella

	* [r11375] iterative davidson diagonalization

2011-05-26 03:18  vondele

	* [r11373] Automatic minor version update to 2.2.234

2011-05-25 09:36  ibethune

	* [r11372] Remove excessive dbcsr_error_set/stop in low level
	  routines.

2011-05-25 03:18  vondele

	* [r11369] Automatic minor version update to 2.2.233

2011-05-24 03:18  vondele

	* [r11367] Automatic minor version update to 2.2.232

2011-05-23 18:44  urban

	* [r11366] Removes INTENT from a pointer dummy argument.

2011-05-23 09:11  urban

	* [r11364] Correctly handle ISO_C_BINDING-based pointer reshaping
	  of 0-sizes arrays.

2011-05-23 03:18  vondele

	* [r11362] Automatic minor version update to 2.2.231

2011-05-22 22:52  urban

	* [r11361] Bugfix: Corrects distribution generation when
	  multiplying incompatibly-sized matrices.

2011-05-22 03:18  vondele

	* [r11359] Automatic minor version update to 2.2.230

2011-05-21 03:17  vondele

	* [r11357] Automatic minor version update to 2.2.229

2011-05-20 17:51  vondele

	* [r11356] bug fix for larger systems

2011-05-20 16:07  tlaino

	* [r11355] prettify

2011-05-20 15:56  tlaino

	* [r11350] bug fix

2011-05-20 10:52  urban

	* [r11349] Cleanup of DBCSR multiplication internals. Records
	  timing and data volume of transfers.

2011-05-20 03:18  vondele

	* [r11347] Automatic minor version update to 2.2.228

2011-05-19 14:51  urban

	* [r11346] Simplifies and improves scaling of local index
	  transpose.

2011-05-19 03:17  vondele

	* [r11344] Automatic minor version update to 2.2.227

2011-05-18 03:18  vondele

	* [r11338] Automatic minor version update to 2.2.226

2011-05-17 03:17  vondele

	* [r11333] Automatic minor version update to 2.2.225

2011-05-16 16:00  urban

	* [r11332] Cleanup to go along with NAG fix.

2011-05-16 14:52  urban

	* [r11331] OpenMP.

2011-05-16 13:22  urban

	* [r11330] NAG debugging is usefull so we try to make it work...

2011-05-16 11:52  vondele

	* [r11329] adjust write format

2011-05-16 03:17  vondele

	* [r11327] Automatic minor version update to 2.2.224

2011-05-15 03:18  vondele

	* [r11325] Automatic minor version update to 2.2.223

2011-05-14 06:40  krack

	* [r11324] bug fix (store the initial position, if needed, and keep
	  it)

2011-05-14 03:17  vondele

	* [r11322] Automatic minor version update to 2.2.222

2011-05-13 03:17  vondele

	* [r11320] Automatic minor version update to 2.2.221

2011-05-12 14:15  vondele

	* [r11319] test

2011-05-12 13:26  vondele

	* [r11318] test

2011-05-12 07:51  urban

	* [r11317] Simplification of selection, fixes non-MPI version.

2011-05-12 07:32  vondele

	* [r11316] fix quotes (Yannick)

2011-05-12 07:29  urban

	* [r11315] Fixes broken assumption of CRAY pointer support.

2011-05-12 07:14  marcella

	* [r11314] bug fix

2011-05-12 03:17  vondele

	* [r11312] Automatic minor version update to 2.2.220

2011-05-11 20:26  urban

	* [r11311] Pointer cleanups.
	  Cleans up MPI memory calling routines, leverages ISO_C_BINDING to
	  make some pointer tricks safer, and more updates to the
	  installation instructions.

2011-05-11 18:22  vondele

	* [r11310] replace some IO with mp.

2011-05-11 10:45  vondele

	* [r11309] add timeset/timestop

2011-05-11 10:44  vondele

	* [r11308] prettify

2011-05-11 03:18  vondele

	* [r11306] Automatic minor version update to 2.2.219

2011-05-10 21:38  urban

	* [r11305] Cleanup of INSTALL file.
	  Line breaks at 80 columns, adds some missing compile defines,
	  code
	  generation options (ssmp, psmp), and libsmm library.

2011-05-10 15:42  tlaino

	* [r11302] bug fix

2011-05-10 10:38  urban

	* [r11301] Performs the right MPI initialization when using
	  OpenMPI.

2011-05-10 03:17  vondele

	* [r11298] Automatic minor version update to 2.2.218

2011-05-09 20:44  urban

	* [r11297] We should only use funneled mode for threaded MPI (only
	  one thread may
	  call MPI routines and it must always be the same one).

2011-05-09 15:11  jgh

	* [r11295] Nullify all pointers on type initiation (GTH PP type)

2011-05-09 14:51  jgh

	* [r11294] Deallocate all memory from PP, avoid memory leaks

2011-05-09 13:46  jgh

	* [r11293] Complete rewrite of local pseudopotential integrals
	  using a general scheme.
	  Introduction of extended GTH potentials into the Quickstep code
	  Added regtests in QS and ATOM

2011-05-09 03:17  vondele

	* [r11290] Automatic minor version update to 2.2.217

2011-05-08 03:18  vondele

	* [r11288] Automatic minor version update to 2.2.216

2011-05-07 03:17  vondele

	* [r11286] Automatic minor version update to 2.2.215

2011-05-06 03:17  vondele

	* [r11284] Automatic minor version update to 2.2.214

2011-05-05 07:44  urban

	* [r11283] Prettifies the HMC(?) commit.

2011-05-05 07:42  urban

	* [r11282] CP2K DBCSR testing expanded to allow all 4 data types.
	  Fixes DBCSR CUDA for remaining 3 data types.

2011-05-05 03:17  vondele

	* [r11280] Automatic minor version update to 2.2.213

2011-05-04 03:17  vondele

	* [r11278] Automatic minor version update to 2.2.212

2011-05-03 12:54  mmcgrath

	* [r11277] Allows HMC to run with a restart file.

2011-05-03 03:17  vondele

	* [r11274] Automatic minor version update to 2.2.211

2011-05-02 10:50  tlaino

	* [r11273] Noam's patch (QMMM) + reset 16 tests

2011-05-02 03:18  vondele

	* [r11270] Automatic minor version update to 2.2.210

2011-05-01 03:18  vondele

	* [r11268] Automatic minor version update to 2.2.209

2011-04-30 03:17  vondele

	* [r11266] Automatic minor version update to 2.2.208

2011-04-29 14:03  urban

	* [r11265] Some OpenMP cleanups: use omp walltime when available
	  for CP2K m_walltime, avoids most ATOMIC uses, audit of
	  multi-thread access to shared dbcsr_data_obj structures, and a
	  few other cleanups.

2011-04-29 13:48  vondele

	* [r11264] more options

2011-04-29 12:16  vondele

	* [r11263] also fix the group size for farming

2011-04-29 11:03  vondele

	* [r11262] check on group size, and shift average option

2011-04-29 03:18  vondele

	* [r11260] Automatic minor version update to 2.2.207

2011-04-28 14:13  krack

	* [r11259] * Avoid overwriting of spline output for nonbonded 1-4
	  interactions
	  * Align printout

2011-04-28 12:11  ibethune

	* [r11258] Report memory usage on AIX/BlueGene correctly

2011-04-28 07:08  urban

	* [r11257] Fixes OpenMP memory leak.

2011-04-28 03:18  vondele

	* [r11255] Automatic minor version update to 2.2.206

2011-04-27 16:38  urban

	* [r11254] Changes module USEage and preprocessor directives to
	  ensure proper dependency generation.

2011-04-27 12:29  ibethune

	* [r11253] Removed MPI calls from within OpenMP parallel regions.
	  Now precomputed and stored in the fft_scratch type

2011-04-27 09:21  urban

	* [r11252] Prettify.

2011-04-27 09:18  urban

	* [r11251] Some CUDA clenaup: splits kernel for different types
	  into separate files, removes dead code.

2011-04-27 03:17  vondele

	* [r11249] Automatic minor version update to 2.2.205

2011-04-26 03:18  vondele

	* [r11246] Automatic minor version update to 2.2.204

2011-04-25 03:17  vondele

	* [r11244] Automatic minor version update to 2.2.203

2011-04-24 03:17  vondele

	* [r11242] Automatic minor version update to 2.2.202

2011-04-23 13:37  mmcgrath

	* [r11241] Also need the routine for the input

2011-04-23 09:34  mmcgrath

	* [r11240] Added the hybrid Monte Carlo move. Initial tests look
	  good, but further tests being done.

2011-04-23 03:18  vondele

	* [r11237] Automatic minor version update to 2.2.201

2011-04-22 12:49  urban

	* [r11236] Removes uneeded CPP flag and adds some DBCSR CUDA
	  documentation.

2011-04-22 12:06  urban

	* [r11235] Allows CUDA compilation with compilers having no
	  ISO_C_BINDING module.

2011-04-22 03:18  vondele

	* [r11233] Automatic minor version update to 2.2.200

2011-04-22 02:00  ikuo

	* [r11232] Fix pointer errors from qmmm wall skin.

2011-04-21 21:53  ikuo

	* [r11231] allow for variable skins for different dimensions option
	  for qmmm walls

2011-04-21 21:13  urban

	* [r11230] Adds precise POSIX timing when using CUDA or
	  OMP_GET_WTIME with OpenMP.

2011-04-21 14:03  tlaino

	* [r11229] use m_flush instead of FLUSH

2011-04-21 03:17  vondele

	* [r11225] Automatic minor version update to 2.2.199

2011-04-20 22:42  urban

	* [r11224] Moves DBCSR config initialization and resolves circular
	  dependency.

2011-04-20 21:06  marcella

	* [r11222] g-space mixing with spin polarization: mix total density
	  and spin density

2011-04-20 16:28  tlaino

	* [r11220] prettify + bug fix (Matt)

2011-04-20 14:18  urban

	* [r11219] Experimental CUDA support for DBCSR.

2011-04-20 12:54  urban

	* [r11218] Fixes & cleans up existing CUDA code.

2011-04-20 12:33  vondele

	* [r11217] bug fix for wrong NR to H mapping

2011-04-20 10:48  marcella

	* [r11213] bug fix: stress tensor with gapw, one term was missing

2011-04-20 07:29  vondele

	* [r11212] fix for cell files containing one line (Yannick)

2011-04-20 03:18  vondele

	* [r11210] Automatic minor version update to 2.2.198

2011-04-19 09:54  vondele

	* [r11209] eps_filter adjustments

2011-04-19 08:05  urban

	* [r11208] Corrects INTENT attribute.

2011-04-19 06:57  vondele

	* [r11207] some cleanup and infrastructure for testing of sp code

2011-04-19 06:28  tlaino

	* [r11206] speed-up patch nr.1 (Noam) + reset regs

2011-04-19 03:17  vondele

	* [r11203] Automatic minor version update to 2.2.197

2011-04-18 14:00  urban

	* [r11202] Adds operation to fill an existing symmetric matrix with
	  another matrix.
	  
	  The matrix is declared as having no symmetry but should be
	  mathematically symmetric.

2011-04-18 03:18  vondele

	* [r11200] Automatic minor version update to 2.2.196

2011-04-17 03:17  vondele

	* [r11198] Automatic minor version update to 2.2.195

2011-04-16 03:17  vondele

	* [r11196] Automatic minor version update to 2.2.194

2011-04-15 16:08  vondele

	* [r11195] silence debug write

2011-04-15 09:22  urban

	* [r11194] Adds utility routines for matching scalar types when
	  using mixed-precision
	  calculations.

2011-04-15 08:48  marcella

	* [r11193] x,y,z, components of x-ray absorption transition moments
	  not squared in output

2011-04-15 03:17  vondele

	* [r11191] Automatic minor version update to 2.2.193

2011-04-14 21:04  urban

	* [r11190] Forgotten file.

2011-04-14 14:34  urban

	* [r11189] Rewrite of desymmetrize_deep.
	  
	  Support added for all data types and detransposing transposed
	  blocks. Adds CP2K interface. Removes hack from qs_ls code.

2011-04-14 03:17  vondele

	* [r11187] Automatic minor version update to 2.2.192

2011-04-13 11:42  vondele

	* [r11186] allow for clustering

2011-04-13 10:39  vondele

	* [r11185] allow for a unit number in dbcsr_print

2011-04-13 06:50  marcella

	* [r11184] comment unused code

2011-04-13 03:17  vondele

	* [r11182] Automatic minor version update to 2.2.191

2011-04-12 14:47  tlaino

	* [r11181] bug fix

2011-04-12 03:17  vondele

	* [r11178] Automatic minor version update to 2.2.190

2011-04-11 03:17  vondele

	* [r11176] Automatic minor version update to 2.2.189

2011-04-10 03:17  vondele

	* [r11174] Automatic minor version update to 2.2.188

2011-04-09 03:17  vondele

	* [r11172] Automatic minor version update to 2.2.187

2011-04-08 03:17  vondele

	* [r11170] Automatic minor version update to 2.2.186

2011-04-07 03:18  vondele

	* [r11168] Automatic minor version update to 2.2.185

2011-04-06 07:20  urban

	* [r11167] Adds missing file.

2011-04-06 03:18  vondele

	* [r11165] Automatic minor version update to 2.2.184

2011-04-05 21:32  urban

	* [r11164] DBCSR configuration support.
	  
	  Adds configuration support for DBCSR options with a few
	  preliminary options. Adds CP2K input section for configuring
	  DBCSR and adds regtests to test several options.

2011-04-05 14:34  urban

	* [r11163] Skip copies/zeros for 0-length arrays.

2011-04-05 10:36  urban

	* [r11162] Introduces changes to support more special memory types.
	  
	  Multiple memory types can be used instead of a single "special"
	  MPI
	  memory type. The DBCSR matrix may have different data and index
	  memory types. New routines for managing special memory for
	  greater
	  code-sharing.

2011-04-05 03:18  vondele

	* [r11159] Automatic minor version update to 2.2.183

2011-04-03 03:18  vondele

	* [r11157] Automatic minor version update to 2.2.181

2011-04-02 03:17  vondele

	* [r11155] Automatic minor version update to 2.2.180

2011-04-01 09:00  vondele

	* [r11154] improved bound for the max number of blocks

2011-04-01 07:52  marcella

	* [r11152] bug fix

2011-04-01 03:17  vondele

	* [r11150] Automatic minor version update to 2.2.179

2011-03-31 18:48  vondele

	* [r11149] improved safe guard

2011-03-31 10:25  krack

	* [r11148] Fix error check. 2 regtests reset.

2011-03-31 09:02  krack

	* [r11147] Non-harmonic core-shell spring potential. 1 regtest
	  input added.

2011-03-31 03:17  vondele

	* [r11145] Automatic minor version update to 2.2.178

2011-03-30 12:29  tlaino

	* [r11144] bug fix

2011-03-30 11:49  krack

	* [r11143] Consider the first energy as reference energy for the
	  conserved quantity at MD startup, if not explicitly specified in
	  the input.
	  Less cryptic output label.

2011-03-30 10:01  tlaino

	* [r11142] print when coordinates are translated (if requested)

2011-03-30 09:16  krack

	* [r11141] prettify

2011-03-30 06:38  tlaino

	* [r11140] redefine the box used for fitting (Daniele)

2011-03-30 03:17  vondele

	* [r11138] Automatic minor version update to 2.2.177

2011-03-29 08:39  tlaino

	* [r11137] not needed

2011-03-29 05:37  tlaino

	* [r11136] checks for unimplemented

2011-03-29 05:13  tlaino

	* [r11135] bug fix (when using QM/MM)

2011-03-29 03:17  vondele

	* [r11133] Automatic minor version update to 2.2.176

2011-03-28 19:26  urban

	* [r11132] Resets FLOP count for each local multiplication.

2011-03-28 14:10  urban

	* [r11130] Bugfix.

2011-03-28 13:18  vondele

	* [r11129] implement rehashing and use smaller initial hash table
	  size

2011-03-28 12:59  urban

	* [r11128] Fixes memory leak. Removes debugging print.

2011-03-28 10:49  urban

	* [r11126] Updates local multiplication based on hierarchical
	  multiplication of submatrices.
	  
	  All matrices are now normalized to be fully non-transposed before
	  multiplication. Also, scaling and cropping is now part of
	  pre-processing. CP2K now matches matrices if the dimensions are
	  incompatible.
	  
	  Conversion to partially dense matrices is supported again and
	  with better support for threads.
	  
	  Adds support for libsmm small block multiplication.
	  
	  Better separation of data multiplication drivers (BLAS, libsmm,
	  compiler, PLASMA).

2011-03-28 03:17  vondele

	* [r11124] Automatic minor version update to 2.2.175

2011-03-27 03:18  vondele

	* [r11122] Automatic minor version update to 2.2.174

2011-03-26 04:18  vondele

	* [r11120] Automatic minor version update to 2.2.173

2011-03-25 04:18  vondele

	* [r11118] Automatic minor version update to 2.2.172

2011-03-24 04:18  vondele

	* [r11116] Automatic minor version update to 2.2.171

2011-03-23 22:52  krack

	* [r11115] * Binary restart for shell thermostats
	  * 4 new regtest inputs
	  * In principle the SAVE_MEM feature has now become obsolete

2011-03-23 04:17  vondele

	* [r11113] Automatic minor version update to 2.2.170

2011-03-22 04:17  vondele

	* [r11111] Automatic minor version update to 2.2.169

2011-03-21 04:17  vondele

	* [r11109] Automatic minor version update to 2.2.168

2011-03-20 04:18  vondele

	* [r11107] Automatic minor version update to 2.2.167

2011-03-19 04:17  vondele

	* [r11105] Automatic minor version update to 2.2.166

2011-03-18 15:00  vondele

	* [r11103] prettifix

2011-03-18 04:17  vondele

	* [r11101] Automatic minor version update to 2.2.165

2011-03-17 15:44  marcella

	* [r11100] occupation for the admm orbitals

2011-03-17 12:05  marcella

	* [r11098] better restart with fractional occupation numbers

2011-03-17 04:18  vondele

	* [r11096] Automatic minor version update to 2.2.164

2011-03-16 15:41  krack

	* [r11095] * Bug fix: nullify pointer, forgotten IOSTAT check
	  * 3 regtest inputs added to check the switching from ASCII to
	  binary restart file

2011-03-16 09:46  krack

	* [r11094] Bug fix for spin-polarised calculations, 1 regtest reset

2011-03-16 04:17  vondele

	* [r11092] Automatic minor version update to 2.2.163

2011-03-15 16:31  krack

	* [r11091] * Enable automatic U ramping with DFT+U
	  * 2 new regtest inputs added

2011-03-15 08:26  tlaino

	* [r11088] adding stride option to dump cube file

2011-03-15 04:17  vondele

	* [r11086] Automatic minor version update to 2.2.162

2011-03-14 22:11  marcella

	* [r11085] make pretty

2011-03-14 22:10  marcella

	* [r11084] bug fix

2011-03-14 16:48  marcella

	* [r11079] smearing + lsd: by default one common Fermi level is
	  computed for spin up and spin down, thus allowing not integer
	  numbers of electron per spin channel (but always keeping an
	  integer and constant total number of electrons). To force a given
	  difference in number of electrons between spin up and spin down,
	  when smearing is active, use the keyword FIXED_MAGNETIC_MOMENT of
	  section SMEAR. This difference does not need to be an integer.

2011-03-14 04:18  vondele

	* [r11074] Automatic minor version update to 2.2.161

2011-03-13 04:18  vondele

	* [r11072] Automatic minor version update to 2.2.160

2011-03-12 04:17  vondele

	* [r11070] Automatic minor version update to 2.2.159

2011-03-11 04:17  vondele

	* [r11068] Automatic minor version update to 2.2.158

2011-03-10 04:17  vondele

	* [r11066] Automatic minor version update to 2.2.157

2011-03-09 04:17  vondele

	* [r11064] Automatic minor version update to 2.2.156

2011-03-08 19:52  marcella

	* [r11063] pretty

2011-03-08 19:40  marcella

	* [r11062] some more cleaning of qs_ks_methods

2011-03-08 15:49  tlaino

	* [r11061] prettify

2011-03-08 10:13  marcella

	* [r11060] some more cleaning of qs_ks_methods

2011-03-08 04:17  vondele

	* [r11058] Automatic minor version update to 2.2.155

2011-03-07 12:05  jgh

	* [r11057] Pseudopotential and basis set optimization: some
	  cleaning and improvements.

2011-03-07 09:22  marcella

	* [r11056] some more cleaning of qs_ks_methods

2011-03-07 04:18  vondele

	* [r11054] Automatic minor version update to 2.2.154

2011-03-06 22:17  marcella

	* [r11053] some more cleaning of qs_ks_methods

2011-03-06 21:45  tlaino

	* [r11052] prettify

2011-03-06 04:18  vondele

	* [r11050] Automatic minor version update to 2.2.153

2011-03-05 12:25  marcella

	* [r11049] some cleaning of the qs_ks_methods module

2011-03-05 04:17  vondele

	* [r11047] Automatic minor version update to 2.2.152

2011-03-04 12:56  mmcgrath

	* [r11045] Added checks for some input values requiring values for
	  each type of molecule.

2011-03-04 11:17  krack

	* [r11044] No need to update the external COORD section when a
	  binary restart file is written

2011-03-04 04:18  vondele

	* [r11042] Automatic minor version update to 2.2.151

2011-03-03 16:55  krack

	* [r11041] Further speedup for the initialisation with MASSIVE
	  thermostats by using a binary search.

2011-03-03 15:00  krack

	* [r11040] * Quadratically scaling loop removed in the startup for
	  MASSIVE thermostats.
	  This loop was an annoying bottleneck for systems with a large
	  number of atoms,
	  e.g. consuming ~2400s for a 768000 atom system with a massive
	  thermostat instead of a few seconds.
	  * More timers added

2011-03-03 07:30  krack

	* [r11039] Missing IMPLICIT NONE added. More timers added for a
	  better traceback.

2011-03-03 04:17  vondele

	* [r11037] Automatic minor version update to 2.2.150

2011-03-02 15:25  vondele

	* [r11036] Added new load balancing routine (experimental) for HFX.
	  [Michael Steinlechner]

2011-03-02 04:17  vondele

	* [r11034] Automatic minor version update to 2.2.149

2011-03-01 10:05  krack

	* [r11033] Write also Nose thermostat information to the binary
	  restart file, if SPLIT_RESTART_FILE is requested.

2011-03-01 04:18  vondele

	* [r11031] Automatic minor version update to 2.2.148

2011-02-28 04:17  vondele

	* [r11029] Automatic minor version update to 2.2.147

2011-02-27 04:17  vondele

	* [r11027] Automatic minor version update to 2.2.146

2011-02-26 04:18  vondele

	* [r11025] Automatic minor version update to 2.2.145

2011-02-25 13:54  krack

	* [r11024] Getting rid of old/wrong stop_program calls (finished)

2011-02-25 12:48  tlaino

	* [r11023] minor cleans

2011-02-25 04:18  vondele

	* [r11020] Automatic minor version update to 2.2.144

2011-02-25 00:40  tlaino

	* [r11019] re-adding the cp_common_uses.h from where it has been
	  removed

2011-02-25 00:02  tlaino

	* [r11017] using the same convention for routineN

2011-02-24 20:08  krack

	* [r11016] Getting rid of old/wrong stop_program calls (3rd stage)

2011-02-24 15:00  krack

	* [r11013] Getting rid of old/wrong stop_program calls (2nd stage)

2011-02-24 12:06  krack

	* [r11012] Getting rid of old/wrong stop_program calls (1st stage)

2011-02-24 04:18  vondele

	* [r11010] Automatic minor version update to 2.2.143

2011-02-23 20:38  krack

	* [r11009] Fix for substring out-of-bounds error

2011-02-23 19:31  krack

	* [r11008] * Optional splitting of the CP2K restart file enabled.
	  This allows that
	  selected input sections, which are growing with the number of
	  atoms
	  in the system, are written to another restart file in binary
	  format
	  instead of the default restart file in human readable ASCII
	  format.
	  This split of the restart file may provide significant memory
	  savings
	  and an accelerated I/O for systems with a very large number of
	  atoms.
	  Currently, only the COORD and VELOCITY sections can be written in
	  binary format. The sections for massive thermostats are not yet
	  included (to be done).
	  * New MOUDULE file input_cp2k_binary_restarts.F added
	  * MODULE termination revised:
	  (1) external_controls moved from MODULE termination to the new
	  MODULE file
	  cp_external_controls to keep MODULE termination low-level
	  (2) specific stop_ routines added which allow a low-level program
	  termination
	  with a detailed error recovery.
	  * MODULE cp_files revised:
	  (1) enhanced error recovery
	  (2) some minor bugs fixed
	  (3) File preconnection managemant added
	  * 16 new regtest inputs added for binary restart testing

2011-02-23 18:25  tlaino

	* [r11007] added a minimum intermeta step counter: this fixes the
	  strange behavior for min_disp as
	  reported by Pablo. The original behavior is still preserved by
	  the default values.

2011-02-23 04:18  vondele

	* [r11005] Automatic minor version update to 2.2.142

2011-02-22 16:59  tlaino

	* [r11004] unimplemented feature

2011-02-22 16:59  tlaino

	* [r11003] prettify

2011-02-22 04:17  vondele

	* [r11001] Automatic minor version update to 2.2.141

2011-02-21 06:14  tlaino

	* [r11000] bug fix

2011-02-21 04:18  vondele

	* [r10998] Automatic minor version update to 2.2.140

2011-02-20 04:18  vondele

	* [r10996] Automatic minor version update to 2.2.139

2011-02-19 04:18  vondele

	* [r10994] Automatic minor version update to 2.2.138

2011-02-18 04:18  vondele

	* [r10992] Automatic minor version update to 2.2.137

2011-02-17 04:18  vondele

	* [r10990] Automatic minor version update to 2.2.136

2011-02-16 14:29  krack

	* [r10989] * Getting rid of old stop_memory routine
	  * All buggy uses of new routine fixed
	  * Minimize allocation size for work array in reallocate

2011-02-16 04:18  vondele

	* [r10987] Automatic minor version update to 2.2.135

2011-02-15 04:18  vondele

	* [r10985] Automatic minor version update to 2.2.134

2011-02-14 13:31  ibethune

	* [r10984] Fix intermittent bug - column map access not thread-safe
	  (thanks Urban!)

2011-02-14 04:18  vondele

	* [r10982] Automatic minor version update to 2.2.133

2011-02-13 04:18  vondele

	* [r10980] Automatic minor version update to 2.2.132

2011-02-12 15:37  tlaino

	* [r10978] allow the usage of any combination of the energies in a
	  MIXED_ENV as possible collective variable: inspired by the
	  request of Peter for using IP (ionization potential) as CV. This
	  allows an infinite combination of energy functions - while
	  propagating the EOM on just a single Hamiltonian (see example
	  committed).

2011-02-12 04:17  vondele

	* [r10976] Automatic minor version update to 2.2.131

2011-02-11 04:18  vondele

	* [r10973] Automatic minor version update to 2.2.130

2011-02-10 13:06  fschiff

	* [r10972] bug

2011-02-10 12:26  fschiff

	* [r10971] fixing same bug in different place

2011-02-10 10:28  krack

	* [r10968] prettify

2011-02-10 10:22  fschiff

	* [r10967] bug fix to get the correct sign of the determinant

2011-02-10 08:42  fschiff

	* [r10966] make type public for g95

2011-02-10 08:29  fschiff

	* [r10965] missing type file

2011-02-10 08:24  fschiff

	* [r10964] better to use cvs add for new routines

2011-02-10 08:15  fschiff

	* [r10963] forgot to add the new routines. Sorry for that

2011-02-10 08:12  fschiff

	* [r10962] forgot to add the new routines. Sorry for that

2011-02-10 07:57  fschiff

	* [r10961] adding possibilty for static, periodic electric fields
	  in DFT calculations using Berry phase

2011-02-10 04:18  vondele

	* [r10959] Automatic minor version update to 2.2.129

2011-02-09 15:33  krack

	* [r10958] bugfix for COORD section dump with scaled coordinates

2011-02-09 04:18  vondele

	* [r10956] Automatic minor version update to 2.2.128

2011-02-08 10:28  krack

	* [r10955] Allow proper restart with an external &COORD section
	  (only relevant for COORD_FILE_FORMAT cp2k)

2011-02-08 04:18  vondele

	* [r10953] Automatic minor version update to 2.2.127

2011-02-07 19:45  vondele

	* [r10952] fix eigenvalues for the admm/pdos combination.

2011-02-07 04:18  vondele

	* [r10950] Automatic minor version update to 2.2.126

2011-02-06 04:18  vondele

	* [r10948] Automatic minor version update to 2.2.125

2011-02-05 04:17  vondele

	* [r10946] Automatic minor version update to 2.2.124

2011-02-04 12:49  urban

	* [r10944] Bugfix (did not affect CP2K): complex matrix is
	  correctly transposed.

2011-02-04 04:18  vondele

	* [r10942] Automatic minor version update to 2.2.123

2011-02-03 04:18  vondele

	* [r10940] Automatic minor version update to 2.2.122

2011-02-02 04:18  vondele

	* [r10937] Automatic minor version update to 2.2.121

2011-02-01 04:18  vondele

	* [r10934] Automatic minor version update to 2.2.120

2011-01-31 17:24  tlaino

	* [r10933] enable constraints on the cell for NPT_F

2011-01-31 15:45  krack

	* [r10932] * enable relax_multip also for OTdiag
	  * bug fix for OTdiag with smear
	  * one regtest added and one reset

2011-01-31 07:52  vondele

	* [r10931] fix compilation with __FFTMKL (Jiang, Hao)

2011-01-31 04:18  vondele

	* [r10929] Automatic minor version update to 2.2.119

2011-01-30 04:17  vondele

	* [r10927] Automatic minor version update to 2.2.118

2011-01-29 04:18  vondele

	* [r10924] Automatic minor version update to 2.2.117

2011-01-28 10:57  vondele

	* [r10922] some cleanup

2011-01-28 04:18  vondele

	* [r10920] Automatic minor version update to 2.2.116

2011-01-27 21:23  vondele

	* [r10919] cleanup

2011-01-27 21:21  krack

	* [r10918] Fix out-of-bounds runtime error

2011-01-27 19:51  vondele

	* [r10917] remove unused subroutine

2011-01-27 16:14  krack

	* [r10916] * Option RELAX_MULTIPLICITY added. It allows the dynamic
	  change of multiplicity during a UKS calculation if
	  diagonalization with ADDED_MOS is used
	  * Routine print_message revised

2011-01-27 04:18  vondele

	* [r10914] Automatic minor version update to 2.2.115

2011-01-26 10:11  vondele

	* [r10911] some ref_cell cleanup

2011-01-26 04:17  vondele

	* [r10909] Automatic minor version update to 2.2.114

2011-01-25 16:33  tlaino

	* [r10908] bug fix

2011-01-25 04:17  vondele

	* [r10906] Automatic minor version update to 2.2.113

2011-01-24 04:17  vondele

	* [r10904] Automatic minor version update to 2.2.112

2011-01-23 04:17  vondele

	* [r10902] Automatic minor version update to 2.2.111

2011-01-22 04:17  vondele

	* [r10900] Automatic minor version update to 2.2.110

2011-01-21 04:17  vondele

	* [r10898] Automatic minor version update to 2.2.109

2011-01-20 23:16  tlaino

	* [r10897] bug fix

2011-01-20 04:17  vondele

	* [r10895] Automatic minor version update to 2.2.108

2011-01-19 08:38  marcella

	* [r10892] bug fix: check for associated pointer

2011-01-19 08:07  marcella

	* [r10891] check for associated pointer

2011-01-19 04:17  vondele

	* [r10889] Automatic minor version update to 2.2.107

2011-01-18 04:17  vondele

	* [r10886] Automatic minor version update to 2.2.106

2011-01-17 16:50  krack

	* [r10885] * Optionally read CP2K COORD section from an external
	  file
	  * Allow to disable all exclusion list for Fist

2011-01-17 04:18  vondele

	* [r10883] Automatic minor version update to 2.2.105

2011-01-16 04:18  vondele

	* [r10881] Automatic minor version update to 2.2.104

2011-01-15 04:18  vondele

	* [r10879] Automatic minor version update to 2.2.103

2011-01-14 04:18  vondele

	* [r10877] Automatic minor version update to 2.2.102

2011-01-13 20:19  vondele

	* [r10876] remove unused bogus function

2011-01-13 07:30  krack

	* [r10875] * Speedup atomic coordinate dump for large systems by
	  reducing the number of string operations
	  * Optionally add molecule number

2011-01-13 04:18  vondele

	* [r10873] Automatic minor version update to 2.2.101

2011-01-12 17:00  urban

	* [r10872] OpenMP bugfix: protects thread-shared variable access.

2011-01-12 04:18  vondele

	* [r10870] Automatic minor version update to 2.2.100

2011-01-11 17:01  krack

	* [r10867] * Optionally recognize also the 6th column of the COORD
	  section as residue name (CONNECTIVITY user)
	  * Allow parsing of fractions also in the COORD section like
	  elsewhere in the input
	  * Improved error handling for input errors in the COORD section
	  * 3 regtests updated

2011-01-11 04:17  vondele

	* [r10862] Automatic minor version update to 2.2.99

2011-01-10 20:10  vondele

	* [r10861] some cleanup

2011-01-10 18:35  vondele

	* [r10860] small speedup

2011-01-10 12:22  urban

	* [r10859] Avoids unnecessary existing diagonal-block transposes
	  (slightly changes numerical behavior) for multiplies with a
	  symmetric product matrix.

2011-01-10 11:56  urban

	* [r10858] Removes the need for detransposing blocks for CP2K after
	  a multiply.

2011-01-10 04:18  vondele

	* [r10856] Automatic minor version update to 2.2.98

2011-01-09 04:18  vondele

	* [r10854] Automatic minor version update to 2.2.97

2011-01-08 04:18  vondele

	* [r10852] Automatic minor version update to 2.2.96

2011-01-07 04:18  vondele

	* [r10849] Automatic minor version update to 2.2.95

2011-01-06 10:03  urban

	* [r10848] Changes a FORALL loop to a DO loop for better compiler
	  compatability.

2011-01-06 08:23  tlaino

	* [r10847] define a fft_kinds and some more re-organization

2011-01-06 04:18  vondele

	* [r10845] Automatic minor version update to 2.2.94

2011-01-05 07:18  tlaino

	* [r10844] prettify

2011-01-05 04:18  vondele

	* [r10842] Automatic minor version update to 2.2.93

2011-01-04 23:58  tlaino

	* [r10841] a kind module to make the lib more robust; prettifying

2011-01-04 20:25  vondele

	* [r10840] more unused vars

2011-01-04 19:32  vondele

	* [r10839] remove more unused vars

2011-01-04 18:23  vondele

	* [r10838] remove some unused variables

2011-01-04 18:09  tlaino

	* [r10837] fix bug

2011-01-04 15:35  vondele

	* [r10835] Add a linear scaling SCF procedure based on the density
	  matrix.

2011-01-04 13:01  vondele

	* [r10834] avoid hard-coded limit

2011-01-04 12:11  vondele

	* [r10833] only access pointers when needed

2011-01-04 04:18  vondele

	* [r10831] Automatic minor version update to 2.2.92

2011-01-03 11:42  vondele

	* [r10830] group timing calls and more debug output

2011-01-03 04:18  vondele

	* [r10828] Automatic minor version update to 2.2.91

2011-01-02 04:18  vondele

	* [r10826] Automatic minor version update to 2.2.90

2011-01-01 04:17  vondele

	* [r10824] Automatic minor version update to 2.2.89

2010-12-31 16:04  tlaino

	* [r10823] update 2011

2010-12-31 04:18  vondele

	* [r10820] Automatic minor version update to 2.2.88

2010-12-30 04:18  vondele

	* [r10818] Automatic minor version update to 2.2.87

2010-12-29 13:21  tlaino

	* [r10817] Allow different charges for same atomic kinds + regs.

2010-12-29 04:18  vondele

	* [r10815] Automatic minor version update to 2.2.86

2010-12-28 04:18  vondele

	* [r10812] Automatic minor version update to 2.2.85

2010-12-27 04:17  vondele

	* [r10810] Automatic minor version update to 2.2.84

2010-12-26 04:18  vondele

	* [r10808] Automatic minor version update to 2.2.83

2010-12-25 04:18  vondele

	* [r10806] Automatic minor version update to 2.2.82

2010-12-24 04:18  vondele

	* [r10804] Automatic minor version update to 2.2.81

2010-12-23 22:42  tlaino

	* [r10803] prettify

2010-12-23 18:41  vondele

	* [r10802] bug fix

2010-12-23 04:18  vondele

	* [r10799] Automatic minor version update to 2.2.80

2010-12-22 16:44  tlaino

	* [r10794] removing dead code

2010-12-22 14:52  tlaino

	* [r10793] making the memory progression factor input driven (by
	  default 1.2)

2010-12-22 14:41  krack

	* [r10792] Revert buffer growth progression factor to 1.2

2010-12-22 13:27  vondele

	* [r10791] bug fix for basis none as first kind

2010-12-22 08:32  vondele

	* [r10785] by default do not echo a full list of host names / pids

2010-12-22 08:13  tlaino

	* [r10784] allow FARMING to have a restart file name depending on
	  the project name of
	  input driven. Added a &RESTART print_key section and a
	  RESTART_FILE_NAME keyword.
	  Moved RESTART logical flag to DO_RESTART.

2010-12-22 04:18  vondele

	* [r10782] Automatic minor version update to 2.2.79

2010-12-21 19:48  krack

	* [r10781] Connectivity option user added (can be handy to reduce
	  the memory allocations for large systems).

2010-12-21 15:34  tlaino

	* [r10779] getting rid of hardcoded I/O unit 6 - assigning proper
	  IO control

2010-12-21 04:18  vondele

	* [r10771] Automatic minor version update to 2.2.78

2010-12-20 04:18  vondele

	* [r10758] Automatic minor version update to 2.2.77

2010-12-19 04:18  vondele

	* [r10755] Automatic minor version update to 2.2.76

2010-12-18 04:18  vondele

	* [r10752] Automatic minor version update to 2.2.75

2010-12-17 08:44  urban

	* [r10746] Reinstates patch.

2010-12-17 07:47  vondele

	* [r10745] remove unused param

2010-12-17 04:18  vondele

	* [r10743] Automatic minor version update to 2.2.74

2010-12-16 22:22  urban

	* [r10742] Bugfix: also resets the declared size of the data area.

2010-12-16 21:15  vondele

	* [r10741] further streamlining

2010-12-16 18:56  vondele

	* [r10740] change data_type to integer

2010-12-16 16:30  vondele

	* [r10739] some more debug/careful only asserts

2010-12-16 16:14  vondele

	* [r10738] revert buggy patch

2010-12-16 15:04  urban

	* [r10737] Quickly creates a new index upon matrix creation to
	  avoid an expensive
	  finalize.

2010-12-16 04:18  vondele

	* [r10735] Automatic minor version update to 2.2.73

2010-12-15 12:53  urban

	* [r10733] Fixes memory leak.

2010-12-15 04:18  vondele

	* [r10730] Automatic minor version update to 2.2.72

2010-12-14 15:21  vondele

	* [r10728] (re)enable the molecular distribution

2010-12-14 10:23  vondele

	* [r10726] move eam data out of particle_type

2010-12-14 04:18  vondele

	* [r10721] Automatic minor version update to 2.2.71

2010-12-13 16:05  vondele

	* [r10720] more local goodies (Urban)

2010-12-13 14:16  vondele

	* [r10719] improved block local indexing (Urban)

2010-12-13 12:15  vondele

	* [r10717] change the mic pair list build, and optimize the subcell
	  choice

2010-12-13 04:18  vondele

	* [r10712] Automatic minor version update to 2.2.70

2010-12-12 14:59  vondele

	* [r10711] small speedup

2010-12-12 04:18  vondele

	* [r10709] Automatic minor version update to 2.2.69

2010-12-11 04:18  vondele

	* [r10707] Automatic minor version update to 2.2.68

2010-12-10 20:51  vondele

	* [r10706] use ids also for tmp arrays

2010-12-10 20:09  vondele

	* [r10705] one more string variable to an id.

2010-12-10 15:48  vondele

	* [r10704] alloc lmi only when needed

2010-12-10 13:16  vondele

	* [r10703] only alloc lci if needed

2010-12-10 08:20  urban

	* [r10701] Matrix rows/columns local to a process grid row/column
	  are computed on-demand and for all grid rows/columns.

2010-12-10 04:17  vondele

	* [r10699] Automatic minor version update to 2.2.67

2010-12-09 20:09  vondele

	* [r10698] distribution_2d cleanup. It is actually almost not
	  needed.

2010-12-09 18:27  vondele

	* [r10697] dbcsr_dist added to qs_env

2010-12-09 14:06  vondele

	* [r10696] timeset

2010-12-09 12:45  vondele

	* [r10695] fix typo in timeset

2010-12-09 09:13  vondele

	* [r10692] prettify

2010-12-09 09:07  vondele

	* [r10691] use a heap

2010-12-09 04:18  vondele

	* [r10689] Automatic minor version update to 2.2.66

2010-12-08 20:58  vondele

	* [r10688] no need anymore for a finalize after a create

2010-12-08 20:35  vondele

	* [r10687] more timeset/timestop

2010-12-08 20:23  urban

	* [r10686] Finalizes matrix on creation (making it immediately
	  "valid").

2010-12-08 20:17  vondele

	* [r10685] prettify

2010-12-08 15:49  vondele

	* [r10684] more of the same

2010-12-08 15:38  vondele

	* [r10683] fix for an early return

2010-12-08 14:29  vondele

	* [r10682] some more timeset/timestop

2010-12-08 13:56  vondele

	* [r10681] more spaces

2010-12-08 04:18  vondele

	* [r10679] Automatic minor version update to 2.2.65

2010-12-07 09:01  vondele

	* [r10678] add timeset

2010-12-07 04:18  vondele

	* [r10676] Automatic minor version update to 2.2.64

2010-12-06 04:18  vondele

	* [r10674] Automatic minor version update to 2.2.63

2010-12-05 15:44  vondele

	* [r10673] use the string_table more often

2010-12-05 04:18  vondele

	* [r10671] Automatic minor version update to 2.2.62

2010-12-04 04:17  vondele

	* [r10669] Automatic minor version update to 2.2.61

2010-12-03 04:17  vondele

	* [r10667] Automatic minor version update to 2.2.60

2010-12-02 18:00  tlaino

	* [r10666] addendum at the masterpiece: kind of cherry on the
	  cake..

2010-12-02 15:28  vondele

	* [r10665] move the exclusions from the particle_type to fist.

2010-12-02 14:50  urban

	* [r10664] Removes "special memory" argument in cp_dbcsr_create.

2010-12-02 13:34  vondele

	* [r10663] only build exclusion lists if the method is fist

2010-12-02 11:16  vondele

	* [r10662] remove unused routine

2010-12-02 08:54  krack

	* [r10661] Bug fix: correct index

2010-12-02 08:45  vondele

	* [r10660] another case where ignore_topology can be used

2010-12-02 04:18  vondele

	* [r10658] Automatic minor version update to 2.2.59

2010-12-01 18:36  krack

	* [r10657] Enable Band (NEB) runs with the core-shell model

2010-12-01 16:00  urban

	* [r10656] Fixes unbalanced MPI allocate/deallocate. Removes some
	  direct accesses to subtypes.

2010-12-01 10:39  vondele

	* [r10655] exchange cutoff takes the default into account if not
	  specified explicitly

2010-12-01 10:26  tlaino

	* [r10654] still some missing infos..

2010-12-01 10:15  tlaino

	* [r10653] RCUT is sufficient as a name - specifying possible side
	  effects

2010-12-01 09:49  vondele

	* [r10652] allow for explicitly setting ewald_r_cut

2010-12-01 09:28  vondele

	* [r10651] revert unfinished bit

2010-12-01 09:25  vondele

	* [r10650] Fix bug for SE and EWALD for systems with molecular
	  topology

2010-12-01 04:18  vondele

	* [r10648] Automatic minor version update to 2.2.58

2010-11-30 11:34  hforbert

	* [r10647] Improve source code formatting (prettify by hand)
	  [lukasz]

2010-11-30 07:20  vondele

	* [r10646] pretty

2010-11-30 04:18  vondele

	* [r10644] Automatic minor version update to 2.2.57

2010-11-29 18:11  vondele

	* [r10643] code around a NAG checking bug

2010-11-29 16:27  vondele

	* [r10642] double nullify

2010-11-29 16:13  vondele

	* [r10641] use cp_dbcsr_alloc_block_from_nbl in qs_kinetic /
	  qs_overlap.

2010-11-29 15:50  vondele

	* [r10640] more undefined pointers

2010-11-29 14:56  vondele

	* [r10639] remove commented code

2010-11-29 14:25  urban

	* [r10638] 2nd half of commit.

2010-11-29 14:19  vondele

	* [r10637] add a needed nullify

2010-11-29 13:53  vondele

	* [r10636] fix missing nullify

2010-11-29 13:52  vondele

	* [r10635] avoid recursive IO

2010-11-29 13:06  vondele

	* [r10634] test

2010-11-29 12:57  urban

	* [r10633] Reverts Friday's patch changing
	  allocate-from-nonbond-list behavior.

2010-11-29 11:33  vondele

	* [r10632] protect against div by zero

2010-11-29 11:22  vondele

	* [r10631] init

2010-11-29 04:18  vondele

	* [r10629] Automatic minor version update to 2.2.56

2010-11-28 04:18  vondele

	* [r10627] Automatic minor version update to 2.2.55

2010-11-27 04:18  vondele

	* [r10625] Automatic minor version update to 2.2.54

2010-11-26 19:25  vondele

	* [r10624] *** empty log message ***

2010-11-26 16:37  vondele

	* [r10623] more efficient way to add blocks to a matrix

2010-11-26 16:25  vondele

	* [r10622] prettify

2010-11-26 15:04  jgh

	* [r10621] Code cleaning

2010-11-26 13:32  hforbert

	* [r10620] Added printing of bead and centroid velocities in PINT.
	  [lukasz]

2010-11-26 09:03  fschiff

	* [r10617] bug fix in the computation of Wda (found by Hanning
	  Chen)

2010-11-26 04:18  vondele

	* [r10615] Automatic minor version update to 2.2.53

2010-11-25 16:39  vondele

	* [r10613] improve output

2010-11-25 16:24  vondele

	* [r10612] bug fix uninit var

2010-11-25 13:28  vondele

	* [r10611] one more obvious candidate

2010-11-25 12:49  vondele

	* [r10610] skip some calculation of some unused neighbor lists

2010-11-25 11:41  vondele

	* [r10609] clarify nl names

2010-11-25 04:18  vondele

	* [r10607] Automatic minor version update to 2.2.52

2010-11-24 21:02  vondele

	* [r10606] missing dbcsr_error_stop

2010-11-24 14:19  vondele

	* [r10605] faster norms

2010-11-24 11:28  vondele

	* [r10604] helper function

2010-11-24 08:36  vondele

	* [r10602] more flexible memory choice

2010-11-24 04:18  vondele

	* [r10600] Automatic minor version update to 2.2.51

2010-11-23 20:41  vondele

	* [r10599] fix bogus dealloc

2010-11-23 17:53  tlaino

	* [r10597] cleaning and putting some order: dbcsr has its own lib
	  directory as well as fft : first this is a way to acknowledge the
	  work done
	  so far by Urban and Valery for dbcsr (its own directory
	  indipendent from the rest) and second helps to keep separated
	  things that
	  nothing have to do (fft, dbcsr, rest of lib content).
	  Finally, this should also make easier to use these libs in other
	  codes.

2010-11-23 16:37  tlaino

	* [r10596] additional checks on allocation

2010-11-23 16:25  ibethune

	* [r10595] Bugfix: non-ortho grids and OpenMP out-of-bounds
	  (reported by Florian)

2010-11-23 15:54  vondele

	* [r10593] speed up summing of replicated matrices, restricting to
	  matrices with the same layout everywhere

2010-11-23 13:34  vondele

	* [r10592] add timeset/timestop

2010-11-23 12:24  vondele

	* [r10591] refactoring mopac guess

2010-11-23 09:01  vondele

	* [r10590] bug fix passing uninit variables

2010-11-23 04:18  vondele

	* [r10588] Automatic minor version update to 2.2.50

2010-11-22 20:24  vondele

	* [r10587] some cleanup on where to compute matrix_w

2010-11-22 20:14  vondele

	* [r10586] fix bug in passing unallocated arrays around

2010-11-22 18:01  tlaino

	* [r10585] change definition of mm_radius (Toon - 0032)

2010-11-22 17:58  tlaino

	* [r10584] restore previous version (this commit still requires
	  some checks)

2010-11-22 17:46  tlaino

	* [r10583] change definition in mm_radius (Toon 0032)

2010-11-22 16:19  vondele

	* [r10581] cleanup has_unit_metric. Fixes a bug with semi-empirical
	  and replica_env

2010-11-22 15:01  vondele

	* [r10580] cleanup

2010-11-22 14:37  vondele

	* [r10579] remove comment

2010-11-22 14:10  vondele

	* [r10578] more careful mod

2010-11-22 12:34  vondele

	* [r10577] more careful mod

2010-11-22 04:18  vondele

	* [r10575] Automatic minor version update to 2.2.49

2010-11-21 18:52  krack

	* [r10574] Reduce code redundancy

2010-11-21 04:18  vondele

	* [r10572] Automatic minor version update to 2.2.48

2010-11-20 04:18  vondele

	* [r10570] Automatic minor version update to 2.2.47

2010-11-19 12:04  vondele

	* [r10569] careful_mod

2010-11-19 11:43  jgh

	* [r10568] Code cleaning.

2010-11-19 11:21  vondele

	* [r10567] a couple of more error_set/stop

2010-11-19 11:02  vondele

	* [r10566] local indexing in the multiply (Urban)

2010-11-19 09:44  vondele

	* [r10565] on the fly filtering (urban)

2010-11-19 08:03  krack

	* [r10564] Add unit

2010-11-19 04:18  vondele

	* [r10562] Automatic minor version update to 2.2.46

2010-11-18 23:20  tlaino

	* [r10560] Rename scale_cutoff to shift_cutoff; changing few
	  variables' names (Toon - 0026/0027)

2010-11-18 23:01  tlaino

	* [r10559] Fix a bug in case there is only ei and no vdw; Compute
	  the correction factor correctly in nonbonding correction (Toon -
	  0023 / 0025)

2010-11-18 22:36  tlaino

	* [r10558] NONBONDED14: The option is no longer useful. Moreover
	  cp2k would probably crash
	  when nonbonded and nonbonded14 flags would be different in an
	  input
	  file. There are no tests either that have such combination of
	  input flags. Moreover: The name in the source code and the input
	  are made consistent. It
	  was also confusing to have a nonbonded flag and a nonbonded
	  section
	  in the FORCEFIELD section.

2010-11-18 22:15  tlaino

	* [r10557] more clean-up ( Toon 0019 - 0020 )

2010-11-18 21:38  tlaino

	* [r10556] fix of the bug fix : group should be taken from the
	  right communicator, not just one randomly.
	  Cleanup of several routines.

2010-11-18 20:19  tlaino

	* [r10555] New reference energy for ei14 + prettify ( Toon -
	  0012/0013 )

2010-11-18 19:47  tlaino

	* [r10554] his is the critical patch that reroutes the onfo
	  evaluation. The
	  contribution is simply removed from fist_intra_force, and changes
	  a few
	  other things to activate the onfo evaluation on
	  fist_nonbond_force.F.
	  Many preparative patches are now effectively used. There is still
	  a lot
	  of stuff to be removed, which will be done in later patches
	  (Toon).

2010-11-18 18:57  tlaino

	* [r10553] Introduce potparm14 in nonbonding environment; Refer to
	  14 potentials in fist_nonbond_force.F; Include scaling
	  information in neighbor list output.

2010-11-18 18:12  tlaino

	* [r10552] keep track of the onfos in the neighbor lists (Toon -
	  0007)

2010-11-18 16:01  vondele

	* [r10551] merge less frequently, with larger stacks

2010-11-18 15:50  krack

	* [r10550] Quick bug fix: provide a valid MPI communicator

2010-11-18 13:45  tlaino

	* [r10549] Bond correction in Ewald summation based on NL (Toon/Teo
	  - 0006).
	  Note: The bonded correction does no longer use the exclusion
	  lists directly.
	  Therefore it can only be done if there are neighborlists, i.e.
	  when the
	  NONBONDED option in the input is set. This affects four tests
	  with the
	  'NONBONDED F' options is used, but the changes are consistent.
	  There
	  are five other tests with tiny numerical changes. The output of
	  Fist/regtest-5/N3dye_vib_bfgs.inp is rather different because the
	  convergence criteria is met at another point. The values at
	  corresponding iterations are the same.

2010-11-18 13:12  vondele

	* [r10548] more careful_mod's and using it

2010-11-18 05:31  vondele

	* [r10547] declare loop bound shared

2010-11-18 04:18  vondele

	* [r10545] Automatic minor version update to 2.2.45

2010-11-17 17:43  tlaino

	* [r10544] simplify code in NL (Toon-Teo 0005)

2010-11-17 16:52  tlaino

	* [r10543] reduce repetitive code (Toon 0004)

2010-11-17 16:04  tlaino

	* [r10542] Translate exclusion lists to binary scaling factors
	  (Toon).
	  
	  The force evaluation should no longer depend directly on the
	  exclusion
	  lists, but rather take this information from the neighborlists.
	  This
	  change is done in a few steps and it makes easier to access all
	  1-4
	  information in the nonbonded code. Instead of binary exclusion (0
	  or 1)
	  I introduce a scaling factor from 0 to 1. For now these factors
	  are
	  just 0 or 1, but that will change in later patches. Interactions
	  that
	  are scaled to zero are excluded. This information is only stored
	  for
	  1-2, 1-3 and 1-4 interactions. This will be used to take into
	  account
	  the EI_SCALE14 and VDW_SCALE14 options. We may add similar
	  EI_SCALE13
	  etc. options in future.

2010-11-17 15:41  vondele

	* [r10541] count flops correctly

2010-11-17 14:11  tlaino

	* [r10540] fix double counting bug in onfo generator (Toon 0002)

2010-11-17 13:01  tlaino

	* [r10539] Use print_key for MS_RESTART. The mode selective part
	  wrote restart files always to MS_RESTART.
	  This makes the tests very sensitive to the order, probably
	  unintentionally, which is problematic for a regtest script that
	  does all jobs in parallel. (Toon - 0001)

2010-11-17 10:40  tlaino

	* [r10538] prettify

2010-11-17 04:18  vondele

	* [r10536] Automatic minor version update to 2.2.44

2010-11-16 19:26  vondele

	* [r10535] use correct positions for output

2010-11-16 10:50  marcella

	* [r10532] PDOS: it is now possible to specify more lists of atoms
	  to compute more local DOS in the same run. The input has been
	  changed

2010-11-16 10:05  vondele

	* [r10531] add BASIC_SPATIAL_OPTIMIZATION to the mess

2010-11-16 08:35  vondele

	* [r10530] add unit

2010-11-16 04:17  vondele

	* [r10528] Automatic minor version update to 2.2.43

2010-11-15 17:56  tlaino

	* [r10527] bug fix (Carlo)

2010-11-15 17:00  krack

	* [r10526] Automatic fit of the (default) 5th- and 3rd-order
	  polynomials for BUCK4RANGES

2010-11-15 04:17  vondele

	* [r10524] Automatic minor version update to 2.2.42

2010-11-14 04:18  vondele

	* [r10522] Automatic minor version update to 2.2.41

2010-11-13 04:17  vondele

	* [r10520] Automatic minor version update to 2.2.40

2010-11-12 10:03  krack

	* [r10519] Print correct shell/core atom index
	  Print also grand total force

2010-11-12 09:19  krack

	* [r10518] Cleaning (just one kind variable)

2010-11-12 04:18  vondele

	* [r10516] Automatic minor version update to 2.2.39

2010-11-11 16:23  vondele

	* [r10514] no need for a pointer

2010-11-11 15:04  vondele

	* [r10513] fix template

2010-11-11 14:13  vondele

	* [r10511] refactoring

2010-11-11 10:16  tlaino

	* [r10509] fix severe syntax errors !

2010-11-11 09:09  vondele

	* [r10508] bug fix to assign proper cost to bins

2010-11-11 04:18  vondele

	* [r10505] Automatic minor version update to 2.2.38

2010-11-10 04:17  vondele

	* [r10503] Automatic minor version update to 2.2.37

2010-11-09 04:17  vondele

	* [r10501] Automatic minor version update to 2.2.36

2010-11-08 04:18  vondele

	* [r10498] Automatic minor version update to 2.2.35

2010-11-07 04:18  vondele

	* [r10496] Automatic minor version update to 2.2.34

2010-11-06 17:32  krack

	* [r10495] Bug fixes for forces and stress tensor of core-shell
	  model.
	  Finally, analytical and numerical forces are in agreement.

2010-11-06 04:18  vondele

	* [r10492] Automatic minor version update to 2.2.33

2010-11-05 04:18  vondele

	* [r10489] Automatic minor version update to 2.2.32

2010-11-04 20:04  vondele

	* [r10488] no need for pointers here

2010-11-04 04:18  vondele

	* [r10486] Automatic minor version update to 2.2.31

2010-11-03 15:20  vondele

	* [r10484] ugh ?

2010-11-03 15:07  urban

	* [r10483] In finalization, the reshuffle flag now forces
	  reshuffling and compacting of existing data. Changes
	  cp_dbcsr_filter to force reshuffling (and packing) of matrix
	  data.
	  Limits unnecessary memory use when moving matrix data into work
	  data arrays.

2010-11-03 04:18  vondele

	* [r10481] Automatic minor version update to 2.2.30

2010-11-02 11:54  vondele

	* [r10480] workaround for a compiler bug

2010-11-02 04:18  vondele

	* [r10477] Automatic minor version update to 2.2.29

2010-11-01 15:57  jgh

	* [r10476] Fix bug in response basis for unoccupied l states

2010-11-01 10:21  vondele

	* [r10474] workaround to load a statically linked binary on some
	  architectures.

2010-11-01 04:18  vondele

	* [r10472] Automatic minor version update to 2.2.28

2010-10-31 13:01  vondele

	* [r10471] code proper test

2010-10-31 12:38  vondele

	* [r10470] improve timings output, do not allow timer names with
	  spaces

2010-10-31 12:37  vondele

	* [r10469] prettify

2010-10-31 04:18  vondele

	* [r10467] Automatic minor version update to 2.2.27

2010-10-30 16:19  jgh

	* [r10466] Remove duplicated coordinates from QS neighbor list
	  generation.
	  Remove further left overs from old SM type.

2010-10-30 03:18  vondele

	* [r10464] Automatic minor version update to 2.2.26

2010-10-29 16:35  jgh

	* [r10463] Reintroduce old type of screening in ppnl neighbor list

2010-10-29 16:11  jgh

	* [r10462] Remove old sparse matrix type completely.

2010-10-29 15:15  jgh

	* [r10461] Remove not used functions depending on old SM type.

2010-10-29 13:26  jgh

	* [r10460] Adjusted printing of information.

2010-10-29 03:18  vondele

	* [r10458] Automatic minor version update to 2.2.25

2010-10-28 18:25  vondele

	* [r10457] adjust format

2010-10-28 08:01  vondele

	* [r10454] refactoring

2010-10-28 05:38  tlaino

	* [r10453] Some people just like interfaces.. even when there is no
	  real necessity..

2010-10-28 03:17  vondele

	* [r10451] Automatic minor version update to 2.2.24

2010-10-27 19:50  vondele

	* [r10450] no need to be public

2010-10-27 19:17  vondele

	* [r10449] remove some write-only data

2010-10-27 18:57  vondele

	* [r10448] More bytes.... easy enough.

2010-10-27 15:09  vondele

	* [r10447] save a couple of bytes

2010-10-27 11:14  jgh

	* [r10444] Really fix memory leaks.

2010-10-27 10:56  vondele

	* [r10443] Broyden update (Kurt)

2010-10-27 08:06  jgh

	* [r10442] Memory Leaks: deallocate all intermediate arrays

2010-10-27 03:18  vondele

	* [r10439] Automatic minor version update to 2.2.23

2010-10-26 15:33  urban

	* [r10432] Bugfix: Skips already-issued partial multiplications.

2010-10-26 13:05  jgh

	* [r10427] Forgot new routine.

2010-10-26 12:59  jgh

	* [r10426] Final synchronization with qs_refactoring branch.
	  New QS neighbor list code (1 general routine instead of 5 special
	  ones).
	  New QS neighbor list iterator.
	  Removed N^2 search loops and N^2 memory in neighbor list
	  construction
	  and in neighbor list 3-center iterators.
	  Refactoring of overlap integrals (1 general routine replaces 4
	  old routines).
	  Refactoring of qs_core_hamiltonian - now only contains calls to
	  operator routines.
	  Many other small refactoring steps.
	  Reset 130 regtests due to numerics - changed 3 regtests to get
	  better numerical stability.

2010-10-26 08:00  tlaino

	* [r10425] removing unused type

2010-10-26 03:18  vondele

	* [r10423] Automatic minor version update to 2.2.22

2010-10-25 21:33  ibethune

	* [r10422] Loop reordering and OpenMP buffer packing in FFT
	  transpose -> speedup for psmp case

2010-10-25 16:11  jgh

	* [r10421] Back porting the qs_refactoring branch. A series of
	  smaller changes
	  and additions.

2010-10-25 14:13  jgh

	* [r10416] Removing KG options and DKH code (options are still
	  available in Version 2.1 branch);
	  This is the first step of back porting the qs_refactoring branch.

2010-10-25 12:31  vondele

	* [r10415] introduce a SKIP_LOAD_BALANCE_DISTRIBUTED option

2010-10-25 03:17  vondele

	* [r10411] Automatic minor version update to 2.2.21

2010-10-24 03:17  vondele

	* [r10409] Automatic minor version update to 2.2.20

2010-10-23 03:17  vondele

	* [r10407] Automatic minor version update to 2.2.19

2010-10-22 12:04  urban

	* [r10406] Fixes behavior of block multiplication with non-zero
	  product block but zero left/right blocks.

2010-10-22 05:52  tlaino

	* [r10393] dipole damping to the polarizable FF in FIST + regtests
	  (Fist+Rodolphe).
	  Next to come QM/MMpol.

2010-10-22 05:23  tlaino

	* [r10387] repulsion damping + new non-bonded pot (hard to have
	  with the genpot feature) + regtests (Toon)

2010-10-22 04:30  tlaino

	* [r10386] prettify

2010-10-22 03:18  vondele

	* [r10383] Automatic minor version update to 2.2.18

2010-10-21 03:17  vondele

	* [r10381] Automatic minor version update to 2.2.17

2010-10-20 10:56  ibethune

	* [r10376] Re-introduce FFT buffer changes (and OpenMP) with fix
	  for grids with unequal dimensiona

2010-10-20 03:17  vondele

	* [r10374] Automatic minor version update to 2.2.16

2010-10-19 15:12  ibethune

	* [r10373] Backing out last two commits for bug-fix...

2010-10-19 11:31  ibethune

	* [r10372] Added back in OpenMP for xz_to_yz, yz_to_xz

2010-10-19 08:12  ibethune

	* [r10371] Replaced pessimistic buffer allocation in fft_scratch
	  with correct size (OpenMP removed for now)

2010-10-19 03:17  vondele

	* [r10369] Automatic minor version update to 2.2.15

2010-10-18 14:39  jgh

	* [r10368] Output response basis in Quickstep normalization

2010-10-18 03:17  vondele

	* [r10364] Automatic minor version update to 2.2.14

2010-10-17 03:18  vondele

	* [r10355] Automatic minor version update to 2.2.13

2010-10-16 15:37  marcella

	* [r10354] xas scf output

2010-10-16 03:17  vondele

	* [r10352] Automatic minor version update to 2.2.12

2010-10-15 15:19  vondele

	* [r10351] prettify

2010-10-15 14:59  vondele

	* [r10350] refine the element extraction scheme (so CA is C if a
	  psf is in use)

2010-10-15 08:46  vondele

	* [r10349] fix a potential overflow

2010-10-15 03:17  vondele

	* [r10347] Automatic minor version update to 2.2.11

2010-10-14 14:14  hforbert

	* [r10340] Moved helium_update to input_cp2k_restarts, removed
	  restart-related
	  code from PINT and HELIUM, helium restart test added, pint/qs
	  test fixed.
	  [lukasz]

2010-10-14 12:25  vondele

	* [r10339] use the safer (faster) (proper) (?) data size

2010-10-14 03:18  vondele

	* [r10337] Automatic minor version update to 2.2.10

2010-10-13 21:30  urban

	* [r10336] The _put and _get_p routines no longer ignore zero-sized
	  blocks.

2010-10-13 14:48  vondele

	* [r10335] prettify

2010-10-13 13:48  hforbert

	* [r10329] PINT restart bug fixed and 2 tests added, He 'rotation'
	  restoring in
	  presampling added and thus reset 2 tests, some tests/Pimd
	  cleanups.
	  [lukasz]

2010-10-13 06:36  vondele

	* [r10328] clean up the trace facility, introducting TRACE_MAX for
	  pruning the traces on-the-fly

2010-10-13 03:17  vondele

	* [r10326] Automatic minor version update to 2.2.9

2010-10-12 15:20  vondele

	* [r10324] harden grid types

2010-10-12 12:12  vondele

	* [r10323] improved heuristics guaranting a linear scaling
	  allocation for sparse matrices

2010-10-12 11:55  vondele

	* [r10322] prettify

2010-10-12 11:16  vondele

	* [r10321] fix some asserts that should fail but only were
	  warnings.

2010-10-12 11:15  jgh

	* [r10320] Reduced memory usage in neighbor lists construction
	  using subgrids

2010-10-12 10:05  vondele

	* [r10319] fix if condition in dbcsr_add

2010-10-12 09:01  vondele

	* [r10318] make untranspose available as a workaround

2010-10-12 03:17  vondele

	* [r10316] Automatic minor version update to 2.2.8

2010-10-11 14:46  urban

	* [r10315] Sane default for work matrix setup in OpenMP;
	  Ensures work matrix indices always exist in finalize even if
	  empty;
	  Heuristics for size of index that is built on-the-fly in the
	  multiply;
	  More appropriate size query in _add; also OpenMP-parallelizes
	  _add.

2010-10-11 12:11  urban

	* [r10314] Improves heuristic for product data size; fixes OpenMP
	  scope declaration.

2010-10-11 03:17  vondele

	* [r10312] Automatic minor version update to 2.2.7

2010-10-10 08:27  vondele

	* [r10311] improve density/charges output

2010-10-10 08:27  vondele

	* [r10310] fix incorrect format string

2010-10-10 03:17  vondele

	* [r10308] Automatic minor version update to 2.2.6

2010-10-09 18:09  vondele

	* [r10307] mark an overflow

2010-10-09 18:08  vondele

	* [r10306] no need for a pointer here

2010-10-09 07:52  vondele

	* [r10305] remove a couple of timesets again, in too often called
	  functions

2010-10-09 03:17  vondele

	* [r10303] Automatic minor version update to 2.2.5

2010-10-08 20:31  vondele

	* [r10302] add a couple more timeset/timestop

2010-10-08 11:50  urban

	* [r10299] Heuristics to guess the product matrix size in multiply
	  (instead of
	  max. possible).

2010-10-08 11:03  vondele

	* [r10298] duplicate bugs in duplicate code... fix the same loop

2010-10-08 07:11  urban

	* [r10296] Removes debugging WRITE.

2010-10-08 06:38  jgh

	* [r10295] Replace quadratic search loop

2010-10-08 03:17  vondele

	* [r10292] Automatic minor version update to 2.2.4

2010-10-07 20:24  vondele

	* [r10291] some small refactoring of the initial guess routine.

2010-10-07 19:55  urban

	* [r10290] Finalize removes deleted blocks. Filtering supports
	  quick (just removes
	  blocks from index) and slow (also reduces data storage; use the
	  "thorough"
	  argument) filtering.

2010-10-07 06:07  vondele

	* [r10288] use nelectron_total from qs_env

2010-10-07 03:17  vondele

	* [r10286] Automatic minor version update to 2.2.3

2010-10-06 20:17  vondele

	* [r10285] store nelectron_total and nelectron_spin in qs_env

2010-10-06 13:12  vondele

	* [r10284] test

2010-10-06 11:53  vondele

	* [r10281] Automatic minor version update to 2.2.2

2010-10-06 11:40  vondele

	* [r10280] bump the version of cp2k trunk to 2.2

2010-10-06 03:18  vondele

	* [r10278] Automatic minor version update to 2.1.396

2010-10-05 03:19  vondele

	* [r10277] Automatic minor version update to 2.1.395

2010-10-04 03:17  vondele

	* [r10276] Automatic minor version update to 2.1.394

2010-10-03 03:18  vondele

	* [r10275] Automatic minor version update to 2.1.393

2010-10-02 03:17  vondele

	* [r10274] Automatic minor version update to 2.1.392

2010-10-01 07:47  vondele

	* [r10270] facility to optimize parameters in input files

2010-10-01 03:18  vondele

	* [r10269] Automatic minor version update to 2.1.391

2010-09-30 08:33  vweber

	* [r10267] get rid of transposition status from dbcsr get/put and
	  iterator (never used).
	  fix a reshape.

2010-09-30 03:17  vondele

	* [r10266] Automatic minor version update to 2.1.390

2010-09-29 15:07  urban

	* [r10265] Adds assertions that blocks returned from DBCSR routines
	  are not transposed.

2010-09-29 12:26  vweber

	* [r10257] added a performance driver for dbcsr multiply, fixed
	  flop count in multiply.

2010-09-29 03:17  vondele

	* [r10256] Automatic minor version update to 2.1.389

2010-09-28 20:49  urban

	* [r10255] Free speedups for all! Leaves left matrix data
	  "transposed" when it is set to be transposed.

2010-09-28 03:17  vondele

	* [r10254] Automatic minor version update to 2.1.388

2010-09-27 22:26  urban

	* [r10253] Fixes & reduces critical section occurrence for dynamic
	  OpenMP iterators.

2010-09-27 13:38  vweber

	* [r10251] dbcsr_multiply: loop over the local A rows (as before).

2010-09-27 07:33  urban

	* [r10250] Adds declared shared flag to iterator in OpenMP section.

2010-09-27 03:18  vondele

	* [r10249] Automatic minor version update to 2.1.387

2010-09-26 13:17  vondele

	* [r10245] add a comment on a memory bottleneck

2010-09-26 03:17  vondele

	* [r10244] Automatic minor version update to 2.1.386

2010-09-25 07:51  vondele

	* [r10243] some more timeset

2010-09-25 03:17  vondele

	* [r10242] Automatic minor version update to 2.1.385

2010-09-24 15:11  vondele

	* [r10241] small speedup of pw_grid_sort

2010-09-24 13:26  vondele

	* [r10240] skip molecule type info if each atom is its own mol

2010-09-24 13:24  vondele

	* [r10239] add another timeset

2010-09-24 13:11  vondele

	* [r10238] slightly adjust format

2010-09-24 12:08  vondele

	* [r10237] s_sm_pools is nearly dead, but KG rules.

2010-09-24 11:42  vweber

	* [r10234] fixed wrong logic for antisymmetric matrices, added
	  (anti)hermitian types and added tester for dbcsr_add

2010-09-24 10:08  vondele

	* [r10232] Broyden update, improvements, and getting rid of isnan
	  (Kurt Baarman)

2010-09-24 03:17  vondele

	* [r10231] Automatic minor version update to 2.1.384

2010-09-23 20:58  urban

	* [r10230] Removes obvious max 65k block (and hence atom) limit in
	  DBCSR.

2010-09-23 17:26  tlaino

	* [r10229] fix bug

2010-09-23 16:34  tlaino

	* [r10228] restore compilation with NAG

2010-09-23 13:37  tlaino

	* [r10227] prettify

2010-09-23 03:17  vondele

	* [r10226] Automatic minor version update to 2.1.383

2010-09-22 12:11  vweber

	* [r10225] cleanup dbcsr tester

2010-09-22 03:17  vondele

	* [r10224] Automatic minor version update to 2.1.382

2010-09-21 12:24  vweber

	* [r10217] fix lower limits in dbcsr_multiply (block limits), reset
	  few regtests (numerics) and reduced nbr of SCF in ZrF2-PM6.inp

2010-09-21 03:17  vondele

	* [r10216] Automatic minor version update to 2.1.381

2010-09-20 13:39  vondele

	* [r10215] improved broyden minimization with adaptive sigma (Kurt
	  Baarman)

2010-09-20 13:38  urban

	* [r10214] Picky CVS...

2010-09-20 13:06  urban

	* [r10213] Splits dbcsr_scale into separate scale-by-scalar and
	  scale-by-matrix
	  subroutines.
	  Fixes block scaling.
	  Adds rank-2 block printing.

2010-09-20 03:17  vondele

	* [r10212] Automatic minor version update to 2.1.380

2010-09-19 03:18  vondele

	* [r10211] Automatic minor version update to 2.1.379

2010-09-18 03:18  vondele

	* [r10210] Automatic minor version update to 2.1.378

2010-09-17 13:14  vweber

	* [r10203] fix upper limits for the multiply. clean matrix types
	  and transposition status.

2010-09-17 03:18  vondele

	* [r10201] Automatic minor version update to 2.1.377

2010-09-16 10:36  vweber

	* [r10195] fix few bugs in multiply and block access.

2010-09-16 03:17  vondele

	* [r10194] Automatic minor version update to 2.1.376

2010-09-15 09:21  urban

	* [r10192] Avoid unecessary memory copies (and honor the argument)
	  when doing
	  transposes.

2010-09-15 03:17  vondele

	* [r10191] Automatic minor version update to 2.1.375

2010-09-14 08:29  urban

	* [r10188] Changes the default 2-D distribution
	  (BASIC_OPTIMIZATION=TRUE and
	  COST_MODEL=COST_MODEL BLOCK_COUNT in the
	  FORCE/EVAL/DFT/QS/DISTRIBUTION
	  section).
	  Free speedups for everyone!

2010-09-14 03:17  vondele

	* [r10187] Automatic minor version update to 2.1.374

2010-09-13 22:11  marcella

	* [r10185] change default for direct mixing in XAS

2010-09-13 13:58  vondele

	* [r10184] clean

2010-09-13 03:17  vondele

	* [r10183] Automatic minor version update to 2.1.373

2010-09-12 03:17  vondele

	* [r10182] Automatic minor version update to 2.1.372

2010-09-11 03:17  vondele

	* [r10181] Automatic minor version update to 2.1.371

2010-09-10 14:10  vweber

	* [r10180] added a test for dbcsr

2010-09-10 07:50  ibethune

	* [r10179] Fix out-of-bounds issue (psmp only)

2010-09-10 03:17  vondele

	* [r10178] Automatic minor version update to 2.1.370

2010-09-09 13:34  urban

	* [r10177] Another explicit deallocation of thread-private
	  allocatable array.

2010-09-09 11:59  urban

	* [r10176] Adds explicit per-thread data resizing (in multiply).
	  make_images: converts some OMP SINGLE statements to OMP MASTER
	  and adds
	  a necessary critical section; explicitly deallocates
	  thread-private
	  allocatable arrays (fixes gfortran 4.3 OpenMP memory leaks).

2010-09-09 11:41  vondele

	* [r10175] deallocate proper var

2010-09-09 03:17  vondele

	* [r10174] Automatic minor version update to 2.1.369

2010-09-08 11:21  krack

	* [r10173] Write headline to *.stress files

2010-09-08 03:17  vondele

	* [r10172] Automatic minor version update to 2.1.368

2010-09-07 11:32  krack

	* [r10166] * Enable geometry optimisation for core-shell model
	  * 3 new regtests added to check the forces calculated with the
	  core-shell model
	  * Printing of FIST forces updated

2010-09-07 08:39  vweber

	* [r10165] move test matrix generations into libdbcsr

2010-09-07 03:17  vondele

	* [r10164] Automatic minor version update to 2.1.367

2010-09-06 13:28  krack

	* [r10163] * Bug fix for forces and stress tensor of the core-shell
	  model
	  * Forces were wrong since April 2010
	  * Reset of 46 regtests

2010-09-06 11:12  krack

	* [r10162] prettify

2010-09-06 03:18  vondele

	* [r10161] Automatic minor version update to 2.1.366

2010-09-05 03:18  vondele

	* [r10160] Automatic minor version update to 2.1.365

2010-09-04 03:18  vondele

	* [r10159] Automatic minor version update to 2.1.364

2010-09-03 22:54  cjmundy

	* [r10158] Bug fix for Dangchang regtests. Agreement with original
	  code.

2010-09-03 20:27  cjmundy

	* [r10157] Self-consistent Quadrupole capability. Regtests ok.
	  Examples coming.

2010-09-03 12:45  krack

	* [r10155] * Two single-precision constants introducing a numerical
	  noise of 1.0E-6 replaced
	  * Reset of 204 regtests due to numerical noise
	  * Please, check new code (especially basic functions) always for
	  a missing "_dp" and use the constants from the module
	  mathconstants preferibly to avoid such annoying incidents

2010-09-03 10:55  urban

	* [r10154] Factors and C-matrix resizing and allows MATMUL non-BLAS
	  GEMMs.

2010-09-03 03:18  vondele

	* [r10153] Automatic minor version update to 2.1.363

2010-09-02 12:25  vweber

	* [r10150] ...the plasma interface.

2010-09-02 12:21  vweber

	* [r10149] added plasma (optional) for gemms in dbcsr

2010-09-02 03:18  vondele

	* [r10148] Automatic minor version update to 2.1.362

2010-09-01 15:08  krack

	* [r10147] Remove commented debug lines

2010-09-01 08:41  ibethune

	* [r10146] OpenMP parallelisation for data packing and unpacking

2010-09-01 03:18  vondele

	* [r10143] Automatic minor version update to 2.1.361

2010-08-31 19:35  vweber

	* [r10142] fix bug with scp and parallel

2010-08-31 09:27  urban

	* [r10141] * Allows creating a matrix from a given matrix template.
	  * Temporary hack in cp_dbcsr_interface & distribution_methods to
	  help use clusters in multiply, test multiply speed.
	  * Adds basic matrix clustering routine to cp_dbcsr_interface.
	  * Clustering fills cluster2particle map; adds two new cluster
	  methods.
	  * Separates "blocked" row & column block sizes/offsets and
	  distribution from the clustered ones: updates block<->cluster
	  mapping methods and adds routines to convert block sizes/offsets
	  and distribution.
	  * Adds a distribution remap libdbcsr distribution_obj and some
	  get/set metods. Needed to get placement of symmetric matrix
	  blocks on the same process as the clustered block.

2010-08-31 03:17  vondele

	* [r10140] Automatic minor version update to 2.1.360

2010-08-30 14:38  vweber

	* [r10139] add examples for dbcsr

2010-08-30 14:13  krack

	* [r10138] Same output for numerical and analytical stress tensor.
	  Enable direct regtest of differences between numerical and
	  analytical stress tensor.

2010-08-30 03:17  vondele

	* [r10135] Automatic minor version update to 2.1.359

2010-08-29 03:17  vondele

	* [r10134] Automatic minor version update to 2.1.358

2010-08-28 03:17  vondele

	* [r10133] Automatic minor version update to 2.1.357

2010-08-27 16:59  jgh

	* [r10130] New default EPS_EIGENVALUE (10^-12) in atom calculation,
	  reset 90+ regtests

2010-08-27 15:57  jgh

	* [r10129] New definition of vdW radius (now in line with manual),
	  bug fix for E9 LRC,
	  13 regtests reset

2010-08-27 13:05  ibethune

	* [r10128] Remove unnecessary zeroing

2010-08-27 07:00  vondele

	* [r10127] add citation for ADMM

2010-08-27 03:18  vondele

	* [r10126] Automatic minor version update to 2.1.356

2010-08-26 14:21  krack

	* [r10125] Missing _dp added. 4 regtest reset due to numerical
	  noise

2010-08-26 12:25  ibethune

	* [r10124] Fix OpenMP data scoping, remove unnecessary barrier

2010-08-26 03:18  vondele

	* [r10123] Automatic minor version update to 2.1.355

2010-08-25 03:17  vondele

	* [r10122] Automatic minor version update to 2.1.354

2010-08-24 03:18  vondele

	* [r10121] Automatic minor version update to 2.1.353

2010-08-23 15:12  tlaino

	* [r10120] angle defined between 0 and Pi

2010-08-23 03:18  vondele

	* [r10119] Automatic minor version update to 2.1.352

2010-08-22 03:18  vondele

	* [r10118] Automatic minor version update to 2.1.351

2010-08-21 03:17  vondele

	* [r10117] Automatic minor version update to 2.1.350

2010-08-20 09:40  krack

	* [r10116] Cell file headline corrected

2010-08-20 03:18  vondele

	* [r10115] Automatic minor version update to 2.1.349

2010-08-19 17:46  vweber

	* [r10114] posibility to cluster atoms with small block on heavier
	  ones.

2010-08-19 06:36  vondele

	* [r10113] allow for initializing the parser with a set of known
	  initial variables

2010-08-19 03:18  vondele

	* [r10112] Automatic minor version update to 2.1.348

2010-08-18 14:52  urban

	* [r10111] Shortens routine names to .LE. 31.

2010-08-18 08:09  jgh

	* [r10109] Simplify setup of DBCSR matrix

2010-08-18 03:18  vondele

	* [r10108] Automatic minor version update to 2.1.347

2010-08-17 18:43  vweber

	* [r10107] bug fix

2010-08-17 03:18  vondele

	* [r10102] Automatic minor version update to 2.1.346

2010-08-16 15:41  vondele

	* [r10098] prettify

2010-08-16 07:20  urban

	* [r10097] Fixes missing OpenMP USE and silences chatty warning.

2010-08-16 03:18  vondele

	* [r10096] Automatic minor version update to 2.1.345

2010-08-15 03:18  vondele

	* [r10095] Automatic minor version update to 2.1.344

2010-08-14 03:17  vondele

	* [r10094] Automatic minor version update to 2.1.343

2010-08-13 15:30  tlaino

	* [r10093] prettify

2010-08-13 11:08  urban

	* [r10090] Preparation for clustering:
	  Moves CP DBCSR type methods into a separate cp_dbcsr_methods
	  module.
	  Adds cp_dbcsr_xlat module for the heavy code using for mappings
	  and translations and cluster_buffers for dealing with the
	  buffering.
	  Buffers as preparation for clustering:
	  * expands buffer capabalities (pointer to data backing instead of
	  implied single master);
	  * allows querying the appropriate data type for an array.
	  Adds buffer flushing to routines that need it.
	  Adds mapping subtypes and some utility functions to lookup
	  cluster<->buffer maps.

2010-08-13 03:17  vondele

	* [r10089] Automatic minor version update to 2.1.342

2010-08-12 14:28  hforbert

	* [r10088] Added possibility to extend number of He environments if
	  more cpus
	  become available after a restart (lukasz).

2010-08-12 09:17  vweber

	* [r10077] added clustering hydrogen on heavier atoms, fixed
	  problems while clustering 0d, 1d and 2d periodic systems and
	  added regs

2010-08-12 03:17  vondele

	* [r10076] Automatic minor version update to 2.1.341

2010-08-11 17:13  tkuehne

	* [r10075] DFTB Stress Tensor (4 regtests are reseatted) + minor
	  typos

2010-08-11 09:36  vweber

	* [r10074] fixed ssmp compilation from previous commit

2010-08-11 08:15  vweber

	* [r10073] improved interface cp/dbcsr, cleanup code (avoiding
	  direct access to dbcsr data directly). and prepared for
	  clustering

2010-08-11 03:18  vondele

	* [r10072] Automatic minor version update to 2.1.340

2010-08-10 13:48  urban

	* [r10071] Update the cluster pricing model.

2010-08-10 12:41  urban

	* [r10069] Corrects sorting order and block indexing for the new
	  basic 2-D
	  distribution optimization.

2010-08-10 12:07  krack

	* [r10068] DOI corrected

2010-08-10 03:18  vondele

	* [r10063] Automatic minor version update to 2.1.339

2010-08-09 11:33  vweber

	* [r10059] cleanup dbcsr interface, split cp_dbcsr_operations
	  module into interface, types and operations.

2010-08-09 08:06  vweber

	* [r10055] cleanup

2010-08-09 03:17  vondele

	* [r10054] Automatic minor version update to 2.1.338

2010-08-08 03:17  vondele

	* [r10053] Automatic minor version update to 2.1.337

2010-08-07 03:18  vondele

	* [r10052] Automatic minor version update to 2.1.336

2010-08-06 14:38  urban

	* [r10048] Uses the basic 2-D distribution generator when
	  clustering is used.

2010-08-06 12:52  urban

	* [r10045] An O(Nlog(N)) 2-D distribution generator
	  (BASIC_OPTIMIZATION in the
	  DFT%QS%DISTRIBUTION section).

2010-08-06 12:50  urban

	* [r10042] Adds a minimum-heap data structure.

2010-08-06 09:31  vweber

	* [r10041] clustering

2010-08-06 03:17  vondele

	* [r10040] Automatic minor version update to 2.1.335

2010-08-05 03:18  vondele

	* [r10039] Automatic minor version update to 2.1.334

2010-08-04 13:28  urban

	* [r10037] Corrects matrix addition in the multiplication
	  (triggered by serial sm+fmfm^T).
	  Also fixes a bug with pointer assignment to transposed blocks
	  (not used).

2010-08-04 03:18  vondele

	* [r10035] Automatic minor version update to 2.1.333

2010-08-03 14:20  krack

	* [r10034] Provide proper headlines for the relevant SCF matrices

2010-08-03 03:18  vondele

	* [r10033] Automatic minor version update to 2.1.332

2010-08-02 07:49  vondele

	* [r10032] Fix intent (Jason).

2010-08-02 03:17  vondele

	* [r10031] Automatic minor version update to 2.1.331

2010-08-01 03:18  vondele

	* [r10030] Automatic minor version update to 2.1.330

2010-07-31 03:18  vondele

	* [r10029] Automatic minor version update to 2.1.329

2010-07-30 10:55  jgh

	* [r10027] Bug fix DFTB dispersion energy, excludes now self term.
	  Mostly constant shift
	  of energy, 12 regtests reset.

2010-07-30 03:18  vondele

	* [r10026] Automatic minor version update to 2.1.328

2010-07-29 03:17  vondele

	* [r10019] Automatic minor version update to 2.1.327

2010-07-28 03:17  vondele

	* [r10017] Automatic minor version update to 2.1.326

2010-07-27 14:01  urban

	* [r10015] * Cleans up dbcsr_finalize. Usually appends new data
	  into existing matrix (resulting in less data copying). More
	  OpenMP parallelization and removes nested parallel sections.
	  * New add_wm_from_matrix, including limiting data to move into
	  work matrix and simplifies use cases.
	  * A few minor changes to support the dbcsr_finalize changes.
	  * Removes a few unnecessary routines.

2010-07-27 07:56  urban

	* [r10011] Shifts the counting of DBCSR image matrices. Resets
	  changed regtests.

2010-07-27 03:18  vondele

	* [r10010] Automatic minor version update to 2.1.325

2010-07-26 13:36  ibethune

	* [r10007] Reintroduce OpenMP changes with out-of-bounds fix

2010-07-26 03:18  vondele

	* [r10006] Automatic minor version update to 2.1.324

2010-07-25 03:18  vondele

	* [r10005] Automatic minor version update to 2.1.323

2010-07-24 03:18  vondele

	* [r10001] Automatic minor version update to 2.1.322

2010-07-23 14:29  vondele

	* [r10000] fix a bug that causes jumps in the potential, affecting
	  runs with cell_ref and changing cells.

2010-07-23 03:17  vondele

	* [r9999] Automatic minor version update to 2.1.321

2010-07-22 03:18  vondele

	* [r9998] Automatic minor version update to 2.1.320

2010-07-21 09:21  tlaino

	* [r9995] prettify

2010-07-21 03:18  vondele

	* [r9994] Automatic minor version update to 2.1.319

2010-07-20 08:33  jgh

	* [r9990] DFT-D3: more control over reference coordination numbers,
	  change of keyword for atoms and new keyword for kinds
	  (ATOM_COORDINATION_NUMBERS,KIND_COORDINATION_NUMBERS),
	  1 new regtest.

2010-07-20 03:18  vondele

	* [r9989] Automatic minor version update to 2.1.318

2010-07-19 13:35  marcella

	* [r9985] extension of gle thermostat for quantum dynamics and path
	  integral
	  r-path variable using RMSD
	  pint restart

2010-07-19 03:18  vondele

	* [r9984] Automatic minor version update to 2.1.317

2010-07-18 03:18  vondele

	* [r9983] Automatic minor version update to 2.1.316

2010-07-17 03:18  vondele

	* [r9982] Automatic minor version update to 2.1.315

2010-07-16 03:18  vondele

	* [r9981] Automatic minor version update to 2.1.314

2010-07-15 03:18  vondele

	* [r9977] Automatic minor version update to 2.1.313

2010-07-14 15:21  hforbert

	* [r9976] Reseting regtest due to changed default helium sampling
	  scheme + io bugfix

2010-07-14 03:18  vondele

	* [r9975] Automatic minor version update to 2.1.312

2010-07-13 13:49  mguidon

	* [r9974] Bug fix memory leak

2010-07-13 13:43  krack

	* [r9973] prettify

2010-07-13 13:41  krack

	* [r9972] bug fix: define natom in the right place (no clue why the
	  automatic regtester did not catch this)

2010-07-13 10:43  mguidon

	* [r9970] - Fixes openMP force calculation (no stress tensor if not
	  requested)
	  - Optimizes atomic pair list build for systems containing > 1000
	  atoms

2010-07-13 10:35  jgh

	* [r9969] Update of DFT-D3 code: tapering function and long range
	  correction,
	  new option to have constant c9 values for E9 energy, new debug
	  option

2010-07-13 03:17  vondele

	* [r9968] Automatic minor version update to 2.1.311

2010-07-12 13:51  mguidon

	* [r9967] revert previous commit (out of bounds issues)

2010-07-12 12:09  krack

	* [r9966] Correct typo

2010-07-12 12:03  hforbert

	* [r9965] W-number calc on multiple procs fixed, direct sfluid
	  density calc added, m distribution modifications added. (lukasz)

2010-07-12 03:18  vondele

	* [r9963] Automatic minor version update to 2.1.310

2010-07-11 20:27  krack

	* [r9962] * Buggy module lowdin removed
	  * New module population_analyses added including a rewrite of the
	  lowdin module
	  * Bug fix for DBCSR sparse matrix printing
	  * More detailed (optional) output for Mulliken and Lowdin added
	  * Docs updated and corrected accordingly

2010-07-11 03:18  vondele

	* [r9961] Automatic minor version update to 2.1.309

2010-07-10 03:17  vondele

	* [r9960] Automatic minor version update to 2.1.308

2010-07-09 14:01  ibethune

	* [r9959] Use new task_list ordering for OpenMP parallelisation of
	  calculate_rho_elec(slightly faster), also
	  fully collapse 3d loop in rs_grid_zero

2010-07-09 08:37  ibethune

	* [r9958] Another fix - use the correct number of threads in all
	  cases

2010-07-09 07:14  vondele

	* [r9957] proper (?) fix, all threads reaching the same barrier.

2010-07-09 05:59  vondele

	* [r9956] fix a deadlock

2010-07-09 03:17  vondele

	* [r9955] Automatic minor version update to 2.1.307

2010-07-08 19:51  vondele

	* [r9954] small omp speedup

2010-07-08 14:43  mguidon

	* [r9953] Remove useless global debug variable

2010-07-08 12:39  mguidon

	* [r9952] Adds print key for detalied load balance information and
	  possibillity to
	  randomize the distribution of bins among processes

2010-07-08 12:36  ibethune

	* [r9951] Removing another OpenMP num_threads clause

2010-07-08 09:11  krack

	* [r9950] Remove unused pointer variable

2010-07-08 07:37  krack

	* [r9949] prettify

2010-07-08 03:18  vondele

	* [r9948] Automatic minor version update to 2.1.306

2010-07-07 14:43  ibethune

	* [r9947] Replacing OpenMP num_threads() clauses with manual thread
	  id checks to work around GOMP bug 44833 -
	  http://gcc.gnu.org/bugzilla/show_bug.cgi?id=44833

2010-07-07 03:17  vondele

	* [r9946] Automatic minor version update to 2.1.305

2010-07-06 03:18  vondele

	* [r9945] Automatic minor version update to 2.1.304

2010-07-05 15:26  vondele

	* [r9944] test

2010-07-05 14:47  vondele

	* [r9942] test

2010-07-05 11:13  vondele

	* [r9941] test

2010-07-05 03:17  vondele

	* [r9940] Automatic minor version update to 2.1.303

2010-07-04 03:17  vondele

	* [r9939] Automatic minor version update to 2.1.302

2010-07-03 03:17  vondele

	* [r9938] Automatic minor version update to 2.1.301

2010-07-02 10:02  vondele

	* [r9937] Added Broyden minimizer with OT (Kurt Baarman)

2010-07-02 09:15  vondele

	* [r9936] fix comment typo

2010-07-02 03:17  vondele

	* [r9935] Automatic minor version update to 2.1.300

2010-07-01 03:17  vondele

	* [r9934] Automatic minor version update to 2.1.299

2010-06-30 07:44  vondele

	* [r9933] Automatic minor version update to 2.1.298

2010-06-28 03:17  vondele

	* [r9932] Automatic minor version update to 2.1.295

2010-06-27 03:17  vondele

	* [r9931] Automatic minor version update to 2.1.294

2010-06-26 03:17  vondele

	* [r9930] Automatic minor version update to 2.1.293

2010-06-25 13:59  mguidon

	* [r9929] Bug fix uninitialized variable

2010-06-25 03:17  vondele

	* [r9928] Automatic minor version update to 2.1.292

2010-06-24 23:27  marcella

	* [r9927] bug fix: variable not inizialized

2010-06-24 21:18  vondele

	* [r9926] add USE_HISTORY to the localize section. Generates an
	  initial rotation of the current orbitals based on a history
	  collected during MD. Seems to yield a modest speedup in the
	  number of iterations needed.

2010-06-24 09:56  urban

	* [r9925] Fixes to dbcsr_verify and dbcsr_checksum.

2010-06-24 08:02  urban

	* [r9924] Proactive bugfix: correct handling of transposed blocks
	  in replicate.

2010-06-24 07:25  urban

	* [r9923] Fixes prettify issue.

2010-06-24 07:06  vondele

	* [r9922] restore original correct comment

2010-06-24 03:17  vondele

	* [r9921] Automatic minor version update to 2.1.291

2010-06-23 20:29  vondele

	* [r9920] prettify

2010-06-23 19:59  vondele

	* [r9919] clarify comment

2010-06-23 14:19  vondele

	* [r9918] spline2 and spline3 also work with blocked pw_grids

2010-06-23 03:17  vondele

	* [r9917] Automatic minor version update to 2.1.290

2010-06-22 10:27  urban

	* [r9916] Removes now uneeded files...

2010-06-22 09:09  urban

	* [r9915] 1. Adds (indirect) block buffers for accessing data
	  blocks
	  (dbcsr_block_buffers.F).
	  * By default there is no change. Buffers are not used unless
	  explicitly enabled.
	  * By using buffers it is now possible to avoid dubious pointer
	  hacks
	  with compilers that do not support pointer rank remapping
	  * CP2K pointers can be indirect (i.e., not point directly into
	  matrix
	  storage: allows for other storage types)
	  2. Moves dbcsr_data_obj and dbcsr_scalar_type methods into
	  separate
	  dbcsr_data_methods.F files. Also adds some helper subroutines
	  (reallocation, ...).
	  3. Updates block buffer support in iterator and block access
	  routines.
	  Iterator now also supports a read-only specification (no buffer
	  flushing).
	  4. Flushing buffers added where needed
	  5. Unifies pointer rank remapping and places it in
	  dbcsr_ptr_util.F.
	  6. Updates a few block operations (1D-2D data copy/transposes).
	  7. Cleans up make dense/undense data

2010-06-22 03:17  vondele

	* [r9914] Automatic minor version update to 2.1.289

2010-06-21 03:17  vondele

	* [r9912] Automatic minor version update to 2.1.288

2010-06-20 03:17  vondele

	* [r9911] Automatic minor version update to 2.1.287

2010-06-19 03:17  vondele

	* [r9910] Automatic minor version update to 2.1.286

2010-06-18 03:17  vondele

	* [r9909] Automatic minor version update to 2.1.285

2010-06-17 03:17  vondele

	* [r9908] Automatic minor version update to 2.1.284

2010-06-16 13:29  jgh

	* [r9906] [CP2K:2757]: Stabilize electron count if CORE_CORRECTION
	  is used (INT->NINT)

2010-06-16 12:13  mguidon

	* [r9905] Fix naming convention according to standard (found with
	  ifort)

2010-06-16 03:17  vondele

	* [r9904] Automatic minor version update to 2.1.283

2010-06-15 19:26  tlaino

	* [r9903] tightening checks

2010-06-15 19:25  tlaino

	* [r9902] bug fix for recursive connectivities

2010-06-15 18:17  tlaino

	* [r9901] fix PDB output for large numbers

2010-06-15 14:49  ibethune

	* [r9900] More OpenMP data scoping fixes (found with Pathscale
	  compiler)

2010-06-15 12:09  mguidon

	* [r9899] Bug fix ray tracer for large supercells

2010-06-15 03:17  vondele

	* [r9898] Automatic minor version update to 2.1.282

2010-06-14 15:12  urban

	* [r9897] Honors declared OpenMP data-sharing declarations.

2010-06-14 09:04  urban

	* [r9895] Persistent thread distribution & more flexible iterator.
	  Thread distribution remains persistent and is based on a matrix's
	  row
	  sizes (better for load balancing).
	  The iterator can now dynamically give whole rows (or blocks) to
	  threads
	  or continue to use the thread distribution as before.

2010-06-14 08:56  urban

	* [r9894] Iterators called by only one thread but that go through
	  the whole matrix
	  alone are set to be not shared.

2010-06-14 08:48  urban

	* [r9893] Can initialize the MPI environment with MPI_Init_Thread.
	  Instead of MPI_Init, the thread-aware MPI_Init_Thread can be
	  called when
	  compiled with OpenMP. However, this option is turned off for the
	  sake of
	  compatibility.

2010-06-14 03:17  vondele

	* [r9892] Automatic minor version update to 2.1.281

2010-06-13 03:17  vondele

	* [r9891] Automatic minor version update to 2.1.280

2010-06-12 03:17  vondele

	* [r9890] Automatic minor version update to 2.1.279

2010-06-11 14:59  vondele

	* [r9889] add link for an input-section-generating applet for GLE

2010-06-11 13:30  urban

	* [r9888] Adds routine to sort the index of a matrix.

2010-06-11 11:46  urban

	* [r9887] Adds utility routines to count/build the row_p index.

2010-06-11 03:17  vondele

	* [r9886] Automatic minor version update to 2.1.278

2010-06-10 11:00  ibethune

	* [r9885] OpenMP for PBE functional evaluation

2010-06-10 03:17  vondele

	* [r9884] Automatic minor version update to 2.1.277

2010-06-09 16:00  urban

	* [r9883] Disallows making matrices dense only by rows or columns.

2010-06-09 03:17  vondele

	* [r9882] Automatic minor version update to 2.1.276

2010-06-08 10:15  urban

	* [r9881] Fixes memory leak; cleanup.

2010-06-08 08:35  urban

	* [r9880] dbcsr_new_transposed: size handling and redistribution
	  specification.

2010-06-08 03:17  vondele

	* [r9879] Automatic minor version update to 2.1.275

2010-06-07 15:23  urban

	* [r9878] Removes unnecessary counting.

2010-06-07 10:13  urban

	* [r9877] More cleanups of symmetric distribution.

2010-06-07 03:17  vondele

	* [r9875] Automatic minor version update to 2.1.274

2010-06-06 03:17  vondele

	* [r9874] Automatic minor version update to 2.1.273

2010-06-05 03:17  vondele

	* [r9873] Automatic minor version update to 2.1.272

2010-06-04 13:29  jgh

	* [r9872] DFT-D3: IMPORTANT BUG FIX: sign of forces was
	  wrong!DFT-D3: IMPORTANT BUG FIX: sign of forces was wrong!DFT-D3:
	  IMPORTANT BUG FIX: sign of forces was wrong!

2010-06-04 13:02  mguidon

	* [r9870] Adds possibility to define a list of BASIS_SET_FILES

2010-06-04 07:54  urban

	* [r9869] Bugfix.

2010-06-04 05:49  urban

	* [r9868] Allow replication of data from only one process.

2010-06-04 03:17  vondele

	* [r9867] Automatic minor version update to 2.1.271

2010-06-03 14:55  mguidon

	* [r9866] Bug fix

2010-06-03 09:21  urban

	* [r9865] Prettify (from several recent commits).

2010-06-03 09:16  urban

	* [r9864] Better accounting for matrix data size.
	  Use BLAS for some larger memory copies.

2010-06-03 03:17  vondele

	* [r9863] Automatic minor version update to 2.1.270

2010-06-02 14:17  jgh

	* [r9862] DFT-D3 correct some typos in parameters (BLYP)

2010-06-02 13:45  mguidon

	* [r9861] Implements analytical stress tensor for HFX and ADMM

2010-06-02 07:20  jgh

	* [r9860] Bug-bug-fix of debug code

2010-06-02 03:17  vondele

	* [r9859] Automatic minor version update to 2.1.269

2010-06-01 16:59  jgh

	* [r9858] Bug fix for debug code in parallel runs.

2010-06-01 12:36  urban

	* [r9857] Completing yesterday's bugfix...

2010-06-01 03:17  vondele

	* [r9856] Automatic minor version update to 2.1.268

2010-05-31 14:24  urban

	* [r9855] Proactive bugfix: correctly determine transposed blocks
	  when using any
	  distribution.

2010-05-31 03:17  vondele

	* [r9854] Automatic minor version update to 2.1.267

2010-05-30 03:17  vondele

	* [r9853] Automatic minor version update to 2.1.266

2010-05-29 03:18  vondele

	* [r9850] Automatic minor version update to 2.1.265

2010-05-28 10:18  krack

	* [r9847] * Standard atomic weights updated to IUPAC 2007
	  (reference given in module periodic_table)
	  * Old weights are still available using the compile flag
	  -D__USE_LEGACY_WEIGHTS (source of these weights is not specified)
	  * Ground state electronic configuration of Ce and Tc corrected
	  * Unused component mass removed from ptable (weight of most
	  abundant isotope)
	  * Elements up to 118 added
	  * Reset of more than 300 regtests

2010-05-28 09:05  urban

	* [r9846] Proactive bugfix: index merging was reset block pointer.
	  New dbcsr_insert_blocks replaces existing blocks or adds new ones
	  to an
	  existing matrix;
	  New mp_recv and mp_send for encapsulated data types and logical
	  renaming
	  of some arguments in mp_irecv;
	  New dbcsr_add_reserved for inserting blocks into existing matrix
	  without
	  using work matrices;
	  New dbcsr_reduce to perform global, row, or column reduction of a
	  matrix;
	  Proactive bugfix: zeroes correct array when resizing.

2010-05-26 11:14  jgh

	* [r9845] Implementation of DFT-D3 (Grimme et al, JCP 132 154104
	  (2010)).
	  C6, C8, C9 terms, analytic gradients, PBC and stress tensor.
	  12 new regtests, 4 old regtests reset.
	  DFT-D3 parameter file dftd3.dat in ../tests/QS
	  WARNING: new keywords DFTD2 (old GRIMME) and DFTD3, DFTD3 is
	  default.
	  WARNING: old inputs will no longer work or use different
	  defaults.
	  WARNING: bug fix for interaction radius (was effectively only
	  half of the
	  promised radius), the default value has been changed.
	  WARNING: new keyword REFERENCE_FUNCTIONAL is used to easily
	  access the
	  functional dependent parameters, no more defaults available.

2010-05-26 03:17  vondele

	* [r9844] Automatic minor version update to 2.1.262

2010-05-25 17:57  tlaino

	* [r9842] bug fix (Juerg). reset 5 tests.

2010-05-25 17:56  tlaino

	* [r9841] prettify

2010-05-25 03:17  vondele

	* [r9840] Automatic minor version update to 2.1.261

2010-05-24 03:17  vondele

	* [r9839] Automatic minor version update to 2.1.260

2010-05-23 03:17  vondele

	* [r9838] Automatic minor version update to 2.1.259

2010-05-22 03:17  vondele

	* [r9837] Automatic minor version update to 2.1.258

2010-05-21 03:17  vondele

	* [r9836] Automatic minor version update to 2.1.257

2010-05-20 13:27  urban

	* [r9835] Removes unnecessary and non-standard VOLATILE attribute.

2010-05-20 03:17  vondele

	* [r9834] Automatic minor version update to 2.1.256

2010-05-19 03:17  vondele

	* [r9833] Automatic minor version update to 2.1.255

2010-05-18 03:17  vondele

	* [r9832] Automatic minor version update to 2.1.254

2010-05-17 03:17  vondele

	* [r9831] Automatic minor version update to 2.1.253

2010-05-16 03:17  vondele

	* [r9830] Automatic minor version update to 2.1.252

2010-05-15 05:42  vondele

	* [r9829] fix a bug where a touch EXIT would not lead to
	  termination.

2010-05-15 03:17  vondele

	* [r9828] Automatic minor version update to 2.1.251

2010-05-14 10:40  mmcgrath

	* [r9827] Formatting change and minor bug fix

2010-05-14 03:17  vondele

	* [r9826] Automatic minor version update to 2.1.250

2010-05-13 03:17  vondele

	* [r9825] Automatic minor version update to 2.1.249

2010-05-12 06:21  tlaino

	* [r9824] fix division by zero

2010-05-12 03:17  vondele

	* [r9823] Automatic minor version update to 2.1.248

2010-05-11 03:17  vondele

	* [r9822] Automatic minor version update to 2.1.247

2010-05-10 03:17  vondele

	* [r9821] Automatic minor version update to 2.1.246

2010-05-09 17:40  tlaino

	* [r9820] bug fixes for MULTIPOLE EWALDS for GAUSSIAN charges +
	  cleaning (Toon). Reset few tests due to numerics (10^-14)

2010-05-09 03:18  vondele

	* [r9819] Automatic minor version update to 2.1.245

2010-05-08 03:17  vondele

	* [r9818] Automatic minor version update to 2.1.244

2010-05-07 16:59  tlaino

	* [r9817] Add the possibility to generate exclusion lists not based
	  on the topology information rather on
	  an automatic way (Toon+Louis). Added regtests. Bug Fix (Teo)

2010-05-07 13:47  tlaino

	* [r9816] prettify

2010-05-07 03:17  vondele

	* [r9815] Automatic minor version update to 2.1.243

2010-05-06 03:17  vondele

	* [r9814] Automatic minor version update to 2.1.242

2010-05-05 11:23  marcella

	* [r9811] XAS + mixing;
	  get rid of hartree_2c and hartree_3c variables;
	  normalization of the reaction path CV

2010-05-05 10:26  krack

	* [r9810] * Print virial
	  * Default geometrical basis set (Clementi) for H enlarged
	  * Cosmetics (printing of exponents, identation)

2010-05-05 09:28  krack

	* [r9809] More detailed headline

2010-05-05 03:17  vondele

	* [r9808] Automatic minor version update to 2.1.241

2010-05-04 03:17  vondele

	* [r9807] Automatic minor version update to 2.1.240

2010-05-03 03:17  vondele

	* [r9806] Automatic minor version update to 2.1.239

2010-05-02 03:17  vondele

	* [r9805] Automatic minor version update to 2.1.238

2010-05-01 03:17  vondele

	* [r9804] Automatic minor version update to 2.1.237

2010-04-30 03:17  vondele

	* [r9803] Automatic minor version update to 2.1.236

2010-04-29 13:16  ibethune

	* [r9801] Added OpenMP parallelisation to integrate_v_rspace, with
	  supporting changes to
	  the task_list structure.

2010-04-29 03:17  vondele

	* [r9800] Automatic minor version update to 2.1.235

2010-04-28 03:17  vondele

	* [r9798] Automatic minor version update to 2.1.234

2010-04-27 03:17  vondele

	* [r9796] Automatic minor version update to 2.1.233

2010-04-26 15:54  ibethune

	* [r9795] Bugfix: OpenMP with non-orthorhombic cells summed rs
	  grids incorrectly in
	  collocate kernels.

2010-04-26 10:47  tlaino

	* [r9794] fix printing (Toon)

2010-04-26 07:59  urban

	* [r9793] Saner default (by default iterations should not honor
	  thread distribution
	  in non-threaded code sections).

2010-04-26 03:17  vondele

	* [r9792] Automatic minor version update to 2.1.232

2010-04-25 03:17  vondele

	* [r9791] Automatic minor version update to 2.1.231

2010-04-24 11:31  marcella

	* [r9790] bug fix

2010-04-24 03:17  vondele

	* [r9789] Automatic minor version update to 2.1.230

2010-04-23 20:07  vondele

	* [r9788] cp_fm_syevd: redistribute small matrices transparently to
	  smaller number of tasks as suggested by an empirical rule based
	  on timings made on rosa.

2010-04-23 03:18  vondele

	* [r9787] Automatic minor version update to 2.1.229

2010-04-22 03:17  vondele

	* [r9786] Automatic minor version update to 2.1.228

2010-04-21 16:46  urban

	* [r9785] Adds dbcsr_copy_submatrix to allow getting a
	  (block-aligned) submatrix of
	  a matrix. Allows shallow and deep copies.
	  More efficient dbcsr_reserve_blocks; allows pre-existing blocks
	  to be
	  listed (they are ignored) and data pointers to be explicitly
	  stated.

2010-04-21 15:16  vondele

	* [r9784] change the Hessian output of BFGS to use
	  cp_fm_write_unformatted for improved efficiency. This changes in
	  a non-backward compatible form the format of the .Hessian files.

2010-04-21 11:50  urban

	* [r9783] Prettify.

2010-04-21 11:46  urban

	* [r9782] Bugfix: had cleaned too many variables from
	  dbcsr_replicate.
	  Preemptive bugfix (assigning array to uninitialized pointer).

2010-04-21 04:47  vondele

	* [r9781] remove unused buffer

2010-04-21 03:17  vondele

	* [r9780] Automatic minor version update to 2.1.227

2010-04-20 22:00  tlaino

	* [r9779] bug fixes for exclusion lists + new bunch of regtests
	  (Toon)

2010-04-20 19:17  vondele

	* [r9778] added timing output to geo_opt

2010-04-20 19:15  vondele

	* [r9777] prettify

2010-04-20 03:17  vondele

	* [r9776] Automatic minor version update to 2.1.226

2010-04-19 13:34  vondele

	* [r9775] The frequency of checking EXIT files is now reduced, only
	  check every 20 seconds. This avoids slowsdowns for fast runs on
	  parallel file systems.

2010-04-19 10:31  fschiff

	* [r9774] fixed wrong usage description

2010-04-19 03:17  vondele

	* [r9773] Automatic minor version update to 2.1.225

2010-04-18 03:17  vondele

	* [r9772] Automatic minor version update to 2.1.224

2010-04-17 03:17  vondele

	* [r9771] Automatic minor version update to 2.1.223

2010-04-16 09:21  urban

	* [r9770] Makes matrix replication faster (now uses allgather),
	  support all data
	  types, and support just row/column replication.

2010-04-16 03:17  vondele

	* [r9769] Automatic minor version update to 2.1.222

2010-04-15 03:17  vondele

	* [r9768] Automatic minor version update to 2.1.221

2010-04-14 12:27  urban

	* [r9767] Prettify.

2010-04-14 03:18  vondele

	* [r9766] Automatic minor version update to 2.1.220

2010-04-13 08:50  urban

	* [r9765] Bulk adding of zeroed blocks into the matrix (bypassing
	  work matrices,
	  dbcsr_reserve_blocks in dbcsr_block_access).
	  Iterator seeking to arbitrary row (dbcsr_iterator_seek); also
	  some
	  iterator cleanups.

2010-04-13 03:17  vondele

	* [r9764] Automatic minor version update to 2.1.219

2010-04-12 18:39  vondele

	* [r9763] remove bogus write

2010-04-12 12:08  vweber

	* [r9762] bug fix with keep_sparsity=F

2010-04-12 08:15  vondele

	* [r9761] fix format

2010-04-12 03:17  vondele

	* [r9760] Automatic minor version update to 2.1.218

2010-04-11 03:17  vondele

	* [r9759] Automatic minor version update to 2.1.217

2010-04-10 03:17  vondele

	* [r9758] Automatic minor version update to 2.1.216

2010-04-09 14:50  vondele

	* [r9756] port of CP2K to Windows using mingw/gfortran (FX Coudert)

2010-04-09 11:53  vweber

	* [r9755] added a simple Lanczos routine to compute extremal
	  eigenvalues and a cheap approximate to the 2-norm. avoid printing
	  0 entry in _print and
	  fix a few bugs.

2010-04-09 09:13  tlaino

	* [r9753] short range part of electrostatic handled analytically
	  (i.e. removed from the spline): this, unfortunately, is causing a
	  slight overhead in the evaluation of the pair potential, but it
	  was necessary to handle different exclusion lists for
	  electrostatic and van der waals : this option was preferred to
	  the one of having two separate NL. Reser most of FIST regtests
	  (error for energy evaluation less than 1.0E-5%.. maybe larger in
	  some cases due to propagation). (Toon and Teo).

2010-04-09 08:43  vweber

	* [r9752] bug fix: _put_block with summation didnt handle propertly
	  transposed block

2010-04-09 03:17  vondele

	* [r9751] Automatic minor version update to 2.1.215

2010-04-08 19:18  vondele

	* [r9749] fix format

2010-04-08 18:11  vondele

	* [r9748] fix citation format

2010-04-08 18:01  vondele

	* [r9747] add one more citation for DDAPC

2010-04-08 03:17  vondele

	* [r9746] Automatic minor version update to 2.1.214

2010-04-07 11:05  vweber

	* [r9745] fix minor bug introduiced with previous commit.

2010-04-07 03:17  vondele

	* [r9742] Automatic minor version update to 2.1.213

2010-04-06 19:54  vondele

	* [r9741] more parallel version of cp_fm_init_random. Resets all
	  tests that rely on this (e.g. OT for initial guess)

2010-04-06 15:58  mguidon

	* [r9740] Fix memory leak

2010-04-06 15:26  vweber

	* [r9739] bug fix for copy of replicated matrices, multiply of sub
	  matrices and added a _init_random.

2010-04-06 13:20  mguidon

	* [r9737] Changes th algorithm for determining the maximum number
	  of neighbors cells needed
	  in a periodic TC HFX calculation

2010-04-06 09:58  vondele

	* [r9735] fix format

2010-04-06 03:17  vondele

	* [r9733] Automatic minor version update to 2.1.212

2010-04-05 03:17  vondele

	* [r9732] Automatic minor version update to 2.1.211

2010-04-04 03:17  vondele

	* [r9731] Automatic minor version update to 2.1.210

2010-04-03 08:51  vondele

	* [r9730] fix a bug with farming and TC

2010-04-03 06:08  marcella

	* [r9728] some cleaning and restructuring of the XAS code to allow
	  the use of mixing and smearing

2010-04-03 03:17  vondele

	* [r9727] Automatic minor version update to 2.1.209

2010-04-02 11:50  vondele

	* [r9726] all for using an OUTPUT_FILE_NAME when farming

2010-04-02 03:17  vondele

	* [r9725] Automatic minor version update to 2.1.208

2010-04-01 03:17  vondele

	* [r9724] Automatic minor version update to 2.1.207

2010-03-31 03:17  vondele

	* [r9723] Automatic minor version update to 2.1.206

2010-03-30 18:07  vondele

	* [r9722] fix farming output to fort.999

2010-03-30 14:31  vweber

	* [r9721] reset eps_matrix_filter to 0 as default.

2010-03-30 08:33  vweber

	* [r9720] removed unused variables.

2010-03-30 03:17  vondele

	* [r9719] Automatic minor version update to 2.1.205

2010-03-29 10:15  vweber

	* [r9718] possibility to do some operations localy for replicated
	  matrices. reset 3 regs from previous commit.

2010-03-29 03:17  vondele

	* [r9717] Automatic minor version update to 2.1.204

2010-03-28 03:17  vondele

	* [r9716] Automatic minor version update to 2.1.203

2010-03-27 15:57  tlaino

	* [r9715] bug fix in printing energy components in parallel (bug
	  reported by Chris)

2010-03-27 04:17  vondele

	* [r9714] Automatic minor version update to 2.1.202

2010-03-26 04:17  vondele

	* [r9713] Automatic minor version update to 2.1.201

2010-03-25 16:46  vweber

	* [r9712] turn on matrix thresholding for S and H.

2010-03-25 16:14  vondele

	* [r9710] machine files for the Cray compiler environment (Roberto
	  Ansaloni)

2010-03-25 15:48  urban

	* [r9709] Absolute real kinds in c_mpi_calls instead of relative
	  sp/dp.

2010-03-25 14:31  urban

	* [r9708] The kind should be applied to a real and not an integer.

2010-03-25 14:27  urban

	* [r9707] Cleanups (removes many unused variables in templates and
	  makes some type
	  conversions explicit).

2010-03-25 08:41  urban

	* [r9706] Limits matrix size dbcsr_finalize. Does a data array
	  resize/copy if
	  the data area is twice as big as required data size instead of
	  avoiding
	  copying data whenever possible.

2010-03-25 08:17  urban

	* [r9705] A few small fixes (Joost).

2010-03-25 04:17  vondele

	* [r9704] Automatic minor version update to 2.1.200

2010-03-24 22:23  tlaino

	* [r9703] remove another wild write..

2010-03-24 22:14  tlaino

	* [r9702] debug print?! (removed..)

2010-03-24 22:14  tlaino

	* [r9701] missing initialization: several tests failing because of
	  a print statement introduced with last juerg commit
	  (uninitialized variable)

2010-03-24 19:26  vondele

	* [r9700] init vars, likely not needed

2010-03-24 19:17  vondele

	* [r9699] fix undef var

2010-03-24 19:13  vondele

	* [r9698] undefined var fix

2010-03-24 18:42  tlaino

	* [r9697] bug fixes (Joost: gfortran -O0/-O2 were giving different
	  results. No other compilers so far had shown this issue)

2010-03-24 18:21  urban

	* [r9696] Fixes memory leak.

2010-03-24 17:13  urban

	* [r9695] Adds missing interfaces for all data types.
	  Moves some subroutines that work on blocks from _operations to
	  _block_operations.

2010-03-24 16:20  vondele

	* [r9694] add an cp_unimplemented for tau functionals and finer xc
	  grids

2010-03-24 15:29  jgh

	* [r9692] Uninitialized variable (Cray compiler/Joost)

2010-03-24 14:30  jgh

	* [r9691] revert back erronous change

2010-03-24 14:25  vondele

	* [r9690] another undefined fix

2010-03-24 14:21  vondele

	* [r9689] more undefined accesses

2010-03-24 14:09  jgh

	* [r9688] Extension of GTH pseudopotential definition, not active
	  in Quickstep code

2010-03-24 14:08  jgh

	* [r9687] Extension of GTH pseudopotential definition, not active
	  in Quickstep code

2010-03-24 13:04  tlaino

	* [r9686] prettify

2010-03-24 10:16  vondele

	* [r9684] recommit the fix for the intel issue

2010-03-24 09:51  ibethune

	* [r9683] BUGFIX: Uninitialised variable / Cray Compiler

2010-03-24 04:17  vondele

	* [r9682] Automatic minor version update to 2.1.199

2010-03-23 22:18  marcella

	* [r9681] uninitialized varialbes, Cray compiler/Joost

2010-03-23 16:03  vondele

	* [r9679] fix OMP bugs

2010-03-23 15:54  vondele

	* [r9678] fix undefined stuff

2010-03-23 15:18  urban

	* [r9677] Bugfixes (Joost/CRAY compiler).

2010-03-23 14:57  urban

	* [r9676] Makes add_block_node work with work_create.
	  Clearer assertions.

2010-03-23 13:42  tlaino

	* [r9675] remove useless code

2010-03-23 12:36  jgh

	* [r9674] Bug fixes (uninitialized varialbes, Cray compiler/Joost)

2010-03-23 12:03  tlaino

	* [r9673] undefined variables (from Joost's mail checked:
	  metadynamics.F, md_run.F and cp_lbfgs.F)

2010-03-23 11:18  tlaino

	* [r9672] prettify

2010-03-23 10:19  vweber

	* [r9671] undef vars

2010-03-23 04:17  vondele

	* [r9670] Automatic minor version update to 2.1.198

2010-03-22 19:56  vondele

	* [r9669] bug fix a swap for inverted frequency range

2010-03-22 16:19  urban

	* [r9668] Better dbcsr_complete_redistribute (FM<->DBCSR).

2010-03-22 04:17  vondele

	* [r9667] Automatic minor version update to 2.1.197

2010-03-21 20:29  tlaino

	* [r9666] clean-up (Toon)

2010-03-21 04:17  vondele

	* [r9665] Automatic minor version update to 2.1.196

2010-03-20 04:17  vondele

	* [r9664] Automatic minor version update to 2.1.195

2010-03-20 04:12  tlaino

	* [r9663] add doxygen comment (Toon)

2010-03-20 04:09  tlaino

	* [r9662] FIST: Guassian charges, dipoles, quadrupoles + regtests
	  (Toon)

2010-03-19 23:02  tlaino

	* [r9661] fix typo

2010-03-19 22:22  tlaino

	* [r9660] prettify

2010-03-19 16:38  urban

	* [r9659] OpenMP bugfix: Restricts parallel environment.

2010-03-19 09:23  urban

	* [r9658] Adds a thread-prepared assertion to dbcsr_add_block_node.

2010-03-19 04:17  vondele

	* [r9657] Automatic minor version update to 2.1.194

2010-03-18 21:52  urban

	* [r9656] Adds range verification to multi-threaded block access.

2010-03-18 13:51  urban

	* [r9655] Removes an assertion that is too sensitive.

2010-03-18 11:05  urban

	* [r9654] Make the code compilable.

2010-03-18 09:12  urban

	* [r9653] Explicitly sets work matrix size.

2010-03-18 08:08  urban

	* [r9652] Reduces memory usage in dbcsr_make_images and adds OpenMP
	  parallelization.
	  Includes some small changes in other subroutines in support of
	  this.

2010-03-18 04:17  vondele

	* [r9651] Automatic minor version update to 2.1.193

2010-03-17 22:16  urban

	* [r9650] Limits the data size of the finalized matrix to actual
	  element count.

2010-03-17 09:52  vweber

	* [r9649] remove printing

2010-03-17 04:17  vondele

	* [r9648] Automatic minor version update to 2.1.192

2010-03-16 22:57  urban

	* [r9647] Restores OpenMP code to local multiply.
	  More thread safety (dbcsr_error, dbcsr_iterator, dbcsr_finalize,
	  ...)
	  and threading bugfixes. Some dbcsr_finalize cleanups.

2010-03-16 10:16  urban

	* [r9646] Prevents pointer data copies when not using Fortran rank
	  remapping (fixes
	  issue with Intel).

2010-03-16 04:17  vondele

	* [r9645] Automatic minor version update to 2.1.191

2010-03-15 12:55  vondele

	* [r9644] demo

2010-03-15 04:17  vondele

	* [r9643] Automatic minor version update to 2.1.190

2010-03-14 04:17  vondele

	* [r9639] Automatic minor version update to 2.1.189

2010-03-13 23:18  marcella

	* [r9637] well tempered metadynamics [Sebastiano Caravati]

2010-03-13 04:17  vondele

	* [r9636] Automatic minor version update to 2.1.188

2010-03-12 15:25  ibethune

	* [r9635] Improved schedule clause

2010-03-12 14:39  tlaino

	* [r9634] prettify

2010-03-12 10:30  urban

	* [r9633] Updates prettify script to accout for files included
	  outside of a
	  subroutine body.
	  Accordingly, the prettify_fooler subroutines are now gone!

2010-03-12 09:03  urban

	* [r9632] Removes unneded subroutine calls.

2010-03-12 04:17  vondele

	* [r9631] Automatic minor version update to 2.1.187

2010-03-11 16:01  vweber

	* [r9630] call routines in cp_fm_diag instead of scalapack ones.

2010-03-11 04:17  vondele

	* [r9629] Automatic minor version update to 2.1.186

2010-03-10 12:51  vondele

	* [r9628] cleanup of eigensolver libtest

2010-03-10 04:17  vondele

	* [r9627] Automatic minor version update to 2.1.185

2010-03-09 14:47  ibethune

	* [r9626] Bugfix : OMP schedule clause can evaluate to zero in some
	  cases

2010-03-09 07:53  urban

	* [r9624] Properly initializes some variables that were previously
	  unused.

2010-03-09 04:17  vondele

	* [r9623] Automatic minor version update to 2.1.184

2010-03-08 18:56  tlaino

	* [r9622] split ELP potential type into a KG and FIST. The KG
	  potential must be therefore
	  defined as KG and not anymore as ELP (Toon).

2010-03-08 15:51  urban

	* [r9621] Bugfix: correctly use complex type.

2010-03-08 14:14  urban

	* [r9618] Cleanup: adds wrapper for isend/irecv for DBCSR data.

2010-03-08 14:06  urban

	* [r9617] Fixes incorrect usage of MPI_BOTTOM (thanks to F. Parpia,
	  IBM).

2010-03-08 13:30  urban

	* [r9616] Nicer check.

2010-03-08 13:04  tlaino

	* [r9615] prettify

2010-03-08 12:23  urban

	* [r9614] Removes trigger of unitialized variable.

2010-03-08 11:26  urban

	* [r9613] * More efficient matrix multiplication with transposed
	  matrices (avoids
	  one alltoall, avoids in-memory block transposes); includes
	  regtest resets.
	  * Makes matrices dense by rows or columns (or both, as before)
	  when
	  product matrix can not be made dense.
	  * More specific specification of matrix type (symmetry is
	  decoupled from
	  the transformation of the mirrored block).
	  * Fixes conversion between canonical <-> triangular storage of
	  symmetric
	  index.
	  * Adds "implication" logical assertion.

2010-03-08 04:17  vondele

	* [r9612] Automatic minor version update to 2.1.183

2010-03-07 04:17  vondele

	* [r9611] Automatic minor version update to 2.1.182

2010-03-06 04:17  vondele

	* [r9610] Automatic minor version update to 2.1.181

2010-03-05 15:51  fschiff

	* [r9609] got rid of two multiplies

2010-03-05 04:17  vondele

	* [r9608] Automatic minor version update to 2.1.180

2010-03-04 04:17  vondele

	* [r9607] Automatic minor version update to 2.1.179

2010-03-03 19:35  tlaino

	* [r9606] bug fix (see [CP2K:2597])

2010-03-03 12:46  jgh

	* [r9603] Bug-bug fix LSD GGA atom code

2010-03-03 11:27  vweber

	* [r9602] libdbcsr is now completly independent from the rest of
	  the code

2010-03-03 04:17  vondele

	* [r9601] Automatic minor version update to 2.1.178

2010-03-02 17:59  vweber

	* [r9600] fix bug with empty matrix

2010-03-02 17:33  tlaino

	* [r9599] bug fix

2010-03-02 16:43  vweber

	* [r9598] make dense matrices that are almost full, more error
	  propagation.

2010-03-02 15:30  urban

	* [r9597] Some fixes for transposed blocks and transposed
	  operations.

2010-03-02 13:43  urban

	* [r9596] Removes mp_minloc, mp_maxloc for complex types.
	  Adds mp_sum and mp_bcast interface for logical scalars.

2010-03-02 12:08  vweber

	* [r9595] added timer for multiply (shows the transpose/matrix
	  shape/symmetry), got rid of the last dependencies in libdbcsr.

2010-03-02 11:14  krack

	* [r9594] print reference also for LSD runs

2010-03-02 10:32  tlaino

	* [r9592] conjugate gradients for POL_SCF in FIST (Toon): next
	  patch will add regtests + reorganization of present ones.. reset
	  due to numerics (10^-9)

2010-03-02 08:56  krack

	* [r9591] fix doc

2010-03-02 04:17  vondele

	* [r9590] Automatic minor version update to 2.1.177

2010-03-01 19:01  tlaino

	* [r9589] bug fix + regtests (Toon)

2010-03-01 15:57  vweber

	* [r9588] fix compilation of smp from previous commit.

2010-03-01 14:34  vweber

	* [r9587] replace old routine to create product distribution,
	  removed unused variables.

2010-03-01 04:17  vondele

	* [r9586] Automatic minor version update to 2.1.176

2010-02-28 04:17  vondele

	* [r9585] Automatic minor version update to 2.1.175

2010-02-27 04:17  vondele

	* [r9584] Automatic minor version update to 2.1.174

2010-02-26 16:15  jgh

	* [r9583] Bufix LSD GGA in atom code, reset regtests

2010-02-26 04:17  vondele

	* [r9581] Automatic minor version update to 2.1.173

2010-02-26 00:10  tlaino

	* [r9580] prettify

2010-02-25 11:05  vweber

	* [r9579] prettify

2010-02-25 10:10  fschiff

	* [r9577] wfn_mix can overwrite the MOs if required (no need to
	  restart for RTP necessary anylonger)

2010-02-25 09:57  vweber

	* [r9576] added filteringfor ot irac + 1 regtest

2010-02-25 04:17  vondele

	* [r9575] Automatic minor version update to 2.1.172

2010-02-24 04:17  vondele

	* [r9574] Automatic minor version update to 2.1.171

2010-02-23 09:51  urban

	* [r9572] More refactoring (multiplication-related). Removes
	  uneeded templates for
	  dbcsr_util.

2010-02-23 09:21  vweber

	* [r9571] use the mos in the dbcsr format for ot

2010-02-23 04:17  vondele

	* [r9570] Automatic minor version update to 2.1.170

2010-02-22 16:15  urban

	* [r9569] Limits too-broad USE statement.

2010-02-22 15:53  urban

	* [r9568] Reinstates DBCSR-related commits from Friday (row/column
	  subcommunicators
	  and product/right matrix incongruity).

2010-02-22 15:43  fschiff

	* [r9567] bug fix restarting RTP, simplification for writing
	  restart history

2010-02-22 15:37  urban

	* [r9566] Refactors DBCSR modules.

2010-02-22 04:17  vondele

	* [r9565] Automatic minor version update to 2.1.169

2010-02-21 04:17  vondele

	* [r9564] Automatic minor version update to 2.1.168

2010-02-20 04:17  vondele

	* [r9563] Automatic minor version update to 2.1.167

2010-02-19 14:43  tlaino

	* [r9562] new two colvars: XYZ_DIAG and XYZ_OUTERDIAG + 2 regtests.

2010-02-19 11:07  urban

	* [r9561] Prettify.

2010-02-19 10:28  urban

	* [r9560] Supports multiplication when product and right matrices
	  have different
	  number of blocked columns.

2010-02-19 09:50  urban

	* [r9559] Makes subcommunicator use optional (workaround for issue
	  with Intel MPI).
	  Default is now not to use subcommunicators.

2010-02-19 09:36  marcella

	* [r9558] update GAPW with gspace mixing: Broyden mixing

2010-02-19 09:13  urban

	* [r9557] Bugfix: Nullify pointer before use.

2010-02-19 07:00  tlaino

	* [r9556] bug fix

2010-02-19 04:17  vondele

	* [r9555] Automatic minor version update to 2.1.166

2010-02-18 16:41  tlaino

	* [r9554] prettify

2010-02-18 13:57  hforbert

	* [r9553] added: PINT & HELIUM restarts from external file
	  (EXT_RESTART),
	  HELIUM%CELL_SIZE input keyword, output of helium forces
	  (instantaneous
	  and averaged), output of helium coordinates from all
	  processors/environments,
	  output of pint_env COM;
	  minor bug fix: mpi logger files not created any more by pint_env
	  updated: comments in src/parallel_rng_types.F

2010-02-18 04:17  vondele

	* [r9551] Automatic minor version update to 2.1.165

2010-02-17 18:48  vweber

	* [r9550] convert mo_derivs to dbcsr

2010-02-17 15:23  urban

	* [r9549] Bugfix: postpones too-early increment.

2010-02-17 13:00  urban

	* [r9545] Less overhead in local multiply.

2010-02-17 11:51  mguidon

	* [r9543] Add missing branch for blocked purification

2010-02-17 10:18  tlaino

	* [r9542] fix doc

2010-02-17 04:17  vondele

	* [r9541] Automatic minor version update to 2.1.164

2010-02-16 14:09  mguidon

	* [r9540] Implements two new energies for denisty matrix blocked
	  ADMM

2010-02-16 11:04  ibethune

	* [r9538] OpenMP parallelisation for QS task collocation

2010-02-16 04:17  vondele

	* [r9537] Automatic minor version update to 2.1.163

2010-02-15 16:12  vweber

	* [r9536] cleaning and more error propagation.

2010-02-15 14:46  urban

	* [r9535] Bugfix: honor sparsity retainment in fm->dbcsr
	  conversion.

2010-02-15 11:08  urban

	* [r9534] Combines data types for MPI transfers in DBCSR multiply.
	  Selectable in
	  code, default is off (i.e., no change).

2010-02-15 07:46  vondele

	* [r9533] small speedup for gfortran

2010-02-15 04:17  vondele

	* [r9532] Automatic minor version update to 2.1.162

2010-02-14 04:17  vondele

	* [r9531] Automatic minor version update to 2.1.161

2010-02-13 13:46  marcella

	* [r9529] First step for GAPW with gspace mixing: kerker_mixing

2010-02-13 04:17  vondele

	* [r9528] Automatic minor version update to 2.1.160

2010-02-12 16:10  vweber

	* [r9527] initating the divorce between libdbcsr and cp_error.
	  removed optional error arguments between cp2k and libdbcsr.
	  added error handling for libdbcsr.

2010-02-12 11:45  fschiff

	* [r9525] improve wfn_mix, marks on molecular states can be set
	  and states for wfnmix can be choosen according to these marks

2010-02-12 04:17  vondele

	* [r9524] Automatic minor version update to 2.1.159

2010-02-11 22:31  tlaino

	* [r9523] bug fix (Csilla)

2010-02-11 14:05  fschiff

	* [r9521] second time new write (sorry for the first breaking sopt)

2010-02-11 13:22  vondele

	* [r9520] add a flush after a finished farming job

2010-02-11 10:43  tlaino

	* [r9519] adding units

2010-02-11 04:17  vondele

	* [r9518] Automatic minor version update to 2.1.158

2010-02-10 17:28  tlaino

	* [r9517] revert back patch (Florian): does not even compile!

2010-02-10 16:49  fschiff

	* [r9516] faster writing of restart WFN (least invasiv version)

2010-02-10 12:01  mguidon

	* [r9515] Implements more ADMM energies

2010-02-10 04:17  vondele

	* [r9513] Automatic minor version update to 2.1.157

2010-02-09 16:30  vweber

	* [r9511] added interface cp2k/libdbcsr with timing, get rid of the
	  corresponding timings in libdbcsr

2010-02-09 16:03  urban

	* [r9510] Avoids testing processor subsets.

2010-02-09 15:41  urban

	* [r9509] Adds libtest for DBCSR testing and benchmarking.
	  * Includes forgetten OBJECTDEFS and fix for g95 leaky exports.

2010-02-09 15:39  urban

	* [r9508] Adds libtest for DBCSR testing and benchmarking.

2010-02-09 11:43  jgh

	* [r9506] atomic code: unrestricted KS und HF methods
	  atomic code: pseudopotential optimization with LSD reference
	  states

2010-02-09 04:17  vondele

	* [r9505] Automatic minor version update to 2.1.156

2010-02-08 08:30  urban

	* [r9504] * Uses process row and process column communicators in
	  matrix multiply.
	  Also used by new hybrid_alltoall, used to transfer data mostly
	  row-wise
	  and column-wise. Helps for single-interface switched networks.
	  * Reuses dbcsr_mp_env in some FM<->DBCSR conversions

2010-02-08 04:17  vondele

	* [r9503] Automatic minor version update to 2.1.155

2010-02-07 04:17  vondele

	* [r9502] Automatic minor version update to 2.1.154

2010-02-06 20:42  urban

	* [r9501] Bugfix: use of uninitialized variable.

2010-02-06 12:14  vondele

	* [r9500] example code for using craypat sampling on specific
	  routines

2010-02-06 04:17  vondele

	* [r9499] Automatic minor version update to 2.1.153

2010-02-05 15:43  mguidon

	* [r9498] Implements some more algorithms, includes a correction to
	  the Kohn-Sham
	  matrix eigenvalues

2010-02-05 10:32  vweber

	* [r9497] rm useless copy fm/dbcsr and added extra timer

2010-02-05 04:17  vondele

	* [r9496] Automatic minor version update to 2.1.152

2010-02-04 17:52  mguidon

	* [r9491] Changes wavefunction fitting into auxiliary density
	  matrix method (ADMM).
	  Adds some keyword for algorithms/methods that still need to be
	  implemented.
	  Adds one method for calculating the Kohn-Sham matrix

2010-02-04 16:18  marcella

	* [r9490] restart file for localization

2010-02-04 12:27  krack

	* [r9488] Adjust format for larger numbers

2010-02-04 04:17  vondele

	* [r9487] Automatic minor version update to 2.1.151

2010-02-03 21:28  vondele

	* [r9486] some more istat checking and a small speedup

2010-02-03 21:18  tlaino

	* [r9485] prettify

2010-02-03 19:59  vondele

	* [r9484] fix buglet in output

2010-02-03 14:30  urban

	* [r9482] Bugfix: local index was incorrectly created from "dense"
	  matrix.

2010-02-03 12:47  urban

	* [r9481] Removes spurious warnings when ghost atoms are used.

2010-02-03 11:42  vweber

	* [r9480] fm->dbcsr (1st step), reset few regs.

2010-02-03 10:29  jgh

	* [r9479] Bug fixes for pseudopotential optimization

2010-02-03 10:28  jgh

	* [r9478] Hybrid functionals using Geminals

2010-02-03 09:28  fschiff

	* [r9477] removed debug write

2010-02-03 09:18  vondele

	* [r9476] check status before deallocate

2010-02-03 08:25  vondele

	* [r9475] jacobi_rot_para: significantly improve scalability, fix
	  some bugs for large numbers of mpi tasks.

2010-02-03 04:17  vondele

	* [r9474] Automatic minor version update to 2.1.150

2010-02-02 12:58  fschiff

	* [r9472] changed defaults, possibility to avoid self consistency
	  check in the first n steps

2010-02-02 04:17  vondele

	* [r9471] Automatic minor version update to 2.1.149

2010-02-01 04:17  vondele

	* [r9470] Automatic minor version update to 2.1.148

2010-01-31 04:17  vondele

	* [r9469] Automatic minor version update to 2.1.147

2010-01-30 07:55  vondele

	* [r9468] check regtester

2010-01-30 04:17  vondele

	* [r9467] Automatic minor version update to 2.1.146

2010-01-29 14:10  mguidon

	* [r9466] dummy commit to keep the regtester busy

2010-01-29 12:50  mguidon

	* [r9465] Some bug fixes and cosmetics:
	  - wrong scalapack block size
	  - B3LYP adds now the correct correction functionals for
	  wfn_fitting
	  - exchange-correlation energy in output contains now the
	  correction term

2010-01-29 10:16  vweber

	* [r9464] ... need to rm that file.

2010-01-29 09:56  vweber

	* [r9463] added a mpi interface to libdbcsr (2nd step)

2010-01-29 04:17  vondele

	* [r9462] Automatic minor version update to 2.1.145

2010-01-28 19:20  tlaino

	* [r9461] bug fix (Toon)

2010-01-28 12:09  urban

	* [r9460] * Enables local conversion to dense matrices (bigger
	  block sizes) for
	  multiplication.
	  * Pre-sizes destination matrix in dbcsr_add.

2010-01-28 04:17  vondele

	* [r9459] Automatic minor version update to 2.1.144

2010-01-27 10:26  vweber

	* [r9457] created a library for the dbcsr modules (1st step)

2010-01-27 08:59  vweber

	* [r9456] removed unused dbcsr procedures.

2010-01-27 04:17  vondele

	* [r9455] Automatic minor version update to 2.1.143

2010-01-26 15:24  vweber

	* [r9454] removed distribution_2d from dbcsr_type.

2010-01-26 15:17  vondele

	* [r9453] fix a use of undefined variables

2010-01-26 11:22  urban

	* [r9452] More correct.

2010-01-26 09:59  marcella

	* [r9450] doxygen info

2010-01-26 09:22  urban

	* [r9449] Fix out-of-bounds write access during creation of
	  BLACS-based distribution.

2010-01-26 08:25  tlaino

	* [r9448] added missing doxygen info with fakes templates: to be
	  fixed.
	  clean debug comments: debug instructions should be contained
	  between if (debug_this_module) ... endif.
	  Since they are comments they are just to be considered as trash.

2010-01-26 04:17  vondele

	* [r9447] Automatic minor version update to 2.1.142

2010-01-25 17:27  marcella

	* [r9446] reaction path colvar using positions of selected atoms as
	  parameters

2010-01-25 14:56  vweber

	* [r9444] added a kind module for dbcsr and got rid of stop_program
	  in dbcsr*F

2010-01-25 09:20  vondele

	* [r9443] fix a line too long for the title format

2010-01-25 04:17  vondele

	* [r9442] Automatic minor version update to 2.1.141

2010-01-24 04:17  vondele

	* [r9441] Automatic minor version update to 2.1.140

2010-01-23 04:17  vondele

	* [r9440] Automatic minor version update to 2.1.139

2010-01-22 16:26  vweber

	* [r9439] bug fix (*gemm)

2010-01-22 09:00  vweber

	* [r9438] fix few bugs. changed interface for few procedures
	  (improved mixed precision capabilities).

2010-01-22 04:17  vondele

	* [r9437] Automatic minor version update to 2.1.138

2010-01-21 16:27  fschiff

	* [r9435] bug fix for multiple envelops (Hanning Chen)

2010-01-21 12:57  tlaino

	* [r9434] keeping argument character length *

2010-01-21 12:23  fschiff

	* [r9433] bug fix memory leak

2010-01-21 10:08  fschiff

	* [r9430] wfn_mix with localized lumos and molecular states

2010-01-21 06:27  tlaino

	* [r9429] prettify

2010-01-21 04:17  vondele

	* [r9428] Automatic minor version update to 2.1.137

2010-01-20 13:27  marcella

	* [r9427] bug fix

2010-01-20 13:19  urban

	* [r9426] Speed improvements:
	  * Option to retain redistributed matrix generated for a multiply
	  (useful
	  for static matrices that are reused in several multiplications,
	  but on
	  account of approx. doubling the storage size for that matrix).
	  * For efficiency, the routine to issue DGEMMs in the local part
	  of the
	  matrix multiply is now a template, generated for each data type,
	  and a
	  few other tweaks.

2010-01-20 13:13  marcella

	* [r9424] new colvar: distance from the reaction path

2010-01-20 12:20  fschiff

	* [r9423] bug fix (KEYWORD NAME and printing of lumos)

2010-01-20 04:17  vondele

	* [r9422] Automatic minor version update to 2.1.136

2010-01-19 20:59  tlaino

	* [r9421] bug fix

2010-01-19 15:34  fschiff

	* [r9420] removed debug write

2010-01-19 09:20  tlaino

	* [r9419] allowing the possibility to combine hills from
	  one-dimensional mtd or to do
	  bias-exchange mtd (Toon). Reset few..

2010-01-19 09:18  tlaino

	* [r9418] prettify

2010-01-19 09:05  fschiff

	* [r9416] restructuring of qs_scf_post_gpw.F, added localization
	  for unoccupied states,
	  tiny fix for occupied states, only important if smearing is
	  active or localization
	  for a energy range is requested

2010-01-19 08:49  tlaino

	* [r9415] adding wall energy to the total energy (Toon)

2010-01-19 06:45  tlaino

	* [r9414] when NT is negative don't add hills anymore (allows fixed
	  bias potential) (Toon)

2010-01-19 04:17  vondele

	* [r9413] Automatic minor version update to 2.1.135

2010-01-19 00:55  tlaino

	* [r9412] cv_energu => epot_s (Toon)

2010-01-19 00:54  tlaino

	* [r9411] improve doc (Too)

2010-01-19 00:29  tlaino

	* [r9410] added an ignore missing critical parameters for FIST
	  (Toon)+regtest

2010-01-18 15:52  tlaino

	* [r9409] computing the hills_env energy and adding to the total
	  energy. Removing unused
	  columns in COLVAR file for non-lagrangian MTD. Added at the end
	  of the line the
	  hills energy (Toon). Resetting all MTD regtests.

2010-01-18 15:26  tlaino

	* [r9408] changing SPAWNED_HILLS_AMPL into SPAWNED_HILLS_SCALE
	  (Toon)

2010-01-18 04:17  vondele

	* [r9407] Automatic minor version update to 2.1.134

2010-01-17 04:17  vondele

	* [r9406] Automatic minor version update to 2.1.133

2010-01-16 04:17  vondele

	* [r9405] Automatic minor version update to 2.1.132

2010-01-15 04:17  vondele

	* [r9403] Automatic minor version update to 2.1.131

2010-01-14 04:17  vondele

	* [r9402] Automatic minor version update to 2.1.130

2010-01-13 17:28  mguidon

	* [r9401] Fixes a bug in force calculation for wavefunction
	  fitting. Previously went
	  wrong if the sparsity pattern of all involved overlap matrices
	  was not
	  identical (i.e. 100 %).

2010-01-13 04:17  vondele

	* [r9400] Automatic minor version update to 2.1.129

2010-01-12 04:17  vondele

	* [r9399] Automatic minor version update to 2.1.128

2010-01-11 14:45  mguidon

	* [r9397] Fixes an issue with the task list optimization. The
	  default behaviour of
	  generate_qs_task_list optimizes the rank order and overwrites the
	  rs descriptor.
	  This leads to problems, if the routine is called several times
	  with different
	  tasks lists as it is done in wave-function fitting.

2010-01-11 14:11  jgh

	* [r9396] Pseudopotential optimization, small improvements

2010-01-11 04:17  vondele

	* [r9395] Automatic minor version update to 2.1.127

2010-01-10 04:17  vondele

	* [r9394] Automatic minor version update to 2.1.126

2010-01-09 04:17  vondele

	* [r9393] Automatic minor version update to 2.1.125

2010-01-08 15:10  urban

	* [r9392] More support for more data types.

2010-01-08 04:17  vondele

	* [r9391] Automatic minor version update to 2.1.124

2010-01-07 18:09  tlaino

	* [r9390] FUES potential (Toon)

2010-01-07 12:42  urban

	* [r9389] Submatrix copy.

2010-01-07 04:17  vondele

	* [r9388] Automatic minor version update to 2.1.123

2010-01-06 04:17  vondele

	* [r9387] Automatic minor version update to 2.1.122

2010-01-05 15:49  vweber

	* [r9386] extended some procedure to complex and fix few bugs.

2010-01-05 09:58  vondele

	* [r9385] dbcsr_nn_mult_lin: remove timeset/timestop from an often
	  called routine

2010-01-05 04:17  vondele

	* [r9382] Automatic minor version update to 2.1.121

2010-01-04 19:27  tlaino

	* [r9381] prettify

2010-01-04 17:48  jgh

	* [r9380] Add new control options for pseudopotential optimizer

2010-01-04 17:46  jgh

	* [r9379] New option for atom calculation (confinement control)

2010-01-04 17:45  jgh

	* [r9378] Add some more flexibility to CIF file reader

2010-01-04 14:59  vweber

	* [r9377] bug fix

2010-01-04 12:16  vweber

	* [r9376] fix bugs with complex data type and added new routines

2010-01-04 11:17  marcella

	* [r9375] clean dbg line

2010-01-04 09:47  tlaino

	* [r9374] more updates

2010-01-04 04:17  vondele

	* [r9373] Automatic minor version update to 2.1.120

2010-01-03 23:41  tlaino

	* [r9372] update for 2010

2010-01-03 22:46  marcella

	* [r9368] rmsd colvar

2010-01-03 04:17  vondele

	* [r9367] Automatic minor version update to 2.1.119

2010-01-02 04:17  vondele

	* [r9366] Automatic minor version update to 2.1.118

2010-01-01 04:17  vondele

	* [r9365] Automatic minor version update to 2.1.117

2009-12-31 04:17  vondele

	* [r9364] Automatic minor version update to 2.1.116

2009-12-30 04:17  vondele

	* [r9363] Automatic minor version update to 2.1.115

2009-12-29 04:17  vondele

	* [r9362] Automatic minor version update to 2.1.114

2009-12-28 04:17  vondele

	* [r9361] Automatic minor version update to 2.1.113

2009-12-27 04:18  vondele

	* [r9360] Automatic minor version update to 2.1.112

2009-12-26 04:17  vondele

	* [r9359] Automatic minor version update to 2.1.111

2009-12-25 04:17  vondele

	* [r9358] Automatic minor version update to 2.1.110

2009-12-24 04:17  vondele

	* [r9357] Automatic minor version update to 2.1.109

2009-12-23 04:17  vondele

	* [r9356] Automatic minor version update to 2.1.108

2009-12-22 11:28  marcella

	* [r9355] bug fix

2009-12-22 04:17  vondele

	* [r9354] Automatic minor version update to 2.1.107

2009-12-21 04:17  vondele

	* [r9353] Automatic minor version update to 2.1.106

2009-12-20 04:17  vondele

	* [r9352] Automatic minor version update to 2.1.105

2009-12-19 04:17  vondele

	* [r9351] Automatic minor version update to 2.1.104

2009-12-18 17:25  tlaino

	* [r9350] fine tuning of exclusion lists through two list sections
	  + 7 new regtests.

2009-12-18 14:06  tlaino

	* [r9349] using cp_unimplemented

2009-12-18 13:24  mguidon

	* [r9348] Wavefunction-Fitting: Generalized way of adding the DFT
	  correction.
	  Currently supports coulomb, short-range and truncated HFX
	  potentials.

2009-12-18 08:30  tlaino

	* [r9344] using CPP

2009-12-18 07:36  tlaino

	* [r9343] fix: missing initialization

2009-12-18 06:23  tlaino

	* [r9342] fix for non associated mapping arrays

2009-12-18 04:17  vondele

	* [r9341] Automatic minor version update to 2.1.103

2009-12-17 04:17  vondele

	* [r9340] Automatic minor version update to 2.1.102

2009-12-16 16:41  urban

	* [r9339] Towards full support for all 4 data types.

2009-12-16 04:17  vondele

	* [r9338] Automatic minor version update to 2.1.101

2009-12-15 21:40  urban

	* [r9337] Towards full support for the four (R/C 4/8) data types.

2009-12-15 08:56  tlaino

	* [r9336] still robodoc?! this is a 2 years old story..

2009-12-15 08:40  tlaino

	* [r9335] prettify

2009-12-15 08:30  urban

	* [r9334] Standardizing non-standard format specifiers ($) that
	  break NAG.

2009-12-15 04:17  vondele

	* [r9333] Automatic minor version update to 2.1.100

2009-12-14 17:53  jgh

	* [r9332] Remove repetitive comment

2009-12-14 17:18  hforbert

	* [r9331] Committing changes to PINT mostly regarding experimental
	  code

2009-12-14 13:08  tlaino

	* [r9330] removed unused file

2009-12-13 13:58  tlaino

	* [r9329] using cpp

2009-12-13 04:17  vondele

	* [r9328] Automatic minor version update to 2.1.99

2009-12-12 19:54  tlaino

	* [r9327] bug fix

2009-12-12 15:39  tlaino

	* [r9326] making the usage of external_control more restrictive..

2009-12-12 14:41  tlaino

	* [r9325] prettify

2009-12-12 04:17  vondele

	* [r9324] Automatic minor version update to 2.1.98

2009-12-11 22:01  marcella

	* [r9323] removed globev from scf_env_do_scf

2009-12-11 12:15  urban

	* [r9322] Forces symmetric matrices to have CP2K-compatible index.

2009-12-11 10:28  urban

	* [r9321] Cleans up adding new transposed blocks to product matrix.

2009-12-11 10:26  urban

	* [r9320] Cleans up adding new transposed blocks to product matrix.

2009-12-11 09:36  ibethune

	* [r9319] Sparse alltoall in rs_pw_transfer_distributed

2009-12-11 06:40  marcella

	* [r9317] output jacobi

2009-12-11 04:17  vondele

	* [r9316] Automatic minor version update to 2.1.97

2009-12-10 14:47  urban

	* [r9315] Better handling of product matrices:
	  * Dropped when unneeded (beta=0 and columns match right
	  matrix) thereby avoiding uneeded data copies and additions
	  * Product symmetry enforced
	  More accurate bounds checking when copying data block-by-block.

2009-12-10 04:17  vondele

	* [r9314] Automatic minor version update to 2.1.96

2009-12-09 16:07  vweber

	* [r9313] new general distribution, fixed few problems in sparse
	  linear algebra.

2009-12-09 13:19  tlaino

	* [r9312] added an A_SCALE for scaling generic A matrix to the
	  system frequency

2009-12-09 04:17  vondele

	* [r9311] Automatic minor version update to 2.1.95

2009-12-09 00:30  tlaino

	* [r9310] removing AUNITS and moving NADD -> NDIM (which is now the
	  real dimension of the GLE matrix)

2009-12-08 13:57  jgh

	* [r9309] Towards OFDFT module for Quickstep (1. step)

2009-12-08 04:17  vondele

	* [r9308] Automatic minor version update to 2.1.94

2009-12-07 16:24  vweber

	* [r9307] changed intent to in for 1d dbcsr_get_block. added
	  hadamard routine.

2009-12-07 14:55  vweber

	* [r9306] new interfaces dbcsr/cfm

2009-12-07 14:14  vweber

	* [r9305] added timer

2009-12-07 12:56  vweber

	* [r9304] makeup

2009-12-07 04:17  vondele

	* [r9303] Automatic minor version update to 2.1.93

2009-12-06 21:39  marcella

	* [r9302] distributed jacobi rotation, reduced memory requirement

2009-12-06 04:17  vondele

	* [r9301] Automatic minor version update to 2.1.92

2009-12-05 04:17  vondele

	* [r9299] Automatic minor version update to 2.1.91

2009-12-04 16:40  vweber

	* [r9296] turned on the dbcsr_multiply_fm_fm_t routine, reset regs
	  mostly small, larger for
	  regtest-nonortho/graphite*.inp and
	  regtest-ot/He2_ddapc_constraint-3.inp (the difference new/old
	  gets negligible for tighter SCF convergence).

2009-12-04 15:47  tlaino

	* [r9295] this was a strange one..

2009-12-04 09:54  vweber

	* [r9294] bugs fix, dbcsr multiply now accepts transpose arguments
	  (gemm style), parallel precond.

2009-12-04 04:17  vondele

	* [r9293] Automatic minor version update to 2.1.90

2009-12-03 12:25  urban

	* [r9292] Proper DGEMM call for transposed product block (currently
	  unused).

2009-12-03 12:01  urban

	* [r9291] Removes unwanted parameter (its uninitialized use caught
	  by valgrind).

2009-12-03 10:59  urban

	* [r9290] Turns off debugging.

2009-12-03 10:55  urban

	* [r9289] Restores commit with bugfix.

2009-12-03 07:13  tlaino

	* [r9287] generic (functional form) external electrostatic field
	  (static) for DFT + regtests.

2009-12-03 04:17  vondele

	* [r9286] Automatic minor version update to 2.1.89

2009-12-02 20:31  tlaino

	* [r9285] removing other leftovers

2009-12-02 20:19  tlaino

	* [r9284] reverting back: this commit makes regtests explode in
	  terms of memory allocation: more than 6GB per simple regtest!

2009-12-02 17:53  tlaino

	* [r9283] removing leftover..

2009-12-02 14:49  urban

	* [r9282] Smarter array resizing in DBCSR multiply.

2009-12-02 13:55  vondele

	* [r9281] add additional flushes in case of error

2009-12-02 13:47  mguidon

	* [r9280] Fix uninitialized logical

2009-12-02 13:37  tlaino

	* [r9279] use default IO and avoid generation of localLog files

2009-12-02 13:37  tlaino

	* [r9278] write only if unit_nr is >0

2009-12-02 13:03  krack

	* [r9277] bug fix

2009-12-02 12:51  marcella

	* [r9276] forgotten commit

2009-12-02 12:40  tlaino

	* [r9275] fix default IO in MC. Using CPP instead of stop_memory

2009-12-02 12:39  tlaino

	* [r9274] indent

2009-12-02 12:39  tlaino

	* [r9273] prettify

2009-12-02 11:52  marcella

	* [r9272] distributed jacobi rotation for the minimization of the
	  spread functional, added input-option to give an energy
	  range where to select the states to be localized

2009-12-02 11:51  ibethune

	* [r9271] Added back in fixed OpenMP for xz_to_yz etc. routines

2009-12-02 11:12  tlaino

	* [r9270] fix wrong IO control for MM_POTENTIAL

2009-12-02 10:31  krack

	* [r9269] print PDOS for all virtual orbitals with nlumo=-1 (like
	  for MO cube file output)

2009-12-02 04:17  vondele

	* [r9268] Automatic minor version update to 2.1.88

2009-12-02 00:10  tlaino

	* [r9267] skip ionode as argument to pw_to_cube. bug fix:
	  determining the right node which is writing (unit_nr>0) (not
	  necessarily node 0!)

2009-12-01 22:59  tlaino

	* [r9266] avoid recursive IO

2009-12-01 21:00  tlaino

	* [r9265] making uniform the usage of the default IO unit: this
	  should avoid wrong syntax and the creation of zillions of
	  localLog files (no need to setup/check anymore the ionode: just
	  call directly from all processors cp_logger_get_default_io_unit
	  and discriminate with output_unit>0 who is writing).

2009-12-01 15:34  urban

	* [r9263] Allows input and output messages in mp_sendrecv to have
	  different sizes.

2009-12-01 14:47  ibethune

	* [r9262] Add in allocate check for call found to be suspect under
	  Pathscale

2009-12-01 14:22  tlaino

	* [r9260] prettify

2009-12-01 12:31  urban

	* [r9258] Cleanup of local multiply (removes uneeded arrays, uses
	  new native block
	  offsets instead of block sizes).

2009-12-01 09:45  vweber

	* [r9257] bugfix in _scale

2009-12-01 04:17  vondele

	* [r9256] Automatic minor version update to 2.1.87

2009-11-30 19:27  vondele

	* [r9255] cleanup unused subroutine

2009-11-30 16:33  urban

	* [r9254] Prospective bugfix in joining blocks for multiplication.

2009-11-30 15:23  urban

	* [r9253] Adds more support for accessing any data type in a DBCSR
	  iterator and in
	  other access functions.

2009-11-30 09:08  vweber

	* [r9252] added row/col offset in the iterator, cleaned explicit
	  build of the offset arrays.

2009-11-30 04:17  vondele

	* [r9251] Automatic minor version update to 2.1.86

2009-11-29 19:37  tlaino

	* [r9250] wildcard for nonbonded interactions in FIST + regtest

2009-11-29 04:17  vondele

	* [r9249] Automatic minor version update to 2.1.85

2009-11-28 04:18  vondele

	* [r9248] Automatic minor version update to 2.1.84

2009-11-27 15:05  tlaino

	* [r9247] bug fix + 1 reset

2009-11-27 14:54  tlaino

	* [r9246] fix: avoid crash due to undefined variable in a print
	  statement

2009-11-27 14:37  tlaino

	* [r9245] initialize array (was making crash few compilers..)

2009-11-27 14:13  tlaino

	* [r9244] prettify

2009-11-27 14:02  urban

	* [r9243] Changes type-specific message passing calls to generic
	  ones.

2009-11-27 13:39  tlaino

	* [r9242] bug fix for core/shell when not all atoms are defined as
	  shell (and a topology structure is present)

2009-11-27 13:29  urban

	* [r9241] Unifies subroutines with different types in
	  message_passing.F by
	  introducing templates. Has interfaces for single/double precision
	  real/complex as well as 32-bit and 64-bit integers.

2009-11-27 10:46  krack

	* [r9238] Recover correct printout of the mixing parameter

2009-11-27 10:40  tlaino

	* [r9236] expanding the possibility of the exclusion lists for vdw
	  and ei + regtest. (useful for forcefield
	  which require 1-2 NB interactions and have also bend terms:
	  suggested from Sergey)

2009-11-27 09:09  vweber

	* [r9235] allocation of density matrices from nbl. added a reg

2009-11-27 04:17  vondele

	* [r9234] Automatic minor version update to 2.1.83

2009-11-27 01:00  tlaino

	* [r9233] bug fix NEB when using MIXED_ENV (handling in a correct
	  way IO) (bug reported by Carlo)

2009-11-26 21:38  tlaino

	* [r9232] prettify

2009-11-26 17:38  jgh

	* [r9231] Atomic initial guess: bug fix for pseudos with complete
	  core shells;
	  reset 3 regtests

2009-11-26 13:30  jgh

	* [r9230] refactoring

2009-11-26 12:37  krack

	* [r9229] Fix format syntax

2009-11-26 11:08  jgh

	* [r9228] refactoring

2009-11-26 08:36  jgh

	* [r9227] More robust sorting for very large grids

2009-11-26 04:17  vondele

	* [r9226] Automatic minor version update to 2.1.82

2009-11-25 16:19  tlaino

	* [r9225] prettify

2009-11-25 13:24  vweber

	* [r9224] bugfix from -GAPW rewrite- (rebuild density coeff).

2009-11-25 13:00  urban

	* [r9223] Prospective bugfix: new blocks should be appended to
	  existing data.

2009-11-25 12:36  marcella

	* [r9222] update gle thermostat (Ceriotti),
	  shell-model bug fix
	  fix in keyword description

2009-11-25 04:17  vondele

	* [r9220] Automatic minor version update to 2.1.81

2009-11-24 21:27  vondele

	* [r9219] Fix OMP bug

2009-11-24 09:55  tlaino

	* [r9218] fix typo

2009-11-24 08:36  jgh

	* [r9217] Return best estimate from Powell routine even if
	  algorithm didn't converge

2009-11-24 04:17  vondele

	* [r9216] Automatic minor version update to 2.1.80

2009-11-23 14:22  tlaino

	* [r9215] out of plane bends + regtests (Louis)

2009-11-23 04:17  vondele

	* [r9214] Automatic minor version update to 2.1.79

2009-11-22 22:23  tlaino

	* [r9213] prettify

2009-11-22 16:11  jgh

	* [r9212] Atom code: out of bounds bug fix

2009-11-22 04:17  vondele

	* [r9211] Automatic minor version update to 2.1.78

2009-11-21 04:17  vondele

	* [r9210] Automatic minor version update to 2.1.77

2009-11-20 15:19  urban

	* [r9209] Resurrects and improves joining of right-matrix blocks in
	  the local
	  multiply.

2009-11-20 14:58  vondele

	* [r9208] prettify

2009-11-20 14:32  vweber

	* [r9207] minor bugfix

2009-11-20 13:07  mguidon

	* [r9206] Add save pointer associations in case a thread has no
	  bins to process

2009-11-20 07:23  vondele

	* [r9205] add recent HFX paper to references

2009-11-20 04:17  vondele

	* [r9204] Automatic minor version update to 2.1.76

2009-11-19 12:19  vweber

	* [r9203] updated some dbcsr utilities (replaced old style loops
	  with iterators)

2009-11-19 04:17  vondele

	* [r9202] Automatic minor version update to 2.1.75

2009-11-18 14:49  urban

	* [r9201] Better OpenMP support with matrices.

2009-11-18 14:10  ibethune

	* [r9200] More OpenMP in the pw2rs part of the routine (missed in
	  earlier commit)

2009-11-18 04:17  vondele

	* [r9199] Automatic minor version update to 2.1.74

2009-11-17 14:20  urban

	* [r9197] Unifies the DBCSR multiplication routines (never uses
	  RMA/one-sided
	  communication now).
	  Unifies creation of row/column matrix images.
	  Adds some needed routines for existing interfaces to to
	  message_passing.
	  Bugfix: use correct (antisymmetric) neighbor list in
	  qs_linres_op.

2009-11-17 12:16  vweber

	* [r9196] bugfix for iterator

2009-11-17 04:17  vondele

	* [r9195] Automatic minor version update to 2.1.73

2009-11-16 04:17  vondele

	* [r9194] Automatic minor version update to 2.1.72

2009-11-15 04:17  vondele

	* [r9193] Automatic minor version update to 2.1.71

2009-11-14 04:17  vondele

	* [r9192] Automatic minor version update to 2.1.70

2009-11-13 04:17  vondele

	* [r9191] Automatic minor version update to 2.1.69

2009-11-12 15:17  vweber

	* [r9190] makeup + fix minor sign error in the occupation function

2009-11-12 11:21  vweber

	* [r9189] bugfix: check for 0-size block while filtering. added
	  occupation function, possibility of filtering the ks matrix and
	  checks for association of density matrix blocks where needed.

2009-11-12 04:17  vondele

	* [r9188] Automatic minor version update to 2.1.68

2009-11-11 15:48  vweber

	* [r9187] added the possibility to filter the overlap matrix.

2009-11-11 13:01  vweber

	* [r9186] create density matrix from -scratch- (not anymore a copy
	  from S). add few finalized where needed.
	  bugfix nze is now reset if a block of the matrix is removed
	  (filter).

2009-11-11 04:17  vondele

	* [r9184] Automatic minor version update to 2.1.67

2009-11-10 14:10  vweber

	* [r9183] merged set_up_op_sm and allocate_w_matrix_blocks

2009-11-10 04:17  vondele

	* [r9182] Automatic minor version update to 2.1.66

2009-11-09 23:56  tlaino

	* [r9180] prettify

2009-11-09 14:35  jgh

	* [r9179] prettify atom routines

2009-11-09 13:18  jgh

	* [r9178] Towards unresticted atomic calculatins

2009-11-09 04:17  vondele

	* [r9177] Automatic minor version update to 2.1.65

2009-11-08 14:07  jgh

	* [r9176] cosmetics

2009-11-08 12:19  marcella

	* [r9175] if smearing is used, localization and wannier centers are
	  restricted to the subset of MOs with integer occupation

2009-11-08 04:17  vondele

	* [r9174] Automatic minor version update to 2.1.64

2009-11-07 04:17  vondele

	* [r9173] Automatic minor version update to 2.1.63

2009-11-06 20:03  vondele

	* [r9172] remove useless comments

2009-11-06 15:33  vweber

	* [r9171] added interfaces dbcsr/scalapack

2009-11-06 15:22  urban

	* [r9170] Bugfix for copy with sparsity retainment.

2009-11-06 14:10  mguidon

	* [r9169] Add missing OpenMP variable scope definition

2009-11-06 13:34  urban

	* [r9168] Retain target matrix sparsity on matrix copy. Retains
	  sparsity in the
	  wavefunction extrapolation. Adds several relevant regtests.

2009-11-06 09:48  krack

	* [r9167] Simplification and 20x speedup for Lowdin (2 regtest
	  reset)

2009-11-06 07:51  urban

	* [r9166] Changes thread distribution to cover all and not just
	  local rows.

2009-11-06 04:17  vondele

	* [r9165] Automatic minor version update to 2.1.62

2009-11-05 16:06  mguidon

	* [r9163] Added preprocessor flags such that cp2k compiles faster
	  when not used with Libint.
	  Changed the generating scripts accordingly (/tools/hfx_tools).

2009-11-05 10:54  tlaino

	* [r9161] bug fix for climbing image NEB: never consider boundary
	  points

2009-11-05 09:45  ibethune

	* [r9160] More OpenMP - additional speedup in rs_pw_transfer

2009-11-05 09:21  vweber

	* [r9159] get rid of dbcsr_allocate_matrix,
	  dbcsr_replicate_matrix_struct and sparsity_id checks.
	  set eps to 1e-6 for linear response calculations (reset few
	  regs).

2009-11-05 09:18  krack

	* [r9158] Simplification and speedup

2009-11-05 08:44  urban

	* [r9157] Speedups: bisection search by column for block and faster
	  fm<->dbcsr
	  conversion (fewer block lookups in dbcsr_complete_redistribute).

2009-11-05 04:17  vondele

	* [r9156] Automatic minor version update to 2.1.61

2009-11-04 16:19  mguidon

	* [r9153] Reanimated the wavefunction fitting code. For my own
	  sanity added 2 meaningless
	  regtests. Stills needs some refactoring.

2009-11-04 13:59  marcella

	* [r9152] bug fix

2009-11-04 09:09  vweber

	* [r9151] removed unused calls to the pool (sm) and useless
	  printing.

2009-11-04 08:04  krack

	* [r9150] * Initial occupation with Lowdin enabled
	  * Method mulliken_block_diag removed
	  * Various bug fixes
	  * 4 regtests added for method Lowdin

2009-11-04 04:17  vondele

	* [r9149] Automatic minor version update to 2.1.60

2009-11-03 16:41  urban

	* [r9148] Prettify, adds another size 0 check, and adds optional
	  row/column block
	  size output parameters to all block retrieval subroutines.

2009-11-03 14:25  ibethune

	* [r9147] OpenMP for rs_pw_transfer_distributed

2009-11-03 13:33  urban

	* [r9146] Bugfixes: uninitialized memory use in output, and
	  pointers to matrices.
	  Stops memory leak in some parallel configurations.

2009-11-03 11:16  marcella

	* [r9143] some changes in the defaulta of mixing, added subspace
	  diagonalization as internal loop

2009-11-03 10:27  vweber

	* [r9142] preallocate work array to avoid many reallocations in
	  redistribute_matrix, fixed matrix_type in copy_fm_to_dbcsr

2009-11-03 04:17  vondele

	* [r9141] Automatic minor version update to 2.1.59

2009-11-02 15:34  urban

	* [r9140] Bugfix: uninitialized variable usage leading to crashes.

2009-11-02 13:51  urban

	* [r9139] Speedup: more efficient fm->dbcsr conversion of
	  (anti)symmetric matrices.

2009-11-02 11:43  ibethune

	* [r9138] Avoid overthreading when nthreads > array dimensions

2009-11-02 10:25  urban

	* [r9137] Bugfix: wrong logic.

2009-11-02 04:17  vondele

	* [r9136] Automatic minor version update to 2.1.58

2009-11-01 04:17  vondele

	* [r9135] Automatic minor version update to 2.1.57

2009-10-31 13:04  jgh

	* [r9134] Bug fix for numeric exchange, 2 regtests reset

2009-10-31 04:17  vondele

	* [r9133] Automatic minor version update to 2.1.56

2009-10-30 16:36  urban

	* [r9132] Better thread safety:
	  * makes compatibility routines thread-safe;
	  * threaded support for iterator;
	  * enhancements to array_types (safety checks), dbcsr_data_obj
	  (2-d data
	  areas)

2009-10-30 15:39  vweber

	* [r9131] replaced dbcsr_get_block_node by dbcsr_get_block_p

2009-10-30 10:08  vweber

	* [r9130] removed unnecessary copies + reset a reg.

2009-10-30 09:04  urban

	* [r9129] Speedup: Preallocation in dbcsr_distribute().

2009-10-30 04:17  vondele

	* [r9128] Automatic minor version update to 2.1.55

2009-10-29 17:56  tlaino

	* [r9127] prettify

2009-10-29 15:34  vweber

	* [r9125] removed some printing

2009-10-29 15:29  vweber

	* [r9124] substitution sm -> dbcsr (9th step)

2009-10-29 10:46  urban

	* [r9123] Speedup: faster FM->DBCSR conversion.

2009-10-29 10:09  urban

	* [r9121] Speedup: Changes BLACS grids parameters to optimize
	  primarily for
	  common factors in process grid dimensions and secondarily for
	  "squareness". Now favors more columns over more rows.

2009-10-29 10:08  urban

	* [r9120] Speedup: Creates more even DBCSR distributions.

2009-10-29 04:17  vondele

	* [r9119] Automatic minor version update to 2.1.54

2009-10-28 15:18  krack

	* [r9118] Allow for special initial orbital occupations with DFT+U

2009-10-28 11:54  tlaino

	* [r9116] bug fix for DOF (single atoms, linear molecules): reset
	  regtests

2009-10-28 04:17  vondele

	* [r9115] Automatic minor version update to 2.1.53

2009-10-27 22:53  tlaino

	* [r9114] prettify

2009-10-27 22:02  urban

	* [r9113] Fixes wrong results due to bad distribution when using
	  processor grids
	  that have dimensions with several different factors.

2009-10-27 14:51  urban

	* [r9112] Removes outdated assertion (fixes failing EP regtests).

2009-10-27 14:17  urban

	* [r9110] Use correct pointer array bounds when calling BLAS
	  (should fix
	  the issues with ifort).

2009-10-27 10:30  vweber

	* [r9108] substitution sm -> dbcsr (8th step), reset regs, fix out
	  of mems

2009-10-27 04:17  vondele

	* [r9107] Automatic minor version update to 2.1.52

2009-10-26 20:18  urban

	* [r9106] New features and sm_fm_multiply bugfix:
	  * addition into existing matrix upon multiplication
	  (C=beta*C+alpha*A*B);
	  * optional sparsity enforcement for product matrix in
	  multiplication;
	  * blocks of any matrix (A, B, C) can be transposed in
	  multiplication;
	  * limits scaling of matrix to specified (non-blocked) column;
	  * native DBCSR implementation of sm_fm_multiplication (fixes
	  recent bugs);
	  * native DBCSR implementation of sm_plus_fm_fm_t (not yet turned
	  on);
	  * moves some subroutines from dbcsr_util to dbcsr_operations.

2009-10-26 20:16  urban

	* [r9105] Bugfix: unitialized pointer use.

2009-10-26 09:53  ibethune

	* [r9104] Added __HAS_NO_OMP_3 define which should be used with
	  compilers that do not support OpenMP 3.0 (only needed if
	  compiling the code for SMP)

2009-10-26 06:56  tlaino

	* [r9103] fix wrong index (Vittorio) + implicit none in all
	  definitions

2009-10-26 04:17  vondele

	* [r9102] Automatic minor version update to 2.1.51

2009-10-25 04:17  vondele

	* [r9097] Automatic minor version update to 2.1.50

2009-10-24 17:06  marcella

	* [r9096] getting rid of full gapw

2009-10-24 03:17  vondele

	* [r9094] Automatic minor version update to 2.1.49

2009-10-23 23:28  tlaino

	* [r9093] completing the implementation of IMPROPERS (possibility
	  to define them in the input file with a proper section); added
	  relevant comments and regtests (Louis).

2009-10-23 13:47  ibethune

	* [r9092] Improved OpenMP for parallel FFT

2009-10-23 09:56  krack

	* [r9091] * Keep iter_delta in "just_energy" SCF iteration steps
	  instead of setting it to 1.0.
	  * SCF energy change printed only between cycles with KS matrix
	  updates
	  * cosmetics

2009-10-23 03:17  vondele

	* [r9090] Automatic minor version update to 2.1.48

2009-10-22 12:08  tlaino

	* [r9089] Added ANGVEL_TOL to remove angular velocities during MD +
	  regtest

2009-10-22 08:04  tlaino

	* [r9088] comvel_tol has no default: the rescaling of comvel is
	  applied only if the keyword is present (not if the value is
	  larger than 0.0 .. so one can specify 0.0 as well as a possible
	  value)

2009-10-22 07:30  tlaino

	* [r9087] prettify

2009-10-22 07:30  tlaino

	* [r9086] indent code

2009-10-22 07:30  tlaino

	* [r9085] remove unused pointer

2009-10-22 03:17  vondele

	* [r9084] Automatic minor version update to 2.1.47

2009-10-21 15:25  vondele

	* [r9083] fix typo

2009-10-21 14:18  ibethune

	* [r9082] Bugfix: take care of the non-parallel case.

2009-10-21 10:38  ibethune

	* [r9081] More efficient memory access pattern for MPI/OMP parallel
	  FFT

2009-10-21 03:17  vondele

	* [r9080] Automatic minor version update to 2.1.46

2009-10-20 15:16  vondele

	* [r9078] cossin: Fixes a that might trigger for the berry phase
	  dipole

2009-10-20 09:35  krack

	* [r9076] Release modulo 2 constraint for the redistribution of the
	  electrons as requested
	  for in the initial atomic guess in the case of spin polarised
	  calculations. The
	  affected regtest inputs are adapted to give the same numbers. ie.
	  the unphysical
	  initial occupations still occur, but the final scaling step
	  ensures a proper
	  electron count anyway.

2009-10-20 03:17  vondele

	* [r9074] Automatic minor version update to 2.1.45

2009-10-19 13:15  krack

	* [r9073] Add optional printout for DFT+U method Mulliken_charges

2009-10-19 11:57  krack

	* [r9072] Enhance printout for DFT+U and avoid printout
	  duplications. Reduce default
	  printout -> QS_SCF 0

2009-10-19 11:00  ibethune

	* [r9071] Fix: some variables missing from OpenMP data scope
	  clauses

2009-10-19 09:42  krack

	* [r9069] Enable another DFT+U variant

2009-10-19 03:17  vondele

	* [r9068] Automatic minor version update to 2.1.44

2009-10-18 19:26  tlaino

	* [r9067] bug fix

2009-10-18 03:17  vondele

	* [r9066] Automatic minor version update to 2.1.43

2009-10-17 19:11  tlaino

	* [r9065] Allowing MOL_SET with AMBER topology

2009-10-17 16:08  jgh

	* [r9064] Cube file output for electrical fields

2009-10-17 16:08  jgh

	* [r9063] Speedup for geminal integrals

2009-10-17 03:17  vondele

	* [r9062] Automatic minor version update to 2.1.42

2009-10-16 16:34  tlaino

	* [r9061] refactoring and bug fixes (assignment element for
	  different topology formats: now the behavior is the same
	  independently of the input format)

2009-10-16 15:16  krack

	* [r9060] Allow to monitor the relevant occupation numbers in DFT+U
	  runs

2009-10-16 12:35  jgh

	* [r9058] Optimization of geminals

2009-10-16 03:17  vondele

	* [r9057] Automatic minor version update to 2.1.41

2009-10-15 12:08  urban

	* [r9056] Speedup of FM->DBCSR conversion, including option to
	  retain the current
	  sparsity of target DBCSR matrix.

2009-10-15 03:17  vondele

	* [r9054] Automatic minor version update to 2.1.40

2009-10-14 03:17  vondele

	* [r9052] Automatic minor version update to 2.1.39

2009-10-13 19:35  vondele

	* [r9051] improve Global print level output

2009-10-13 19:21  vondele

	* [r9050] provide a standard form for m_datum

2009-10-13 12:57  tlaino

	* [r9048] bug fix when computing degrees of freedom for linear
	  molecules (2 atoms or more): additional corrections to DOFs maybe
	  required in the future. Reset quite few regs. Bug reported by
	  Marcel.

2009-10-13 11:21  urban

	* [r9047] Compilation/cosmetic fixes:
	  * Adds missing variable data-sharing attribute declaration in
	  OpenMP
	  parallel section.
	  * Removes spurious warnings in several routines.

2009-10-13 09:25  vweber

	* [r9044] substitution sm -> dbcsr (7th step)

2009-10-13 09:03  vondele

	* [r9043] further cleanups and speedups

2009-10-13 08:21  ibethune

	* [r9042] Added OpenMP directives to parallel FFT (FFTW3 and FFT
	  transposes)

2009-10-13 07:40  urban

	* [r9041] Use thread-safe calls to get data block in an OpenMP
	  environment.

2009-10-13 03:17  vondele

	* [r9040] Automatic minor version update to 2.1.38

2009-10-12 19:21  vondele

	* [r9039] cleanup and speedup of get_current_loads

2009-10-12 15:27  tlaino

	* [r9038] extra kind of BEND term : it is a mixture of the bend
	  term (sixth order series) and stretch-bend interaction, both from
	  the MM3 force field of Allinger (Louis)

2009-10-12 13:42  urban

	* [r9037] Bugfix: Non-square process grid multiplication.

2009-10-12 12:49  mguidon

	* [r9035] Parallelized initial guess of atomic block distribution

2009-10-12 12:07  vondele

	* [r9034] modify preconditioners that should not directly rely on
	  the energy level of the homo. While energy differences are
	  meaningful (energy_gap), absolute values are not. This affects 43
	  regtests. Preconditioner construction now does not need to
	  compute subspace eigenvalues as a result, which speeds up
	  preconditioner generation.

2009-10-12 03:17  vondele

	* [r9033] Automatic minor version update to 2.1.37

2009-10-11 03:17  vondele

	* [r9032] Automatic minor version update to 2.1.36

2009-10-10 21:32  krack

	* [r9031] bug fix: use the right timer entry for adaptive format
	  and keep printout
	  readable by enforcing at least one blank between columns

2009-10-10 07:14  vondele

	* [r9030] remove useless interface

2009-10-10 06:20  vondele

	* [r9029] clean up white space

2009-10-10 03:17  vondele

	* [r9028] Automatic minor version update to 2.1.35

2009-10-09 03:17  vondele

	* [r9027] Automatic minor version update to 2.1.34

2009-10-08 14:11  fschiff

	* [r9026] possibility to avoid calculation of distances in pbc

2009-10-08 10:24  vondele

	* [r9025] adjust format of the timings dynamically for longer runs

2009-10-08 03:17  vondele

	* [r9024] Automatic minor version update to 2.1.33

2009-10-07 14:33  krack

	* [r9023] HOMO-LUMO gap printout (diag.)

2009-10-07 03:17  vondele

	* [r9022] Automatic minor version update to 2.1.32

2009-10-06 15:03  mguidon

	* [r9021] Refactored interface to Libint library. Disentangle
	  ISO_C_BINDING capabilities
	  from compiler. Makes it possible to use different compiler
	  versions of the same company
	  that have different ISO_C_BINDING support. In addition it is now
	  easily possible
	  to use higher angular momenta than what is provided with default
	  Libint settings. See
	  /tools/hfx_tools/libint_tools/README_LIBINT for more information.

2009-10-06 12:44  tlaino

	* [r9019] setting to -1 IO unit when not present..

2009-10-06 12:43  tlaino

	* [r9018] simplify logic

2009-10-06 11:08  krack

	* [r9016] prettified

2009-10-06 11:07  krack

	* [r9015] Allow for more output concerning the atomic guess

2009-10-06 08:02  krack

	* [r9013] Enable the use of a libint library compiled for larger l
	  values

2009-10-06 03:17  vondele

	* [r9012] Automatic minor version update to 2.1.31

2009-10-05 10:36  tlaino

	* [r9011] fix typo (Louis Vanduyfhuys)

2009-10-02 03:17  vondele

	* [r9010] Automatic minor version update to 2.1.27

2009-10-01 03:18  vondele

	* [r9008] Automatic minor version update to 2.1.26

2009-09-30 03:17  vondele

	* [r9006] Automatic minor version update to 2.1.25

2009-09-29 07:27  tlaino

	* [r8997] removing get_nhc_temperature not used anymore

2009-09-29 07:19  tlaino

	* [r8996] irrelevant changes

2009-09-29 03:17  vondele

	* [r8995] Automatic minor version update to 2.1.24

2009-09-28 21:48  marcella

	* [r8993] Langevin equation with colored noise for constant
	  temperature MD

2009-09-28 15:41  marcella

	* [r8991] bug fix : smearing and G-space mixing working with LSD

2009-09-28 03:17  vondele

	* [r8990] Automatic minor version update to 2.1.23

2009-09-27 03:17  vondele

	* [r8989] Automatic minor version update to 2.1.22

2009-09-26 07:19  marcella

	* [r8988] removed unused stuff

2009-09-26 03:17  vondele

	* [r8987] Automatic minor version update to 2.1.21

2009-09-25 10:26  tlaino

	* [r8984] added units for COMVEL_TOL. fix typo.

2009-09-25 08:40  marcella

	* [r8977] pretty

2009-09-25 08:35  marcella

	* [r8974] this gives the possibility to redistribute the electrons
	  among the atomic orbitals
	  while generating the first density matrix for scf_guess ATOMIC;
	  might
	  be useful for charged systems, and systems with spatially
	  localized spin polarization

2009-09-25 03:17  vondele

	* [r8973] Automatic minor version update to 2.1.20

2009-09-24 03:17  vondele

	* [r8972] Automatic minor version update to 2.1.19

2009-09-23 03:17  vondele

	* [r8971] Automatic minor version update to 2.1.18

2009-09-22 21:34  tlaino

	* [r8970] Fix error conditions and messages.

2009-09-22 13:56  urban

	* [r8969] Bugfix: Fixes misplaced assertion in block retrieval.

2009-09-22 03:18  vondele

	* [r8968] Automatic minor version update to 2.1.17

2009-09-21 15:38  urban

	* [r8967] Retain SM structure upon reallocation from DBCSR.
	  dbcsr_get_block() ignores zero-sizes requests

2009-09-21 03:17  vondele

	* [r8966] Automatic minor version update to 2.1.16

2009-09-20 03:17  vondele

	* [r8965] Automatic minor version update to 2.1.15

2009-09-19 03:17  vondele

	* [r8964] Automatic minor version update to 2.1.14

2009-09-18 15:39  urban

	* [r8963] Fix for switched logic.

2009-09-18 14:24  urban

	* [r8962] Quiets spurious warnings.

2009-09-18 14:23  urban

	* [r8961] Zeroes new allocation and finalizes temporary matrix to
	  ensure a valid
	  index.

2009-09-18 12:47  mguidon

	* [r8959] Bug fix screening forces (Thanks Matthias)

2009-09-18 03:19  vondele

	* [r8958] Automatic minor version update to 2.1.13

2009-09-17 22:37  tlaino

	* [r8957] removing more unused dummy arguments + cleaning

2009-09-17 15:05  urban

	* [r8956] Matrix multiplication to be limited to a submatrix (only
	  a specified range
	  of rows or columns is multiplied).

2009-09-17 08:16  ibethune

	* [r8955] Bugfix: FFTW3 now generates plans with SSE (if available)
	  by default as originally intended. Reset several regtests.

2009-09-17 05:27  tlaino

	* [r8954] removing entire set (excluding cp_error_type) of unused
	  dummy arg from constraint* modules.

2009-09-17 03:17  vondele

	* [r8953] Automatic minor version update to 2.1.12

2009-09-16 22:58  tlaino

	* [r8952] making comments more informative ;-)

2009-09-16 17:35  vondele

	* [r8951] remove uninformative comments

2009-09-16 15:45  tlaino

	* [r8950] adaptive hills in metadynamics (Rachel)

2009-09-16 07:19  vondele

	* [r8948] remove unused dummy arg

2009-09-16 03:17  vondele

	* [r8947] Automatic minor version update to 2.1.11

2009-09-15 03:18  vondele

	* [r8946] Automatic minor version update to 2.1.10

2009-09-14 19:16  urban

	* [r8945] Bugfix: non-scalapack FM-to-DBCSR conversion.

2009-09-14 03:17  vondele

	* [r8944] Automatic minor version update to 2.1.9

2009-09-13 03:17  vondele

	* [r8943] Automatic minor version update to 2.1.8

2009-09-12 15:56  vondele

	* [r8942] remove unused dummy arguments

2009-09-12 15:44  vondele

	* [r8940] remove unused dummy arg para_env

2009-09-12 15:41  krack

	* [r8939] CP2K version and year tag added to XML dump for automatic
	  manual generation

2009-09-12 14:42  vondele

	* [r8938] remove unused dummy arguments

2009-09-12 14:34  vondele

	* [r8937] remove unused variables

2009-09-12 09:23  vondele

	* [r8936] remove unused dummy arguments

2009-09-12 03:17  vondele

	* [r8935] Automatic minor version update to 2.1.7

2009-09-11 18:59  vondele

	* [r8934] cleaning unused arguments

2009-09-11 17:59  vondele

	* [r8933] fix too short fmt string

2009-09-11 03:17  vondele

	* [r8932] Automatic minor version update to 2.1.6

2009-09-10 14:27  urban

	* [r8931] Bugfix: fix divide-by-0 (now correctly handles zero-sized
	  array when
	  creating blocks of an array).

2009-09-10 13:49  fschiff

	* [r8930] speed up

2009-09-10 09:14  vondele

	* [r8929] Automatic minor version update to 2.1.5

2009-09-10 09:11  mguidon

	* [r8928] Hide memory leak at a place where nobody can see it

2009-09-10 08:38  fschiff

	* [r8927] bug fix

2009-09-09 12:24  urban

	* [r8926] Adds copy_dbcsr_to_fm() as emulation of copy_sm_to_fm().

2009-09-09 12:06  urban

	* [r8925] Bugfix: removes extraneous data copy

2009-09-09 09:04  fschiff

	* [r8924] assert for nonconverged EMD, m_flush for output

2009-09-09 07:56  vondele

	* [r8923] make FLUSH_SHOULD_FLUSH by default true

2009-09-09 04:36  vondele

	* [r8922] Automatic minor version update to 2.1.4

2009-09-09 04:36  vondele

	* [r8921] Automatic minor version update to 2.1.

2009-09-08 14:26  vweber

	* [r8920] fix bug + new routines

2009-09-08 13:38  vondele

	* [r8919] Automatic minor version update to 2.1.2

2009-09-08 13:35  vondele

	* [r8918] prepare for nightly minor version update

2009-09-08 13:33  urban

	* [r8917] Implements a redistribution of a dbcsr matrix to any
	  distribution that is
	  compatible with the global matrix row and column dimensions.

2009-09-08 07:14  tlaino

	* [r8916] simplify

2009-09-08 00:07  tlaino

	* [r8915] bug fix for multiple_unit_cell + 1 regtest

2009-09-08 00:07  tlaino

	* [r8914] added missing recursive specifications + avoid possible
	  division by zero

2009-09-07 23:33  tlaino

	* [r8913] prettify. By the way: two regtests are failing since a
	  couple of days: h2o_ot_precond_13.inp and no_ot_precond_13.inp
	  with an out of bounds.. [check routine dbcsr_util.F]

2009-09-07 18:23  vondele

	* [r8912] remove bogus labels

2009-09-07 18:06  vondele

	* [r8911] clean some unused dummy arguments

2009-09-04 20:51  vondele

	* [r8910] Fix an EMD restart bug after a should-stop.

2009-09-04 15:05  vweber

	* [r8909] template for dbcsr_methods

2009-09-04 13:33  vweber

	* [r8908] interface for the product dbcsr times fm

2009-09-04 11:36  urban

	* [r8907] Avoids NAG bug with pointer redirects.

2009-09-04 11:12  urban

	* [r8906] Fixes memory leak in fm->dbcsr copy subroutine.

2009-09-04 11:09  vweber

	* [r8905] cleaning

2009-09-04 09:24  urban

	* [r8904] Adds copy_fm_to_dbcsr() as emulation of copy_fm_to_sm.

2009-09-04 09:07  urban

	* [r8903] Expands get/put block capabilities of DBCSR matrices.

2009-09-04 08:11  mguidon

	* [r8901] cosmetics

2009-09-03 19:36  tlaino

	* [r8900] prettify

2009-09-03 10:30  fschiff

	* [r8898] bug fix for EMD initial step and restart

2009-09-03 07:00  mcgrath

	* [r8897] Write out coordinates in the correct units

2009-09-02 17:29  mguidon

	* [r8896] Allows NAG to finish regtests before Judgement Day

2009-09-02 16:34  urban

	* [r8895] Removes multi-thread access protection code from btrees.
	  They can now be
	  used per-thread.

2009-09-02 16:08  mguidon

	* [r8894] Introduce density matrix screening at atomic level. Only
	  kicks in for systems
	  containing thousands of atoms

2009-09-01 16:07  urban

	* [r8893] Removes some internal pointers in DBCSR matrices,
	  sprinkles some
	  needed dbcsr_init()s and fixes a few bugs.

2009-09-01 13:08  vweber

	* [r8892] substitution sm -> dbcsr (6th step). reset few regs
	  (<1e-10).

2009-09-01 11:37  urban

	* [r8891] Bugfix: memory leak & uninitialized variable.

2009-09-01 10:05  urban

	* [r8890] Restores most recent commit of fschiff.

2009-09-01 09:25  urban

	* [r8889] Oops, reverts previous commit reversions. Sorry.

2009-09-01 09:12  fschiff

	* [r8888] bug fix STEP_START_VAL for ehrenfest, speedup ehrenfest,
	  avoid unnecessary
	  computation of kohn-sham matrix

2009-09-01 09:03  urban

	* [r8887] Reuses data more often instead of copying and allows
	  dbcsr matrices to be
	  easily recycled.
	  Forces the SM<->DBCSR conversions to be local without using any
	  message
	  passing.

2009-08-31 17:32  mguidon

	* [r8886] Fix of previous commit (Sorry, forgot that other people
	  are using the compression as well)

2009-08-31 13:06  mguidon

	* [r8884] Added in-core forces for Ehrenfest MD (+ 2 regs)

2009-08-28 12:31  vweber

	* [r8883] pretty and reset 2 regs (<1e-12)

2009-08-28 12:02  vweber

	* [r8882] substitution sm -> dbcsr (5th step). fix few bugs. reset
	  few regs.

2009-08-28 08:20  fschiff

	* [r8880] bug fix for restart

2009-08-27 11:38  mguidon

	* [r8879] Bug fix dynamic blocks

2009-08-27 09:27  urban

	* [r8878] Allows marking matrix replication upon creation.
	  Allows keeping existing pointer when allocating a
	  real_matrix_type with a
	  pre-associated pointer

2009-08-27 06:39  tlaino

	* [r8877] updating runtime info before printing footer

2009-08-26 21:39  tlaino

	* [r8876] bug fix for lithium parameters PM3

2009-08-26 21:02  tlaino

	* [r8875] prettify

2009-08-26 20:54  urban

	* [r8874] Changes trace to return the correct value if the matrix
	  is replicated.
	  Adds subroutine to replace blocks in a matrix with another
	  matrix.
	  Adds subroutine for distribution of replicated matrix.
	  Bugfix: uninitialized reference counter in mutable data type &
	  others.

2009-08-26 15:16  vondele

	* [r8873] add & fix a timeset

2009-08-26 12:06  urban

	* [r8872] Allows summation of replicated matrices.
	  Enables summation of individual blocks when adding into an
	  existing array.
	  Shortens too-long subroutine name.

2009-08-26 10:51  mguidon

	* [r8870] Restore innermost screening level for derivatives

2009-08-26 10:05  mguidon

	* [r8867] Minor OpenMP adds

2009-08-26 09:22  mguidon

	* [r8866] Rounding issue in bit-container detection

2009-08-25 23:17  tlaino

	* [r8865] prettify

2009-08-25 11:38  vweber

	* [r8864] substitution sm -> dbcsr (4th step). removed unused
	  routines.

2009-08-25 10:52  mguidon

	* [r8863] Bug fix dynamic block size

2009-08-24 13:21  vweber

	* [r8862] fixed bugs.

2009-08-24 12:17  vondele

	* [r8861] update cost model, and introduce parameters for energies
	  and forces

2009-08-24 11:37  mguidon

	* [r8860] Significantly improves load balance by using dynamically
	  adjusted block sizes

2009-08-24 11:35  fschiff

	* [r8858] further speed up for EMD, bug fix in the forces

2009-08-24 10:07  urban

	* [r8857] Prettify.

2009-08-24 09:22  urban

	* [r8856] Adds subroutine to replicate a dbcsr matrix on all
	  processors.
	  Removes some unused error parameters from the core library
	  routines.

2009-08-21 18:15  vweber

	* [r8855] ubstitution sm -> dbcsr (3th step)

2009-08-21 13:17  mguidon

	* [r8854] Enables to dynamically distribute processing of bins
	  among involved threads in order to reduce load imbalance on one
	  single MPI process.

2009-08-20 17:48  vondele

	* [r8852] fix OMP build

2009-08-20 16:34  fschiff

	* [r8851] removed some loops

2009-08-20 15:12  fschiff

	* [r8850] 4 cp_fm_gemms less

2009-08-20 13:11  vweber

	* [r8849] ubstitution sm -> dbcsr (2th step)

2009-08-20 10:55  urban

	* [r8848] Removes obselete warnings, properly calculates
	  coordinates for block
	  storage for non-symmetric matrices, and forces data to remain
	  local (to
	  work with KG).

2009-08-20 07:58  urban

	* [r8847] Decripples SM->DBCSR conversion for non-symmetric
	  matrices and properly
	  handles unassociated data blocks.

2009-08-19 11:25  vondele

	* [r8846] speedup HFX avoiding generation of array temporaries

2009-08-19 09:14  urban

	* [r8845] Bugfix for empty matrix.

2009-08-19 08:28  urban

	* [r8844] Implements get_block_node by emulating the old block
	  node. Moves the SM/FM
	  conversion routines into the proper file.

2009-08-18 22:16  tlaino

	* [r8843] reproducibility patch for BLACS (Noam) + some refactoring
	  of duplications of variables (teo)

2009-08-18 16:29  jgh

	* [r8842] Geminals: kind and set radii

2009-08-18 13:06  vweber

	* [r8841] fixed unset ptr and removed dead code

2009-08-18 13:05  urban

	* [r8840] Makes the SM-compatibilty routines for dbcsr matrices
	  even more compatible
	  by using pointers.

2009-08-18 12:12  urban

	* [r8839] Enables SM-equivalent block storage in dbcsr matrices.

2009-08-18 12:06  vweber

	* [r8838] undef pointer

2009-08-18 11:06  vweber

	* [r8837] ubstitution sm -> dbcsr (1th step)

2009-08-17 14:19  urban

	* [r8836] Adds SM-compatible array methods for dbcsr matrices.

2009-08-17 12:34  vweber

	* [r8835] substitution sm -> dbcsr (0th step)

2009-08-17 11:31  tlaino

	* [r8834] prettify and reshuffle the order of few arguments (first
	  the ones used for the definition of subsequent arguments)

2009-08-17 11:03  jgh

	* [r8833] Geminals: new functionality and many bug fixes

2009-08-16 18:05  urban

	* [r8832] Bugfixes in dealing with transposed block storage.

2009-08-14 13:42  urban

	* [r8831] Fixes conversion to sparse_matrix_type to always create
	  upper-triangular
	  form for (anti)symmetric matrices.

2009-08-14 12:56  urban

	* [r8830] Updates the functions to get stored block coordinates and
	  other block info
	  and corrects flag usage for transposed block storage.

2009-08-14 09:22  vweber

	* [r8829] fix bugs

2009-08-13 16:19  mguidon

	* [r8828] Refactoring HFX code

2009-08-13 13:05  tlaino

	* [r8827] prettify

2009-08-13 12:10  urban

	* [r8826] Adds some emulation of current sparse_matrix_type:
	  dbcsr_add_block_node,
	  dbcsr_allocate_matrix, dbcsr_replicate_matrix_structure

2009-08-13 10:16  tlaino

	* [r8825] fix of the fix of the fix: if rij->0 (and i/=j) the
	  derivative is nan (0/0): Provide the correct limit instead.

2009-08-13 10:04  vondele

	* [r8824] restore the marcella's definition of coordination number

2009-08-13 10:01  tlaino

	* [r8823] fix of the bug fix: sumup contribution for r->0 and set
	  the correct limit of the derivative

2009-08-13 09:29  vweber

	* [r8822] added new functions, fix bugs and cleaning.

2009-08-13 08:43  marcella

	* [r8821] bug fix

2009-08-12 20:56  urban

	* [r8820] Adds mutable data based on btrees to allow greater
	  flexibility for data
	  access during construction and to better emulate current sparse
	  matrix
	  access.
	  Changes some interfaces to block access.

2009-08-12 09:17  vweber

	* [r8819] fixed unassociate ptr

2009-08-12 09:09  urban

	* [r8818] Bugfix: memory leak.

2009-08-11 16:40  urban

	* [r8817] Bugfix: matrix copy routine

2009-08-11 16:21  urban

	* [r8815] Different possibilites for obtaining data in a rank-2
	  pointer from matrix:
	  * 2003 Fortran pointer rank remapping (updates NAG arch file)
	  * fast but unsafe assumption that assumed-size arguments are not
	  copiedr
	  upon procedure call (if it works for MPI non-blocking
	  subroutines, why
	  not here...)
	  * safe but slow copy (with settable option even slower option to
	  force
	  contiguous pointers)
	  * can also obtain data as a rank-1 array

2009-08-11 14:05  vweber

	* [r8813] new loop (btriu) + added initialization before gemm.

2009-08-11 13:27  tlaino

	* [r8810] prettify

2009-08-11 13:15  mguidon

	* [r8809] New HFX code

2009-08-11 12:29  vweber

	* [r8808] added new iteration loop to the filter routine.

2009-08-11 11:58  vondele

	* [r8805] fix bug in coordination colvar (Holger Hesske)

2009-08-11 11:37  urban

	* [r8800] Type for 1-D and 2-D arrays of dbcsr matrices.
	  Bugfixes: iterator, checks in finalization, empty index
	  generation.

2009-08-10 20:49  urban

	* [r8798] Wraps DBCSR structure into an object to facilitate calls
	  from CP2K proper.
	  Bugfix of wrong specified INTENTs.

2009-08-10 13:59  urban

	* [r8794] Small fix: allows getting current block number in
	  iterator.

2009-08-10 13:29  urban

	* [r8792] Bugfixes: memory leaks, uninitialized values.

2009-08-10 13:02  urban

	* [r8791] Individual data block removal/insertion, iteration over
	  all blocks, and
	  some fixes.

2009-08-10 11:43  vweber

	* [r8790] sparse initial guess: replacing full matrices by sparce
	  ones. added filter.

2009-08-07 15:16  marcella

	* [r8788] fixing and cleaning in shell model

2009-08-07 13:30  tlaino

	* [r8785] prettify

2009-08-07 12:24  marcella

	* [r8784] cleaned unused

2009-08-07 12:23  marcella

	* [r8783] cleaned unused

2009-08-07 12:03  marcella

	* [r8778] improved Lanczos algorithm

2009-08-07 09:34  urban

	* [r8776] Convert new sparse matrix to current sparse matrix.

2009-08-06 10:54  vondele

	* [r8774] fix types, intent and get rid of a tmp

2009-08-06 10:02  vweber

	* [r8773] added a sparse initial guess for ot + reg.

2009-08-06 06:20  tlaino

	* [r8771] bug fix: this was crashing just with a sopt..

2009-08-05 23:02  tlaino

	* [r8770] prettify

2009-08-05 20:33  urban

	* [r8769] Bugfix: an extraneous variable prevented OpenMP
	  compilation.

2009-08-05 20:32  urban

	* [r8768] DBCSR matrices now support 4 datatypes (real and complex
	  in both single
	  and double precision types). To allow better OpenMP performance,
	  they
	  support completely independent data and index areas when building
	  a matrix.
	  
	  Adds absolute real_4 and real_8 kinds.
	  
	  Adds single/double precision versions to the message passing
	  interfaces
	  required for the DBCSR multi-precision code.
	  
	  Refactors some code.

2009-08-05 14:43  vweber

	* [r8767] avoiding reallocation.

2009-08-05 12:45  fschiff

	* [r8766] Load balancing on the HFX in core steps possible in RTP

2009-08-05 11:07  fschiff

	* [r8765] RTP and EMD can be restarted with EXT_RESTART.
	  INITIAL_WFN will be RT_RESTART

2009-08-05 10:45  tlaino

	* [r8764] prettify

2009-08-05 08:41  fschiff

	* [r8763] added reference, STEP_START_VAL from MD can now be used
	  with RTP

2009-07-29 10:58  jgh

	* [r8755] Bug fix for case with several atoms

2009-07-28 13:14  jgh

	* [r8753] Geminal code: optimizer and energy, bug fixes

2009-07-27 15:29  tlaino

	* [r8752] mmmhh... removed too much.. restored popt compilation..

2009-07-27 13:29  tlaino

	* [r8751] removing more unused variables

2009-07-27 13:29  tlaino

	* [r8750] moving function into the proper module

2009-07-27 10:26  tlaino

	* [r8749] removing unused variables

2009-07-27 10:10  tlaino

	* [r8748] removing more unused variables

2009-07-27 09:50  tlaino

	* [r8747] fix uninitialized failures variables

2009-07-27 09:29  tlaino

	* [r8746] use last_ener_id_nr and remove unused variables

2009-07-27 09:28  tlaino

	* [r8745] doxify k290 and move a mathlib routine in mathlib module

2009-07-27 06:04  tlaino

	* [r8744] fix instantiation for integer linked list

2009-07-24 23:08  tlaino

	* [r8743] removing more unused variables (improved prettify)

2009-07-24 15:02  fschiff

	* [r8742] bug fix

2009-07-24 14:34  vondele

	* [r8741] Confession?

2009-07-24 10:09  fschiff

	* [r8739] Changed the default quantity in extraoplation,
	  improvement for arnoldi,
	  in two RTP successive steps subspace is reused, might need
	  slightly tighter
	  EXP_ACCURACY

2009-07-24 06:54  vondele

	* [r8738] adjust file name format for large mols

2009-07-23 19:33  vondele

	* [r8737] optionally restrict the range of cubes dumped for
	  molecular states

2009-07-23 19:04  vondele

	* [r8736] make particles required for pw_to_cube

2009-07-23 17:51  vondele

	* [r8735] speed up localization based on the crazy rotations by
	  using a pade based exponential

2009-07-23 15:26  tlaino

	* [r8734] zeroing argument with INTENT(OUT) (used in a
	  self-referring loop)

2009-07-23 15:23  tlaino

	* [r8733] fix initialization

2009-07-23 12:54  vondele

	* [r8732] bug fix (florian)

2009-07-23 11:28  vondele

	* [r8731] add timeset

2009-07-23 11:27  vondele

	* [r8730] pade exponential of a real matrix (Florian)

2009-07-22 20:01  vondele

	* [r8729] tiny speedup

2009-07-22 19:58  vondele

	* [r8728] typo

2009-07-21 09:06  vondele

	* [r8727] wfn_mix lsd version

2009-07-21 07:36  vondele

	* [r8726] wfn_mix for generating restart files in an excited state

2009-07-20 12:35  jgh

	* [r8725] Exploit symmetry in geminal integrals.

2009-07-20 09:06  vondele

	* [r8724] typo

2009-07-18 14:57  jgh

	* [r8723] Derivatives of geminal integrals, initial version of
	  routines to calculate
	  full integral matrices

2009-07-17 08:33  vweber

	* [r8722] mixed precision technics for building / applying
	  preconditioner. provide dp accuracy.
	  fixed an overflow in distribution_optimize. added regs

2009-07-14 11:28  vweber

	* [r8721] fix bug introduiced by commit 2009-05-07 20:33
	  src/task_list_methods.F.
	  added a reg.

2009-07-10 11:44  vweber

	* [r8720] biftech: removed computation of the u-update and avoid
	  multiply when possible, fixed few bugs.

2009-07-09 10:30  vweber

	* [r8719] prepare for parallel biftech and added timers

2009-07-09 07:16  jgh

	* [r8718] Read in Cholesky factors instead of exponent matrix,
	  ensures positive definitness

2009-07-08 14:32  tlaino

	* [r8717] prettify

2009-07-08 14:12  vweber

	* [r8716] added BIF THEChnology (serial so far) and new
	  functionalities for the sparse library. removed ainv routines.

2009-07-08 11:46  urban

	* [r8715] Some more OpenMP work and reduction of data transferred
	  in the initial
	  data rescattering.

2009-07-07 22:45  tlaino

	* [r8714] skip INCLUDE in interface function definition

2009-07-07 22:44  tlaino

	* [r8713] fix intent for ks

2009-07-07 22:42  tlaino

	* [r8712] fix non-nullified pointer

2009-07-07 12:47  urban

	* [r8711] Adds OpenMP support in local multiplication of DBCSR
	  matrices.
	  
	  Also enables less memory use when building matrices.

2009-07-07 09:35  jgh

	* [r8710] prettify ....

2009-07-07 09:09  jgh

	* [r8709] Input for Geminal basis sets; 1 new regtest

2009-07-03 11:43  jgh

	* [r8704] Geminal integral code, test and performance routines.

2009-07-02 16:34  tlaino

	* [r8703] prettify

2009-07-02 16:07  mguidon

	* [r8702] Bug fix sparsity of overlap matrix

2009-07-01 22:59  tlaino

	* [r8701] fix uninitialized variable

2009-07-01 20:43  tlaino

	* [r8698] extend __NO_STATM_ACCESS to m_memory_details (pathscale
	  gives problems for STREAM access)

2009-07-01 19:15  tlaino

	* [r8697] prettify

2009-07-01 17:54  mguidon

	* [r8695] New storage format for replicated density/fock matrices

2009-07-01 17:41  urban

	* [r8694] Reduces memory usage and prevents reallocation in local
	  multiply.

2009-07-01 15:54  tlaino

	* [r8693] fix not initialized variables

2009-07-01 15:28  urban

	* [r8692] Prevents divide-by-zero in FLOP/s calculation.

2009-07-01 15:26  tlaino

	* [r8691] fix correct limit for coordination number (when rij->r0)
	  (function and derivative)

2009-06-30 16:06  urban

	* [r8689] Adds lookup methods for DBCSR matrices.
	  Updates matrix distributions in the preconditioner.

2009-06-30 14:12  tlaino

	* [r8688] prettify

2009-06-30 13:04  urban

	* [r8687] Local DBCSR multiply: joins consecutive blocks in rows of
	  the right matrix.

2009-06-30 09:20  tlaino

	* [r8686] bug fix

2009-06-29 14:43  urban

	* [r8683] Scalable parallel DBCSR multiplication (part 2/3):
	  More general (all matrices may have the same processor grid) and
	  changes
	  to matrix distributions. Adds desymmetrize operation.

2009-06-29 13:09  fschiff

	* [r8681] changed criterium for adaptive arnoldi. Uses EXP_ACCURACY
	  as target accuracy

2009-06-26 14:53  marcella

	* [r8679] change name of key

2009-06-26 14:02  tlaino

	* [r8678] recovering omp lost while cleaning (!?!)

2009-06-26 13:59  tlaino

	* [r8677] using error and fail if allocation fails..

2009-06-26 11:03  tlaino

	* [r8676] propagating error in pw_methods

2009-06-26 06:51  marcella

	* [r8675] bug fix

2009-06-26 06:28  tlaino

	* [r8674] fix output for moment of inertia eigenvalues + vib header

2009-06-25 17:14  tlaino

	* [r8672] bug fix HFX when using QMMM. Added 1 regtest.

2009-06-25 17:13  tlaino

	* [r8671] no need to nullify!

2009-06-25 12:44  urban

	* [r8670] Adds GFLOP/s output to sparse matrix multiply test.

2009-06-25 12:18  tlaino

	* [r8669] typo

2009-06-25 12:12  tlaino

	* [r8668] prettify

2009-06-25 07:12  urban

	* [r8667] DBCSR matrix multiply: enhanced scalability (limited to
	  square processor
	  grids).
	  Bugfix: correct transposed index generation when there is no
	  local data.

2009-06-23 08:48  tlaino

	* [r8665] bug fix + footer for replica files in BAND

2009-06-22 19:09  tlaino

	* [r8664] initialize error (where it was not..) with a default
	  one..

2009-06-22 18:19  tlaino

	* [r8663] collecting runtime informations in cp2k_info and
	  factorize CP2K header and footer in header.F

2009-06-22 17:31  tlaino

	* [r8662] accessing data from the right module

2009-06-22 17:13  tlaino

	* [r8661] -) skip ionode in banner (iw unit is enough)
	  -) new banner for QMMM and TDDFT
	  -) additional output infos for NEB

2009-06-22 14:27  tlaino

	* [r8658] prettify

2009-06-22 13:19  vondele

	* [r8657] add m_memory_details, to return some info about system
	  memory. Currently only implemented for systems with /proc/meminfo
	  and compilers with STREAM I/O.

2009-06-22 10:21  tlaino

	* [r8656] more info in BAND output

2009-06-22 09:47  tlaino

	* [r8655] bug fix

2009-06-22 08:03  vondele

	* [r8654] timeset per bin

2009-06-19 23:51  tlaino

	* [r8653] few major changes in NEB output: each NEB replica has now
	  an indipendent output not
	  shared with anyother replica (this applies to all possible files
	  produced for each
	  image): this allows for example to save .wfn indipendent files
	  along a NEB and adds
	  full flexibility w.r.t. the std cp2k functionalities.
	  Bug fix when reading xyz coordinate. Added new banner.

2009-06-19 15:39  fschiff

	* [r8652] padding for increasing matrix size

2009-06-19 13:50  tlaino

	* [r8651] allowing the possibility to change the distribution of
	  proc in the construction
	  of replica_env for BAND and VIBRATIONAL_ANALYSIS in order to have
	  procs
	  topologically close working on the same replica. Change the
	  default which is
	  now blocked instead of interlayered. Possibility to recover old
	  style through
	  keyword.

2009-06-19 11:33  vweber

	* [r8650] sparse package: more functionalites.

2009-06-19 09:07  marcella

	* [r8648] block-krylov as iterative diagonalization method
	  stm_images modified to take into account occupation numbers and
	  Fermi energy
	  qs_pdos bug fix

2009-06-19 08:11  fschiff

	* [r8647] bug fixes

2009-06-19 00:28  tlaino

	* [r8646] removing unimplemented, never regtested stuff

2009-06-18 23:57  tlaino

	* [r8645] fix for exclude 1-4 + regtests.

2009-06-18 18:46  tlaino

	* [r8644] Allow multiple_unit_cell also for connectivity. Added 1
	  regtest based on
	  AMBER.

2009-06-18 13:38  vweber

	* [r8642] added more functionalities for the sparse package.

2009-06-18 13:05  tlaino

	* [r8641] prettify

2009-06-18 11:38  fschiff

	* [r8640] serious bug fix for periodic moments with LSD (different
	  number of MOS)
	  improvement for RTP moments
	  adaptive arnoldi scheme implemented
	  improved parallel performance of arnoldi using combined matrices

2009-06-18 09:22  vondele

	* [r8639] avoid expensive cp_fm_init_random

2009-06-18 07:55  urban

	* [r8638] Enables arbitrary processor grid, further reduces link
	  contention, bugfix
	  in creating transposed indices, and updates file comments and
	  versions.

2009-06-17 15:21  tlaino

	* [r8637] bug fix for multiple_unit_cell when using scaled
	  coordinates in &COORD

2009-06-17 11:31  vondele

	* [r8636] use m_walltime instead of m_cputime

2009-06-16 15:31  mguidon

	* [r8635] Some Memory savings when screening on initial density
	  matrix

2009-06-16 10:58  vondele

	* [r8634] added cp_fm_gemm

2009-06-15 11:13  urban

	* [r8630] Fix omission: Reduce memory contention.

2009-06-14 16:27  urban

	* [r8629] Scalable parallel DBCSR multiplication (part 1/3):
	  reduces communication
	  steps to O(sqrt(#proc)).
	  
	  Also standardizes to MP_STD_REAL the recently added types in
	  message_passing.F and cleans up output of the recently added
	  DBCSR code.

2009-06-12 11:24  jgh

	* [r8628] Increase numerical stability of Whittaker function.
	  Avoids problems
	  with GAPW calculations with basis sets with very large exponents.
	  Some regtests reset for numerics.

2009-06-12 06:39  vondele

	* [r8626] test

2009-06-12 06:31  vondele

	* [r8623] add comment about MPI_ADDRESS_KIND

2009-06-12 06:20  vondele

	* [r8622] add 2009

2009-06-11 17:25  urban

	* [r8620] Bugfix: Fixes type mismatch in calls to some MPI
	  routines.

2009-06-11 11:16  jgh

	* [r8619] GAPW memory distribution (atomic integrals); regtests
	  reset for numerics; 1 new regtest

2009-06-11 08:40  urban

	* [r8618] Bugfix: Use MPI_INFO_NULL instead of 0 as argument in MPI
	  window creation.

2009-06-10 14:15  fschiff

	* [r8616] bug fix for higher order pade approximation and
	  prettyfication of the subroutine

2009-06-10 07:39  krack

	* [r8615] Complex variants added for matrix inversion

2009-06-09 22:08  tlaino

	* [r8614] d0 -> dp. Import dp from kinds where possible (1 central
	  definition of dp)

2009-06-09 21:13  tlaino

	* [r8613] #ifdef accepts just 1 MACRO (all the rest is skipped..).
	  is #if defined [..] what you meant?

2009-06-09 17:21  urban

	* [r8612] Prettify.

2009-06-09 17:10  urban

	* [r8611] Enables memory allocation through the MPI library.
	  
	  Fixes for the optional C interface to MPI memory allocation and
	  enables
	  MPI-allocated memory to be used for the DBCSR multiplication
	  routines.

2009-06-09 12:22  urban

	* [r8610] Adds OpenMP support to DBCSR local multiplication
	  
	  Makes B-trees OpenMP-safe, fixes a bug in B-tree addition, and
	  enables
	  OpenMP in local DBCSR multiplication.

2009-06-09 10:41  urban

	* [r8609] Bugfix: uninitialized value and bounds error

2009-06-09 08:38  mguidon

	* [r8608] Bug fix parallel distribution for MD

2009-06-08 21:02  tlaino

	* [r8606] additional bug fix for localization WANNIER CENTERS for
	  non-orthorombic cells (fix
	  bug in qs_loc_utils, restore original qs_loc_methods). These
	  fixes resolve all
	  known issues connected with the localization of wannier centers
	  in PBC with non-orthorombic
	  cells.

2009-06-08 09:46  tlaino

	* [r8605] bug fix

2009-06-08 09:15  tlaino

	* [r8604] CENTER_COORDINATES becomes a section with the possibility
	  to specify the
	  centering point. Default is kept like before: cell/2.

2009-06-08 07:10  jgh

	* [r8603] Reduce memory usage of GAPW; Some EPR regtests reset

2009-06-07 08:29  tlaino

	* [r8602] recovering single precision compilation (Axel)

2009-06-07 08:27  tlaino

	* [r8601] Regtests MUST be run always! No exceptions! fix 15 of
	  them failing after
	  last modification of this file.

2009-06-06 22:38  tlaino

	* [r8600] prettify

2009-06-06 17:19  fawzi

	* [r8599] more checks on topology read from psf, dealloc metadyn
	  rng only if allocated (avoid crash with energy run+langevin)

2009-06-06 15:24  vondele

	* [r8598] fix OMP compilation (OpenMP directives may not appear in
	  PURE or ELEMENTAL procedures)

2009-06-06 14:37  jgh

	* [r8597] GAPW bug fix for atoms with number of projectors < local
	  basis set; 1 regtest added

2009-06-06 07:02  tlaino

	* [r8596] skip factor of 2

2009-06-05 22:45  cjmundy

	* [r8595] Bug fixes for SCP-NDDO periodic.
	  
	  1) bug in charge constraint (scp_nddo_utils.F)
	  2) Made ewald_gks follow MIC
	  
	  Both ewalds agree rather well.
	  
	  ice regtests in SE/regtest/ice-scp*.inp

2009-06-05 20:58  tlaino

	* [r8592] allow for general order storage when doing eam

2009-06-05 17:11  tlaino

	* [r8591] bug fix localization WANNIER CENTERS for non-orthorombic
	  cells.

2009-06-05 14:05  krack

	* [r8590] Input keyword KERKER_MIN added; adapt output format on
	  purpose

2009-06-05 13:00  tlaino

	* [r8589] bug fix in parallel for molecular dipoles and adding an
	  unimplemented feature
	  when molecular dipoles are asked with nspin>1 that stops the
	  execution.

2009-06-05 10:28  tlaino

	* [r8588] more refactoring (grouping common pieces of code)

2009-06-05 10:13  jgh

	* [r8587] bug fix for self-interaction of DFTB dispersion energy; 2
	  regtests reset

2009-06-05 09:40  tlaino

	* [r8586] collecting common instructions

2009-06-05 07:45  tlaino

	* [r8585] enable REFERENCE, REF_POINT, PERIODIC for
	  MOLECULAR_DIPOLES.

2009-06-04 18:03  tlaino

	* [r8583] bug fix for self-interaction between periodic images for
	  DDAPC model.

2009-06-04 18:02  tlaino

	* [r8582] prettify

2009-06-04 16:16  jgh

	* [r8581] Bug fix for Grimme dispersion correction (see previous
	  commit for
	  nonbonded forcefields); two new regtests added

2009-06-04 14:11  tlaino

	* [r8580] severe bug fix: fix self-interaction between atomic
	  periodic images. This bug
	  affects all FIST (all potentials included manybody (EAM only))
	  runs where the
	  cutoff interaction radius is larger than cell/2. Reset 70 + 2
	  additional new ones.

2009-06-04 11:28  vweber

	* [r8579] fix bug with more mos than states (atomic guess).

2009-06-03 12:53  tlaino

	* [r8578] building FIST neighbor lists according the largest cutoff
	  between the FF and
	  the SR Ewald part (this allows to have always teh accuracy
	  requested for the
	  SR Ewald term [see &EWALD%EWALD_ACCURACY] ). Reset few regtests
	  and modify
	  others is order to avoid reset..

2009-06-02 19:17  tlaino

	* [r8577] skip forces evaluation in debug module

2009-06-02 17:43  tlaino

	* [r8576] using less than 32 characters name functions

2009-06-02 14:35  vweber

	* [r8575] removed the build of the shifted kinetic matrix within
	  the sainv routine.
	  fixed parallel bugs with the full sainv. reset 2 regs.

2009-06-02 11:13  tlaino

	* [r8574] prettify

2009-06-02 11:09  vweber

	* [r8573] (Urban Borstnik/Valery Weber) Adds framework for the
	  Stable Approximate
	  Inverse preconditioner, selectable with
	  PRECONDITIONER SPARSE_STABLE_APPROXIMATE_INVERSE
	  in the DFT/SCF/OT section, but with no regtest yet.
	  
	  (Urban Borstnik) Fixes known bugs and adds other changes; in
	  midst of
	  changing use of distribution and image distributions; adds a
	  BCSR_CODE
	  testing option to the DFT/QS input section.
	  
	  Added two regs, more coming.

2009-06-01 16:51  tlaino

	* [r8572] more on stress tensor for LR-SE

2009-06-01 15:29  tlaino

	* [r8571] more debug output

2009-06-01 12:00  tlaino

	* [r8570] bug fix derivatives for KDSO-D only

2009-06-01 10:06  tlaino

	* [r8569] towards stress tensor for periodic SE (not yet working).
	  Some refactorings and
	  fixes.

2009-05-29 23:15  tlaino

	* [r8568] severe bug fix: stress tensor n CELL_OPT and NPT_F not
	  symmetric. Bug affects cell optimizations and NPT_F with QS only.
	  reset 1 regtest.

2009-05-29 21:40  tlaino

	* [r8567] cosmetics

2009-05-29 20:17  tlaino

	* [r8566] bug fix

2009-05-29 18:07  tlaino

	* [r8565] bug fix stress tensor for multipoles. bug fix for the
	  evaluation of the numerical stress tensor (we compute it now
	  directly fully symmetric). added a control key do_stress in few
	  SE routines. added regtests for stress tensor multipoles. Added a
	  check in the debug routine.

2009-05-29 13:50  krack

	* [r8564] Moving input control keys for mixing to
	  scf_control%mixing to allow for an early readin and digest of the
	  input
	  parameters for mixing. Heavy weight data structures for mixing
	  remain in the mixing_store.

2009-05-29 10:08  tlaino

	* [r8562] Towards SPME and PME for MULTIPOLES : reordering
	  instructions in order to factor the Ewald method out of the
	  polarization.
	  Added output for convergence in POL FF (activate the ITER_INFO
	  print_key).

2009-05-29 05:24  tlaino

	* [r8561] bug fix

2009-05-28 23:11  tlaino

	* [r8560] more on printout cosmetics (some left parts..)

2009-05-28 22:16  tlaino

	* [r8559] major rewriting of the driver for numerical stress tensor
	  (now all force_eval provide
	  the numerical stress tensor and not just QS!). Rewriting of the
	  cp2k debug routine, with
	  an additional DEBUG_STRESS_TENSOR flag. Reorganization between
	  modules and cosmetics.

2009-05-28 17:30  tlaino

	* [r8558] added ALIGN_FRAMES to the neb section

2009-05-27 13:44  krack

	* [r8557] printout cosmetics

2009-05-27 08:43  krack

	* [r8555] bug fix for ADDED_MOS with LSD

2009-05-26 14:14  krack

	* [r8553] bug fix for LSD using g-space mixing

2009-05-26 13:04  krack

	* [r8551] Syncronize g- and r-space density after g-space mixing.
	  Faster convergence is obtained now, especially for Broyden
	  mixing.

2009-05-24 22:55  tlaino

	* [r8548] bug fix: missing initialization

2009-05-24 22:09  tlaino

	* [r8547] prettify

2009-05-24 21:22  marcella

	* [r8546] new colvar: gyration radius

2009-05-24 19:53  marcella

	* [r8544] - cleaning of the routines for the g-space mixing
	  - small change in the scf output to include the energy difference
	  after each scf iteration
	  - lowdin orthogonalization to be used to predict the new MOS
	  after a geo_opt or MD step in case smearing is active

2009-05-22 19:24  tlaino

	* [r8543] added MULTIPLE_UNIT_CELL : allows to perform the
	  calculation on a system which
	  is a multiple (in X,Y,Z) of the unit cell provided.. added 1
	  regtest.

2009-05-22 12:09  tlaino

	* [r8542] adding TARGET_LIMIT to the constraint/restraint growth

2009-05-22 10:47  tlaino

	* [r8541] modify output tag

2009-05-22 10:04  tlaino

	* [r8540] major rewriting of the ewald multipole bonded correction:
	  re-use already existing code (avoid useless duplications), avoid
	  too many useless allocation (re-use what is already available)
	  and introducing a proper variable to switch on the evaluation of
	  the correction. regtests ok.

2009-05-22 07:21  tlaino

	* [r8537] remove debug statements

2009-05-22 07:16  tlaino

	* [r8536] prettify

2009-05-22 07:15  tlaino

	* [r8535] plese use standard fortran..

2009-05-22 04:53  cjmundy

	* [r8534] Memory leak fix.

2009-05-21 22:43  cjmundy

	* [r8533] GKS taught me something

2009-05-21 22:34  cjmundy

	* [r8532] Intel will compile anything?! Bug fix found with G95

2009-05-21 20:49  cjmundy

	* [r8531] Dang-Change polarization (scalar isotropic
	  polarizability) now in.
	  Fist/regtest/*dangchang.inp are the regtests and sampling input.
	  Joost: POLARIZABILITY->DIPOLE, IPOL_SCP->POL_SCF in the input.

2009-05-21 07:17  tlaino

	* [r8529] recovering single precision compilation (Axel)

2009-05-21 06:50  tlaino

	* [r8528] fix typos (Axel)

2009-05-20 15:38  vweber

	* [r8527] added optional variable for triplet calculations

2009-05-20 12:02  vweber

	* [r8526] pso term for issc, fix some bugs related to gapw second
	  derivative of the XC. added regs.

2009-05-19 21:10  tlaino

	* [r8524] support for XTL file format (MSI)

2009-05-19 17:02  jgh

	* [r8523] Updates for atomic code: calculate response basis
	  contractions,
	  allow for different weights for different electronic states and
	  methods
	  in optimization (pseudopotential and basis)

2009-05-19 09:36  vweber

	* [r8520] fix bug (wrong skipout) and add few more regs for tddft.

2009-05-19 06:33  tlaino

	* [r8519] modify header

2009-05-18 21:31  tlaino

	* [r8518] SE LR correction (4th part) : several bug fixes

2009-05-14 14:01  vweber

	* [r8517] added regs for tddft.

2009-05-13 08:49  tlaino

	* [r8515] prettify

2009-05-13 07:56  vweber

	* [r8514] (Urban Borstnik) Fixes divide-by-zero issue in printout
	  with short runs.

2009-05-12 20:10  vweber

	* [r8513] (Urban Borstnik) Fixes memory leaks and multiple frees.

2009-05-12 16:14  tlaino

	* [r8512] prettify

2009-05-12 16:13  tlaino

	* [r8511] restoring NAG compiler (which was broken after the last
	  commit)

2009-05-12 16:12  tlaino

	* [r8510] module name must match filename (this is case sensitive!)

2009-05-12 15:38  vweber

	* [r8509] (Urban Borstnik) Initial distributed block compressed
	  sparse row matrix
	  routines to support linear scaling.
	  (Some fixed come shortly.)

2009-05-11 17:52  tlaino

	* [r8508] SE LR correction (3rd part) : bunch of bug fixes..

2009-05-10 20:40  tlaino

	* [r8507] SE LR correction (2nd part)

2009-05-08 15:18  jgh

	* [r8506] prettify

2009-05-08 14:38  jgh

	* [r8505] New functionality in Atom code (pseudo potential
	  optimization); some new defaults and bug fixes;
	  All regtests for atom code reset

2009-05-07 18:33  vondele

	* [r8504] use rank reordering on the multigrids to improve load
	  balancing (Iain Bethune).

2009-05-07 18:32  vondele

	* [r8503] cite aspc where used

2009-05-07 13:06  tlaino

	* [r8502] prettify

2009-05-07 13:02  fschiff

	* [r8501] bug fix in pade_only imaginary (thanks to Matthias)
	  added Arnoldi subspace algorithm for the matrix exponential

2009-05-07 12:07  tlaino

	* [r8500] SE LR correction (1st part)

2009-05-07 07:07  tlaino

	* [r8499] taper for SE LR correction

2009-05-06 17:37  tlaino

	* [r8498] additional NL for SE LR correction

2009-05-06 14:48  tlaino

	* [r8497] reorganization and splitting of the nddo_module in
	  different modules. Getting rid of one level call
	  in qs_semi_empirical_hamiltonian..
	  Preliminary framework for NDDO-long range correction (not yet
	  there)

2009-05-04 19:33  tlaino

	* [r8496] fix iteration index for optimizers (cell, geo and rot)
	  when restarting (bug reported
	  by Maricarmen).

2009-05-04 12:26  krack

	* [r8495] bug fix for optional argument

2009-05-02 12:30  vondele

	* [r8494] fix intent

2009-05-02 12:27  vondele

	* [r8493] fix number of arguments

2009-05-01 11:47  jgh

	* [r8492] NDDO pseudodiagonalization using block-Jacobi (1
	  regtest);
	  NDDO stress tensor (1 regtest), might need further testing

2009-04-30 20:31  vweber

	* [r8490] added wrapping for 3 centers integral. reset few regs.

2009-04-30 18:41  vondele

	* [r8489] fix definition to match call sites

2009-04-30 09:53  vweber

	* [r8488] another array initalizations

2009-04-30 07:31  vweber

	* [r8487] fix array inits.

2009-04-29 22:12  tlaino

	* [r8485] clean and pretty

2009-04-29 15:07  cjmundy

	* [r8484] Modifications due to GKS. Now both multipole ewalds give
	  same results to
	  micro-hartree.

2009-04-27 16:11  jgh

	* [r8480] SCP NDDO dispersion terms: add energy to total energy
	  correctly

2009-04-27 14:59  jgh

	* [r8479] Fix bug (typo) in Slater screening integral derivatives

2009-04-27 11:17  jgh

	* [r8478] SCP NDDO/EWALD: initialize pointer

2009-04-27 07:42  vweber

	* [r8477] fix bug indexing array.

2009-04-26 16:09  jgh

	* [r8476] Bug fix for NDDO LSD with EWALD

2009-04-25 20:40  marcella

	* [r8474] diis excluded when mixing in reciprocal space is active
	  different options for cholesky reduce to improve scaling

2009-04-25 08:41  tlaino

	* [r8473] fix dangling pointer

2009-04-24 16:33  tlaino

	* [r8472] prettify and clean

2009-04-24 15:43  jgh

	* [r8471] Bug in self-terms of EWALD_GKS (still needs work)

2009-04-24 14:21  mguidon

	* [r8470] Many changes related to parallel scaling. New way of
	  distributing the load, introducing
	  pair lists in order to quickly walk through set_pair
	  intercations, removing expensive
	  hash calls... All in all this scales now up to 64k processes :-).
	  Still many things left to be cleaned...

2009-04-23 07:42  fschiff

	* [r8468] forgot to commit a file (sorry for that)

2009-04-22 17:16  tlaino

	* [r8466] prettify

2009-04-22 16:05  fschiff

	* [r8464] first implementation of a nonperiodic laser_fild (be
	  carful when using,
	  it is not symmetric around the center of the cell, will be
	  changed)
	  added restart of the complex RTP wavefunction
	  new estimate for the 2-norm of the matrix used for the
	  exponential

2009-04-22 13:00  vweber

	* [r8463] add new integral code for the l and p operators.

2009-04-21 06:00  tlaino

	* [r8462] CUDA: Initial support for double precision BLAS (Ben)

2009-04-18 13:15  jgh

	* [r8461] prettify

2009-04-18 13:01  jgh

	* [r8460] Wavefunction extrapolation for methods with unit overlap
	  matrix; 4 new regtests;
	  reset 57 regtests

2009-04-17 17:16  tlaino

	* [r8459] prettify

2009-04-17 14:58  vweber

	* [r8458] get rid of the do i=1,-1 and rm a useless write

2009-04-17 14:31  vweber

	* [r8457] reduce memory usage while building chi (one center),
	  reset regs.

2009-04-17 12:46  vweber

	* [r8456] fix out of bound while writting restart file.

2009-04-17 11:45  vondele

	* [r8455] introduce technology for rank remapping of the multigrids
	  (Iain Bethune)

2009-04-17 09:28  jgh

	* [r8454] SCP NDDO dispersion added

2009-04-16 18:37  tlaino

	* [r8452] fix initialization

2009-04-16 18:06  tlaino

	* [r8451] prettify

2009-04-16 18:06  tlaino

	* [r8450] bug fix in parallel

2009-04-16 15:58  vondele

	* [r8449] fix io bug in parallel

2009-04-16 14:23  vondele

	* [r8448] fix end of record error (gfortran) by providing an
	  explicit format

2009-04-16 12:20  vweber

	* [r8446] fix diassociated pointer

2009-04-16 09:03  marcella

	* [r8445] cube files are now by default overwritten (MO_CUBES,
	  WANNIER_CUBES, TOT_DENSITY_CUBE,...),
	  in order to have them appended instead, one can add the
	  keyword APPEND in the corresponding print subsection

2009-04-16 07:01  jgh

	* [r8444] SCP dispersion moved to energy calculation

2009-04-15 12:36  vondele

	* [r8443] more potential overflows

2009-04-15 12:28  vondele

	* [r8442] fix a potential overflow

2009-04-15 12:16  vondele

	* [r8441] fix a bug in image cell generation of large systems or
	  short R_c

2009-04-15 07:55  tlaino

	* [r8440] remove duplications, refactoring, fix few more loop
	  indexes, added minimal doxygen info

2009-04-14 18:57  tlaino

	* [r8439] speed-up int_gks: using the fortran hierarchy (by
	  column). Collecting common
	  pieces of codes: avoid duplications. Reset 1 for numerics
	  (1.0E-14)

2009-04-14 18:07  tlaino

	* [r8438] clean duplications

2009-04-14 16:54  jgh

	* [r8436] NDDO: make it possible to combine PNNL/SLATER integrals
	  with standard EWALD code

2009-04-14 16:13  tlaino

	* [r8435] get rid of the alternative integral order and use the
	  CP2K (MOPAC) instead.
	  Make GKS compatible to d-orbitals (array sizes). Avoid too many
	  zerofications: speed-up.

2009-04-14 13:36  tlaino

	* [r8434] stop if ewald_gks is requested with other
	  parameterization than PNNL

2009-04-14 13:36  tlaino

	* [r8433] added debug

2009-04-14 12:14  tlaino

	* [r8432] prettify

2009-04-14 11:08  tlaino

	* [r8431] setting default integral_screening to Slater for PNNL and
	  print a warning
	  message in case of different input setup

2009-04-14 10:50  tlaino

	* [r8430] remove duplications

2009-04-14 10:35  tlaino

	* [r8429] building multipoles only when strictly required..

2009-04-14 09:42  tlaino

	* [r8428] bug fix. refactoring + reset

2009-04-13 17:51  tlaino

	* [r8426] clean

2009-04-13 17:38  tlaino

	* [r8425] prettify

2009-04-13 12:19  cjmundy

	* [r8423] Committed a refined parameter set by GKS. Now benchmarks
	  are established.

2009-04-13 10:17  jgh

	* [r8422] Set consistent cutoff for EWALD_GKS

2009-04-13 10:16  jgh

	* [r8421] prettify

2009-04-13 09:26  jgh

	* [r8420] Atom code: prepare for parameter optimizations

2009-04-13 09:17  jgh

	* [r8419] SE Code initialization: special case for hydrogen with sp
	  basis

2009-04-13 08:47  jgh

	* [r8418] Allow for DKH(0) and DKH(1) calcuations in atom code;
	  reset 2 tests

2009-04-10 15:13  jgh

	* [r8416] Calculate local pseudopotential on real space grid, 2 new
	  regtests;
	  Virial not yet working, default is to use analytic integrals.

2009-04-10 13:54  tlaino

	* [r8415] clean

2009-04-10 13:50  tlaino

	* [r8414] prettify

2009-04-10 12:49  jgh

	* [r8413] New NDDO integral routine using Slater screening

2009-04-10 12:45  jgh

	* [r8412] New integral routines using Slater type screening and
	  Ewald; 2 regtests

2009-04-08 14:05  marcella

	* [r8411] bug fix

2009-04-08 09:45  tlaino

	* [r8409] prettify

2009-04-08 08:33  vweber

	* [r8408] deallocate matrices when possible, and get rid of the
	  last get_submatrix.

2009-04-08 05:40  marcella

	* [r8406] added a second type of broyden mixing

2009-04-08 05:39  marcella

	* [r8405] a second type of broyden mixing is added

2009-04-08 05:36  marcella

	* [r8404] bug fix in force_rho_core on atoms of gpw_type in a gapw
	  calculation, one regtest added

2009-04-06 21:44  marcella

	* [r8398] pretty

2009-04-06 21:36  marcella

	* [r8396] the syevx (partial spectrum) scalapack routine is
	  replaced by
	  syevd (full spectrum), which in most of the cases seems to be
	  faster and to require significantly smaller memory allocation.
	  Some regtests need to be resetted

2009-04-03 11:46  tlaino

	* [r8395] prettify

2009-04-03 08:53  vweber

	* [r8394] added flushing.

2009-04-03 08:04  vweber

	* [r8393] fixed diassociated pointer arising with gapw and no soft
	  function.

2009-04-02 15:01  vweber

	* [r8390] fix initialization of a variable.

2009-04-02 13:56  vweber

	* [r8389] spherical harmonics: reduced mem usage. reset some regs.

2009-04-01 16:47  jgh

	* [r8387] Remove debug code

2009-04-01 16:24  jgh

	* [r8386] Speedup of local pseudopotential terms; reset 60 regtests

2009-04-01 12:42  mguidon

	* [r8385] Broadcast the data for trucanted HFX instead of loading
	  it from file

2009-04-01 06:07  tlaino

	* [r8384] bug fix

2009-03-31 21:07  tlaino

	* [r8383] fix initialization

2009-03-31 20:51  mguidon

	* [r8382] Added possibility to run alpha and beta spin seperately,
	  which saves half of the memory used for the density/fock
	  matrices. New routines for calculating the exchange energy.
	  Removed expensive way of calculating the sparsity of the KS
	  matrix. Adiabatically rescaled hybrids still need refactoring

2009-03-31 15:29  vondele

	* [r8381] Consistent printout, no printing for forces, speedup
	  message passing of list/mat stats.

2009-03-31 14:59  vondele

	* [r8380] join mp_max

2009-03-31 14:53  vondele

	* [r8379] added mp_max_8iv

2009-03-31 13:57  vondele

	* [r8378] join mp_sum of scalars

2009-03-30 15:29  vondele

	* [r8377] add timings to derivate code

2009-03-30 14:13  vondele

	* [r8376] prettify

2009-03-30 13:21  vondele

	* [r8375] updated get_is_assoc_atomic_block

2009-03-30 10:16  vondele

	* [r8374] get_is_assoc_atomic_block: refactored communication step
	  out of hfx_energy_potential, to b further rewritten

2009-03-30 07:17  vondele

	* [r8373] sync before/after a ring of isendrecv to speedup the
	  message passing, and avoid flooding
	  of busy nodes by messages from a chain of idle neighbors

2009-03-27 19:06  vondele

	* [r8371] fix potential overflow

2009-03-27 17:05  mguidon

	* [r8369] Only store upper diagonal fock and density matrices and
	  introduce some caching to make index conversion cache friendly

2009-03-27 11:17  vondele

	* [r8368] bug fix accounting of sync

2009-03-26 21:28  tlaino

	* [r8367] skipping comments for pre-processing (bug reported by
	  Joost)

2009-03-26 17:16  tlaino

	* [r8366] more info during debug runs

2009-03-26 12:08  tlaino

	* [r8365] avoid name (realspace_grid_descriptor_p_type) longer than
	  31 (NAG prints thousands
	  of warnings..): therefore, moving
	  realspace_grid_descriptor_p_type into
	  realspace_grid_desc_p_type and for consistency
	  realspace_grid_descriptor_type into
	  realspace_grid_desc_type.

2009-03-26 11:25  vondele

	* [r8364] and prettify

2009-03-26 11:15  vondele

	* [r8363] cleanup: remove rs_pools and split rs_grid_type (Iain
	  Bethune)

2009-03-26 10:09  tlaino

	* [r8362] enabling max_multipole for SE + regtest

2009-03-26 10:01  tlaino

	* [r8361] prettify

2009-03-26 09:13  vondele

	* [r8360] force flush at end

2009-03-25 19:18  vondele

	* [r8359] fix overflow for large systems

2009-03-25 15:24  tlaino

	* [r8358] expand $VAR as well as ${VAR}

2009-03-25 14:26  mguidon

	* [r8357] Hash-Table Index based on natom^3 instead of natom^2,
	  which leads to better load balance for huge number of CPUs

2009-03-25 11:56  mguidon

	* [r8356] Not needed any longer

2009-03-25 11:51  mguidon

	* [r8355] Cleanup, removed everything related to Monte Carlo load
	  balancing

2009-03-25 09:57  tlaino

	* [r8354] consistent exchange cutoff for SE; more debug info and
	  adding the nuclear
	  multipolar contribution for SE to the proper energy term. Reset 2
	  for numerics

2009-03-25 07:51  tlaino

	* [r8353] prettify

2009-03-24 09:15  fschiff

	* [r8351] parallel bug fix, removed remaining debug write
	  statements

2009-03-23 17:18  jgh

	* [r8350] Bug fix for GAPW (more local bsf than projectors), added
	  one regtest

2009-03-23 10:59  tlaino

	* [r8349] prettify

2009-03-23 10:45  vweber

	* [r8348] added restart for the linear response.

2009-03-23 10:44  fawzi

	* [r8347] skip mp_sum for 0 dim arrays so that bogus checks of
	  array overlap in mpi
	  implementations like impi do not fail if the compiler uses a
	  unique object for
	  all 0 dim arrays.

2009-03-20 09:01  tlaino

	* [r8346] bug fix

2009-03-20 09:00  tlaino

	* [r8345] prettify

2009-03-19 18:41  tlaino

	* [r8344] providing the possibility to give in input cell angles,
	  useful for specifying
	  non-orthorombic cells together with the ABC keyword.

2009-03-19 15:55  vondele

	* [r8343] refined the way the radius is computed / fitted. In
	  particular alway keep it positive

2009-03-19 13:22  vweber

	* [r8342] merged linres and ot preconditioners. reset few regs.

2009-03-19 13:16  vondele

	* [r8341] improve Powell fitting of screening functions.

2009-03-19 08:59  tlaino

	* [r8338] bug fix

2009-03-19 00:17  tlaino

	* [r8337] prettify

2009-03-19 00:02  tlaino

	* [r8336] better choice of defaults for Coulomb and Exchange
	  cutoffs (SE).

2009-03-18 20:58  vondele

	* [r8335] improved hfx load balance. Still needs serious
	  refactoring.

2009-03-18 19:39  vondele

	* [r8334] Fix OMP compile

2009-03-18 16:14  tlaino

	* [r8332] bug fix: using QMMM with AMBER native prmtop files
	  (reported by Marius Retegan)

2009-03-18 15:26  tlaino

	* [r8331] changing default units (from bohr -> angstrom)

2009-03-18 15:16  marcella

	* [r8330] deleted useless and disturbing change in the restart
	  file, sorry

2009-03-18 13:24  fschiff

	* [r8328] bug fix, wrong index for paralell runs

2009-03-18 12:51  vondele

	* [r8327] Fix value of param for orig B97 (Yun Ding).

2009-03-18 09:40  tlaino

	* [r8326] prettify

2009-03-18 09:13  fschiff

	* [r8323] some more rtp and EMD, new propagators, new matrix
	  exponential,
	  enabled periodic dipole, aspc extrapolation for all methods

2009-03-17 16:28  tlaino

	* [r8322] deleting the several RC_* keywords in SE and introducing
	  several section for
	  COULOMB, EXCHANGE and SCREENING. Added different taper functions
	  for the
	  several terms and reset 2 regtests for numerics.

2009-03-17 15:59  vweber

	* [r8321] issc: added response for the fc, pso and efg
	  perturbations. added regs for the responses.

2009-03-16 22:01  marcella

	* [r8320] bug fix

2009-03-16 20:57  vweber

	* [r8319] added electric field gradient integrals, rm useless
	  arrays, reset a reg.

2009-03-16 16:01  marcella

	* [r8314] Some extension to the calculation of STM images, to allow
	  the use of different
	  tip orbitals

2009-03-16 16:00  marcella

	* [r8313] Some changes in the available mixing procedures, as used
	  in QS calculations together
	  with diagonalization algorithms. A new section DFT%SCF%MIXING
	  should be used to specify type of mixing and related parameters.
	  Note: the old keyword DFT%SCF%MIXING is no longer valid.
	  The available mixing procedures are at the moment the direct
	  density-matrix mixing (default),
	  the Kerker damping, the Pulay mixing, and the Broyden mixing. The
	  latter three are performed in
	  reciprocal space, therefore are applied only to the density
	  already on the grid.
	  When OT is used, no mixing is active. Some regtest are added

2009-03-16 12:53  tlaino

	* [r8310] prettify

2009-03-16 12:49  jgh

	* [r8309] Atom code: small updates and bug fixes; added 103
	  regtests

2009-03-16 12:46  tlaino

	* [r8308] adjust order of arguments and propagate error

2009-03-13 15:32  vweber

	* [r8305] issc: added paramagnetic spin-orbit integrals and generic
	  Obara-Saika reccurence relation for any integral that can be
	  derived from 1/r_C. added checksum to the reg.

2009-03-12 15:48  tlaino

	* [r8304] bug fix

2009-03-12 14:06  vweber

	* [r8303] issc: added fermi contact integral and 1 reg.

2009-03-11 22:52  tlaino

	* [r8302] KDSO-D SE integrals + regtests

2009-03-11 18:10  tlaino

	* [r8301] bug fix [CP2K:1858]

2009-03-09 18:09  tlaino

	* [r8300] split molden routine into a separate module and dumping
	  in case of standard
	  vibrational analysis all frequencies (either positive or
	  negative)

2009-03-09 10:05  krack

	* [r8299] Bibliography updated and citations added

2009-03-06 15:20  mguidon

	* [r8297] LSD bug fix

2009-03-06 14:00  vondele

	* [r8296] cleanup: get rid of useless and untested map_paa

2009-03-06 13:18  vweber

	* [r8295] possibility to select states for the response
	  calculation. added regs.

2009-03-06 11:39  tlaino

	* [r8293] bug fix for LANGEVIN on COLVARS. Added a print_key to
	  print COLVARS temperature
	  in an extended scheme. 1 more regtest (Fabio Sterpone and Teo)

2009-03-05 20:04  tlaino

	* [r8291] prettify

2009-03-05 16:41  fschiff

	* [r8289] first implementation of Ehrenfest MD and real time
	  propagation in cp2k

2009-03-04 18:24  mguidon

	* [r8288] The total available memory for integral storage per
	  MPI-process is now shared by all openMP threads

2009-03-04 13:34  fawzi

	* [r8287] uniforming length of strings in character array

2009-03-04 12:44  fawzi

	* [r8286] b97 xc functional

2009-03-03 18:22  tlaino

	* [r8285] more on semiempirical periodic and reorganization of
	  structures.

2009-03-03 15:29  mguidon

	* [r8284] Some additional printing, more clear MAX_MEMORY handling

2009-03-03 11:12  tlaino

	* [r8283] prettify

2009-03-03 10:53  vweber

	* [r8282] removed unused arrays. reduced printing for regs

2009-03-02 11:44  jgh

	* [r8277] Some speedups for erfc integrals

2009-02-27 15:50  tlaino

	* [r8276] prettify

2009-02-27 13:08  hforbert

	* [r8275] Changing replica row/col order for forces in PINT so
	  topologically close processors should work on the same replica
	  (other methods that use replica may want to give this a try as
	  well)

2009-02-27 10:11  vweber

	* [r8274] new gauge build. clean some printing. fix memory leaks.
	  modify 2 regs for the new gauge

2009-02-26 20:59  tlaino

	* [r8273] bug fix for IO

2009-02-26 18:58  tlaino

	* [r8271] new colvar: angle between two planes + 2 regtests

2009-02-26 16:28  fawzi

	* [r8269] adding key to print the electronic kinetic energy

2009-02-26 16:26  fawzi

	* [r8268] improving flushing on termination when flush_should_flush
	  is false

2009-02-26 11:59  tlaino

	* [r8265] bug fix charge for non-periodic

2009-02-26 09:03  vondele

	* [r8264] fix omp statements

2009-02-26 08:19  jgh

	* [r8263] Refactoring of qs_core_hamiltonian;
	  New scheme to calculate all electron nuclear potential (ERFC),
	  we use the same method as for the local PP, meaning the 3-center
	  lists are gone, except for the DKH and FULL_GAPW cases.
	  99 regtests were reset

2009-02-25 16:07  vondele

	* [r8262] reorganise print order of HFX_MEM_INFO

2009-02-25 14:34  tlaino

	* [r8261] recursive IO not allowed

2009-02-25 11:59  vondele

	* [r8260] a couple more timesets

2009-02-25 11:03  vondele

	* [r8259] qs_init_subsys does not need to be public

2009-02-24 23:42  tlaino

	* [r8258] bug fix evaluation of charge in MOMENTS (bug reported by
	  Ondrej)

2009-02-24 13:07  mguidon

	* [r8257] minor openMP issue

2009-02-23 16:37  tlaino

	* [r8255] prettify

2009-02-23 14:04  jgh

	* [r8254] GAPW rewrite: 2nd step;
	  Improved code for projector build up (up to 10x speedup);
	  Compressed storage of projector matrices -> reduced memory usage;
	  Matrix multiplies using compressed storage -> speedup;
	  Compressed matrix size for rho_atom current matrices -> reduce
	  memory and speedup;
	  Some additional intermediate memory needed in rho_atom for
	  density may increase
	  memory for special cases (MOLOPT basis).
	  50 regtests reset.

2009-02-23 13:53  hforbert

	* [r8253] Added comment about possible MPI standard violation

2009-02-23 13:46  hforbert

	* [r8252] For PINT be more flexible with the number of CPUs per
	  bead and change default to try to determine a reasonable value

2009-02-23 12:34  vweber

	* [r8251] makeup.

2009-02-23 07:30  tlaino

	* [r8249] restoting single precision compilation (Axel)

2009-02-22 16:58  vondele

	* [r8248] add more timings, to trace memory usage

2009-02-22 16:57  vondele

	* [r8247] prettify

2009-02-21 15:16  jgh

	* [r8246] Activate DKH code also for atomic calculations

2009-02-21 15:06  jgh

	* [r8245] Relativistic DKH code (up to order 3) Jens Thar;
	  Regtests will follow shortly.

2009-02-20 22:43  vweber

	* [r8244] new projection: avoiding soft-soft contribution for the
	  response. reducing the number of dgemm when possible (juerg).
	  reset some regs.

2009-02-18 22:22  vweber

	* [r8241] new projection: mv symmetric add outside loops. use dger
	  instead of 2 dgemm when possible.
	  better align routines for the screened pair list. reset few regs.

2009-02-18 20:38  vondele

	* [r8240] reapply FFT patch with buf fixed (non-linked-in FFTW
	  properly detected)

2009-02-18 16:32  vondele

	* [r8239] revert FFT patch, for fixing a bug.

2009-02-18 16:30  tlaino

	* [r8238] removing commented OMP instructions

2009-02-18 16:20  tlaino

	* [r8237] EAM reference

2009-02-18 15:56  tlaino

	* [r8236] EAM for alloys (Daniele e teo) + regtest

2009-02-18 10:52  vweber

	* [r8235] linear solver: check for convergence already with the
	  very first residual. reset some regs.

2009-02-17 20:06  tlaino

	* [r8234] bug fix (cell_opt for triclinic cells) + propagating
	  error..

2009-02-17 12:47  vondele

	* [r8233] Another FFT patch (Iain) to have planning also for 3D
	  FFTs

2009-02-17 09:45  krack

	* [r8232] Print occupation numbers (PDOS file)

2009-02-16 21:30  vweber

	* [r8231] New projection: screening the projectors, some speedup
	  (up to 2x). reset few regs.

2009-02-16 16:55  krack

	* [r8230] update MO occupations even if OT is active

2009-02-16 14:59  krack

	* [r8229] * print Fermi instead of HOMO energy to provide an useful
	  output even if smearing is applied
	  * print additional header line to describe each column
	  * print atomic kind name instead of kind number which is already
	  specified with the PDOS file name

2009-02-16 12:47  marcella

	* [r8227] dos projected over a list of atoms specified from input

2009-02-13 15:00  vweber

	* [r8225] fixed previously introduiced printing bug (Samuele).

2009-02-11 21:22  vweber

	* [r8224] rm printing+cleanup

2009-02-11 18:57  vweber

	* [r8222] speedup gauge build (2-3x).

2009-02-11 13:33  vondele

	* [r8221] count maxvals, fix unit for powell, remove syncs, improve
	  format (Manuel & Joost)

2009-02-11 12:11  vondele

	* [r8220] a similar speedup for qs_ks_atom

2009-02-11 09:45  vondele

	* [r8219] prettify

2009-02-11 09:43  vondele

	* [r8218] clarify error

2009-02-10 20:54  vweber

	* [r8217] new projection: up to 10x faster (blanc). timing slightly
	  worst that old code.

2009-02-10 11:43  mguidon

	* [r8216] Fix memory leak

2009-02-10 10:12  vweber

	* [r8215] first step toward indirect spin-spin coupling.

2009-02-09 12:50  vondele

	* [r8214] fixes to the hfx load balancing routines.

2009-02-09 11:53  jgh

	* [r8213] First step in GAPW code rewrite;
	  New projecton routine (similar to PPNL);
	  Solves a long standing bug (small cells) and uses less agressive
	  screening;
	  Uses less memory (some 3-center lists are gone) but is slower
	  than the old code;
	  53 regtests reset.

2009-02-08 14:36  vondele

	* [r8212] rename a variable for clarity (number_of_integrals ->
	  number_of_atomic_quartets)

2009-02-06 10:21  vweber

	* [r8205] rm useless keywords. cleaned the nmr/epr regs.

2009-02-05 18:32  vondele

	* [r8204] report maximum memsize over all mpi tasks

2009-02-05 14:08  vweber

	* [r8203] new cg. reset nmr/epr regs.

2009-02-05 14:07  tlaino

	* [r8202] prettify

2009-02-05 11:22  vondele

	* [r8201] some more mem info

2009-02-05 11:15  vondele

	* [r8200] bug fix for many cores

2009-02-04 19:51  fawzi

	* [r8199] delta_dipole without jumps in the periodic case, and
	  always correct (even if other dipoles are calculated)

2009-02-04 19:11  tlaino

	* [r8198] prettify

2009-02-04 19:10  tlaino

	* [r8197] allowing the parsing of XYZ files that miss the title..

2009-02-04 19:10  tlaino

	* [r8196] several bug fixes for reading "non-standard" amber
	  structures ;-)

2009-02-04 19:09  tlaino

	* [r8195] enabling the parsing of white lines..

2009-02-04 13:59  mguidon

	* [r8194] Remove hard coded threshold in calculation of pair charge
	  distribution radii

2009-02-03 19:43  mguidon

	* [r8189] Removed all screening matrices and replaced them by
	  functions. A lot of optimizations

2009-02-03 18:00  hforbert

	* [r8188] Added subroutine to write the trajectory of PINT runs
	  (positions only atm)
	  cleaned pint_calc_f a bit
	  changed/hopefully fixed output of centroid trajectory if the
	  format is not xmol

2009-02-03 14:41  vweber

	* [r8187] makeup

2009-02-03 13:51  hforbert

	* [r8186] removed leftover x array in splines, x values have been
	  always been a
	  uniform grid for the past couple years. the variable x part of
	  the spline
	  routines was removed a long time ago...

2009-02-03 13:19  hforbert

	* [r8185] overflow in lcm function should now only happen if the
	  true result would overflow

2009-02-03 06:34  marcella

	* [r8180] added a new form of walls for colvars

2009-02-02 21:36  tlaino

	* [r8179] bug fix : fixed atoms + metadynamics

2009-02-02 20:25  vondele

	* [r8178] TRACE_MASTER, for tracing only on the master node

2009-02-02 20:00  vondele

	* [r8177] get total memory used by accessing /proc/self/statm

2009-02-02 16:40  tlaino

	* [r8176] bug fix input preprocessor

2009-02-01 12:49  vondele

	* [r8175] fix an OMP bug

2009-02-01 10:30  marcella

	* [r8174] bug fix in comparing names of kinds

2009-01-30 17:38  vondele

	* [r8173] faster (parallel) FFTs by using improved plans with FFTW.
	  See FFTW_PLAN_TYPE in &GLOBAL.

2009-01-30 13:27  vweber

	* [r8172] added timeset.

2009-01-30 11:15  vondele

	* [r8171] prettify

2009-01-29 12:46  vweber

	* [r8167] rm useless reallocation and zeroing. rm not used
	  keywords. cleaning.

2009-01-28 18:48  tlaino

	* [r8166] setting default POTENTIAL to GTH (to make more user
	  friendly the specification
	  of the potential through the corresponding &POTENTIAL section).

2009-01-28 11:36  vweber

	* [r8165] rm few useless arrays.

2009-01-28 09:07  jgh

	* [r8162] Initial guess ATOMIC now uses the output of an atomic
	  calculation;
	  SE and DFTB still use the old type diagonal guess;
	  SE uses always MOPAC style and DFTB an equivalent approach;
	  445 regtests reset

2009-01-28 07:38  tlaino

	* [r8161] added unit for wannier_centers (and ions+centers as well)
	  (reported by Axel)

2009-01-27 14:49  vweber

	* [r8160] fixed nics for gapw. added 1 reg.

2009-01-27 14:03  vweber

	* [r8159] added interpolation of the soft part of the shifts. rm
	  unused variables. added 2 regs.

2009-01-27 13:24  tlaino

	* [r8158] bug fix: section LOCALIZE should never allow an explicit
	  calculation of the LOCALIZATION
	  properties. LOCALIZE section controls the flow of the calculation
	  during localization,
	  but whether performing localization or not is only controlled by
	  the single
	  quantities based on localization (bug reported by Axel).

2009-01-26 14:45  vweber

	* [r8156] rm calls to fm_get/set_submatrix. added unit for the
	  common_center key. fixed bug with pointers.

2009-01-26 10:52  tlaino

	* [r8155] for GHOST atoms the specification of the POTENTIAL is
	  redundant (and never
	  used). When an atom is GHOST, the POTENTIAL specification in the
	  input file
	  can now be skipped (suggested by Ondrej). Adapting 11 regtests.

2009-01-26 10:13  tlaino

	* [r8154] bug fix doc

2009-01-26 09:21  vondele

	* [r8153] no vdw interaction with ghost atoms.

2009-01-26 08:56  vondele

	* [r8152] replace system call with a rename, this works around an
	  infiband/fork problem observed on blanc / sara.

2009-01-23 10:04  jgh

	* [r8149] Complete rewrite of non-local pseudopotential part:
	  minimal overhead of redundant integrals in parallel,
	  new screening of overlap integrals and matrix products,
	  minimize flop count by early contraction of integrals,
	  Energies will change by sub-micro Hartrees for default settings,
	  229 regtests reset

2009-01-22 18:55  tlaino

	* [r8148] removing duplication. restructuring. prettifying. towards
	  a unified version
	  of metadynamics.

2009-01-22 14:42  tlaino

	* [r8147] Metadynamics: allowing langevin dynamics for COLVARS (see
	  biblio). (Fabio Sterpone)
	  Step 1. added 1 regtest.

2009-01-20 18:01  tlaino

	* [r8143] bug fix OMP (Mauro Boero)

2009-01-16 15:30  vweber

	* [r8142] speeding & cleaning up

2009-01-16 09:56  vweber

	* [r8141] clarified error message (Samuele)

2009-01-16 07:41  vondele

	* [r8140] bug fix: use proper max_set

2009-01-15 15:42  vondele

	* [r8139] bug fix for aux basis

2009-01-15 14:37  vondele

	* [r8138] fix buglet

2009-01-15 08:15  tlaino

	* [r8136] prettify

2009-01-14 21:12  vweber

	* [r8135] rm FULL keyword

2009-01-14 20:57  vweber

	* [r8134] rm FULL keyword

2009-01-14 13:16  jgh

	* [r8133] Towards a calculation of the density gradients using
	  collocation, not yet fully working

2009-01-13 21:51  tlaino

	* [r8132] minor fixes (Axel)

2009-01-13 21:37  tlaino

	* [r8131] fixing single precision issues

2009-01-13 16:06  vweber

	* [r8129] fixed description.

2009-01-13 15:31  vweber

	* [r8128] cleaning + rm useless gauge.

2009-01-13 11:23  tlaino

	* [r8127] COLVAR population analysis + regtest (Fabio Sterpone)

2009-01-10 08:31  vondele

	* [r8123] adjust time format for slow/fast scf iterations

2009-01-09 18:32  vondele

	* [r8122] bug fix the periodic (berry phase) dipole, also affects
	  vibrational intensities.

2009-01-09 13:28  vweber

	* [r8121] speedup

2009-01-08 18:47  tlaino

	* [r8118] prettify

2009-01-08 17:56  mguidon

	* [r8117] Correct array assignment

2009-01-08 15:55  vweber

	* [r8115] speedup

2009-01-08 14:09  mguidon

	* [r8112] Some more exchange holes used for Hartree-Fock long range
	  correction

2009-01-08 08:46  krack

	* [r8111] init variables (helps to avoid the printing of garbage)

2009-01-07 18:02  tlaino

	* [r8110] using dp instead of fixed double precision

2009-01-07 17:24  mguidon

	* [r8109] Added laplace dependent truncated becke roussel exchange
	  hole

2009-01-07 13:26  vweber

	* [r8108] rm useless initializations.

2009-01-07 12:33  mguidon

	* [r8107] Speedup pbc calls by a factor of 2-3

2009-01-06 10:05  tlaino

	* [r8106] prettify

2009-01-06 09:34  vweber

	* [r8105] init variables and added regs for charged systems.

2009-01-06 09:33  jgh

	* [r8104] Bug fix for last commit: remove test in order not to
	  trigger problems for
	  non-tau functionals

2009-01-05 19:00  tlaino

	* [r8103] prettify

2009-01-05 18:03  jgh

	* [r8102] Nullify pointers, No INTENT for pointer variables

2009-01-05 17:49  jgh

	* [r8101] Tau-functionals for GAPW: stability problem remains, one
	  new regtest

2009-01-05 12:46  jgh

	* [r8100] Atom code: numerical and semi-analytical exchange

2009-01-04 15:11  vondele

	* [r8098] more New Year stuff

2009-01-04 08:46  vondele

	* [r8097] init vars

2008-12-31 11:16  tlaino

	* [r8096] Update for 2009

2008-12-31 08:38  tlaino

	* [r8095] Added CIF (Christallographic Information File) as
	  possible format for inputting
	  coordinates. Added 1 regtest and a couple of CIF examples. Only
	  the basics of
	  the CIF have been implemented (it may require improvements in the
	  future).

2008-12-31 08:36  tlaino

	* [r8094] added a parser_reset to allow pattern searching from
	  begin of file

2008-12-23 10:32  vondele

	* [r8093] make ASPC extrapolation the default (~120 resets for
	  regtests)

2008-12-23 08:44  vondele

	* [r8092] add screening bufs to the accounting

2008-12-22 14:13  jgh

	* [r8091] Atom code: refactoring and adding more functionality

2008-12-21 10:57  tlaino

	* [r8090] clarifying error message

2008-12-21 10:18  tlaino

	* [r8089] PNNL semi-empirical parameterization (chris)

2008-12-21 08:37  tlaino

	* [r8088] allowing p-orbitals on H for SE calculations (requires a
	  parameterization that
	  support p-orbitals for H).

2008-12-19 14:35  tlaino

	* [r8086] prettify

2008-12-19 13:35  vweber

	* [r8085] reset the chi for gapw and some regs.

2008-12-19 13:27  tlaino

	* [r8084] prettify

2008-12-19 09:42  fschiff

	* [r8083] added the possibility to specify a frequency range, if
	  modes are tracked
	  by involved atoms

2008-12-18 12:17  tlaino

	* [r8081] bug fix

2008-12-18 10:35  vweber

	* [r8080] added igaim and 2 regs. reset some nmr/epr gapw regs
	  (numerics). huge speedup on gauge build
	  and more than 10 faster for the computation of gapw shift.

2008-12-17 16:38  tlaino

	* [r8078] patch for a PGI bug (just the N-one)

2008-12-15 23:54  tlaino

	* [r8051] clarifying error message

2008-12-12 20:42  tlaino

	* [r8048] new colvar: COMBINE_COLVAR allows for defining a COLVAR
	  as a combination of COLVARS.
	  The combination is fully arbitrary and uses the mathematical
	  parser. added 1 regtest.

2008-12-11 16:10  mguidon

	* [r8045] Completely rewritten periodic hfx code. Regtests and long
	  range correction functionals to be followed

2008-12-11 14:02  vondele

	* [r8039] make the objective function for 2d optimization user
	  selectable (Pekka Manninen)

2008-12-09 20:13  tlaino

	* [r8038] If parameters for GRIMME are not setup abort the run..

2008-12-09 20:13  tlaino

	* [r8037] no need to check further element.. at this level it is
	  already assigned..

2008-12-09 20:12  tlaino

	* [r8036] more robust and consistent assignment of element for
	  KINDS!

2008-12-09 20:12  tlaino

	* [r8035] added the possibility to avoid the failure and report
	  back the error

2008-12-09 16:18  tlaino

	* [r8034] VIRTUAL SITE constraint. This allows for TIP4P water.
	  Added regtest (patch
	  created by Marcel Baer and fixed/cleaned by Teo). Reset few
	  regtests for numerics.

2008-12-05 20:41  tlaino

	* [r8032] uniforming atom_names between different possible input
	  channels

2008-12-05 20:21  tlaino

	* [r8031] bug fix: element passed as argument should not be
	  modified.

2008-12-05 19:50  tlaino

	* [r8030] read amber labels in upper case

2008-12-05 19:50  tlaino

	* [r8029] check on atoms' name more robust

2008-12-05 19:13  tlaino

	* [r8028] refactoring

2008-12-05 19:04  tlaino

	* [r8027] allowing lower_to_upper in the general interface

2008-12-05 18:01  tlaino

	* [r8026] prettify

2008-12-05 17:36  vondele

	* [r8025] refactoring + small speedup (Iain Bethune)

2008-12-05 16:20  tlaino

	* [r8024] bug fix [CP2K:1612]

2008-12-05 15:10  tlaino

	* [r8022] factoring

2008-12-05 08:28  tlaino

	* [r8017] update doc

2008-12-03 09:43  tlaino

	* [r8016] using s2a

2008-12-03 09:00  tlaino

	* [r8015] NVE with a thermostat? questionable.. but you are using
	  in the test file
	  a rescaling (TEMP_TOL) with a CSVR. These are the same thing!
	  Maybe you
	  want to thermalize too much...

2008-12-03 08:33  marcella

	* [r8013] it is needed the thermalization of the shells even if the
	  run is not nvt or npt

2008-12-03 07:35  tlaino

	* [r8011] removing garbage

2008-12-03 07:35  tlaino

	* [r8010] removing wrong doxygen header

2008-12-03 01:36  tlaino

	* [r8009] another step towards periodic SE

2008-12-03 00:33  tlaino

	* [r8008] restore original versions

2008-12-02 23:32  tlaino

	* [r8007] additional control for longrange in SE

2008-12-02 22:17  marcella

	* [r8005] More stability for the shell model; now it is possible to
	  optimize the shell positions through
	  a CG minimization of the shell-core forces. The optimization can
	  be performed in initialization
	  and along an MD run, when the shell-core distance exceeds a given
	  threshold.
	  Additional tools to initialize and/or maintain
	  different regions of the system at different temperatures (useful
	  in simulations of cascades
	  and heat dissipation).
	  Implementation of the ZBL universal scattering potential, usually
	  added to the force field
	  when collisions are simulated, and anyway useful to avoid the
	  collapsing of opposite charges.
	  Some cleaning
	  Some regtest changes in Fist/regtest-7

2008-12-01 23:03  tlaino

	* [r8003] prettify

2008-12-01 22:21  cjmundy

	* [r8002] Restructured computing the constraint

2008-12-01 18:34  tlaino

	* [r8001] refactoring

2008-12-01 14:39  tlaino

	* [r8000] prettify

2008-12-01 14:37  tlaino

	* [r7999] restoring cell info in DCD lost with a subsequent commit,
	  changing unformatted write.

2008-11-27 17:17  tlaino

	* [r7998] pretttttttttification

2008-11-27 16:42  vondele

	* [r7997] robustify

2008-11-27 15:03  tlaino

	* [r7996] handling the DOF for the newly added colvar

2008-11-27 14:53  tlaino

	* [r7995] using less than 32 characters for function names

2008-11-27 14:44  tlaino

	* [r7994] simplify logics

2008-11-27 14:25  fschiff

	* [r7993] added reference to the input

2008-11-27 14:05  tlaino

	* [r7992] refactoring

2008-11-27 14:04  tlaino

	* [r7990] bug fix (wrong index in a do loop) + pretty

2008-11-27 13:20  fschiff

	* [r7989] Added REACTION_PATH collecitive variable, added regtest
	  in SE/regtest-3,

2008-11-27 07:06  tlaino

	* [r7985] pretty

2008-11-27 04:50  cjmundy

	* [r7984] Seems that dbm are already zeroed in add_block_node

2008-11-27 04:43  cjmundy

	* [r7983] Parallel bug fix dor SCP-NDDO. Thanks Teo. Sorry for the
	  troubles. Passed regtests. Added new regtests.

2008-11-26 16:18  tlaino

	* [r7980] propagating error and CPP statements

2008-11-26 14:40  tlaino

	* [r7979] adding an unimplemented for SCP-NDDO when using
	  d-orbitals

2008-11-26 14:39  tlaino

	* [r7978] more I/O for detailed_energy when SCP

2008-11-26 13:44  tlaino

	* [r7977] some more refactoring

2008-11-26 11:17  tlaino

	* [r7976] fix a bug introduce with last SCP_NDDO commit: never
	  allocate inside a neighbor list
	  loop (sensible slow down).. moreover in parallel it was just
	  creating a mess of memory leaks.
	  Using a scratch array instead. Fix test_type for the newly added
	  regtest.

2008-11-26 07:08  tlaino

	* [r7975] prettify

2008-11-25 23:35  cjmundy

	* [r7971] SCP-NDDO functionality. Passed regtests. reset SCP-DFT
	  regtests, added
	  1 regtest SE/regtest/dimer.inp

2008-11-25 18:10  tlaino

	* [r7970] completing cleaning of cell type (subcells has nothing to
	  do with cell information,
	  moved into DFT section).

2008-11-25 13:51  tlaino

	* [r7969] added units of measure for MOTION%PRINT { TRAJECTORY,
	  VELOCITIES, FORCES }.
	  Fully flexibl: all units (available in CP2K) can be combined with
	  whatever
	  output format. Cleaned unused code in write_particle_coordinates.

2008-11-25 10:57  tlaino

	* [r7967] clening cell type (removing everything connected with
	  units), removing UNIT keyword
	  from CELL and defining instead units for A,B,C and ABC keywords.
	  Added two more keywords
	  in COORD sections to have an additional and disentangled control
	  of the units for
	  coordinates (UNIT and SCALEDi, in &COORD). Updating the whole
	  bunch of regtests and
	  resetting 9 for numerics (due to the new conversion tools).
	  Several cosmetics.
	  Moreover, this commit introduces the possibility to have units
	  also for output quantities.
	  At the moment this keyword is supported only for a minor set of
	  print_keys (essentially
	  the ones that were depending on the unit_of_length of cell). More
	  work to do on this
	  topic in the future.

2008-11-24 16:45  tlaino

	* [r7966] removing useless code

2008-11-24 16:20  tlaino

	* [r7965] cleaning: please avoid to put comments and statements in
	  the definition of arguments.
	  The prettify will anyway get rid of them putting all the
	  statements at then end.
	  The doxygen header is there for this specific purpose: variable
	  definition,
	  history track, comments. Please, use doxygen instead of sparse
	  comments.

2008-11-24 13:46  tlaino

	* [r7964] cleaning unused

2008-11-24 13:15  tlaino

	* [r7963] more units + disentagling radius units from cell units

2008-11-24 12:29  krack

	* [r7962] Improve doc

2008-11-24 10:42  tlaino

	* [r7961] cosmetics

2008-11-21 23:30  tlaino

	* [r7953] [CP2K:1562] bug fix

2008-11-20 15:56  tlaino

	* [r7951] adding the possibility to specify a list of atoms on
	  which to apply the external
	  potential (Marcel,teo)

2008-11-20 15:25  tlaino

	* [r7948] prettify

2008-11-20 10:46  mcgrath

	* [r7945] Added a way to bias the swapping of molecules between
	  boxes.

2008-11-19 16:59  vondele

	* [r7942] remove unused ncoa_sum

2008-11-19 16:22  vondele

	* [r7941] provide a single routine to compute cube_center and use
	  in integrate, collocate and task_list_methods.
	  This avoids an out of bounds where different numerics and
	  rounding lead to an off by one error.

2008-11-19 15:51  vondele

	* [r7940] fix another KG bug

2008-11-19 12:52  tlaino

	* [r7939] adding units and clarifying doc

2008-11-18 20:53  vondele

	* [r7938] some more cleanups

2008-11-18 14:58  tlaino

	* [r7937] using CPP instead..

2008-11-18 14:41  tlaino

	* [r7936] check stat, when present in an allocate. Propagating
	  error and fix a bug with
	  an uninitialized error in _transfer.

2008-11-17 13:56  tlaino

	* [r7935] prettify

2008-11-15 16:28  jgh

	* [r7933] Input for dispersion type is now an integer list type

2008-11-15 14:24  jgh

	* [r7932] Prepare for DKH modules (Jens Thar)

2008-11-13 21:06  tlaino

	* [r7929] improved fixed atom constraints

2008-11-13 18:26  tlaino

	* [r7928] bug fix (Rachel)

2008-11-12 23:58  tlaino

	* [r7921] prettify

2008-11-12 13:57  fawzi

	* [r7918] corrected bug in density collocation, honors requested
	  radius even when small

2008-11-12 08:24  tlaino

	* [r7916] refactoring

2008-11-11 21:59  tlaino

	* [r7914] bug fix (Daniele)

2008-11-11 16:32  tlaino

	* [r7913] [CP2K:1546] intercepting a possible error in the input
	  setup

2008-11-11 14:34  fschiff

	* [r7912] parallelized bfgs optimizer (diagonalization, matrix
	  multiplication)

2008-11-11 14:00  tlaino

	* [r7911] changes applied without regtesting or possibly forgotten
	  to reset regtests.
	  the problem was reported 24 hours ago and no actions were taken.
	  going back to the previous version.

2008-11-11 13:18  tlaino

	* [r7910] bug fix cell info dcd

2008-11-10 23:36  tlaino

	* [r7907] prettify

2008-11-10 23:31  tlaino

	* [r7906] fix cell info for DCD

2008-11-10 13:17  fawzi

	* [r7905] fixing gauss colloc wings code for upper border>ngrid
	  points

2008-11-07 21:54  cjmundy

	* [r7904] A little debugging. Still not used, but getting there...

2008-11-07 18:09  tlaino

	* [r7903] correct the analytical form of the dihedral potential

2008-11-07 11:42  vondele

	* [r7902] correct timesets

2008-11-07 09:50  tlaino

	* [r7901] more timings

2008-11-07 09:30  tlaino

	* [r7900] clarify error message

2008-11-07 08:44  tlaino

	* [r7899] prettify

2008-11-06 23:00  tlaino

	* [r7897] 3rd and final step for natively supporting AMBER files:
	  parsing the force field
	  information. added 1 regtest.

2008-11-06 22:24  cjmundy

	* [r7895] Routines for the future SCP-DFT with appropriate
	  screening. These routines
	  (in scp_coeff_types.F) are not hooked up yet, but compile. The
	  new routines
	  allow one to replicated the SCP coefficient structure when
	  needed.

2008-11-06 13:56  vondele

	* [r7894] refactoring ppnl out of core hamiltonian

2008-11-06 11:21  vondele

	* [r7893] bug fix buffer size (Iain Bethune)

2008-11-06 08:28  tlaino

	* [r7892] workaround for PGI compiler bug (local variable CAN have
	  the same name
	  of the argument of a called function)

2008-11-06 07:52  vondele

	* [r7891] cleaning part 1 (remove OMP)

2008-11-06 07:50  vondele

	* [r7890] prettify

2008-11-06 07:47  tlaino

	* [r7889] fix doc [CP2K:1531]

2008-11-02 16:55  tlaino

	* [r7888] fix doc + cosmetics

2008-11-02 15:24  vondele

	* [r7887] add citation for Grimme vdW parameters

2008-10-31 23:09  ikuo

	* [r7886] Fix GEMC file without the right XC_FUNCTIONAL expanded
	  subsection.

2008-10-31 11:53  jgh

	* [r7885] Iterative diagonalization using OT, 3 new regtests

2008-10-30 18:55  vondele

	* [r7882] enhance cell output

2008-10-30 14:42  marcella

	* [r7880] update dt_fact only when needed

2008-10-30 13:49  tlaino

	* [r7879] removing dt_fact from thermostat type (ok in the csvr and
	  nhc type). dt_fact
	  can simply be a real and no need to have a pointer. access the
	  proper dt_fact
	  through the use of a set_thermostats routine. again 1 reset for
	  instabilities

2008-10-30 10:26  marcella

	* [r7874] modified algorithm for the variable time step, this
	  version should be more stable.
	  bug fix and cleaning in integrator_utils.F.
	  added optional check on the maximum core-shell separation in
	  shell-model (regtest to be added).
	  
	  ps: I obtain wrong results with Linux-x86-64-g95.sopt for some
	  regtest of
	  the type Fist/regtest-10/nh3-meta* and
	  Fist/regtest-4/nh3-meta-walks*; I believe this
	  is due to numerical instabilities, since this commit should not
	  affect those tests.
	  I could not find any other explanation.

2008-10-30 02:07  tlaino

	* [r7871] remove last duplication on units: all units
	  transformation are now handled by
	  a unique module cp_units.

2008-10-29 23:51  tlaino

	* [r7870] remove duplication: keeping only 1 unit converter. reset
	  16 regtests for numerics.

2008-10-29 17:42  tlaino

	* [r7869] doxygen info

2008-10-29 17:39  tlaino

	* [r7868] 2nd step for natively supporting AMBER files: reading the
	  connectivity from the
	  topology file. These routines are linear scaling, MUCH faster
	  than leap2fist.
	  Once the forcefield part will be parsed the leap2fist will be
	  totally obsolete.
	  Added 1 regtest using the connectivity of AMBER topology file.

2008-10-29 17:36  tlaino

	* [r7867] functions for sorting arrays of strings as well as
	  matrices (sorting per row)

2008-10-29 17:35  tlaino

	* [r7866] cleaning

2008-10-29 15:25  vondele

	* [r7865] disable flushing files by default. FLUSH_SHOULD_FLUSH in
	  &GLOBAL can be used to switch to the old behavior.

2008-10-29 09:46  vondele

	* [r7864] more timesets

2008-10-29 08:10  vondele

	* [r7862] rename KEEP_SHAPE in KEEP_ANGLES, since the shape of the
	  unit cell is not preserved.

2008-10-28 22:41  tlaino

	* [r7861] 1st step for natively supporting AMBER files: reading
	  coordinates in amber format.
	  added 1 regtest.

2008-10-28 16:58  tlaino

	* [r7859] using new function to detect if an atom is hydrogen

2008-10-28 16:35  tlaino

	* [r7858] bug fix (axel)

2008-10-28 15:51  tlaino

	* [r7857] doxygen

2008-10-28 15:18  tlaino

	* [r7856] propagating error

2008-10-28 14:28  tlaino

	* [r7855] deleting character arrays in input_constants.F : this
	  module should contain
	  only integer/real definitions of constants. Enumeration types
	  should be
	  handled differently, not with character arrays.

2008-10-28 11:56  tlaino

	* [r7854] refactoring for format keyword (collecting common calls
	  and getting rid of useless
	  arrays). Added XYZ as an alias for XMOL (Axel)

2008-10-28 11:52  tlaino

	* [r7853] fix doc

2008-10-28 11:01  tlaino

	* [r7852] modified short-range dumping term for SCC-DFTB (hydrogen
	  bonds) + regtest (axel,teo)

2008-10-28 10:01  vondele

	* [r7851] fix a potentially serious bug in the generation of the
	  ppnl lists. If the atomic kinds have very different radii, terms
	  could be missing in the neighbor lists, leading to jumps in the
	  constant of motion.

2008-10-28 09:54  vondele

	* [r7850] do not specify intent on pointer (F2003, rejected by
	  pathscale)

2008-10-28 09:39  vondele

	* [r7849] prettified

2008-10-28 07:35  tlaino

	* [r7847] cleaning

2008-10-28 07:32  krack

	* [r7846] * no automatic change of the MIXING parameter for the
	  Hartree damping with ASPC
	  * correct doc for thermostat

2008-10-28 05:25  ikuo

	* [r7845] Fix cases where VIRIAL_TEMP is not specified explicitly.

2008-10-27 17:31  tlaino

	* [r7844] zeroing out of the loop

2008-10-27 13:33  tlaino

	* [r7843] refactoring and splitting of tapering function

2008-10-26 12:48  tlaino

	* [r7840] ASPC output under control of a print_key

2008-10-26 11:46  tlaino

	* [r7839] set a better default for write frequency of MD restart
	  (now by default every 20 steps)

2008-10-26 09:09  tlaino

	* [r7836] tuning print level

2008-10-25 12:15  vondele

	* [r7835] don't forget to check stat if allocate/deallocate has
	  STAT=stat

2008-10-25 09:31  vondele

	* [r7834] don't forget to check stat if allocate/deallocate has
	  STAT=stat

2008-10-25 09:27  vondele

	* [r7833] add forgotten STAT=stat

2008-10-25 07:29  vondele

	* [r7832] added ref, fixed reference manager for longer sources,
	  extended s2a

2008-10-24 18:10  tlaino

	* [r7831] bug fix for periodic colvar for constraints as well
	  (refactoring restraints for
	  periodic)

2008-10-24 17:21  tlaino

	* [r7830] fix doc

2008-10-24 17:08  vondele

	* [r7829] bug fix

2008-10-24 17:08  tlaino

	* [r7828] bug fix when applying restraints to a torsion colvar
	  (issue with periodicity). 2 reset.

2008-10-24 17:06  tlaino

	* [r7827] zero only the local components

2008-10-24 15:13  krack

	* [r7826] typos in doc corrected

2008-10-24 12:21  tlaino

	* [r7824] cosmetics: some more changes to conform to the doxygen
	  style

2008-10-24 11:48  vondele

	* [r7822] additional machine file (Pekka Manninen, PRACE)

2008-10-24 11:09  tlaino

	* [r7820] printing available units in the online reference manual

2008-10-24 11:08  tlaino

	* [r7819] cosmetics

2008-10-24 08:38  tlaino

	* [r7818] cleaning

2008-10-23 17:17  tlaino

	* [r7809] print a warning on screen if we do ASPC and specify
	  MIXING (it will be anyway
	  ignored). Cleaning.

2008-10-23 17:06  tlaino

	* [r7808] cleaning: weird thing.. two sets of integers.. absolutely
	  duplicated.. one
	  used to specify the input the other one to control the flow of
	  the code.. why that???
	  we need a unique set of integers for specifying the input and
	  codeflow control..

2008-10-23 15:56  jgh

	* [r7807] New SCF section DIAGONALIZATION, move Jacobi keywords to
	  this section

2008-10-23 14:33  tlaino

	* [r7806] for cell optimizations MD based we want to exploit the
	  information collected
	  during the previous run (i.e. avoid to start always a clean MD
	  run but use
	  the information coming from the old ones (thermostats and
	  velocities)).

2008-10-23 13:45  tlaino

	* [r7805] restructuring the LOCALIZE keyword. LOCALIZATION%LOCALIZE
	  moved into DFT%LOCALIZE
	  the content of DFT%PRINT%LOCALIZATION%LOCALIZE%PRINT has been
	  moved into DFT%PRINT%LOCALIZATION
	  the LOCALIZE%PRINT of the XAS and LINRES has been moved into the
	  corresponding XAS%PRINT and
	  LINRES%PRINT. Regtests adapted. more details on [CP2K:1470] and
	  following..

2008-10-23 09:24  tlaino

	* [r7804] metadynamics: refactoring + splitting into an utilities
	  file.

2008-10-23 08:28  tlaino

	* [r7803] more units: picometer, nanometer, joulemol

2008-10-23 07:28  vondele

	* [r7802] some more timesets

2008-10-22 17:57  tlaino

	* [r7800] computing the dispersion before of the SCF run so at the
	  end of the SCF run
	  we have energies to print on screen.. fix TEST_TYPE which was set
	  to a wrong value.

2008-10-22 16:18  tlaino

	* [r7799] fix wrong write format; closing output_unit; prettify

2008-10-22 16:17  tlaino

	* [r7798] prettify

2008-10-22 15:46  tlaino

	* [r7797] prettify

2008-10-22 15:40  jgh

	* [r7795] DFT-D method implemented; This is a pair potential for
	  dispersion,
	  the functional form by S. Grimme JCC 27 1787 (2006) is used; one
	  new regtest

2008-10-22 13:23  vondele

	* [r7794] remove strange use

2008-10-22 12:50  vondele

	* [r7793] bug fix: temporarily switch back to the previous
	  generalised collocate which circumvents a bug with
	  non-orthorhombic cells and load_balance_distributed

2008-10-22 06:11  tlaino

	* [r7791] prettify

2008-10-22 06:02  marcella

	* [r7790] original version

2008-10-21 22:30  marcella

	* [r7789] time consistent with the trajectory in reftraj

2008-10-21 21:03  tlaino

	* [r7788] bug fix

2008-10-21 18:27  vondele

	* [r7787] bug fix for a corner case with distributed grids, and a
	  little cleanup

2008-10-20 22:52  tlaino

	* [r7785] bug fix: undefined

2008-10-20 18:03  tlaino

	* [r7784] multiple walkers in metadynamics + 4 new regtests (teo,
	  axel, giacomo)

2008-10-20 12:59  vondele

	* [r7780] added TPSS reference

2008-10-20 11:01  vondele

	* [r7779] make the locating the chemical potential more robust

2008-10-20 08:49  vondele

	* [r7778] bug fix added MOs and LSD

2008-10-19 23:45  tlaino

	* [r7777] cleaning localization section (not a print_key anymore..
	  just a standard section).. moving
	  MOLECULAR_STATES inside the localization section (since
	  MOLECULAR_STATES was depending on
	  LOCALIZATION). general cleaning and cosmetics..

2008-10-19 19:14  tlaino

	* [r7775] avoid writing wannier centers at the 0 step of an MD
	  restart.. [CP2K:1462] (reported by Axel)

2008-10-19 18:56  tlaino

	* [r7774] cleaning

2008-10-19 15:19  tlaino

	* [r7773] generalizing the calculation of total_dipole and
	  molecular_dipoles (this last one
	  still to be fully implemented) using the wannier centers.
	  TOTAL_DIPOLE and MOLECULAR_DIPOLES
	  are now two print_keys (more flexibility) and for both is
	  possible to specify PERIODIC,
	  REFERENCE POINT. For MOLECULAR_DIPOLES these part of the input
	  have been specified but
	  not implemented (a warning is printed on screen when used..)
	  Getting rid of useless keys and prettify.

2008-10-19 15:18  tlaino

	* [r7772] bug fix

2008-10-19 13:09  tlaino

	* [r7771] prettify

2008-10-19 12:56  tlaino

	* [r7770] prettify

2008-10-19 12:29  tlaino

	* [r7768] cosmetics

2008-10-19 12:01  tlaino

	* [r7767] poisoning the blas calls of few more new and old files
	  (where the USE was deleted). The
	  only non-poisoned files are the ones which require a proper and
	  carefully re-organization of
	  the blas calls. More important: the module f77_blas is now ONLY
	  poisoning the blas calls
	  but does not cure them. A proper USE to f77_blas_generic and
	  f77_blas_extra has to be
	  explicitly imported where needed. This makes the compilation
	  process faster (avoiding to
	  import a module where it is not needed).

2008-10-18 21:23  tlaino

	* [r7766] bug fix

2008-10-18 20:21  tlaino

	* [r7765] bug fix

2008-10-18 18:08  tlaino

	* [r7764] Setting to .FALSE. the default of averages of colvars

2008-10-18 16:12  tlaino

	* [r7763] RESTART for AVERAGES

2008-10-18 16:04  tlaino

	* [r7762] prettify

2008-10-18 15:30  jgh

	* [r7761] DFTB dispersion neighbor list updated

2008-10-18 14:02  tlaino

	* [r7760] prettify

2008-10-18 13:52  jgh

	* [r7759] Simplify interface to fist_nonbond

2008-10-18 11:02  tlaino

	* [r7758] bug fix

2008-10-17 21:24  tlaino

	* [r7757] bug fix (wrong collection of timing data in parallel) +
	  cosmetics.

2008-10-17 20:15  tlaino

	* [r7756] cleaning flops: not needed anymore..

2008-10-17 19:23  vondele

	* [r7755] further timeset/timestop cleaning, now only a single form
	  is available

2008-10-17 18:00  tlaino

	* [r7754] changing BECKE_CONSTRAINT into BECKE_RESTRAINT. This
	  makes the all the sections (DDAPC,
	  MULLIKEN and BECKE) imposing restraints/constraints on WFN look
	  with similar structure.

2008-10-17 13:48  tlaino

	* [r7748] prettify

2008-10-17 13:44  jgh

	* [r7747] DKH input and setup only

2008-10-17 12:49  tlaino

	* [r7743] controlling the AVERAGES

2008-10-17 11:15  vondele

	* [r7739] new timing report (total time, self time, and variation
	  between MPI tasks).
	  requires updated do_regtest

2008-10-17 09:46  tlaino

	* [r7737] using the MD section all over the MD module, no need to
	  have general section in
	  inner part of MD engine. Added an AVERAGES section.

2008-10-17 09:45  tlaino

	* [r7736] workaround for NAG compiler bug

2008-10-17 06:31  vondele

	* [r7735] bug fix parallel mode selective vibrational analysis with
	  bfgs or molden initial guess

2008-10-16 21:56  tlaino

	* [r7734] fix localLogs for L-BFGS and redirecting restart
	  informations to file (if required)
	  (problem reported by Axel [CP2K:1445]).

2008-10-16 15:59  tlaino

	* [r7732] removing dinosaurs

2008-10-16 15:27  tlaino

	* [r7731] refactoring: creation of a simulation_parameters_type and
	  methods.
	  creation of an md_vel_utils. cosmetics.

2008-10-16 11:18  vondele

	* [r7729] compress the output of the distribution of the neighbor
	  lists and overlap matrix

2008-10-16 08:05  tlaino

	* [r7728] cleaning and getting rid of the md_defaults (defaults
	  should always go
	  as long as possible in the input specification)

2008-10-16 06:14  vondele

	* [r7727] prettified

2008-10-16 05:07  vondele

	* [r7726] bug fix yesterdays commit for distributed non-ortho grids

2008-10-15 22:33  tlaino

	* [r7725] bug fix

2008-10-15 21:35  tlaino

	* [r7724] prettify

2008-10-15 20:30  vondele

	* [r7723] greatly (measured up to 40% speedup) improve the parallel
	  efficiency of the FFT (Iain Bethune)

2008-10-15 18:03  tlaino

	* [r7722] .. and of course let's print all (includign defaults) if
	  the user is asking
	  for them..

2008-10-15 17:42  tlaino

	* [r7721] (1) section_vals_write dumps only explicit
	  keywords/sections. This makes finally
	  consistent the usage of a restart file as input file (which was
	  an dramatic issue
	  due to the way we use keywords to enable methodologies (which may
	  possibly be
	  yet defaults)).
	  (2) Moreover this reduces dramatically the amount of I/O due to
	  the writing of
	  the restart file.
	  (3) added the "explicit" field to the section_vals_val_get call.
	  
	  bug fix for KG (wrong interpretation of keywords, solved with the
	  new clean restart).

2008-10-15 16:05  tlaino

	* [r7720] bug fix for iteration number in L-BFGS optimizer (bug
	  reported by Pietro)

2008-10-15 15:34  tlaino

	* [r7719] prettify

2008-10-15 15:32  vondele

	* [r7717] added a cost model for generalised tasks and refined the
	  other (not much effect)

2008-10-15 12:59  tlaino

	* [r7716] clean-up timestop usage: clean also the template file

2008-10-15 11:11  vondele

	* [r7714] clean-up timestop usage

2008-10-15 09:48  tlaino

	* [r7713] prettify

2008-10-15 09:48  tlaino

	* [r7712] making uniform the call of timeset

2008-10-15 09:42  tlaino

	* [r7711] clean-up timeset usage: modifying templates files as
	  well...

2008-10-15 09:36  tlaino

	* [r7710] more efficient averaging of COLVAR infos

2008-10-15 09:27  vondele

	* [r7709] clean-up timeset usage

2008-10-15 08:36  tlaino

	* [r7708] remove write statement

2008-10-15 06:18  tlaino

	* [r7707] fix IO

2008-10-14 17:14  tlaino

	* [r7706] clarify doc

2008-10-14 17:14  tlaino

	* [r7705] cosmetics

2008-10-14 14:26  tlaino

	* [r7704] refactoring and collecting common keywords to langevin in
	  a langevin section..

2008-10-14 13:31  tlaino

	* [r7703] prettify

2008-10-14 13:30  tlaino

	* [r7702] bug fix

2008-10-14 12:52  tlaino

	* [r7701] making more human comprehensible some error messages

2008-10-13 17:05  tlaino

	* [r7700] cosmetics

2008-10-13 11:13  tlaino

	* [r7699] more biblio references

2008-10-13 10:17  tlaino

	* [r7698] units for WW

2008-10-13 08:57  vondele

	* [r7697] assign generalised tasks to CPUs optimally

2008-10-12 22:05  tlaino

	* [r7696] averages of COLVARs. reset 5 regtests due to numerics
	  (<10^-12)

2008-10-12 10:56  tlaino

	* [r7695] collecting all md related IO on screen in a common
	  routine..

2008-10-11 22:51  tlaino

	* [r7694] left justify field 13:16 [CP2K:1438]

2008-10-11 16:27  tlaino

	* [r7693] bug fix: undefined pointer..

2008-10-11 16:22  tlaino

	* [r7692] Band methods in a subspace of collective variables
	  [towards minimum free energy paths (not yet working)]

2008-10-11 15:30  tlaino

	* [r7691] refactoring

2008-10-10 15:57  jgh

	* [r7690] Correct typo

2008-10-10 12:52  vondele

	* [r7687] prettified versions

2008-10-10 12:50  vondele

	* [r7686] Non-blocking communication in the halo exchange (Iain)

2008-10-10 08:00  vondele

	* [r7685] additional functionality (Iain)

2008-10-09 16:25  tlaino

	* [r7684] prettify

2008-10-09 14:28  vondele

	* [r7683] remove unneeded mp_syncs

2008-10-09 14:07  vondele

	* [r7682] improved load balancing using tasks on distributed grids

2008-10-08 20:37  tlaino

	* [r7681] tool for statistical analysis (to be fully validated yet)

2008-10-08 17:07  marcella

	* [r7680] bug fix

2008-10-08 16:59  tlaino

	* [r7679] bug fix

2008-10-08 15:30  marcella

	* [r7675] variable timestep: displacment check also on the
	  shell-core motion, scaling factor computed in a separated routine

2008-10-08 06:55  vondele

	* [r7670] add scatter

2008-10-07 14:38  tlaino

	* [r7669] Improving docs for EAM file format (full description of
	  the file format),
	  allowing a different number of points than the hardcoded 5000 for
	  EAM potentials:
	  this means that the old CU.pot may break old cp2k version. Please
	  use an updated
	  one! Most important: using the parser to parse the EAM pot file

2008-10-07 14:24  marcella

	* [r7667] bug fix

2008-10-06 22:53  tlaino

	* [r7664] the MD module must be kept slim. bug fix for updating
	  time for use in metadyn.

2008-10-06 15:44  tlaino

	* [r7662] remove useless variables

2008-10-06 14:07  tlaino

	* [r7661] prettify

2008-10-06 14:01  tlaino

	* [r7660] retrieving shorter name

2008-10-06 13:56  marcella

	* [r7658] fix memoryleak

2008-10-06 13:43  tlaino

	* [r7657] remove duplications!

2008-10-06 13:24  marcella

	* [r7656] Collective variable coordination number: added the
	  possibility to define this CV as product of two
	  coordination numbers, i.e. Cn_i = \sum_j f1(r_{ij})*[\sum_k
	  f2(r_{jk})]

2008-10-06 10:16  tlaino

	* [r7655] prettify

2008-10-06 00:43  tlaino

	* [r7654] DIPOLE for mixed_env (allowing intensities for mixed_env
	  in the vibrational module) + regtests.
	  some cosmetics.

2008-10-05 17:29  tlaino

	* [r7653] major rewriting of cp_results: storing properties in a
	  specific type (getting
	  rid of storing numbers on strings) + cosmetics.

2008-10-05 08:52  tlaino

	* [r7652] fix uninitialized value

2008-10-03 17:35  tlaino

	* [r7651] Enabling VIBRATIONAL_ANALYSIS and BAND methods with MIXED
	  force_env + regtests.

2008-10-03 16:32  tlaino

	* [r7650] properly initialize in parallel

2008-10-03 14:08  vondele

	* [r7649] encode alternate task locations

2008-10-03 12:41  vondele

	* [r7648] generalise collocation/integration

2008-10-03 10:20  tlaino

	* [r7647] tune print_key EWALD INFO (low->medium)

2008-10-03 07:46  vondele

	* [r7646] prettified

2008-10-03 07:42  vondele

	* [r7645] distribute_tasks: enhanced comments, declared a few more
	  variables int_8

2008-10-02 20:21  tlaino

	* [r7644] fine tuning of the rotational analysis output.

2008-10-02 16:52  tlaino

	* [r7643] bug fix (problem reported by Matt)

2008-10-02 16:52  tlaino

	* [r7642] bug fix

2008-10-02 15:34  vondele

	* [r7641] further refactoring, with some speedup

2008-10-02 12:19  tlaino

	* [r7633] "almost bug" fixes [CP2K:1389]

2008-10-02 10:16  krack

	* [r7632] Update links

2008-10-02 07:36  vondele

	* [r7630] further refactoring

2008-10-02 07:31  tlaino

	* [r7629] making taper consistent with the construction of the NL
	  (fix input doc). Reset 5.

2008-10-01 17:09  tlaino

	* [r7627] typo

2008-10-01 17:09  tlaino

	* [r7626] bug fix

2008-10-01 16:07  tlaino

	* [r7625] cosmetics

2008-10-01 14:20  vondele

	* [r7624] further refactoring

2008-10-01 13:38  tlaino

	* [r7623] biblio update

2008-10-01 12:57  vondele

	* [r7622] some refactoring

2008-10-01 11:27  tlaino

	* [r7621] evaluation of shortrange integrals+derivatives 10% faster

2008-10-01 10:30  tlaino

	* [r7620] tiny speedup (2%)

2008-09-30 15:11  tlaino

	* [r7619] more 10 % faster

2008-09-30 15:10  tlaino

	* [r7618] fix typo (Axel)

2008-09-30 11:45  tlaino

	* [r7617] another small speedup

2008-09-30 09:14  tlaino

	* [r7616] more 25% faster..

2008-09-30 08:40  tlaino

	* [r7615] prettify

2008-09-30 08:39  tlaino

	* [r7614] bug fix

2008-09-30 08:16  fschiff

	* [r7613] bug fix for pw_grid ray distribution

2008-09-29 19:12  tlaino

	* [r7612] SE 300% faster

2008-09-29 16:18  tlaino

	* [r7611] global twopi and prettify

2008-09-29 16:05  tlaino

	* [r7610] bug fix

2008-09-29 15:47  tlaino

	* [r7609] Recovering single-precision (Axel)

2008-09-29 14:39  tlaino

	* [r7608] prettify

2008-09-29 11:34  jgh

	* [r7606] Atom code: density fit, Powell optimizer, over-complete
	  basis

2008-09-28 12:47  vondele

	* [r7605] prettified

2008-09-28 10:57  vondele

	* [r7604] added a generalized low spin ROKS scheme

2008-09-27 13:28  tlaino

	* [r7602] bug fix doc

2008-09-24 13:31  krack

	* [r7600] * variable precision output for eigenvectors too
	  * allow printing of a subset of MO eigenvalues or eigenvectors
	  * 2 regtests updated

2008-09-24 09:35  tlaino

	* [r7599] doc typo

2008-09-24 00:33  tlaino

	* [r7598] periodic SE + regtests

2008-09-23 23:05  tlaino

	* [r7597] bug fixes for parallel

2008-09-23 13:28  tlaino

	* [r7596] bug fix + 1 regtest

2008-09-22 19:23  tlaino

	* [r7595] bug fix + 1 regtest

2008-09-22 19:22  tlaino

	* [r7594] bug fix: out of bound..

2008-09-22 12:29  krack

	* [r7592] Generate non-fixed MO occupation numbers only if MO
	  eigenvalues are available

2008-09-22 10:12  krack

	* [r7591] more compact printout of MO eigenvalues and occupation
	  numbers with variable
	  precision, if no MO eigenvector printing is requested

2008-09-20 14:31  jgh

	* [r7589] Atom code: tau functionals, timings, more output

2008-09-19 11:15  vondele

	* [r7584] allow cp_fm_syevx to compute only eigenvalues, not
	  eigenvectors.
	  This turns out not to be faster than syevd computing eigenvalues
	  and eigenvectors.

2008-09-18 12:57  krack

	* [r7580] Adapt type names in XML dump

2008-09-18 12:38  vondele

	* [r7579] nondiagonal energy term added for rotation.

2008-09-17 15:18  tlaino

	* [r7578] String Method (Rodolphe and Teo).

2008-09-14 18:28  tlaino

	* [r7575] using physcon! reset 38.

2008-09-14 15:52  tlaino

	* [r7574] bug fix core-core for periodic SE (reset few regtests due
	  to numerics)

2008-09-14 13:09  vondele

	* [r7573] use cpu_time directly

2008-09-13 15:35  tlaino

	* [r7572] more for SE periodic

2008-09-13 06:59  vondele

	* [r7571] and prettify

2008-09-13 06:53  vondele

	* [r7570] check the status variable if is present (not just for the
	  last allocate)

2008-09-08 15:41  vondele

	* [r7567] improved pw2rs halo swap (Iain Bethune)

2008-09-08 08:01  vondele

	* [r7566] hack around a problem with added MOs, atomic guess, and
	  OT.

2008-09-08 07:01  vondele

	* [r7565] call set_mo_occupation more consistently after the
	  generation of new MOs. Fixes a problem with restarting from a
	  converged run, but with different temperature

2008-09-07 14:30  vondele

	* [r7564] further preparations for smear and OT.

2008-09-06 12:37  vondele

	* [r7563] keep a copy of the expectation values as needed

2008-09-06 12:03  vondele

	* [r7562] pretty version

2008-09-06 11:59  vondele

	* [r7561] moving the matrix_dedc_scaled and occupation numbers
	  around.

2008-09-06 11:17  vondele

	* [r7560] slightly more efficient (1 log instead of 2) for the
	  entropy, should still be free of arithmetic exceptions.

2008-09-06 08:31  vondele

	* [r7559] cleanup: consistently use routineN in timeset

2008-09-05 23:20  tlaino

	* [r7557] prettify

2008-09-05 23:15  tlaino

	* [r7556] removing debug statement

2008-09-05 21:10  marcella

	* [r7555] bug fix: gapw with not paw atoms. The missing energy
	  terms have been included
	  Not paw atoms in standard gapw calculations are now
	  treated by the gpw scheme, i.e. collocating the entire density
	  and core density onto the gobal grid,
	  while the local nad compensation densities are not computed for
	  these atoms.
	  When full-gapw calculations are performed, instead, the gapw
	  scheme is used, i.e. the
	  compensation densities are computed and the missing terms are
	  included via the 2-center
	  and 3-center gapw terms.
	  Finally, even if kinds with soft basis set are present, it is
	  possible to force
	  the use of a gapw scheme for all the kinds by the keyword
	  FORCE_PAW. In this case
	  the local densities are computed, even if soft and hard local
	  densities should be equal.
	  (useful for testing)
	  Some regtest need resetting, some new regtests have been added

2008-09-05 20:36  vondele

	* [r7553] redefine mo_derivs [mo_derivs_i=(dEdC_i)/(2 f_i)], and
	  use everywhere correctly.
	  Bring the rescaling in qs_ot_scf for later use.
	  One incorrect regtest and its dependencies reset (added_mos,
	  without rotation, and with full_all)

2008-09-05 15:45  vondele

	* [r7552] optimized rs2pw halo swap and rs grid distribution
	  heuristics (Iain Bethune)

2008-09-05 13:02  vondele

	* [r7551] towards a combination of OT and smear (not yet
	  functional).

2008-09-05 12:12  krack

	* [r7550] Avoid arithmetic exceptions

2008-09-04 15:06  krack

	* [r7545] XML to HTML output and processing updated:
	  * Printout of references added
	  * HTML output is now written to multiple HTML files (one per
	  section) in a
	  tree structure which perfectly matches the input tree structure
	  * much faster loading of the(nojavascript) HTML pages now (as
	  fast as the direct HTML output)
	  * "linked" input path added
	  * XML 1.0 -> 2.0
	  * XSL processor changed from Xalan to saxon which can deal with
	  XML 2.0

2008-09-04 15:00  krack

	* [r7544] XML to HTML output and processing updated:
	  * Printout of references added
	  * HTML output is now written to multiple HTML files (one per
	  section) in a
	  tree structure which perfectly matches the input tree structure
	  * much faster loading of the(nojavascript) HTML pages now (as
	  fast as the direct HTML output)
	  * "linked" input path added
	  * XML 1.0 -> 2.0
	  * XSL processor changed from Xalan to saxon which can deal with
	  XML 2.0

2008-09-04 14:21  tlaino

	* [r7543] prettify

2008-09-04 12:33  vondele

	* [r7542] add electronic entropy to the energy in the case of
	  fermi-dirac smearing

2008-09-04 09:24  vondele

	* [r7541] more robust check for the occupation numbers

2008-09-04 09:10  vondele

	* [r7540] correct lfomo calculation

2008-09-04 09:05  vondele

	* [r7539] add fermi_utils.
	  use it in set_mo_occupation.

2008-09-03 18:52  vondele

	* [r7537] added a list of occupation numbers as a method to smear.
	  With fixed fractional occupations OT now works.

2008-09-03 14:52  tlaino

	* [r7536] improved efficiency of parser (less buffering operations
	  as really needed)

2008-09-03 11:11  tlaino

	* [r7535] printing a summary of the restarted information (when
	  restarting..) (suggested by axel & giacomo)

2008-09-03 08:02  tlaino

	* [r7534] repeatable/required information for sections in html and
	  section describe [CP2K:1326]

2008-09-02 15:52  jgh

	* [r7532] Atom code (alpha version): more basis sets, functionals,
	  potentials

2008-09-02 13:12  krack

	* [r7530] * Allow for different types of smearing (first stage)
	  * New input section &SMEAR introduced
	  * Fermi-Dirac smearing added (not yet fully tested, regtest will
	  follow)
	  * Regtests updated, reset of one regtest

2008-09-02 11:29  tlaino

	* [r7529] bug fix (juerg)

2008-09-02 10:20  tlaino

	* [r7528] prettify

2008-09-02 10:07  krack

	* [r7527] Improved documentation of the pseudo-diagonalization
	  method

2008-09-01 16:44  tlaino

	* [r7525] NEB references

2008-09-01 16:43  tlaino

	* [r7524] book references may not have the volume field (check if
	  present VL)

2008-09-01 15:11  tlaino

	* [r7523] units for R0 (rachel)

2008-08-31 17:47  tlaino

	* [r7522] speed up for strings broadcast (on a test case of 35 MB
	  file the speedup is
	  roughly 900%).

2008-08-31 14:51  tlaino

	* [r7521] bug fix (getting rid of the backspace call).

2008-08-29 17:27  tlaino

	* [r7520] bug fix for keyword DESCRIBE. added 1 ["extreme" use of
	  parser] regtest.

2008-08-29 14:29  tlaino

	* [r7519] Prettify

2008-08-29 13:57  jgh

	* [r7518] Atom code: more options for basis set and xc functionals

2008-08-29 12:28  tlaino

	* [r7517] Workaround for the circular dependency timings <->
	  message_passing. Prettify.

2008-08-29 10:07  vondele

	* [r7516] tiny speedup

2008-08-28 17:04  tlaino

	* [r7515] Buffer for the parser. Restored the parser calls in the
	  reftraj module. Few
	  bug fixes in the parser itself. Cosmetics.

2008-08-26 21:20  tlaino

	* [r7513] Allowing parser to parse range of integer with the
	  notation: INT1..INT2.
	  Some cleaning and restructuring of the parser itself. Removed the
	  keyword
	  RANGE (do not need one now that we can parse range of integer at
	  the parser
	  level) from few input sections and reformatted consequently the
	  regtests.
	  cosmetics.

2008-08-26 07:02  tlaino

	* [r7512] framework periodic SE. 1.

2008-08-25 13:20  jgh

	* [r7511] A first (small) step towards an atomic code in CP2K

2008-08-22 11:06  vondele

	* [r7510] QS now automatically selects a blocked (real and gspace)
	  pw grid layout for the FFTs if this is expected to be more
	  efficient.
	  up to 100% speedup of the FFTs for large number of CPUs.

2008-08-21 14:01  vondele

	* [r7508] update comment

2008-08-21 13:44  vondele

	* [r7507] make pw_grid_blocked also a ternary at the input level

2008-08-21 12:35  vondele

	* [r7506] pw_env_methods: check what methods can not deal with
	  ray-distributed pw_grids

2008-08-21 07:49  vondele

	* [r7505] make the blocked argument of pw_grid_setup take three
	  values (true, false, free)

2008-08-20 19:10  vondele

	* [r7504] bug fix fft routines for ray/blocked (Juerg)

2008-08-20 13:27  vondele

	* [r7503] allow for doing debug mode FFTs in libtest

2008-08-20 12:38  vondele

	* [r7502] add a limit to the size of the fft scratch space

2008-08-20 07:17  vondele

	* [r7501] fix #elif

2008-08-19 19:12  vondele

	* [r7500] added a pw_grid_layout_all option
	  print the flag spherical explicitly

2008-08-19 15:12  vondele

	* [r7498] generalized libtest of the pw_transfer routines (several
	  options, but not all, being tested in LIBTEST/test_pw.inp)

2008-08-18 16:15  tlaino

	* [r7496] Making ddapc fully independent: it won't appear in the
	  timings unless
	  DDAPC charges are requested (please, next time be careful when
	  doing
	  this kind of splitting reorganization!).

2008-08-18 13:53  vondele

	* [r7495] add a comment (why ddapc always appears in timings now)

2008-08-18 11:53  vondele

	* [r7494] New collocate/integrate core routines for a number of
	  compilers, controlled at compile time with -D__GRID_CORE=X
	  (X=1..6).
	  Using specialized routines up to l=9.
	  Resets (a few hundred regtests).

2008-08-15 08:38  vondele

	* [r7493] bug fix fft routines for ray&blocked FFTs (Juerg)
	  add an input keyword to enable blocking with QS
	  changed pw_grid printout

2008-08-15 07:42  tlaino

	* [r7491] use consistently rootpi and oorootpi instead of SQRT(pi)
	  and 1/SQRT(pi). reset
	  regtests for numerics..

2008-08-14 18:24  tlaino

	* [r7490] SPHERICAL CUTOFF for QMMM interactions both for SE and
	  GPW/GAPW. 1 more regtest.

2008-08-14 14:26  tlaino

	* [r7489] prettify

2008-08-14 11:34  krack

	* [r7488] ALLOCATABLE changed to POINTER attribute in derived type.
	  Otherwise the
	  AIX V11.1 compiler is crashing with an unspecific error message
	  in
	  force_env_types.F. Therefore I suggest not (yet) to use
	  ALLOCATABLES within
	  derived types to keep robustness.

2008-08-13 15:15  krack

	* [r7485] * constant rootpi corrected
	  * mathconstants cleaned
	  * 152 regtests reset, but more resets might be needed
	  * numerical differences (mostly QS inputs) are 1.0E-5 or less
	  * The ten largest differences are
	  1.38002651e-05
	  1.35046137e-05
	  6.15851759e-06
	  2.07762843e-06
	  1.91037461e-06
	  1.86513782e-06
	  2.67201873e-07
	  1.56685460e-07
	  6.65610387e-08
	  6.09160603e-08

2008-08-13 13:20  tlaino

	* [r7484] make symmetry OFF by default. must be explicitely
	  required to get symmetry info.

2008-08-13 10:00  krack

	* [r7482] * Goodbye eps0
	  * 16 regtests reset

2008-08-12 17:59  tlaino

	* [r7480] moving DDAPC main control routine to its original
	  location (more logical) since
	  its doing more than just applying restraints/constraints.
	  cleaning wrong doxygen
	  statements.

2008-08-12 15:55  fschiff

	* [r7478] made the DDAPC_RESTRAINT repeatable, shifted
	  TYPE_OF_DENSITY from
	  DENSITY_FITTING. Its now possible to use multiple restraints, but
	  still only one constraint

2008-08-12 13:52  krack

	* [r7477] * eps0 set to 1/(4*pi) (stage 1 of getting rid of eps0)
	  * 275 regtests reset

2008-08-12 08:52  tlaino

	* [r7476] prettify

2008-08-12 07:39  fschiff

	* [r7475] changed distance difference colvar in a linear
	  combination of twodifferences
	  now additives are possible via the new coefficient (-1 for
	  difference as it was)

2008-08-10 11:17  tlaino

	* [r7473] REAL -> REAL(KIND=dp)

2008-08-09 20:48  krack

	* [r7471] * 4 new regtest inputs for QS stress tensor (GPW) added
	  * new print section for variable format stress_tensor output
	  which allows for trace check in the regtester
	  * physcon updated and cleaned and new variables applied in other
	  routines

2008-08-08 08:23  fschiff

	* [r7470] #assert in case of none cubic cells for 0D periodicity

2008-08-07 12:59  krack

	* [r7469] * stress tensor for LSD with GGA functionals enabled
	  * regtests will follow

2008-07-29 12:26  tlaino

	* [r7468] bug fix

2008-07-29 10:14  tlaino

	* [r7467] restoring original version (fix a bug introduced with
	  last commit)

2008-07-28 11:43  tlaino

	* [r7465] more robust check to fix (?) a numerical problem..

2008-07-27 12:23  tlaino

	* [r7464] FIX SSMP compilation

2008-07-24 12:23  vondele

	* [r7462] new defaults for eps_ppnl and eps_ppl, now 'properly'
	  consistent with eps_default.
	  Saves memory and speeds up core_hamiltonian. 329 regtests reset.

2008-07-24 04:17  vondele

	* [r7461] added a mixed_bend_stretch term

2008-07-23 18:03  tlaino

	* [r7460] converting unformatted write into formatted to avoid
	  splitting into new lines
	  of html code with few compilers (intel) (report:Ilya).

2008-07-22 13:42  vondele

	* [r7459] bug fix where using the default error leads to an abort
	  due to a wrong file name being generated (cell_opt, but also
	  others)

2008-07-22 10:16  vondele

	* [r7458] use a format to write basis and potential section (list
	  directed can yield comma separated values (pathscale), which
	  causes the parser to choke)

2008-07-22 08:09  vondele

	* [r7457] greatly speed up the calculation of lumos
	  (qs_ot_eigensolver uses now an outer_scf approach)

2008-07-21 15:20  vondele

	* [r7456] fix cell_opt with keep_shape for orthorhombic cells so as
	  to retain off-diagonal 0.0 exactly (speedup QS)

2008-07-21 14:47  vondele

	* [r7455] forcefields: added cubic and morse functional forms for
	  bonds, and cubic for bends

2008-07-20 01:55  vweber

	* [r7454] posibility to choose the number of mixing step during the
	  scf (1reg).

2008-07-16 09:53  marcella

	* [r7453] pretty

2008-07-16 07:00  marcella

	* [r7447] localization of molecular orbitals extended to
	  non-orthorhombic cells: should work with the Jacobi and the crazy
	  angle method

2008-07-15 18:03  tlaino

	* [r7446] bug fix for shortrange + bug fix in the definition of the
	  taper. 2 resets.
	  shortrange SE : validated.

2008-07-15 11:14  tlaino

	* [r7445] splitting qs_scf_post and preparing the new module for
	  SE. very few functionality
	  have been checked/implemented.. will be implemented as they will
	  become needed.
	  at the moment MO_CUBES can be printed also for SE.. lots of work
	  still to do..

2008-07-14 12:54  vondele

	* [r7444] bug fix fft scratch routines using is_equal, which
	  triggered with QS an NPT.

2008-07-11 20:06  tlaino

	* [r7443] new taper function.. reset+1 new reg.

2008-07-11 15:35  vondele

	* [r7442] fix bug affecting flexible cell stuff iff started from an
	  orthorombic cell.
	  get_hinv: deleted
	  use cell_init to update a cell instead (which properly changes
	  the cell%orthorombic field)
	  reset 13 regtests

2008-07-07 09:53  krack

	* [r7441] * Alternative DFT+U methods LOWDIN and MULLIKEN added
	  based on the respective
	  population analyses
	  * Default method is now MULLIKEN instead of MULLIKEN_CHARGES
	  * Forces for method LOWDIN not yet fully implemented
	  * 4 new regtests added

2008-07-04 13:31  tlaino

	* [r7440] bug fix for multipoles potential, electrostatic field and
	  e.f.g

2008-07-04 12:24  tlaino

	* [r7439] bug fix reftraj msd parallel

2008-07-04 09:55  marcella

	* [r7438] small fixing in printing output

2008-07-04 07:32  vondele

	* [r7437] new keyword for STM image generation

2008-07-03 08:45  vweber

	* [r7436] switch to inverse overlap if precond not spd for direct
	  method

2008-07-02 14:03  vondele

	* [r7435] remove intent on pointers

2008-06-30 06:50  vondele

	* [r7434] added HFX bibliography entry

2008-06-27 13:50  vweber

	* [r7433] missing initialization.

2008-06-27 11:47  tlaino

	* [r7432] [CP2K:1136]: bug fix

2008-06-25 11:55  vweber

	* [r7431] posibility to use different solvers with
	  full_single_inverse and full_s_inverse preconds. activate few
	  regs.
	  fixed a bug with the s_inverse precond and reset a reg.

2008-06-24 14:31  vweber

	* [r7429] added direct solver (Cholesky factorization-triangular
	  solve) for the preconditioner.
	  activated regs.

2008-06-22 09:41  tlaino

	* [r7425] Bug fixes for potential, electrostatic field and
	  electrostatic field gradients (fully debugged,
	  there is only a minor issue still to be solved on the energy).

2008-06-20 14:45  vweber

	* [r7423] toward a sparse inner solver for preconditioning.

2008-06-20 14:00  tlaino

	* [r7422] minor recoding for multipoles + bug fixes for fields

2008-06-16 23:31  tlaino

	* [r7421] [CP2K:1113]: bug fix for topology reorder + 1 regtest
	  (QMMM+LINK) (csilla, teo)

2008-06-11 15:53  tlaino

	* [r7420] bug fix for g=0 term for ch-qu. debug code for
	  field+field gradients. reset numerics (1E-13).

2008-06-10 16:26  tlaino

	* [r7419] stop_program -> cp_unimplemented

2008-06-10 16:21  tlaino

	* [r7418] refactoring and cleaning

2008-06-10 14:55  tlaino

	* [r7417] bug fix

2008-06-10 13:44  vondele

	* [r7416] clean up

2008-06-10 12:59  tlaino

	* [r7415] Multipoles Ewald validated FORCES and bug fix for
	  multipolar tensors..

2008-06-10 09:11  tlaino

	* [r7414] allowing reftraj also with standard XYZ (not necessarily
	  with timings informations).
	  added 1 regtest.

2008-06-08 09:55  tlaino

	* [r7413] common debug routine for all methods (specific one for
	  FIST in the attic in case
	  someone may need it in the future) to make things easier and less
	  complicated.
	  Cleaning plenty of regtests in the FIST regtest directory.

2008-06-07 15:13  tlaino

	* [r7411] EWALD multipole dbgd and validated. Next to come forces
	  and e.f.g. added
	  several regtests to cover all possibilities of multipoles
	  interaction

2008-06-07 14:34  vondele

	* [r7410] Fix a bug with calculate_wavefunction for certain
	  distributed grids, plus small cleanup

2008-06-05 13:39  jgh

	* [r7403] Preconditioners for S=1 (NDDO); remove copy from
	  SINGLE_INVERSE;
	  6 new regtests

2008-06-05 11:28  vweber

	* [r7402] data parallel for expansion coefficients.

2008-06-04 12:47  tlaino

	* [r7401] prettify

2008-06-04 12:07  jgh

	* [r7400] Starting to code for S=1 preconditioners

2008-06-03 19:33  vondele

	* [r7399] clean-up of the molecular_states cubes (better file
	  names, title, and includes coordinates)

2008-06-03 18:46  vondele

	* [r7398] add citations for molecular_states

2008-06-03 17:51  vondele

	* [r7397] molecular states do not require molecular_dipoles

2008-06-03 09:11  vweber

	* [r7396] removed useless variables.

2008-06-02 15:17  fschiff

	* [r7395] print only absolut value of Hab

2008-06-02 14:51  vweber

	* [r7394] reorder matrix allocations and cleaning.

2008-05-31 07:26  vondele

	* [r7393] clarify error message

2008-05-30 10:52  vweber

	* [r7391] fixed bug with small systems in large box (null pointer).
	  added reg.

2008-05-30 09:24  mguidon

	* [r7390] Towards Wavefuntion-fitting, added Pulay forces for GPW

2008-05-29 09:18  vweber

	* [r7389] changed calculation of the aniso part of the CS.

2008-05-28 14:59  vweber

	* [r7388] cleanup. added regs nmr with both gapw/gpw.

2008-05-28 14:27  fschiff

	* [r7387] bug fix for LDA calculations

2008-05-28 12:22  vweber

	* [r7386] reduced mem needs.

2008-05-26 19:15  tlaino

	* [r7385] bug fix EAM + MASS in KIND for FIST (bug report: Carlo
	  Pignedoli). Added regtest

2008-05-26 11:39  jgh

	* [r7382] Generalize definition of alat to cases with a1(1)=0

2008-05-25 16:36  jgh

	* [r7381] Bugfixes for Dnh symmetry and atoms

2008-05-24 15:11  jgh

	* [r7379] Activate symmetry print section

2008-05-24 14:27  jgh

	* [r7376] Input for SYMMETRY keyword

2008-05-24 08:23  tlaino

	* [r7375] prettify

2008-05-23 15:29  jgh

	* [r7374] Molecular and Crystal symmetry routines; Only symmetry
	  detection is
	  activated, no effect on any calculations; regtests will be
	  updated soon.

2008-05-23 07:42  vondele

	* [r7373] valgrind bug fix for basis set none

2008-05-23 06:31  vondele

	* [r7372] allow for basis_set NONE in gpw calculations

2008-05-21 14:40  vweber

	* [r7370] removed useless arrays

2008-05-20 09:09  vondele

	* [r7369] code around a gfortran bug (and avoiding copy-in of
	  arrays)

2008-05-19 18:15  tlaino

	* [r7368] bug fix analytical derivatives core-core interactions
	  PM6. Covering all
	  special cases of core-core interactions with regtests (thanks to
	  joost for
	  the bug report).

2008-05-19 17:06  tlaino

	* [r7367] avoiding recursive I/O

2008-05-19 17:06  tlaino

	* [r7366] cleanup

2008-05-19 07:19  jgh

	* [r7365] Avoid SCF cycles when MAX_SCF=0

2008-05-16 14:44  vweber

	* [r7364] reduced mem usage and cleaning.

2008-05-15 19:57  tlaino

	* [r7363] SE integral in-core storage with the possibility of
	  compression. Default
	  now is storing integrals during the SCF cycle in memory (without
	  compression, which
	  is optional). Added few regtests.

2008-05-15 11:27  vweber

	* [r7362] reduced mem needs by 3.

2008-05-14 12:41  tlaino

	* [r7361] bug fix periodic colvar

2008-05-13 15:52  mguidon

	* [r7360] Fix memory leak introduced by previous commit

2008-05-13 14:47  tlaino

	* [r7359] .. and avoid printing parameters that are zero.

2008-05-13 14:41  tlaino

	* [r7358] printing also the VDW core-core parameters (if requested)

2008-05-13 14:41  tlaino

	* [r7357] fix mistype

2008-05-13 11:39  tlaino

	* [r7356] walkaround a NAG compiler bug

2008-05-13 11:14  vweber

	* [r7354] added boxing for orbital centers, and 2 regs.

2008-05-13 10:49  tlaino

	* [r7353] prettify

2008-05-13 09:53  mguidon

	* [r7352] Towards Wavefuntion-fitting, GPW energy/potential done,
	  derivatives wrt basis set origin and GAPW version still missing

2008-05-13 09:20  tlaino

	* [r7351] bug fix when matrix is allocated with size 0

2008-05-12 19:24  tlaino

	* [r7350] bug fix for Klopman-Ohno core term and added more
	  regtests. PM6 fully validated on
	  a set of 2021 molecules (see [CP2K:1030] on
	  http://groups.google.com/group/cp2k)

2008-05-12 11:06  tlaino

	* [r7349] bug fix for the Klopman-Ohno CORE term + unification.
	  reset for numerics.

2008-05-11 10:14  tlaino

	* [r7346] handling 0 electrons with OT

2008-05-11 09:31  tlaino

	* [r7345] more on d

2008-05-10 15:57  tlaino

	* [r7344] clean

2008-05-10 15:56  tlaino

	* [r7343] Enabling d-orbitals for more elements (and checking at
	  the parameterization
	  level if the element owns d-orbitals or not). Added the
	  possibility to skip
	  the taper function if RC_RANGE if <= 0.

2008-05-10 12:35  tlaino

	* [r7342] more units

2008-05-09 11:19  tlaino

	* [r7340] reorganization of the amn coefficients: starting for
	  periodic SE

2008-05-08 16:14  tlaino

	* [r7339] speedup

2008-05-08 10:21  tlaino

	* [r7338] optimizing SE derivatives

2008-05-06 05:13  tlaino

	* [r7337] fix doc

2008-05-05 20:56  vweber

	* [r7336] fixed a problem with the orbital centers and qmmm, added
	  3 regs.

2008-05-05 17:07  tlaino

	* [r7335] patching few problems for the fftsgl build (Ben Levine)

2008-05-05 16:59  tlaino

	* [r7334] cosmetics

2008-05-01 19:25  tlaino

	* [r7332] bug fix MOPAC_GUESS for UHF calculations. Added 1 regtest

2008-05-01 18:52  tlaino

	* [r7331] Analytical derivatives for d-orbitals! added few more
	  regtests together with
	  a "real" chemistry example.

2008-05-01 18:46  tlaino

	* [r7330] OT with MOPAC_GUESS

2008-05-01 16:09  tlaino

	* [r7328] new driver for sp (SE) analytical integrals derivatives
	  (1 reset: numerics)

2008-05-01 14:30  tlaino

	* [r7327] new integral framework invertion routine (handling also
	  d-orbitals). cleaning.

2008-05-01 07:15  tlaino

	* [r7326] bug fix for single precision. few more checks to
	  intercept arithmetic exceptions. cosmetics.

2008-04-30 15:29  tlaino

	* [r7325] bug fix for the 1center 2electron SE fock matrix. Bug fix
	  for f0sd and g2sd
	  parameters. PM6 validated (on HCl, BrCl and TiO) and working!
	  Added 3 more
	  regtests. The bug fix restores also the rotational invariant
	  property.

2008-04-30 07:38  tlaino

	* [r7324] prettify

2008-04-29 22:48  tlaino

	* [r7323] bug fix for periodic colvar (Rachel and Gerard). Reset 2
	  regtests.

2008-04-28 19:36  tlaino

	* [r7320] bug fix for mopac guess

2008-04-28 14:52  tlaino

	* [r7319] MOPAC_GUESS for initial guess

2008-04-28 10:51  tlaino

	* [r7318] prettify.. and there was no bug to fix..

2008-04-28 10:40  marcella

	* [r7317] bug fix

2008-04-28 09:10  marcella

	* [r7316] prettify

2008-04-28 08:50  marcella

	* [r7314] added the possibility to run MD with a variable timestep,
	  in order
	  to avoid "too large" atomic displacements between two consecutive
	  force evaluations

2008-04-28 07:50  vondele

	* [r7313] add a timing to monitor excessive io

2008-04-27 09:04  tlaino

	* [r7312] rewriting last commit removing the useless skip_this,
	  because:
	  (1) READ statements cut&pasted from the parser routine without
	  any kind of check
	  on the IOSTAT variables are extremely dangerous (do not provide
	  any sort of
	  feedback to users in case of end of file or errors in reading)..
	  Let's use
	  instead specific parser routines together with the bcast.. (these
	  routines
	  will provide at least error messages to the USER)
	  (2) The main MD loop MUST be kept lean and simple in the logic
	  (all the part
	  of the reftraj frame skipping can be directly moved with a much
	  more simple logic
	  into the reftraj integrator).. this makes the code easier to read
	  (3) output routines and free energies routines have been created
	  in such
	  a way that they shouldn't care about the type of ensemble..
	  and they must stay in this way.. if you think to introduce an
	  ensemble
	  variable in that part of the code very probably the modification
	  can be done
	  in a better, different way..
	  regtests OK.

2008-04-27 07:41  tlaino

	* [r7311] bug fix

2008-04-25 13:42  tlaino

	* [r7310] restoring lost (?!) code..

2008-04-25 08:54  tlaino

	* [r7309] cleaning

2008-04-25 08:21  marcella

	* [r7306] input for reftraj: using the standard parser for reading
	  trajectories of large systems
	  becomes very slow in parallel jobs, due to broadcasting of each
	  line. It is better to read
	  the entire snapshot and broadcast only once per snapshot.
	  bug fix in printing "time" in the xyz-trajectory file

2008-04-24 17:23  tlaino

	* [r7305] using symmetry also for the radial part of d-orbitals (no
	  need to have duplicated code)

2008-04-24 16:39  tlaino

	* [r7304] Deleting old code for generating SE integrals for
	  sp-basis and added a totally new
	  driver for this task (more compact and easy to understand):
	  finally some order also
	  in this part of the code.. next to come convert as well the
	  analytical gradients and
	  add analytical gradients for d-orbitals..

2008-04-24 14:14  vweber

	* [r7303] chi tensor was merged with nmr. dont need this guy
	  anyore.

2008-04-24 13:51  vweber

	* [r7302] fix previously introduiced bugs, distributed grid and
	  merge chi tensor from epr/nmr.

2008-04-23 18:14  tlaino

	* [r7300] bug fix in assigning ELEMENT to KIND (Chris)

2008-04-23 13:42  tlaino

	* [r7299] bug fix for periodic colvar in extended metadynamics.
	  Added units to LAMBDA.

2008-04-22 10:00  tlaino

	* [r7297] bug fix

2008-04-22 09:43  tlaino

	* [r7296] bug fix

2008-04-21 19:10  tlaino

	* [r7295] adding units to values of parameters for generic
	  functions (in case the values
	  of the parameters have different units you can split them in
	  different lines i.e.
	  making values and parameters repeatable).

2008-04-21 14:56  tlaino

	* [r7294] grouping common pieces of code (cleaning)

2008-04-21 14:56  tlaino

	* [r7293] fix memory leak

2008-04-21 12:47  tlaino

	* [r7292] another mistype fix (Juerg)

2008-04-21 12:10  tlaino

	* [r7291] update info

2008-04-21 12:04  vweber

	* [r7290] prepare for qmmm

2008-04-21 05:32  tlaino

	* [r7289] cleaning and completing the printing of SE parameters
	  (original and derived)

2008-04-21 05:32  tlaino

	* [r7288] new version for SE code

2008-04-20 20:43  tlaino

	* [r7287] YEAHHHHHHHHHHHHH!!!!!! d-orbitals for SE working!! ;)..
	  added 2 regtests (expect few more
	  bug fixes/improvements in the near future)..

2008-04-20 19:31  tlaino

	* [r7286] this was a VERY nasty one (was giving so far (for s and p
	  orbitals ) the right numbers but for a mutual
	  EXACT compensation of errors). The phase of the angular
	  interaction should be
	  the correct one now.. resetting for numerics few regtests
	  (10^-14-10^-11).

2008-04-20 15:32  tlaino

	* [r7285] first bunch of bug fixes for d-orbitals. Bug fix for
	  coefficient sto-6g.
	  Added the possibility to read the factor N (STO-NG) from input.
	  several
	  reset fue to numerics.

2008-04-20 12:44  tlaino

	* [r7284] bug fixes (uninitialized variables and string length
	  issue) (valgrind)

2008-04-18 15:38  tlaino

	* [r7283] prettify

2008-04-18 14:58  vweber

	* [r7282] updated the epr module for the new -centers-. activated
	  some regs.

2008-04-18 12:17  vweber

	* [r7280] new possible -centers- for the orbital dependent
	  response. reset few regs.

2008-04-18 10:10  tlaino

	* [r7279] prettify

2008-04-18 08:19  jgh

	* [r7277] Improved scalability for FFT (not yet fully accessible)

2008-04-17 14:28  tlaino

	* [r7276] cleaning

2008-04-17 14:11  mguidon

	* [r7275] Some more cleaning

2008-04-17 10:11  tlaino

	* [r7274] NEW SE fock matrices construction drivers (major
	  rewriting). Massive cleaning
	  and restructures for d-orbitals. Several regtests resetted.

2008-04-17 09:38  tlaino

	* [r7273] d0 -> _dp

2008-04-17 06:30  tlaino

	* [r7272] prettify

2008-04-16 16:03  mguidon

	* [r7271] Cosmetix

2008-04-16 15:04  mguidon

	* [r7270] Full source cleaning. Splitting the 300K beast into
	  several smaller pieces

2008-04-15 17:48  cjmundy

	* [r7265] Bug fix for parallel. Thanks Teo!
	  (a left over debug statement can blend in and look innocent,
	  but cause a day of intense stress.)

2008-04-14 06:47  tlaino

	* [r7263] new 1-center fock driver for semi-emprical code

2008-04-13 03:07  vweber

	* [r7261] rm useless files from previous merging.

2008-04-11 22:25  tlaino

	* [r7260] prettify and cleaning

2008-04-11 21:06  cjmundy

	* [r7258] Getting out the Raid and killing roaches: coulomb2_new is
	  now not screened. Some cleaning.
	  Added print key to print SAB_SCP neighbor lists for dispersion.
	  The "cutoff" for the
	  neigborlist is now gives converged results for 64 waters. Reset
	  regtests

2008-04-11 19:31  tlaino

	* [r7257] bug fix (NAG compiler)

2008-04-11 15:55  tlaino

	* [r7255] cleaning

2008-04-11 15:49  tlaino

	* [r7254] prettify

2008-04-11 15:36  vweber

	* [r7253] some cleaning. merged epr/nmr common routines.

2008-04-11 12:38  marcella

	* [r7252] bug fix

2008-04-11 11:21  mguidon

	* [r7251] HFX module extension: Added possibility to write
	  compressed ERI's to disk

2008-04-11 08:02  tlaino

	* [r7250] tiny fix

2008-04-11 07:22  tlaino

	* [r7249] prettify

2008-04-10 22:56  cjmundy

	* [r7247] Bug fix. Reset regtests.

2008-04-10 20:52  tlaino

	* [r7246] d-orbitals SE: few more structures needed and some
	  recoding (cosmetics).

2008-04-10 09:40  tlaino

	* [r7244] Taking out of the SCF the core-core energy/forces
	  interactions for SE. reset
	  few regtests.

2008-04-10 05:59  tlaino

	* [r7243] changing a variable name to bypass a PGI compiler bug!

2008-04-09 17:47  tlaino

	* [r7240] uniforming numerical VS analytical module for computing
	  integral derivatives.
	  Added 1 regtest.

2008-04-09 13:34  tlaino

	* [r7238] new integral module: generalization to spd basis set.
	  Reset few SE regtests.
	  Expect few more relevant changes for SE integrals.

2008-04-09 12:24  tlaino

	* [r7237] prettify

2008-04-08 16:56  tlaino

	* [r7236] mistyping (?)

2008-04-08 11:48  vweber

	* [r7234] Reinout: epr update, added a reg and reset the other.

2008-04-08 09:03  vondele

	* [r7233] adjust name to avoid clash with mpif.h

2008-04-06 07:32  vondele

	* [r7232] added IRAC reference

2008-04-04 19:17  tlaino

	* [r7231] creating wrappers for numerical/analytical integral
	  modules.

2008-04-04 16:56  tlaino

	* [r7230] getting rid of the taper function as module variables
	  (transforming into a
	  derived type)

2008-04-04 14:31  tlaino

	* [r7228] allocate structures in dft_control, only when strictly
	  required. making SE, DFTB, XAS, TDDFPT
	  sections required (better don't have sections accessible to avoid
	  unconsistencies with restarts).
	  Cosmetics.

2008-04-04 09:35  tlaino

	* [r7227] prettify

2008-04-04 08:53  vweber

	* [r7226] fix few bugs, starting merging epr/nmr common routines.
	  reset nmr regs.

2008-04-03 15:00  tlaino

	* [r7225] one step more towards working d-orbitals: calpar_d is
	  fully checked and validated.
	  Using Klopman-Ohno parameters for integral modules. error
	  propagated deeper in more
	  routines.

2008-04-02 05:32  vondele

	* [r7224] correct DOIs

2008-04-01 18:47  tlaino

	* [r7223] keeping only unique parameters between MNDO and MNDOd.
	  Adding references for
	  all SE methods. Fixing references printout for all methods using
	  the QS section.
	  Additional reference for gapw. MNDOd : validated for element not
	  containing d-orbitals (added more elements for plain MNDO and
	  MNDODd). 1 more
	  regtest.

2008-04-01 16:06  tlaino

	* [r7222] prettify

2008-04-01 09:35  vondele

	* [r7221] more INTENT(INOUT)'s

2008-04-01 09:01  vondele

	* [r7220] fix intent

2008-03-31 20:35  vondele

	* [r7219] improved ray distributed FFT (Juerg)

2008-03-31 19:47  tlaino

	* [r7218] more stuff on d-orbitals (not yet working) for
	  semi-empirical + MNDO-d parameters.
	  Reset 1 regtest due to a missing "dp" in the PDDG parameters.

2008-03-31 19:43  vondele

	* [r7217] use int_8 to get the correct order for the sort

2008-03-31 14:00  tlaino

	* [r7216] prettify

2008-03-31 10:54  mguidon

	* [r7215] bug fix

2008-03-31 04:45  tlaino

	* [r7214] CODATA 2006. reset most of regtests. CODATA 1998 still
	  available through a pre-processing flag.

2008-03-30 09:12  fawzi

	* [r7212] * matt fix for distribued non-ortho routines
	  * activating new non ortho routines

2008-03-29 17:14  tlaino

	* [r7211] bug fix: fix not associated pointer

2008-03-29 10:12  fawzi

	* [r7209] handle oversized rsgrids (by collocating just on a
	  subpiece)

2008-03-28 22:03  tlaino

	* [r7208] .. still breaking the regtest.. reverting to stable
	  version..

2008-03-28 17:29  fawzi

	* [r7207] * switching to faster non ortho routines (for real :)
	  * fixing bug in new routines in qs_collocate_density introduced
	  in 1.178

2008-03-28 16:49  tlaino

	* [r7206] prettify

2008-03-28 16:16  mguidon

	* [r7205] Added OMP critical section

2008-03-28 16:10  mguidon

	* [r7203] Enabled hybrid OpenMP/MPI version for exact exchange.
	  Works with gfortran and XLF

2008-03-28 15:10  tlaino

	* [r7202] reverting to previous file version: the last change was
	  breaking the regtest.

2008-03-28 14:19  fawzi

	* [r7201] introducing int_4 for more future proofing.

2008-03-28 14:18  fawzi

	* [r7200] switch to new routine for non ortho case.

2008-03-28 08:23  tlaino

	* [r7199] bug fix

2008-03-27 17:42  tlaino

	* [r7198] making a uniform behavior between all SE methods.

2008-03-27 15:05  tlaino

	* [r7197] Further refactoring/cleaning duplications and preliminary
	  structures for d-orbitals in
	  semi-empirical.

2008-03-27 07:55  vondele

	* [r7196] fix overflow in libtest

2008-03-26 22:47  tlaino

	* [r7194] prettify

2008-03-26 13:09  vondele

	* [r7193] remove unused dummy input sections

2008-03-25 22:35  ikuo

	* [r7192] typo in keyword

2008-03-25 19:06  vondele

	* [r7191] some more unused dummies

2008-03-25 18:33  vondele

	* [r7190] some never dereferenced stuff cleaned away

2008-03-25 17:50  tlaino

	* [r7189] handling exception: qm or mm region defined but this is
	  not a qmmm run..

2008-03-25 17:15  tlaino

	* [r7188] genpot functions et similia now can be 250 chars long.

2008-03-25 14:17  vondele

	* [r7187] removing unused globenv arguments

2008-03-25 13:43  vondele

	* [r7186] cleaning some of the ~1000 unused dummy arguments.

2008-03-24 14:42  tlaino

	* [r7184] propagating error into the integral functions and further
	  splitting of the
	  utility subroutines.

2008-03-23 22:51  tlaino

	* [r7183] new integral driver with few resets due to numerics
	  (10^-12).

2008-03-22 22:50  tlaino

	* [r7182] bug fix (Matt) for non-orthorombic cells (see [CP2K:867]
	  on the google group)

2008-03-21 20:15  tlaino

	* [r7181] new keyword to tune the convergence in cell optimization
	  run: PRESSURE_TOLERANCE.
	  added 1 regtest. internal pressure printed also with the
	  information at each
	  optimization step.

2008-03-21 17:44  tlaino

	* [r7180] sync with template

2008-03-21 17:42  tlaino

	* [r7179] bug fix: the otuput of the restart file in BFGS and
	  L-BFGS were not sync with
	  the present iteration step. Merged all output routine between all
	  methods and
	  made them more uniform. no further modifications to the
	  BFGS/L-BFGS/CG should be
	  required, modify instead the main output functions (this will
	  make the modification
	  available to all geometry optimizers). few regtests reset.

2008-03-21 11:09  tlaino

	* [r7178] clarifying error message

2008-03-21 07:09  vondele

	* [r7176] introduce a halo reduction scheme with subpatch exception
	  handling (Fawzi, Matt, Joost). Resets about 400 regtests

2008-03-21 00:33  tlaino

	* [r7175] bug fix for cell optimization. cell optimization
	  validated for QS/FIST methods.

2008-03-21 00:32  tlaino

	* [r7174] if forces are not computed let's give back a virial type
	  zeroed. few minor fixes.

2008-03-20 14:14  tlaino

	* [r7173] merging STRIDE and STRIDE_XYZ (no need to have two
	  different keywords): STRIDE
	  can now accept 1 or 3 numbers to specify the stride for the
	  printing of the whole
	  set of cube files.

2008-03-20 13:37  krack

	* [r7172] bug fix (crash on IBM AIX due to an attempt to close an
	  unit which wasn't
	  opened previously)

2008-03-20 11:02  krack

	* [r7171] * cube file output of the total (all-)electron density
	  including the
	  hard part for PAW atoms. Thus only GAPW runs are affected.
	  * doc updated (with caveats), one new keyword: TOTAL_DENSITY
	  * two new regtests added
	  * unit typo corrected in regtest H2O-xrd.inp

2008-03-20 08:15  marcella

	* [r7170] modified pdos-file header

2008-03-19 19:46  tlaino

	* [r7169] more units for keywords

2008-03-19 19:32  tlaino

	* [r7168] minor bug fix in line search 2pnt (both geo and dimer).
	  reset few regtests.

2008-03-19 17:35  ikuo

	* [r7167] Add CN6B atom prm type for charmm

2008-03-19 16:38  jgh

	* [r7166] Set correct return code if library is not available

2008-03-19 00:29  tlaino

	* [r7165] 2pnt line search for geo_opt in CG. bug fix in 2pnt line
	  search for dimer.
	  Enabling the cell_opt based on MD. added 2 regtests and resetted
	  the ones testing
	  dimer.

2008-03-18 14:47  tlaino

	* [r7163] fix doc

2008-03-18 12:05  mguidon

	* [r7162] Bug fix severe memory-leak on AIX platforms

2008-03-18 08:18  tlaino

	* [r7161] new pressure units: atm, kbar, Pa, MPa, GPa.

2008-03-18 07:37  tlaino

	* [r7160] force_eval initialization should never depend on the run
	  type. deleting
	  dft_control%forces, this is not how we decide if we want forces
	  or not..

2008-03-17 18:02  tlaino

	* [r7159] Cell Optimization algorithm (both at 0K and at finite
	  temperature): the
	  finite temperature algorithm is still partially unimplemented
	  (requires an
	  update for the geometry optimizer utilities, to come in the next
	  days).
	  Added 2 regtests (Teo e Paolo).

2008-03-17 17:58  tlaino

	* [r7158] bug fix memory leak

2008-03-17 14:23  krack

	* [r7157] * X-ray routine unitized, preparing for trial cube file
	  output of the total
	  electronic density
	  * missing regtest for X-ray diffraction spectrum added
	  * citations updated
	  * bug fix in XRD printout

2008-03-17 13:32  tlaino

	* [r7156] bug fixes

2008-03-17 12:37  tlaino

	* [r7155] bug fix calculation of intensities in parallel.

2008-03-17 06:28  tlaino

	* [r7154] cleaning

2008-03-15 17:04  tlaino

	* [r7151] Reorganization of the averages during an MD (creation of
	  an environment added
	  in md_env). Refactoring of the md_env module into md_env and
	  md_ener. Cleaning.

2008-03-15 15:47  vondele

	* [r7150] bug fix in deadlock detection

2008-03-15 15:39  vondele

	* [r7149] deal with the case where some CPUs have no matrix
	  elements (e.g. minimal basis H atom on 8 CPUs)

2008-03-15 15:30  tlaino

	* [r7148] clarify docs (Axel)

2008-03-14 16:41  tlaino

	* [r7147] deep cleaning

2008-03-14 06:52  tlaino

	* [r7146] prettify

2008-03-14 01:42  fawzi

	* [r7145] * faster subpatch colloc/int (more callgrind opts)
	  * preparing to replace non orthorombic routine

2008-03-13 22:00  tlaino

	* [r7144] Reorganizing the restart of the SCP. The restart
	  information are now collected in the
	  same file used for the WFN. This avoids possible bugs using
	  multiple force_eval and also
	  make the SCP restart much more coherent with the rest of the code
	  (SCP and WFN are essentially
	  two faces of the same thing). regtests ok.

2008-03-13 15:50  tlaino

	* [r7143] cleaning unused stuff left around..

2008-03-13 15:27  marcella

	* [r7142] control output of force evn info

2008-03-13 14:04  tlaino

	* [r7140] improving the line search for dimer method (if curvature
	  is positive let's move of the
	  maximum allowed step to escape the convex region AFAP.

2008-03-12 21:53  tlaino

	* [r7137] format (Axel)

2008-03-12 19:06  tlaino

	* [r7136] bug fix SE analytical gradients (thanks to Diederica for
	  bug report). added 1 regtest.

2008-03-11 23:19  tlaino

	* [r7134] remove possibly redundant and non-documented output
	  (Axel)

2008-03-11 17:46  tlaino

	* [r7133] External termination can be tuned with the
	  PROJECT_NAME.EXIT or as usual just
	  with EXIT (as well as inner loops like MD with EXIT_MD or
	  <PROJECT_NAME>.EXIT_MD..
	  this applied also to other generic kind of loops).
	  The possibility to tune the exit with the project_name is helpful
	  when running different projects in a same directory.

2008-03-11 17:42  tlaino

	* [r7132] using termination

2008-03-11 15:14  krack

	* [r7130] recover correct routine names

2008-03-11 14:51  fawzi

	* [r7128] speedup of subpatch colloc/integrate (callgrind
	  optimizations)

2008-03-11 14:36  tlaino

	* [r7127] bug fixes for DIMER method (tested on several systems as
	  well as 2D potentials). Added 2 regtests.

2008-03-11 14:35  tlaino

	* [r7126] splitting and reorganization for next commit

2008-03-11 07:42  vondele

	* [r7125] fix ncpu*natom**2 integer overflow, untested. (Matt &
	  Joost)

2008-03-10 15:59  vondele

	* [r7124] add a rs_pw_transfer test section to the library tests

2008-03-10 13:02  vondele

	* [r7123] cleanup library_tests and regtest it

2008-03-07 13:13  marcella

	* [r7122] comment

2008-03-07 13:10  marcella

	* [r7121] prettify

2008-03-07 13:03  marcella

	* [r7120] pdos extended to unoccupied mos

2008-03-07 09:48  tlaino

	* [r7118] temporarily reverting to the original color also for
	  section link

2008-03-07 09:38  tlaino

	* [r7117] minor fix: allowing gromos topology with CHARMM parameter
	  file (now fully disentangled..
	  you can use gromos topology + charmm paremters and viceversa)

2008-03-06 19:31  tlaino

	* [r7116] Bug Fix (thanks to Marcel Baer for raising the problem)
	  for gromos, connectivity and parmtype
	  have been disentangled (as it was in the original version,
	  deprecated with
	  a recent cleanup). Added 1 regtest covering this feature.

2008-03-06 18:40  tlaino

	* [r7115] prettify

2008-03-06 18:37  fawzi

	* [r7114] wrapped boundaries, integration with subpatch
	  collocation/integration

2008-03-06 11:27  tlaino

	* [r7113] valgrind bug fix: uninitialized value for SCP

2008-03-06 10:12  tlaino

	* [r7112] moved to the Attic unused module

2008-03-06 10:09  tlaino

	* [r7111] routine name: please follow always the standard name
	  convention!

2008-03-06 09:39  krack

	* [r7110] Adapt format for trajectory reading

2008-03-06 09:16  krack

	* [r7109] * Fix: allow Molden to parse trajectory xyz files, too,
	  even if the time is printed
	  * Corrected: "timestep" changed to "time", since the
	  instantaneous time not the timestep is printed

2008-03-06 07:33  tlaino

	* [r7106] adding a specifc option to have the extended title for
	  xmol (don't rely on the fact
	  that a value of a certain variable is zero.. it may not be
	  consistent)

2008-03-06 02:20  tlaino

	* [r7105] cuda patch: performs the scatter/gather operation on the
	  gpu before/after the fft (Ben Levine)

2008-03-06 01:57  tlaino

	* [r7104] Splitting pw_types into pw_types and pw_methods (pw_types
	  contains only types and
	  create/retain/release).

2008-03-05 14:13  krack

	* [r7103] remove useless timestep output for xmol output, if the
	  timestep is zero (e.g.
	  for GEO_OPT runs), which even prevents a correct parsing of the
	  energy by
	  MOLDEN.

2008-03-04 12:36  marcella

	* [r7101] pdos: append option

2008-03-04 08:45  marcella

	* [r7100] pdos: possibile print out of the angular momentum
	  components

2008-03-03 20:32  vondele

	* [r7099] allow for distributed rs grids also if pw grids are block
	  distributed (Matt Watkins)

2008-03-03 20:29  vondele

	* [r7098] prettify

2008-03-03 15:25  jgh

	* [r7097] Bug fix for block distributed FFT (scratch memory
	  allocation estimates)

2008-03-03 15:02  vondele

	* [r7096] workaround a stack size issue (Matt)

2008-03-03 04:16  tlaino

	* [r7094] bug fix memory leaks

2008-03-02 22:24  tlaino

	* [r7093] Converting comments into more meaningful(?) conditionals,
	  depending on the keyword enabling_color_tags.

2008-03-02 22:04  tlaino

	* [r7092] added the possibility to have an EXTERNAL_POTENTIAL.
	  Cleaning and Prettifying. added 1 regtest.

2008-03-02 20:05  vondele

	* [r7091] first patch towards ray parallel FFT

2008-03-02 16:40  vondele

	* [r7090] only enable once a meaningful (?) part of the code is
	  converted, now it is useless

2008-03-02 16:35  vondele

	* [r7089] by default use the same distribution_layout at all grid
	  levels for distributed multigrids

2008-03-01 16:33  tlaino

	* [r7088] robodoc -> doxygen

2008-03-01 15:19  tlaino

	* [r7087] collecting blank characters in kinds

2008-03-01 15:19  tlaino

	* [r7086] parser cleanup (Axel)

2008-03-01 09:24  tlaino

	* [r7085] references also for XML (Axel)

2008-02-29 23:31  tlaino

	* [r7083] bug fix

2008-02-29 20:54  marcella

	* [r7081] - projected dos: memory leaks

2008-02-29 20:05  marcella

	* [r7079] - calulation and printout of projected-dos (per kind and
	  angular momentum), a couple of regtest are added
	  - cleaning of some unused variables

2008-02-29 13:02  fschiff

	* [r7078] splitted vibr.anal. subroutine added several restart
	  possibilities to mode_selective
	  added a new choice for the selected vector (INVOLVED_ATOMS) which
	  allows to look for the
	  mode most localized on the selected atoms, removed preconditioner
	  (it's useless)

2008-02-29 08:42  fschiff

	* [r7077] bug fix for wavelet solver in parallel (bug in
	  redistribution of the density
	  bug fix for replica env (simultanious access of more than one
	  replica to restart crashed the calculation)

2008-02-28 22:10  cjmundy

	* [r7076] Initialized arrays to avoid underflow in unaccessed parts
	  of arrays. Doesn't effect results,
	  just looked scary and dangerous. Perhaps I am getting old.

2008-02-28 08:05  tlaino

	* [r7075] removed unused variables

2008-02-28 00:21  cjmundy

	* [r7074] Reactivated logger printing.

2008-02-27 21:30  tlaino

	* [r7073] moving pre-processing debug instructions into conditional
	  using the debug_this_module variable.
	  Prettify.

2008-02-27 20:07  tlaino

	* [r7072] bug fixes (out of string length)

2008-02-27 17:58  tlaino

	* [r7071] More pre-processing instruction for the input parser
	  (possibility to define
	  variables and use them within the input; IF construct; PRINT).
	  Added a regtest
	  testing several properties of the pre-processor (Axel)

2008-02-27 17:08  tlaino

	* [r7070] bug fixes (un-nullified pointers and un-initialized
	  variables) for SCP. changed
	  settings for regtests to have faster checkings..

2008-02-27 14:54  tlaino

	* [r7068] printing cell also in the geometry optimizer

2008-02-27 12:02  tlaino

	* [r7067] prettify

2008-02-27 11:47  tlaino

	* [r7066] typo (Axel)

2008-02-27 08:21  vondele

	* [r7065] code around limitations in the MPI e.g. on the XT3

2008-02-27 08:01  vondele

	* [r7064] Bug fix for distributed non-ortho grids (thanks Fawzi)..

2008-02-26 11:58  tlaino

	* [r7060] bug fix for mutual coexistence of restraints and
	  constraints

2008-02-25 16:11  tlaino

	* [r7059] Patch for improving the CUDA FFT performance and added
	  the infrastructure for
	  managing the GPU's memory (Ben Levine).

2008-02-25 09:20  tlaino

	* [r7058] prettify

2008-02-25 07:02  fawzi

	* [r7057] First version of subpatch collocate/integrate

2008-02-24 17:12  tlaino

	* [r7056] Intercepting user errors in case of mixed_env
	  calculations in the force_eval_order

2008-02-24 16:34  tlaino

	* [r7055] prettify

2008-02-24 14:17  tlaino

	* [r7054] bug fix: uppercase

2008-02-24 14:17  tlaino

	* [r7053] prettify

2008-02-23 23:22  cjmundy

	* [r7052] Make use of outer scf for much more efficient SCP.
	  Passed regtests on intel.
	  Almost have a set of stable tests to commit.

2008-02-23 22:05  tlaino

	* [r7051] Possibility to have growing target constraints for
	  collective section (teo and axel).
	  Added 1 regtest. On my way back to zurich..

2008-02-23 17:54  tlaino

	* [r7050] minors: typo in keyword and added the control over the
	  roll loop (introduced the
	  roll_tol keyword)

2008-02-23 01:12  tlaino

	* [r7049] Added a pre-processing capability to the parser (Axel and
	  Teo). It is possible
	  now to include files in the input file of cp2k using the keyword
	  @INCLUDE <FILENAME>.
	  This can be quite helpful for building sort of standard
	  template/libraries of part
	  of the input files (like for the GENPOT feaure) instead of
	  repeting the same amount
	  of lines thousands of times.
	  Possibly other pre processing instructions will follow in the
	  near future..
	  Added 1 regtest.

2008-02-22 17:21  tlaino

	* [r7048] QMMM quadratic walls (Ben Levine). added 1 regtest.

2008-02-22 17:20  tlaino

	* [r7047] prettify

2008-02-22 16:44  cjmundy

	* [r7046] Ooops... Forgot this file

2008-02-22 16:25  cjmundy

	* [r7045] Fixed parallel restarts.
	  cleaning

2008-02-22 16:24  cjmundy

	* [r7044] cleaning

2008-02-22 15:24  tlaino

	* [r7043] prettify

2008-02-21 19:02  cjmundy

	* [r7041] Fixed memory leaks.

2008-02-21 04:47  tlaino

	* [r7040] single precision issues for SHAKE and the XC

2008-02-21 04:47  tlaino

	* [r7039] sigle precision issue: making numerically stable
	  pw_grid_assign and pw_grid_count.

2008-02-20 22:51  cjmundy

	* [r7038] Initializing SCP guess to zero

2008-02-20 05:59  tlaino

	* [r7037] bug fix for PGI compiler!

2008-02-19 23:44  tlaino

	* [r7034] bug fixes (Axel)

2008-02-19 22:59  tlaino

	* [r7033] Cuda (NV graphic card) support for cp2k (Ben Levine).
	  This initial support
	  includes the possibility to perform FFT on graphic cards.
	  Requires update of the
	  makefile and of the entire cp2k tree. added 1 arch file.

2008-02-19 20:10  tlaino

	* [r7032] minutes.. but since we care about avoiding writing not
	  useful stuff
	  in the restart file.. let's avoid also the writing of the
	  barostat%print section
	  if not used..

2008-02-19 20:07  tlaino

	* [r7031] just two lines to avoid writing the THERMOSTAT section of
	  BAROSTAT if thermostat
	  is not used

2008-02-19 15:53  tlaino

	* [r7030] Restoring original files.. it's not just a matter of used
	  or unused variables. The logic
	  of this part should not be changed! if you do so.. please let's
	  discuss about it first!

2008-02-19 15:33  marcella

	* [r7029] cleaned unused variables

2008-02-19 14:47  marcella

	* [r7027] handling restart of the barostat even when the thermostat
	  of the barostat is not defined,
	  output of the barostat temperature independent from the presence
	  of the thermostat,
	  rescaling of barostat temperature to keep it within a given range

2008-02-19 12:06  tlaino

	* [r7026] cell dimensions can now be read from an external file
	  (cp2k native and namd
	  xsc format). 1 more regtest. important: cell print_key does not
	  depend anylonger
	  on the type of ensemble. It can be printed, on request, also
	  during a non NP* ensemble (this is
	  useful for example for post-processing or visualization). Changed
	  the print_level
	  of cell and velocities (default now is high_print_level).

2008-02-19 03:56  tlaino

	* [r7025] Print_key to print temperatures of the different
	  thermostatting regions. Better logical
	  organization of the print_key to print barostat and thermostat
	  energy: have been moved in
	  the barostat and in the thermostat region (these keywords do not
	  make sense in the MD%PRINT
	  since in case of new thermostats or of the already present like
	  shells and coeffs would
	  require a continuous modification of the code.. this is not OO..
	  while having a proper
	  print_key in the thermostat or barostat section makes everything
	  much easier).
	  Cleaned further the velocity verlet (the ensemble MUST be method
	  indipendent!).
	  Possible further splitting in the next commits..

2008-02-17 20:49  marcella

	* [r7023] isotropic npe ensemble, optional printing of thermostat
	  and barostat energies to external output files,
	  control on the correct definition of the thermostat if nvt or npt

2008-02-17 14:16  tlaino

	* [r7022] bug fix (valgrind)

2008-02-17 13:22  tlaino

	* [r7021] prettify

2008-02-16 20:13  cjmundy

	* [r7020] More functionality for the SCP:
	  -Multipole dipersion debugged
	  -Restarts
	  -New and improved input
	  -Add scp neighbor list for the dispersion (bit of a hack now)
	  -Passied Intel regtests (that's all I have)
	  -Examples to come

2008-02-16 07:57  tlaino

	* [r7019] bug fix for reflective walls (both lagrangian and non).
	  reset 1 regtest (thanks
	  to Minghsun and Axel for the help.. and to the lights of broadway
	  that are
	  just amazing at 3 o'clock in the night!)

2008-02-16 02:31  tlaino

	* [r7018] Restoring some of the logic lost with the last changes on
	  thermostats.
	  Specifically (1) thermostat section cannot be created with
	  required=.FALSE. .
	  (2) There's no need to read the same keyword N times. 1 is
	  enough.
	  cosmetics

2008-02-15 17:30  marcella

	* [r7016] New ensamble that allows to run at constant pressure
	  without thermostats, some regtests are added
	  I called it npe, to distinguish it from the cp2k nph, which is
	  used for shockwaves.
	  With npe at the moment the thermostat for the barostat is not
	  active, but its initial temperature
	  can be given as input, and it is possible to rescale it step by
	  step by using an annealing factor.
	  Also the annealing of the ion vellocities is possible by npe.

2008-02-14 23:35  tlaino

	* [r7015] bug-fix: more accurate control for non-periodic
	  calculations (Ben)

2008-02-14 18:34  tlaino

	* [r7014] bug fix for mixed_env logger (weird behaviors) [Ben
	  Levine]

2008-02-14 18:32  tlaino

	* [r7013] prettify

2008-02-14 16:20  vondele

	* [r7012] bug fix for distributed non-orthorombic grids (Fawzi)

2008-02-13 18:02  fawzi

	* [r7011] always print restart (if requested) when exiting with
	  touch EXIT, making compiler happy

2008-02-13 17:52  tlaino

	* [r7010] bug fix for MOLECULAR_DIPOLES: last commit on
	  localization broke the calculation
	  of molecular dipoles. Restored. Also:
	  1) bug fix for molecular dipoles for GAPW
	  2) unforming MOLECULAR_DIPOLES to TOTAL_DIPOLE (printed only if
	  explicitly required)

2008-02-12 14:29  fawzi

	* [r7009] add work dir to f77_interface arguments, add destroy env
	  to shell commands.

2008-02-08 17:45  tlaino

	* [r7006] doc typo

2008-02-08 17:41  tlaino

	* [r7005] as marcella pointed out annealing+thermo does not make
	  sense. deleted.

2008-02-08 14:34  marcella

	* [r7004] warning if annealing with thermostat

2008-02-08 14:25  tlaino

	* [r7003] another uniforming issue: using parser for IO operations
	  with reftraj

2008-02-08 12:16  marcella

	* [r7002] warning if annealing with thermostat

2008-02-08 01:33  tlaino

	* [r7001] Further re-organization after the last common code
	  grouping: splitting integrator
	  and keeping in the integrator module only the INTEGRATORS.
	  Getting rid of the
	  module variables (tmp and old) which become now local to each
	  integrator. prettify..

2008-02-08 00:46  tlaino

	* [r7000] typo bug fix

2008-02-08 00:08  marcella

	* [r6999] small changes

2008-02-07 20:22  tlaino

	* [r6998] prettify

2008-02-07 20:10  tlaino

	* [r6997] bug fix for variable cell in reftraj (energies were not
	  consistent with the
	  ones of the trajectory, now they are..). Assumed the same
	  convention for the
	  name of the keyword specifying file_names. added 2 regtest for
	  reftraj variable cell.

2008-02-07 18:29  marcella

	* [r6995] integrator module: some new function to better organize
	  repeated operation
	  realspace_grid_cube: added the possibility to use different
	  stride in x y and z
	  reftraj : new option to check the displaced atoms

2008-02-07 11:34  tlaino

	* [r6994] bug fix when print V_HARTREE (thanks to Gianluca and
	  Marcella)

2008-02-06 22:49  tlaino

	* [r6993] bug fix when dumping the PSF (correct resid)

2008-02-06 19:47  tlaino

	* [r6992] topology generator more robust.. introduced the
	  possibility to specify isolated atom
	  in the topology generation and added the possibility to use field
	  81-> of pdb for
	  reading charge (charge_extended) without any problem of number of
	  digits allowed.
	  Added 1 regtest for GENERATE%ISOLATED_ATOMS

2008-02-06 10:23  tlaino

	* [r6991] typo bug fix (thanks Ben!)

2008-02-06 09:55  tlaino

	* [r6990] speed-up of FIST setup

2008-02-06 09:48  tlaino

	* [r6988] prettify

2008-02-04 20:50  fawzi

	* [r6987] added cd/pwd, made setpos return change

2008-02-04 16:40  jgh

	* [r6986] Add additional variation on core energy in NDDO/PM6

2008-02-04 09:46  mguidon

	* [r6980] Cosmetics + added two regtests

2008-02-03 18:57  tlaino

	* [r6978] bug fix

2008-02-03 15:50  tlaino

	* [r6975] PM6 core-core analytical derivatives

2008-02-03 13:42  tlaino

	* [r6974] cleaning md_energies.. collecting common routines
	  removing useless code..

2008-02-02 22:14  tlaino

	* [r6973] bug fix

2008-02-02 20:45  tlaino

	* [r6972] Enabling the printing of STRESS tensor also during
	  GEO_OPT

2008-02-02 15:39  tlaino

	* [r6971] recovered original

2008-02-02 15:36  tlaino

	* [r6970] Converted more obsolete routine's documentation into
	  doxygen. Please avoid to
	  use since now on the spurious "Purpose:" statement, use doxygen
	  style instead.

2008-02-02 11:36  tlaino

	* [r6969] added missing separator lines in the very first doxygen
	  statements

2008-02-02 11:34  tlaino

	* [r6968] more on doxygen (avoiding two white spaces at the
	  beginning of the doxygen instruction)

2008-02-02 11:20  tlaino

	* [r6967] new doxygen templates for module header and
	  subroutine/functions/types header

2008-02-01 20:56  tlaino

	* [r6966] pretty again the full source and cosmetics

2008-02-01 19:55  fawzi

	* [r6965] the doxygen swat team (fawzi & teo) kisses robodoc
	  goodbye

2008-02-01 19:13  tlaino

	* [r6964] converted templates to doxygen

2008-02-01 19:02  fawzi

	* [r6963] converted templates to doxygen

2008-02-01 18:30  fawzi

	* [r6962] fixing robodoc comment for conversion

2008-02-01 18:27  tlaino

	* [r6961] fixing robodoc comment for conversion

2008-02-01 18:24  fawzi

	* [r6960] fixing robodoc comment for conversion

2008-02-01 18:15  fawzi

	* [r6959] fixing robodoc comment for conversion

2008-02-01 18:08  fawzi

	* [r6958] fixing robodoc comment for conversion

2008-02-01 18:05  tlaino

	* [r6957] fixing robodoc comment for conversion

2008-02-01 17:50  fawzi

	* [r6955] fixing robodoc comment for conversion

2008-02-01 17:20  tlaino

	* [r6952] fixing robodoc comment for conversion

2008-02-01 17:13  fawzi

	* [r6951] fixing robodoc comment for conversion

2008-02-01 16:49  tlaino

	* [r6949] fixing robodoc for doxygen converter

2008-02-01 16:42  fawzi

	* [r6948] fixing robodoc comment for conversion

2008-02-01 16:32  tlaino

	* [r6947] fixing robodoc comment for conversion

2008-02-01 16:22  fawzi

	* [r6946] fixing robodoc comment for conversion

2008-02-01 16:07  fawzi

	* [r6945] fixing robodoc for doxgen converter

2008-02-01 14:16  jgh

	* [r6943] Bug fix in hard core repulsion (PM6/NDDO)

2008-02-01 13:56  mguidon

	* [r6942] Towards MCY3 sic-hybrid: Added scaled B88 long-range
	  functional. Thats it. regtests to follow

2008-02-01 13:42  tlaino

	* [r6941] prettifying also the lib source files..

2008-02-01 13:01  tlaino

	* [r6940] some splitting and getting ready to use the
	  dyn_coeff_type also for other purposes..

2008-02-01 09:44  vondele

	* [r6939] updated prettify script parses these files as well.

2008-01-31 23:13  tlaino

	* [r6937] bug fix: this timeset/timestop needs a different tag

2008-01-31 22:59  tlaino

	* [r6936] let's use everywhere routineP instead of routineN in CPP
	  macros.. the number of arguments
	  must include at least the failure_level.. one may want possibly
	  to use it with a different
	  failure_level.. all the rest 'routineP,error,failure' IMO can
	  go..

2008-01-31 22:54  vondele

	* [r6935] switch more timeset to use routineN

2008-01-31 22:25  vondele

	* [r6934] cosmetics

2008-01-31 22:16  vondele

	* [r6933] use routineN instead of routineP in timeset.

2008-01-31 22:10  vondele

	* [r6932] use routineP instead on routineN in
	  CPPostcondition/CPPrecondition.
	  At this point it would make sense to have just a single argument
	  in CPPostcondition (just the condition),
	  the others are almost always
	  'cp_failure_level,routineP,error,failure'

2008-01-31 21:34  tlaino

	* [r6931] Few more routine names sync.. the prettifier fails in
	  prettfying the following
	  sources (at least.. possibly more): hartree_local_2centers.F,
	  hartree_local_3centers.F,
	  hartree_local_methods.F, scp_hartree_1center.F (reason for the
	  failure in prettifying: UNKNOWN).

2008-01-31 17:38  tlaino

	* [r6929] getting rid of few more non-standard ways of defining
	  routine name..

2008-01-31 17:22  vondele

	* [r6928] new prettify script (Fawzi) now deals correctly with all
	  sources. A 'make pretty' target has been added to the Makefile.

2008-01-31 16:52  jgh

	* [r6927] A few updates for PM6/NDDO

2008-01-31 11:29  tlaino

	* [r6922] cosmetics and added few regtests for PM6

2008-01-31 09:02  mguidon

	* [r6920] Towards MCY3 sic-hybrid: Make it MD-able, i.e. scale HF
	  forces

2008-01-30 22:15  tlaino

	* [r6919] restoring the original format.. (this kind of formats
	  should be kept through the prettifier)

2008-01-30 20:15  tlaino

	* [r6918] few bug fixes for NAG pdbg

2008-01-30 20:15  tlaino

	* [r6917] just to keep some order..

2008-01-30 18:56  jgh

	* [r6916] A further step towards NDDO/PM6

2008-01-30 16:14  vondele

	* [r6914] bug fix for the case of distributed grids and DFTB

2008-01-30 15:17  mguidon

	* [r6912] Remove stupid code line (thanks to teo)

2008-01-30 15:05  fawzi

	* [r6911] more commands, improved asyncrhonous execution, reduced
	  accuracy (for intel compiler)

2008-01-30 13:38  mguidon

	* [r6909] Clean up and bug fix

2008-01-30 12:34  mguidon

	* [r6908] single precision issue

2008-01-30 12:30  mguidon

	* [r6907] Towards MCY3 sic-hybrid: Glued everything together

2008-01-28 20:36  tlaino

	* [r6904] bug fix

2008-01-28 16:13  tlaino

	* [r6903] PM6 atomic parameters (thanks to Samuele Giani).
	  Propagating error also into
	  more allocate/deallocate subroutine.

2008-01-28 12:13  tlaino

	* [r6902] Restart for the Dimer Method (restarting the dimer vector
	  orientation)

2008-01-27 17:00  tlaino

	* [r6899] DBL -> REAL(.. , KIND=dp) . Of course in the last commit
	  message "getting rid" -> "getting ready" ( ;-) sorry about my
	  dyslexia)

2008-01-27 16:35  tlaino

	* [r6898] monthly cleaning.. getting rid to the full source
	  cleaning planned in 1 week.
	  Prettified source code and, as usual, moved to the Attic not
	  compiled or
	  compiled but never USEd modules.

2008-01-27 14:20  tlaino

	* [r6897] Updating 2007 -> 2008

2008-01-27 14:20  tlaino

	* [r6896] uniforming to the general rule of having module/program
	  name equal to the
	  file name

2008-01-27 10:49  krack

	* [r6895] previous change breaks doxygen/latex, now cpp and doxygen
	  do not complain

2008-01-26 20:30  tlaino

	* [r6894] bug fix

2008-01-26 19:26  tlaino

	* [r6893] Restart for BSSE runs + plus 2 new regtests (BSSE
	  restart)

2008-01-26 15:15  tlaino

	* [r6892] bug fix for backup of print_keys active at different
	  iteration levels during the
	  same run (regtest added).

2008-01-26 14:31  tlaino

	* [r6891] remove white space after // at the end of the comment
	  line to make also CPP happy about CP2K!

2008-01-25 21:42  tlaino

	* [r6890] reset to the original..

2008-01-25 21:28  cjmundy

	* [r6889] Initial commit of the Self-consitent Polarization (SCP)
	  method that should
	  sit on top of CP2K's force fields. Now is hooked up with QS.
	  Dispersion
	  corrections in the making. Regtests to come.

2008-01-25 16:52  tlaino

	* [r6888] typo

2008-01-25 16:37  tlaino

	* [r6887] cleaning globenv (we don't need project_name in it).
	  Changing also the
	  behavior of the coord_file_name and conn_file_name.. when cp2k
	  tries to read
	  them they must be present in the input (otherwise an error pops
	  up).

2008-01-25 16:00  tlaino

	* [r6886] Doing an ENERGY_FORCE also for FIST. MC should seriously
	  be re-written..
	  it wastes an incredibly amount of time in the IO routines. Note:
	  since few
	  recent commits one of the regtests of MC (GEMC_swap) may fail
	  with NAG/sdbg because the
	  amount of time spent in that routine is increased enourmously
	  without any
	  clear reason.. the code looks clean (at valgrind).. Mostly sure
	  is a problem of how NAG
	  handles enormous amount of allocation/deallocation (the first are
	  very fast
	  and they get slower and slower.. sigh..) Anyway those
	  alloc/dealloc should be
	  avoided.. we don't need to dump all the time an input on the file
	  and then
	  re-read it..

2008-01-25 08:23  jgh

	* [r6885] Prepare for new NDDO parametrizations

2008-01-24 19:12  tlaino

	* [r6884] bug fix: cleaning the spurious translational components..

2008-01-24 18:20  tlaino

	* [r6883] making TDDFPT section required

2008-01-24 15:57  tlaino

	* [r6881] splitting iteration_types and introducing a further
	  control when adding iteration levels..

2008-01-24 12:33  mguidon

	* [r6880] Towards MCY3 sic-hybrid: Implement scaling of
	  xc-potential (GPW/GAPW). Keep local copy of Fock-Matrix before
	  adding to Kohn-Sham matrix

2008-01-24 00:53  tlaino

	* [r6878] Converting the EACH keyword of the print_key sections
	  into a section and
	  specifying a more easy to use way for the iteration levels. The
	  section
	  EACH constains several keywords each attached to a specific each
	  iteration
	  level (for example the keyword MD that maps the steps of the MD,
	  and so on).
	  This should make more user-friendly how to decide the frequency
	  of a print
	  for each print_key..
	  fixed a bug in the routine deciding if the present iteration
	  level was the
	  right one to print (i.e. in the old version very weird behavior
	  with the
	  keyword EACH was possible). Now it should show the right
	  behavior.
	  Regtests have been properly adapted.
	  Moreover the iteration_info has been fully made consistent with
	  retain/release.
	  This makes impossible (gives an error at run time) to create
	  iteration levels and
	  destroy them randomly or even keep them allocated. They must be
	  destroyed in the
	  inverse order they are created. When adding a new iteration level
	  remember to update also the create_print_key_section properly (if
	  not so a
	  runtime error will pop up..).
	  RESTART keywords have been also specified with the
	  add_last_numeric specification
	  (this was essentially the behavior we had so far, though induced
	  by the bug
	  present in the code, much more clean now).

2008-01-23 16:52  jgh

	* [r6876] STO-nG parameters for 6s-7i functions

2008-01-23 14:55  tlaino

	* [r6875] add_last is just an integer.. converting and cleaning for
	  add_last..

2008-01-23 13:04  tlaino

	* [r6872] getting rid of the limitation on the size of the
	  iteration_level and deleting
	  unuseful subroutine.. prettify

2008-01-23 13:04  vondele

	* [r6871] fix a memory leak for tiny systems on a large number of
	  CPUs

2008-01-23 12:53  vondele

	* [r6870] turn a warning into a failure, and collect some info if
	  this happens again.

2008-01-23 11:24  mguidon

	* [r6868] Towards MCY3 sic-hybrid: Enable mixed Gaussian/Longrange
	  Hartree-Fock calculations

2008-01-23 07:54  vondele

	* [r6867] small cleanup/speedup (Matt)

2008-01-22 22:48  tlaino

	* [r6866] re-grouping the iteration levels for GEO_OPT

2008-01-22 22:48  tlaino

	* [r6865] added a LOG_PRINT_KEY keyword to the print_key section

2008-01-22 16:35  tlaino

	* [r6862] single precision issue

2008-01-22 16:25  mguidon

	* [r6860] Towards MCY3 sic-hybrid: Prepare CP2K for adiabatic LYP
	  functional. Bug fix in scaled lsd-LYP -> reset 1 regtest.

2008-01-22 15:40  tlaino

	* [r6859] cleaning

2008-01-22 15:27  vondele

	* [r6858] fix a bug with ghost atoms as pointed out by Matt

2008-01-22 14:46  tlaino

	* [r6857] Dimer Method working with Conjugate Gradients. Added 1
	  regtest. (Luca Bellucci and Teo)

2008-01-22 13:04  vondele

	* [r6855] also sort tasks in the serial case, reset a large number
	  or regtests

2008-01-22 10:51  tlaino

	* [r6853] single precision issue

2008-01-22 10:23  mguidon

	* [r6852] Towards MCY3 sic-hybrid: Enable Gaussian interaction
	  potential in two electron integrals

2008-01-21 22:17  tlaino

	* [r6851] bug fix in parallel

2008-01-21 20:10  vondele

	* [r6850] Sort tasks with respect to grid levels (some cache
	  related speedup) and fix a buglet in realspace_task_selection.

2008-01-21 16:31  mguidon

	* [r6849] Bug fix parallel + removing stupid call from previous
	  commit

2008-01-21 16:27  tlaino

	* [r6848] prettify

2008-01-21 16:26  tlaino

	* [r6847] bug fix: compilers may complain that unit is not
	  attached..

2008-01-21 16:26  tlaino

	* [r6846] typo

2008-01-21 14:57  mguidon

	* [r6843] Towards MCY3 sic-hybrid: Enable multiple x_hf envs

2008-01-21 14:48  fawzi

	* [r6841] first version of cp2k interactive shell (input control
	  not yet in)

2008-01-21 14:47  fawzi

	* [r6840] correct eof iostat on g95

2008-01-21 14:45  fawzi

	* [r6839] better error message

2008-01-21 11:27  vondele

	* [r6838] increase tolerance for check

2008-01-21 11:14  tlaino

	* [r6837] added a print_key for COM and cleaned the printed of the
	  COM info during an MD

2008-01-21 11:13  tlaino

	* [r6836] bug fix

2008-01-21 09:22  krack

	* [r6835] bug fix: missing if clause added for parallel runs

2008-01-21 02:08  tlaino

	* [r6834] cleaning and bug fix a possible problem when checking the
	  existence of the
	  FF parameter file. Making more consistent the parmtype and
	  parameter_file_name
	  in the input. prettify!

2008-01-20 23:25  tlaino

	* [r6833] Dimer Method: first version working with SD. Cleaning and
	  reorganization of the
	  CG section, gopt_para_type. Getting rid of useless variables.
	  Introduced the
	  possibility to have different kind of LS (still not fully
	  implemented). Added
	  1 regtest for Dimer Method with SD. Common message to all geo_opt
	  methods in case
	  of non-convergence. Prettifying.

2008-01-20 10:23  tlaino

	* [r6832] bug fix for HESSIAN print_key (BFGS). Added 2 regtests.

2008-01-19 20:28  tlaino

	* [r6831] Yet another step towards the Dimer Method. Moreover
	  geometry optimizers are
	  now allowed to call each others recursively.. some splitting of
	  files in a more
	  logical way.. cleaning and prettifying.. (Luca Bellucci and Teo)

2008-01-18 20:24  vondele

	* [r6830] print more digits, update keyword

2008-01-18 18:54  tlaino

	* [r6829] splitting lbfgs from geo_opt module. Introducing a common
	  header for geometry
	  optimizations. Introducing a cp_rot_opt for dimer method. Several
	  other adjustments.
	  Prettifying on the way to the dimer method (Luca Bellucci & Teo)

2008-01-18 15:43  krack

	* [r6828] some blanks removed

2008-01-18 14:35  tlaino

	* [r6827] bug fixing the hardcoded reference to GEO_OPT.. it must
	  be flexible indeed.. few
	  more structures useful for dimer (Luca Bellucci e Teodoro Laino,
	  even if not mentioned
	  in the future is implicit that the Dimer Method is mainly
	  controlled by Luca).

2008-01-18 13:46  tlaino

	* [r6826] more on dimer method.. not yet there...

2008-01-18 12:53  krack

	* [r6825] improved Doxygen style for example routines (details
	  follow by mail)

2008-01-18 10:27  vondele

	* [r6824] improve the load balancing using a task cost model

2008-01-17 15:52  vondele

	* [r6823] add an additional load-balancing step to rs_get_my_tasks,
	  further improving scalability (Matt Watkins and Joost).

2008-01-16 22:18  tlaino

	* [r6822] bug fix for fft_usage for FIST, bug fix in
	  pw_grid_setup_init (typo bug), clarifying
	  input keyword. Prettify. For more details about the bug see
	  http://groups.google.com/group/cp2k
	  message [CP2K:582] "cp2k chooses unsupported FFT grids". Thanks
	  to Axel for bug report

2008-01-16 20:59  vondele

	* [r6821] tighter neighbor bound calculation (2 CPUs are enough to
	  distribute)

2008-01-16 20:43  vondele

	* [r6820] bug fix the bug fix

2008-01-16 19:46  vondele

	* [r6819] bug fix for nodes without tasks

2008-01-16 18:40  vondele

	* [r6818] some renaming of tasks into cores

2008-01-16 14:37  vondele

	* [r6817] add an additional heuristic for deciding between
	  distributed/replicated, and made a parameter input-able

2008-01-16 13:20  mguidon

	* [r6815] New hybrid functional rCAM-B3LYP. See /QS/regtest-hybrid
	  for usage

2008-01-16 13:09  vondele

	* [r6814] fix a memory leak

2008-01-16 11:02  krack

	* [r6813] bug fix (thanks to NAG and Teo)

2008-01-16 10:24  krack

	* [r6812] debug lines removed

2008-01-16 09:59  krack

	* [r6811] * DFT+U method based on Mulliken charges added
	  * New regtest directory with four new regtests for DFT+U method
	  added
	  * New header style readable by Doxygen introduced

2008-01-15 19:17  vondele

	* [r6810] generate collocate/integrate task lists (QS) only once
	  per ionic step

2008-01-15 19:14  vondele

	* [r6809] remove debug statement

2008-01-15 19:00  tlaino

	* [r6808] bug fix (bug reported by Axel, see [CP2K:579] on google
	  group).

2008-01-14 22:33  tlaino

	* [r6806] Dimer Method to optimize transition states (not yet fully
	  working) [Luca Bellucci and Teo].
	  Cleaning of the geo_opt types and splitting the content into
	  several indipendent files.
	  Final working version of the Dimer method in the next days.

2008-01-14 20:31  tlaino

	* [r6805] bug fix

2008-01-14 19:20  vondele

	* [r6804] make the rs_grid distribution input tunable (see
	  tests/QS/regtest-gpw-4/H2O-8.inp) for an example.

2008-01-14 12:20  vondele

	* [r6803] remove some unused subroutines

2008-01-14 09:09  vondele

	* [r6802] document that FFTSG is required for the wavelet solver

2008-01-14 07:55  vondele

	* [r6800] a somewhat cleaner approach to avoid the overlapping
	  boundary problem.

2008-01-13 09:50  tlaino

	* [r6799] bug fix (valgrind)

2008-01-12 21:10  tlaino

	* [r6798] cleaning the global section (deleting never used
	  keywords), assigning few more
	  supported_feature to keys/sections. Splitting the input_cp2k into
	  three files.
	  cosmetics.

2008-01-11 18:24  vondele

	* [r6797] factor redistribute_matrix out of rs_get_my_tasks

2008-01-11 17:09  tlaino

	* [r6796] shortcuts for B3LYP and PBE0 (added 2 regtests).

2008-01-11 15:15  tlaino

	* [r6795] prettify

2008-01-11 14:52  tlaino

	* [r6794] my 2c to keep down the entropy

2008-01-11 14:27  vondele

	* [r6793] extra comment

2008-01-11 14:15  vondele

	* [r6792] add a few more headers

2008-01-11 14:05  vondele

	* [r6791] remove more out-dated OMP statements

2008-01-11 13:33  tlaino

	* [r6790] ... and keep the OMP version still working

2008-01-11 13:20  vondele

	* [r6789] remove debug

2008-01-11 12:05  vondele

	* [r6788] Full rewrite of the integrate and collocate routines.
	  Now, distributed realspace grids are the default for QS as well,
	  leading to significant memory reduction and some speedup in
	  parallel.
	  Expect follow-up commits in this area and report problems.
	  Several regtests reset due to numerics.
	  (Matt Watkins and Joost)

2008-01-11 11:57  tlaino

	* [r6787] ... and of course deleting useless keywords..

2008-01-11 11:31  tlaino

	* [r6786] Merging the two localization modules (deleted the old
	  ones and imported the
	  features present in the old one into the new one). Now there's
	  only 1 section
	  controlling the localization of the MOs (LOCALIZE). No regtests
	  resetted. Heavy
	  cleaning of the new localization module. Deleted old structures
	  and converted all
	  IO operations to the new input (direct print_key control).
	  Updated the regtests
	  with the new input. prettify.

2008-01-10 01:26  tlaino

	* [r6785] libint_wrapper also for XLF v10.1. Expanding the
	  documentation on how to use the
	  libint also for XLF v10.1. XLF v11 (still not available on our
	  IBM machine) should fully
	  support the ISO_C_BINDING.

2008-01-09 17:57  tlaino

	* [r6784] Here it is: this commit enables the fixing of only 1 or
	  more atomic components
	  instead of the fully atomic position. The default is always fix
	  the full atomic
	  position (you can tune your choice with the keyword
	  COMPONENTS_TO_FIX). Added
	  few regtests and reset few more due to numerics after
	  reorganization of few instructions.

2008-01-09 14:47  mguidon

	* [r6783] Prepare cp2k for long-range Becke88 and Hartree-Fock
	  calculations with mixed interaction potentials

2008-01-09 14:05  mguidon

	* [r6782] PGI misunderstood the latest commit

2008-01-08 11:33  tlaino

	* [r6781] colors for supported/unsupported keywords/sections in the
	  input reference (samuele giani)

2008-01-08 09:31  tlaino

	* [r6780] collecting common pieces of code

2008-01-07 19:57  vondele

	* [r6779] a sort for int_8

2008-01-07 16:02  mguidon

	* [r6777] Some NAG issues

2008-01-07 14:48  krack

	* [r6776] redundant code line removed

2008-01-07 14:43  tlaino

	* [r6775] Platform to get ready for a CP2K release. Once a method
	  (specified through sections/keywords)
	  is considered to be stable to be included in a CP2K release it's
	  enough to specify
	  for the keyword/section (supported_feature=.true., by default it
	  is false). If not, when
	  testing a release version of CP2K and one tries to access that
	  feature the program
	  will abort.
	  In the next days/weeks there will be few cleaning/setting for the
	  whole input sections
	  in such a way to be ready for the first release of CP2K.

2008-01-07 14:10  mguidon

	* [r6774] Prettify

2008-01-07 14:09  mguidon

	* [r6773] NAG issue with bit operators

2008-01-06 08:22  tlaino

	* [r6772] adapt to 2008 also templates

2008-01-05 15:55  tlaino

	* [r6771] bug fix NAG sdbg (juerg)

2008-01-03 10:13  vondele

	* [r6770] adapt to 2008

2007-12-24 10:01  jgh

	* [r6768] Safeguard bug: parallel runs with equal reference and
	  comp. grids;
	  This may appear for MT Poisson solver with small cutoff
	  enhancement;
	  Changed 5 regtest to avoid triggering bug

2007-12-23 12:54  tlaino

	* [r6767] bug fix from last commit on refactoring nr.6 : valgrind
	  catch uninitialized integer.

2007-12-22 19:38  tlaino

	* [r6766] making prettifiable. Prettifying..

2007-12-22 18:58  tlaino

	* [r6765] manual cleaning

2007-12-22 18:45  tlaino

	* [r6764] Synchronizing templates and source files

2007-12-22 18:16  tlaino

	* [r6762] Full source cleaning: (1) Moved to the attic the files
	  not compiled or compiled
	  and not used. (2) Prettified all prettifiable files.

2007-12-22 17:23  marcella

	* [r6761] bug fix: wrong index order in coping matrices

2007-12-22 15:25  jgh

	* [r6760] Fix bug from last commit, reset two more tests

2007-12-22 14:22  jgh

	* [r6759] Plane wave grid refactoring: 6th round;
	  pw_grid_setup is only remaining setup routine;
	  redefining pw_grid for MT Poisson solver leads to small numerical
	  changes;
	  34 regtests reset

2007-12-21 20:39  marcella

	* [r6755] some cleaning in the localization input

2007-12-21 14:36  tlaino

	* [r6754] now that OUTER_SCF has a section parameter let's use it..
	  no need to check
	  anymore if the section is present..

2007-12-21 13:46  krack

	* [r6753] section parameter added to OUTER_SCF section; consistent
	  subroutine names

2007-12-21 12:18  tlaino

	* [r6752] slim the logic.. cp_print_key_unit_nr should always be a
	  global call (shared
	  by all processors.. nothing guarantees that the node that will
	  write the print_key
	  is also the default ionode..).. bug fix..

2007-12-21 10:44  krack

	* [r6751] missing ionode check added

2007-12-21 08:59  tlaino

	* [r6750] moving the keyword PRINTOUT_IF_SHOULDSTOP in the print
	  section of the MD and
	  renamed FORCE_LAST. Cleaning md_env (no need to have another
	  logical there) and
	  making the force_last option slightly more lean.

2007-12-21 07:04  marcella

	* [r6747] write output if shouldstop only if explicitly required

2007-12-20 19:34  vondele

	* [r6746] try to recover gracefully from warnings (failing asserts
	  at cp_warning_level) without yielding wrong results or core
	  dumps. Basically, we (try to) not to return failure==.TRUE. if
	  the error level is <=CP_WARNING_LEVEL.

2007-12-20 19:23  tlaino

	* [r6745] bug fix for memory leaks

2007-12-20 10:57  tlaino

	* [r6743] typo in a comment

2007-12-20 10:48  tlaino

	* [r6742] get rid of hardcoded values and really use the input
	  parameters..

2007-12-20 08:11  tlaino

	* [r6739] of course stop in case of external requests/walltime..

2007-12-20 07:58  tlaino

	* [r6738] The behavior of the MD module should be uniform with all
	  the methods, if a
	  job has reached the end of the walltime or there's an external
	  request for
	  stopping during an MD just exit and do nothing..

2007-12-20 06:54  marcella

	* [r6736] cleaned localization for core level spectroscpy

2007-12-19 15:54  vondele

	* [r6735] code around a bug in the pw_pools, where allocating too
	  many grids from the pool apparently leads to incorrect
	  (uninitialized) energies being returned from the poisson solver

2007-12-19 12:29  tlaino

	* [r6734] Bug fix for element P (AM1 method) (thanks to Minghsun Ho
	  and Axel for the bug report). added
	  1 more regtest.

2007-12-18 17:34  tlaino

	* [r6733] Bug fix (thanks to Rachel Glaves for addressing the
	  problem). one more regtest.

2007-12-18 15:45  tlaino

	* [r6732] bug fix for TEST_TYPES (typo error for recent normal mode
	  checking regtest). Using
	  a parameter for flagging the length of description, so in case
	  we'll need to change it in the
	  future it will be much easier.

2007-12-18 13:55  tlaino

	* [r6731] bug fix: deallocate after the very last usage. added
	  intensities to the std_output

2007-12-18 13:19  fschiff

	* [r6728] Added intensities to the normal mode analysis
	  Added cp_result_type, which can be used to pass small amount of
	  data of arbitrary shape
	  to high level types (here dipole => rep_env)

2007-12-18 09:16  vondele

	* [r6727] added blocking to speed up pw_hfx

2007-12-17 21:51  tlaino

	* [r6726] Framework and main modules for fixed/induced DIPOLES and
	  QUADRUPOLES in classical FF.
	  Requires still debugging the energy/forces/fields/stress.

2007-12-17 20:42  vondele

	* [r6725] compute HF exchange using a realspace/gspace approach if
	  requested

2007-12-17 08:32  tlaino

	* [r6724] cosmetics

2007-12-17 08:30  tlaino

	* [r6723] extending the % convention also to section_type

2007-12-16 17:23  tlaino

	* [r6722] added a further check for the real-space term of Ewald
	  (EWALD|PME|SPME). If the
	  cutoff of the real-space part is larger than the cutoff of the
	  nonbonded potential
	  let's use this one as cutoff for NL. Added keyword:
	  EWALD_ACCURACY that affects
	  only the calculation of the real-space cutoff. no other effects
	  on the Ewald
	  parameters (in future this can be extended for computing optimal
	  Ewald parameters).

2007-12-16 12:58  tlaino

	* [r6721] cleaning ewald_env (removed useless stuff)

2007-12-14 17:34  tlaino

	* [r6719] bug fix in parallel

2007-12-14 16:07  tlaino

	* [r6718] moving STRESS_TENSOR from MD section to FORCE_EVAL
	  section: is the most adapt place
	  for specifying how to compute the stress tensor.

2007-12-14 15:17  tlaino

	* [r6717] Skip the calculation of the stress tensor in FIST if not
	  requested.

2007-12-14 11:47  tlaino

	* [r6716] Now also FIST behaves as all other methods in CP2K. To
	  access the stress tensor
	  you need to specify the keyword "STRESS_TENSOR" in the MD
	  section. PV_AVA and PV_NUM and
	  others have been deleted. STRESS_TENSOR takes care of them. This
	  is the biggest user
	  change regarding keywords controlling stress tensor.
	  Next commit will make FIST possible skipping the computation of
	  the stress tensor in order
	  to save some cpu time if stress tensor is not needed.

2007-12-13 21:29  vweber

	* [r6714] sweat one, Reinout Declerck.

2007-12-13 10:15  vweber

	* [r6713] fixed few bugs and activated epr-regs

2007-12-12 17:28  fawzi

	* [r6712] efficient handling of dense polynomial of degree up to l

2007-12-12 15:01  tlaino

	* [r6711] Few more strings (and comparisons/assignments/operations)
	  out of here..

2007-12-12 12:36  tlaino

	* [r6710] getting rid of few more archaic comparisons between
	  strings. cleaning.

2007-12-11 21:34  tlaino

	* [r6709] generalizing FIST dipole for non-orthorhombic cells.
	  introduced the possibility
	  to have also the non-periodic dipole. and a bug fix on the
	  original version of the
	  berry phase dipole. cosmetics around.

2007-12-11 18:39  tlaino

	* [r6708] Modifying the CHARMM FF (input driven) in such a way to
	  reflect exactly the
	  CHARMM convention. This means that force constants of bonds,
	  bends, UB and
	  impropers do not need to be multiplied by 2 anylonger. The
	  documentation has been
	  corrected as well in order to reflect how it is handled in the
	  code. This should
	  avoid possible bugs when people cut and paste CHARMM parameters
	  directly in
	  the input. (the problem was rised by Toon Verstraelen). Input in
	  regtests modify
	  properly. regtest ok.

2007-12-10 21:46  tlaino

	* [r6707] Eval_Interp_Spl3_pbc and Eval_d_Interp_Spl3_pbc: patch
	  for distributed grids (by Reinout Declerck)

2007-12-09 10:06  tlaino

	* [r6705] bug fix

2007-12-07 12:25  tlaino

	* [r6703] making more stable the spline generation (reset 24
	  regtests) few more FIST regtests OK
	  in single precision.

2007-12-07 11:16  tlaino

	* [r6702] fix format

2007-12-07 06:28  tlaino

	* [r6701] fix most of FIST regtests in single precision (at the
	  moment 530 regtests OK, 244 FAILED).

2007-12-06 23:07  tlaino

	* [r6700] fix hardcoded double precision

2007-12-06 22:14  tlaino

	* [r6699] cleaning read_atomic_kind: BASIS and POTENTIAL are
	  allocated only if strictly
	  necessary, moreover in atomic_kind_check we perform only check:
	  it's too late
	  for assignment.. general cosmetics.

2007-12-06 10:39  tlaino

	* [r6697] again single precision compilation

2007-12-06 00:07  tlaino

	* [r6696] using calcpar for AM1, PM3 and MDNO parameters (this
	  already showed a bug
	  in just 1 parameter of Si). reset most of semi-empirical regtests
	  (10^-6 rel err).
	  Added 1 regtest for PDDG.

2007-12-05 19:40  tlaino

	* [r6695] There's no need to keep this blas function here..

2007-12-05 19:14  tlaino

	* [r6694] NEW: Recife Model 1 (RM1). Cleaning further the
	  semiempirical module.

2007-12-05 16:41  mguidon

	* [r6693] Enable -D__UGLY_BUT_FAST for pgi on x86-64 linux machines

2007-12-05 15:54  mguidon

	* [r6692] removed INTENT(inout) statement

2007-12-05 15:24  mguidon

	* [r6690] Added hack for PGI on linux 64 bit machines

2007-12-05 13:33  vweber

	* [r6689] Reinout Declerck: epr update.

2007-12-05 12:36  vondele

	* [r6687] start to adopt f77_blas where this is easy (round 1 of
	  n).

2007-12-05 12:08  marcella

	* [r6684] fixed broken xas, better handling of printing in xas,
	  some more options for the calculation of msd.

2007-12-05 11:50  vondele

	* [r6683] start making use of f77_blas where it is trivial

2007-12-04 18:48  tlaino

	* [r6681] once more x single precision (Axel)

2007-12-04 17:45  tlaino

	* [r6680] few more fixes (by Axel) for single precision
	  compilation.

2007-12-04 16:30  tlaino

	* [r6679] including cp_common_uses into all other possible files
	  (kept out only the ones
	  involving the dependences of cp_log_handling and
	  cp_error_handling)

2007-12-04 14:39  tlaino

	* [r6678] __sp_lib becomes __SGL and completing the list of blas
	  calls in f77_blas_poison

2007-12-04 14:28  mguidon

	* [r6677] Cleanup

2007-12-04 13:37  mguidon

	* [r6676] Based load balance on cpu-timings

2007-12-04 09:51  vweber

	* [r6674] added new set of gauges.

2007-12-04 09:34  tlaino

	* [r6673] getting rid of DBLE and converting few integers in proper
	  ._dp

2007-12-04 00:44  tlaino

	* [r6672] smallest representable floating number should depend on
	  the precision (single or double)

2007-12-04 00:10  tlaino

	* [r6671] recovery for single precision compilation

2007-12-03 22:58  tlaino

	* [r6670] Axel's patch for allowing compilation in single precision

2007-12-03 19:06  tlaino

	* [r6668] getting rid of all hardcoded double precision (*D*, *d*)
	  in the definition of numbers
	  (thanks to Axel for pointing the problem). At the same time
	  getting rif of few tab characters..

2007-12-03 17:29  tlaino

	* [r6667] getting rid of the real*8

2007-12-03 11:30  vondele

	* [r6664] work around a pgi bug

2007-12-03 07:17  vondele

	* [r6663] remove debug code

2007-12-03 01:47  tlaino

	* [r6662] bug fix OMP (the code was broken few commits ago)

2007-12-02 21:21  vondele

	* [r6661] fix incorrect usage of mp_stop / mp_abort.

2007-12-02 17:04  vondele

	* [r6660] remove get_error_unit

2007-12-02 14:12  vondele

	* [r6659] remove parallel_include.F

2007-12-02 06:43  vondele

	* [r6658] for the time being, restrict to more simple geometries

2007-12-01 17:48  tlaino

	* [r6657] citing the potential references and adding 1 regtest
	  validating the qmmm module
	  using manybody potential (tersoff)

2007-12-01 17:47  jgh

	* [r6656] Refactoring PW grid code: 5th round
	  Renamed function pw_axpy - replaces pw_sumup, pw_add,
	  pw_subtract, pw_reduce

2007-12-01 16:43  tlaino

	* [r6655] Reorganizing/splitting the manybody potential module. In
	  the manybody_potential
	  we should keep only the main driver for computing energy and
	  forces. Each manybody
	  potential should have an his own module.

2007-12-01 13:42  tlaino

	* [r6654] added references

2007-12-01 12:06  tlaino

	* [r6653] splitting input_cp2k_motion module

2007-12-01 10:36  tlaino

	* [r6652] SEGNAME as alias to MOLNAME and WIDTH as alias to SCALE
	  (added more info about the
	  nature of the history dependent term in metadynamics)

2007-11-30 20:13  vondele

	* [r6651] further bug fixing for differences in rank between pw and
	  rs communicators

2007-11-30 18:05  tlaino

	* [r6650] bug fix (see [CP2K:426] for more details)

2007-11-30 17:53  vondele

	* [r6649] same arguments for MAX

2007-11-30 17:42  tlaino

	* [r6648] Never convert potatoes into cabbages. More tight control
	  over cp_units. bug fix
	  in cp_unit_release.

2007-11-30 16:14  vondele

	* [r6647] fix a bug in the case tasks are reordered between
	  communicators, and add some additional checks.

2007-11-30 14:47  tlaino

	* [r6646] cp_units initialization according the label of the units
	  (not hardcoded to integers). added
	  ps (picoseconds) to the time units. added unit in SCALE
	  (metadynamics)

2007-11-30 12:02  tlaino

	* [r6645] check for out of bounds index (molecule index provided in
	  input)

2007-11-30 11:27  tlaino

	* [r6644] mapping library for GROMOS topology files

2007-11-30 10:36  tlaino

	* [r6643] ALWAYS give the element name to atoms, indipendently of
	  gromos/charmm/amber input files.
	  getting rid of a write_gromos topology routine that was just a
	  dump_psf (no duplicated code)

2007-11-28 18:01  jgh

	* [r6642] Refactoring PW grid code: 4th round
	  Remove coeff_type and all its special routines
	  Use pw_p_type instead

2007-11-28 14:19  fschiff

	* [r6640] added the possibility to track all modes (the program
	  finds) in a specified range
	  added additional output

2007-11-26 16:25  mguidon

	* [r6637] Some cache miss optimizations

2007-11-26 12:43  tlaino

	* [r6636] Using allocatable for dummies arguments is an extension
	  (maybe 200N?) .
	  One of the arguments was passed without the ALLOCATABLE specs and
	  this was giving
	  an error of a non allocated array. Converting the "global" arrays
	  to pointed and
	  keping only the local one as allocatable.
	  
	  Everytime a unit is requested, once the IO is finished the unit
	  must be given back.
	  
	  Molden output cannot be embedded in an if like (IF (ionode)..)
	  this does not guarantee
	  that on a specific machine the output will really be there. Once
	  you've asked for the IO
	  unit you know who will print the info just checking if the unit
	  is > 0 on a local processor.
	  
	  TEST_TYPE 8 does not work because the output is different.
	  Changed the output and added also
	  another TEST_TYPE for this specific run.
	  
	  Strange behavior observed in parallel.

2007-11-26 10:19  fschiff

	* [r6633] Implemented a Divdson algorithm for mode selective
	  vibrational analysis.
	  In addition vibrational analysis without frozen atoms writes now
	  an output file in
	  molden format

2007-11-26 10:09  jgh

	* [r6632] Refactoring PW grid code: 3rd round;
	  Delete linear interpolation and 2 point finite difference
	  derivatives;
	  Change (and reset) one regtest

2007-11-24 16:09  vondele

	* [r6631] added some bin timing code

2007-11-24 11:35  tlaino

	* [r6630] cosmetics

2007-11-24 11:03  vondele

	* [r6629] slightly improve m_walltime

2007-11-23 17:28  jgh

	* [r6628] NDDO short range integrals fo EWALD;
	  some new parameter sets added

2007-11-23 15:20  mguidon

	* [r6627] Replaced some mp_sum by mp_send/receive

2007-11-23 14:39  tlaino

	* [r6626] prettify

2007-11-23 14:31  tlaino

	* [r6625] cosmetics

2007-11-23 09:02  vweber

	* [r6624] fix problem with small exponent, add a reg and reset few
	  regs.

2007-11-22 17:58  tlaino

	* [r6623] merging spme for DFTB and MM. optional arguments in
	  get_patch are non-optional
	  anymore (let's save some cpu time avoiding those checks)

2007-11-22 17:46  jgh

	* [r6622] Prepare for periodic NDDO code: Input and program flow

2007-11-22 15:35  tlaino

	* [r6621] remove files not compiled anymore..

2007-11-22 14:40  jgh

	* [r6620] Refactoring PW grid: 2nd round

2007-11-22 12:11  vweber

	* [r6619] rm duplicated code and fix bug with atom without soft
	  function.

2007-11-22 11:16  tlaino

	* [r6618] bug fix nonbonded forces for multiple potentials (added 1
	  regtest)

2007-11-22 09:12  tlaino

	* [r6617] cleaning

2007-11-21 17:03  mguidon

	* [r6616] Make NAG happy

2007-11-21 15:35  mguidon

	* [r6615] Removed all calls to gto_get_basis

2007-11-21 14:39  vweber

	* [r6614] a memory leak... thanks Teo.

2007-11-21 12:48  vweber

	* [r6611] fixed parallel nb-list and summation of the distributed
	  sigma_local.
	  added max_iter and eps_localization to the localization modules.

2007-11-21 11:32  tlaino

	* [r6610] Bug fix by G. Bussi for CSVR thermostat. reset 3
	  regtests.

2007-11-21 11:20  tlaino

	* [r6609] Bug fix FIST neighbot lists when interaction radius is
	  larger than cell size. Reset 62 regtests due
	  to WRONG RESULTS. Fist neighbor lists were rewritten on the basis
	  of the original ones (similar to the
	  QS ones), with identical numerical results between the old and
	  the new version.
	  Keep your eyes opened for QS calculations. I suspect that the
	  same kind of bug is present also in QS NL.
	  The bug was addressed by Toon.

2007-11-21 07:15  tlaino

	* [r6608] fist neighbor lists speed-up: use correct indexes for
	  subcell and cell. resetting 1 regtest (numerics)

2007-11-20 23:34  vondele

	* [r6607] A first result of the new domain decomposition:
	  use correct border size for distributed grids for SPME/PME,
	  make distributed grids the default. SPME up to 2 times faster in
	  parallel.

2007-11-20 21:21  vondele

	* [r6606] Introduce a new (fully 3D) domain decomposition for the
	  realspace grids,
	  which improves memory usage and/or scalability of classical and
	  DFT
	  parts of the code (Matt Watkins & Joost)

2007-11-20 13:46  mguidon

	* [r6605] Minor optimization of far-field screening

2007-11-20 12:00  mguidon

	* [r6604] Replacing mp_sum with mp_send/receive for communication
	  of density/KS matrix.

2007-11-19 16:44  tlaino

	* [r6603] Towards a correct poisson solver for MM/DFTB
	  calculations. Regtests are not failing
	  anymore but results are still wrong. next commit should fix the
	  proper behavior of
	  poisson solver.

2007-11-19 11:41  jgh

	* [r6602] Refactoring PW grid: 1st round

2007-11-18 19:37  tlaino

	* [r6601] doc typo (Toon)

2007-11-18 19:37  tlaino

	* [r6600] bug fix

2007-11-18 09:59  tlaino

	* [r6599] bug fix

2007-11-18 09:57  tlaino

	* [r6598] bug fix (valgrind)

2007-11-17 15:40  tlaino

	* [r6597] Thermostats: general thermostatting regions. for more
	  info see message [CP2K:398] on the CP2K google forum
	  (http://groups.google.com/group/cp2k). Added several more
	  regtests covering the new functionalities.

2007-11-16 14:43  mguidon

	* [r6595] Make PGI happy

2007-11-16 14:20  mguidon

	* [r6593] Improved load balance of load balance for large number of
	  cpus

2007-11-16 12:09  tlaino

	* [r6592] getting rid of useless/redundat code

2007-11-15 18:10  mguidon

	* [r6591] Allow for smaller compression cache sizes. Saves some
	  memory

2007-11-15 15:45  tlaino

	* [r6590] bug fix for au_e unit pointed by Toon.

2007-11-15 14:20  tlaino

	* [r6588] cosmetics

2007-11-15 14:11  tlaino

	* [r6587] fix typo (Toon)

2007-11-15 13:19  vweber

	* [r6586] nmr update, added local chi

2007-11-15 12:25  vweber

	* [r6585] out of bound for atom without soft function.

2007-11-15 11:25  mguidon

	* [r6583] Added some hacks to use LIBINT with other compilers than
	  G95. See /cp2k/libint_tools for details

2007-11-14 15:55  jgh

	* [r6580] Improved electrical moments, 2 new regtests

2007-11-14 13:19  vondele

	* [r6578] make the calculation of max radius more tight, and more
	  correct in the case of non-orthorhombic grids.
	  If this triggers an 'called with radius too large' error message,
	  send me a testcase. Passes regtest (with a few resets due to
	  numerics).

2007-11-12 10:37  tlaino

	* [r6555] parallel bug fixes

2007-11-12 09:40  vondele

	* [r6553] improve description for the wavelet solver, change
	  default scf

2007-11-12 09:21  vondele

	* [r6551] fix a few uninitialised values in parallel

2007-11-12 00:21  tlaino

	* [r6550] (1) updating for all geometry optimizer the dump of
	  information when converged.
	  (2) BFGS: broadcast is NOT necessary. gradients are correct on
	  all processors and not only on master!
	  (3) common function to print trajectory (enabled all possible
	  trajectory formats, before it was available only xyz!),
	  the printing of the trajectory now is exactly the same as within
	  the MD module. reset regtest due to the new introduced
	  energy calculation at the converged geometry. added 1 with FORMAT
	  DCD for trajectory.
	  Cleaning the geo modules.

2007-11-11 12:56  tlaino

	* [r6549] delete output_utilities and merged the relevant
	  subroutines in cp_log_handling.. deleted the
	  print_warning function (there's already one doing the same job,
	  one more duplication out)..

2007-11-10 19:47  tlaino

	* [r6548] Getting metadyn forces out of
	  force_env_calc_energy_forces (now metadyn belongs completely to
	  the MD module)..
	  Rewriting the WALLS in metadynamics (structure and the way they
	  influence the metadynamics). adding QUADRATIC and GAUSSIAN walls.
	  bug fix REFLECTIVE walls for non-lagrangian metadynamics.
	  resetteg metadynamics regtests (due to the right counter of the
	  metadyn steps
	  compared to the MD steps, before it was implicitely wrong).
	  adding 5 new one testing walls.

2007-11-10 08:12  vondele

	* [r6547] further ref

2007-11-09 20:58  vondele

	* [r6546] ewald, pme and spme references

2007-11-09 17:16  tlaino

	* [r6544] minor cleanup (Toon)

2007-11-09 14:23  tlaino

	* [r6543] possibly zeroing the angular velocity when random
	  velocity are applied and no
	  periodic boundary conditions present (Toon). Added 6 regtests.

2007-11-08 20:39  vondele

	* [r6542] update a DOI

2007-11-08 16:03  vondele

	* [r6539] add more references for some of the common density
	  functionals

2007-11-08 15:29  tlaino

	* [r6538] heavy cleaning/rewriting of the md and md_util module
	  (Toon Verstraelen and Teo). reset due to numerics of 3 regtests.

2007-11-08 09:03  vondele

	* [r6536] add citations

2007-11-07 19:36  tlaino

	* [r6535] bug fix for HILLS in periodic defined colvar

2007-11-07 18:26  tlaino

	* [r6534] redifining torsion colvar for metadynamics and other free
	  energy calculations as
	  a function with range in (-pi,pi]

2007-11-07 14:01  vondele

	* [r6533] bug fix for perdew zunger local correlation

2007-11-07 13:37  fschiff

	* [r6532] bug fix for parallel runs using a large number of Cpu's

2007-11-07 09:46  tlaino

	* [r6531] Bug: the poisson solver NONPERIODIC for the MM/DFTB
	  contains a bug. Just added a stop
	  when one tries to use it in MM/DFTB calculations. 4 regtest
	  failing (2 old + 2 newly added that
	  can be useful to debug, since the energy of an isolated charge +1
	  must be zero). the fixing
	  will have a low priority.

2007-11-06 15:25  mguidon

	* [r6530] Added possibility to scale compression threshold

2007-11-06 09:01  vondele

	* [r6528] prettify files that might be touched with the forthcoming
	  realspace grid update

2007-11-05 07:39  tlaino

	* [r6527] fix memory leak

2007-11-04 18:28  tlaino

	* [r6526] bug fix

2007-11-02 17:31  mguidon

	* [r6525] Bug fix periodic far field screening

2007-11-02 15:48  tlaino

	* [r6524] centering

2007-11-02 12:25  vondele

	* [r6522] cube output also for non-orthorhombic cells and
	  distributed grids (Matt Watkins & Joost)

2007-11-01 12:44  vondele

	* [r6520] bug fix distributed non-orthorhombic grids (Matt Watkins
	  & Joost)

2007-10-31 17:47  tlaino

	* [r6518] Enabling PIMD with all possible kind of energy/forces
	  evaluations.. added 1 regtest. resetted other due to numerics
	  after the last modifications to md_ener

2007-10-31 13:20  tlaino

	* [r6516] cleaning

2007-10-31 11:35  marcella

	* [r6514] zero md_ener, bug fix in comvel_control, cleaning,
	  extension of reftraj going on
	  CVS
	  ----------------------------------------------------------------------

2007-10-31 01:32  tlaino

	* [r6512] prettifying, restoring the logic of the qmmm variable in
	  md_energies/md_conserved_qnty,
	  deleting/moving comments.

2007-10-30 17:15  marcella

	* [r6511] temperature_control reads the temperature stored in
	  md_ener

2007-10-30 11:30  marcella

	* [r6509] The calculation of constant, kinetic energy, and
	  temperature has been moved to a separated routine that is now
	  independent from the routine writing out the md output. Values of
	  energies and temperatures are stored in md_ener, in order to be
	  availabe for later use.

2007-10-29 15:56  tlaino

	* [r6507] STOP DUPLICATNG CODE!!! NOW!!!

2007-10-29 15:25  tlaino

	* [r6506] Bug fixing in Rome: the wild use of retain/release.. bug
	  fix memory leaks.. reftraj type was
	  never deallocated. prettifying

2007-10-29 10:02  krack

	* [r6504] enable handling of long basis set lines

2007-10-29 09:57  krack

	* [r6503] BFGS: print final requested structure data

2007-10-29 06:53  marcella

	* [r6499] Change in reftraj set up to extend to more options for
	  the analysis of trajectories

2007-10-27 18:18  vondele

	* [r6496] improve the format of the citations in the input manual

2007-10-26 09:29  vondele

	* [r6492] fix bug with article numbers,
	  add molopt reference,
	  provide link to all references from the main input page

2007-10-24 19:46  vondele

	* [r6482] work around one more problem on XT3 (pgf90 defines __PGI)

2007-10-24 18:16  vondele

	* [r6479] New Makefile for CP2K. Finally we will be able to remove
	  forpar.
	  
	  Requires an update of the makefiles and tools subdirectories of
	  CP2K.
	  Forpar contained several bugs and had become unmaintainable,
	  dependencies
	  have been incorrect (causing e.g. miscompilations that needed
	  realcleans).
	  This is now solved by using makedepf90 to correctly generate
	  dependencies.
	  In order to avoid recompilation cascades the makefile has now
	  been changed
	  so that dependencies are through *.mod files. If a compiler
	  doesn't touch
	  a mod file that is unchanged (e.g. g95) compilation cascades are
	  avoided.
	  The corrected dependencies, however, still imply that more files
	  need recompilation if a module changes (but a clean or realclean
	  should never be needed).
	  Several more cleanups have been made in the makefile. In
	  particular, it is now possible to avoid
	  completely cpp steps if the Fortran compiler can deal with cpp
	  statements.
	  By putting the correct flags for the Fortran compiler
	  (e.g. '-free -fpp' or '-ffree-form -cpp') and defining 'CPP = '
	  in the arch
	  file a build can be speed up significantly (now the fastest CP2K
	  build is 30sec.
	  vs 75sec. before the update (on opt3)).
	  
	  All unneeded files in the cp2k/tools directory (including forpar)
	  will be removed soon.

2007-10-24 15:29  mguidon

	* [r6477] A few optimizations

2007-10-24 10:23  mguidon

	* [r6474] Added compression algorithm for electron repulsion
	  integrals

2007-10-23 13:00  vondele

	* [r6472] another cpp-ishly correct piece of code ?

2007-10-23 08:50  vondele

	* [r6471] fix cpp-ism

2007-10-23 06:35  vondele

	* [r6470] remove now meaningless comments

2007-10-23 00:04  tlaino

	* [r6469] modifying the interface to the MKL FFT library so that it
	  works with MPI (Ben Levine)

2007-10-23 00:03  tlaino

	* [r6468] The CMPLX function is not permitted in an initialization
	  expression (part of F2063. In
	  2063 Zefram Cochrane will make the Earth's first warp flight..
	  waiting for this event
	  let's just avoid the CMPLX function ;-) )

2007-10-22 14:03  vondele

	* [r6467] make math constants really constants, introduce the
	  inverse factorial, use in gamma

2007-10-22 13:47  vondele

	* [r6466] cite wavelet references in code path

2007-10-20 12:08  vondele

	* [r6465] add syncs to clarify timings

2007-10-19 21:17  vondele

	* [r6464] speedup a surprisingly hot loop

2007-10-19 20:41  vondele

	* [r6463] speedup HFX forces (3.3x)

2007-10-19 18:55  vondele

	* [r6462] fix typo / add implicit none

2007-10-18 19:13  tlaino

	* [r6461] removing just debug statements..

2007-10-18 18:28  tlaino

	* [r6460] restoring writing..

2007-10-18 18:11  vondele

	* [r6459] remove a 'CPPostconditionNoFail' which caused failure
	  (?!).
	  Having a non-commensurate number of CPUs now works fine.

2007-10-18 16:42  vondele

	* [r6458] some output for a replica to mpi process correspondence.

2007-10-18 16:30  tlaino

	* [r6457] getting rid of few useless preprocessing instructions
	  from source files..
	  separating output from qs_core_hamiltonian and just indenting
	  properly..
	  cosmetics around..

2007-10-18 14:25  vondele

	* [r6456] avoid a rep env write

2007-10-18 13:56  vondele

	* [r6455] bug fix in the convergence criteria for geometry
	  optimisation

2007-10-18 13:28  vondele

	* [r6454] fix lbfgs to check convergence from the default criteria
	  first, add a reference for lbfgs.

2007-10-18 09:06  vondele

	* [r6453] remove debug statement

2007-10-18 07:49  vondele

	* [r6452] change GEO_OPT section layout, putting method specific
	  options in their own section.

2007-10-17 22:27  tlaino

	* [r6451] article information printed through the bibliography
	  reference.. no need to have
	  duplicated info in the output

2007-10-17 22:26  tlaino

	* [r6450] generalizing output to non-symmetric sparse matrix

2007-10-17 16:44  fawzi

	* [r6448] added cp_str_cache

2007-10-17 16:42  fawzi

	* [r6447] preliminary version of the string cache (to improve input
	  reading and storage of multilines of characters).

2007-10-17 16:40  fawzi

	* [r6446] added function to get eof iostat code.
	  prettify

2007-10-17 15:36  vondele

	* [r6445] FFTW3 based runs now automatically import wisdom from
	  /etc/fftw/wisdom. This can speedup FFTs significantly (best seen
	  so far is 3.5x faster for a 128x128x128 grid), though we only
	  really benefit for serial runs due to the large overheads in the
	  parallel case. Added FFTW_WISDOM_FILE_NAME to provide a
	  non-default wisdom file, which can be generated with the tool
	  fftw-wisdom (part of an fftw installation) and the script
	  cp2k/tools/cp2k-wisdom.

2007-10-16 19:03  tlaino

	* [r6443] bug fix PRINT_LEVEL DEBUG for GAPW

2007-10-16 16:16  tlaino

	* [r6442] BUG fix for GAPW + GPW_TYPE.. now it works.. added a
	  regtest. Cleaning the
	  modules: atomic_kind_types and soft_basis_set.

2007-10-16 10:33  fschiff

	* [r6441] bug fix for pgi, some more clean up

2007-10-16 08:55  fschiff

	* [r6440] fixed hardcoded double precision, got rid of the abinit
	  exchange related stuff

2007-10-15 23:30  tlaino

	* [r6439] bug fix intent of few variables and proper
	  initialization..

2007-10-15 18:39  tlaino

	* [r6438] bug fix for computing dipole with localization.. added
	  the value in debye

2007-10-15 15:44  fschiff

	* [r6437] added allowed values for SCF_TYPE

2007-10-15 12:57  vondele

	* [r6435] add references for the isokinetic ensemble

2007-10-15 10:42  vondele

	* [r6434] Plug memory leak

2007-10-15 10:07  vondele

	* [r6433] No intent for pointers (this is F2003)

2007-10-15 10:01  tlaino

	* [r6432] CSVR thermostat validated by G.Bussi: it works (global
	  CSVR thermostat from 2-3 times
	  faster than global Nose-Hoover, with other regions (MASSIVE or
	  MOLECULE) it may be slower
	  than Nose-Hoover)

2007-10-15 09:26  fschiff

	* [r6431] additional constants for wavelet solver

2007-10-15 09:25  fschiff

	* [r6430] added a new poisson solver for free and surface (XZ)
	  boundary conditions (needs about 3 Angstrom of vacuum at the
	  boundary
	  solver is based on wavelets (see refernces)

2007-10-12 23:56  tlaino

	* [r6429] OMP for full cp2k (Ben Levine). full regtest passes (with
	  few instability problems
	  for regtests UKS-NMR and few of DDAPC). together with a weekly
	  valgrind regtest
	  there will be also another regtest to check that OMP version will
	  be preserved
	  through developments..

2007-10-12 12:42  marcella

	* [r6428] bug fix

2007-10-12 09:10  tlaino

	* [r6427] Using the global random number generator in a real global
	  way.. this generator
	  should always be called from all processors.. if a local one is
	  needed, please,
	  create one locally and destroy it immediately after..

2007-10-12 07:51  tlaino

	* [r6426] adding references for available thermostats

2007-10-11 22:11  tlaino

	* [r6425] CSVR thermostat enabled and working.. added several
	  regtests covering all the
	  possibility of usage of the new thermostat (particles, KG coeffs,
	  shells, barostat)..
	  Finally: LAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAND!..

2007-10-11 15:21  tlaino

	* [r6424] More for CSVR thermostat: completed all the files
	  handling the MD of the thermostat,
	  the generation of the random numbers in parallel and several
	  other pieces of cosmetics around..
	  Now I can say that: LAAAAAAAAAAAAAAAAAAAAAAAAAAAAAND! ;-) still
	  something to debug though..

2007-10-11 13:44  marcella

	* [r6422] restart for the shell%NOSE%THERMOSTAT

2007-10-10 14:24  tlaino

	* [r6420] Merging thermostat for (particles and particle_set) and
	  (shells and shell_set)
	  useless duplicated code.. now we've just one thermostat to be
	  applied for
	  thermostating particles and shells.. no changes in regtest..

2007-10-09 17:20  tlaino

	* [r6419] Activating the new files for CSVR.. deleting redundant
	  code from the restart
	  part of the thermostat and bug fixing the restart for thermostat
	  for the new
	  CSVR type. cosmetics..

2007-10-09 11:40  tlaino

	* [r6418] Removing useless info from simpar.. no need to have
	  information stored about
	  thermostat in the simulation parameter type.. renaming
	  nhc_mapping into a more
	  consisten filename.. adding two more files for CSVR thermostat..
	  Due to numerics
	  1 reset..

2007-10-08 19:34  tlaino

	* [r6417] Grouping more structures common to different kind of
	  thermostats.. reset 1 regtest
	  due to numerics.. removing other useless pieces of code..

2007-10-07 18:15  tlaino

	* [r6416] cosmetics.. merging two files + renaming. get rid of
	  nose_shell logical.
	  Redundancy not needed anylonger..

2007-10-06 18:02  tlaino

	* [r6415] Duplicated code: RAAAAAAAAAAAAAAAAAAAAAAAAUS! I wonder
	  whether this story
	  (the one of duplicating code instead of reusing old one) will
	  never end..

2007-10-06 17:20  tlaino

	* [r6414] Getting rid of nose-hoover in integrator. Generalization
	  to thermostat. Getting
	  rid of interface for NH and getting rid of old never used code.
	  Ready to activate
	  the CSVR thermostat.

2007-10-05 18:42  tlaino

	* [r6413] bug fix save_mem with geo_opt

2007-10-05 18:16  tlaino

	* [r6412] Molecular or massive thermostats are not allowed with
	  intermolecular constraints. Bug Fixed 2 regtests.
	  Added a stop in such cases.

2007-10-05 14:26  tlaino

	* [r6411] Common initialization for particles and shell.. adding
	  one more regtest for
	  REGION molecule for shell and resetting one regtest from the last
	  commit.

2007-10-04 08:41  marcella

	* [r6408] monitoring of the temperature of each kind

2007-10-04 06:43  vondele

	* [r6407] fix some strange CR problem on AIX

2007-10-03 20:06  tlaino

	* [r6406] patch provided by G. Bussi for DOF equal 1

2007-10-03 14:51  tlaino

	* [r6405] Removing useless code.. 1 regtest failing in parallel
	  after the last commit regarding
	  the shell thermostat. in my opinion the thermostat over shells
	  (massive and
	  molecule) is wrong. they give different numbers in serial and in
	  parallel and this is
	  already an indication that something is wrong. the only correct
	  one is the global.

2007-10-03 12:40  marcella

	* [r6404] thermostat region molecule for shell-core motion

2007-10-03 09:43  krack

	* [r6403] * check alloc/dealloc status
	  * write final particle information, if converged

2007-10-02 18:44  tlaino

	* [r6402] cleaning the parallelization schemes of thermostats and
	  splitting the
	  mapping part common to all "possible" kind of thermostats..

2007-10-02 08:32  tlaino

	* [r6400] restructuring the thermostats (one for each kind of
	  system to thermostat).. several cleaning around..
	  still in the middle of the ocean.. no Land yet..

2007-10-01 11:45  mguidon

	* [r6399] Activate MC keywords

2007-10-01 07:19  tlaino

	* [r6398] patch by Axel

2007-09-30 12:26  tlaino

	* [r6397] fix uninitialized variable (valgrind regtest)

2007-09-30 12:24  tlaino

	* [r6396] less than 31 characters..

2007-09-28 17:23  tlaino

	* [r6395] Getting rid of the array of force_env in MD:
	  RAAAAAAAAAAAAAAAAAAAAAAAAAUS!

2007-09-28 15:33  mguidon

	* [r6393] Clean up. New input structure.

2007-09-28 15:32  mguidon

	* [r6392] Clean up. New input structure

2007-09-28 13:44  vondele

	* [r6390] add a master-slave mode for doing farming.
	  This improves load balance in the case where jobs of unequal
	  lenght are being processed.
	  Allow for dependencies between jobs in a farming run.

2007-09-28 13:39  vondele

	* [r6389] add additional flags for a version string,
	  a license printout, add CP2K to the references,
	  and adjust some CP2K| output.
	  This requires a Makefile update.

2007-09-28 13:02  tlaino

	* [r6388] Array of SUBSYS: RAAAAAAAAAAAAUS!..

2007-09-27 15:01  tlaino

	* [r6384] Splitting md_run and grouping similar pieces of code.
	  getting rid of different ways of
	  computing kinetic energy. 1 test resetted for parallel runs due
	  to numerics.

2007-09-27 12:34  mguidon

	* [r6383] Correction of wrong default value

2007-09-27 10:39  mguidon

	* [r6380] Removed old Monte Carlo distribution for HFX

2007-09-27 08:20  mguidon

	* [r6379] Missing CYCLE statement inserted

2007-09-26 16:39  mguidon

	* [r6377] Faster load balancing using Monte-Carlo/binning

2007-09-26 16:17  tlaino

	* [r6376] cleaning the MD driver: two level of calls and no
	  interfaces

2007-09-26 15:23  vondele

	* [r6375] rename GEOOPT section in GEO_OPT (consistent with
	  RUN_TYPE GEO_OPT)

2007-09-26 14:47  vondele

	* [r6374] reduce amount of output on error

2007-09-26 14:44  vondele

	* [r6373] and of course don't make it the default

2007-09-26 14:43  vondele

	* [r6372] add note about our linking to FFTACML being buggy

2007-09-26 14:02  vondele

	* [r6371] IMPORTANT: change in FFT library selection and setup,
	  these can affect total energies significantly if you have been
	  using an FFT library different from FFTSG. All FFT libraries now
	  use by default the same grids, meaning that total energies now
	  become 'independent' from the FFT library selected (and equal to
	  FFTSG results, but with the extra speed that comes from using
	  more optimised libraries). For simulations that need capabilities
	  beyond the default FFTSG grids or that need to reproduce old
	  results, use EXTENDED_FFT_LENGTHS. Furthermore, FFTW is now the
	  default for PREFERRED_FFT_LIBRARY. If PREFERRED_FFT_LIBRARY is
	  not available, FFTW is tried first, and FFTSG is used as a
	  fall-back. The distinction between FFTW2 and FFTW3 on the
	  input-file level is gone. No regtests require reset if FFTW was
	  not linked into the code (otherwise the number depends on having
	  either FFTW2 or FFTW3 linked in).

2007-09-26 12:45  tlaino

	* [r6370] still to early for CSVR thermostat ;-)

2007-09-26 12:28  tlaino

	* [r6368] Reorganizing thermostat and barostat - PART 1. For a
	  complete description see message:
	  "THERMOSTATs - PART 1" on cp2k google group.

2007-09-24 13:58  vondele

	* [r6366] fix a (serious) bug in the generation of the initial
	  velocities, i.e. use proper mass-weighting (provided by T.
	  Kuehne). Reset 276 regtests

2007-09-22 13:23  vondele

	* [r6360] make EXTRAPOLATION PS, EXTRAPOLATION_ORDER 3 the default
	  (101 resets)

2007-09-22 08:20  vondele

	* [r6359] make map_consistent the default, reset 302 regtests

2007-09-21 07:17  vondele

	* [r6358] try flushing the output unit in case of error, so that
	  error messages hopefully make it to the user

2007-09-21 06:08  vondele

	* [r6357] add two more spaces in the title line

2007-09-20 10:33  tlaino

	* [r6356] Bug fix for impropers in multi_psf and ading unit to
	  targets (HBONDS constraint)

2007-09-20 09:21  vondele

	* [r6355] only write the wfn restart every 20th step during the SCF
	  and at convergence

2007-09-20 09:04  vondele

	* [r6354] change in wfn restart: if wfn_restart_file_name is not
	  set, restart from the file generated by the wfn restart print key
	  (which can be the default one). All regtests, with modifications
	  to the input, passing.

2007-09-19 21:45  tlaino

	* [r6353] adding a check for possible user errors in input file in
	  specifying the configurations

2007-09-19 18:41  tlaino

	* [r6352] Making more general the bsse calculation. possibility to
	  specify charge and
	  multiplicity for the different combinatorial combinations giving
	  contribution
	  to the counterpoise correction. Added a regtest provided by Chris
	  Adriaanse
	  testing the new feature. Resetting two regtest from the last
	  commmit on
	  the FIST neighbor lists.

2007-09-19 11:07  tlaino

	* [r6351] New memory distribution for FIST neighbor lists. 2
	  regtests resetted due to
	  numerics (10^-14).

2007-09-19 10:02  mguidon

	* [r6350] Messed up keyword description

2007-09-19 09:53  mguidon

	* [r6349] Added optional correction to screening on initial density
	  matrix

2007-09-17 11:46  mguidon

	* [r6347] Cleaning

2007-09-16 12:43  tlaino

	* [r6346] further check when reading XYZ files..

2007-09-15 10:54  tlaino

	* [r6345] fixing and restoring a check for pint_types..

2007-09-15 10:06  tlaino

	* [r6344] [CP2K:240] Introducing a RESTART_DEFAULT that controls
	  the default value for all restartable
	  quantities.. The specification of a single RESTART_qnty overrides
	  the default value.. totally
	  clear for users..

2007-09-14 14:27  tlaino

	* [r6343] bug fix intermolecular constraints (zero MD step)..
	  cleaning the lagrange
	  multipliers.. minor fixes for IO info in SHAKE and RATTLE..

2007-09-14 12:51  mguidon

	* [r6342] Bug fix load balance. Screening on initial density
	  matrix.

2007-09-14 05:42  tlaino

	* [r6341] Guillaume and Axel: string too short. avoiding
	  segmentation faults on some
	  platforms. Resetting due to numerics (last change on
	  preconditioner) .

2007-09-13 20:45  vondele

	* [r6340] make the preconditioner more robust with respect to very
	  bad initial guesses

2007-09-12 12:22  vweber

	* [r6337] Reinout Declerck: fixed few bugs with gapw-current for
	  uks.

2007-09-11 14:23  mguidon

	* [r6335] Added far field screening for periodic boundary
	  conditions

2007-09-10 08:33  vondele

	* [r6333] add a couple of the dftb references to the bibliography

2007-09-09 23:55  tlaino

	* [r6332] if name fail in searching dftb parameter look for
	  element.. Motion section
	  applies only for the main force_eval.. normalized the structure
	  of the qmmm..
	  prettifying and cleaning..

2007-09-07 10:52  vondele

	* [r6331] remove link to old file

2007-09-06 14:36  vondele

	* [r6329] fix silent print level buglet in parallel

2007-09-05 19:45  vondele

	* [r6328] fix a case of uninit charge in dump_psf

2007-09-05 18:49  vondele

	* [r6327] 'fix' psf and pdb dump file names

2007-09-05 12:20  vondele

	* [r6326] adapt format for longer simulations

2007-09-05 06:04  ikuo

	* [r6325] Change to accept multiple lines of WEIGHTS in the POINTS
	  subsection.

2007-09-01 18:39  vondele

	* [r6324] some more added exit codes

2007-09-01 12:31  vondele

	* [r6323] generate non-zero exit code ($?) on error, gets rid of
	  memory leak messages after failed runs with g95.

2007-08-31 17:25  tlaino

	* [r6322] Extending EXCLUDE_QM and EXCLUDE_MM to all type of
	  constraints. Adding warning
	  when molname is not defined. Adding warning if
	  temperature_tolerance is ON .
	  Adding 8 more regtests for all constraints and both exclude
	  possibilities.

2007-08-30 23:49  tlaino

	* [r6321] Axel: typos bug fix.

2007-08-30 18:40  tlaino

	* [r6320] bug fix thermostat (nvt) with fixed atom constraints (now
	  it works both with global and molecule NHCOPT option): resetted 2
	  regtests added 2 new.

2007-08-30 16:45  tlaino

	* [r6319] fixing for other compilers and typo

2007-08-30 16:06  mguidon

	* [r6318] Added far field screening for 1/R and erfc(wR)/R in non
	  periodic case.

2007-08-29 11:12  vweber

	* [r6317] fixed few reference to disassociated pointers.

2007-08-28 13:38  vweber

	* [r6316] Reinout Declerck: epr update.

2007-08-28 13:13  tlaino

	* [r6315] fixing input references..

2007-08-28 09:14  vweber

	* [r6314] fixed performance problem that shows up with the nag
	  compiler.

2007-08-27 11:44  tlaino

	* [r6313] clarifying output

2007-08-26 21:00  tlaino

	* [r6312] cleaning unuseful things.. redundant..

2007-08-26 14:16  tlaino

	* [r6311] Philly's bug fix ;-).. contraint init bug fix for g3x3 nd
	  g4x6 (just a reminder: md_run needs
	  some cleaning in the core-shell part)

2007-08-23 19:42  vondele

	* [r6310] avoid system to move file

2007-08-23 18:16  vondele

	* [r6309] avoid system call for file move, use rename

2007-08-23 15:00  vondele

	* [r6308] fix lsd bug

2007-08-22 09:39  vondele

	* [r6307] warning only on source

2007-08-18 21:03  tlaino

	* [r6306] Bug fix for SE analytical derivatives when TAPER was ON.
	  Bug fix for SE-QMMM: never
	  switch on the TAPER function.. added 1 regtest

2007-08-18 14:34  tlaino

	* [r6305] Bug fix for RESTART keyword.. Bug fix for restart
	  metadynamics.. resetting few regtests of metadynamics..

2007-08-18 08:42  tlaino

	* [r6304] Syncronizing templates and instantiation files with
	  source files! Again:
	  the files generated through templates must NEVER be modified. The
	  corresponding
	  templates and instantiations are the ones to be changed..
	  Propagating error in few more routines..

2007-08-17 18:39  tlaino

	* [r6303] possibility to dump the translational vector for qmmm
	  runs (complete structure
	  with restart capabilities).. fixing format for link atoms
	  (Axel)..

2007-08-17 13:41  tlaino

	* [r6302] patch for printout of COLVAR (more space for integers)
	  (Axel)

2007-08-17 12:08  tlaino

	* [r6301] Bug fix in restart Metadynamics.. Bug fix in
	  restart_history (avoiding the proliferation
	  of numerous restart files for QS MDs).. reset 2 regtests added 3
	  more..

2007-08-17 08:09  vondele

	* [r6299] added a new SIC method, explicitly on a set of orbitals
	  (Chris Adriaanse & Joost).

2007-08-17 06:46  vondele

	* [r6298] add comment about an intel checking bug

2007-08-16 18:50  vondele

	* [r6296] fix typo

2007-08-16 18:25  vondele

	* [r6295] one more

2007-08-16 17:57  vondele

	* [r6294] make a few more filenames and file_names of
	  default_string_length

2007-08-16 17:18  tlaino

	* [r6293] QMMM walls..

2007-08-16 17:05  vondele

	* [r6292] use intrinsic posix module to move file. Needs updated
	  arch files...

2007-08-15 13:56  vondele

	* [r6290] bug fix

2007-08-15 11:33  vondele

	* [r6289] reference for MT psolver

2007-08-15 07:33  vondele

	* [r6288] adding Toukmaji_CPC_95_73 as a citation for the ewald
	  section. BTW, it doesn't seem to talk about SPME.

2007-08-15 05:08  vondele

	* [r6287] change the default ewald alpha to 0.35A^-1. Only a number
	  of DFTB regtested used the 'wrong' old default, these have been
	  reset.

2007-08-14 20:37  vondele

	* [r6285] allow for setting different number of points on the SPME
	  and EWALD grids (gmax)

2007-08-13 22:53  tlaino

	* [r6283] new syntax for mapping section of free energy methods..
	  fixing info for dumping
	  particles info for several methods.. added 1 regtest..

2007-08-13 14:52  vondele

	* [r6280] fix for parsing 1E-7 as a real number

2007-08-13 08:42  vondele

	* [r6279] added a reference manager and the seed of a bibliography

2007-08-11 14:41  tlaino

	* [r6278] fixing uninitialized variable and formatting a
	  preprocessing instruction..

2007-08-10 18:28  tlaino

	* [r6277] BASIS section and POTENTIAL section in KIND section. now
	  information on basis sets
	  and potentials can be directly put in the input file.. since now
	  they will always
	  be present in the .restart file. added 1 regtest. regtest OK.

2007-08-10 17:58  vondele

	* [r6276] extend s2a

2007-08-10 14:13  mguidon

	* [r6275] Optimized periodic code

2007-08-10 12:24  tlaino

	* [r6273] Adding warning for RESTART wfn when the file is missing
	  as usual continues with
	  GUESS ATOMIC but prints a warning on the screen.. a bug fix in
	  cp_assert.. removing
	  error in the several cp_assert warning.. we use it only for
	  warning not for
	  changing the error status..

2007-08-09 15:56  tlaino

	* [r6270] RESTART of the wavefunction becomes a print_key. fully
	  support available only for QS.
	  no regtest reseted. Added also a RESTART_HISTORY print_key in the
	  same spirit of the
	  restart print_keys of the motion section. A small problem only
	  present in pint_types.F
	  (see e-mail sent to the group) but I warned already Fawzi about
	  that.

2007-08-08 13:53  mguidon

	* [r6269] Special branch for ncpu=1

2007-08-07 22:33  tlaino

	* [r6268] uninitialized value in topology_types and a possible bug
	  fix in input_cp2k_restarts.. prettifying..

2007-08-07 18:57  ikuo

	* [r6267] Add support to read in velocity for G96 format. [mbaer]

2007-08-07 15:43  hforbert

	* [r6266] bug fix in normal mode pint md

2007-08-07 15:29  mguidon

	* [r6265] Bug fix for int_8 datatype.

2007-08-07 13:06  tlaino

	* [r6264] clarifying error message..

2007-08-07 11:46  vweber

	* [r6263] loop skipout extended to the 3 different DMs used.

2007-08-07 10:08  tlaino

	* [r6262] IO bug fix

2007-08-07 09:14  tlaino

	* [r6261] adding units to keywords in SPLINE section.. modified
	  properly regtests.. no resetting..

2007-08-07 08:14  tlaino

	* [r6260] Improved the tersoff algorithm (making it linear scaling
	  w.r.t. the definition of
	  cutoff parameter (useful for multiple_potentials)).
	  Retrieving the original pair_potential and deleting the newly
	  introduced
	  pair_potential type. Most of new potentials type can be obtained
	  with the GENPOT potential (even piecewise defined functions).
	  Only types that
	  cannot be created with GENPOT can be introduced in FIST.
	  Adding potentials increases the work load for maintaining that
	  part of the code.
	  The idea is to keep it light and flexible.

2007-08-06 22:28  marcella

	* [r6259] bug fix and new nonbond-potential type

2007-08-06 16:42  mguidon

	* [r6258] New parallelization scheme. Load balance parallelized.

2007-08-06 09:24  tlaino

	* [r6257] Units for RMIN and RMAX.. 1 reset

2007-08-06 08:22  tlaino

	* [r6256] FIST: nonbonded interactions can now be defined with
	  piecewise functions! added regtest..

2007-08-05 07:56  tlaino

	* [r6255] bug fix GROMOS: uninitialized variables..

2007-08-04 21:48  tlaino

	* [r6254] Bug fix for BMHFT potential.. splitting GAPW regtests for
	  valgrind.. resetting 1 regtest..

2007-08-04 20:34  tlaino

	* [r6253] removing pwarning using cp_assert instead.. splitting the
	  setup of the shell
	  potential from the pack of the charges (was making no sense!)..
	  cleaning..

2007-08-04 15:22  tlaino

	* [r6252] Multiple-potential: heavy rewriting of the
	  pair_potential.. rewriting of the manybody (now
	  pair contribution are treated as a pair_potential and not within
	  the manybody part)..
	  The Multiple-potential enables the specification of nonbonded
	  potential in an additive way
	  overlapping several different functional forms.. Regtest OK!..
	  added few more examples...
	  General cleaning especially for the manybody part and the
	  pair_potential terms..

2007-08-03 22:42  ikuo

	* [r6249] Update GROMOS topology and force field. Fix some memory
	  leaks in GROMOS routines. [mbaer]

2007-08-02 23:56  marcella

	* [r6247] bug fix

2007-08-02 20:56  marcella

	* [r6245] shell model: nose thermostat for core-shell motion

2007-08-02 13:56  mguidon

	* [r6244] Refactored permutation of buffers

2007-08-02 12:47  tlaino

	* [r6243] reorganizing spline for EAM.. now into EAM type..

2007-08-01 15:58  tlaino

	* [r6242] bug fix

2007-08-01 13:23  tlaino

	* [r6241] bug fix

2007-07-30 19:25  tlaino

	* [r6240] computing the derivative of the unbiased free energy in
	  the AC scheme.. some
	  cleaning and reorganization..

2007-07-27 17:54  tlaino

	* [r6239] restoring changes lost with a previous commit..

2007-07-27 11:42  tlaino

	* [r6238] bug fix.. resetting regtests..

2007-07-27 08:43  tlaino

	* [r6237] prettify

2007-07-27 08:26  tlaino

	* [r6236] cleaning and prettifying..

2007-07-27 07:05  vondele

	* [r6235] improve printout for keywords with units but no default
	  value.

2007-07-26 22:07  marcella

	* [r6233] thermostat for the core-shell motion

2007-07-26 19:21  ikuo

	* [r6232] Change CASE in dist_colvar to use defined constants.

2007-07-26 17:26  vondele

	* [r6231] improve printout of defaults

2007-07-26 17:14  ikuo

	* [r6229] Modify dist colvar to each component separately [M.Baer]

2007-07-25 16:15  tlaino

	* [r6228] property sections fixed (Francesco).. cleaning..

2007-07-24 15:02  tlaino

	* [r6227] bug fix

2007-07-24 14:16  tlaino

	* [r6226] bug fix format (Axel)

2007-07-24 07:04  vondele

	* [r6225] avoid tmp

2007-07-23 20:20  tlaino

	* [r6224] Introduced the KINDS_FROM and KINDS_TO in the
	  COORDINATION colvar section. Possibility
	  to define coordination using atomic KINDS. 1 more regtest added..

2007-07-23 15:46  tlaino

	* [r6223] adding a REFTRAJ section (possiblity to specify the
	  filename and whether you want
	  energy computed or not on the external trajectory (at the moment
	  supported only xyz
	  format)). Added an iteration level to the metadynamics and
	  corrected the ugly
	  way of outputting HILLS. Added two more regtests on REFTRAJ.

2007-07-22 13:47  tlaino

	* [r6222] bug fix

2007-07-21 14:27  jgh

	* [r6219] Post-SCF calculation of electric moments;
	  Arbitrary moments for non-periodic systems - only dipoles for
	  periodic;
	  Initial implementation - not fully tested

2007-07-20 19:13  tlaino

	* [r6218] opening the log file only when it's strictly needed..
	  added a GEO_CHECK keyword
	  for FIST neighbor_lists.

2007-07-20 17:33  ikuo

	* [r6217] Read in multiple ATOMS line in POINTS subsection

2007-07-20 15:16  tlaino

	* [r6216] cleaning pair_potential.F file splitting the file..
	  Handling the case of very small
	  values of r (< 0.01, not in the spline range..). This is done
	  just to avoid the segmentation
	  fault (especially for free energy calculations) and a warning
	  message will be printed
	  on the screen when this happens.

2007-07-19 17:22  tlaino

	* [r6215] removing wrong comment

2007-07-19 16:49  tlaino

	* [r6214] Clarifying error statement and bug fix for
	  clean_intra_force for bends (reset 1 regtest)

2007-07-19 12:47  vondele

	* [r6213] added basis_molopt_quantities to aid basis_molopt
	  optimization

2007-07-17 13:04  vondele

	* [r6211] bug fix: pass an initialized error

2007-07-16 21:13  tlaino

	* [r6209] getting rid of another SAVE

2007-07-16 16:07  tlaino

	* [r6206] Converting few write(*, into write(output_unit,

2007-07-16 15:04  tlaino

	* [r6205] IO bug fix

2007-07-16 13:49  tlaino

	* [r6204] Removing SAVE variables NP* ensemble and few other
	  places.. removed empty file from regtest

2007-07-15 00:16  tlaino

	* [r6200] deallocation -> needed to get around a pgi bug.. now
	  memory is stable when using
	  executables compiled with pgi..

2007-07-13 11:16  vondele

	* [r6198] bug fix for npt_f restarts: write cell vectors as columns
	  of hmat in the restart.

2007-07-12 22:15  tlaino

	* [r6197] Cleaning the qs_scf_post module.. few IO bug fixes..
	  introducing also the
	  possibility to have the wannier centers as in CPMD in
	  IONS+CENTERS format..
	  These can be dumped in xyz/dcd/atomic... added a regtest..

2007-07-11 20:40  vondele

	* [r6195] generalise / allow nptf

2007-07-11 17:45  vondele

	* [r6193] faster nonortho routines

2007-07-11 14:12  tlaino

	* [r6192] bug fix [CP2K:162].. cleaning..

2007-07-11 10:30  vondele

	* [r6191] remove unused var

2007-07-10 10:08  vweber

	* [r6190] fixed a problem with number of angular grid points
	  (gapw).

2007-07-09 18:27  vondele

	* [r6189] avoid tmp

2007-07-09 08:44  tlaino

	* [r6188] valgrind regtest: bug fix

2007-07-06 14:48  vondele

	* [r6187] bug fix non-ortho

2007-07-06 07:29  tlaino

	* [r6186] Patching filename lengths and intel machine file (Axel
	  K).. prettifying

2007-07-05 21:33  vweber

	* [r6185] added the on the fly localization for the OT/IRAC.

2007-07-05 20:03  vondele

	* [r6183] cleanup

2007-07-05 18:23  vweber

	* [r6182] bug fixed with n-list

2007-07-05 17:27  tlaino

	* [r6181] can't use thermostat if there are no degrees of freedom..
	  stopping to avoid an
	  arithmetic exception.

2007-07-05 16:31  tlaino

	* [r6180] Shell Model : few more bug fix (in nonbonded and
	  electrostatics).. now working..
	  added few more debugging regtests and reorganized all
	  shell-models input files
	  in a new directory (teo & marcella)

2007-07-05 08:35  tlaino

	* [r6179] procedure name too long (more than 32 chars)

2007-07-03 17:04  tlaino

	* [r6178] no need to zero ..

2007-07-03 16:18  tlaino

	* [r6177] cleaning.. removing unused keyword.. adding the
	  possibility to start_from_scratch the
	  neighbor_lists in fist at each MD step.. added also the
	  verlet_skin keyword..

2007-07-03 12:37  vondele

	* [r6176] general unit cells for QS: initial implementation (proof
	  of concept implementation for debugging only)

2007-07-02 21:51  vweber

	* [r6175] updated the IRAC algorithm: matrix multiply count is now
	  optimal, added a Cholesky and Lowdin orthogonalization functions.
	  updated the regs as well.

2007-07-02 16:19  tlaino

	* [r6174] enabling local log.. fixing format and reducing the
	  maximum number of
	  iterations for shake...

2007-07-02 13:53  tlaino

	* [r6173] just cosmetics...

2007-07-02 13:15  mguidon

	* [r6171] Implemented periodic boundary conditions. Optimized DGEMM
	  calls in forces.

2007-07-02 00:46  tlaino

	* [r6169] Bug fix for NP* runs mixed_env.. Bug fix for general NP*
	  and NVT : making sections
	  BAROSTAT and NOSE non-optional.. this means that in NP* and NVT
	  runs the header of the
	  BAROSTAT or NOSE must always be present (default values can
	  always be used)..
	  this avoids possible bugs when using in cascade restarts produced
	  with MD with
	  different ensembles..
	  added one more regtest for restart mixed_env NPT..

2007-06-28 08:05  vondele

	* [r6167] dftb bug fix

2007-06-24 08:35  tlaino

	* [r6165] Optional unique spline generation for LJ. Cleaning and
	  rewriting the setup of
	  pair_potential splines. bug fix for shell-modell cutoff of
	  electrostatics..
	  Cosmetics. Few regtests resetted.

2007-06-21 14:46  vweber

	* [r6163] fixed problems with uks-linres precond and state dep
	  perturbation for uks (too many loops over spin).

2007-06-19 21:35  tlaino

	* [r6162] NAG compiler issue.. cannot use a symbol before its
	  declaration

2007-06-19 20:29  vweber

	* [r6161] fixed problem with a matrix dimension for uks, allocation
	  has been moved inside
	  the DO-loop over the spin. This should fix [ Bug #10092 ]
	  submitted by
	  fxcoudert lonnngggg time back.

2007-06-19 15:00  fawzi

	* [r6160] removed stupid tipole, added total density

2007-06-19 13:14  vondele

	* [r6159] Improve contraction step and data passing (speedup 35%)
	  derivates still need to be done

2007-06-18 15:55  marcella

	* [r6157] bug fix

2007-06-18 11:09  vweber

	* [r6155] Toward uks-nmr. linres convergence improved. Still more
	  bugs in qs_loc_xxx (i.e. [ Bug #10092 ])

2007-06-15 13:32  vondele

	* [r6154] speed up computation of the screening matrices

2007-06-15 11:46  marcella

	* [r6152] optional deallocation of some input sections to save some
	  memory with large systems

2007-06-14 09:36  vweber

	* [r6148] NMR-gapw: bugs fixed, clean up, chi-gapw done
	  analytically (should work now).

2007-06-13 16:06  mguidon

	* [r6147] Truncated procedure names

2007-06-13 11:06  mguidon

	* [r6146] Consitency with KS-matrix sparseness. New load balance
	  algorithm in parallel

2007-06-12 18:46  tlaino

	* [r6145] small speed-up for non-shell models and cosmetics

2007-06-12 14:22  tlaino

	* [r6144] Saving memory for 1 single force_eval (marcella)..

2007-06-12 00:08  tlaino

	* [r6143] bug fix in fist_debug (we need to update the neighbor
	  lists when computing
	  the numerical stress tensor..) and cleanings.. reset 1 newly
	  added regtest
	  in order to avoid to compare a null energy..

2007-06-11 19:59  tlaino

	* [r6142] bug fix for generate (when two hydrogens are connected
	  each other) (Toon)

2007-06-11 09:40  tlaino

	* [r6141] more cleaning..

2007-06-11 09:04  tlaino

	* [r6140] mainly cleaning and some tiny further optimization..

2007-06-10 23:26  tlaino

	* [r6139] bug fixes for fist debug module.. updated the module to
	  the most relevant
	  modification of the FIST FF.. bug fix for a wrong index in fist
	  neighbor list..
	  regtests ok..

2007-06-09 13:14  vondele

	* [r6137] bug fix: do not access force arrays if they may not be
	  associated

2007-06-08 23:59  tlaino

	* [r6136] bug fix for fist_debug and added a regtest FIST DEBUG

2007-06-08 11:26  jgh

	* [r6134] Add new SCF options, SCF_GUESS=NONE, and 0 SCF
	  iterations,
	  Caution: this allows for non-save runs and might have wrong
	  energy printings

2007-06-07 15:27  tlaino

	* [r6132] Cleaning pair_potential routine and type and bug fix in
	  the 0 of the electrostatic pot on spline
	  for shell_model. reset 1 regtest

2007-06-07 11:55  jgh

	* [r6131] move pw_grid_create_copy_no_pbc from pw_grids to
	  qmmm_pw_grid, one REGTEST reset

2007-06-07 07:16  vondele

	* [r6129] fix name

2007-06-06 19:17  vondele

	* [r6128] remove unused array

2007-06-06 19:05  vondele

	* [r6127] tiny opt

2007-06-06 11:28  tlaino

	* [r6126] handling CMAP extension in the CHARMM pot file and also
	  the official continuation
	  line "-" @#$!@#$@#!!!!

2007-06-06 09:10  vondele

	* [r6125] parser speedup

2007-06-05 22:20  tlaino

	* [r6124] Three pigeons with one bean: Bug fix for PBC in coorect
	  bondend electrostatics both
	  energy/forces and virial.. Major rewriting of the routine and
	  heavy cleaning.. Bug
	  fix in stress tensor for shell models (was missing the
	  contribution from shell_i//core_i)..
	  Added two more regtests to check that PBC will be preserved since
	  now on..
	  Resetted several regtests due to numerics (and to the bug fix for
	  the shell model)

2007-06-05 14:57  mguidon

	* [r6122] Use symmetry in Fock-Matrix construction, makes it a bit
	  faster

2007-06-05 14:43  vondele

	* [r6121] tiny opt

2007-06-05 13:37  vondele

	* [r6120] an even clearer hint

2007-06-05 11:03  mguidon

	* [r6119] Optimization of memory-handling

2007-06-05 10:34  tlaino

	* [r6118] Patching the point in colvar (introducing the possibility
	  to have different
	  weights) (Patch by Rodolphe Vuilleumier). added new regtest.

2007-06-05 09:19  vondele

	* [r6117] tighten code (50% speedup)

2007-06-04 22:21  tlaino

	* [r6116] bug fix for full periodicity (problem pointed out by
	  Joost).. Bug fix in the
	  stress tensor for internal forcefields (bonds, angles,
	  torsions).. were not
	  PBC friendly.. resetting 1 regtest and added a new one checking
	  we are really
	  fully periodic.

2007-06-04 17:11  tlaino

	* [r6115] Accepting patch of Toon [CP2K:119] about the exclusion
	  lists vs periodic replicas.. added a new regtest

2007-06-04 16:00  mguidon

	* [r6114] Included all 8 symmetries in eri calculation.
	  Speedup/Memory-Reduction 400%

2007-05-31 14:42  vweber

	* [r6113] NMR: fixed a bug of the type a(1:i)=b(1:j) with i.ne.j in
	  general.

2007-05-30 22:26  tlaino

	* [r6112] possible bug fix.. IN/OUT argument referenced on same
	  variable.. not all
	  compilers create in this case a temporary variable..

2007-05-30 15:13  mguidon

	* [r6111] Clean up, bug fix in buffer-handling

2007-05-30 11:25  mguidon

	* [r6108] Bug fix in RESTART procedure for system with nmo = 0

2007-05-30 08:16  vondele

	* [r6106] deal with zero orbitals

2007-05-30 07:53  vondele

	* [r6105] init to zero (needed for metadyn on mixed envs)

2007-05-29 17:41  vondele

	* [r6104] fix format

2007-05-29 12:51  mguidon

	* [r6102] Added LSD to HSE functional

2007-05-29 09:49  tlaino

	* [r6101] Bug fix in restart for fixed atoms when atoms are on
	  different kind of molecules..
	  added two more regtests.

2007-05-28 18:17  vweber

	* [r6098] Protected the nmr modules against non-working keywords
	  and added comments in the reference manual. Can now do NICS
	  calculations.

2007-05-27 21:44  vweber

	* [r6093] New OT is now working with diis.

2007-05-27 20:29  tlaino

	* [r6092] print_key feature also for MOLECULAR_DIPOLES.. changed
	  the global common_iteration_level
	  for the print_key localization in order to make it uniform to the
	  others..

2007-05-27 19:53  tlaino

	* [r6091] clean..

2007-05-27 16:48  tlaino

	* [r6090] Fixing MOLECULAR_STATES: enabling control and dump info
	  file via the
	  print_key structure (Limin).. some cleaning..

2007-05-27 13:30  tlaino

	* [r6089] Bug fix in counting the number of fixed atoms
	  restraints.. allowing
	  the keywords LIST and RANGE in the fixed_atom section to be
	  repeatable..
	  some cleaning around..

2007-05-27 11:21  tlaino

	* [r6088] Added a rescale_forces section for bad starting
	  geometries.. new regtest..
	  changes NMR regtest -> too slow for NAG and memcheck

2007-05-27 10:16  vondele

	* [r6087] add check for RESPA and NVE

2007-05-25 11:09  vweber

	* [r6085] Clean up. Fixed few bugs with parallel. Chemical shifts
	  are now computed by adding up g-coeffs (seems more accurate).
	  Interpolation based method to compute the shifts needs to be
	  carefuly parallelized (problem with the distributed grid).

2007-05-24 08:33  vweber

	* [r6084] Fixed an array out of bound and a problem with parallel
	  printing.
	  CV:
	  ----------------------------------------------------------------------

2007-05-23 20:22  vweber

	* [r6080] clean up, remove dead code and dbg printing, deallocate
	  arrays, added checksum for regtest.

2007-05-23 10:24  mguidon

	* [r6079] Added screening to HFX forces, screening matrices now
	  stored in buffer.

2007-05-23 07:59  krack

	* [r6078] use pi from mathconstants to avoid ACOS function in
	  declaration which is
	  not tolerated by the NAG compiler

2007-05-22 14:38  vondele

	* [r6077] prettify

2007-05-22 14:18  vondele

	* [r6076] added screening on the density matrix for the forces

2007-05-22 13:14  mguidon

	* [r6073] New exchange functional, used for HSE hybrid

2007-05-20 11:53  tlaino

	* [r6072] bug fixes for target values in constraints/restraints..
	  bug fix when using fixed atoms
	  together with colvar.. added a new regtest..

2007-05-19 20:06  vondele

	* [r6071] added separate screening for the forces and use new
	  defaults

2007-05-19 15:21  jgh

	* [r6070] Remove print_section from pw_grid_type

2007-05-19 09:21  vondele

	* [r6069] use some earlier screening opportunities

2007-05-19 09:01  vondele

	* [r6068] add one more year

2007-05-18 20:17  vondele

	* [r6067] more accurate screening taking contraction into account

2007-05-18 17:10  tlaino

	* [r6066] bug fix : recursive IO..

2007-05-17 17:15  jgh

	* [r6065] Bug fix for new realspace grid printing, parallel case

2007-05-17 14:56  jgh

	* [r6064] Disentangle realspace grid type from input/output
	  information

2007-05-17 14:50  jgh

	* [r6063] Bug fix in two center terms (RD)

2007-05-16 19:15  vweber

	* [r6062] Update the periodic "dipole" moment matrices for the l1
	  norm based localization scheme.
	  This allows the computation of the Wannier centers, spreads and
	  molecular dipole moments.

2007-05-16 16:38  jgh

	* [r6061] Commenting out old code (L operator), correcting last
	  commit (RD)

2007-05-16 16:18  jgh

	* [r6060] Disentangle realspace routines and task management; use
	  rs pools in pme/spme

2007-05-16 16:17  jgh

	* [r6059] Use GFlops performance measure in fft3d_s

2007-05-16 15:19  vweber

	* [r6058] fixed few bugs and changed mic in integral codes.

2007-05-14 19:28  vweber

	* [r6057] removed some printing, fixed few bugs, added checksum for
	  regtest... toward working nmr

2007-05-14 12:32  vondele

	* [r6056] use abs for easier mixing of vdw parameters

2007-05-13 17:12  tlaino

	* [r6055] Making restart more robust to possible (and
	  unpredictable) kind of uses .. Fixing doc for metadynamics...

2007-05-13 06:08  tlaino

	* [r6054] removing keyword..

2007-05-12 15:40  jgh

	* [r6053] DFTB: enable preconditioners with OT

2007-05-12 13:34  tlaino

	* [r6049] printing biased and weighting energy for alchemical
	  change free energy calculations.
	  1 more regtest.

2007-05-12 13:01  jgh

	* [r6048] DFTB: enable restarts and MO printing, 3 new regtests

2007-05-10 18:55  vondele

	* [r6044] Limin: bug fix

2007-05-10 09:09  tlaino

	* [r6039] fixing uninitialized variable in xas(valgrind).. just a
	  tag: today valgrind gives
	  a cleaned regtest. 0 error reported.

2007-05-09 11:30  fawzi

	* [r6038] prettify

2007-05-09 11:29  fawzi

	* [r6037] fixing parallel bug

2007-05-09 11:29  fawzi

	* [r6036] fixing cp_print_key_log in parallel

2007-05-09 09:29  tlaino

	* [r6035] bug fix BFGS + restraints.. added regtest

2007-05-09 06:59  tlaino

	* [r6034] Introducing the keyword NGROUP.. if not specified a
	  number of groups equal to
	  the number of subforce_eval defined is tried to be allocated..

2007-05-09 06:15  tlaino

	* [r6033] bug fix in the documentation... thanks joost..

2007-05-08 17:50  tlaino

	* [r6032] Making the sub_force_eval really parallel for mixed_env..
	  all defined sub_force_eval
	  can be run in parallel, exactly like the replica_env.. added one
	  more regtest for
	  group_partition ..

2007-05-07 11:15  vondele

	* [r6028] Valery Weber : New OT algorithm imposing orthogonality
	  with refinement. New localization based on L1 norm. (Regtests to
	  follow)

2007-05-07 10:44  vondele

	* [r6027] compute_screening_matrices: bugfix and speedup (10x)

2007-05-06 23:58  tlaino

	* [r6026] Virial for mixed_env.. enabling the use of NP* ensemble
	  for mixed force eval calculations..
	  removing old stuff and cleaning.. new regtest for NPT mixed force
	  eval..

2007-05-06 21:56  vondele

	* [r6025] 100% speedup for in-core calculations

2007-05-06 20:57  vondele

	* [r6024] bugfix: uninit. iw (any takers for a 16-way valgrind?)

2007-05-06 18:43  vondele

	* [r6023] few percent speedup

2007-05-06 17:00  tlaino

	* [r6022] Trying to make a little bit more logical the behavior of
	  pv_availability..
	  It was always set to .false. even fot methods like FIST that
	  normally have
	  the stress tensor.. more dangerous was the fact that you may have
	  requested
	  an NPT run for a method that had no virial and the run was
	  smoothly going through..
	  This has been now changed. If virial is not available the job
	  stops. One
	  may decide to force the run switching on the keyword PV_AVA..
	  Added a regtest for QS and NPT_I (just to check the calculation
	  of the virial for QS)

2007-05-05 20:21  vondele

	* [r6021] added cost model for the 2d distribution for HFX memory
	  use, including sparsity.

2007-05-05 11:28  vondele

	* [r6018] factor out compute_screening_matrices and prettify

2007-05-05 09:22  tlaino

	* [r6017] Heavy cleaning of the force_field pack.. introducing the
	  possibility to skip both
	  nonbonded and nonbonded14 and keeping only internal FF. Added two
	  new regtests
	  (computing only internal FF)

2007-05-04 20:47  tlaino

	* [r6016] bug fix for stress tensor with tersoff.. modified a
	  regtest (from 3 md to 10 md steps) since
	  this bug was introduced recently and it was passing through the
	  regtest (3 md steps not enough?! )

2007-05-04 17:11  tlaino

	* [r6015] [CP2K:40] bug fix on printing HILLS..

2007-05-04 13:58  vondele

	* [r6013] add more printouts

2007-05-04 12:51  vondele

	* [r6012] further cost model

2007-05-03 21:09  tlaino

	* [r6010] bug fix for FE AC.. added regtest.. some cosmetics..

2007-05-03 14:48  fschiff

	* [r6007] added difference constraint for MD runs

2007-05-03 08:57  marcella

	* [r6006] format

2007-05-03 08:49  marcella

	* [r6005] bug fix

2007-05-02 22:02  marcella

	* [r6001] control of kinetic energy and center of mass velocity,
	  for more stability
	  with the shell model. NHCOPT GLOBAL seems to be stable also
	  without rescaling, otherwise the
	  internal core-shell kinetic energy needs to be controlled.

2007-05-02 19:32  tlaino

	* [r6000] bug fix..

2007-05-02 16:46  tlaino

	* [r5999] Bug fix in the extended lagrange metadynamics.. Inserted
	  also a TEMP_TOL parameter
	  and added a new regtest.

2007-05-02 15:52  fschiff

	* [r5998] accelerated becke_constraint, added the possibility to
	  constrain charge differences and made it possible to run MD with
	  becke constraint (differences therefor not yet implemented)

2007-05-02 11:36  vondele

	* [r5997] fix routineN

2007-05-02 08:41  vondele

	* [r5996] add a few syncs to make timings clearer

2007-05-02 07:38  vondele

	* [r5995] fix parallel output issue

2007-05-01 19:02  vondele

	* [r5993] improve parallel 2d particle distribution for QS. Use
	  monte carlo annealing to globally optimize distribution, and
	  introduce different cost models. 50 percent QS speedup in best
	  case. Reset a number of regtests due to numerics introduced by
	  different particle orderings.

2007-04-30 06:29  vondele

	* [r5988] prettify

2007-04-30 06:25  vondele

	* [r5987] delete unused module

2007-04-28 13:59  tlaino

	* [r5985] Core of the FE-PQA: general mapping in force_env mixed..
	  added new regtest

2007-04-28 07:18  tlaino

	* [r5984] Deleting the debug write statement.. cleaning debug
	  code.. prettifying

2007-04-27 21:12  ikuo

	* [r5982] Add smearing for qparm colvar

2007-04-27 19:05  ikuo

	* [r5981] Add hydronium collective variable by Marcel Baer

2007-04-27 16:55  marcella

	* [r5980] still on the way for more stability with the shell-model

2007-04-27 06:31  vondele

	* [r5979] clean irrelevant comments away

2007-04-25 23:25  tlaino

	* [r5978] Bug fix: comments on groups.google.com/group/cp2k
	  messages 1-6. Thanks to Nick for
	  having pointed out the problem.. now trajectory/velocities/forces
	  etc.. can have different
	  format in the output files..

2007-04-25 13:17  vondele

	* [r5976] avoid segfaults in the restricted case, skip MO
	  calculation because it is unclear what we do

2007-04-25 08:35  tlaino

	* [r5975] intent bug fixed

2007-04-24 18:39  fawzi

	* [r5974] fixing leak

2007-04-24 14:03  tlaino

	* [r5973] fix arithmetic exception

2007-04-24 12:40  fawzi

	* [r5971] reactivated ep e2 correction, placeholder for rotations

2007-04-24 06:28  tlaino

	* [r5969] Removing specific DCONJG and DABS, DSQRT.. using instead
	  the generic ones..
	  Removing the hardcoded double precision i.e. 0d0 -> 0._dp..
	  Using a default parameter for alpha to fix an undefined problem
	  (though alpha is not yet used)..
	  Prettifying..

2007-04-23 23:00  ikuo

	* [r5968] Fix failed regtest. Comment out alpha which is not in
	  used right now...

2007-04-23 21:29  tlaino

	* [r5967] restoring the proper print level for constraint_info.. we
	  don't want this
	  to be printed for every run.. most of the time it is a useless
	  info..

2007-04-23 20:56  ikuo

	* [r5965] Bug fix for force in colvar qparm

2007-04-20 19:19  jgh

	* [r5961] Fix memory leaks in DFTB, add 26 regtests;
	  ToDo list: validate dispersion code, debug stress tensor, allow
	  for NPT_F

2007-04-20 12:37  vondele

	* [r5959] added coeff in the ddapc restraint

2007-04-19 21:29  tlaino

	* [r5958] more and more closer to FE-PQA: empowering the MOL_SET
	  feature.. now it is
	  possible define a single molecule with PSFs of fragments and
	  explicitly declaring
	  the further terms.. This required a reorganization both of the
	  input sections (sorry about that guys!!)
	  but also in the code handling the reading and post_process of PSF
	  info.
	  Few regtests added and changed the syntax of MOL_SET in MC
	  regtests.
	  Added also a preliminary section of the new free energy
	  method(not working yet).

2007-04-19 17:46  ikuo

	* [r5957] update docs for input

2007-04-18 17:33  jgh

	* [r5956] Update for Stress tensor: GAPW_XC, GAPW, DFTB

2007-04-17 22:29  fawzi

	* [r5955] reverting my bad fix

2007-04-17 18:49  fawzi

	* [r5954] stopping on closing unit 6

2007-04-17 18:49  fawzi

	* [r5953] added the possibility to fix the whole filename

2007-04-17 17:38  fawzi

	* [r5951] better description of filename, fixed possible closing of
	  std_out

2007-04-17 16:18  fawzi

	* [r5950] better handling of filename

2007-04-17 11:14  tlaino

	* [r5949] consistent restart for mixed_env (on the basis of the
	  real subsys defined)..
	  inserting an accurate way of computing the derivative of compiled
	  functions.
	  regtest ok..

2007-04-16 21:26  tlaino

	* [r5948] One step more towards FE-PQA: General mixing for
	  multiple_force_env..
	  you can specify exactly as for the general potential
	  a mathematical expression mixing the two or more force_eval.
	  Keeping the one implemented
	  as special cases though they could be reproduced exactly with the
	  general mixing..
	  regtest ok .. added one..

2007-04-16 12:55  tlaino

	* [r5947] handling subsystem coherently with the keyword
	  multiple_subsys..
	  Generalizing mixed_force_env to several force_eval and not just
	  two..
	  regtest ok.

2007-04-16 06:15  vondele

	* [r5946] more force env mixing flavours

2007-04-15 22:26  tlaino

	* [r5943] [Patch #1963]: Axel Kohlmeyer: correct timings on cray
	  xt3

2007-04-15 22:07  tlaino

	* [r5942] Mainly getting ready for FE-PQA: (1) A more logical setup
	  for multiple_force_eval..
	  In order to save memory and also to avoid a duplication of
	  information it is possible to specify
	  just one single subsys when using multiple_force_eval (or
	  different)..
	  This has been possible cleaning all referential calls to subsys..
	  The code now
	  is definitely more structured and cleaned.
	  At the same time several cleanings have been performed while I
	  was working on the single files.
	  In particular structure_data has been moved to the print section
	  of the motion section.
	  Same for printing subcell.. moved to the respective section in MM
	  and DFT.
	  This reflects the idea of making logical partitions in the input.
	  Heavy lifting around (not just cosmetics).. regtest ok!

2007-04-15 15:53  mcgrath

	* [r5940] Multiple temperatures for the virial

2007-04-14 13:37  vondele

	* [r5938] to keep this case within the format of psf

2007-04-12 17:56  mcgrath

	* [r5936] Virial calculation gives the correct converged answer for
	  SPC/E

2007-04-12 07:06  jgh

	* [r5935] Use generic name for CONJG

2007-04-11 21:22  ikuo

	* [r5933] Add crystalline order parameter based on spherical
	  harmonics. JCP 96 4665 (1992). (Nir and Will)

2007-04-11 17:44  tlaino

	* [r5932] just some cleaning..

2007-04-11 14:27  vondele

	* [r5931] added bisection as a way to minimize the outer_scf loop
	  in 1D cases

2007-04-11 08:49  tlaino

	* [r5930] Tsunami effects (1): Analytical gradients for
	  semiempirical methods. Nuclear gradients
	  are since now on computed analytically. Also the value of the
	  integral for the limiting
	  case of two atoms aligned in the z direction is computed EXACTLY
	  without performing
	  the limit. The old situation can be restored disablign the
	  keyword ANALYTICAL_GRADIENTS in the
	  SE section.
	  New regtests. Resetting few qm/mm SE because of the use of
	  analytical expressions.
	  waiting for the next wave...

2007-04-09 11:56  tlaino

	* [r5929] Earthquake in QM/MM module: Introduced a new type:
	  Coulomb and deleted all the
	  *_qmmm_grid related stuff (was there only for debugging the
	  original version, not used
	  anymore and was only requiring time to keep it updated!!!)..
	  Implemented electrostatic coupling (Only Coulomb at the moment)
	  for the
	  semiempirical..
	  Analytical version for drotnuc -> drotnuc_ana .. once also the
	  drotint will be available
	  analytical gradients will be able to be used also in pure SE
	  runs..
	  Few more regtests added for qm/mm SE.. deleted regtest with
	  "ECOUPL GRID"..
	  Several clean around...
	  Beware to Tsunami.. regtest passed..

2007-04-08 16:30  tlaino

	* [r5928] bug fix dumping topology during qmmm runs

2007-04-06 14:35  jgh

	* [r5927] Full stress tensor for GPW, towards stress tensor for
	  GAPW

2007-04-06 04:15  ikuo

	* [r5926] Change default output level for LAGRANGE_MULTIPLIERS to
	  high

2007-04-04 06:26  jgh

	* [r5925] DFTB: more debugging + SPME + dispersion

2007-04-03 13:03  tlaino

	* [r5924] skipping ddap charges for unimplemented methods..

2007-04-03 11:27  gtb

	* [r5921] Bug fix: generalized the initialization according to
	  grid_level
	  MC_KG test now works

2007-04-03 09:36  tlaino

	* [r5920] bug fix in KG: the initialization of maxradius for KG was
	  not grid_level dependent.
	  resetting one regtest.

2007-04-02 15:45  tlaino

	* [r5919] fixing problem of floating point exception.. (for few
	  compilers..)

2007-03-31 15:32  jgh

	* [r5917] Allow for print out of matrices without GTO basis set
	  defined

2007-03-31 14:35  jgh

	* [r5916] Towards working DFTB: 15 regtests reset due to numerics

2007-03-31 14:25  jgh

	* [r5915] Bug [010756] fix for exponentiation of negative number
	  with real, not allowed by some compilers

2007-03-30 15:03  vondele

	* [r5914] new resp default

2007-03-29 10:49  marcella

	* [r5912] shell-core exclusion + some printing and fixing

2007-03-29 07:56  vondele

	* [r5911] fix for systems where default integer is integer*8

2007-03-28 08:08  vondele

	* [r5910] make the count variable a longer int

2007-03-28 07:14  vondele

	* [r5909] do not paper over bugs elsewhere

2007-03-27 17:57  tlaino

	* [r5908] core-shell interaction potential: modification of the
	  potential due to the
	  very short distance between the two particles. Not enabled yet
	  (the honour to do that is of Marcella ;-) )..

2007-03-27 10:32  vondele

	* [r5907] added a default restraint to zero to the heavy atoms in
	  the resp fit

2007-03-27 09:36  vondele

	* [r5906] have the cray trace print out memory usage

2007-03-26 14:01  tlaino

	* [r5905] just changing name to the qmmm module and getting ready
	  for the qmmm semiempirical/dftb module..

2007-03-26 12:21  tlaino

	* [r5904] Forcing all colvar to use PBC. Bug fix in the defaults
	  for tersoff potential.
	  new regtest for Si with defaults.

2007-03-25 22:38  tlaino

	* [r5903] Grouping conditional statements for shell models in order
	  to reduce the
	  logical checks for normal runs (no-shell models)

2007-03-25 21:46  tlaino

	* [r5902] New fast implementation of tersoff potential. parallel
	  version ok. have been
	  fixed also few potential bugs in the original implementation
	  (spec. in manybody_energy, and
	  the use of hardcoded minimum cell images). Deleted the keyword
	  RCUT, being redundant
	  (RCUT=R+D).
	  Few modifications were necessary in the neighbors lists in order
	  to have an
	  efficient parallel version. The modifications enable also the
	  possibility
	  to use tersoff potential together with other different kind of
	  force_fields.
	  Regtest ok.

2007-03-24 20:10  marcella

	* [r5901] bug fix

2007-03-23 20:42  vondele

	* [r5898] improved (faster, more accurate) QS grid kernels. QS
	  >200% faster in favourable cases.

2007-03-23 15:06  vondele

	* [r5896] clarify

2007-03-23 13:01  tlaino

	* [r5895] Introducing the possibility to have colvar based on
	  combination of atoms..
	  Bug fix for rotation_bond colvar.. Introducing few more regtests
	  for the
	  new colvar feature.. reset 1 regtest.

2007-03-23 09:17  jgh

	* [r5894] DFTB code: not working, debugging and regtests will be
	  comming soon

2007-03-21 19:21  vondele

	* [r5892] removed

2007-03-21 17:17  ikuo

	* [r5890] Fix bug for parallel Tersoff

2007-03-21 15:27  marcella

	* [r5888] shell_model npt

2007-03-21 15:21  vondele

	* [r5886] new (faster) algorithm to compute the Gaussian radius,
	  affects basically all QS regtests to within ~ EPS_DEFAULT

2007-03-21 04:49  ikuo

	* [r5885] clean up manybody_potential
	  re-fix units problem in dcd. (Note: units for force is not yet
	  fixed)

2007-03-20 21:48  ikuo

	* [r5883] Move EAM and Tersoff into manybody_potentail

2007-03-20 18:40  marcella

	* [r5882] shell model nvt

2007-03-20 17:23  tlaino

	* [r5881] getting rid of the buffer overflow on output unit for few
	  compilers..

2007-03-20 15:33  mguidon

	* [r5880] optimized derivatives code

2007-03-19 19:27  tlaino

	* [r5879] Bug fix in counting external constraints for 3x3 and
	  4x6.. resetting 4 regtests.
	  Possibility to dump lagrange multipliers (X Chris) activating the
	  proper print_key.
	  Cleaning debug comments left in the code.. and moving psf, xyz,
	  and pdb out of the regtest directory
	  of fist.

2007-03-19 11:31  mguidon

	* [r5877] Bug fix

2007-03-16 17:18  mguidon

	* [r5874] bug fix

2007-03-16 15:40  mguidon

	* [r5872] Improved screening, optimized code for energy

2007-03-15 22:28  marcella

	* [r5871] erased dbg line

2007-03-15 13:46  marcella

	* [r5869] bug fix

2007-03-14 21:21  mcgrath

	* [r5868] Added the calculation of the second virial coefficient
	  and regtest, as well as a few minor changes elsewhere

2007-03-14 16:51  marcella

	* [r5867] cleaning

2007-03-14 16:41  marcella

	* [r5866] cleaning

2007-03-14 16:39  marcella

	* [r5865] cleaning

2007-03-14 12:52  marcella

	* [r5863] shell-model NVE running

2007-03-14 09:33  mguidon

	* [r5862] Bug fix

2007-03-13 13:03  mguidon

	* [r5859] optimized code

2007-03-13 06:22  ikuo

	* [r5858] Add virial for Tersoff

2007-03-12 20:28  ikuo

	* [r5857] bug fix for tersoff

2007-03-11 05:11  ikuo

	* [r5854] Tersoff potential with correct energy and forces

2007-03-09 13:57  fschiff

	* [r5852] Added a method for calculating diabatic electronic
	  couplint matrix element for ET reactions
	  by using constraint DFT (either DDAPC or new implemented Becke
	  weight population)

2007-03-09 12:20  marcella

	* [r5851] commented debug

2007-03-08 20:00  ikuo

	* [r5850] remove debug write statements

2007-03-08 07:24  ikuo

	* [r5846] remove central force water potential
	  add tersoff potential (energy is okay, force is wrong)

2007-03-06 19:08  marcella

	* [r5842] shell-model NVE

2007-03-01 14:21  tlaino

	* [r5827] allowing keywords ATOMS_FROM and ATOMS_TO being
	  repeatable.. (Chris)

2007-03-01 14:21  mguidon

	* [r5826] Added HF-Exchange derivatives (forces)

2007-03-01 09:23  tlaino

	* [r5822] fixing double precision in pair_potential and
	  prettifying.. added also a regtest
	  cf_genpot.inp...

2007-03-01 09:14  tlaino

	* [r5821] Fixing a problem with GENPOT when negative floating point
	  are raised to real powers

2007-02-28 21:47  ikuo

	* [r5819] Add central force water (Nir Goldman and Chris Mundy)

2007-02-27 19:02  jgh

	* [r5815] Refactoring

2007-02-27 09:22  jgh

	* [r5814] Bug 010494 : Add error and idir to the OpenMP shared
	  clause

2007-02-26 13:40  tlaino

	* [r5812] moving unused slater_koster routines to the Attic.
	  Allowing the compilation of ai_overlap_k
	  (for future use) in order to avoid the shift to the Attic.

2007-02-26 13:29  vondele

	* [r5811] added RESP charges

2007-02-25 14:07  tlaino

	* [r5810] Fixing the IO problem posted by G.Fluder [Support
	  #102425]: Removed the keyword
	  DUMP_TOP and added two more print_keywords DUMP_PDB and DUMP_PSF,
	  controlling in a better way the
	  output in parallel runs. bug fixed and changed in few regtest the
	  DUMP_TOP keyword with the two added new.

2007-02-23 18:23  jgh

	* [r5809] Remove EHT method, was not operational

2007-02-23 17:23  tlaino

	* [r5808] Cleaning some different keyword names for files....
	  CONN_FILE moved into CONN_FILE_NAME, COORD_FILE into
	  COORD_FILE_NAME and PARMFILE into PARM_FILE_NAME.
	  Previous keywords are still allowed as variants (just for
	  compatibility with older input files).
	  Changed regtests in order to have them representing always the
	  official name convention.

2007-02-23 10:31  jgh

	* [r5807] g-tensor code (Reinout Declerck), not yet operational;
	  activates a NAG compiler bug - fails two Fist regtests

2007-02-22 15:43  vondele

	* [r5806] fix format

2007-02-21 19:15  tlaino

	* [r5805] S-Wave Gaussian Expansion library (GEEP). More regtests.
	  clean and reorganized the directory
	  of QMMM regtests in order to avoid multiple copies of same files.

2007-02-21 12:24  jgh

	* [r5804] Correct print out of units

2007-02-19 19:42  vondele

	* [r5800] added initial version for testing

2007-02-19 18:53  jgh

	* [r5799] Overlap integrals for k-points, just added the code for
	  future use

2007-02-19 13:33  tlaino

	* [r5798] On a medium system size (80K atoms) distributed is
	  slightly slower than replicated (< 10% for spme only).
	  Therefore restoring replicated as default and in case of memory
	  problems the distributed grid option is
	  the right choice to go for..

2007-02-17 21:45  tlaino

	* [r5797] Adding two keywords in the HBONDS constraint section:
	  EXCLUDE_QM and EXCLUDE_MM.
	  These logicals exclude (if true) QM or MM bonds (respectively)
	  involving hydrogens from the list of
	  shake (requested feature by G.L.).. 2 new regtests.

2007-02-16 15:52  tlaino

	* [r5795] Making distributed grids PME/SPME the default. Collecting
	  common routines to
	  PME and SPME (avoiding duplication of similar code). Changing
	  name of
	  input_poisson into input_cp2k_poisson in order to have files with
	  the same
	  name structure. Resetting few regtests of Fist due to the default
	  change.

2007-02-16 15:47  tlaino

	* [r5794] jgh: Bug fix for distributed PME grids

2007-02-15 13:32  jgh

	* [r5793] Bug fix for distributed SPME grids

2007-02-15 11:02  marcella

	* [r5792] another step towards adiabatic shell-model

2007-02-14 17:54  jgh

	* [r5789] Remove noise from atomic Coulomb potential on grid (for r
	  << 1);
	  All GAPW regtests reset, energy changed 10^-12 or less

2007-02-14 08:49  vondele

	* [r5784] do not stop on predictions, but on actually reached
	  tolerances

2007-02-13 20:14  tlaino

	* [r5783] Making the output of the constraint_info more clear and
	  useful..

2007-02-13 17:12  jgh

	* [r5779] Distributed real space grids for PME/SPME
	  Activate this option in EWALD section with "RS_GRID DISTRIBUTED"
	  Default is replicated grids
	  Further tests necessary
	  Improved speed and regtests are comming soon

2007-02-13 17:05  tlaino

	* [r5778] doubly NEB now available..

2007-02-13 15:28  jgh

	* [r5777] Use physical constants in hyperfine module (Reinout
	  Declerck)

2007-02-13 08:39  tlaino

	* [r5776] fix output format

2007-02-12 19:15  vondele

	* [r5775] send optical conductivity output to file

2007-02-12 18:22  jgh

	* [r5774] Remove overlap potential option from KG code, it did not
	  work.

2007-02-12 16:28  jgh

	* [r5772] More compact output for atomic and molecule kinds

2007-02-12 15:03  jgh

	* [r5770] Bug fix in wfn extrapolation/reorthogonalization for NDDO
	  methods

2007-02-12 09:48  jgh

	* [r5768] Exclude spin-resticted calculations from EPR hyperfine
	  tensor routine

2007-02-12 07:45  tlaino

	* [r5767] Fix out of bounds.. regtests still running?!?

2007-02-11 19:34  marcella

	* [r5766] shell-model : intrashell forces

2007-02-11 17:52  jgh

	* [r5765] EPR hyperfine coupling tensor, needs approval from
	  Reinout Declerck

2007-02-11 15:37  jgh

	* [r5764] Debug option for additional output in EFG

2007-02-10 08:20  vondele

	* [r5763] fix/clarify mo_cubes

2007-02-09 16:17  marcella

	* [r5762] bug fix

2007-02-08 15:38  marcella

	* [r5761] bug fix: check missing

2007-02-08 11:48  mguidon

	* [r5755] Added lsd part for vwn3, vwn5

2007-02-08 11:47  mguidon

	* [r5754] Added lsd part

2007-02-06 18:56  vondele

	* [r5751] clarify docs

2007-02-06 18:49  marcella

	* [r5750] some more stuff for shell model

2007-02-06 14:20  vondele

	* [r5749] Valery Weber: bug fix

2007-02-06 10:36  tlaino

	* [r5748] Moving NEB section into a more general BAND section..
	  reorganized also the
	  optimization keywords for band methods. regtests ok. Preparing
	  for the string method.

2007-02-02 15:45  vondele

	* [r5747] include 2007

2007-01-31 19:51  vondele

	* [r5746] comment more buggy OMP

2007-01-31 10:32  vondele

	* [r5745] comment incorrect OMP blocks

2007-01-30 21:34  vondele

	* [r5744] comment OMP clauses leading to wrong results already with
	  just one thread (as observed with ifort 9.1 20060707). More
	  troubles with more threads (sounds logical doesn't it ;-) are not
	  yet 'fixed'

2007-01-30 12:20  vondele

	* [r5743] get OMP code to compile again: OMP apparently still
	  *very* buggy

2007-01-26 19:58  vondele

	* [r5742] A try at improving the efficiency of the trace, and its
	  output in parallel (see TRACE_SYNC). In order for the latter to
	  work properly, all processes in the MPI group must call
	  timeset/timestop simultaneously (which most runtypes in CP2K do).

2007-01-26 16:24  marcella

	* [r5741] set up for shell-model potentials

2007-01-26 08:41  vondele

	* [r5740] cleanup: remove unused stuff

2007-01-25 09:25  vondele

	* [r5739] add a forgotten m_mov, which however can be made
	  functional

2007-01-24 19:10  tlaino

	* [r5738] new constraint.. |d1|-|d2|=const.. new regtest..

2007-01-22 12:27  tkuehne

	* [r5733] Bug fix

2007-01-19 13:14  tlaino

	* [r5732] Here we go.. Let's start with free energy methods.. UI
	  implementation..
	  One more regtest. Reorganization of themetadynamic section. now
	  all free
	  energy method are contained in a main folder called free_energy.
	  Cleaning the regtest directory of fist, moving the xyz, pot and
	  psf files
	  in correspondingly appropriate directories. few more work on free
	  energy
	  calculations later on..

2007-01-19 11:51  mguidon

	* [r5730] bug fix in xc_vwn, added vwn3, b3lyp now works

2007-01-18 19:54  ikuo

	* [r5729] Check for return status for system calls

2007-01-18 14:41  vondele

	* [r5727] Manuel Guidon: Add some hybrid functionals (see
	  tests/QS/regtest-hybrid)

2007-01-16 14:17  vondele

	* [r5725] code around a pgf90 bug ('conflicting' module names and
	  local variables?)

2007-01-15 19:26  tlaino

	* [r5724] some cleaning in the neb. one step further towards the
	  minimum free energy path driver (still few more work to do.. )

2007-01-15 09:50  tlaino

	* [r5723] resetting a regtest due to cvs problem with the server
	  (created a wrong reference)..
	  Moving the print of the COM info into a print_key section..

2007-01-14 00:24  tlaino

	* [r5722] Intermolecular restraints - new regtests.. some more
	  detailed IO for constraints and restraints.

2007-01-13 12:43  tlaino

	* [r5721] bug fix for energy restraints (parallel version)

2007-01-12 17:55  tlaino

	* [r5720] IO bug fix in vibrational module

2007-01-12 17:03  tlaino

	* [r5719] INTERmolecular constraints working... three regtests
	  showing how to use them..
	  have fun..

2007-01-11 23:58  ikuo

	* [r5717] Fix print to ionode

2007-01-11 23:36  ikuo

	* [r5716] write only to ionode

2007-01-11 17:35  tlaino

	* [r5715] avoiding an arithmetic exception with few pseudo
	  potential files..

2007-01-11 11:32  tlaino

	* [r5714] more splitting and reorganization of constraints related
	  stuff..

2007-01-10 15:53  tlaino

	* [r5713] Reorganizing intramolecular constraints. First setup of
	  intermolecular constraints.
	  Next commit will provide full availability to intermolecular
	  constraints.
	  Resetting one regtest due to numerical (10^-12) (mostly due to
	  the algebraic reorganization
	  of the existing constraints modules)

2007-01-10 08:07  vondele

	* [r5712] init

2007-01-08 11:21  jgh

	* [r5711] [Bug #9665] allow for FFTs larger than 1024. Default FFT
	  (FFTSG) still limited to 1024

2007-01-07 12:12  tlaino

	* [r5710] in case a degenerate torsion (three atoms aligned) is
	  defined skip the energy contribution (singular).

2007-01-05 09:55  tlaino

	* [r5709] Bug fix for restraint fixed atoms

2007-01-04 21:46  tlaino

	* [r5708] [Bug#9859] RESTART inconsistent with keyword UNITS.
	  FIXED.

2007-01-04 20:33  tlaino

	* [r5707] bug fix restart restraint position

2007-01-04 19:18  ikuo

	* [r5706] missing openmp defines

2007-01-04 09:39  vondele

	* [r5705] prettify

2007-01-03 20:17  vondele

	* [r5702] MD with ghost atoms (?)

2007-01-03 20:08  vondele

	* [r5701] Happy New Year !

2006-12-27 23:44  tlaino

	* [r5700] fastest NEB dump restart

2006-12-25 17:40  tlaino

	* [r5699] more timings..

2006-12-20 23:28  tlaino

	* [r5698] bug fix for intramolecular forcefield bond/bends/ub/etc..
	  bug pointed by Toon Verstraelen.
	  Added one more regtest.

2006-12-20 15:56  tlaino

	* [r5697] fixing a problem with molnames in PDB where the molname
	  field is empty.

2006-12-20 14:32  tlaino

	* [r5696] Bug fix when using/reading UB (thanks to Toon
	  Verstraelen). The provided patched by Toon
	  has been applied with few changes. Added also the possibility to
	  use different
	  potential KIND for UB. Added two more regtests.

2006-12-20 13:08  tlaino

	* [r5695] Bugfix for nonbonded14 when using GENPOT. The bug was
	  pointed out by Toon Verstraelen.
	  added a new regtest.

2006-12-19 23:56  tlaino

	* [r5694] fixing recursive IO for nag compiler

2006-12-19 16:18  tlaino

	* [r5693] printing information for restraints. cosmetics on dumped
	  info..

2006-12-19 09:51  tlaino

	* [r5692] just cosmetics and a small bug fix

2006-12-18 16:58  tlaino

	* [r5691] bug fix when using colvar,g3x3,g4x6 constraints together
	  with fixed_atoms constraints.
	  Fixed and added also 3 more regtests.

2006-12-17 17:01  tlaino

	* [r5687] [Bug #9719] restarting a classical md with restraints.
	  Bug fixed and fixed concurrently
	  also another bug for fixed atoms constraints. The bug was
	  implicit also for colvar restraints. Fixed that one too. Added
	  few more regtests. Resetting one.

2006-12-17 11:10  vondele

	* [r5686] added timing

2006-12-17 09:44  vondele

	* [r5685] test

2006-12-16 19:05  vondele

	* [r5683] change case to avoid recompiles

2006-12-16 16:50  tlaino

	* [r5682] splitting teh print of the initial coordinates for
	  replicas in neb with another
	  logical keyword.. useful for very large systems (default is not
	  print information
	  about coordinates of starting replica)

2006-12-16 15:16  tlaino

	* [r5681] allowing the external control on neb

2006-12-16 12:21  tlaino

	* [r5680] Changing the format of few fields in the NEB IO.

2006-12-16 11:33  tlaino

	* [r5679] just few print statements

2006-12-15 16:40  vondele

	* [r5678] bug fix

2006-12-15 15:00  vondele

	* [r5677] remove binding label

2006-12-15 14:37  vondele

	* [r5673] Manuel Guidon: added Hartree-Fock exchange (initial
	  version), requires libint

2006-12-13 13:57  vondele

	* [r5671] fix needed to build on NEC
	  (Francois-Xavier.Coudert@lcp.u-psud)

2006-12-13 00:58  tlaino

	* [r5667] [Bug #9704] Ghost atom and localization doesn't mix well.
	  Fixed.
	  You can avoid to specify the POTENTIAL for ghost atom but in case
	  it is a GPW calculation you've to force
	  the GPW_TYPE in the KIND section. Bug fix in the documentation.

2006-12-12 22:00  vondele

	* [r5666] bug fix

2006-12-12 09:38  tlaino

	* [r5665] bug fix

2006-12-12 01:02  tlaino

	* [r5664] introduced the possibility to have constraints applied at
	  the zero step of an MD.
	  this will avoid the jump in the energy between the zero and the
	  first step. default is
	  to apply constraints only at the 1st md step.

2006-12-11 14:22  tlaino

	* [r5663] [Bug #9682] constraint during a QMMM not defined on QM
	  molecules. Fixed.
	  provided a new regtest.

2006-12-11 12:24  krack

	* [r5662] copy alpha to beta MOs for ROKS/OT/atomic

2006-12-10 20:05  tlaino

	* [r5661] fixing a bug in derivatives using sic with decoupling and
	  recoupling.. only derivatives affected..

2006-12-10 18:33  tlaino

	* [r5660] fixing proliferation of localLog in parallel..

2006-12-10 14:28  tlaino

	* [r5659] speeding up one regtest and modifying the optimizer in
	  order to perform the LS, if a new collection of DIIS
	  point is requested.

2006-12-09 21:20  marcella

	* [r5658] bug fix

2006-12-09 20:37  tlaino

	* [r5657] Improving the geometry optimizer for NEB. resetting few
	  regtests.

2006-12-09 10:10  tlaino

	* [r5655] Handling better "bad" starting geometry in NEB

2006-12-09 07:05  tlaino

	* [r5654] first release of different implementations of NEB
	  available. Added regtests.
	  Correcting few other places where error was still optional.
	  prettifying all submitted files.

2006-12-08 23:27  krack

	* [r5652] enable atomic guess for ROKS/OT

2006-12-08 19:41  marcella

	* [r5651] modified print level for kg_gpw moments

2006-12-08 14:25  tlaino

	* [r5649] bug fix in format and for undefined charges (not always
	  triggered)

2006-12-08 12:29  tlaino

	* [r5648] [Bug #9656]. Bug fixed for DDAP charges. Introduced on
	  revision 1.15. Fixed.
	  resetted 11 regtests.

2006-12-08 10:01  vondele

	* [r5647] fix one bug in the ionic forces with sic & ddapc

2006-12-07 19:16  mcgrath

	* [r5646] Bug fix for swap moves with multiple molecule types

2006-12-07 16:28  vondele

	* [r5645] fix a bug is ddapc spin rest

2006-12-07 14:09  vondele

	* [r5644] even more

2006-12-07 14:08  vondele

	* [r5643] clarify manual

2006-12-06 15:00  hforbert

	* [r5642] Changed ALL functions with side effects to subroutines in
	  pint_types.F,
	  though FUNCTIONs with side-effects are not unusual for cp2k (see
	  for example
	  the recursive I/O complaint in the previous commit for this
	  file.)
	  Maybe we should use the keyword PURE with pure functions more
	  often...
	  [Remark: POINTER arguments never seem to be able to have INTENT
	  anything
	  in Fortran 90/95, and in the functions in question they would
	  have been
	  intent(inout)]

2006-12-06 12:46  krack

	* [r5639] work-around for NAG compiler to avoid crash of the TDDFT
	  regtests

2006-12-06 10:24  vondele

	* [r5635] switch to yahoo search. Yahoo seems to index our input
	  pages, whereas google does not

2006-12-06 09:59  tlaino

	* [r5634] [Bug #9639]: DCD output conflicts with npt. Fixed.

2006-12-06 09:20  tlaino

	* [r5632] documentation bug fix (thanks to Toon V.)
	  added a cleaning script to the Pimd regtest directory

2006-12-06 09:10  krack

	* [r5631] fixes w.r.t. the complains of the NAG compiler:
	  There were several issues in pint_types to worry about:
	  
	  1) misusage of FUNCTION in Fortran 90/95:
	  Changing the components of an INTENT(IN)-POINTER to a FUNCTION is
	  possible/permissible, but it is also a very bad programming
	  style,
	  since it might cause most-likely side-effects which are difficult
	  to trace.
	  
	  2) recursive I/O due to a FUNCTION call in the parameter list of
	  a
	  WRITE statement
	  
	  3) missing initializations: pint_env%v_tol and pint_env%randomG

2006-12-05 18:39  hforbert

	* [r5630] Fix memory leaks

2006-12-05 17:54  tlaino

	* [r5629] bug fix in comparing strings

2006-12-05 17:20  hforbert

	* [r5628] g95 is a lot pickier than my compiler i see....
	  Let's see if this fix works (and I'll try to install g95 here so
	  i
	  can test locally with it)

2006-12-05 15:14  hforbert

	* [r5627] Bug fix (?) for output in parallel
	  Added simple regtest for path integrals
	  (might not work with more than 4 cpus atm)
	  [again these changes were tested with an intel compiler, I hope
	  I just get unlucky last time with the g95 not liking it...]

2006-12-05 14:49  tlaino

	* [r5626] bug fix (thanks to Toon Verstraelen).

2006-12-05 14:28  tlaino

	* [r5625] by-passing a bug in PGI compiler. A wish for christmas:"a
	  bugfree compiler to
	  use on the CRAY". Thanks.

2006-12-05 09:47  vondele

	* [r5624] fix intent (Vittorio)

2006-12-05 09:37  vondele

	* [r5623] init

2006-12-04 21:58  mcgrath

	* [r5622] Parallel bug...thanks Chris

2006-12-04 18:54  vondele

	* [r5621] test

2006-12-04 16:01  vondele

	* [r5620] fix build

2006-12-04 15:35  hforbert

	* [r5619] First round of Pathintegral MD changes. (mostly got
	  Fawzi's stuff working,
	  added normal modes and virial kinetic energy estimator. at the
	  moment still
	  lightly tested and I planned to arrange things later but they
	  have now
	  (minimal?) functionality. I will check in a regtest soon.

2006-12-01 11:58  vondele

	* [r5618] bug fix

2006-12-01 10:47  tlaino

	* [r5617] bug fix for DCD format when restarting MD

2006-12-01 08:11  tlaino

	* [r5616] bug fix (openMP) [thanks to Francois-Xavier Coudert]: Fix
	  OpenMP build; this patch adds back the necessary OpenMP
	  directives to make it build again

2006-11-30 19:54  tlaino

	* [r5615] supporting cell info in the DCD format (thanks Axel for
	  the proper format!)

2006-11-30 16:45  tlaino

	* [r5614] fixing a print statement

2006-11-29 23:55  tlaino

	* [r5611] bug fixes and added more checks..

2006-11-29 21:39  mcgrath

	* [r5609] Some changes for starting a simulation with no molecules
	  in one of the boxes

2006-11-29 17:24  tlaino

	* [r5608] improving further the parallelization scheme of qm/mm
	  runs.. resetted few regtests
	  for numerics.. added a regtest to keep under control the old
	  parallelization scheme

2006-11-29 10:39  tkuehne

	* [r5607] Bug fix + minor change

2006-11-29 10:10  tlaino

	* [r5606] just cosmetics and few more timings

2006-11-28 17:31  tlaino

	* [r5604] New parallelization scheme for QM/MM.. switched default
	  to the new scheme and
	  resetting few regtests for numerical errors..

2006-11-28 13:04  tlaino

	* [r5603] bug fix

2006-11-27 00:51  vondele

	* [r5602] remove quadratic input reading stuff, most useful for
	  millions of atoms. Can the MacBook have a bit more memory please?
	  Also, battery life is short while regtesting ;-)

2006-11-24 19:54  vondele

	* [r5601] cleanup: remove unused module

2006-11-24 15:44  tkuehne

	* [r5600] Move code into a subroutine

2006-11-24 14:58  mcgrath

	* [r5599] Makes sense to write the step number with the coordinates

2006-11-24 05:37  tlaino

	* [r5597] fixing memory leaks

2006-11-24 01:10  tlaino

	* [r5596] qm/mm with semiempirical.. a the moment working only the
	  mechanical coupling..
	  added two more regtests for qm/mm semiempirical..

2006-11-23 15:30  tlaino

	* [r5595] bug fix undefined variable.. bug fix for qmmm_env
	  (optional) not checked if present:
	  mm_subsys present in the constraint section in a FIST run.. now
	  simply ignores this keyword

2006-11-23 12:42  tkuehne

	* [r5593] More robust history restart

2006-11-22 21:00  mcgrath

	* [r5592] More error messages.

2006-11-22 14:33  jgh

	* [r5591] Add gyromagnetic info to periodic table
	  Routine to calculate 1/G^2 * d^2F(G)/dG^2

2006-11-22 14:32  jgh

	* [r5590] Check for atan2(0,0) (Reinout Declerck)

2006-11-22 12:48  krack

	* [r5587] substitute special tokens also in the section description

2006-11-22 00:29  ikuo

	* [r5585] fix mol_set nonfo error

2006-11-21 21:38  mcgrath

	* [r5583] A few minor changes, and some more helpful error messages

2006-11-21 13:25  tlaino

	* [r5581] being just slightly more clever..

2006-11-20 12:19  tkuehne

	* [r5577] Reinserting lost natom

2006-11-20 11:18  tlaino

	* [r5576] undefined variable (Vittorio) and related bug fix..
	  regtest ok..

2006-11-20 08:12  vondele

	* [r5575] bug fixing 4 undefined variables (thanks Vittorio)

2006-11-18 02:10  tlaino

	* [r5573] major rewriting of the DDAPC module. now it is possible
	  to use at the same
	  time different methods based on ddapc. introduced also the
	  possibility to choose
	  the spin density fitting instead of the full density. Few
	  regtests were resetted due to
	  numerics. One new regtest on the spin density...

2006-11-17 17:01  tlaino

	* [r5572] So... since we choose to go for the machine dependent
	  (NC) "highly OS indipendent" move file routine
	  why not DELETE the old move_file? With a combination of grep and
	  vi
	  takes only 3 minutes.. or.. and I have no problem to admit my
	  limitations...
	  is there some logic behind the need to have around doubled
	  routines with the same
	  functionality?
	  If YES I really want to hear this logic..

2006-11-17 13:53  marcella

	* [r5570] this should make kg_gpw more stable in parallel

2006-11-17 13:45  tkuehne

	* [r5569] NC

2006-11-16 23:28  mcgrath

	* [r5568] Left a debugging flag in

2006-11-16 20:09  ikuo

	* [r5567] Update for GROMOS. Now use off diagnol parameters for
	  nonbond. [M.Baer]

2006-11-16 17:59  vondele

	* [r5566] test

2006-11-16 17:36  mcgrath

	* [r5565] A bug for restarting GEMC with an empty box

2006-11-16 15:25  vondele

	* [r5564] new CG defaults (reset a number of regtests)

2006-11-16 14:10  vondele

	* [r5563] fix format (Vittorio)

2006-11-16 11:13  vondele

	* [r5561] clarify

2006-11-15 20:43  mcgrath

	* [r5560] Print out some info consitently

2006-11-15 15:48  vondele

	* [r5559] spaces

2006-11-14 21:46  tlaino

	* [r5557] bug fix for SIC when using DDAPC for decoupling..
	  changing the definition of ch_dens in DDAPC and
	  resetting 10 regtests due to numerics (10^-9 - 10^-14)...

2006-11-14 17:56  marcella

	* [r5555] bug fix in the multiple grids option with kg_gpw

2006-11-14 16:38  vondele

	* [r5553] enhanced flexibility for fixed_atoms mm_subsys/qm_subsys
	  (teo)

2006-11-14 14:56  marcella

	* [r5552] bug fix in the multiple grids option with kg_gpw

2006-11-14 14:25  vondele

	* [r5551] print imomm atom coordinates if first printout, tighter
	  check on PSF format (Teo)

2006-11-14 12:46  vondele

	* [r5550] clarify output

2006-11-13 17:51  vondele

	* [r5549] Attempt to clarify the units

2006-11-13 17:45  vondele

	* [r5548] improve docs

2006-11-11 00:13  ikuo

	* [r5547] Unit conversion for simpar%p_ext

2006-11-10 11:22  marcella

	* [r5544] should fix the new leaks

2006-11-10 07:28  marcella

	* [r5542] bug fix and possibility to have charged molecules in
	  kg_gpw

2006-11-09 18:27  tlaino

	* [r5541] moving last tbmd file to the attic..

2006-11-09 13:13  tkuehne

	* [r5538] OBJECTDEFS cp2k_runs.F environment.F f77_interface.F
	  force_env_methods.F force_env_types.F input_constants.F
	  input_cp2k.F : Remove TBMD
	  
	  harris_metropolis_mc.F input_cp2k_tbmd.F tbmd_energy.F
	  tbmd_energy_types.F tbmd_environment.F
	  tbmd_environment_methods.F tbmd_force.F tbmd_force_types.F
	  tbmd_hamiltonian.F tbmd_main.F tbmd_model_types.F
	  tbmd_neighbor_list_env_types.F tbmd_neighbor_list_types.F
	  tbmd_neighbor_lists.F : Move to Attic

2006-11-08 16:26  tkuehne

	* [r5536] OBJECTDEFS cp2k_runs.F cp_control_utils.F
	  input_constants.F
	  input_cp2k_check.F input_cp2k_dft.F input_cp2k_motion.F
	  input_cp2k_restarts.F md.F scf_control_types.F: Remove HMC
	  qs_wf_history_methods.F: ASPC start-up with outer-loop

2006-11-08 12:25  tlaino

	* [r5535] deleting unimplemented keywords

2006-11-08 11:26  tlaino

	* [r5534] adding a shortcut NONE to the XC_FUNCTIONAL.. added a
	  regtest.. xc section and xc_functional section now must be always
	  provided (not for semiempirical run).

2006-11-07 22:28  mcgrath

	* [r5533] bug and an extra write statement

2006-11-07 19:42  mcgrath

	* [r5532] Caught a bug...input doesn't like empty force_envs, but
	  sometimes we get them in MC, so I have to work around it

2006-11-07 19:03  tlaino

	* [r5530] Adding the possibility to specify molecule name in the
	  constraint section
	  instead of molecule index. Added the same possibility also to the
	  HBONDS section.
	  All constraints are now equivalent (from a keyword point of
	  view)..
	  Added 4 new regtests..

2006-11-06 23:07  mcgrath

	* [r5529] Whoops..forgot a file

2006-11-06 22:54  mcgrath

	* [r5528] Some more error checking and changes to allow GEMC_NpT

2006-11-06 18:00  mcgrath

	* [r5525] Forgot the file with the keywords.

2006-11-06 14:49  tlaino

	* [r5524] Enabling the possibility don't have a XC section (if not
	  defined the section, no XC functional
	  used)... ;-)

2006-11-05 21:10  mcgrath

	* [r5523] Major MC changes...should now be able to do multiple
	  molecule types...still testing for a few more weeks

2006-11-05 16:28  mcgrath

	* [r5521] Bug fix...temporary, since this week I hope to commit a
	  major change that will obviate the need for this routine.

2006-11-03 10:09  vondele

	* [r5518] allow for printing the forces for all methods in a
	  consistent way

2006-11-02 17:33  ikuo

	* [r5517] Add missing case npt_i for QS

2006-11-02 14:11  vondele

	* [r5516] move to low print level

2006-11-02 14:05  vondele

	* [r5515] bug fix

2006-11-02 10:20  fschiff

	* [r5514] added a printkey MIXED_ENERGIES to print the potential
	  energies of the two SUB_FORCE_EVALS

2006-11-02 09:41  vondele

	* [r5509] move detailed energy to the high print level

2006-11-01 14:27  fschiff

	* [r5508] Added a new METHOD to calculate MD simulations with a
	  linear combination of two FORCE_EVAL

2006-10-31 08:12  tlaino

	* [r5507] my 2c to contain the entropy of the universe

2006-10-30 16:29  tkuehne

	* [r5505] Write the wavefunction history into files

2006-10-30 09:12  vondele

	* [r5504] improved full_single_inverse preconditioner (doesn't need
	  diagonalisation, but is fairly efficient)

2006-10-27 23:09  tlaino

	* [r5503] bug fix + an additional regtest (testing the bug fix)

2006-10-27 18:05  tlaino

	* [r5502] Introducing a general potential functional form for
	  nonbonded. ANY kind of
	  potential can now be used inside cp2k providing the analytical
	  expression
	  in the input file. 2 more regtests added that give exactly the
	  same result
	  as two other present regtest. Enjoy!

2006-10-27 07:58  vondele

	* [r5501] do not re-add m_walltime, it is not needed anymore

2006-10-27 07:51  vondele

	* [r5500] update: REQUIRES UPDATED ARCH FILES... (forpar madness
	  bit). Now uses a very quick and accurate way to measure the total
	  memory allocated by CP2K (accessing a g95 internal variable using
	  the ISO_C_BINDING stuff). Most importantly, this allows for using
	  GLOBAL%TRACE without slowing down program execution. The module
	  variable total_memory can also be read at any point to know
	  exactly how many bytes are allocated at a given point.

2006-10-26 14:04  krack

	* [r5498] fix for IBM 10.x compiler (thanks to Thomas Kuehne).
	  Please, avoid such
	  awkward write statements in the future (if clause corrected).

2006-10-26 13:53  krack

	* [r5497] fix for IBM 10.x compiler (thanks to Thomas Kuehne).
	  Please, avoid such
	  awkward write statements in the future.

2006-10-26 13:23  krack

	* [r5496] fix: missing initialization of mass_sum (requested by NAG
	  compiler, 158 regtests failed last night)

2006-10-26 05:12  vondele

	* [r5495] workaround a NAG 5.1 optimizer bug (Vittorio using prism
	  theory).

2006-10-26 05:11  vondele

	* [r5494] input enhancement

2006-10-25 23:35  ikuo

	* [r5493] set mass and charge for molecule_kind

2006-10-25 17:22  mcgrath

	* [r5489] And a change in the source to make MOL_SET work for MC

2006-10-25 12:46  marcella

	* [r5486] bug fix on the potential form

2006-10-25 06:00  ikuo

	* [r5485] fix mol_set

2006-10-24 20:55  tlaino

	* [r5484] some cleaning around the potential energy functions in
	  FIST.. specifying the analytical
	  form in the autogenerated reference input manual (PLEASE Always
	  keep it up2date!)..

2006-10-24 17:41  tlaino

	* [r5483] few bug fixes and improved features in damping PSF
	  information..
	  more tight control on masses and charges (for the classical part
	  of the code)
	  Possibility to read masses (for fist) also from the KIND section
	  (ignored before)..
	  regtest ok.

2006-10-23 13:21  marcella

	* [r5482] Two potentials for ionic materials

2006-10-23 08:21  vondele

	* [r5481] cleaner dealloc (Vittorio)

2006-10-23 08:09  vondele

	* [r5480] better choice in case of error

2006-10-19 08:29  tkuehne

	* [r5479] qs_wf_history_methods.F, scf_control_types.F,
	  harris_metropolis_mc.F, input_cp2k_dft.F: MAX/EPS_SCF_HIST
	  input_cp2k_motion.F, itegrator.F, md.F: Shadow-Gamma
	  qs_scf.F: Misc

2006-10-18 11:26  tlaino

	* [r5477] bug fix

2006-10-17 12:17  tlaino

	* [r5476] bug fix in output message

2006-10-17 05:41  ikuo

	* [r5475] analytical pressure for GGA (Jochen Schmidt)

2006-10-16 12:04  tlaino

	* [r5474] reverting to the original preprocessing instruction

2006-10-15 22:28  tlaino

	* [r5473] -)cleaned the element definition.. added a list of most
	  common amber and charmm kinds
	  with the corresponding element definition (in most cases now the
	  element specification in the pdb
	  is not needed anymore).
	  -)bug fix for qm/mm gapw due to a problem in the definition of
	  the element
	  -)added a new regtest for QS
	  -)some cleaning around

2006-10-15 19:37  mcgrath

	* [r5472] Had to change something for the PSF changed that occured
	  this week.

2006-10-13 17:37  fawzi

	* [r5471] potential (never executed) bug fixed, thanks to Jochen
	  Schmidt

2006-10-13 09:11  tlaino

	* [r5467] making uniform (everywhere but not the imported GPL
	  routine) the copyright
	  and program header at the beginning of files.. now should be easy
	  to change it
	  automatically each year..

2006-10-12 23:56  tlaino

	* [r5466] bug fix for xyz in generating molname.. making consistent
	  both XYZ and &COORD section
	  regarding the molname.. if not provided in &COORD unique names
	  are given to the molnames..
	  In this way (unless one does not provide molnames explicitly)
	  even with connectivity generate
	  we will create molecules with one atom inside.. this makes
	  everything more consistent and avoids
	  to specify connectivity off for QS in parallel runs (the result
	  will be the same)..

2006-10-12 15:15  tlaino

	* [r5464] linear scaling algorithm and bug fix for assigning
	  constraints of bonds in which
	  one of the element is hydrogen (the bug was popping up everytime
	  there was a bond
	  between two hydrogens defined (like in fast water of amber)).
	  Resetting regtest due to numerics (reorganized the order in which
	  shake is applying constraints
	  and the error is of the order of the shake tolerance 10^-6)

2006-10-11 14:30  vondele

	* [r5461] Improved FULL_ALL preconditioner, now very much
	  recommended

2006-10-11 13:24  tlaino

	* [r5460] fixing PSF format (now youcan specify UPSF (unformatted
	  read of PSF) and PSF (formatted read) with CHARMM style)
	  extension to the EXT CHARMM support.
	  added a regtest for PSF EXT

2006-10-11 13:22  tlaino

	* [r5459] this is the real bug fix..

2006-10-11 12:30  tlaino

	* [r5458] ops.. the last one was a wrong bug fix.. restoring the
	  original version..

2006-10-11 11:23  tlaino

	* [r5457] bug fix: uninitialized variables

2006-10-10 11:58  tlaino

	* [r5454] bug fix in the topology generation. Added regtest.

2006-10-09 14:05  vondele

	* [r5453] do not check by default

2006-10-09 13:54  vondele

	* [r5452] improved numerical stability of FULL_ALL preconditioner

2006-10-08 18:58  krack

	* [r5450] mp_sum call removed (as suggested by Fawzi)

2006-10-07 16:01  tlaino

	* [r5449] in a QM/MM MD run dump the temperature of the MM system
	  and of the QM system as well..

2006-10-06 19:42  mcgrath

	* [r5448] Bug fix...not sure why it didn't show up earlier.

2006-10-06 14:28  tlaino

	* [r5447] new reordering algorithm (to activate it use REORDER in
	  GENERATE section). i
	  Now we can handle scrambled (but really scrambled PDB)
	  reorder them, match similar kind of molecules (on the basis of
	  the connectivity only
	  , built starting from coordinates) and pack everything in a very
	  efficient way.
	  The matching could fail only for very high symmetric molecular
	  system.. in that case
	  two different molecules will be created instead of one.. added
	  few more regtests..
	  
	  Last: bug fix in the dipole of FIST.

2006-10-05 09:12  tlaino

	* [r5446] bug fix in the documentation

2006-10-05 09:09  tlaino

	* [r5445] bug fix in the documentation

2006-10-04 15:48  krack

	* [r5444] constraints for Langevin dynamics, optionally printout of
	  structure data during MD

2006-09-29 12:05  jgh

	* [r5441] Correct sign of force contributions

2006-09-27 18:46  vondele

	* [r5440] bug fix

2006-09-27 18:24  vondele

	* [r5439] bug fix (otherwise ranks do not match)

2006-09-27 14:14  vondele

	* [r5438] find equivalent molecules

2006-09-27 14:09  krack

	* [r5437] improve printout

2006-09-26 07:44  vondele

	* [r5434] added mp_xxxxxx timings in message_passing. Can be
	  controlled with __mp_timeset__ (see message_passing.F)

2006-09-25 15:42  tlaino

	* [r5433] added the possibility to have frozen atoms in the
	  vibrational analysis module.
	  Specify FIXED_ATOM as usual through the constraint section.
	  added a new regtest.

2006-09-25 15:40  tlaino

	* [r5432] fixing I/O in parallel

2006-09-25 10:55  tlaino

	* [r5431] fixing output in parallel

2006-09-25 08:51  vondele

	* [r5430] some refactoring and documentation

2006-09-25 08:07  tkuehne

	* [r5429] Noisy Gamma

2006-09-24 12:53  tlaino

	* [r5428] print energy in force_eval only if forces are not
	  required (on the contrary the total energy
	  should already be printed).

2006-09-24 12:41  tlaino

	* [r5427] print energy in force_env after all terms have been
	  computed!

2006-09-24 12:01  tlaino

	* [r5426] bug fix for format

2006-09-24 11:08  tlaino

	* [r5425] added a print statement for the energy in
	  force_env_calc_energy specifieng
	  the force_env to which that energy corresponds. fixing write
	  statements and
	  info for qm/mm and fist runs.

2006-09-23 17:20  tlaino

	* [r5424] more robust algorithm for the generation of the topology.
	  checks consistency with
	  molecule name read from PDB. Added a problematic regtest (thanks
	  Joost!) and modified
	  one. Added further checks. No changes in regtests!

2006-09-22 13:35  tlaino

	* [r5423] bug fix in integer_to_string

2006-09-22 13:35  tlaino

	* [r5422] added more comments instead of a simple abort..

2006-09-21 17:24  vondele

	* [r5421] fix uninitialised error usage

2006-09-21 16:56  tlaino

	* [r5420] adding a comment in the code regarding the reason why we
	  should not
	  dump the restart file in case of external exit

2006-09-21 16:28  vondele

	* [r5419] make error non-optional almost everywhere

2006-09-21 16:18  tlaino

	* [r5418] change name type..

2006-09-21 15:07  tlaino

	* [r5417] several bug fixes around with the generation of the
	  connectivity.. Cosmetics in topology_types.

2006-09-21 14:14  tlaino

	* [r5416] fixing bug in the format

2006-09-21 11:44  tlaino

	* [r5415] adding reference for qm/mm periodic calculations

2006-09-21 11:43  tlaino

	* [r5414] dumping more detailed info to the force field pack output
	  (if missing parameters)

2006-09-21 09:07  vondele

	* [r5413] prettify

2006-09-21 09:07  vondele

	* [r5412] try to get back to the old error behavior (try 2)

2006-09-20 22:11  ikuo

	* [r5410] Units bug for VEL file (Marcel Baer)

2006-09-20 20:27  tlaino

	* [r5409] Splitting force_fields.F file into multiple file..
	  modified the forcefield pack.
	  Whenever a forcefield parameter is not found it is stored and
	  instead of aborting
	  immediately the setup goes through all the forcefield_pack and
	  abort at the end providing
	  a list of all missing parameters (it aborts if between them
	  there's a crucial one).

2006-09-20 14:53  vondele

	* [r5407] periodic QMMM only: Fix the definition of the zero of
	  energy the classical potential, useful for eigenstates and
	  charged systems (Teo & Joost)

2006-09-20 04:58  vondele

	* [r5406] add an additional check and stop on failure

2006-09-19 18:57  vondele

	* [r5405] improve docs

2006-09-19 18:22  tlaino

	* [r5404] bug fix in the documentation

2006-09-19 16:37  tlaino

	* [r5403] New connectivity generation routines (linear scaling)..
	  faster.. more robust than older one.
	  Introduced the possibility to modify by input the generated
	  connectivity.
	  Bug fix in the neighbour_lists of fist for non-periodic case.
	  cleaned the particle_types routine.
	  created a new section in the subsys section called GENERATE that
	  contains all
	  possible parameters influencing the generation of the
	  connectivity.
	  
	  Added few more regtests for the new connectivity generator.

2006-09-19 07:23  vondele

	* [r5402] improve the cross-linking of the html manual, hoping to
	  get it more google friendly.

2006-09-16 18:12  fawzi

	* [r5396] * added shake to force_env and f77 interface
	  * f77_int.h with interface declarations for f77_in_low
	  * cpot runs
	  * replica bugfix with less procs than replicas
	  * ep bugfixes

2006-09-16 18:03  vondele

	* [r5395] Fix incredibly slow write (many minutes for just 60
	  atoms) of the Hessian, which showed up only for some filesystems
	  (e.g. Lustre, but not the standard Linux ones). This changes the
	  format of the anyway binary HESSIAN file. Thanks to Daniele
	  Passerone for reporting the issue.

2006-09-16 17:11  vondele

	* [r5394] at least write the checkpoint, even if mem is not
	  available

2006-09-16 17:09  vondele

	* [r5393] added more timings and header

2006-09-16 16:38  krack

	* [r5392] Avoid boundary check for zero-size arrays

2006-09-15 17:29  vondele

	* [r5391] just comments

2006-09-15 10:42  tlaino

	* [r5390] splitting topology_generate_* into another file.. added a
	  key to dump info
	  during topology generation (generate_info)

2006-09-15 09:22  marcella

	* [r5389] XAS restart and parallel KG_GPW with pauli potential

2006-09-15 07:16  tlaino

	* [r5388] bug fix -> check bond control

2006-09-14 22:08  tlaino

	* [r5387] Topology connectivity NOW really linear scaling. Splitted
	  few routines to avoid
	  messiness in one single file..
	  Topology connectivity pack completely rewritten.. used a more
	  robust, fast (linear
	  scaling) algorithm. Bug fix in torsion (the old
	  topology_connectivity_pack had a bug
	  in the torsion/improper section.. and potentially also for angle)
	  -> reset one regtest..

2006-09-13 12:07  vondele

	* [r5386] moved the abort to the serial bit of code

2006-09-13 12:03  vondele

	* [r5385] added a flush after the end

2006-09-13 12:00  vondele

	* [r5384] remove unused var

2006-09-13 09:19  vondele

	* [r5383] farming: added group idle times

2006-09-11 17:56  vondele

	* [r5382] small speedup

2006-09-11 11:55  vondele

	* [r5381] speedup topology_coordinate_pack (x3.5 on big.inp)

2006-09-11 11:22  vondele

	* [r5380] fix parallel regression

2006-09-11 09:21  vondele

	* [r5379] small speedup

2006-09-11 08:56  vondele

	* [r5378] speedup: do not update particle positions if no restraint
	  has been used

2006-09-11 08:20  vondele

	* [r5377] speed up of nonbonded forces (15%)

2006-09-10 08:57  tlaino

	* [r5376] Removing cell variable from write_particle_coordinates..
	  keeping up2date the template file

2006-09-09 17:48  vondele

	* [r5375] added a string table

2006-09-08 17:46  vondele

	* [r5374] prepare for large systems ;-)

2006-09-08 14:14  vondele

	* [r5373] improve docs

2006-09-07 14:50  vondele

	* [r5372] add store logical to the print section

2006-09-07 08:04  vondele

	* [r5371] fixes some string lengths (Vittorio Zecca)

2006-09-07 07:17  vondele

	* [r5370] fix printing

2006-09-07 06:25  vondele

	* [r5369] test

2006-09-06 14:31  tlaino

	* [r5368] Correcting restarts for GEOi_OPT: no dumping of velocity
	  section (this may cause troubles when
	  using this restart file (from geo_opt) for an MD run).. If a
	  velocity section is provided
	  in a GEO_OPT run it is cleaned and removed (velocities not used
	  for GEO_OPT).
	  Resetting one regtest (showing this bad behavior) ..

2006-09-06 09:24  tlaino

	* [r5367] avoiding quadratic loop in the settings of the neighbor
	  lists

2006-09-05 19:26  vondele

	* [r5366] small speedup

2006-09-05 14:50  vondele

	* [r5365] speedup: avoid a quadratic reallocate

2006-09-05 14:03  vondele

	* [r5364] speedup parser (25% on fist_init for JAC)

2006-09-05 12:28  vondele

	* [r5363] speedup spline generation (4x), reset 4 tests due to
	  numerics

2006-09-05 08:36  vondele

	* [r5362] speed up fist_init by 30%

2006-09-04 13:08  vondele

	* [r5361] tiny speedup

2006-09-04 11:36  vondele

	* [r5360] speedup (2x) + fixes precision bug (real*4->real*8: 3
	  reset regtests)

2006-09-04 08:46  vondele

	* [r5359] speedup qmmm_forces (2x)

2006-09-03 16:48  vondele

	* [r5358] speed up ewald_ddapc (10x)

2006-09-03 13:00  vondele

	* [r5357] speed up by a bit (up to 20% ?)

2006-09-03 07:47  vondele

	* [r5356] last valgrind error

2006-09-01 14:56  tlaino

	* [r5355] bug fix for qm/mm runs.. undefined jump with valgrind->
	  fixed

2006-09-01 13:11  vondele

	* [r5354] speed up ddapc forces (30x) (changes numerics of 11
	  regtests)

2006-09-01 12:35  tlaino

	* [r5353] bug fix: great!! the parallel regtester works!! ;-)

2006-09-01 09:55  tlaino

	* [r5352] just cosmetics

2006-08-31 16:50  vondele

	* [r5351] yet another test

2006-08-31 15:45  vondele

	* [r5350] test

2006-08-31 14:45  vondele

	* [r5349] test

2006-08-31 13:10  tlaino

	* [r5347] Using kahan sums for the DDAPC charges.. resetting 13
	  regtests due to changes in
	  precision number

2006-08-31 10:22  tlaino

	* [r5346] moving the mathematical routines from util in mathlib.
	  cleaning util.
	  cleaning cp_ddapc_methods and introducing the possibility to ask
	  for
	  the condition number of the A matrix used for the bloechl
	  charges.

2006-08-30 13:31  tlaino

	* [r5344] added the possibility to print also the electron-electron
	  coulomb energy
	  in a gpw runif the detailed_energy keyword is enabled.

2006-08-30 11:03  tlaino

	* [r5343] restoring NDIGITS keyword for printing QS forces.. added
	  a comment on numerical
	  precision of results using replica_env in parallel

2006-08-29 23:14  tlaino

	* [r5342] bug fix fixed atoms in parallel... bug fix restraints in
	  parallel... I will reset old results for parallel execution
	  of restraints regtests after the first automatic regtester..
	  deleting a write statement from the force_fields.F file..

2006-08-29 19:52  vondele

	* [r5340] fix bug in fixed_atoms constraint in parallel for geoopt
	  (reset 4 regtests)

2006-08-29 18:57  vondele

	* [r5339] fix parallel farming issue

2006-08-29 13:57  vondele

	* [r5337] added a few timings

2006-08-29 13:30  tlaino

	* [r5336] vibrational analysis works also in parallel now using
	  replica_env.. GREAT!

2006-08-29 11:46  vondele

	* [r5335] fix init bug

2006-08-29 11:01  tlaino

	* [r5334] removing remaining leaks in MC regtests

2006-08-29 08:49  vondele

	* [r5333] some further globenv cleaning

2006-08-29 07:50  vondele

	* [r5332] root_section to lowercase

2006-08-28 18:57  vondele

	* [r5331] further globenv removal

2006-08-28 18:37  tlaino

	* [r5330] VIBRATIONAL ANALYSIS module, to perform frequency and
	  normal modes analysis. three new regtests.
	  bug fix in qmmm_topology_util.F due to the removal of scr from
	  globenv
	  Fixed memory leaks in EP.

2006-08-28 14:59  vondele

	* [r5329] remove globenv%scr

2006-08-28 07:33  jgh

	* [r5327] Remove memory leaks

2006-08-27 14:08  jgh

	* [r5326] Fixing KG_GPW input for non-GGA functionals

2006-08-25 15:54  vondele

	* [r5325] towards cleaning scr from globenv

2006-08-25 14:16  vondele

	* [r5324] remove para_env from globenv

2006-08-24 19:29  vondele

	* [r5323] removes input_file (root_section) from globenv.
	  Introduces some as yet unresolved memory leaks in MC and EP.

2006-08-24 06:05  marcella

	* [r5322] checking integrals for chi

2006-08-23 16:32  vondele

	* [r5321] towards removing para_env from globenv

2006-08-23 13:57  tlaino

	* [r5320] Implicit solvation model: general multipole Onsager
	  model. Added two regtests.
	  few modification around for DDAPC charges. and few modifications
	  for the debug
	  module. regtest ok.

2006-08-23 10:22  vondele

	* [r5319] Get rid of group,mepos,num_pe,source,ionode in globenv,
	  use para_env instead

2006-08-23 07:05  vondele

	* [r5318] document the FFT libs

2006-08-23 06:58  vondele

	* [r5317] recover printing physcon at medium level

2006-08-23 06:45  vondele

	* [r5316] remove input_file_name from globenv, add &STORE to the
	  print section

2006-08-21 23:04  ikuo

	* [r5315] Bug fix

2006-08-21 08:27  jgh

	* [r5313] Off molecule orthogonalization potential for KG_GPW
	  method
	  one new regtest

2006-08-20 10:36  vondele

	* [r5312] reprettify

2006-08-18 16:23  tlaino

	* [r5311] new banners for qm/mm and tddft avoiding the problem of
	  the cpp..

2006-08-18 16:00  marcella

	* [r5310] ...hunting for bugs

2006-08-18 15:42  vondele

	* [r5309] prettify

2006-08-18 13:12  tlaino

	* [r5307] Moved the previous implementation of RESTRAINT to one
	  using COLVAR.. restraints are
	  defined like contraints (see new regtests based on restraints).
	  Added several regtests
	  testing the different kind of restraints.
	  New definition for torsional angle that avoids the singularity at
	  pi or -pi.
	  Due to this new defition 2 regtests resetted.
	  Also the 2 regtests using restraints have been resetted due to
	  the different implementation.

2006-08-17 08:05  tlaino

	* [r5306] bug fix.. undefined general unit..

2006-08-16 22:52  tlaino

	* [r5305] allowing for a generic unit..

2006-08-14 15:39  tlaino

	* [r5303] Moving CONSTRAINTS section into MOTION section..
	  converting corresponding regtests.. no changes in results.
	  Ready to convert RESTRAINTS with the same structure of
	  CONSTRAINTS.

2006-08-14 09:57  tlaino

	* [r5301] Bug fix for the construction of the neighbor lists: this
	  was an old bug present
	  in the old implementation. We need to skip the the excluded atoms
	  from the neighbor
	  pairs only if they belong to the same image cell, otherwise they
	  will be a neighbor pair..
	  Resetted 24 regtests (errors in the range 10^-3 - 10^-8)...

2006-08-13 15:55  vondele

	* [r5300] remove double use (it is time to prettify the code again)

2006-08-12 19:17  vondele

	* [r5299] small speedup

2006-08-12 14:22  jgh

	* [r5298] Fix bug in parallel execution

2006-08-12 06:49  vondele

	* [r5297] less mystery and more speed ?!

2006-08-11 17:16  tlaino

	* [r5296] New algorithm for FIST neighbor lists.. Memory lighter
	  and faster. Adding JAC input file
	  to the regtest. Parallel version ok. Regtests OK!

2006-08-11 15:33  marcella

	* [r5295] on the way towards NMR, plus some cleaning

2006-08-11 08:37  fawzi

	* [r5293] plugging leaks

2006-08-10 20:28  fawzi

	* [r5292] bugfix mc_kg some leaks, recursive f_env dealloc (ep).

2006-08-10 19:12  fawzi

	* [r5291] adding forgotten eip and tbmd in f77_interface

2006-08-10 18:37  fawzi

	* [r5290] bugfix: ensure globenv presence in finalize

2006-08-10 17:54  fawzi

	* [r5289] fixing f77 interface, using it also in cp2k_runs.

2006-08-09 12:57  fschiff

	* [r5288] bug fix for nve_respa integrator (parallel version)

2006-08-09 07:38  fschiff

	* [r5287] bug fix respa

2006-08-08 17:51  vondele

	* [r5284] test

2006-08-08 10:50  fschiff

	* [r5283] RESPA: change in the input structure &GLOBAL PROGRAM ->
	  &FORCE_EVAL METHOD (see regtests).
	  Two fully independent FORCE_EVAL sections can be around in the
	  input. Thus, do not directly access them from the
	  globenv%input_file, but use the force_env(section) directly.
	  force_env section now contains a flag indicating what method it
	  describes. All now incorrect uses of globenv%input_file fixed.

2006-08-08 08:37  jgh

	* [r5282] remove not allowed pointer assignment

2006-08-07 15:46  ikuo

	* [r5281] Bug fix for harmonic_eval

2006-08-06 09:44  vondele

	* [r5280] fix build

2006-08-04 16:09  fawzi

	* [r5279] corrected missing comma

2006-08-04 15:55  fawzi

	* [r5275] * f77_int_low to calc forces at low level
	  * working replica_env
	  * ressurrected ep (now only at e0 level)

2006-08-04 13:29  jgh

	* [r5274] New routine for atom overlap

2006-08-03 21:10  ikuo

	* [r5273] Add harmonic restraints

2006-08-03 14:20  jgh

	* [r5272] Separate qs_overlap and sparse_matrix output routines

2006-08-01 21:03  ikuo

	* [r5271] bug fix, change multiplicity from real to integer

2006-08-01 18:01  ikuo

	* [r5270] bug fix, change multiplicity from real to integer

2006-08-01 17:30  ikuo

	* [r5267] add a restraint section and put flexible partitionng
	  inside the restraint section
	  no special output for the restraint section yet

2006-08-01 09:23  krack

	* [r5266] bug fix for convergence check

2006-07-29 12:35  jgh

	* [r5265] Molecular dipoles and quadrupoles for KG_GPW method

2006-07-26 16:48  ikuo

	* [r5264] shorten subroutine name for absoft compiler

2006-07-26 10:15  vondele

	* [r5262] Further RESPA improvements (Florian Schiffmann)

2006-07-20 23:50  ikuo

	* [r5261] Fix for Intel compiler bug

2006-07-20 17:42  ikuo

	* [r5260] fix so openmp can compile

2006-07-20 14:50  krack

	* [r5259] is_fixed cleaned from particle_type; new regtest
	  H2O-fixed.inp added

2006-07-19 15:15  tlaino

	* [r5258] bug fix for the use of multiple force_eval_section..
	  Introducing the RESPA
	  in the NVE ensemble (by Florian Schiffmann). regtest ok!

2006-07-19 14:11  krack

	* [r5257] Fix atom constraint for BFGS geo. opt. recovered. Another
	  lost&found story.

2006-07-18 08:45  jgh

	* [r5255] fix memory leak

2006-07-18 06:25  tlaino

	* [r5254] introducing the possibility to have multiple force_eval
	  sections in the input files.
	  preliminary setup for RESPA. regtest ok.

2006-07-17 09:50  marcella

	* [r5252] GAPW current density

2006-07-15 16:10  fawzi

	* [r5250] fixing bug (specification inconsistent with default) nvar
	  2->3

2006-07-14 14:51  jgh

	* [r5248] remove debug statement

2006-07-11 18:31  ikuo

	* [r5246] quick and dirty openmp calls for fist force

2006-07-10 16:40  jgh

	* [r5245] New defaults for single molecule box sizes
	  EPS_KG_GPW can be set from input
	  Reset of KG_GPW regtests

2006-07-09 12:43  jgh

	* [r5242] Huge reduction of memory requirements for basis sets with
	  large l quantum number
	  'static' allocation of auxilliray arrays replaced by
	  allocation/deallocation
	  when needed
	  recalculation of index array causes some overhead

2006-07-07 15:52  jgh

	* [r5241] Printing changes

2006-07-07 14:55  jgh

	* [r5240] print out of XC references

2006-07-07 13:27  jgh

	* [r5239] Printing of grid information slightly changed

2006-07-07 12:42  jgh

	* [r5238] Update and bug fixes for electric field gradient

2006-07-06 16:02  ikuo

	* [r5237] bug fix for 1-4 pv numerical

2006-07-06 15:34  ikuo

	* [r5236] Bug fix for missing 1-4 interaction in force numerics

2006-07-04 13:45  tlaino

	* [r5235] index bug fix.. bug fix in generate_particle_distances
	  (now uses the molecule definition)
	  and more comments on an exit error..

2006-07-01 00:19  fawzi

	* [r5234] cosmetic: nicer info printout

2006-07-01 00:00  fawzi

	* [r5233] added scaling to PW92 (to be able to do PBE1W), removed
	  the *_new from the names of many xc modules.

2006-06-30 22:55  ikuo

	* [r5232] Fixes for gromos routines

2006-06-30 21:52  krack

	* [r5231] Enable selective MO printout during the SCF (EACH option)

2006-06-29 17:01  ikuo

	* [r5228] Energy match for bond,bends,torsion to gromos

2006-06-27 17:17  krack

	* [r5227] Use last MOs for transformation and level-shift when ROKS
	  (no OT) is restarted

2006-06-27 16:31  krack

	* [r5226] Use last MOs for transformation when ROKS (no OT) is
	  restarted

2006-06-27 15:05  krack

	* [r5225] Enable restart with smaller MO set

2006-06-26 23:26  ikuo

	* [r5224] Fixes and changes to read in GROMOS topology file and
	  coordinate.
	  Units are not correct yet....

2006-06-23 15:04  krack

	* [r5222] bug fix for GHOST potential in connection with
	  GAPW/all-electron (BSSE calculation)
	  * GHOST potential removed for POTENTIAL keyword
	  * Now the keyword GHOST has to be added to the KIND section to
	  make all atoms
	  of that kind ghost atoms
	  * Reset of regtest inputs for BSSE due to print format change
	  (numbers are
	  still the same)
	  * Generic basis set names added as alias names in the
	  EMSL_BASIS_SETS file,
	  e.g. now 6-311++G(3df,3pd) can be specified in the input if
	  quoted:
	  BASIS_SET "6-311++G(3df,3pd)"
	  * Typo in basis set name corrected: 6-311++G3df2pd ->
	  6-311++G3df3pd

2006-06-21 20:45  ikuo

	* [r5221] Put things in place to have GROMOS support soon...

2006-06-21 12:22  marcella

	* [r5220] more generalized collocation

2006-06-21 11:00  marcella

	* [r5219] generalized collocate: general functions of the
	  primitives
	  can be defined in qs_modify_pab_block and
	  put on the grid using the same scheme adopted for tau

2006-06-19 11:59  tkuehne

	* [r5218] Corrected output

2006-06-16 15:51  marcella

	* [r5217] angmom integrals : some corrections
	  nmr : in part rewritten
	  some other details and a little cleaning in the other routines

2006-06-16 08:42  tkuehne

	* [r5216] harris_metropolis_mc.F: Introduction of LPSC, which
	  corrects for the phase space
	  violation due to using Langevin Dynamics. Mainly this is
	  the usual Metropolis acceptance criterion, but based on
	  what Matt calls "the difference of differences", since
	  an auxilliary Markov Chain is used in combination with
	  Langevin Dynamics.
	  qs_scf_post.F & qs_wf_history_methods.F: Minor modifications

2006-06-16 08:40  krack

	* [r5215] shortcut BP for the Becke88/Perdew86 XC functional added
	  for convenience
	  (just to be consistent with the shortcuts of the PP library)

2006-06-14 21:01  ikuo

	* [r5214] Missing init causing problem when running without
	  "DO_HILLS"

2006-06-12 07:11  tlaino

	* [r5213] printing energy in geoemtry optimizer if the starting
	  geometry is yet fully converged..
	  Fixing the length of the filenames (making consistent with the
	  long_char type defined
	  in the input)

2006-06-06 13:31  vondele

	* [r5210] add the derivative of the system dipole (non-berry phase
	  way) to the output

2006-06-06 12:49  marcella

	* [r5209] fix for large exponents in the construction of the
	  reduced 3-center lists

2006-06-06 11:23  vondele

	* [r5208] standard timings

2006-06-05 16:04  mcgrath

	* [r5207] added the input into fist_force_env...not sure why it
	  wasn't done before,
	  but it makes handling the .dat file easier for MC

2006-06-03 06:29  vondele

	* [r5205] Clean format

2006-06-02 17:27  tlaino

	* [r5204] Bug fix for onfo. Fully conforming amber results to FIST
	  results. introduced few
	  other tests to check that amber and fist will be compatible in
	  the future..
	  Introduced a keyword to avoid the scaling of the cutoff energy..
	  In this way amber and fist
	  give exactly the same results.A bug fix also in the
	  topology_converter.. cosmetics around..
	  reset 2 regtests.

2006-06-02 12:17  vondele

	* [r5202] bug fix, close files

2006-06-02 11:16  jgh

	* [r5201] bug fix: initialization

2006-06-01 19:46  vondele

	* [r5198] allows for elec_conf in the &KIND section to influence
	  the atomic initial guess

2006-05-31 17:51  tlaino

	* [r5195] Bug fix for 1-4, 1-3 and 1-2 interactions..
	  Bug fix for double counting of 1-4 interactions..
	  Bug fix for spline_bond_env (1-4) generations..
	  Introduced a test used to compare the single energy contributions
	  with AMBER.
	  The energies agree 10^-6-10^-7 if the e_cutoff_coul and
	  e_cutoff_vdw is not
	  taken into account..
	  Infact AMBER defines the electrostatic and VDW potential with a
	  step at the
	  r_cutoff.. (the potential does NOT go to zero at the cutoff
	  distance)
	  Since we want a potential that is zero at the r_cutoff
	  our values will be still in agreement with the ones of AMBER
	  though not
	  numerically similar (just shifted of a fixed amount)
	  FIST and AMBER synchronized..

2006-05-31 14:26  tlaino

	* [r5194] fully compatible (w.r.t. translation) geometry optimizer
	  for QM/MM runs..
	  reset regtests due to numerics..

2006-05-31 12:50  tlaino

	* [r5193] bug fix for geometry optimization (coming up for qm/mm
	  calculations)
	  introducing a new test_type for geometry optimizations,
	  converting the relative
	  tests (QM/MM and FIST) to this new test_type.. few resets in Fist
	  due to the new
	  spline scheme (these tests were never resetted).. few resets in
	  QM/MM due to an
	  increase of the steps of geometry optimizer..

2006-05-30 19:28  vondele

	* [r5190] bug fix for access beyond last char of string

2006-05-24 20:41  tlaino

	* [r5188] bug fix for onfo for 5 and 6 membered rings.. introducing
	  2 specific regtests..
	  resetting loop.inp.

2006-05-19 12:37  krack

	* [r5181] * keyword RESTRICTED removed
	  * now simply ROKS has to be specified instead of LSD + RESTRICTED
	  * DFT printout rearranged/cleaned
	  * test inputs updated due to the removal of the RESTRICTED
	  keyword

2006-05-19 10:13  tlaino

	* [r5180] Splitting tbmd from the fist_neighbor_list_type ..

2006-05-18 23:26  marcella

	* [r5179] small details for xas

2006-05-18 22:01  cjmundy

	* [r5178] Finally Jochen's Virial for LDA. Sorry for the delay.
	  
	  Things to deal with next:
	  -interesting sign in qs_force.F must be fixed
	  -GGA implementation next to come
	  -passed regtests

2006-05-17 15:47  tlaino

	* [r5177] bug fix in the sorting trick

2006-05-17 09:12  tlaino

	* [r5176] fixing recursive IO

2006-05-17 08:08  tlaino

	* [r5175] fixing wrong warning on torsion definitions..

2006-05-17 01:12  tlaino

	* [r5174] new linear scaling (LS) setup for FIST: major rewriting
	  for:
	  (1) spline_nonbond_env and spline_bond_env
	  (2) check_molecule (improved also the checks, introduced also a
	  check on the topology of bonds) (LS)
	  (3) the molecule exclusion list (LS)
	  (4) connectivity_pack
	  Making faster the reading of PSF and XYZ..
	  getting rid of the keyword LOW_STORAGE for the spline.. the new
	  algorithm (LS) evaluated directly
	  the only unique splines (much faster)
	  Regtest OK only numerics (10^-12) for one test (loop.inp)
	  
	  Next step neighboring lists..

2006-05-12 19:40  tlaino

	* [r5171] completely new routine core generate_molecule (scales
	  linearly w.r.t. syste size) (Joost/Teo)

2006-05-12 16:26  tlaino

	* [r5170] low_storage keyword for splines

2006-05-12 15:32  tlaino

	* [r5169] gettinig rid of the exclude_vdw_index and
	  exclude_ei_index (are just a waste of memory)

2006-05-12 13:11  krack

	* [r5168] Lost trajectory printout for BFGS recovered

2006-05-12 09:35  vondele

	* [r5167] more wrong formats (Vittorio Zecca)

2006-05-12 08:55  tlaino

	* [r5166] new exclusion list types.. (memory light version)

2006-05-12 07:57  vondele

	* [r5165] another spme speedup

2006-05-11 20:14  vondele

	* [r5162] small speedup (& 1E-16 change in numerics for two
	  regtests)

2006-05-11 17:39  marcella

	* [r5161] bug fix localization with isolated system

2006-05-11 17:14  vondele

	* [r5160] add forgotten file (sorry)

2006-05-11 15:33  vondele

	* [r5159] fix for few splines

2006-05-11 15:32  vondele

	* [r5158] giving a hint that spline generation is taking place

2006-05-11 15:15  tlaino

	* [r5157] introducing the keyword constraint_info; introducing a
	  check in the construction of the neighboring lists
	  to control that no atoms fall below the distance used for splines

2006-05-11 10:57  vondele

	* [r5156] fixed format (thanks Vittorio Zecca)

2006-05-11 10:15  vondele

	* [r5155] another round of speedups

2006-05-10 22:50  marcella

	* [r5154] missing conversion factor

2006-05-10 20:44  vondele

	* [r5153] two further speedups

2006-05-10 15:54  tlaino

	* [r5151] getting rid of the comparison between strings in the
	  potential evaluation

2006-05-10 15:08  vondele

	* [r5150] further speedup (make sure no temporaries are generated,
	  especially function calls)

2006-05-10 14:43  tlaino

	* [r5149] bug fix

2006-05-10 13:22  vondele

	* [r5148] and another one

2006-05-10 13:21  vondele

	* [r5147] forgot one file

2006-05-10 13:09  vondele

	* [r5146] another large FIST speedup

2006-05-10 12:03  tlaino

	* [r5145] speeding up further the topology_constraint_pack (thanks
	  Joost)

2006-05-10 10:37  vondele

	* [r5143] fix bug if charge/occup is not associated

2006-05-10 10:20  tlaino

	* [r5142] introducing r0_nb (to have more flexibility for starting
	  systems with close contact atoms)
	  slightly modified the fist neighbouring list (light loop
	  optimization)

2006-05-10 10:16  tlaino

	* [r5141] fixing the format of the dump psf file..

2006-05-09 14:37  marcella

	* [r5136] typo in debug forces

2006-05-09 13:45  krack

	* [r5135] remove redundant printout

2006-05-09 12:44  jgh

	* [r5134] Bug fix: missing coordinates added

2006-05-09 11:34  tlaino

	* [r5133] speed-up FIST list builds by a factor of (more than) 5
	  (JAC) (Joost) (teo)

2006-05-09 11:01  vondele

	* [r5132] fix multiple atoms per line for mulliken constraint

2006-05-09 10:20  krack

	* [r5131] ROKS for semiempirical methods (SE), i.e. for orthogonal
	  basis sets in general

2006-05-09 09:34  tlaino

	* [r5130] cleanup (fixing default for subcells)

2006-05-09 08:03  tlaino

	* [r5129] more debug printout

2006-05-09 08:02  tlaino

	* [r5128] cleanup (Joost)

2006-05-08 19:58  vondele

	* [r5127] small fist list cleanup and speedup

2006-05-08 18:56  vondele

	* [r5126] further doc fixes

2006-05-08 18:39  vondele

	* [r5125] further qmmm docs

2006-05-08 17:42  tlaino

	* [r5123] bug fix and cleaning of not useful printout (too much
	  printout difficult even to understand)..

2006-05-08 16:15  vondele

	* [r5121] allow for longer enum descriptors

2006-05-08 16:11  vondele

	* [r5120] added docs for qmmm coupling

2006-05-08 15:48  vondele

	* [r5118] added SCF_GUESS docs

2006-05-08 15:42  krack

	* [r5116] Cholesky for ROKS

2006-05-08 15:09  vondele

	* [r5115] if a RESTART file is not present, fall back to SCF_GUESS
	  ATOMIC if SCF_GUESS RESTART is specified. Corrupted or invalid
	  restarts still cause an abort.

2006-05-08 14:59  marcella

	* [r5114] ops

2006-05-08 14:11  jgh

	* [r5113] New property: EFG, not yet fully tested

2006-05-08 14:09  marcella

	* [r5112] periodic de-coupling/re-coupling by multipole for
	  QMMM-GAPW

2006-05-08 13:37  vondele

	* [r5111] fixed a bug in the interaction between the homos
	  printkey, OT and forces

2006-05-08 13:15  krack

	* [r5109] proper index for serial run

2006-05-08 12:43  marcella

	* [r5108] fix get_qs_env call

2006-05-08 12:28  krack

	* [r5107] Level-shifting for ROKS

2006-05-08 11:32  marcella

	* [r5106] missing term in the qmmm electrostatic energy with gapw

2006-05-08 11:15  tlaino

	* [r5105] getting rid of a nested function call

2006-05-07 20:32  tlaino

	* [r5104] fixing the length of few residual filenames (left the
	  internally generated filenames to defautl_string_length while the
	  one possibly parsed from the input file to default_path_length)..
	  (thanks Axel)

2006-05-07 17:49  tlaino

	* [r5103] fixing performance of the build-up of the exclusion list.

2006-05-07 08:58  vondele

	* [r5102] fix farming newinput

2006-05-07 08:31  vondele

	* [r5101] tiny speedup

2006-05-06 19:48  tlaino

	* [r5100] (1) New spline implementation (light-memory and enhancing
	  the control on the
	  accuracy of the spline generation).
	  (2) bug fix due to numerics in the IPBV potential
	  (3) cleaned the splines.F removed all unused routines
	  (4) bug fix in e_coulomb (pair_potential.F)
	  (5) setting the defaults of force_field_types directly in the
	  input
	  (6) elimination of multiple allocations of splines. Now the
	  splines allocated are UNIQUE
	  (Teo and Joost)

2006-05-05 19:12  marcella

	* [r5098] better cutoff for the 3 center lists; background charge;
	  rho0 instead than rho_core for the qm/mm interaction

2006-05-05 18:12  vondele

	* [r5097] added further timings/memory checks

2006-05-04 16:23  krack

	* [r5096] * Printing in qs_environment postponed, i.e. moved to the
	  end until all input
	  parameters are settled.
	  * Printout of ATOMIC KIND and BASIS SET/POTENTIAL "homogenised"
	  * qs_scf: first step for level-shift with ROKS

2006-05-04 14:39  tlaino

	* [r5095] getting rid of old stuff..

2006-05-04 14:17  vondele

	* [r5094] added more options to the localization keyword

2006-05-03 17:54  krack

	* [r5090] High-spin ROKS recovered: energies and forces coincide
	  now with OT RESTRICTED
	  for doublets) and with Gaussian03 for higher multiplicities.
	  General ROKS: unfinished

2006-05-03 15:26  jgh

	* [r5089] remove debug code

2006-05-03 06:43  jgh

	* [r5088] Test for orthorhombic cells in grids routines

2006-05-03 06:42  jgh

	* [r5087] Bug fix: add orthorhombic to cell_clone

2006-05-02 06:26  tlaino

	* [r5086] redundant logical statements

2006-05-01 22:15  ikuo

	* [r5085] Redundant call, save 5% for fist_intra_control for large
	  systems

2006-05-01 18:26  ikuo

	* [r5084] fix for constraint pack. (thanks axel and guillaume)

2006-05-01 18:15  ikuo

	* [r5082] Change the default to something more reasonable

2006-05-01 17:11  tlaino

	* [r5081] reorganizing BFGS geometry optimizer.. now all methods
	  share the same routine
	  to evaluate energies and forces.. ready for general constraints..
	  reset due of numerical differences (10^-11 -> 10^-13) of few
	  regtests

2006-05-01 14:53  tlaino

	* [r5080] fix

2006-05-01 14:52  tlaino

	* [r5079] fixing memory leaks

2006-05-01 11:48  tlaino

	* [r5078] reorganizing CG geometry optimizer.. no changes in
	  regtests

2006-04-30 16:40  tlaino

	* [r5077] no need to scale coordinates in the geometry optimizer..
	  starting to prepare
	  everything to reorganize the geometry optimizers with
	  constraints.

2006-04-29 23:40  tlaino

	* [r5076] avoiding the repetition of the first/last info line
	  during a restart in ener, traj., vel., forc., files..

2006-04-28 20:28  tlaino

	* [r5075] Bug fix in the definition of colvar constraints for the
	  number of degree
	  of freedom.. resetting relative regtests..
	  New collective variable, rotation of a bond w.r.t. another bond.

2006-04-28 19:54  tlaino

	* [r5074] Getting rid of the EXCLUDE_CHECK keyword, you don't need
	  to provide anymore
	  this keyword.. the check is performed automatically. optimized
	  the construction
	  of the molecule list to build the excluded lists. regtest ok!

2006-04-28 06:15  krack

	* [r5073] Remove redundant printout

2006-04-27 19:36  vondele

	* [r5072] restore printout

2006-04-27 08:35  vondele

	* [r5071] fist_nonbond_force.F, fist_neighbor_list_types.F: speed
	  up force_nonbond by a large factor (8 times for g95, 2 times for
	  nag).
	  fist_neighbor_lists.F, fist_neighbor_list_types.F, kinds.F:
	  remove unused cell of the fist_neighbor_node

2006-04-26 10:18  krack

	* [r5069] bug fix in FIST and TBMD neighbor lists

2006-04-26 10:13  tlaino

	* [r5068] let's abort instead of continuing the calculation..

2006-04-25 15:52  krack

	* [r5067] Check allocation/association status before DEALLOCATE
	  Cosmetics

2006-04-25 11:22  tlaino

	* [r5066] Modified QM_CELL into CELL..

2006-04-25 07:39  vondele

	* [r5065] docs

2006-04-24 15:55  vondele

	* [r5064] added initial implementation of flexible partitioning

2006-04-24 09:44  tlaino

	* [r5063] Introducing a section QM_CELL in QMMM section..modified
	  all regtests of QMMM properly.
	  no changes in regtests.

2006-04-21 17:08  tlaino

	* [r5062] bug fix for introduced memory leaks..

2006-04-21 14:58  tlaino

	* [r5061] bug fix for constraints, rewriting of the setup part..
	  deleted completely all
	  the reordering part.. introduced the possibility to don't specify
	  a target
	  value (for all the set of same molecules the target values are
	  evaluated for
	  the first molecule of the set, and those values used for all
	  kinds..)
	  introduced a regtest to test this feature..

2006-04-21 13:44  mcgrath

	* [r5060] Removed an obsolete keyword, changed it so data files now
	  append instead of replace

2006-04-21 09:15  vondele

	* [r5059] keep OMP running (thanks Francois-Xavier.Coudert@lcp)

2006-04-21 07:30  vondele

	* [r5058] improve docs

2006-04-20 22:02  marcella

	* [r5057] format

2006-04-20 20:51  marcella

	* [r5056] bugs fixed in chi calculation
	  simple work-around to reduce memory allocations in GAPW at low
	  accuracy cost

2006-04-20 20:41  tlaino

	* [r5055] bug fix for uninitialized type.. fixing memory leaks..

2006-04-20 17:31  tlaino

	* [r5053] General scheme for using collective variables as
	  constraint, based on
	  an iterative procedure, available both for NV* and NP*
	  ensembles.. added new regtests (constraint
	  angle and torsion).. resetting the regtests based on constraints
	  due to a different use
	  of the functional form of the distance..
	  Deleted the old scheme for distance constraint.. some cleaning
	  around..
	  To add a new constraint only the functional form and the
	  derivatives need to be provided
	  in colvar_methods.. no other modifications to the routines
	  constraints* need to be done..

2006-04-19 14:56  vondele

	* [r5052] improved typo output (ranking and limit)

2006-04-19 11:29  vondele

	* [r5051] treat both '#' and '!' as comment characters in the
	  input. Being able to avoid '#' is useful for cpp'ing input files
	  with comments

2006-04-18 14:52  vondele

	* [r5049] enhance print key

2006-04-17 18:21  krack

	* [r5048] ROKS revised and enabled again; needs testing for
	  validation.

2006-04-15 13:39  vondele

	* [r5047] fix against non-standard system_clocks.

2006-04-14 17:26  ikuo

	* [r5046] Add kjmol to cp_units for input parsing

2006-04-14 10:24  krack

	* [r5045] disable g space mixing when DIIS is active

2006-04-13 21:16  ikuo

	* [r5044] Fix torsion input nmul problem
	  add regtest for FF through the input

2006-04-13 16:50  tlaino

	* [r5042] change variable name to override a compilation bug in pgi
	  6.1.1 (cray)

2006-04-13 15:25  krack

	* [r5041] simple g-space mixing plus Kerker damping

2006-04-13 08:23  vondele

	* [r5039] added some more docs

2006-04-13 06:56  tlaino

	* [r5038] not really "bug fix".. changing the behavior of
	  get_atomic_kind: in principle could create problems
	  not easily traceable

2006-04-12 16:39  krack

	* [r5033] line break

2006-04-12 16:08  krack

	* [r5032] use (p)dsyevd for full matrices

2006-04-12 16:06  krack

	* [r5031] * timeset added
	  * more informative error message

2006-04-12 15:25  vondele

	* [r5030] fix fix

2006-04-12 15:23  vondele

	* [r5029] took away flops

2006-04-12 13:48  krack

	* [r5028] ROKS disabled, since the simple approach doesn't work
	  with DFT

2006-04-12 13:46  vondele

	* [r5027] added sub timesets in force_field_pack

2006-04-12 10:12  vondele

	* [r5023] saves about 20% of memory in Fist neighbor lists

2006-04-11 16:02  tlaino

	* [r5021] Bug fix in the qeff part of force_fields.F
	  Making the input case insensitive in the definition of the
	  forcefield
	  Changing accordingly few input.. Charges now needs to be
	  specified in one of the many
	  possible ways.. But it must be defined anyway!

2006-04-11 15:57  krack

	* [r5019] * Restricted open Kohn-Sham (high-spin ROKS)
	  * Calculation of spin contamination adapted
	  * Unused Jacobi routines removed
	  * Cosmetics

2006-04-07 17:02  marcella

	* [r5018] calculation of chi

2006-04-06 22:35  mcgrath

	* [r5016] Forgot a write statement

2006-04-06 18:25  mcgrath

	* [r5015] Some new stuff for the GEMC that the new input
	  broke...seems to be working, but hunting down memory leaks not
	  found by g95 or valgrind

2006-04-05 16:09  marcella

	* [r5014] sm x fm : problems with antisymmetric matrix

2006-04-05 14:40  mcgrath

	* [r5013] Gotten rid of the dat file in the GEMC volume move

2006-04-04 19:21  mcgrath

	* [r5011] Bug fix..not all the cell info got copied over

2006-04-04 17:54  cjmundy

	* [r5010] Moderate changes:
	  1) All box_change logicals now deleted. Everything is
	  handled through force_env_set_cell.
	  2) No more crazy wrting out a new input file
	  for MC. Internally performed.
	  3) cell_type no longer part of ewald_env_type
	  
	  Passed my regtests and Matthew's.

2006-04-04 16:48  tlaino

	* [r5009] Modified bondparm_factor -> bondparm_factor^2; introduced
	  a molname_generated logical
	  to allow skipping the dumping of the molname if internally
	  generated by cp2k;
	  changed also molname definition: deleted the reference to resname
	  (because restart after restart
	  was making to explode the dimension of the molname string) ; Bug
	  fix for input_val_types.F for long_char:
	  if the string read was exactly of the same dimension as
	  default_string_length was cutting the last character.
	  REGTEST OK!

2006-04-04 16:19  marcella

	* [r5007] lists with all the pairs : a-b and b-a

2006-04-04 10:40  tlaino

	* [r5006] bug fix in topology_util.F (generation of bond_list),
	  cleaning of commented (old) stuff in topology.F

2006-04-03 15:53  fawzi

	* [r5005] simple wannier total dipole

2006-03-31 20:23  mcgrath

	* [r5002] Parallel bug (broadcast variable before using)

2006-03-31 19:16  ikuo

	* [r5001] Workaround for intel compiler error on 9.0.31 and 9.0.32
	  when using debug

2006-03-31 13:45  vondele

	* [r5000] avoid breaking single precision builds

2006-03-31 13:44  vondele

	* [r4999] use _dp instead of D0

2006-03-31 10:56  vondele

	* [r4997] added accurate_sum as a way to sum a lot of number
	  without loss in precision. Based on Kahan's algorithm. reset of
	  150 regtests.

2006-03-30 17:37  tkuehne

	* [r4996] EIP: Changes to make the mentally ill compiler-guys from
	  Oxford and it's idolizers happy.
	  TBMD: Changes due to the checkerboard matrix-distibution + minor
	  bugfix

2006-03-30 15:44  krack

	* [r4993] correct Lebedev grid check for maximum angular momentum
	  number

2006-03-30 14:53  vondele

	* [r4991] added another OT preconditioner (still experimental)

2006-03-30 12:07  tlaino

	* [r4987] fixing use_ref_cell and cell_ref for use_qmmm in
	  force_env_get

2006-03-30 07:36  vondele

	* [r4982] init the name field of STOs

2006-03-30 07:17  ikuo

	* [r4981] Bug Fix: out of bound error on ex_bond_list when using
	  coords with XYZ w/o molecule info

2006-03-29 15:10  mcgrath

	* [r4980] Slightly altered regest and .dat file stuff for biasing
	  with more complex forcefields

2006-03-29 12:54  vondele

	* [r4979] keep a restart history

2006-03-29 12:21  vondele

	* [r4978] formating

2006-03-29 10:18  marcella

	* [r4977] print statement fixed

2006-03-29 08:54  vondele

	* [r4976] add 'WAIT' to the external control. Useful if CP2K is
	  going too fast.

2006-03-29 08:22  vondele

	* [r4975] cycle through a number of bak files. This implementation
	  is kind of a hack.

2006-03-29 07:14  krack

	* [r4974] work-around for PGI compiler (silly)

2006-03-28 15:25  jgh

	* [r4973] PDDG/PM3 method debugged; Si/P/S/ parameters for AM1
	  fixed

2006-03-28 12:47  mcgrath

	* [r4972] Forgot some write statements

2006-03-28 12:40  jgh

	* [r4970] Update and bug fix for parallel SE (still problems with
	  Coulomb)

2006-03-27 21:32  mcgrath

	* [r4969] Slight change for KG dat files

2006-03-27 17:00  mcgrath

	* [r4968] Fixed a leak and had to add the print level to the MC
	  structure

2006-03-27 15:34  mcgrath

	* [r4967] A couple changes for creating new force_env

2006-03-27 09:55  marcella

	* [r4965] changes to handle non symmetric sparse matrix

2006-03-27 07:56  vondele

	* [r4964] rebuild preconditioner for selected OT cases, reortho wfn
	  for outer scf

2006-03-26 17:16  jgh

	* [r4962] Deleted Ewald method for semi-empirical;
	  Separated Exchange and Coulomb integrals;
	  Independent cutoff function for Coulomb integrals;
	  Added Jorgensens PDDG/PM3 parametrization

2006-03-25 12:52  vondele

	* [r4961] further make_basis cleaning (has been split), resulting
	  in some speedup and cleaner code

2006-03-24 20:50  mcgrath

	* [r4960] Small change to writing out .dat files

2006-03-24 19:13  tkuehne

	* [r4959] Minor adaptions due my last check-in

2006-03-24 16:14  vondele

	* [r4958] cleanup of make_basis, now free of explicit scalapack
	  calls

2006-03-24 16:14  vondele

	* [r4957] fix bug in cp_fm_syrk

2006-03-24 12:04  vondele

	* [r4956] cleaning: avoid passing large parts of the input around.
	  In particular, only pass the required bit of the print section to
	  the grid related procedures.

2006-03-24 09:47  krack

	* [r4953] printout of special structure data recovered

2006-03-24 01:27  ikuo

	* [r4952] Fix bend problem for linear molecules (ie CO2)
	  Debug version not yet fixed

2006-03-23 20:33  vondele

	* [r4951] upcase some input keywords

2006-03-23 20:04  vondele

	* [r4950] have gaussian_gridlevels retain its own printkey (no
	  kidding)

2006-03-23 15:04  vondele

	* [r4949] remove OPTIONAL as the argument is not

2006-03-23 13:50  vondele

	* [r4948] doc typo

2006-03-23 13:31  vondele

	* [r4947] added new dates

2006-03-23 13:26  vondele

	* [r4946] made interface of calculate_wavefunction independent of
	  qs_env

2006-03-23 08:36  vondele

	* [r4945] prettify

2006-03-22 23:28  marcella

	* [r4943] bug fixed in the analytic calculation of the p operator

2006-03-22 18:10  ikuo

	* [r4942] Turn MOL_CHECK on as default
	  Add more descriptions for CONNECTIVITY keyword

2006-03-22 18:06  ikuo

	* [r4941] Fix DCD dump on some platforms

2006-03-22 13:13  vondele

	* [r4939] do nothing in the other scf

2006-03-22 10:18  vondele

	* [r4937] use google to search the input

2006-03-21 23:35  ikuo

	* [r4934] fix wrong printkey name

2006-03-21 19:20  tlaino

	* [r4933] bug fix for element name (thanks will!)

2006-03-21 13:19  tlaino

	* [r4932] added DOI for qmmm

2006-03-21 11:45  tlaino

	* [r4931] bug fix for element name -> wrong element in xyz (if
	  element was not defined)
	  correct the reference of the qmmm paper

2006-03-21 09:23  tkuehne

	* [r4930] Harris: FILL_WF_HISTORY & HARRIS_MAX_SCF added
	  HMC: MULTIPLE_SCF added

2006-03-21 08:33  vondele

	* [r4929] fix some write statements

2006-03-20 14:26  krack

	* [r4928] allow default_path_length for cwd

2006-03-20 12:31  vondele

	* [r4927] improved slightly the html manual

2006-03-20 12:02  vondele

	* [r4926] test

2006-03-20 10:38  vondele

	* [r4923] test

2006-03-19 17:04  cjmundy

	* [r4921] Fixed bug that effects DEBUG with FIST. Somehow, these
	  routines
	  became out of sync. Needed to add an optional logical
	  "orthorhombic"
	  to init_cell. This allows us to use general routines when
	  computing
	  numerical stress.

2006-03-18 09:09  ikuo

	* [r4920] Remove molname from particle set

2006-03-17 18:29  marcella

	* [r4917] again about chi for gpw

2006-03-17 17:49  jgh

	* [r4916] POISSON_DFT and POISSON_MM section combined to section
	  POISSON
	  -> change of input!
	  Included screening section into POISSON
	  Bug fix for Poisson solver ANALYTIC0D
	  Further improvements for SE with Ewald

2006-03-17 12:46  vondele

	* [r4914] s2 cons added

2006-03-17 10:58  tkuehne

	* [r4913] TBMD: Modifications due to Juerg's change of the
	  eigensolver interface.

2006-03-17 00:55  marcella

	* [r4912] analytic moment operator and chi for gpw (not complete
	  yet)

2006-03-16 19:40  tkuehne

	* [r4911] HMC: Bug fix

2006-03-16 18:26  tkuehne

	* [r4910] HMC: DRLD added

2006-03-16 17:22  ikuo

	* [r4909] fix input torsion

2006-03-16 00:20  ikuo

	* [r4908] fix multiple call of PSF reader and MOL_SET section

2006-03-15 23:32  marcella

	* [r4907] xas module reorganized, small changes in linres

2006-03-15 19:51  ikuo

	* [r4905] Compile for DEC

2006-03-15 11:20  vondele

	* [r4896] fixed write statements for outer scf

2006-03-14 21:21  ikuo

	* [r4893] Oops, undo of molname remove from particle_set

2006-03-14 20:08  ikuo

	* [r4892] remove molname from particle_set

2006-03-14 19:01  ikuo

	* [r4891] fix input nonbond14, ub

2006-03-14 16:41  jgh

	* [r4890] SE update: long range Coulomb not yet working

2006-03-14 16:29  tkuehne

	* [r4889] Integrating EIP into HMC.

2006-03-14 15:04  vondele

	* [r4888] date

2006-03-14 12:58  krack

	* [r4886] fix: too many continuation lines for the NAG compiler

2006-03-14 12:42  vondele

	* [r4885] added the CPC & OT papers to the QS header, added doi
	  links

2006-03-14 12:13  vondele

	* [r4884] added geo opt restarts

2006-03-14 09:36  vondele

	* [r4883] moved &MOTION &MD &PRINT &RESTART to &MOTION &PRINT
	  &RESTART

2006-03-13 19:57  marcella

	* [r4881] change reallocate in allocate

2006-03-13 16:46  krack

	* [r4880] * removal of the NR random number generator (ran2,gasdev)
	  * some calls to RANDOM_NUMBER in test routines are still present,
	  but
	  could be substituted by CALL random_numbers(array,rng_stream)
	  from
	  parallel_rng_types
	  * reset of many test inputs, since all initial velocities have
	  changed
	  * some cleaning in libtests

2006-03-13 15:48  vondele

	* [r4879] made a subroutine md_output that groups output related
	  stuff during MD, to allow for easy timing... writing of restarts
	  during classical had better be done only once in a long while

2006-03-13 15:06  vondele

	* [r4876] fix multiple atoms case

2006-03-13 14:54  vondele

	* [r4875] doc typo

2006-03-13 11:54  marcella

	* [r4874] some cleaning

2006-03-13 09:09  vondele

	* [r4872] added a standard m_walltime that is less likely to roll

2006-03-13 07:24  vondele

	* [r4871] added an outer scf

2006-03-12 15:12  jgh

	* [r4870] Explicitely refer to the different SE methods, no ranges

2006-03-12 01:07  krack

	* [r4869] empty if clause removed

2006-03-10 17:06  krack

	* [r4867] bug fix for reference cell to guarantee a clean MD
	  restart

2006-03-10 14:27  tlaino

	* [r4866] moved colvar and structures internal to colvar to
	  pointers (only allocated when used)..
	  modified the global and local structures for constraint in order
	  to have colvar related to
	  molecules to use later to evaluate constraints..
	  introduced the colvar_p_type.
	  Next thing to do a general solver for contraint equations
	  (working on it now!)

2006-03-10 12:39  krack

	* [r4865] forgotten debug write statement removed

2006-03-10 11:03  vondele

	* [r4864] last round of cleaning before adding new functionality

2006-03-10 10:02  vondele

	* [r4863] another round of cleaning, move these parts that expect a
	  converged wavefunction out of the scf_do function into scf(

2006-03-08 16:58  krack

	* [r4862] * RNG restart via string record recovered
	  * Optionally uniform or normal Gaussian distributed random number
	  stream
	  for the parallel random number generator (RNG)
	  * Gaussian distribution for Langevin -> reset of affected regtest
	  inputs
	  * max. line length of parser set to 500 = 2*default_path_length
	  for
	  RNG record (433)
	  * libtest routine and input updated/extended
	  * writing "ES25.16" instead of "*" format for real numbers in the
	  restart

2006-03-08 16:17  vondele

	* [r4861] split init_scf_run in init_scf_run and init_scf_loop

2006-03-08 15:12  vondele

	* [r4860] added additional test is syevd

2006-03-08 13:46  tkuehne

	* [r4859] Removal of EII_necessary out of the SCF cylcle

2006-03-08 13:36  jgh

	* [r4857] Simple_eigensolver: use cp_fm_syevd for cases with nmo >
	  nao/4

2006-03-08 13:35  jgh

	* [r4856] use plane_distance

2006-03-08 13:35  jgh

	* [r4855] Use older routine plane_distance, remove
	  cell_min_image_radius

2006-03-08 12:48  vondele

	* [r4854] further small layout improvements

2006-03-08 10:01  vondele

	* [r4853] move calculate_ecore_overlap / self out of the scf

2006-03-07 19:21  vondele

	* [r4852] give calculate_first_density_matrix its own routine
	  qs_initial_guess.F.
	  Lots of further refactoring potential here ...

2006-03-07 17:50  jgh

	* [r4851] New input defaults for SE parameters
	  Consistent calculation of interaction range
	  Reset and update all regtests

2006-03-07 17:25  tkuehne

	* [r4849] HMC: FILL_WF_HISTORY and SOFT_START options added as
	  suggested by Fawzi

2006-03-07 17:23  tkuehne

	* [r4848] Relocating the frozen density block out of the SCF loop,
	  as correctly annotated by Joost

2006-03-07 16:04  tlaino

	* [r4847] reorganization of the constraint treatment.. splitting of
	  routines in different files

2006-03-07 15:50  vondele

	* [r4846] further cleaning, making the use of scf_env a little more
	  consistent (i.e. towards the usual create/retain/release
	  touch-and-feel).

2006-03-07 09:47  jgh

	* [r4844] Semi-empirical code now fully parallel

2006-03-07 09:19  tlaino

	* [r4843] Bug fix in metadynamics read input section.. doesn't
	  affect regtests.
	  deleting unused keywords in 2 regtests input files.

2006-03-07 07:43  vondele

	* [r4841] small changes

2006-03-06 19:29  vondele

	* [r4840] split off qs_scf_print_summary

2006-03-06 18:31  vondele

	* [r4839] another small clarification of ot

2006-03-06 17:59  vondele

	* [r4838] factor out qs_scf_loop_do_general_diag,
	  qs_scf_loop_do_special_diag , qs_scf_loop_do_ot

2006-03-06 15:10  vondele

	* [r4837] take a big blob of printing out of the scf loop, move to
	  qs_scf_loop_print

2006-03-06 14:17  vondele

	* [r4836] another round of cleaning

2006-03-06 13:15  tkuehne

	* [r4835] Minor revision

2006-03-06 11:45  vondele

	* [r4831] moves update_wfi in qs_scf, the new location seems more
	  logical. This influences the extrapolation of the density matrix
	  for the initial guess during MD, but should matter very little at
	  SCF convergence, regtests typically affected by O(1E-10)-O(1E-14)
	  (34 resets).

2006-03-06 11:11  vondele

	* [r4830] further fine tuning of typo_match

2006-03-06 10:30  marcella

	* [r4829] possibly wrong deallocation

2006-03-06 09:43  tkuehne

	* [r4828] Missing DEALLOCATE added

2006-03-06 08:04  tlaino

	* [r4827] recovered a deprecated constrained.. All bonds involving
	  hydrogen can be constrained with the
	  &HBONDS section. Added two descriptive regtests

2006-03-06 00:22  tkuehne

	* [r4826] Make EIP "parallel consistent", e.g. the parallel version
	  produces the same results, since the whole calculation is
	  redundantly performed on all CPUs.

2006-03-05 21:55  gtb

	* [r4824] kg biasing for mc qs

2006-03-05 13:34  vondele

	* [r4823] shorten TRACE output line

2006-03-05 13:03  vondele

	* [r4822] adding timeset/timestop pairs and some reformatting

2006-03-05 11:33  vondele

	* [r4821] another round of qs_scf cleaning (split off
	  qs_scf_methods)

2006-03-05 09:45  vondele

	* [r4820] added a constraint ddapc functional form

2006-03-04 12:09  vondele

	* [r4819] made external_file a variant for restart_file_name in the
	  &EXT_RESTART block

2006-03-04 10:25  vondele

	* [r4818] updated some comments

2006-03-04 08:32  vondele

	* [r4817] unified print of the various atomic charges

2006-03-03 20:57  tlaino

	* [r4816] bug fix for forces ddapc restraint.. reset regtest

2006-03-03 19:14  tlaino

	* [r4815] removing a write statement

2006-03-03 18:56  tlaino

	* [r4814] bug fix for energy definition in DDAPC restraint (thanx
	  Joost).. resetting regtest
	  moving the section involving ddapc to "normal" section and
	  explicitely defining a print_key
	  section for controlling otuput

2006-03-03 15:05  vondele

	* [r4813] better bug fix

2006-03-03 15:03  vondele

	* [r4812] bug fix

2006-03-03 14:35  tlaino

	* [r4811] Activated the restraint on DDAPC inserted also a regtest
	  testing the new feature

2006-03-03 13:18  vondele

	* [r4810] remove unused cp_keywords.F

2006-03-03 08:45  tkuehne

	* [r4809] Copy & Paste Error

2006-03-02 22:04  tkuehne

	* [r4808] Minor changes on HMC & TBMD + EIP

2006-03-02 20:55  vondele

	* [r4807] prettify

2006-03-02 20:38  vondele

	* [r4806] some additional timeset/timestop's

2006-03-02 20:25  vondele

	* [r4805] typo

2006-03-02 20:15  vondele

	* [r4804] cleanup, getting rid of checkpoint_handler

2006-03-02 19:20  vondele

	* [r4803] some cleaning

2006-03-02 17:26  jgh

	* [r4801] Test for semi-empiric not EWALD within SE

2006-03-02 14:24  tlaino

	* [r4800] Another little step towards a more logical setup for
	  CONSTRAINTS.. changing the old syntax
	  now constraint of distances, and etc. are accessed only through
	  the collective variables (COLVAR)
	  definition. bug fix for printing in cp_ddapc.F

2006-03-02 12:27  vondele

	* [r4799] first version of a framework for ddapc restraints.

2006-03-02 11:26  tlaino

	* [r4798] bug fix

2006-03-02 09:02  vondele

	* [r4797] lowdin only at high print levels

2006-03-01 13:19  tlaino

	* [r4796] reorganizing the three geometry optimizer..

2006-03-01 11:19  vondele

	* [r4795] spell checker refinements and bug fixes

2006-02-28 13:17  vondele

	* [r4794] added spell checker (thanks Teo).

2006-02-28 09:20  vondele

	* [r4793] use FFTW definition of number of flops (log_2 instead of
	  log)

2006-02-27 16:36  tlaino

	* [r4792] bug fix NONE coupling (QMMM) + PERIODIC keyword. new
	  regtest

2006-02-27 10:37  tlaino

	* [r4791] ops.. real bug fix for dipole.. and deleting the RESTART
	  keyword from GLOBAL section.
	  not used anymore!

2006-02-27 09:38  tlaino

	* [r4790] bug fix for uninitialized variable in calc_dipole

2006-02-26 10:34  fawzi

	* [r4789] oops forgotten to add...

2006-02-25 21:38  fawzi

	* [r4788] * better handling if incompatible (old) inputs
	  * path integral (not yet fully there)

2006-02-25 14:24  jgh

	* [r4787] Correct(?) cutoff to N for non-orthorhombic cells

2006-02-25 14:23  jgh

	* [r4786] replace dlog and dexp

2006-02-25 14:23  jgh

	* [r4785] towards the FFT MKL interface

2006-02-24 22:22  mcgrath

	* [r4784] Bug with pointers that weren't referenced in GEMC

2006-02-24 21:50  krack

	* [r4783] Bug fix for PERIODIC NONE also for FIST, KG, and TBMD
	  Printing of MM neighbor list like for QS

2006-02-24 20:33  tlaino

	* [r4782] fixing backslash for PGI compiler

2006-02-24 20:23  vondele

	* [r4781] also get the LDA case right

2006-02-24 20:03  tlaino

	* [r4780] writing 8 instead of 10 integers per line in dumped input
	  file/ restart file

2006-02-24 16:19  marcella

	* [r4779] current density on the pw grid
	  nmr shift in gpw, the g=0 component is still missing
	  restart lsd from lda

2006-02-24 15:56  vondele

	* [r4778] add some version info the the restart file (this might be
	  useful later)

2006-02-24 14:56  vondele

	* [r4777] try to fail more gracefully upon external restart format
	  differences

2006-02-24 11:17  krack

	* [r4776] * Bug fix for PERIODIC NONE
	  * Full QS neighbor list printing recovered (parallel output)
	  * GAPW neighbor list print keys added
	  * Reset for all QMMM/QS regtest inputs
	  * Subcell bug fix for general (non-orthorhombic) cells also
	  introduced in the QS neighbor list (FIST was already fixed)

2006-02-24 09:36  vondele

	* [r4775] if external control forces an exit, do not append to
	  trajectories of write MD restarts

2006-02-22 20:28  vondele

	* [r4774] more input doc

2006-02-22 20:16  vondele

	* [r4773] input docs

2006-02-22 14:55  vondele

	* [r4772] lowdin charges

2006-02-22 14:08  jgh

	* [r4770] Make code single precision ready (check random number
	  generators!)

2006-02-22 14:02  tlaino

	* [r4769] bug fix in the manual for metadynamics

2006-02-22 11:06  tlaino

	* [r4768] tiny bug dix in restart (no effects at all).. moving the
	  trasnformation.F file to the attic (it was not used all around)

2006-02-22 10:13  vondele

	* [r4766] switch cp_fm_power to use syevd

2006-02-22 09:18  jgh

	* [r4765] Making code NAG -nan save

2006-02-22 07:46  tlaino

	* [r4764] ops.. this is correct!

2006-02-22 07:44  tlaino

	* [r4763] re-organized metadynamics module. Using the same layout
	  restart as the rest of cp2k,
	  both for extended langrange and standard metadynamics.
	  Added more regtests testig the several restart features.
	  Bug fix in restart for coordinates (now really using long_char
	  type).
	  Bug fix while updating restarts in linked list access. no
	  differences in regtests with NAG compiler

2006-02-21 16:55  jgh

	* [r4761] Bug fixes for __FFTSGL code

2006-02-21 16:19  jgh

	* [r4760] Remove debug code

2006-02-21 16:08  jgh

	* [r4759] A bug fix - size of scrtach array
	  single precision FFT version

2006-02-21 12:20  jgh

	* [r4757] Bug fix serial version

2006-02-21 11:18  jgh

	* [r4756] Better and correct handling of extreme parallel cases

2006-02-21 09:19  tlaino

	* [r4755] using shorter names

2006-02-20 19:04  tlaino

	* [r4754] fixing uninitialized variable for long_char type and a
	  bug fix in _get_rest of the linked list

2006-02-20 14:08  tlaino

	* [r4753] zeroing cmplx

2006-02-20 13:48  tlaino

	* [r4752] zeroing only when it's really needed

2006-02-20 09:55  vondele

	* [r4748] added print key mo_magnitude

2006-02-20 08:33  tlaino

	* [r4747] bug fix for FFT (new interface) coming out when processes
	  have no g-vector (ngpts_cut_local = 0)

2006-02-19 20:32  tlaino

	* [r4745] Fixing the mismatching of the echoed input and input.
	  Using list-directed IO to print numbers
	  in the echoed and in the restart file. Resetting corresponding
	  regtests

2006-02-18 17:47  tlaino

	* [r4744] Some cleaning

2006-02-17 22:02  vondele

	* [r4743] test

2006-02-17 17:41  jgh

	* [r4741] Protect FFT calls in case of zero length and/or
	  transforms

2006-02-17 10:25  vondele

	* [r4740] added standard and generic interfaces to BLAS. USE
	  f77_blas will provide all of these. I.e. BL_COPY will now resolve
	  to DCOPY,SCOPY,ZCOPY,CCOPY,DSCOPY,SDCOPY,ZCCOPY,CZCOPY.

2006-02-17 10:19  vondele

	* [r4739] small input documentation patch

2006-02-17 10:04  tlaino

	* [r4738] qmmm restart bug fix for labels..

2006-02-17 07:52  vondele

	* [r4736] prettify

2006-02-17 00:42  cjmundy

	* [r4735] *Fixed some print keys for writing through md_energies.F
	  *Fixed compile time failure for xas_tp_scf.F. Blessed is the DEC

2006-02-16 20:10  vondele

	* [r4732] mulliken

2006-02-16 19:20  mcgrath

	* [r4731] Trying to get KG biasing working with QS...

2006-02-16 18:55  mcgrath

	* [r4730] Small bugfix for SILENT runs

2006-02-16 14:06  vondele

	* [r4729] Clean up the output in case of an error in the input. The
	  input description won't be printed anymore, a link to the webpage
	  is given instead.

2006-02-16 10:16  vondele

	* [r4728] explicitly include the fftw_f77.i file, as it is not
	  installed with fftw2, and can only be found in the build
	  directories, which might not be present

2006-02-16 08:57  krack

	* [r4727] default printout of BASIC_DATA_TYPES and PHYSCON
	  disabled; minor adjust in format

2006-02-15 16:20  jgh

	* [r4725] New keyword ALLTOALL_SGL : allows for reduced precision
	  communication

2006-02-15 10:05  vondele

	* [r4722] test

2006-02-15 09:55  vondele

	* [r4721] test

2006-02-15 09:26  jgh

	* [r4718] New Interface for FFT Libraries
	  - creates a F77 type library in "lib"
	  - supported libraries:
	  Goedecker : default
	  FFTW2 : needs path to include file!
	  FFTW3 : needs path to include file! (only FFTW2 or FFTW3 can be
	  used - conflicts)
	  ACML : seems only to work for serial jobs, unknown bug for
	  parallel
	  ESSL :
	  MKL and SCILIB : only stubbs at the moment
	  - Makefile has been changed!
	  - reset of some regtests, very small differences in energy

2006-02-15 00:42  marcella

	* [r4717] erase dbg write

2006-02-15 00:38  marcella

	* [r4716] bug fix, print gpw current, nmr variables

2006-02-15 00:01  tlaino

	* [r4715] speed-up of the dump vals_write

2006-02-14 19:22  tlaino

	* [r4714] speeding-up also the dumping of rngs..

2006-02-14 16:05  tlaino

	* [r4713] another little problem with format in dumping restart
	  file
	  avoiding the writing of RNG info if not required

2006-02-14 15:45  tlaino

	* [r4712] ops... recoverign old pw_pool_types..

2006-02-14 15:43  tlaino

	* [r4711] tiny bug fix for format

2006-02-14 14:21  krack

	* [r4710] bug fix: IF block bypassing an mp_sum removed

2006-02-14 10:47  tlaino

	* [r4709] New routines (faster) to dump coordinates and velocities
	  in the restart file

2006-02-14 08:17  vondele

	* [r4708] one more step towards no localLog files at MEDIUM

2006-02-14 07:55  vondele

	* [r4707] towards getting no localLog files below IOLEVEL HIGH

2006-02-13 16:04  vondele

	* [r4702] make the qm neighborhoodlist non-periodic (for qmmm)

2006-02-13 14:46  vondele

	* [r4701] rename read_cell_new into read_cell

2006-02-13 14:11  vondele

	* [r4700] remove init_qmmm_cell

2006-02-13 11:28  tlaino

	* [r4699] Getting rid of ATOM module

2006-02-13 08:43  tlaino

	* [r4698] new collective variable distance plane-point..
	  Implemented also the angle (was not implemented)

2006-02-10 16:21  tkuehne

	* [r4695] TBMD: A subtle change

2006-02-10 12:32  tkuehne

	* [r4694] TBMD: Things are getting better

2006-02-10 09:44  tkuehne

	* [r4693] TBMD: Kinetic energy calculation

2006-02-09 18:46  tkuehne

	* [r4692] TBMD: Debugging

2006-02-09 18:25  fawzi

	* [r4691] oops

2006-02-09 17:47  fawzi

	* [r4690] removed required from LD_STEPS
	  now HMC section does not "appear" in the restart if not needed

2006-02-09 13:52  tkuehne

	* [r4689] TBMD: Print functions for the sparsity pattern of the
	  H-Matrix and it's derivatives

2006-02-09 13:51  tkuehne

	* [r4688] TBMD: Bug fixes - Much better now, but still wrong.

2006-02-09 13:49  tkuehne

	* [r4687] Bug fix in the FIST neighbor-list output

2006-02-08 18:44  tkuehne

	* [r4686] TBMD: Debugging of the tbmd neighbor-list

2006-02-08 14:37  tkuehne

	* [r4685] TBMD: Quick & Dirty printing routines. For debugging
	  purposes only!

2006-02-08 10:19  tkuehne

	* [r4684] TBMD: Print functions & first debugging

2006-02-08 08:18  tlaino

	* [r4683] bug fix for systems with no-electrons + DEBUG_PRINT
	  LEVEL,
	  general bug fix for multipole decoupling as PRINT_key..
	  no changes in results

2006-02-07 22:51  tlaino

	* [r4682] Fixing the PDB format in output

2006-02-07 19:19  vondele

	* [r4681] fix recent units bug in the time written to file

2006-02-07 15:44  tlaino

	* [r4680] bu fix for multipole decoupling scheme, resetting the
	  proper regtests

2006-02-07 15:37  marcella

	* [r4679] gpw current density: collocated on the grid but not yet
	  tested

2006-02-07 14:17  fawzi

	* [r4678] fixed spurius "," in format

2006-02-07 11:10  fawzi

	* [r4676] fixes wrt. long strings in input

2006-02-07 11:09  fawzi

	* [r4675] cp_error_check now more verbose in case of failure

2006-02-07 11:09  fawzi

	* [r4674] reverting my previous bugfix

2006-02-07 10:33  vondele

	* [r4673] prettified

2006-02-07 00:14  fawzi

	* [r4672] long chars (i.e. whole lines of input) are now supported
	  by the input.

2006-02-06 14:52  vondele

	* [r4668] fix presence of both moduleN and module_name

2006-02-03 15:44  tlaino

	* [r4662] tiny bug fix..

2006-02-03 13:48  tlaino

	* [r4661] bug fix for parallel runs

2006-02-03 10:51  tlaino

	* [r4659] recovering the original modifications

2006-02-03 09:10  marcella

	* [r4658] sorry

2006-02-03 09:04  tlaino

	* [r4657] Temporary bug fix

2006-02-03 08:16  tlaino

	* [r4656] Step number and time bug fix for restart.. fixing also
	  the accepted_md and md_approach_nr in the HMC module

2006-02-02 23:37  marcella

	* [r4655] towards the calculation of the response current density

2006-02-02 15:34  tkuehne

	* [r4654] TBMD: First numbers are produced, but nothing conserved
	  and everything way too high

2006-02-02 15:32  krack

	* [r4653] avoid init with lcleb=0

2006-02-02 11:25  tlaino

	* [r4651] Some cleaning and bug fixes for the dumping of the
	  restart file.. still few things to fix..

2006-02-02 08:26  vondele

	* [r4650] try to change permission

2006-02-01 23:07  vondele

	* [r4649] ener file also at low print level

2006-02-01 22:53  tlaino

	* [r4648] Memory leaks fixed.. also a format string coming out on
	  IBM

2006-02-01 21:13  tlaino

	* [r4647] NEW RESTART format.. in this commit unfortunately there
	  are several bug fixes:
	  (1) bug fix to the cp_linked_list
	  (2) bug fixes to attributes of several input sections
	  (3) double initialization of velocities if restart only POS
	  
	  The new restart format is now completed.. the old routines have
	  been deleted and should for
	  no reason be restored.. 2 fancy things:
	  The restart file can be used both as restart file than as input
	  file (this thanks to the new input style)
	  
	  The corresponding regtests have been resetted because of numerics
	  of the order of 10^-9/10^-10 (with exclusion of the
	  pos restart and the following depending on it) since in the old
	  MD module the velocity were initialized 2 times if restart
	  POS was required.. now just 1 time..
	  
	  The next checkins will be more towards the cleaning of the old
	  RESTART stuff..

2006-02-01 19:44  marcella

	* [r4645] to avoid zero projectors in GAPW and change default to
	  no-(2 and 3)-center terms

2006-02-01 18:44  tkuehne

	* [r4644] A subtlety with the neighbor-lists is now solved

2006-02-01 16:09  tkuehne

	* [r4643] Band-structure term evaluation

2006-02-01 12:09  vondele

	* [r4642] add some commented code to check parallel load balancing

2006-02-01 12:08  vondele

	* [r4641] do not add apples to oranges

2006-02-01 11:56  tkuehne

	* [r4640] TBMD: Memory leak

2006-02-01 09:40  tkuehne

	* [r4639] TBMD: Bug fixes

2006-02-01 07:50  tlaino

	* [r4637] This is the correct bug fix.. If you want to use
	  checkpoint utility you MUST BE SURE that the error
	  variable is correctly propagated.. oytherwise it is useless.. in
	  this case those calls should be deleted!

2006-02-01 07:43  vondele

	* [r4636] fix failing regtests

2006-01-31 20:29  cjmundy

	* [r4633] Leak fix

2006-01-31 17:16  tkuehne

	* [r4632] Dummy check-in

2006-01-31 15:44  tkuehne

	* [r4631] Diagonalization & matrix pools

2006-01-31 15:37  mcgrath

	* [r4630] Needed to change some values...should work with ref_cell
	  now

2006-01-31 14:59  vondele

	* [r4629] mostly some cleaning of pointer variables that were not
	  dereferenced

2006-01-31 14:21  cjmundy

	* [r4628] * Ref grid now works
	  * NP* now included in QS velocity verlet options
	  * Towards a single method to compute cut-offs
	  (consistently for all methods)
	  * Regtests fine EXCEPT for MC_QS.inp. I will
	  work with Matt to find this.

2006-01-31 11:21  marcella

	* [r4627] cleaning

2006-01-31 09:53  tkuehne

	* [r4626] Initialization of blacs_env & para_env

2006-01-30 21:26  tkuehne

	* [r4625] BLACS

2006-01-30 17:07  tkuehne

	* [r4624] Not working yet: The matrix pool stuff is missing!

2006-01-30 15:28  tkuehne

	* [r4623] Init stuff for diagonalization

2006-01-30 10:29  tkuehne

	* [r4622] Pair-repulsion term added.

2006-01-28 19:36  tlaino

	* [r4620] minor rewriting of the generate_molecule routine for
	  qm/mm cleaning the logical syntax..
	  writing under print_keys for qmmm_topology_util

2006-01-28 00:28  tlaino

	* [r4618] Improved the check on the generate_molecule routine

2006-01-27 17:46  mcgrath

	* [r4617] NAG complained...added the error type to more MC routines

2006-01-27 11:45  tlaino

	* [r4614] bug fix for parallel run..

2006-01-27 11:42  vondele

	* [r4613] mo_orthonormality to the high print level

2006-01-27 11:08  tkuehne

	* [r4612] Minor change

2006-01-27 09:51  tlaino

	* [r4611] another tiny bug fix..

2006-01-27 08:15  tlaino

	* [r4610] bug fix.. restored the parallel version

2006-01-27 00:07  tkuehne

	* [r4609] The missing scaling functions
	  No guarantee, but Hamiltonian- & dH/dR-Matrix built-up for Si1000
	  in ~0.01sec@1CPU!

2006-01-26 22:12  tkuehne

	* [r4608] dH/dR matrix construction: done!

2006-01-26 22:05  tlaino

	* [r4607] Preliminary operations to set-up the new restart file
	  structure..
	  Introduced the iterations levels into GEO-OPT
	  Deleted mdio structure.. Starting from now the file will be
	  opened only using the
	  cp_print_key_unit_nr ..
	  Very soon the RESTART file..

2006-01-26 20:26  tkuehne

	* [r4606] Tight-Binding: Mostly done!

2006-01-26 19:30  gtb

	* [r4605] paralleb bug

2006-01-26 18:45  vondele

	* [r4604] autoclean test

2006-01-26 18:43  vondele

	* [r4603] test

2006-01-26 17:26  tkuehne

	* [r4602] Bug fix

2006-01-26 16:39  fawzi

	* [r4601] fixing missing argument declarations

2006-01-26 16:11  tlaino

	* [r4600] fixed bug, echo now after check of input (f)

2006-01-26 15:55  krack

	* [r4599] improved handling of special keywords for XML dump

2006-01-26 15:45  tkuehne

	* [r4598] Initialization & allocation of the H & dH/dR matrices

2006-01-26 15:43  marcella

	* [r4597] memory leaks

2006-01-26 15:42  tlaino

	* [r4596] fixing misnamed variable(fawzi)

2006-01-26 14:28  marcella

	* [r4594] move line

2006-01-26 13:39  fawzi

	* [r4592] moved input check to its own module.

2006-01-26 13:37  tkuehne

	* [r4591] TBMD neighbor-list traversal for Hamiltonian built-up

2006-01-26 13:33  fawzi

	* [r4590] * echo input as input works and does the same thing.
	  * various input related fixes
	  * external restart
	  * first (non working) replica types

2006-01-26 11:11  marcella

	* [r4589] fix of some integrals and new options for xas

2006-01-26 08:05  vondele

	* [r4587] fix a TAB issue

2006-01-26 00:46  gtb

	* [r4586] bug fix in parallel version

2006-01-25 18:12  vondele

	* [r4585] adjust format for printing

2006-01-25 17:43  krack

	* [r4584] include _SECTION_PARAMETERS_ and _DEFAULT_PARAMETERS_
	  into the XML dump

2006-01-25 16:26  tkuehne

	* [r4582] Debugging and memory leak finxing

2006-01-25 14:19  vondele

	* [r4581] moved mm_dipole to medium print level

2006-01-25 14:08  vondele

	* [r4580] fix output handling for debug in parallel

2006-01-25 11:20  tkuehne

	* [r4577] Still TBMD neighbor-lists

2006-01-25 08:39  tlaino

	* [r4576] bug fix for lbfgs in parallel

2006-01-25 05:51  tlaino

	* [r4575] removing force_env_list: because we have another
	  force_env_p_types (i.e. sub_force_env) and
	  particularly because since its introduction March 2004 it was
	  never used!

2006-01-24 23:33  tkuehne

	* [r4574] Fix of a memory leak in the TBMD input section

2006-01-24 23:06  mcgrath

	* [r4573] Minor fixes to produce identical pdbg and sdbg results
	  (old way was not wrong, just different) and stop some Intel
	  complaints

2006-01-24 19:00  tlaino

	* [r4570] Keeping up-2-date The template and instantiation files..
	  removing unused routines.
	  Please don't forget NEVER modify .F files produced with
	  .instantiation files.

2006-01-24 18:41  vondele

	* [r4569] fix build

2006-01-24 18:17  tkuehne

	* [r4568] TBMD neighbor-lists

2006-01-24 18:00  marcella

	* [r4567] linres preconditioners

2006-01-24 17:55  vondele

	* [r4566] allow fixing atoms by molecule name (this name is
	  non-trivial right now)

2006-01-24 15:13  fawzi

	* [r4562] xc tpss pbe rewrite

2006-01-24 15:12  fawzi

	* [r4561] bugfix (non initialization)

2006-01-24 12:44  vondele

	* [r4560] md.F: bugfix for printout, format fix
	  topology_pdf.F: make timestop last call

2006-01-24 12:34  tkuehne

	* [r4559] Changes concerning the Tight-Binding neighbor-list +
	  corresoponding print key

2006-01-24 11:22  tlaino

	* [r4558] Geometry optimizer now have the all the same output
	  structures (BFGS,LBFGS,CG)..
	  The output of the history of the geometry can be controlled with
	  the new input structure
	  It is realted to the TRAJECTORY print_key in &MOTION

2006-01-24 09:27  vondele

	* [r4557] improved error message (still not good)

2006-01-24 09:12  tkuehne

	* [r4556] Most of the TBMD neighbor-list stuff

2006-01-24 09:12  tlaino

	* [r4555] restoring the original timings in SCF

2006-01-23 18:03  tlaino

	* [r4552] Removed the old parser and also some unused I/O routines

2006-01-23 16:44  tlaino

	* [r4551] keyword bug fix

2006-01-23 16:25  tlaino

	* [r4550] little not really bug fix: added a threshold to overcome
	  the fact that sometime the timing
	  function produces a division by zero..

2006-01-23 15:38  tkuehne

	* [r4549] Minor bug fixes for the fortcoming tight-binding method.

2006-01-23 14:34  tlaino

	* [r4548] few bug fix for library_test

2006-01-23 13:53  vondele

	* [r4546] topology_psf.F: printkey bug fix when dumping psfs,
	  topology.F: fix comment

2006-01-23 13:23  tlaino

	* [r4545] Enabling the TEST program of CP2K.. converted
	  corresponding inpt file

2006-01-23 13:22  vondele

	* [r4544] further improved error messages

2006-01-23 12:45  vondele

	* [r4543] clarify error message

2006-01-23 10:44  tkuehne

	* [r4542] Changes due to Teo's new print-keys

2006-01-23 10:17  krack

	* [r4541] print xray data file name to stdout

2006-01-23 00:07  tlaino

	* [r4539] Old print keys COMPLETELY REMOVED. WOOOOOOOOOF!!!!

2006-01-22 20:27  krack

	* [r4538] just a small bug fix before Jack the dull boy continues
	  to check in

2006-01-22 20:01  tlaino

	* [r4537] More print_keys.. "all work and no play make Jack a dull
	  boy" (cultural reference)

2006-01-22 15:50  vondele

	* [r4536] explain why debug printing might change slightly results
	  (part 1/n)

2006-01-22 15:32  tlaino

	* [r4535] more print_keys.. bug fix for printing orthonormality in
	  KG_GPW and in SEMIEMPIRICAL

2006-01-22 09:32  tlaino

	* [r4534] setting the remaining print_keys to false to allow
	  debugging..

2006-01-22 07:34  tlaino

	* [r4533] More print_keys converted.. added headers for qm/mm and
	  TDDFPT .. some cleaning

2006-01-21 20:44  vondele

	* [r4532] tried to clarify warning

2006-01-21 14:28  tlaino

	* [r4531] print_keys, fixed and reorganized (moved few print_keys
	  in subsys).. some cleaning around..
	  cosmetics!

2006-01-20 18:01  marcella

	* [r4530] test with s-1

2006-01-20 15:20  tkuehne

	* [r4529] Additional stuff tight-binding

2006-01-20 10:06  tlaino

	* [r4528] simulation_cell -> cell_types :: just style. other
	  print_keys converted.
	  Error variable propagated further
	  Cosmetics

2006-01-19 20:35  tkuehne

	* [r4527] Orthogonal Tight-Binding initialization

2006-01-19 20:21  vondele

	* [r4526] what about using data parsed in the input?

2006-01-19 17:56  marcella

	* [r4525] if linres write ground state energy first

2006-01-19 16:38  krack

	* [r4524] arg local dropped, since it seems to be misplaced and
	  causes unnecessarily
	  warnings

2006-01-19 16:35  tlaino

	* [r4523] fixing label to be more precise

2006-01-19 16:34  tlaino

	* [r4522] tiny bug fix ;-)

2006-01-19 16:34  krack

	* [r4521] print also grid size

2006-01-19 15:06  tlaino

	* [r4520] Little step towards the new restart format. Introduced
	  the correct dipole computed with the berry phase for Fist

2006-01-19 07:44  tlaino

	* [r4519] Introducing a Pwarning function connected to the print
	  WARNINGS section.. Please
	  Use this function everytime you want to advice the user for
	  something to take care of..
	  Fixing the new input style for the pritning of the cube files...
	  removed other print_keys..
	  MAking more logical the print of LOCALIZATION and
	  MOLECULAR_STATES.. Now both these section have a CUBES
	  section controlling the printing of their cube files.. don't need
	  anymore to insert also &MO_CUBES (it
	  was completely unlogical!!!!)
	  Making the name of the file opened by cp_print_key_unit_nr more
	  friendly..

2006-01-18 18:33  tkuehne

	* [r4517] Orthogonal Tight-Binding

2006-01-18 15:22  tlaino

	* [r4515] If a geometry is provided with PDB or XYZ file (the only
	  two supported external format) and also a &COORD section
	  is present, than &COORD overwrites all information regarding
	  position of the atoms..

2006-01-18 14:42  tlaino

	* [r4514] New print_key section to control the timings print at the
	  end of cp2k.
	  THRESHOLD defines the % of cputime above which the rouinte will
	  be printed.

2006-01-18 14:20  krack

	* [r4513] - correct unit conversion for XML output (from CP2K
	  units)
	  - reasonable default values for CUTOFF, TIMESTEP, and TEMPERATURE
	  due
	  to appropriate unit conversion

2006-01-18 10:18  tlaino

	* [r4511] Removing the PM* printing definition. Now the information
	  to print something during the execution can be accessed with the
	  new input style. converted also the cputime print and the memory
	  print.
	  Main change: diffusion of the error variable through the cp2k
	  code..
	  Very soon the restart ;-) Stay Tuned!

2006-01-17 21:47  tkuehne

	* [r4510] Several new types for orthogonal tight-binding

2006-01-17 16:32  fawzi

	* [r4509] * quick fix for pbe lsd bug

2006-01-17 16:04  tkuehne

	* [r4508] Removal of the objectionable string

2006-01-17 13:24  tkuehne

	* [r4507] Orthogonal Tight-Binding

2006-01-17 13:16  tkuehne

	* [r4506] Orthogonal Tight-Binding

2006-01-17 11:35  marcella

	* [r4505] linres: optimization algorithm

2006-01-16 22:37  tlaino

	* [r4504] Using the new parser to pars XYZ and PDB

2006-01-13 19:42  marcella

	* [r4501] nmr variables

2006-01-13 15:28  fawzi

	* [r4500] fixing bug wrt. number of points (thanks Joost)

2006-01-13 14:25  fawzi

	* [r4499] stride in cube output, mulliken charges back to medium
	  print level.

2006-01-12 20:11  mcgrath

	* [r4497] Set up for production, not EV curves

2006-01-11 16:49  fawzi

	* [r4493] * cleaned up grid initialization
	  * fine xc grid option
	  * removed slater_koster files from OBJECTDEFS

2006-01-11 09:56  vondele

	* [r4492] fix printing of MOs

2006-01-11 00:37  marcella

	* [r4490] variables for nmr

2006-01-10 17:21  tlaino

	* [r4488] rewriting of the move_file routine and moved to cp_files
	  for consistency (Thanks Joost).

2006-01-10 13:01  tlaino

	* [r4487] Removing the specific system call.. Please don't use
	  nevermore this.. In few architectures like catamount
	  the SYSTEM CALL cannot be used! Let's try to be as independent as
	  possible from these system calls.

2006-01-10 11:57  vondele

	* [r4486] remove a repeated output

2006-01-10 06:57  vondele

	* [r4485] indicate the 'real fix', i.e. there should be no system
	  calls out of the machine files as the
	  argument to system is intrinsically non-portable.

2006-01-09 21:43  krack

	* [r4484] Printing adjusted and WRITE statements merged.
	  I hope this will work with all compilers.

2006-01-09 21:41  krack

	* [r4483] This is a quick hack to allow the CALL SYSTEM() in this
	  module using the NAG compiler.
	  Such a direct call to SYSTEM should not be used here and a
	  m_system wrapper
	  routine in the machine(_arch) modules would be an appropriate
	  solution.

2006-01-09 15:46  mcgrath

	* [r4481] Some info doesn't need to be printed for SILENT
	  runs...want to do the same for pw_grids, but globenv isn't
	  there...

2006-01-09 15:25  vondele

	* [r4480] please no output that is quadratic with system size below
	  high print level !

2006-01-09 11:33  tlaino

	* [r4478] CG bug fix and improvements (resetting CG tests).
	  WALLTIME now available to all
	  RUN_TYPE. Cleaning of the old WALLTIME.

2006-01-09 10:01  vondele

	* [r4476] introduce SVD for GAPW, and new default radii for H.

2006-01-08 23:00  tlaino

	* [r4475] fixing external_control termination for lbfgs and cg
	  optimization

2006-01-08 11:08  tlaino

	* [r4474] fixing timings

2006-01-07 00:39  ikuo

	* [r4473] Move md restart file over to ".bak" instead of overwrite

2006-01-06 23:47  mcgrath

	* [r4472] Bug fix for AVBMC with biasing

2006-01-05 00:31  ikuo

	* [r4468] OpenMP fix for DEC

2006-01-04 22:30  ikuo

	* [r4467] Make it Dec compiler friendly.

2006-01-04 15:47  vondele

	* [r4463] some cleaning to help the prettifier

2006-01-04 15:10  vondele

	* [r4461] must be zen ... separated colvars from metadynamics and
	  moved them to subsys

2006-01-03 22:28  tlaino

	* [r4458] fixing memory leaks

2006-01-03 00:35  tlaino

	* [r4457] still some more work..

2006-01-02 23:10  tlaino

	* [r4456] minor rewriting of few loops for speed-up in parallel..
	  resetting related regtests due to
	  10^-11 numerical differences (reordering of instructions)

2006-01-02 12:52  vondele

	* [r4455] remove redundant ncolvar from input

2006-01-02 12:03  vondele

	* [r4454] removed TABs

2006-01-02 11:56  vondele

	* [r4453] prettified

2006-01-02 11:50  vondele

	* [r4452] giving colvar its own section for later reuse

2006-01-02 09:28  vondele

	* [r4451] separating colvar and metadynamics a bit (towards
	  constraints). Prettifying these files

2006-01-01 22:23  tlaino

	* [r4450] get few print keys working: atomic_coordinates,
	  interatomic_distances, mo_cartesian, mo_occupation,
	  mo_eigenvalues, mo_eigenvectors.

2006-01-01 21:05  vondele

	* [r4449] get rid of cp_prep_globals.h, use cp_common_uses.h
	  instead (this avoids confusion about which procedures to include
	  from cp_log_handling and cp_error_handling, which depended on the
	  preprocessor settings). If you've locally modified
	  cp_prep_globals.h you'll need to put those changes in
	  cp_common_uses.h

2006-01-01 20:16  vondele

	* [r4448] add IMPLICIT NONE

2005-12-30 19:48  tlaino

	* [r4447] bugfix for parallel execution for fist and qm/mm. ->
	  parser_get_next_line

2005-12-30 13:50  vondele

	* [r4446] bugfix: seems like res can be returned uninitialised,
	  leading to random printing

2005-12-30 08:16  tlaino

	* [r4445] bug fix for real part of the ewald contribution.. regtest
	  ok!

2005-12-30 06:24  tlaino

	* [r4444] (1) Fast construction of the Gterm for the periodic QM/MM
	  and also for the bloechl decoupling.
	  An additional keyword (for this last case) was added
	  (ANALYTICAL_GTERM) to retrieve the original
	  Ewald scheme. No analytical procedure has been left for QM/MM (th
	  enew interpolated Spline is highly reliable and
	  accurate).
	  (2) Bug fix printing atomic density derived charges without
	  providing the section
	  Added relative new regtests

2005-12-27 19:56  tlaino

	* [r4443] Get rid of the old parser. Improving the reading of the
	  CHARMM FF. commenting the code that was using the old parser in
	  the ATOM module.

2005-12-27 10:09  tlaino

	* [r4442] bug fix for setting element in atomic kind.. downgrading
	  failure_level to warning_level for check_subsys_element

2005-12-23 16:00  krack

	* [r4441] Works now also in parallel

2005-12-23 15:59  krack

	* [r4440] gsq_max argument added (controls the maximum g vector
	  modulus calculated)

2005-12-23 15:58  krack

	* [r4439] Print format adjusted (only GAPW)

2005-12-23 15:57  krack

	* [r4438] wrapper routines for mpi_gatherv added

2005-12-23 13:32  tchassai

	* [r4437] Bugfix

2005-12-23 13:30  fawzi

	* [r4436] higher print level for mulliken and xray

2005-12-23 00:05  marcella

	* [r4435] linres: a small step forward

2005-12-22 22:52  fawzi

	* [r4434] print_keys
	  smaller bugfixes

2005-12-21 10:06  marcella

	* [r4433] in the process of getting linres working

2005-12-21 08:38  tlaino

	* [r4432] Oooppssss.. bug fix for the latest introduced bug ;-)
	  Still memory leak free ;-)

2005-12-20 23:12  tlaino

	* [r4431] Cleaning memory leaks for KG and MC.. CP2K again memory
	  leak free ;-)

2005-12-20 22:01  mcgrath

	* [r4430] Still more tweaking with the .dat file...can't wait until
	  I can remove it

2005-12-20 21:36  tlaino

	* [r4429] Many things: (1) globenv is now only a POINTER (2) bug
	  fix for GEEP_LIB and added
	  the expansion up to 18 gaussian (16, 17 and 18 with the
	  corresponding regtests)
	  (3) First major clean to remove all TARGET declaration in CP2K..
	  (4) removed two
	  files: coefficient_lists.F and density_types.F (not used in
	  cp2k). Still more work to
	  clean completely the TARGET declaration

2005-12-20 16:53  gtb

	* [r4428] looking for a possible leak

2005-12-20 08:19  vondele

	* [r4427] switch to walltime for mpi performance

2005-12-19 19:34  mcgrath

	* [r4426] Got MC working with KG

2005-12-19 15:45  vondele

	* [r4424] use input constants for the integration grids

2005-12-19 15:22  tlaino

	* [r4423] bug fix

2005-12-19 11:10  tlaino

	* [r4422] Recovering Metadynamics in CP2K.. bug fix for waring in
	  set_atm_mass and set_bond also for CP2K and
	  not only for QM/MM. Some cosmetics around

2005-12-18 07:15  cjmundy

	* [r4418] *More Joost trickery.
	  *Regtests should work again

2005-12-18 06:43  cjmundy

	* [r4417] Added PV_AVA and PV_NUM

2005-12-18 05:08  cjmundy

	* [r4416] *Reinstated PV_AVA and PV_NUM for QS virial (almost
	  there)
	  *Fixed bug in qs_environment_methods
	  *Used an idea of Joost to fix timings problem when grids rebuilt

2005-12-17 18:53  tlaino

	* [r4415] bug fix for the warning in set_atm_mass and create_bond..

2005-12-17 12:52  tlaino

	* [r4414] Bug fix for argon_nph_plastic.. since PV_AVA e PV_NUM
	  were not used I cleaned them
	  simpar struct.. Resetting other two missed bloechl regtest and
	  also KG new regtest due to
	  a wrong choice of test type

2005-12-17 10:10  tlaino

	* [r4413] Bug fix for CMPLX.. resetting one failing regtest. bug
	  fix for some printing information..
	  Give a default name for PROJECT.

2005-12-17 09:18  krack

	* [r4412] switching back to old output file handling
	  printing now only on request

2005-12-17 08:02  krack

	* [r4411] fix memory leak

2005-12-17 01:57  gtb

	* [r4409] NEW functionality:
	  Standard KG calculation of only energy without forces
	  is now impmemented

2005-12-16 18:31  mcgrath

	* [r4408] Fixed a parallel bug

2005-12-16 17:13  tlaino

	* [r4407] Major rewriting of the routines for multipole decoupling
	  (speed-up), bug fix for bounds_local in qmmm
	  printing information for the periodic qmmm run. Resetting the
	  corresponding tests giving wrong results
	  for numerics

2005-12-16 16:05  vondele

	* [r4406] back to more reasonable formats

2005-12-16 16:02  vondele

	* [r4405] do not use random additional precision (at least on
	  location less)

2005-12-16 15:48  krack

	* [r4404] new property: calculation of coherent X-ray diffraction
	  spectrum
	  first stage: slow (not optimized)

2005-12-16 15:46  vondele

	* [r4403] new input bug fix, do not use distributed grids randomly

2005-12-16 15:44  krack

	* [r4402] (re)allocate function for complex data type

2005-12-16 15:38  krack

	* [r4401] typo

2005-12-16 06:38  cjmundy

	* [r4400] *Towards viral for QS. Still need term for XC.
	  Jochen Schmidt wrote these routines
	  *Fixed bug in input_cp2k_subsys. Now
	  reads in when units are in Bohr
	  *Will committ test for pressure soon.
	  *Added Virial to KG_TYPES

2005-12-16 06:35  cjmundy

	* [r4399] Numerical pressure calculation for QS

2005-12-15 20:08  mcgrath

	* [r4397] Cleaned up some stuff, got MC working with Fist again.

2005-12-14 22:03  vondele

	* [r4396] fix last newinput runtime failure

2005-12-14 13:53  tchassai

	* [r4394] Replace temp with query.

2005-12-14 13:51  tchassai

	* [r4393] Fix memory leaks.

2005-12-13 21:36  marcella

	* [r4392] memory leaks?

2005-12-13 21:27  marcella

	* [r4391] some memory leaks

2005-12-13 18:58  marcella

	* [r4388] some fixing for KG

2005-12-13 17:28  mcgrath

	* [r4386] This should take care of the memory leak

2005-12-13 16:29  mcgrath

	* [r4380] Volume stuff working again.

2005-12-13 15:58  tchassai

	* [r4378] Refactoring: replace temp with query (See Fowler p. 120)

2005-12-13 15:46  tchassai

	* [r4377] Fix tddfpt triplets.

2005-12-13 14:14  marcella

	* [r4376] get rid of globenv from qs_scf_post

2005-12-13 12:54  tlaino

	* [r4375] bug fix

2005-12-13 12:14  tlaino

	* [r4374] timings in routines

2005-12-13 11:38  marcella

	* [r4373] restored printing of MOS on output unit

2005-12-13 09:14  tchassai

	* [r4372] - Replaced globenv%ionode with ionode got from logger
	  - Replaced globenv%scr with output_unit got from logger
	  - Try to fix memory leak

2005-12-13 09:03  tchassai

	* [r4371] DFT%METHOD --> DFT%QS%METHOD

2005-12-13 09:01  tchassai

	* [r4370] Make module more consistent.

2005-12-13 08:18  tchassai

	* [r4369] Reactivated Lumo printing in scf_post_calculation for
	  tddfpt fix.

2005-12-13 08:11  vondele

	* [r4368] test

2005-12-12 15:11  krack

	* [r4367] Set an upper limit for the subcell grid size

2005-12-12 12:39  tlaino

	* [r4364] bug fix for charge. Faster implementation

2005-12-12 11:18  marcella

	* [r4363] bug fixed to restart with different number of MOs

2005-12-09 19:46  mcgrath

	* [r4362] MC now completely uses new input (sorry for the
	  delay...flu :-( )

2005-12-09 14:13  mcgrath

	* [r4360] fixed one regtest

2005-12-08 15:56  marcella

	* [r4356] fixed bug in the calculation of the max radius for the 3
	  center lists

2005-12-08 13:17  tlaino

	* [r4355] bug fix for multipole decouple (now uses the real
	  numerical charge). resetting the related regtest

2005-12-07 21:12  marcella

	* [r4354] xes emission

2005-12-07 20:38  marcella

	* [r4353] x-ray emission

2005-12-07 19:33  ikuo

	* [r4352] Cleanup for DEC and fix all warnings on DEC compiler

2005-12-07 18:20  tlaino

	* [r4351] bug fix in ewald decoupling part

2005-12-07 18:19  tlaino

	* [r4350] removing a write statement

2005-12-07 18:17  tlaino

	* [r4349] making more rational the settings of eps_default and so
	  on.. cleaning eps_pgf_aux not used anymore..

2005-12-07 15:58  gtb

	* [r4347] fixing for new input

2005-12-07 13:50  tchassai

	* [r4346] Remove duplicated code.

2005-12-07 12:54  tchassai

	* [r4345] Deleted two now superfluous comments.

2005-12-07 12:51  tchassai

	* [r4344] Refactoring of the xc_rho_set_and_dset_create routine
	  with small
	  changes in xc_rho_set_types.F and kg_kxc.F to adapt.

2005-12-07 12:47  tchassai

	* [r4343] Added accessor functions that can be used instead of
	  passing a variable
	  in the parameter list of a subroutine call. This should make the
	  code
	  a lot simpler. See xc_rho_set_and_dset_create in xc.F as an
	  example.

2005-12-07 11:18  vondele

	* [r4342] make name shorter than 32 characters

2005-12-07 10:05  tlaino

	* [r4341] bug fix for PGI

2005-12-06 14:00  vondele

	* [r4340] quick fix for breakage

2005-12-06 12:17  vondele

	* [r4339] first steps towards fixing metadynamics

2005-12-05 18:55  tkuehne

	* [r4338] harris_functional.F & harris_force.F: Preparation for
	  LSDA

2005-12-05 16:46  tlaino

	* [r4337] Semi-empirical ALMOST working..

2005-12-05 13:47  tlaino

	* [r4336] ops.. now FIST is really fully recovered..

2005-12-04 22:21  fawzi

	* [r4334] leaks --

2005-12-04 18:06  fawzi

	* [r4333] fixing hcth 120, checking for extra param in input files

2005-12-04 09:56  tlaino

	* [r4330] bug fix new input..FIST is fully recovered.. YAHOOOOO ;-)
	  !

2005-12-03 21:45  fawzi

	* [r4329] fixing default values of nose,...

2005-12-03 19:18  vondele

	* [r4327] just spelling

2005-12-03 14:45  vondele

	* [r4324] test regtest

2005-12-03 13:34  vondele

	* [r4323] test regtest

2005-12-02 18:52  tlaino

	* [r4322] bug fix for qm/mm .. regtest fully recovered.. YAHOOOOOO!
	  ;-)

2005-12-02 15:34  tkuehne

	* [r4319] harris_force.F: Memory leak fix

2005-12-02 14:38  fawzi

	* [r4317] fist mostly fixed

2005-12-02 14:10  vondele

	* [r4312] fixing regtests

2005-12-02 13:57  tkuehne

	* [r4311] harris_force.F: Emergency hack to employ Fawzi's changes
	  into my newest version,
	  which I unfortunately check-in before I recognized it. Sorry!
	  The failing harris regtests are going to be resetted.

2005-12-02 12:41  tkuehne

	* [r4310] Formerly harris_force_Final.F

2005-12-02 12:41  tkuehne

	* [r4309] Formerly harris_force_clean.F

2005-12-02 12:40  tkuehne

	* [r4308] Formerly harris_force_FK.F

2005-12-02 12:40  tkuehne

	* [r4307] Step5, formerly harris_force_save.F

2005-12-02 12:39  tkuehne

	* [r4306] Step4, formerly harris_force_CVS.F

2005-12-02 12:38  tkuehne

	* [r4305] harris_force.F: Step2 of my major cleanup

2005-12-02 12:35  tkuehne

	* [r4304] harris_force.F: First cleanup

2005-12-02 12:32  tkuehne

	* [r4303] harris_functional.F: Minor changing concerning the right
	  denisties
	  qs_force.F: removal of fast flag
	  qs_ks_methods.F: Making calc_rho_tot_gspace public

2005-12-02 12:20  marcella

	* [r4302] options for GAPW

2005-12-02 11:07  tkuehne

	* [r4301] qs_scf.F: Changes due to the FD Debug mode for the Harris
	  forces
	  qs_rho_methods.F: Additional multiplier for the function
	  diff_rho_type
	  harris_functional.F: Minor cleanup
	  harris_env_type: Removal of Hamiltonian & KS matrices
	  harris_force.F: Some refactoring

2005-12-02 10:11  tchassai

	* [r4299] Workaround for PGI 6.0.5

2005-12-02 09:39  vondele

	* [r4298] regtest test

2005-12-02 00:24  fawzi

	* [r4296] * new input
	  * poisson_env
	  * pw_grid_create
	  * many bugfixes & cleanup (so many that I forgot which)
	  * not all regtest pass !
	  * xc: use section instead of xc_control

2005-11-29 17:46  tlaino

	* [r4291] removing write statement

2005-11-29 12:53  vondele

	* [r4290] added error on inverting matrices of 0 size, could maybe
	  kindly return as well.

2005-11-29 12:40  tchassai

	* [r4289] Towards a self-consistent application of accurate xc
	  potentials:
	  - expanded xc_control_type
	  - defined new input_constants
	  - added an read_xc_potential_section routine

2005-11-29 12:37  tchassai

	* [r4288] Refactoring: extraced calc_v_sic_rspace routine

2005-11-28 20:35  ikuo

	* [r4287] fix for dsyevd call on dec for cxml lib

2005-11-28 16:41  tlaino

	* [r4286] just cosmetics

2005-11-28 16:40  tlaino

	* [r4285] Bug fix for input coordinate as XYZ and no PSF provided

2005-11-28 13:58  vondele

	* [r4284] rename another eps

2005-11-28 13:48  vondele

	* [r4283] renamed eps to something more clear

2005-11-28 11:29  vondele

	* [r4282] fix format

2005-11-28 11:26  vondele

	* [r4281] fix format

2005-11-25 20:23  cjmundy

	* [r4280] New formate for nph calculations. CMASS is given
	  for the cell mass instead of TIMECON. This is better
	  for two reasons:
	  1) Now can directly compare against CPMD
	  2) TIMCON => npt % mass required a target
	  temperature which is not known in nph.
	  REGTEST will not change for the example.

2005-11-25 14:35  vondele

	* [r4278] jgh: add general formula for the whittaker functions

2005-11-24 22:23  cjmundy

	* [r4277] Replaced with initialize_extended_system_types,
	  init_extended_variables, nhc_mapping

2005-11-23 21:34  marcella

	* [r4275] check for linear depencies in the basis set before
	  constructing the gapw projectors

2005-11-23 21:15  marcella

	* [r4274] changing LMAXN1 from input it is possible to truncate
	  the local density expansions to a lmax quantum number lower than
	  2*maxlgto

2005-11-23 17:19  tkuehne

	* [r4273] harris_force.F: New function harris_calc_nsc_force
	  harris_env_types.F: New structure-fields rho_diff, matrix_h &
	  matrix_ks
	  harris_force_types.F: f_nsc added
	  harris_functional.F: matrix_h & matrix_ks replicated
	  qs_ks_methods.F: NSC_flag added
	  qs_scf.F: cleanup

2005-11-23 10:04  gtb

	* [r4271] modified for compilation with xlf 8.1.1.

2005-11-22 18:18  vondele

	* [r4269] some more digits needed for GAPW

2005-11-21 13:27  vondele

	* [r4265] easier temperature format

2005-11-21 13:01  vondele

	* [r4262] full_all preconditioner with LSD

2005-11-18 13:51  vondele

	* [r4261] clean mult_prec away (use full_all instead).

2005-11-18 12:33  vondele

	* [r4260] revives FULL_ALL preconditioner, useful with all-electron
	  calculations

2005-11-18 11:12  marcella

	* [r4259] bug fix

2005-11-17 13:16  marcella

	* [r4258] fixed bug in read_constraint_section_new

2005-11-17 08:52  vondele

	* [r4256] fix KG regression

2005-11-16 22:07  cjmundy

	* [r4254] Cleaining and fixing a bug in the shock code (nph).
	  Regtest
	  for this example will change.

2005-11-16 22:05  cjmundy

	* [r4253] Change the name of files to better reflect the modules
	  purpose.
	  Should make navigating the thermostat initialization, etc much
	  more transparent. We'll see.

2005-11-16 21:49  marcella

	* [r4252] smearing in xas

2005-11-15 16:06  marcella

	* [r4249] preconditioner for lumos

2005-11-14 23:04  mcgrath

	* [r4248] MC should be working with all one component systems now

2005-11-11 19:50  marcella

	* [r4247] better not to divide by zero

2005-11-11 19:34  marcella

	* [r4246] hcth name

2005-11-11 15:23  marcella

	* [r4245] new input hcth

2005-11-11 12:47  tchassai

	* [r4243] Added repeat=.TRUE. in COORDS section to allow for the
	  specification of
	  two or more atoms of the same species in the new input

2005-11-10 18:30  marcella

	* [r4242] work around for local density lmax too small,
	  should be better fixed

2005-11-10 00:49  marcella

	* [r4241] 2 problems with gapw maybe solved
	  xas calculation updated

2005-11-09 23:08  mcgrath

	* [r4240] A new move that makes a general angle change on a
	  molecule, and some clean-up

2005-11-08 16:24  fawzi

	* [r4237] search_object -> parser_search_string and made it public.

2005-11-08 13:19  tlaino

	* [r4236] syntax bug fixed

2005-11-08 09:52  vondele

	* [r4235] further testing

2005-11-08 09:46  vondele

	* [r4234] further testing

2005-11-08 08:57  vondele

	* [r4233] further testing

2005-11-07 23:53  marcella

	* [r4232] xas: list of atoms

2005-11-07 23:48  marcella

	* [r4231] xas: list of atoms

2005-11-07 23:40  marcella

	* [r4230] xas: list of atoms

2005-11-07 22:27  marcella

	* [r4229] list cubes fixed

2005-11-07 20:32  tlaino

	* [r4228] All the memory management of the pair_potential_type as
	  been completely rewritten.
	  Now only the specific potential required in FIST calculation is
	  allocated.
	  Less memory usage (particularly for jobs with many kinds). Some
	  cosmetics around.. WOOF ;-)

2005-11-07 18:57  tlaino

	* [r4227] Bug fix: Commenting the STOP statement otherwise CP2K
	  doesn't exit nicely !!

2005-11-07 18:25  marcella

	* [r4226] cp_assert included tocompile on IBM

2005-11-07 16:33  mcgrath

	* [r4225] Bug fix.

2005-11-07 14:41  vondele

	* [r4224] further testing

2005-11-07 14:00  vondele

	* [r4223] further testing, restore build

2005-11-07 13:59  vondele

	* [r4222] break

2005-11-07 13:27  vondele

	* [r4221] further testing

2005-11-07 13:05  krack

	* [r4219] Allow to specify the number of computed eigenvectors for
	  the expert driver

2005-11-07 12:42  vondele

	* [r4218] further testing

2005-11-07 12:38  vondele

	* [r4217] further testing

2005-11-07 12:24  vondele

	* [r4216] further testing

2005-11-07 11:36  vondele

	* [r4215] further testing

2005-11-07 11:09  vondele

	* [r4214] further testing

2005-11-07 10:24  vondele

	* [r4213] remove leak for testing

2005-11-07 10:23  vondele

	* [r4212] introduce leak for testing

2005-11-07 09:56  vondele

	* [r4211] idem

2005-11-07 09:21  vondele

	* [r4210] Hello, world!
	  Some testing

2005-11-04 22:39  tlaino

	* [r4209] New Input TOPOLOGY section.. Moved all checkings with
	  strings to checkings with INTEGER..

2005-11-04 17:33  mcgrath

	* [r4208] Added a few references for people interested in MC.

2005-11-04 16:57  mcgrath

	* [r4207] A new move (AVBMC swap move), some clean up and
	  documentation, and using the RNG implimented by Matthias

2005-11-04 15:49  tlaino

	* [r4206] Bug fix for CONNECTIVITY OFF..
	  cleaning the GROMOS stuff (not used)
	  Bug fix for parallel execution in fist_neighbour_lists

2005-11-04 11:11  fawzi

	* [r4205] more bugfix for big procs numbers.

2005-11-03 19:08  fawzi

	* [r4204] fixed bug with many many processors

2005-11-02 23:15  tkuehne

	* [r4203] force_env_methods.F, harris_metropolis_mc.F, qs_scf.F &
	  qs_energy.F: Changes to
	  make the over-pedantic NAG compiler content.
	  qs_core_energies.F: cosmetic

2005-11-02 20:36  marcella

	* [r4202] sd for linres

2005-11-02 16:16  tkuehne

	* [r4201] md.F & input_constants.F: Prettify HMC output
	  harris_metropolis_mc.F: Subtle changes & cleanup

2005-11-02 00:15  tkuehne

	* [r4198] harris_metropolis_mc.F & md.F: Dump HMC informations into
	  a .hmc-file
	  md_energies.F: cosmetic

2005-11-01 11:15  tchassai

	* [r4197] Removed comments

2005-11-01 11:14  tchassai

	* [r4196] Fix for broken TD-DFT triplets

2005-10-31 18:41  tkuehne

	* [r4195] harris_metropolis_mc.F: Some subtle changes & cleanup
	  input_cp2k_motion.f & md.F: New keywords
	  qs_wf_history_methods.F: Bugfix

2005-10-26 10:54  tlaino

	* [r4194] bug fix in printing the potential in parallel execution

2005-10-26 08:10  tchassai

	* [r4193] - new input cleanup in tddfpt
	  - preparing to apply xc potentials self-consistently

2005-10-25 14:11  tkuehne

	* [r4192] Revised new and old input for HMC.

2005-10-25 13:35  tchassai

	* [r4191] -Cosmetic changes
	  -Bugfix for bug that appears with NAG sdbg

2005-10-24 14:53  tlaino

	* [r4190] bug fix for parallel execution

2005-10-21 12:31  fawzi

	* [r4189] checking for too long timer names
	  removed superfluos routine_name from timer_env

2005-10-21 12:30  fawzi

	* [r4188] removing 95 spaces in the middle of the timer name.

2005-10-20 13:17  tlaino

	* [r4187] Just cosmetics

2005-10-19 17:51  fawzi

	* [r4183] split input_cp2k.F in smaller files (input_cp2k_dft.F
	  input_cp2k_mm.F input_cp2k_motion.F input_cp2k_qmmm.F
	  input_cp2k_subsys.F input_cp2k.F)

2005-10-19 17:46  fawzi

	* [r4182] binsearch for faster lookup in timeset.

2005-10-19 12:32  tchassai

	* [r4177] Bugfix for the DESCRIBE 4 level.

2005-10-15 12:55  tkuehne

	* [r4175] some minor changes

2005-10-14 10:56  tkuehne

	* [r4173] qs_wf_history_methods.F: Deallocation is removed from the
	  copy routine
	  harris_metropolis_mc.F: Some preliminary changes, since I'm going
	  into holidays!
	  md.F, input_constants.F, cp_control_utils.F: Resurgence of LDMC

2005-10-14 09:16  krack

	* [r4172] Redundant copy of the new density matrix removed

2005-10-12 23:58  tkuehne

	* [r4171] harris_metropolis_mc.F: Possibility to properly restart &
	  to restore the state of the wavefunction history,
	  which is necessary in combination with certain extrapolations if
	  a MD approach is rejected.
	  qs_wf_history_methods.F: New functions - wfi_duplicate_history,
	  wfi_copy_history, wfs_duplicate_snapshot &
	  wfs_copy_snapshot, BUT not yet fully tested!
	  kg_gpw_fm_mol_types.F: Additional function -
	  duplicate_kg_fm_p_type
	  qs_wf_history_types.F: making the function wfi_clear_history
	  public.
	  scf_control_types.F: Frozen density extrapolation in the old
	  input
	  qs_rho_methods.F: Code perfection
	  pw_pool_types.F: syntax

2005-10-12 15:20  tlaino

	* [r4170] bug fix

2005-10-12 15:03  tchassai

	* [r4168] Workaround for bug in PGI 6.0.5

2005-10-12 14:31  tlaino

	* [r4167] now we read also XYZ from external file as we do for pdb

2005-10-12 09:46  tkuehne

	* [r4166] qs_wf_history_methods.F: NO reorthogonalization in the
	  frozen extrapolation if OT is used, and some minor cleanup.
	  input_cp2k.F & input_constants.F: ASPC in new input style
	  harris_force.F & qs_scf.F: Minor cleanup

2005-10-11 14:12  krack

	* [r4165] MODULE name corrected; dclock is used instead of
	  mpi_wtime

2005-10-10 21:22  mcgrath

	* [r4164] Some cleaning...one step closer to general MC routines

2005-10-10 20:21  ikuo

	* [r4163] bug fix for PARA_RES, and 1-4 kind list

2005-10-10 16:28  vondele

	* [r4162] switch to cp_common_uses to make sure things work
	  independent of the setting of the long lines expansion

2005-10-09 09:37  tlaino

	* [r4158] using the xt3 machine file.. making everything consistent

2005-10-08 05:55  tlaino

	* [r4154] New optimization method: a Mixed steeepest descent
	  conjugate gradients method.
	  Bug fix for periodic qmmm potential
	  Bug fix for keyword silent
	  GEO_OPT section parsed wth the new input

2005-10-07 13:03  fawzi

	* [r4153] removing an ABS too much

2005-10-07 10:03  krack

	* [r4152] Always stable predictor-corrector (ASPC);
	  new SCF section keyword MAX_SCF_ASPC addded

2005-10-07 09:19  krack

	* [r4151] function name binomial_coefficient changed to simply
	  binomial

2005-10-05 23:36  marcella

	* [r4148] sx arch

2005-10-05 23:35  marcella

	* [r4147] SX arch

2005-10-05 16:38  ikuo

	* [r4146] Add wildcard matching for improper torsion potential

2005-10-04 22:19  tkuehne

	* [r4145] Redesign & elimination of a few subtle bugs, which
	  formerly results in low acceptance ratios

2005-10-04 20:24  mcgrath

	* [r4144] updating documentation

2005-10-04 19:58  mcgrath

	* [r4143] guess someone already made external_control general

2005-10-04 19:29  mcgrath

	* [r4142] Changed to mc_run.F so we no longer need this

2005-10-04 19:28  mcgrath

	* [r4141] Fixed a bug in the volume move (recently introduced) and
	  cleaned up some stuff

2005-10-02 17:06  tkuehne

	* [r4140] Doing the velocity initialization right, with kindly
	  support of Fawzi

2005-09-30 17:29  fawzi

	* [r4139] mainly robustness and being nice toward the enduser.

2005-09-30 15:56  ikuo

	* [r4138] change KIND=dp to KIND=sp

2005-09-30 13:40  krack

	* [r4137] fix for optional argument binary

2005-09-30 07:17  krack

	* [r4136] fix: KIND=4 changed to KIND=dp (even if 4 might refer to
	  sp)

2005-09-29 21:20  ikuo

	* [r4135] DCD binary trajectory format

2005-09-29 21:19  ikuo

	* [r4134] Remove duplicate USE for DEC

2005-09-27 18:06  ikuo

	* [r4133] Fix restart all for QMMM in regards to no cell_ref yet

2005-09-27 18:04  ikuo

	* [r4132] Temporary workaround for Intel and array dimension 0
	  problem

2005-09-27 18:03  ikuo

	* [r4131] de-tab

2005-09-27 13:24  marcella

	* [r4129] problems compiling on IBM

2005-09-27 07:45  tkuehne

	* [r4128] Changing the denotation from LDMC to HMC.

2005-09-26 19:02  tkuehne

	* [r4127] harris_metropolis_mc.F: Hybrid MC (HMC) implementation
	  qs_environment_types.F, qs_environment.F, qs_scf.F, qs_force.F:
	  Harris flag
	  util.F: cosmetics

2005-09-26 15:48  fawzi

	* [r4126] energy drift now in kelvin per atom

2005-09-26 15:46  fawzi

	* [r4125] .crd -> .xyz

2005-09-26 15:19  tkuehne

	* [r4124] force_env_methods.F, input_keyword_types.F: Substitution
	  of a single CPPrecondition through an equivalent
	  cp_assert, to quiten g95 without using Joost's quick-fix.

2005-09-26 13:46  tkuehne

	* [r4123] cp_subsystem_methods.F, input_keyword_types.F:
	  Substitution of CPAssert through cp_assert
	  input_keyword_types.F: Shorten a too long line

2005-09-22 16:12  vondele

	* [r4119] shorten long line

2005-09-21 17:06  fawzi

	* [r4116] new Splines (also pbc)
	  minor fixes

2005-09-21 05:18  marcella

	* [r4115] missing default in qs_control

2005-09-21 00:25  ikuo

	* [r4114] Remove leftover debuging output

2005-09-20 20:29  ikuo

	* [r4113] Make new atomic kind is qeff is different for FIST

2005-09-20 20:28  ikuo

	* [r4112] Add FIST debug

2005-09-20 17:30  tkuehne

	* [r4111] harris_metropolis_mc.F, OBJECTDEFS, md.F,
	  input_constants.F, input_cp2k.F,
	  cp_control_utils.F, cp2k_runs.F: Langevin Dynamics Monte Carlo
	  md_conserved_quantities.F: Change of the conserved quantity of
	  the Langevin dynamics to
	  total energy.
	  md_run.F: Making the functions temperature_control &
	  external_control public
	  wiener_process.F: Option to suppress the storage of the random
	  number genarator state
	  force_env_types.F: Syntax cosmetics

2005-09-20 09:45  tlaino

	* [r4110] bug fix for variable error.

2005-09-19 10:55  tlaino

	* [r4108] problem with cpp macro fixed!

2005-09-16 17:16  mcgrath

	* [r4107] Make sure we're not writing to screen for every processor

2005-09-16 17:00  mcgrath

	* [r4105] The start of massive MC work...restructured and cleaned
	  up some things, added a discrete volume move

2005-09-15 12:23  tkuehne

	* [r4104] Minor change

2005-09-15 11:40  tkuehne

	* [r4102] cp_control_utils.F, input_constants.F, input_cp2k.F,
	  qs_rho_methods.F, qs_scf.F,
	  qs_wf_history_methods.F, qs_wf_history_types.F: Frozen density
	  approximation.
	  Activated through EXTRAPOLATION FROZEN
	  qs_collocate_density.F: Minor changes to make Fawzi happy

2005-09-13 10:28  tlaino

	* [r4099] Introduced the Fumi-Tosi potential for NaCl.. Clean all
	  the amber stuff (not used!)

2005-09-13 09:46  tkuehne

	* [r4098] New function diff_qs_rho, which allocates a new
	  qs_rho_type instance that contains the difference of two
	  qs_rho_type variabes

2005-09-09 11:19  tlaino

	* [r4097] Bug fix capital/lower letters in atomic types
	  (PSF+CHARMM)

2005-09-02 20:04  tlaino

	* [r4095] One step further towards the periodic QM/MM calculation.
	  Introduced the possibility
	  to recouple the periodic images. Introduced the super_cell in
	  qs_env.

2005-09-02 14:19  tkuehne

	* [r4094] qs_scf.F: Making the Harris functional working together
	  with OT.
	  harris_functional.F, harris_force.F: Some cleanup

2005-09-02 12:21  tlaino

	* [r4092] Bloechl decoupling scheme finished and working (energy
	  and derivatives). added regtests.

2005-09-02 12:21  tlaino

	* [r4091] allow dynamics for one atom only..

2005-09-01 10:23  krack

	* [r4090] default print levels for new DFT section adjusted

2005-08-31 07:28  marcella

	* [r4086] fixing bug

2005-08-30 21:51  krack

	* [r4085] forgotten NULLIFY added

2005-08-30 21:23  krack

	* [r4084] Eigensolver test added

2005-08-30 14:43  marcella

	* [r4081] displaced mpi sum

2005-08-28 13:09  marcella

	* [r4080] bug fixing

2005-08-27 18:22  marcella

	* [r4079] lists for gapw_xc

2005-08-27 11:54  marcella

	* [r4077] sorry, this should fix the regtests

2005-08-26 19:57  marcella

	* [r4076] GAPW_XC finished but not tested yet

2005-08-25 22:16  marcella

	* [r4072] GAPW only for the XC terms

2005-08-25 19:30  marcella

	* [r4071] displaced nullify

2005-08-25 18:08  marcella

	* [r4070] skip 2 and 3 center term if equal exponents of
	  compensation charges

2005-08-25 17:56  marcella

	* [r4069] reduced projector basis for diffuse basis sets

2005-08-25 17:31  tkuehne

	* [r4068] qs_debug.F, force_env_methods.F, qs_energy.F: Making
	  Theo's FD routine ready for the non-selfconsistent
	  case. Btw. the QS forces aren't consistent with the
	  according energies, if not iterated to self-consistency!
	  harris_force.F: MPI parallelized
	  Fast flag added to suppress double calculations among the Harris
	  force terms.
	  harris_functional.F: Fast flag added to suppress double
	  calculations among the Harris energy terms.
	  qs_force.F: Substitution of the QS force calculation through the
	  Harris forces, resulting in a speed-up
	  of about 2.5 against QS in a MD. The forces are still not good
	  enough and the energy isn't
	  conserved without any thermostat, so don't use it in a MD!
	  Major Bugfix: What is stored in the force structure and also what
	  is printed out on the
	  screen are all gradients not forces!
	  The harris regtests are for this reason once again reseted:
	  Sorry!
	  qs_energy_types.F: Additional functions to duplicate and copy
	  qs_energy_type variables.
	  qs_scf.F: The Harris energy is written back into the energy
	  environment.
	  qs_integrate_potential.F: syntax

2005-08-24 09:53  vondele

	* [r4065] skip optical conductivity if homo.ne.nmo

2005-08-23 15:51  fawzi

	* [r4063] pw_spline, fine2coarse, coarse2fine,...:
	  * correct loop unrolling
	  * better preconditioners
	  * parallel bugfixes
	  * defaults suboptimal, should change them.

2005-08-23 15:47  fawzi

	* [r4062] using new cp_common_uses.h to avoid forgetting some
	  includes

2005-08-23 15:45  fawzi

	* [r4061] * bugfixes for a big number of parallel processes
	  * zeroout in cp_cfm_diag to let the parallel debug version with
	  the nag compiler pass trough

2005-08-23 09:30  marcella

	* [r4060] back to former factor

2005-08-23 07:36  marcella

	* [r4059] too long line

2005-08-23 07:36  marcella

	* [r4058] cube files for xas

2005-08-23 05:12  ikuo

	* [r4056] Break the long line into two lines.

2005-08-22 22:19  ikuo

	* [r4055] Added VDW_SCALE14

2005-08-22 19:43  tkuehne

	* [r4053] qs_collocate_density.F: Function which collocates atomic
	  charge densities from the unused
	  aux_basis_set directly onto a grid/multigrid.
	  qs_scf.F: New initial guess (activated through the keyword
	  DENSITIES), which collocates fitted
	  charge densities.
	  basis_set_types.F: Prints out which norm is used.
	  atomic_kind_type.F: In case the DENSITIES initial guess is used
	  and already fitted and
	  normalized charge densities are used, the QS normaliztion is
	  skipped.
	  qs_force.F: The Harris forces are written into the force
	  environment in case the harris
	  functional is used.
	  harris_force.F: Even better harris forces.
	  harris_functional.F: Removal of commented code.
	  Some very minor syntax changes in the following files:
	  cube_utils.F
	  kg_densities.F
	  molecular_states.F
	  pws.F
	  qs_core_hamiltonian.F
	  qs_energy.F

2005-08-22 18:14  ikuo

	* [r4052] BUGFIX add missing 1-4 electrostatic interaction

2005-08-18 14:22  marcella

	* [r4050] minor changes

2005-08-16 22:14  marcella

	* [r4049] ot with ae: only slight improvement in convergence by
	  usina multiple preconditioner

2005-08-15 08:14  marcella

	* [r4048] fix TARGET for new version of nag compiler (Fawzi)

2005-08-11 19:35  tlaino

	* [r4047] Introducing a new decoupling methods for periodic images
	  (still not working).. and going towards a fully
	  functional qmmm periodic electrostatic (still more work!)

2005-08-11 19:02  marcella

	* [r4046] prepare minimizer for linres

2005-08-11 12:19  fawzi

	* [r4045] fixing justenergy with qmmm

2005-08-11 12:08  tlaino

	* [r4044] fit_charge working in parallel.

2005-08-11 09:26  tlaino

	* [r4043] clenaing memory leaks for fit_charge implementation..
	  towards parallelization of the fitting charge procedure

2005-08-09 11:47  tlaino

	* [r4042] Possibility to ge the charges using the density derived
	  scheme even just for a QS calculation

2005-08-08 22:51  tlaino

	* [r4041] Reorganization of the routine to evaluate qm/mm energy
	  (clean few dependencies)..
	  Newroutine to evaluate charges according the Bloechl scheme.
	  still some work to do.

2005-08-03 19:51  marcella

	* [r4040] DeltaSCF : homo = homo+1

2005-08-03 12:28  marcella

	* [r4039] change occupation for dscf

2005-08-03 11:23  tkuehne

	* [r4038] harris_force.F: Memory leak fix
	  Elimination of the double calculation of f_V

2005-08-03 10:24  marcella

	* [r4037] DeltaSCF for the first excited state

2005-08-03 09:02  tlaino

	* [r4036] Introducing a new section PROPERTIES. moving LINRES there
	  and adding a FIT_CHARGE section

2005-08-03 07:55  marcella

	* [r4035] fix memory release

2005-08-02 15:42  tlaino

	* [r4034] Bug fix in derivatives for periodic qm/mm

2005-08-02 10:31  marcella

	* [r4033] added USE cp_to_string, to be able to compile on the sp4

2005-08-02 10:11  tkuehne

	* [r4032] harris_force.F: Correction of the harris force
	  calculation. The forces due to the sum of
	  eigenvalues are calculated with rho_out since the sum of
	  eigenvalues is
	  equal to Tr(H_in*rho_out) in the harris energy functional.
	  Probably the Pulay forces are improper for the time being, since
	  the forces
	  are now calculated in a mixed force expression with rho_in &
	  rho_out.
	  qs_scf.F: removal of commented code
	  input_cp2k.F: Keyword CORE_CORRECTION changed into the new input
	  style
	  hartree.F & harris_force_types.F: Syntax cosmetics

2005-07-28 15:02  vondele

	* [r4031] more consistent format

2005-07-27 16:58  ikuo

	* [r4027] remove extra USE for DEC compiler

2005-07-27 16:57  ikuo

	* [r4026] add missing openmp declaration

2005-07-27 12:01  fawzi

	* [r4025] - fine2coarse loop unrolling
	  - removed commented code
	  - fixed leak in ewald

2005-07-26 15:51  tkuehne

	* [r4024] Modification for the use in combination with the Harris
	  energy functional.

2005-07-26 09:20  krack

	* [r4022] Langevin dynamics using a new MD restart file layout.
	  A MD run is uniquely restartable even with a different number
	  of processes, since a Wiener process is setup for each particle.
	  The status of the Wiener process is written to the MD restart
	  file
	  for each particle.

2005-07-26 08:17  vondele

	* [r4021] further fixing for writing to unconnected units

2005-07-25 07:10  fawzi

	* [r4020] fixing pme in parallel.

2005-07-25 06:11  vondele

	* [r4019] write 'element not found' warning only if iw is a
	  connected IO unit.
	  This warning still needs a proper fix.

2005-07-23 08:38  krack

	* [r4015] logical section elements changed to attributes
	  root element name changed to CP2K_INPUT

2005-07-22 10:59  tlaino

	* [r4014] Preliminary stuff for periodic QM/MM calculations. Not
	  working yet.

2005-07-22 10:05  tlaino

	* [r4013] Bug fix for read_qs_section: problems with default! We
	  need care with the conversion to the new input!
	  Regtest is OK!

2005-07-22 00:24  tlaino

	* [r4010] BUG fix introduced with the read_qs_section:
	  READ_QS_SECTION MUST BE CHECKED TO BE
	  LOGICALLY EQUIVALENT TO THE OLD ROUTINE READING QS SECTION. BY
	  NOW IT IS NOT!

2005-07-21 17:00  fawzi

	* [r4008] fixing ewald_none bug

2005-07-21 16:51  marcella

	* [r4007] check of the type of the orbital
	  using the sto basis set

2005-07-21 16:37  marcella

	* [r4006] check of the type of the orbital
	  using the sto basis set

2005-07-21 14:05  fawzi

	* [r4005] adding forgotten mp_sum for forges of added or moved
	  atoms

2005-07-21 12:21  fawzi

	* [r4004] small bugfix: < -> <=, < -> <=

2005-07-21 11:42  fawzi

	* [r4003] * parallel QMMM (slow, but mostly working)
	  * added cp_common_uses.h to remove uses of cp_log and error
	  * fixing small bug wrt. to parallel PME (that still does not work
	  due to a bug introduced by someone else previously)

2005-07-21 10:33  marcella

	* [r4002] default values

2005-07-20 13:56  marcella

	* [r3998] new input QS and MGRID
	  the subsection INTERPOLATOR in MGRID is not done
	  who knows what it is for should fix it

2005-07-19 22:44  marcella

	* [r3997] fixing for IBM-SP4

2005-07-19 22:34  marcella

	* [r3996] set name deleted

2005-07-19 22:20  marcella

	* [r3995] expansion in gto of a sto basis set

2005-07-19 08:31  krack

	* [r3993] --xml flag description added

2005-07-17 20:38  krack

	* [r3988] Logical XML elements changed to attributes

2005-07-17 08:18  tlaino

	* [r3986] 2,3,4 gaussians for GEEP library

2005-07-17 07:09  tlaino

	* [r3984] Ops.. I forgot to submit the new file..

2005-07-16 19:52  tlaino

	* [r3983] GEEP library for qm/mm..

2005-07-15 09:37  marcella

	* [r3979] operators for nmr caculations

2005-07-15 09:31  krack

	* [r3978] special characters in keywords removed

2005-07-15 08:04  krack

	* [r3977] bug fix

2005-07-14 13:06  fawzi

	* [r3975] Preliminary modifications towards the parallel qm/mm..

2005-07-14 12:59  tkuehne

	* [r3974] harris_functional.F: removal of an unnecessary get_qs_env
	  & cleanup
	  qs_core_hamiltonian.F: Modification to feed in the right density
	  for the harris force calculation
	  harris_force.F: Bug Fix - call of build_core_hamiltonian_matrix
	  with density rho_in.
	  Cleanup

2005-07-14 08:52  tlaino

	* [r3973] Bug fix for run_type names and program names.. to be
	  consistent..

2005-07-13 16:17  tlaino

	* [r3971] Preparing for derivatives during BSSE

2005-07-13 14:52  krack

	* [r3969] - bug fix for XML output generated using the --xml flag:
	  allow for lists of default values
	  - some cleaning

2005-07-12 16:25  tkuehne

	* [r3968] qs_force_types.F: cleanup
	  qs_environment_types.F: syntax cosmetic
	  qs_core_hamiltonian.F: major cleanup
	  harris_functional.F: minor perfecting
	  harris_force.F: fix of a major memory leak with kindly support of
	  Fawzi
	  removal of unnecessary grid initializations & returns
	  fix of an other memory leak
	  minor perfecting
	  cleanup
	  harris_env_types.F: cleanup
	  harris_energy_types.F: cleanup

2005-07-12 07:02  tlaino

	* [r3966] New keyword "CENTER0" to avoid the centering of the
	  system even @ the start of the run.

2005-07-12 06:36  tkuehne

	* [r3965] OBJECTDEFS: inclusion of harris_energy_types.F
	  harris_energy_types.F: The harris energy type, what was formerly
	  more or less harris_env_types.F
	  harris_force.F: Changes due to the new harris environment
	  Removal of unnecessary grid initilaizations and returns
	  Removal of unused code
	  harris_force_types.F: Changes due to the new harris environment
	  Introduction of additional fields of type real
	  harris_functional.F: Changes due to the new harris environment
	  Creation of a duplicate of the rho structure for the force
	  calculation
	  qs_core_hamiltonian.F: Additional harris flag
	  qs_environment.F: Changes due to the new harris environment
	  Hook in of harris_energy & harris_force into harris_env
	  qs_environment_types.F: Insertion of harris_env into qs_env
	  qs_force.F: Changes due to the new harris environment
	  qs_force_types.F: Small syntax cosmetics
	  qs_rho_methods.F: Additional duplicate_rho_type function
	  qs_scf.F: Changes due to the new harris environment
	  xc.F: Fix of a forgotten TD-DFT check
	  harris_env_types.F: The new harris environment

2005-07-11 16:34  fawzi

	* [r3964] in WRITE if UNIT=, then FMT= (no positional arguments
	  after labeled ones)

2005-07-11 14:51  marcella

	* [r3963] some allocations and initializations for lr calculations

2005-07-08 19:49  tlaino

	* [r3962] making more general the correction radius implementation
	  for the qm/mm.. added
	  also a new regtest.

2005-07-07 22:37  tlaino

	* [r3959] bug fix introduced with the previous commit and cleaning
	  of the qmmm_gaussian compensate gaussian

2005-07-07 14:29  tlaino

	* [r3958] minor changes

2005-07-07 14:04  tlaino

	* [r3957] few bug fixes.. no changes in regtests...

2005-07-07 11:40  tlaino

	* [r3956] cleaning the qmmm and the bsse.. Cosmetics around.. non
	  changes in regtest

2005-07-06 23:13  tlaino

	* [r3954] completing the N-body BSSE calculation..

2005-07-05 23:07  tlaino

	* [r3953] Deleting the globenv%program_name and globenv%run_type..
	  moved to an integer variables...
	  cosmetics around.. regtest ok!

2005-07-04 12:02  marcella

	* [r3952] missing keyword release

2005-07-04 10:23  krack

	* [r3951] bug fix: OPTIONAL harris flag

2005-07-04 08:31  krack

	* [r3950] T. Kuehne: Minor changes due to the integration of the
	  harris energy and force
	  type into the QS environment. Integration of the harris energy
	  and force types
	  into the QS environment

2005-07-04 08:30  krack

	* [r3949] T. Kuehne: Integration of the harris force computation

2005-07-04 08:29  krack

	* [r3948] T. Kuehne: Optional result variables are added in the
	  parameter list

2005-07-04 08:28  krack

	* [r3947] - Added calculation of the self and overlap energies
	  - Rearrangement of the USE statements

2005-07-04 08:28  krack

	* [r3946] T. Kuehne: Harris force type

2005-07-04 08:27  krack

	* [r3945] T. Kuehne: Simple straightforward force calculation to
	  the harris energy functional

2005-07-04 08:26  krack

	* [r3944] harris_force & harris_force_types added

2005-07-04 06:55  krack

	* [r3942] FUNCTION changed to SUBROUTINE (to allow for all checks
	  again with NAG 64Bit)

2005-07-01 19:54  tlaino

	* [r3941] preliminary stuff on bsse...

2005-07-01 13:39  marcella

	* [r3940] comments

2005-07-01 10:16  marcella

	* [r3939] First steps towards linear response calculations

2005-06-26 22:24  tlaino

	* [r3938] bug fix for derivatives grid case "compatible".. regtest
	  ok!

2005-06-23 19:16  tlaino

	* [r3936] possibility to dump out the unidimensional MM potential..
	  GRID potential in the COMPATIBILITY case
	  modified to reproduce the correct behaviour of the CPMD
	  implementation.

2005-06-23 14:06  tchassai

	* [r3935] Fix SAOP for excitations from unrestricted wavefunctions.

2005-06-23 13:00  tchassai

	* [r3934] Fix displaying of beta spin contributions to excitation
	  energies

2005-06-23 06:48  krack

	* [r3932] Thomas Kuehne: Minor speed improvement and removal of
	  debug code

2005-06-23 06:47  krack

	* [r3931] Thomas Kuehne: duplicate_qs_force added

2005-06-22 21:40  marcella

	* [r3930] change added_mos only if xas_calculation

2005-06-22 20:17  vondele

	* [r3929] Bug fix, don't add a random number of added mos

2005-06-22 16:52  tlaino

	* [r3927] bug fix for ONFO.. rewriting of few stuff.. organizing
	  the cp_subsystem_methods.F ..

2005-06-22 12:09  marcella

	* [r3926] allocation of fm matrixes for localization without pools

2005-06-21 22:23  marcella

	* [r3923] xas : number of states

2005-06-21 07:56  krack

	* [r3922] OpenMP: variable name corrected

2005-06-21 07:55  krack

	* [r3921] OpenMP: missing line continuation added

2005-06-21 07:54  krack

	* [r3920] OpenMP: missing declaration added

2005-06-20 21:40  ikuo

	* [r3919] Move routines around so we can put in EBO potential next
	  for FIST.

2005-06-20 17:31  ikuo

	* [r3918] Remove extra USE statement for DEC and sort

2005-06-20 17:12  ikuo

	* [r3917] Remove extra USE statement for DEC and sort

2005-06-17 17:59  marcella

	* [r3916] selection of the states in xas

2005-06-17 15:48  tchassai

	* [r3915] Improve output of LSD excitation energies.

2005-06-17 15:46  tchassai

	* [r3914] Bugfix for LSD excitation energies.

2005-06-17 12:41  krack

	* [r3913] Thomas Kuehne:
	  * Transition from the sum of eigenvalues to tr(K*P) to simplify
	  possible force calculations.
	  * print -> write
	  * even more cleanup and removal of debug code

2005-06-17 09:44  vondele

	* [r3912] wrap long line

2005-06-16 20:56  vondele

	* [r3911] other cases of recursive IO

2005-06-16 20:44  vondele

	* [r3910] remove Non-standard recursive I/O

2005-06-16 07:00  marcella

	* [r3909] clean up

2005-06-15 23:36  marcella

	* [r3908] wrong index

2005-06-15 21:48  ikuo

	* [r3907] check molecules of the same kind have same number of atom
	  and name sequence

2005-06-15 21:07  vondele

	* [r3906] bugfix, used ivalue but not set

2005-06-15 21:00  vondele

	* [r3905] set ierr

2005-06-15 20:48  vondele

	* [r3904] iargc isn't external

2005-06-15 20:37  marcella

	* [r3903] ops

2005-06-15 19:43  marcella

	* [r3902] I deleted too much

2005-06-15 17:42  fawzi

	* [r3901] smaller fixes in input (missing nullify)
	  Thomas Kuehne:
	  -cleanup
	  -error handling
	  -energy_calculation -> energy_summation
	  -trace(K*P) <-> energy_calculation

2005-06-15 16:06  marcella

	* [r3900] more on the input

2005-06-14 17:59  fawzi

	* [r3899] Thomas Kuehne, Harris Functional, First checkin

2005-06-14 17:57  fawzi

	* [r3898] Thomas Kuehne, Harris Functional, first checkin

2005-06-14 15:26  tchassai

	* [r3897] Bugfix (only visible with ifort and SUN)

2005-06-14 14:12  marcella

	* [r3895] bug fix

2005-06-14 11:28  tchassai

	* [r3894] Bugfix in 2nd derivative

2005-06-14 07:48  marcella

	* [r3893] new input for the dft xc global sections

2005-06-13 20:37  marcella

	* [r3890] parsing in newinput style dft xc and global sections

2005-06-13 09:26  vondele

	* [r3889] added zero atoms check

2005-06-10 17:57  marcella

	* [r3888] same as before

2005-06-10 17:54  marcella

	* [r3887] new input xc, not yet tested!!!

2005-06-10 15:57  vondele

	* [r3886] print force components only at 'HIGH'

2005-06-10 15:30  vondele

	* [r3885] subcell info only for 'HIGH'

2005-06-10 15:25  vondele

	* [r3884] distribution_2d output only for 'HIGH'

2005-06-10 15:06  vondele

	* [r3883] reduced topology output at 'medium' print level, full
	  topology generation printout still available at 'high'

2005-06-10 10:28  tlaino

	* [r3881] New QM/MM keyword: COMPATIBILITY. Since now this keyword
	  will be set to default. It enables the use of the same radii
	  used in the QM/MM of CPMD. So the migration from that code to
	  ours is really smmoth for the setting of the calculation.

2005-06-08 19:15  marcella

	* [r3878] small things

2005-06-08 10:21  marcella

	* [r3876] write the oscillator strengths

2005-06-08 10:17  fawzi

	* [r3875] small bugfixing

2005-06-07 22:46  marcella

	* [r3874] xas oscillator strengths by dipole in velocity form

2005-06-07 09:48  marcella

	* [r3873] deallocation

2005-06-07 08:14  vondele

	* [r3872] a version that works with -C=undefined if -D__DEBUG__

2005-06-06 23:01  tlaino

	* [r3871] Cleaning the topology%constraint.. not needed anymore..

2005-06-06 22:59  tlaino

	* [r3870] reverting to the WORKING version of input_cp2k!!

2005-06-06 22:13  marcella

	* [r3869] bugs fix

2005-06-06 18:30  marcella

	* [r3868] new option

2005-06-06 17:55  mcgrath

	* [r3867] Small bug fix.

2005-06-06 17:16  tlaino

	* [r3866] FORCEFIELD section converted to the new input style..
	  still few more and we'll be ready for the big movement..
	  some cleaning around..

2005-06-06 14:38  marcella

	* [r3864] missing nullify

2005-06-06 14:07  marcella

	* [r3863] more indexes for cp_fm_gemm

2005-06-06 12:00  marcella

	* [r3862] undefined variable

2005-06-06 11:29  marcella

	* [r3861] deleted unuseful variable

2005-06-06 06:45  vondele

	* [r3860] init ierr

2005-06-05 13:17  vondele

	* [r3857] unused USE

2005-06-05 13:13  vondele

	* [r3856] changed name from forward/backward to rs2pw/pw2rs

2005-06-05 12:21  vondele

	* [r3855] clean unused pointers

2005-06-05 12:06  vondele

	* [r3854] first try to disentangle pw_env_rebuild

2005-06-05 09:38  vondele

	* [r3853] prettified

2005-06-05 09:26  vondele

	* [r3852] prettified

2005-06-04 15:30  gtb

	* [r3851] modified to make fist running again on pgi-4. Regtest ok

2005-06-04 09:22  vondele

	* [r3850] removed one unneeded rs_pool_create_rs, added timeset

2005-06-04 09:21  vondele

	* [r3849] unused USE

2005-06-04 09:21  vondele

	* [r3848] added timeset

2005-06-03 17:19  marcella

	* [r3845] xas tp is done

2005-06-03 16:35  vondele

	* [r3842] now right version for rs_pw_to_cube

2005-06-03 14:15  vondele

	* [r3841] repeated line

2005-06-03 13:50  fawzi

	* [r3840] using global logger for input check

2005-06-03 13:43  vondele

	* [r3839] bug fix : don't use print level before it is defined,
	  define a default

2005-06-03 11:27  vondele

	* [r3837] allow for a rs grid argument to rs_pw_to_cube & and use
	  the pools to save some memory

2005-06-02 18:13  gtb

	* [r3836] fix build with pgi. Regtest working

2005-06-02 09:00  vondele

	* [r3835] fix build with xlf

2005-06-01 20:53  tlaino

	* [r3834] Default unit of measure specified in the html manual and
	  in the xml description.

2005-06-01 14:48  tlaino

	* [r3833] cleaning memory leaks introduced with the cp_unit.F ..
	  Bug fix for conversion cm^-1 to time a.u.
	  Bug fix for MD output.

2005-06-01 13:18  tlaino

	* [r3831] parsing with the new input style of &MD_NEW section.. In
	  the future it will be converted to &MD, but for the moment we
	  prefer to keep two different names.. cosmetics around.. bug fix
	  in cp_units.F.. Using the unit conversion of the new input style
	  within the KG regtest. Regtest is OK! (sorry gloria.. I deleted
	  what you introduced yesterday.. now it is consistent!!!)

2005-06-01 08:14  marcella

	* [r3830] xas with transition potential almost finished

2005-05-31 16:41  gtb

	* [r3829] bug fix in converting units

2005-05-31 09:52  krack

	* [r3827] Cray XT3 machine file

2005-05-31 09:22  tlaino

	* [r3826] few bug fixes for the parsing of the units...

2005-05-30 17:44  ikuo

	* [r3825] Fix water_3_g3x3.inp regtest results...

2005-05-30 14:15  fawzi

	* [r3824] using strings instead of the unit type to define the unit
	  of a keyword.

2005-05-30 09:32  krack

	* [r3823] call mpi_dims_create only if needed

2005-05-30 04:30  ikuo

	* [r3822] Forgot to uncomment the potential fix...

2005-05-29 03:39  ikuo

	* [r3821] Fix regtest failures from last checkin!!!! I think...

2005-05-27 21:36  ikuo

	* [r3820] Speedup for topology_generate_bond

2005-05-27 09:48  fawzi

	* [r3819] fixing stupid bug (sorry I had messed up the files to
	  check in)

2005-05-27 07:31  tchassai

	* [r3818] Bugfix for LSD + TD-DFRT + SAOP

2005-05-26 14:06  tchassai

	* [r3817] Bugfix. Xalpha LSD doesn't really need the norm of the
	  gradient :)

2005-05-26 00:38  cjmundy

	* [r3816] Even more crazy-stupid stuff.
	  I feel as if I am on an archeological
	  dig and uncovering relics from a distant
	  past.

2005-05-26 00:29  cjmundy

	* [r3815] More stupid NHC stuff pertaining to
	  definitions of molecules, etc.

2005-05-25 19:57  marcella

	* [r3814] some small bug fixing

2005-05-25 09:55  fawzi

	* [r3813] adding handling of units to the input

2005-05-25 05:57  cjmundy

	* [r3812] Rewritten RATTLE_ROLL. Works
	  correctly in parallel for all
	  constraint types. Regtests for
	  NPT with constraints reset.

2005-05-24 10:50  tlaino

	* [r3809] compilation bug fix ifort

2005-05-23 10:50  vondele

	* [r3806] bug fix

2005-05-22 14:21  vondele

	* [r3804] Fix parallel bug in pw_get_left_right

2005-05-22 13:04  vondele

	* [r3803] cleanup: remove unused and buggy features (print
	  density_matrix_magnitude, overlap_matrix_magnitude)

2005-05-20 09:53  fawzi

	* [r3802] fixing small shift of center with odd numbered grids
	  (that don't contain the 0)

2005-05-19 20:42  cjmundy

	* [r3801] Bug fix for parallel NPT.
	  Problem was that I was not
	  replicating eta, and f for thermostats.
	  Lets hope NAG is happy now.
	  Serial regtest not effected.

2005-05-19 16:33  tlaino

	* [r3799] bug fix for fix atom constrained with QS..

2005-05-19 13:54  tlaino

	* [r3797] removing debug print statements..

2005-05-19 13:53  tlaino

	* [r3796] Fix atom constraint cleaned! They can be now accessed
	  ONLY with the new input style (see tests in regtest).
	  Several bug fix concearning the $FIXED_ATOM section keywords. Bug
	  fix for annealing. Resetting two tests in regtest for qm/mm
	  inputs using the fix atom keywords: QM_SUBSYS and MM_SUBSYS.

2005-05-19 13:51  tlaino

	* [r3795] Bug fix for qm/mm electrostatic coupling. Error in vector
	  address..

2005-05-18 22:08  cjmundy

	* [r3794] Parallel bug fix for EAM.

2005-05-18 18:42  cjmundy

	* [r3793] Fix parallel restarts for NPT

2005-05-18 17:04  cjmundy

	* [r3792] Parallel bug I introduced

2005-05-18 09:00  tlaino

	* [r3791] Annealing of ions during dynamics.

2005-05-18 06:36  krack

	* [r3790] unused (empty) interfaces removed; usage hint and _dp
	  added

2005-05-17 19:40  tlaino

	* [r3788] Implemented the fixed atom constraint..

2005-05-17 11:48  marcella

	* [r3786] environment for xas calculations

2005-05-16 22:22  ikuo

	* [r3785] FIX NHC parallel restart

2005-05-16 22:21  ikuo

	* [r3784] Remove unused "USE"

2005-05-13 11:27  tlaino

	* [r3783] bug fix..

2005-05-13 08:26  tlaino

	* [r3781] Introduced the possibility to specify a different
	  forcefield for the QM/MM
	  classical nonbonded interactions. new test in regtest..

2005-05-12 21:00  tlaino

	* [r3780] restoring the release of the logger.. this should clean
	  all the memory leaks..

2005-05-12 12:00  tlaino

	* [r3778] bug fix in forcefield for particular qm/mm
	  calculations... regtest is OK...

2005-05-12 06:57  marcella

	* [r3777] replaced file

2005-05-11 20:17  tlaino

	* [r3776] bug fix.. unnullified pointers

2005-05-11 19:17  marcella

	* [r3775] new initialization for xas

2005-05-11 18:45  tlaino

	* [r3774] removing write statement ;-)

2005-05-11 14:30  tlaino

	* [r3771] Introducing the possibility to have atoms different from
	  H as capping atoms.
	  Introducing the pseudo-potential link scheme.
	  Moving zeff from integer to real. We have z as integer.. The Zeff
	  can also be modified for QM/MM calculations.

2005-05-11 13:55  fawzi

	* [r3770] * cleaner handling of non destroyied force_envs in
	  finalize_cp2k
	  * fixed bug in derived interface calc_force with n_el_f=0

2005-05-11 09:27  fawzi

	* [r3767] using atomic units for set_pos instead of angstrom.

2005-05-10 15:25  vondele

	* [r3765] cp_to_string needed for the include

2005-05-07 21:28  krack

	* [r3762] Ambigious interface commented

2005-05-07 20:39  krack

	* [r3761] Error handling added

2005-05-06 09:12  krack

	* [r3760] kind of seed changed from integer to real

2005-05-03 15:37  marcella

	* [r3759] new localization should work

2005-05-02 12:33  krack

	* [r3758] OpenMP directives fixed

2005-05-01 22:07  cjmundy

	* [r3757] * Fixed parallel bug for EAM
	  * Fixed another parallel issue from parallel regtest
	  * I Will fix the rest of FIST parallel issues for
	  regtest by weeks end

2005-05-01 12:54  krack

	* [r3755] Parallel (pseudo)random number generator for multiple
	  random number streams
	  and substreams.

2005-04-29 19:10  marcella

	* [r3754] some more stuff about localization

2005-04-29 17:57  cjmundy

	* [r3753] Zeroing EAM contributions

2005-04-29 17:48  cjmundy

	* [r3752] More leaks fixed.

2005-04-29 12:57  vondele

	* [r3750] fix bug for S**2, introduced last december.

2005-04-29 00:40  cjmundy

	* [r3748] *cleaning
	  *added CALL spline_env_release ( spline_manybody_env, err )

2005-04-28 16:29  cjmundy

	* [r3747] Manybody potential functionality for
	  EAM like potentials

2005-04-28 16:24  cjmundy

	* [r3746] *Modifications for EAM (see regtest)
	  *Spline now part of potparm (not other way around)
	  *some cleaning
	  *Passed regtest locally on intel sdbg and sopt

2005-04-28 14:32  tchassai

	* [r3741] Better fix (consistent with rest of the file)

2005-04-28 13:45  tchassai

	* [r3740] Bug fix for SUN

2005-04-28 12:46  tchassai

	* [r3739] TD-DFPT update:
	  Now supports excitations for the unrestricted case
	  Added a switch for the use of the precontitionder in the Davidson
	  algo.
	  Small bugfix in 2nd derivatives of xc

2005-04-28 11:14  vondele

	* [r3738] possibly improved error messages for parallel runs

2005-04-28 11:05  vondele

	* [r3737] fix the bug fix

2005-04-28 09:17  vondele

	* [r3736] fix parallel bug in output handling

2005-04-28 08:36  vondele

	* [r3735] fix parallel bug (flush on -1)

2005-04-27 13:11  vondele

	* [r3734] fix parallel problem (backspace on unit -1)

2005-04-26 09:37  krack

	* [r3733] bug fix (typo)

2005-04-25 23:22  cjmundy

	* [r3731] NPT_I restarts now working.
	  Thanks, Joost!

2005-04-22 17:31  marcella

	* [r3730] analytic calculation of the pos operator via Berry phase

2005-04-22 08:41  vondele

	* [r3729] reftraj 'ensemble'

2005-04-21 18:30  vondele

	* [r3728] IFC users will need to define __IFC in the arch file.
	  IFORT users get a fewer wraparounds in the wall clock time

2005-04-21 07:47  krack

	* [r3726] Optional printout of Cartesian MOs instead of spherical
	  MOs, e.g.
	  &PRINT
	  CARTESIAN_MOS
	  &END

2005-04-21 07:44  krack

	* [r3725] too many continuation lines

2005-04-21 06:56  krack

	* [r3724] line was too long

2005-04-20 22:06  marcella

	* [r3723] initialization of the new driver for the localization
	  option

2005-04-19 21:22  vondele

	* [r3720] added IPBV bonded contributions (QUARTIC bonds)

2005-04-19 13:23  vondele

	* [r3719] init

2005-04-19 13:09  vondele

	* [r3718] initialize

2005-04-19 12:05  vondele

	* [r3716] possibly fix some of the recent restarting problems

2005-04-18 20:33  vondele

	* [r3714] minor fix: init symbol

2005-04-17 15:30  vondele

	* [r3713] just prettify'ed

2005-04-15 16:28  marcella

	* [r3712] PBE kinetic energy functional

2005-04-15 07:37  vondele

	* [r3711] added timestart/stop

2005-04-14 22:00  cjmundy

	* [r3710] memory leak

2005-04-14 21:42  cjmundy

	* [r3709] *md_energies rid of globals
	  *md_env contains mdio
	  *read in of md *.restart files done on single ionode

2005-04-14 20:03  vondele

	* [r3708] first step of moving mulliken charges to the new print
	  keys

2005-04-13 09:58  vondele

	* [r3707] clarify logic

2005-04-12 13:59  fawzi

	* [r3706] better formatting for dipole, corrected parallel bug.

2005-04-06 19:10  vondele

	* [r3705] unit=-1 fallout

2005-04-06 19:06  vondele

	* [r3704] unit=-1 bug

2005-04-06 19:02  vondele

	* [r3703] fix some parallel regtest bugs (unit=-1 writes)

2005-04-05 12:38  marcella

	* [r3701] failure not initialized

2005-04-05 10:22  marcella

	* [r3699] KG_GPW: extrapolation of the mos

2005-04-01 11:31  jgh

	* [r3697] Allow for z=0 (ghost atoms)

2005-03-31 22:10  marcella

	* [r3696] KG GPW : distribution_2d disactivated

2005-03-31 21:32  mcgrath

	* [r3695] And a new source file, to make things work

2005-03-31 19:23  ikuo

	* [r3691] Write to the ionode only...

2005-03-31 10:04  marcella

	* [r3690] output_unit

2005-03-29 14:07  vondele

	* [r3687] fix OMP compilation for ifort 8.1.24

2005-03-24 13:02  fawzi

	* [r3684] fixing my previous bugfix, now odd number of point (no
	  zero) on the coarsest grid now works.

2005-03-23 21:45  ikuo

	* [r3683] Bug fix for "REORDER" for HF example...

2005-03-23 21:44  ikuo

	* [r3682] Remove duplicate USE statement for DEC compiler...

2005-03-23 15:12  tlaino

	* [r3681] just cosmetics...

2005-03-23 09:44  tlaino

	* [r3679] Bug fix IO in parallel with print level > than LOW

2005-03-22 22:29  tlaino

	* [r3678] just a little rewriting for the parallel stuff..

2005-03-22 15:11  tlaino

	* [r3676] Using the interpolator section within the qm/mm section..
	  modifying regtests accordingly

2005-03-22 14:31  vondele

	* [r3675] NAG upgrade fallout (Release 5.0(361))

2005-03-22 13:04  tlaino

	* [r3674] Introduced a new section interpolator in qmmm_section..
	  QMMM-GRID works in parallel..
	  few more stuff to let QMMM GAUSS work in parallel...

2005-03-21 16:54  tlaino

	* [r3671] another little step towards the QM/MM running parallel..
	  few bug fix around for write statement in parallel..

2005-03-18 16:56  marcella

	* [r3670] broken broadcast

2005-03-17 00:52  fawzi

	* [r3668] correct handling of common_iter_levels, dipol out

2005-03-16 23:57  tlaino

	* [r3667] bug fix in the generation of the MT reference grid.
	  regtests ok!

2005-03-16 21:09  tlaino

	* [r3666] bug fix : missing f_shift...

2005-03-16 17:31  tlaino

	* [r3665] bug fix for commensurated grids with odd numbers of
	  points.

2005-03-16 10:00  tlaino

	* [r3664] Removing the last memory leaks for semiempirical tests...
	  There should be no other leaks around.

2005-03-16 01:02  tlaino

	* [r3663] Martyna-Tuckerman works in parallel both the full
	  decoupling and the 2D.

2005-03-16 01:01  tlaino

	* [r3662] bug fix in pw_sumup

2005-03-15 17:08  tlaino

	* [r3660] correctly handle case with no electron (ncol==0)

2005-03-15 16:00  tlaino

	* [r3659] modifying the template according the previous version of
	  cp_lbfgs_optimizer_gopt.F

2005-03-15 14:47  krack

	* [r3657] XML output of the internal CP2K input structure.
	  Usage: cp2k.sopt --xml will generate the XML file cp2k_input.xml

2005-03-15 10:14  fawzi

	* [r3656] stupid dipole

2005-03-15 10:14  fawzi

	* [r3655] normalize fortran of template generation

2005-03-15 08:54  tlaino

	* [r3654] Fixing memory leak in GAPW

2005-03-14 23:41  tlaino

	* [r3652] metadynamics section available with the new input. Added
	  PERIODIC keyword in the poisson section to specify what kind of
	  poisson solver one wants to use. few modification to MT for
	  parallel runs (still not working yet!). Several bug fix after the
	  change of the cell into pointer... Regtest OK.

2005-03-14 20:41  marcella

	* [r3651] GAPW : loop over spins internal

2005-03-14 13:47  tchassai

	* [r3650] Enable SAOP,LB,GLLB for GAPW-TDDFPT.

2005-03-14 13:46  tchassai

	* [r3649] New options "LB" and "GLLB" for orbital eigenvalue
	  correction in
	  conjunction with TDDFPT.

2005-03-14 10:08  tlaino

	* [r3648] Bug fix for cell type.. deleting the INTENT, TARGET...
	  making POINTER...

2005-03-13 13:56  tlaino

	* [r3646] Martyna-Tuckerman 2D periodicity working now. new regtest
	  on a surface slab.
	  I modify the qs_scf in order to set the stating density to be
	  zero (just skip all the ATOMIC contribution)
	  if the number of electron is zero. First modifications to make MT
	  working in parallel.

2005-03-12 11:25  tlaino

	* [r3645] Fixing an unitialized array before the relocation..

2005-03-11 21:56  tlaino

	* [r3644] bug fix in OT for ortho and non-ortho basis set ...
	  Resetting the two SE regtests because their reference was built
	  with a faked source code..

2005-03-11 08:57  tlaino

	* [r3642] Fixing unsassociated pointer in KG stuff and few
	  modifications for Martyna-Tuckerman 2D.

2005-03-10 22:52  marcella

	* [r3640] external control

2005-03-10 21:46  marcella

	* [r3639] XAS_TP: energies with halfhole

2005-03-10 17:25  vondele

	* [r3637] further leak fixes

2005-03-10 16:56  vondele

	* [r3636] fix some memory leaks

2005-03-10 16:45  tlaino

	* [r3635] Martyna-Tuckerman 3D decoupling periodic images. Moved
	  the green function to be defined as complex...
	  2 New regtests.

2005-03-10 15:08  vondele

	* [r3633] first step in making OT compatible with SE, and
	  restarting of SE

2005-03-09 22:03  marcella

	* [r3631] external exit for qs_scf and bfgs_optimizer

2005-03-09 18:37  marcella

	* [r3630] not uniform occupation number

2005-03-09 17:45  vondele

	* [r3629] just zero, so it is initialised in case of overlap

2005-03-09 17:06  vondele

	* [r3626] make sure data is initialised in all cases

2005-03-09 13:06  marcella

	* [r3625] faster initialization for kg_gpw

2005-03-09 09:38  vondele

	* [r3623] keep up-to-date

2005-03-09 00:06  marcella

	* [r3622] initialization cube_info for pw_env_mol

2005-03-08 15:56  vondele

	* [r3621] speed up writing of restart files using blocked message
	  passing

2005-03-07 19:01  mcgrath

	* [r3620] making the dat file a little more general...eventually it
	  will disappear, but not yet

2005-03-07 15:46  jgh

	* [r3619] Fix some OpenMP related bugs (compilation only)

2005-03-07 14:32  tlaino

	* [r3618] more bug fix for martyna-tuckerman poisson solver

2005-03-07 12:58  tlaino

	* [r3617] Bug fix in reading &CELL with the new input.. added
	  PERIODIC keyword...
	  Preparing environment for Martyna-Tuckerman Poisson solver...

2005-03-07 11:53  tchassai

	* [r3616] Towards gapw saop potential.

2005-03-07 09:09  tchassai

	* [r3615] Cosmetic changes to the output of TDDFPT

2005-03-06 23:11  tlaino

	* [r3614] changing name to greens_fn.F greens_fns.F into
	  green_types.F and green_methods.F
	  Poisson section moved to MM section and DFT section...
	  Regtest OK!

2005-03-06 15:46  tlaino

	* [r3610] (first) bug fix for cell_ref (thanks Matt..)

2005-03-06 14:43  mcgrath

	* [r3609] a couple changes for MC regtests

2005-03-06 13:00  tlaino

	* [r3608] Major change: Removing % green from cell.. cell is a
	  pointer (now contains really only information
	  regarding the cell).. it can be clean further, but the major
	  change has been done. Regtest checked and working...
	  A make distclean could be requested..

2005-03-05 09:37  tlaino

	* [r3607] bug fix for section cell_ref name.

2005-03-05 00:06  tlaino

	* [r3606] Moving all variables switching methods in
	  input_constants.F, to fix few cyrcular dependencies and to
	  organize everything a little bit better.. &CELL &END can now be
	  parsed with the new input style. The part of code that will be
	  deleted (once all conversion will be ready) is marked with the
	  labels "!TODEL_START_TL" "!TODEL_END_TL".. Let's try to use the
	  same labels and append instead of TL your initials.. In this way
	  it will be much more easy to find and to remove the unused code..
	  Regtest OK..

2005-03-04 13:33  vondele

	* [r3605] remove some unused variables

2005-03-03 23:27  marcella

	* [r3604] KG_GPW restart

2005-03-02 15:08  mcgrath

	* [r3602] Cleaned up the MC section.

2005-03-02 15:00  tchassai

	* [r3601] Refactoring GAPW XC.

2005-03-02 14:45  tlaino

	* [r3600] Added MC section with the new input style.. Matt, you
	  should fill the missing description fields ;-).

2005-03-02 13:22  jgh

	* [r3599] Add differential operator (velocity dipole), not tested

2005-03-02 08:02  tlaino

	* [r3598] Deleting the OHH_MC stuff... we don't need it anymore..
	  (Matt is ok with this ;-) )

2005-03-02 07:01  krack

	* [r3597] Allow compilation by the latest Intel compiler releases.
	  The parser of the
	  latest release seems to have bugs.

2005-03-01 20:57  marcella

	* [r3596] lpp tf pw86 and pw91 ke functional + lsd: fixed

2005-02-28 20:04  marcella

	* [r3595] energy output for kg_gpw

2005-02-28 14:43  tlaino

	* [r3591] runtime bug fix..

2005-02-25 11:47  marcella

	* [r3584] some factrors for the KE functionals have changed (bug
	  fix)
	  differences in KG energies

2005-02-24 19:57  tlaino

	* [r3583] let's start with the new input with FIST... Passing on
	  the error variable to
	  fist_control_force.. few more dummy things.. I'll go on starting
	  from Monday..
	  In an hour will start my long WE ;-)

2005-02-24 17:17  mcgrath

	* [r3582] Preparing for PBE run

2005-02-24 15:17  vondele

	* [r3581] fix problem with section with the same name overwriting
	  html manual. Using '~' in the file names to avoid bug in mozilla
	  with "%" in file names.

2005-02-24 13:35  tlaino

	* [r3578] adding a useful variant..

2005-02-24 10:24  tlaino

	* [r3577] Print keywords in dft, qmmm and fist... now it is really
	  completed ;-)

2005-02-24 10:23  tlaino

	* [r3576] really using the new input.. removing hardcoded
	  paramaters..

2005-02-24 09:05  tlaino

	* [r3575] fixing memory_leaks... clean again.. ;-)

2005-02-23 22:32  marcella

	* [r3574] AAA bug in the expression of the efactors for PW91

2005-02-23 21:25  vondele

	* [r3573] remove int_group, use para_env instead

2005-02-23 20:44  vondele

	* [r3572] get rid of a basically unused module variable (glob_env)

2005-02-23 15:00  tlaino

	* [r3571] Adding semiempirical keywords and moving the
	  create_xc_fun_section into the input_cp2k.F..
	  All sections are now in this file... next step.. PRINT and IO
	  sections... hopefully ready for tonight...

2005-02-23 12:36  tlaino

	* [r3569] Completed the New Input parsing. GEOOPT, MD, METADYN have
	  been moved into
	  a new section: MOTION at the same level as FORCE_EVAL.

2005-02-23 12:35  tlaino

	* [r3568] Bug fix for not nullified pointers...

2005-02-22 23:47  marcella

	* [r3567] mp_sum

2005-02-22 14:11  tlaino

	* [r3566] Bug fix..

2005-02-21 22:24  tlaino

	* [r3565] bug fix of unassociated pointers...

2005-02-21 21:51  marcella

	* [r3563] perdew86 = p86c

2005-02-21 21:11  tlaino

	* [r3562] Tightening condition on qm/mm links.. no more than one qm
	  atom can be linked to the same MM atom..

2005-02-21 19:55  tlaino

	* [r3560] Bug fix.. no changes in regtests...

2005-02-21 17:21  marcella

	* [r3558] not initialized variable

2005-02-21 16:42  tlaino

	* [r3557] bug fix for addfine2coarse (fawzi).

2005-02-21 16:35  marcella

	* [r3556] KG_GPW: shift of the grid and ATOMIC guess

2005-02-21 13:02  tlaino

	* [r3555] compilation bug fix

2005-02-21 12:50  tlaino

	* [r3554] enabling repeats for QM_KIND mm_index keyword (sometime
	  in qm/mm we have to specify so many atoms that they don't fit in
	  the 240 character maxlong line (parser).. in this way we can just
	  split the atom indexes on multiple lines... without changing the
	  parser)

2005-02-21 12:07  jgh

	* [r3553] analytic integrals for arbitrary moments (r^n)
	  analytic integrals for Berry phase operator exp(-I k.r)
	  analytic integrals for angular momentum
	  integrals are not tested
	  updated comments and corrected typos

2005-02-20 22:14  tlaino

	* [r3551] So many things...
	  (1) Bug fix in QM/MM for PBC (this fix WILL CHANGE the results of
	  H2O-qmmm-gauss-7.inp and C11H24-qmmm-gauss-0.inp). Bug fix for
	  the radius of the atom in the qm/mm link treatment (this fix will
	  not change any test)...
	  (2) Introduced the possibility to freeze only QM or MM atoms in
	  MD and GEO_OPT...
	  (3) Introduced the possibility to have fix_atom in MD or GEO_OPT
	  in Fist and in qm/mm
	  (4) Changed the format of the wite_topology_psf according the
	  real standard of PSF (now these PSFs can be read also through
	  VMD)..
	  (5) Added few more things in the parsing of the input (very few
	  things miss before we can start with the real usage of the new
	  input)...

2005-02-20 20:27  tlaino

	* [r3550] Fixing memory leak..

2005-02-18 17:44  tlaino

	* [r3549] Bug fixing for undefined pointers and initialization
	  variables...

2005-02-18 17:31  marcella

	* [r3548] preparation for the wf extrapolation with KG_GPW

2005-02-17 16:53  marcella

	* [r3546] KG_GPW : correction and forces are now working

2005-02-17 09:03  vondele

	* [r3545] fix build

2005-02-17 08:23  tchassai

	* [r3544] * Cleaned up tddfpt_environment
	  * Added function to find the contribution of occ to virt
	  transitions in
	  a given excited state (this is automatically printed at the end
	  of
	  a tddfpt run)
	  * Cleaner output

2005-02-17 00:36  tlaino

	* [r3543] tiny bug fix.. no changes in regtests...

2005-02-16 14:18  tlaino

	* [r3540] restoring the L-BFGS-B. Updated to the 2.4 version.
	  restored regtest too.
	  w.r.t. the old regtest there is a difference around 10^-10. This
	  is probably
	  due to a different use of lapack/blas...
	  Preliminary setup of the delocalized coordinates. Still some more
	  work to do..

2005-02-16 09:10  tchassai

	* [r3539] Forgot to remove debugging code.

2005-02-16 09:07  tchassai

	* [r3538] Workaround for PGI 5.2 popt.

2005-02-16 08:42  tchassai

	* [r3537] Moved xc_deriv_method_id and xc_rho_smooth_id from
	  dft_control to xc_control

2005-02-15 18:47  marcella

	* [r3535] minor changes

2005-02-15 16:43  gtb

	* [r3533] -MD functionality of KG_POL restored, some bugs removed
	  -KG_POL coeffs parameters read and used from the new input
	  (regtests coming soon)

2005-02-15 16:01  krack

	* [r3532] allow for additional flags added by mpirun

2005-02-15 13:24  tchassai

	* [r3531] normalized

2005-02-15 11:29  tlaino

	* [r3530] deleting the call to EP...

2005-02-15 11:29  tlaino

	* [r3529] cleaning memory leaks in fist for npt.

2005-02-15 10:20  tchassai

	* [r3528] Fixed memory leaks.

2005-02-15 10:04  marcella

	* [r3527] KG_GPW : kinetic energy correction

2005-02-15 09:03  tchassai

	* [r3526] Fix using a different xc functional in the TDDFPT Kernel.

2005-02-15 09:03  krack

	* [r3525] unused component iatom removed from particle_type

2005-02-15 07:59  krack

	* [r3524] build and update of FIST neighbor lists merged

2005-02-14 16:08  ikuo

	* [r3522] Added some description for the topology section.

2005-02-14 07:54  vondele

	* [r3521] this is the file that contains open and close _file

2005-02-14 07:52  vondele

	* [r3520] first try to separate open_file and close_file from
	  qs_parser.
	  Still needs improved error handling (which is difficult due to
	  circular dependencies)

2005-02-13 17:58  tlaino

	* [r3518] Bug fix reading the improper parameters. NO changes in
	  regtest!

2005-02-12 14:29  tlaino

	* [r3513] bug fix for xlf on OS/X.. the last update of the compiler
	  January2005 can build
	  the sopt executable without modification to the source files (the
	  previous bug due to an internal
	  compiler error, has been fixed.. it's never not too much late..)
	  Still few problems (compiler internal errors) with the sdbg
	  version.

2005-02-11 19:50  tlaino

	* [r3509] Added FIST input sections. Will or Chris, could you
	  please complete the missing keywords declaration? Thanx ;-)

2005-02-11 19:27  marcella

	* [r3508] KG GPW calculation of forces

2005-02-11 16:40  fawzi

	* [r3507] sections_vals_val_get now also accepts paths separated by
	  %

2005-02-11 16:13  gtb

	* [r3505] *** empty log message ***

2005-02-11 13:40  vondele

	* [r3503] - move named constants into a new file input_constants to
	  avoid dependencies
	  - introduced path to get sections FORCE_EVAL%DFT%QS%OT now works
	  - OT can use the new input now

2005-02-11 12:04  vondele

	* [r3501] check default value for strict enums

2005-02-11 11:46  vondele

	* [r3500] bolt

2005-02-11 00:07  marcella

	* [r3499] subsys is not an argument of qs_vxc_create. The smoothing
	  through zero_core has been disactivated

2005-02-10 18:37  krack

	* [r3498] NPT should work again (switched back to scaled
	  coordinates)

2005-02-10 16:22  tchassai

	* [r3497] Just a cosmetic thing.

2005-02-10 16:19  vondele

	* [r3496] improved 'SOURCE' behavior

2005-02-10 16:19  tchassai

	* [r3495] Switched tddfpt to the new input.

2005-02-10 16:03  vondele

	* [r3494] remove useless variables

2005-02-10 15:46  tlaino

	* [r3493] deleting some unused stuff...

2005-02-10 14:26  fawzi

	* [r3491] kg new input.

2005-02-10 13:40  vondele

	* [r3490] Introduced new keyword NOCENTER (default=.false.). It is
	  useful if you don't want to center the QM system every step of
	  dynamics (check, etc..). Teo.

2005-02-09 21:38  cjmundy

	* [r3489] Better format for
	  
	  * CPU time [now 2 digits past decimal]
	  * time [now in F20.12 ]

2005-02-09 18:01  cjmundy

	* [r3488] Checks for unwated bends against
	  specified distance constraints.
	  Will affect regtests for NH3_dist_*

2005-02-09 15:45  vondele

	* [r3487] KG NAG proof...

2005-02-09 01:28  ikuo

	* [r3484] Fix for PARA_RES with PSF

2005-02-08 12:32  krack

	* [r3479] FIST neighbor list update, i.e. re-using of the allocated
	  lists and nodes;
	  NPT is still wrong

2005-02-08 12:15  tchassai

	* [r3478] SAOP correction for the eigenvalues in a TDDFT
	  calculation now working.

2005-02-08 07:37  vondele

	* [r3477] fix PGI problem

2005-02-08 01:31  cjmundy

	* [r3476] 4x6 constraints debugged on liquid ammonia
	  * nve
	  * nvt
	  * npt_i
	  Will commit regtest and examples soon

2005-02-07 15:56  krack

	* [r3475] possibly fixes the FIST crash + correct kind
	  specification

2005-02-07 15:50  krack

	* [r3474] possibly fixes the FIST crash

2005-02-07 11:01  krack

	* [r3473] memory usage of FIST neighbor lists reduced again

2005-02-07 10:39  krack

	* [r3472] getting rid of scaled_to_real calls; cleaned

2005-02-07 10:06  tlaino

	* [r3471] Cleaning memory leaks

2005-02-07 05:41  cjmundy

	* [r3470] Debugging 4x6.
	  Almost there.

2005-02-05 00:27  ikuo

	* [r3469] Bug fix for the exclusion list builder...

2005-02-04 18:34  ikuo

	* [r3467] Speed up building the exclusion list for FIST benchmark.

2005-02-04 09:14  vondele

	* [r3465] cleaned the mulliken constraint

2005-02-04 01:45  cjmundy

	* [r3464] Fix for NH3

2005-02-03 20:26  cjmundy

	* [r3461] Bug fixes for distant constraints.
	  e.g. normal SHAKE. g3x3 are fine.

2005-02-03 17:20  ikuo

	* [r3460] Change my mind, will update with a better exclusion list
	  builder next time...

2005-02-03 17:08  ikuo

	* [r3458] Bug fix for building the exclusion list...

2005-02-03 14:59  krack

	* [r3457] qs_neighbor_list_types without GAPW addons in order to
	  remove FIST memory
	  bottleneck. Not really nice in the spirit of CP2K, since FIST and
	  QS should
	  share common data types, but a quick and clean fix which does not
	  break
	  QS/GAPW.

2005-02-03 14:55  krack

	* [r3456] - particle_type cleaned
	  - memory bottleneck in FIST neighbor list build fixed
	  - cosmetics (like unused USE removed)

2005-02-03 01:32  ikuo

	* [r3455] Reduce amount of time to build exlusion list for most
	  cases

2005-02-02 22:34  ikuo

	* [r3454] zero nsgf and nelectron in molecule_kind_set to fix fist
	  load balancing error

2005-02-02 21:53  marcella

	* [r3453] KG_GPW : minor changes

2005-02-01 22:47  ikuo

	* [r3450] Ability to have residues as individual molecules.

2005-02-01 17:44  tlaino

	* [r3449] deleting unused routine... diamat

2005-02-01 16:58  tlaino

	* [r3448] Ops.. reverting to eigenvalueproblems.. need a little bit
	  more investigation...
	  please hold on...

2005-02-01 14:56  tlaino

	* [r3447] deleting eigenvalueproblems.F ..

2005-02-01 14:36  tlaino

	* [r3446] Deleting eigenvalueproblems.F => moving the diagonalise
	  call in NPT_F to diamat ...
	  No changes in regtest! Hey Chris.. Did you check your failure
	  with PME ? ;-)
	  Teo.

2005-02-01 01:21  cjmundy

	* [r3445] uses ODD grid points for small-grid in PME.
	  Makes comparison to EWALD exact in limit.

2005-02-01 01:07  cjmundy

	* [r3444] 2500 spline-points to pass regtest

2005-02-01 01:04  cjmundy

	* [r3443] *Parallel debug routine works (this is
	  used to debug the residues. Will has it
	  done. I have seen it. It is wonderful. He
	  is checking all regtests and finding failures
	  with the QM/MM so we will find these before
	  committing.) We are finally in the dawn of
	  the age of enlightenment.
	  
	  *Parallel slow ewald works

2005-01-31 12:25  tchassai

	* [r3442] Bugfix for 2nd derivatives in the LSD case.

2005-01-30 11:40  tlaino

	* [r3439] Fixing memory leaks.. clean again..

2005-01-29 20:51  vondele

	* [r3438] allow for running small systems on a large number of
	  CPUs, fix the case where some CPUs have no part of the pw grid.

2005-01-29 16:02  vondele

	* [r3437] hopefully true as well

2005-01-29 14:16  tlaino

	* [r3435] Fully generalized qm/mm link scheme. Additional support
	  for generalized charge shift and dipole adjust.
	  More tests in regtests...

2005-01-28 23:21  mcgrath

	* [r3434] forgot to remove a write statement

2005-01-28 23:16  mcgrath

	* [r3433] some more mc testing...things seem fine

2005-01-28 10:25  tchassai

	* [r3432] Fixed memory leak

2005-01-27 20:05  vondele

	* [r3429] useless USe

2005-01-27 19:50  marcella

	* [r3428] GAPW: lsd + mpi problem solved

2005-01-27 19:33  vondele

	* [r3427] remove unused new_h_matrix

2005-01-27 19:25  vondele

	* [r3426] remove useless USE

2005-01-27 08:55  vondele

	* [r3425] init variables

2005-01-26 20:00  marcella

	* [r3424] bug in the 3 center integrals

2005-01-25 20:05  cjmundy

	* [r3421] Returned to 2500 spline points

2005-01-25 18:40  ikuo

	* [r3420] Bug fix for torsion generate.

2005-01-25 13:50  tchassai

	* [r3419] - New way to calculate the singlet excitations using 2
	  times the lda kernel
	  - New tddfpt keyword: lsd_singlets to change to lsd kernel for
	  singlets

2005-01-24 17:52  cjmundy

	* [r3417] Added the IPBV model of water.
	  Should be tested with Matt's version
	  although I think it is fine.
	  
	  Added timings in the MD output

2005-01-22 23:43  tlaino

	* [r3415] Some more printing info... new qm/mm tests to check that
	  every internal potential in the
	  qm/mm bonding region is computed properly..

2005-01-22 18:14  tlaino

	* [r3414] Bug Fix => Now really deletes the contribution to
	  internal forces of qm atoms...

2005-01-22 17:11  marcella

	* [r3413] deallocation soft basis set

2005-01-22 13:34  tlaino

	* [r3410] Restoring the zeroing of the internal force field for the
	  QM fragment...

2005-01-22 11:03  tlaino

	* [r3408] Fixing not initialized arrays...

2005-01-21 23:44  ikuo

	* [r3407] Add wildcard matching for torsion (not available for
	  improper yet).
	  This will change the regtest result for ethene in fist.
	  Fix dangling pointers in clean_intra_fist.

2005-01-21 20:47  marcella

	* [r3404] some corrections

2005-01-20 22:32  ikuo

	* [r3397] update of clean_intra_force

2005-01-20 18:20  ikuo

	* [r3396] Update clean_intra_force_kind to account for constraints

2005-01-19 22:27  tlaino

	* [r3394] introducing an add_env in qmmm_env.. restyling of few
	  subroutines.

2005-01-19 22:16  ikuo

	* [r3393] Fix core dump problem when in parallel

2005-01-18 16:04  tlaino

	* [r3391] fixing memory leak

2005-01-18 10:14  tlaino

	* [r3390] fixing another little bug... the failing test in regtest
	  should be fine now..

2005-01-18 10:07  tlaino

	* [r3389] bug fix. no change in regtest.

2005-01-17 17:11  tlaino

	* [r3387] This is the commit with the real scaling of MM charge
	  involved in qm/mm link. new test in regtest.

2005-01-17 09:32  tlaino

	* [r3383] introduced the possibility to rescale the MM charge on
	  the MM boundary link atom. Default is to rescale the charge by
	  0.0. I.e. the MM link atom has no effective charge.

2005-01-15 09:11  vondele

	* [r3382] init failure=.false.

2005-01-14 21:48  tlaino

	* [r3381] another change in write forces...

2005-01-14 17:33  tlaino

	* [r3380] correct write statement if ewald_type none.

2005-01-14 00:03  ikuo

	* [r3376] bond,bend,ub,tor,14 match to NAMD
	  nonbond(vdw+ele) match to NAMD/PINY

2005-01-13 21:52  tlaino

	* [r3375] making clear few print statements.

2005-01-13 21:50  tlaino

	* [r3374] Added feature HETATM in read_pdb.

2005-01-13 15:58  tchassai

	* [r3372] Improved formatting

2005-01-13 15:55  tchassai

	* [r3371] Small improvement for the Davidson diagonlizer

2005-01-12 14:55  tlaino

	* [r3369] Fixing write units

2005-01-12 14:31  tlaino

	* [r3368] Bugfix .. no changes in regtests.. Added 2 more regtests
	  to tight checks on topology
	  generation.

2005-01-11 16:39  vondele

	* [r3367] follow-up fixes to remove unused variables

2005-01-11 15:21  tlaino

	* [r3366] Added a missing finish_parser..

2005-01-11 14:50  tlaino

	* [r3365] Including another possibility while reading external
	  potential (we could use
	  in QM run atmname not matching with the element defining the
	  kind..)
	  changing the link atom name in H_LINK.

2005-01-11 00:50  tlaino

	* [r3363] IMOMM qm/mm link scheme completed. Debugged derivatives.
	  Bug fix in torsion_list for qm/mm calculations. Cleaning the
	  write output for fist derivatives. 3 more new tests.

2005-01-09 21:04  vondele

	* [r3361] bug fix: in case of the rotation keyword (restricted/sic)
	  do no overwrite the electronic gradient, which is still needed
	  for the ionic forces. Changes all N3-X.inp (X>2) regtests.

2005-01-09 21:00  vondele

	* [r3360] added a debugging hook to print the electronic gradient,
	  taking the lagrangian multipliers into account

2005-01-09 20:43  tlaino

	* [r3359] improving the topology build molecule for qm/mm runs.
	  another little step towards the qm/mm links. added the utility
	  for the IMOMM qm/mm link scheme.

2005-01-09 16:36  tlaino

	* [r3356] Bug Fix (related to the default change of rcut_nb)!

2005-01-09 16:12  vondele

	* [r3355] write to the right output unit

2005-01-08 10:42  vondele

	* [r3354] Changed the format of the MD energies file, in particular
	  the 6th column is
	  back the conserved quantity (as before rev. 1.16 in july). (It
	  the past few
	  months it has been {E-E0}/E0, which is a related number but with
	  a drift that is
	  roughly 1000 times smaller). The 7th column is now the time per
	  MD step in seconds.

2005-01-07 23:30  ikuo

	* [r3353] Fix factor of 2 for improper torsion to match NAMD for
	  improper torsions.
	  Modify default rcut to be 10A if key work RCUT_NB is not set.
	  Will change answer for examples in regtest.

2005-01-07 23:28  ikuo

	* [r3352] Add 1-4 energy term to energy. Will change any energy for
	  some regtest in fist.

2005-01-07 07:51  vondele

	* [r3351] speed improvements for the special case

2005-01-06 21:48  marcella

	* [r3350] minor changes

2005-01-06 21:03  vondele

	* [r3348] some speedup

2005-01-06 17:56  vondele

	* [r3347] kill last two memory leaks in the regtester

2005-01-06 17:53  vondele

	* [r3346] make IMPLICIT NONE, and declare variables

2005-01-06 09:18  marcella

	* [r3345] KG_GPW: new file name

2005-01-06 09:17  marcella

	* [r3344] KG_GPW: collocate molecular densities on the molecular
	  grids

2005-01-06 05:33  cjmundy

	* [r3343] ONEFOUR bugfix

2005-01-05 18:35  ikuo

	* [r3342] torsion iteration 2...

2005-01-05 17:20  ikuo

	* [r3341] torsion iteration 1...

2005-01-05 15:06  tchassai

	* [r3340] Added new control parameter n_restarts

2005-01-05 14:56  tchassai

	* [r3339] Minor changes

2005-01-05 14:14  tchassai

	* [r3338] Added nullify.

2005-01-05 14:06  tchassai

	* [r3337] qs_tddfpt_utils.F: Some cleanup
	  qs_tddfpt_eigensolver.F: Bugfix

2005-01-05 13:44  tchassai

	* [r3336] Major rehaul of the eigensolver:
	  - Block Davidson now really works :)
	  - New initial guess from the K=0 solution
	  - Both Davidson and Lanczos work with restarts
	  
	  cp_control_types.F & cp_control_utils.F: changed max_iter to
	  max_kv
	  qs_tddfpt_module.F & qs_tddfpt_eigensolver.F: lots of changes

2005-01-04 19:06  ikuo

	* [r3335] Missing variables for OpenMP

2005-01-04 13:07  tchassai

	* [r3334] Refactoring tddft

2005-01-04 13:05  tchassai

	* [r3333] Added array initialization.

2005-01-04 10:21  tlaino

	* [r3332] Just cosmetics. Defining a my_qmmmm logical variable.

2005-01-03 17:58  tlaino

	* [r3330] Cleaning and improving the topology part generating
	  molecule for qmmm runs. New test in QS/regtest.

2005-01-03 04:03  cjmundy

	* [r3324] PME fix. Now gives correct EWALD limit
	  when big grid = small grid (which is the
	  beauty of the method).
	  
	  It also checks for non-orthorhombic boxes.
	  
	  Since Teo has also been seemingly interested in this routine,
	  I decided to make sure it really works again. :-)
	  
	  Happy New Year!

2005-01-01 23:39  cjmundy

	* [r3323] More cleaning

2005-01-01 18:47  cjmundy

	* [r3322] cleaned up fist debug routines.
	  Now, all is printed in nice format
	  without silly user input. Use
	  keyword "debug" for simulation type.

2004-12-31 07:19  tlaino

	* [r3321] Cleaning memory leaks for pme :: Deleting the attribute
	  ALLOCATABLE, SAVE...

2004-12-30 17:58  tlaino

	* [r3320] cleaning introduced memory leaks..

2004-12-30 17:14  tlaino

	* [r3319] Let's get (it) started with qm/mm links... IMOMM setup...
	  almost ready...

2004-12-29 22:18  ikuo

	* [r3315] Fix so we can parse directly the downloadable charmm FF

2004-12-29 22:17  ikuo

	* [r3314] Remove duplicate USE statement for DEC compiler

2004-12-29 22:16  ikuo

	* [r3313] Add some some missing output

2004-12-29 13:56  tlaino

	* [r3312] fixed the qmmm mechanical coupling.

2004-12-29 08:47  tlaino

	* [r3311] Introducing qmmm_env_mm_type. Added a qmmm_topology_util
	  module.

2004-12-27 18:11  tlaino

	* [r3310] just cosmetics.. changing qmmm_env_type =>
	  qmmm_env_qm_type ..

2004-12-27 11:13  tlaino

	* [r3305] Now also PME works with Fist... added a regtest in
	  Fist/regtest...
	  tuning the qs_debug routine checking small value of forces

2004-12-26 18:40  vondele

	* [r3304] fixes

2004-12-24 11:13  vondele

	* [r3302] other unreachable stuff

2004-12-24 11:10  vondele

	* [r3301] yet another removeal

2004-12-24 11:04  vondele

	* [r3300] can go as well

2004-12-24 06:34  vondele

	* [r3299] can go

2004-12-23 17:41  tlaino

	* [r3298] Bug Fix for qmmm when using PSF...

2004-12-22 21:51  marcella

	* [r3297] KG_GPW : first part local grids

2004-12-22 21:50  marcella

	* [r3296] KG_GPW: first part local grids

2004-12-22 09:49  tlaino

	* [r3295] Fixing memory leaks introduced after the main commit on
	  cell_ref...

2004-12-22 01:19  tlaino

	* [r3294] Removing 10094 lines of unused code!!! MAKE DISTCLEAN
	  REQUIRED...

2004-12-21 23:07  tlaino

	* [r3293] bug fix after cell_ref...

2004-12-21 21:58  tlaino

	* [r3292] removing file not even in OBJECTDEFS... Teo.

2004-12-21 19:05  cjmundy

	* [r3289] Some cleaning.
	  Cell_ref now cleaner.
	  Hugoniot relation printed for NPH
	  keyword CONNECTIVITY OFF for large structurless systems (faster)
	  PASSED regtest for QS and QMMM

2004-12-21 16:11  gtb

	* [r3288] obsolete

2004-12-21 16:07  gtb

	* [r3287] removing no longer used files

2004-12-21 16:04  gtb

	* [r3286] not longer used routines

2004-12-21 16:01  gtb

	* [r3285] Obsolete routines

2004-12-20 15:33  vondele

	* [r3284] remove stray print statement

2004-12-20 14:30  tlaino

	* [r3283] Note: after updating this file, please make a
	  DISTCLEAN...

2004-12-20 14:25  tlaino

	* [r3282] Fixing Misleading Binary output for Fist or (QM/MM)
	  runs...

2004-12-20 12:41  tchassai

	* [r3281] Debugging GAPW+TDDFT

2004-12-20 12:16  tchassai

	* [r3280] Debugging GAPW+TDDFT

2004-12-20 09:48  vondele

	* [r3279] improved format for timings in the case of long runs

2004-12-19 23:10  tlaino

	* [r3278] Cleaning memory leaks in FIST when using PSF.. again as
	  white.. ;-)))

2004-12-19 21:23  tlaino

	* [r3276] debug_this_module set to .FALSE.

2004-12-19 20:36  tlaino

	* [r3273] Bug Fix for qmmm forces.
	  Introduced new logical variable to control the translation of the
	  system
	  within qm/mm run.
	  Several cosmetics on qm/mm.

2004-12-19 20:35  tlaino

	* [r3272] New routine to debug analytical forces with numerical
	  forces. It can easily be used
	  with every CP2K environment.

2004-12-19 20:33  tlaino

	* [r3271] Bug Fix in using print level HIGH and DEBUG.. Now they
	  work again within QS...

2004-12-19 20:31  tlaino

	* [r3270] removing the qmmm_list (we don't need it anymore)..
	  introducing the logical variable
	  qmmm in fist_env..

2004-12-19 20:29  tlaino

	* [r3269] Bug Fix for ewald in Fist.

2004-12-18 10:38  tchassai

	* [r3266] Refactoring GAPW-TDDFPT

2004-12-17 12:32  tchassai

	* [r3265] Refactoring

2004-12-15 23:06  ikuo

	* [r3264] Fix for matt's read_topology_mc

2004-12-15 17:22  gtb

	* [r3263] fixed some bugs

2004-12-15 10:21  tchassai

	* [r3262] Workaround for PGI and g95 compilers.

2004-12-14 21:31  marcella

	* [r3260] KG_GPW: colmplete scf without kinetic functional
	  correction

2004-12-14 15:01  tchassai

	* [r3259] Debugging gapw-tddft

2004-12-14 14:58  tchassai

	* [r3258] Added _dp that was left out

2004-12-14 14:54  tchassai

	* [r3257] Slightly better Lanczos initial guess and removed memory
	  leak.

2004-12-14 14:51  tchassai

	* [r3256] Refactoring.

2004-12-14 14:49  tchassai

	* [r3255] Nothing new.

2004-12-13 17:27  tlaino

	* [r3254] Adjusting Print options in qm/mm... cosmetics in debug
	  qm/mm

2004-12-13 05:05  cjmundy

	* [r3249] Bygfix for NPT_I with constraints

2004-12-12 23:26  cjmundy

	* [r3248] NPT now works in parallel.
	  Added some output to the screen (Temp of barostat).

2004-12-10 16:39  tchassai

	* [r3247] Some refactoring of 2nd derivative.

2004-12-10 11:57  tlaino

	* [r3246] cleaning memory leaks...

2004-12-09 23:07  marcella

	* [r3245] KG_GPW: ao density mixing

2004-12-09 21:41  marcella

	* [r3244] KG_GPW : solved problems with the initialization

2004-12-08 21:01  marcella

	* [r3243] KG_GPW : diis & co molecule by molecule

2004-12-08 07:45  tchassai

	* [r3242] Removed unused variable ecoul_1c.

2004-12-07 16:05  tlaino

	* [r3241] Bug fix for write_particle_coordinates when using Fist...

2004-12-07 15:45  tlaino

	* [r3240] Bug fix for write_particle_coordinates when using Fist...

2004-12-07 10:49  fawzi

	* [r3239] removing unused module

2004-12-07 09:53  tlaino

	* [r3238] Improving the qm/mm input...

2004-12-07 00:29  tlaino

	* [r3234] Removing memory leaks...

2004-12-06 18:31  tlaino

	* [r3233] Introducing PBC in qm/mm :: results of regtests
	  Ar-qmmm.inp, H2O-qmmm.inp, H2O-qmmm-gauss.inp
	  may change...

2004-12-06 16:20  fawzi

	* [r3232] removing debugging print & cube generation.

2004-12-06 16:02  tlaino

	* [r3231] Bug fix : zeroing grids...

2004-12-06 13:59  tlaino

	* [r3228] FIST:: Bug Fix when using several molecules! Missing
	  molecule_kind
	  FIST:: Bug Fix in one-four interactions! Using %nonbond instead
	  of %nonbond_14

2004-12-06 10:21  jgh

	* [r3227] save intermediate work for SE code

2004-12-05 22:34  tlaino

	* [r3226] Bug Fix in reading PSF.
	  Zeroing several variables in md_ener..

2004-12-05 16:50  fawzi

	* [r3225] * changed ne print keys
	  * connected iteration info to old print level
	  * fixed spline3 transfer without pbc (parallel case not working)
	  * more tests for qmmm_gauss method (energy consistency)

2004-12-05 16:47  fawzi

	* [r3224] cosmetics

2004-12-05 16:46  fawzi

	* [r3223] removed lbfgs optimizer

2004-12-05 16:43  fawzi

	* [r3222] more informative debug output

2004-12-04 23:34  cjmundy

	* [r3221] Found NPT bug and towards
	  damping for NPH.

2004-12-03 12:38  tchassai

	* [r3220] Changed allocation calls to catch istat and added
	  initializations.

2004-12-03 09:47  tchassai

	* [r3219] * Changed options to tddfpt
	  * Rearranged allocation calls so that istat really catches
	  allocation
	  errors.

2004-12-03 09:40  tchassai

	* [r3218] Changed reallocates to allocates.

2004-12-02 22:31  ikuo

	* [r3217] Bug fix for clean_intra_force_kind

2004-12-02 19:18  ikuo

	* [r3215] PSF read update

2004-12-02 15:12  tchassai

	* [r3214] Bugfix for SUN: removed sourcefileref

2004-12-02 14:48  marcella

	* [r3213] KG_GPW : first part eigensolver

2004-12-01 23:02  cjmundy

	* [r3211] Debugged NPH_UNIAXIAL compression from
	  a shock-wave ensemble. Seems to work
	  well. Now, lets shock some biological
	  matter!!!! LJ example will be checked
	  in.

2004-12-01 13:47  vondele

	* [r3210] bug fix : EOF implies iostat<0 (-4001 on pathscale)

2004-11-30 16:40  vondele

	* [r3209] removed

2004-11-30 15:29  marcella

	* [r3208] one more file

2004-11-30 14:56  marcella

	* [r3207] KG_GPW: first density matrix : random guess

2004-11-30 10:05  vondele

	* [r3206] further updates for the single precision version of cp2k.
	  Introduced a mp type that stands for 'maximum precision', it used
	  in a few places where we do a sum over all grid points (hartree,
	  xc, total rho) and is useful to gain 1000 times more accurate
	  total energies in single precision.

2004-11-30 00:37  cjmundy

	* [r3204] bug fix

2004-11-29 22:56  cjmundy

	* [r3203] Big time compile problems on DEC. Multiple declarations
	  of DP was the culprit.
	  
	  Also, added a new ensemble to study systems
	  under a shock compression. Reference is in
	  integrator.F. Its a beautiful application
	  of extended Lagrangians.

2004-11-29 06:34  vondele

	* [r3202] further fixes for generic precision / sgl code

2004-11-28 21:14  tlaino

	* [r3201] bug fix for write_elp_potential when using FIST with
	  PRINT LEVEL DEBUG...

2004-11-28 16:14  vondele

	* [r3200] towards generic floating point precision

2004-11-27 21:32  vondele

	* [r3199] farming enhancements (useful if queuing systems aren't
	  flexible):
	  MAX_STEP (maximum number of jobs / mpi task executed)
	  RESTART (continue where a previous farm left of)
	  CYCLE (cycle through the jobs without end)

2004-11-27 00:23  marcella

	* [r3198] kg_gpw : preparation of the calculation of the density
	  matrix

2004-11-26 20:21  vondele

	* [r3197] a bit of a cleanup of rs_pw_transfer, and a small speedup
	  if run on many CPUs

2004-11-26 10:15  vondele

	* [r3196] just one instance of dp needed

2004-11-26 09:58  vondele

	* [r3195] More consistent use of dp instead of dbl

2004-11-26 07:57  marcella

	* [r3194] fix

2004-11-25 22:39  marcella

	* [r3193] KG_GPW: overlap orthogonalization

2004-11-25 16:21  vondele

	* [r3192] avoid variables called dp

2004-11-25 13:44  vondele

	* [r3191] removed unused stuff

2004-11-25 12:10  vondele

	* [r3190] fix build

2004-11-25 11:01  tchassai

	* [r3189] * Replaced pbl with dp.
	  * Refined reorthogonalization.

2004-11-25 10:55  tchassai

	* [r3188] Removed an unnecessary variable and added
	  initializations.

2004-11-25 10:54  tchassai

	* [r3187] Some cleanup.

2004-11-24 15:54  marcella

	* [r3186] generalized set_mo_occupation

2004-11-23 19:02  marcella

	* [r3184] KG_GPW create kg_scf_env

2004-11-23 10:56  vondele

	* [r3182] further OMP corrections

2004-11-23 10:33  vondele

	* [r3181] add isendrecv_rv

2004-11-23 09:29  tchassai

	* [r3180] * Tolerance parameter added.
	  * Most instances of glob_env removed.

2004-11-22 20:46  vondele

	* [r3178] test

2004-11-22 20:35  marcella

	* [r3177] with INTEGER*8 I have problems in compiling with ifc.
	  Changed into INTEGER

2004-11-22 20:32  marcella

	* [r3176] KG_GPW number of orbitals per molecule

2004-11-22 19:54  cjmundy

	* [r3175] dft%control needs to be in argument list
	  for read_tddfpt_control
	  
	  Constraints functionality in:
	  (distance, 3x3, 4x6, used with
	  SHAKE, RATTLE (NVE, NVT)
	  SHAKE_ROLL, RATTLE_ROLL (NPT_I, NPT_F))
	  
	  Will commit examples

2004-11-22 17:16  tchassai

	* [r3174] Update to incorporate new TDDFT parameter.

2004-11-22 17:15  tchassai

	* [r3173] A new parameter for TDDFT has been added.

2004-11-22 14:36  vondele

	* [r3169] small modifications, hopefully less binary data in the
	  output, more accurate timings, and fewer timing wraps.

2004-11-22 11:42  marcella

	* [r3168] KG_GPW forgot to add these

2004-11-22 11:19  marcella

	* [r3167] KG_GPW initialization of molecular blocks

2004-11-21 13:53  vondele

	* [r3166] fix typos in OT section

2004-11-21 13:50  vondele

	* [r3165] added the OT section to the new input style

2004-11-20 16:49  vondele

	* [r3164] help pgf

2004-11-20 15:33  vondele

	* [r3163] added a few keyword descriptors

2004-11-20 15:16  vondele

	* [r3162] simplify creating an array of equal length strings using
	  a new function s2a.
	  applied to cp2k input creation. Some small reformatting of the
	  html manual.

2004-11-19 16:26  vondele

	* [r3161] and first bug fix..

2004-11-19 16:24  vondele

	* [r3160] introduce a sparsity_id in the sparse matrix types. We
	  assume that the sparse
	  matrix id is determined by the neighbor_list construction (Only
	  QS part done).
	  Now one can easily assert that the sparsity of two matrices has a
	  chance to
	  be the same (see qs_scf).

2004-11-19 13:26  krack

	* [r3159] missing OMP declaration added

2004-11-19 12:50  tchassai

	* [r3158] Still needs debugging :(

2004-11-19 11:46  vondele

	* [r3157] cleaning, remove unused neighbor_lists_from_scratch

2004-11-19 10:16  vondele

	* [r3156] some cleaning of sparse_matrix_types, remove unused
	  functions

2004-11-19 07:56  krack

	* [r3155] numerical check substituted by a logical construct

2004-11-18 15:09  krack

	* [r3154] fist neighbor lists should work for triclinic boxes

2004-11-18 14:15  tchassai

	* [r3153] Fix memory leak.

2004-11-18 12:45  fawzi

	* [r3152] adding get_natom.

2004-11-18 11:13  fawzi

	* [r3150] enumeration keywords now can have a description.

2004-11-18 11:00  tchassai

	* [r3149] Further work on SAOP

2004-11-18 09:33  vondele

	* [r3148] error situation fix. Not very satisfactory I'd say...

2004-11-18 08:57  vondele

	* [r3147] now really fool proof ?

2004-11-18 02:13  fawzi

	* [r3146] fixing CPPostcondition

2004-11-17 22:56  cjmundy

	* [r3145] Working NPT_F.
	  
	  Again, with no
	  contraints.
	  
	  Some compile time
	  issues with qmmm_init.F
	  Had to comment a line.

2004-11-17 17:33  vondele

	* [r3144] enhanced (?) command line parsing

2004-11-17 15:16  tchassai

	* [r3143] Additions for SAOP in TDDFPT.

2004-11-17 12:37  vondele

	* [r3142] give 999 as a default unit, to avoid problems in case the
	  unit is used incorrectly (e.g. errors)

2004-11-17 11:42  vondele

	* [r3141] introduce version information in the manual

2004-11-17 11:00  vondele

	* [r3140] mention that index.html is nice to start

2004-11-17 10:55  vondele

	* [r3139] html manual formatting tweaks

2004-11-17 10:06  vondele

	* [r3138] make sic section up-to-date

2004-11-17 09:57  tchassai

	* [r3137] Added SIC option to new parser

2004-11-17 09:47  tlaino

	* [r3135] Cleaning and going towards a full working qm/mm
	  functionality...
	  Nullifying internal forcefields for qm atoms and so on...
	  Energy is conserved during a dynamic with GRID option.

2004-11-17 05:25  cjmundy

	* [r3134] Bugfix for non-ortho cell in linklists.
	  Must use cell lengths for non-ortho boxes

2004-11-16 18:06  fawzi

	* [r3133] f77 interface update.

2004-11-15 22:52  cjmundy

	* [r3132] Bug fixes for non-ortho cells

2004-11-15 20:22  cjmundy

	* [r3130] uses only one subcell grid.

2004-11-15 15:23  tchassai

	* [r3129] Reverting to 1.30 because the orbitals are not needed
	  here.

2004-11-15 09:23  tchassai

	* [r3123] Towards SAOP in TDDFT

2004-11-13 14:59  vondele

	* [r3122] remove distinction between caps and non-caps on input,
	  small reformatting

2004-11-12 22:36  fawzi

	* [r3121] * globenv passed as pointer more often ;)
	  * new input parsing now used in dft_control if there is the
	  global section in the inpu
	  * added first version of the new xc interface using the new input

2004-11-12 13:57  tchassai

	* [r3120] Corrected typo.

2004-11-12 10:59  fawzi

	* [r3119] moving hmd leak tests to cp2k.F (and removing wrongly
	  declared i)

2004-11-12 10:44  fawzi

	* [r3118] * bugfix when using PW_MODE_LOCAL in pw_grid_setup
	  * leak write on all processes (not only ionode)

2004-11-12 10:24  vondele

	* [r3117] hack needed for parallel runs with ifc 7.1 (looks like
	  ifc 7.1 bug)

2004-11-12 09:01  vondele

	* [r3115] fixes needed to run H (LSD, only serial tested)

2004-11-11 19:36  vondele

	* [r3112] stupid bug (there is no ierr)

2004-11-11 19:11  vondele

	* [r3111] remember ionode after destruction, only needed for __HMD

2004-11-11 17:10  tlaino

	* [r3108] IN/OUT Bug For NPT_I... ;-))))))))

2004-11-11 14:55  tlaino

	* [r3107] bug fix... Now is really ok.. checked with nag on os/x.

2004-11-11 14:44  tlaino

	* [r3106] xlf wrong.. ok.. now it compiles...

2004-11-11 14:42  tlaino

	* [r3105] Last cleanings of memory leaks! Fist is as pure as white!
	  ;-)

2004-11-11 13:12  tlaino

	* [r3104] Other bug fix... towards an overall clean fist... ;-)

2004-11-11 09:51  vondele

	* [r3103] bug fix: get the virial from force_env before using it

2004-11-10 21:57  tlaino

	* [r3102] Bug Fix...

2004-11-10 21:18  tlaino

	* [r3100] Bug Fix... Thanx Joost...

2004-11-10 21:10  cjmundy

	* [r3098] NPT_I functionality. See tests argon_npt.inp
	  No constraints yet.
	  virial_methods.F is a new stand
	  alone to compute pv_tot. Seemed
	  more logical especially with the
	  mind set of MC code

2004-11-10 20:27  tlaino

	* [r3096] Removing memory leaks... please see e-mail in cp2k
	  mailing list for comments...

2004-11-10 19:33  fawzi

	* [r3095] * cp2k commandline flags (try --help)
	  * updated f77_interface (some bugfixes like /= => == ;)
	  * storable timer_env and mp_perf_env to have the correct default
	  stack with multiple force_env with f77_interface
	  * minor input fixes

2004-11-10 18:40  ikuo

	* [r3094] Corresponding set of deallocate routines for
	  intramolecular forces

2004-11-10 15:10  tchassai

	* [r3093] Allow for SIC in OEI and use of parameters for the
	  control types.

2004-11-10 14:08  jgh

	* [r3092] 2nd try

2004-11-10 13:26  jgh

	* [r3087] correct number of active states in initial guess

2004-11-10 13:19  tchassai

	* [r3086] Further Orbital Eigenvalue Improvement.

2004-11-09 20:03  tlaino

	* [r3076] Missing TARGET statement. W FIST!

2004-11-09 18:31  ikuo

	* [r3074] Fist update: torsion power series compilation fix

2004-11-09 16:35  tchassai

	* [r3073] "Preparations for Eigenvalue improvement step in TDDFT"

2004-11-09 01:01  ikuo

	* [r3072] Add psf_write to generate PSF file for use in NAMD/CHARMM

2004-11-08 22:15  ikuo

	* [r3071] Cleanup, not in OBJECTDEFS

2004-11-08 21:51  ikuo

	* [r3070] Fist update... multiplicity torsion

2004-11-08 21:00  vondele

	* [r3069] expand : amat seems to require intent(INOUT)

2004-11-08 20:38  vondele

	* [r3068] OT cleaning part 4(?) : remove unused variables

2004-11-08 19:21  marcella

	* [r3067] GAPW: some optimization

2004-11-08 08:26  tchassai

	* [r3064] Added an separate runtype for TDDFPT:
	  use ELECTRONIC_SPECTRA or SPECTRA to call for a tddft
	  calculation.

2004-11-08 07:58  vondele

	* [r3063] also take out the arguments now

2004-11-07 20:27  vondele

	* [r3062] fix geo opt by fixing the template this time

2004-11-07 19:55  vondele

	* [r3061] OT cleaning part 3 : get rid of settings%ortho_k

2004-11-07 19:08  vondele

	* [r3060] OT cleaning part 2 : get rid of settings%n ,settings%k

2004-11-07 18:05  vondele

	* [r3059] OT cleaning part 1 : remove nblock_row, nblock_col

2004-11-07 17:59  jgh

	* [r3058] towards Ewald sum for NDDO

2004-11-07 15:27  marcella

	* [r3057] omp directives corrections

2004-11-05 21:12  vondele

	* [r3056] allow for GUESS ATOMIC and OT (medium quality initial
	  guess).

2004-11-05 15:37  jgh

	* [r3055] Don't touch original FNx values

2004-11-05 14:20  jgh

	* [r3054] Added special case for AM1 boron core repulsion

2004-11-05 13:46  fawzi

	* [r3053] * made pw_pool safer (now it is more likely that using a
	  pw that is in the pool is catched, resusing as simple array still
	  possible)
	  * fixed qmmm bug (bad merge of files)

2004-11-05 09:31  marcella

	* [r3051] problem in compiling with pgi for names conflict

2004-11-05 01:02  ikuo

	* [r3050] Fist UB fix
	  PDB charge from occup column fix

2004-11-04 23:09  marcella

	* [r3047] more on reduced 3c lists
	  small change in tot energy for gapw regetsts
	  due to the change of the definition of a cutoff radius

2004-11-04 21:24  fawzi

	* [r3046] activating f77 interface, you have to update also the
	  makefile!

2004-11-04 21:23  fawzi

	* [r3045] * f77 interface
	  * restored "easy" cp2k_run
	  * revised force_env interface

2004-11-04 21:21  fawzi

	* [r3044] cosmetic changes

2004-11-04 20:31  fawzi

	* [r3043] handle correctly required values in non required section.

2004-11-04 20:17  marcella

	* [r3042] more about the reduced 3c lists

2004-11-04 20:00  fawzi

	* [r3041] more parallel input bugfixes (at_end non in sync)

2004-11-04 17:55  marcella

	* [r3039] first steps for reduced 3center lists

2004-11-04 15:58  vondele

	* [r3037] experimental average density correction

2004-11-04 14:24  tchassai

	* [r3036] Bugfix.

2004-11-04 14:04  fawzi

	* [r3035] * fixed parallel bug in cp_error_synchronize_error
	  * fixed (many) parallel bug in cp_parser
	  * fixed bug in xc_rho_set_type (uninitialized my_rho_g_local
	  because give_back_pw was moved)

2004-11-03 22:51  cjmundy

	* [r3031] Fist force fields, bug free.

2004-11-02 17:40  fawzi

	* [r3029] xc cleanup phase 1, only the new routines survived (Fawzi
	  & Thomas)

2004-11-02 11:17  vondele

	* [r3028] fix memory leak

2004-11-02 09:06  vondele

	* [r3026] adding experimental interaction corrections

2004-11-01 14:31  tchassai

	* [r3025] Corrected typos.

2004-11-01 07:52  jgh

	* [r3024] de-prettify code

2004-10-31 16:37  jgh

	* [r3023] some code rearrangements (semi_empiric)

2004-10-30 21:23  cjmundy

	* [r3022] All Fist intra are debugged!
	  e.g. bonds, bends, urey-bradley,
	  improper torsions, torsion, one-four

2004-10-30 15:36  jgh

	* [r3021] NDDO: bug fixes + forces(RHF and UHF)

2004-10-29 12:26  tchassai

	* [r3019] Bug fix for memory leak fix.

2004-10-28 21:37  cjmundy

	* [r3016] bug in onefour

2004-10-28 19:53  ikuo

	* [r3015] Add default for nonbond cutoff
	  internal match of Urey-Bradley in debug

2004-10-28 14:33  tchassai

	* [r3012] Cosmetics.

2004-10-28 14:32  tchassai

	* [r3011] Different defaults for TDDFT.

2004-10-28 14:30  tchassai

	* [r3010] Removed some memory leaks.

2004-10-28 12:15  tchassai

	* [r3009] Removed a memory leak.

2004-10-28 12:07  tchassai

	* [r3008] Removed a memory leak.

2004-10-28 10:18  krack

	* [r3007] initialization of character variables

2004-10-28 07:31  tchassai

	* [r3006] Removed memory leak.

2004-10-28 07:27  tchassai

	* [r3005] Removed a memory leak.

2004-10-28 00:06  cjmundy

	* [r3004] Debug routines for Urey-Bradley and One-Fours

2004-10-27 16:16  ikuo

	* [r3000] Output the sum of qeff charges for molecule kind and
	  system total

2004-10-27 09:20  jgh

	* [r2999] LSD NDDO implemented

2004-10-26 22:04  ikuo

	* [r2997] fix to get bond,bend,torsion energy match between cp2k
	  and piny
	  using charmm w/ decalanine example input

2004-10-26 19:16  vondele

	* [r2996] further memory leak fixing

2004-10-26 19:09  vondele

	* [r2995] fix memory leaks

2004-10-26 14:43  tchassai

	* [r2991] Rooting out memory leaks.

2004-10-26 14:26  marcella

	* [r2990] GAPW : fix bug

2004-10-26 06:30  ikuo

	* [r2988] fist update: internal match of numerical and analytical
	  for nonbond, bond,
	  bend, and torsion...

2004-10-25 11:46  marcella

	* [r2987] fix leaks

2004-10-25 10:56  vondele

	* [r2986] Modifications for the HTML input reference manual. Needs
	  .FALSE. -> .TRUE. in the input_cp2k to be generated. Code is
	  still pretty ugly.

2004-10-25 10:12  marcella

	* [r2985] fix some memory leaks

2004-10-23 20:24  vondele

	* [r2984] fix (?) extended_system_type related memory leaks.
	  Extended_system_types init/create/dealloc/.. needs refactoring.

2004-10-23 16:46  vondele

	* [r2983] remove para_env as a useless argument from
	  cp_sm_fm_multiply

2004-10-23 16:21  vondele

	* [r2982] renaming of mo_set%eigenvectors to mo_set%mo_coeff to
	  indicate that they need not to be eigenvectors depending on the
	  SCF method used

2004-10-23 16:08  jgh

	* [r2981] Bug fix for OH and NH core repulsion term

2004-10-23 15:19  vondele

	* [r2980] generalise calculate_w_matrix_2 (rounding differences
	  influence the energy of OT MD runs by 1E-12)

2004-10-23 10:19  vondele

	* [r2979] fix incorrect format (regtest/pyridine.inp)

2004-10-23 09:41  jgh

	* [r2978] NDDO Hamiltonians (AM1,PM3,MNDO) bug fixes, energy seems
	  to be ok

2004-10-23 09:38  vondele

	* [r2977] nullify and deallocate ub_list impr_list onfo_list

2004-10-23 09:08  vondele

	* [r2976] fix memory leak.

2004-10-23 08:49  vondele

	* [r2975] fix introduced memory leaks

2004-10-22 20:13  vondele

	* [r2974] fix implicit type declarations

2004-10-22 19:37  vondele

	* [r2973] further explicit PRIVATE/PUBLIC module entities

2004-10-22 18:49  vondele

	* [r2972] this makes all modules explictly private/public

2004-10-22 15:30  vondele

	* [r2971] first step for introducing restricted calculations

2004-10-22 15:14  krack

	* [r2970] some fixes just to get the psmp version at least compiled

2004-10-22 15:07  gtb

	* [r2968] rewritten initialization part; partially debugged (only
	  KG part)

2004-10-22 09:48  vondele

	* [r2967] added a simplified interface to timeset and timestop

2004-10-22 04:11  cjmundy

	* [r2966] Debugging impropers

2004-10-22 00:13  cjmundy

	* [r2965] cleaning

2004-10-21 23:12  cjmundy

	* [r2964] cleaning

2004-10-21 22:40  cjmundy

	* [r2963] Transcription error for rattle.
	  Does not effect older versions

2004-10-21 20:51  ikuo

	* [r2962] Missing init and dealloc

2004-10-21 18:05  ikuo

	* [r2961] Some more little parts for Fist...

2004-10-21 16:56  tlaino

	* [r2960] Creating the qs_section, tddfpt_section, qmmm_section,
	  cell_section with the new input style. Increasing the length of
	  the usage field in keyword_type.
	  C'mon y'all, lets get the new input_style... ;-)

2004-10-21 15:55  jgh

	* [r2959] Allow for all_electron potential and semi-empirical
	  Hamiltonian

2004-10-21 15:13  jgh

	* [r2958] Further updates for semi-empirical Hamiltonians

2004-10-21 12:58  marcella

	* [r2956] error message if GPW with All-el is required

2004-10-21 11:25  vondele

	* [r2953] partial fix for atom kinds in the COORD section

2004-10-21 11:20  marcella

	* [r2952] GAPW: more accurate interaction radius for the
	  all-electron potential

2004-10-20 13:01  vondele

	* [r2949] Honor the do_rotation keyword, might change the initial
	  guess for OT with non-equivalent orbitals
	  (and thus non-converged energies, e.g. H2O-9.inp).

2004-10-19 22:22  ikuo

	* [r2947] Some more updates...

2004-10-19 19:57  vondele

	* [r2945] a first cleaning try remove trajana

2004-10-19 13:11  fawzi

	* [r2942] - removed pao
	  - removed linked list tests
	  - removed old ouput utilities
	  - new output utilities
	  - new print key based on the new input
	  - updated spline utils (now using the new input)

2004-10-19 10:28  vondele

	* [r2940] small cleanup

2004-10-19 10:26  marcella

	* [r2939] GAPW: bug fix

2004-10-19 10:15  marcella

	* [r2938] GAPW : bug fix

2004-10-19 09:59  marcella

	* [r2937] GAPW : bug fix

2004-10-18 22:28  cjmundy

	* [r2936] ewald_env optional for call to spline_nonbond_control

2004-10-18 18:18  marcella

	* [r2935] GAPW: cleaning

2004-10-18 09:53  tlaino

	* [r2934] Bug Fix! Missing a nullify...

2004-10-18 07:14  vondele

	* [r2932] fix build : wrong integer parameter, and character length

2004-10-17 21:33  cjmundy

	* [r2931] some bugs

2004-10-17 20:49  cjmundy

	* [r2930] All intra forces are in for bio related
	  apps. Fist_debug is updated to test
	  the new torsions.

2004-10-17 19:18  vondele

	* [r2929] fix build (write format)

2004-10-17 17:32  jgh

	* [r2928] Further updates towards semi empirical codes

2004-10-17 12:35  vondele

	* [r2927] remove old init_cell and rename init_cell_new into
	  init_cell

2004-10-17 08:58  vondele

	* [r2926] small format changes

2004-10-16 19:58  vondele

	* [r2925] small fixes for the geo opts, still not optimal, BFGS and
	  LBFGS should be more similar in input and output

2004-10-16 08:08  vondele

	* [r2921] fix compilation : break long lines

2004-10-15 22:58  cjmundy

	* [r2920] some cleaning

2004-10-15 22:26  fawzi

	* [r2919] input bugfix

2004-10-15 20:52  cjmundy

	* [r2918] Debug capabilility for the Bio code push.
	  Should help in debugging forces, virials,
	  neighborlists, for the heinous intra
	  stuff

2004-10-15 10:59  marcella

	* [r2917] GAPW: a bit of speed up

2004-10-14 20:38  jgh

	* [r2916] New stuff for semi empirical Hamiltonians

2004-10-14 20:34  jgh

	* [r2915] further output changes

2004-10-14 20:31  jgh

	* [r2914] change of output style and some other small stuff

2004-10-12 17:24  ikuo

	* [r2913] Memory leak fix: lci

2004-10-11 23:08  ikuo

	* [r2911] Cleaning up and restore functionality of CHARGE_BETA and
	  CHARGE_OCCUP keyword.

2004-10-08 11:48  tchassai

	* [r2909] Rest of the TDDFT GAPW implementation.

2004-10-08 08:34  tchassai

	* [r2908] Bugfix: parser_end was missing

2004-10-07 23:05  ikuo

	* [r2907] Split topology_control into
	  topology_control+force_field_control

2004-10-07 22:35  ikuo

	* [r2906] Split topology_control into
	  topology_control+force_field_control

2004-10-06 22:30  marcella

	* [r2905] GAPW: some fixing

2004-10-06 22:08  marcella

	* [r2904] GAPW : multigrid allocations

2004-10-06 15:48  tchassai

	* [r2903] Bugfix

2004-10-06 14:36  marcella

	* [r2902] 2nd derivatives with GAPW TDDFT almost done

2004-10-06 12:06  tchassai

	* [r2901] Work on the 2nd derivatives for GAPW TDDFT

2004-10-05 15:43  marcella

	* [r2900] spherical harmonics: special cases for cosn and sinn

2004-10-05 12:43  tchassai

	* [r2899] Introduced the logger for writing the warning

2004-10-05 12:43  tchassai

	* [r2898] Removed an unnecessary write statement

2004-10-04 14:32  marcella

	* [r2897] GAPW: deallocations

2004-10-04 10:56  vondele

	* [r2896] split long lines in the output (e.g. vi limit of 2048
	  chars on IBM, but we should try < 80)

2004-10-02 00:46  cjmundy

	* [r2895] Urey-Bradley, Improper Torsions, Onefours, in, not tested

2004-10-01 21:59  ikuo

	* [r2894] More Fist updates...

2004-10-01 19:05  ikuo

	* [r2893] More additions for UB, IMPR, ONFO

2004-10-01 15:57  marcella

	* [r2892] GAPW : keep zet0_s from being to big

2004-10-01 12:08  tchassai

	* [r2891] qs_kpp1_env_methods.F xc.F: refactoring to make assesible
	  to GAPW
	  qs_vxc_atom.F xc_atom.F: preparing to call xc_calc_2nd_deriv_new

2004-10-01 08:46  krack

	* [r2890] back to CPMD version

2004-09-30 23:51  ikuo

	* [r2888] Start of FIST: UB,IMPR, and ONFO

2004-09-30 20:51  ikuo

	* [r2887] Start of 4x6 const

2004-09-30 15:46  tchassai

	* [r2886] xc.F: extracted method to avoid duplicated code for GAPW
	  TDDFT.
	  xc_atom.F: add second density (rho and rho1)
	  qs_vxc_atom.F: moved a subroutine

2004-09-30 13:34  tchassai

	* [r2885] Add density construction for singlets and triplets.

2004-09-30 12:05  tchassai

	* [r2884] Extracted a fill_rho_set_method in anticipation of GAPW
	  TDDFT XC.

2004-09-30 12:02  tchassai

	* [r2883] Removed a unnecessairy stop_program.

2004-09-30 09:17  krack

	* [r2881] Bug fix (wrong prefactors)

2004-09-30 06:52  marcella

	* [r2880] GAPW: read rho0 exponents from input

2004-09-29 20:31  ikuo

	* [r2879] Missing OpenMP declarations...

2004-09-29 17:47  ikuo

	* [r2878] Bug fix for Fist Charge input...

2004-09-29 17:31  marcella

	* [r2877] GAPW: nullify l_info_rho0

2004-09-28 19:24  marcella

	* [r2876] GAPW : duplicate l_info if lmax_rho0 > maxlgto

2004-09-24 19:44  mcgrath

	* [r2875] forgot to change a couple lines back

2004-09-24 19:42  mcgrath

	* [r2874] changing more things with the dat file...

2004-09-24 15:36  ikuo

	* [r2873] Some tweaking for fist charge input.

2004-09-24 15:21  tchassai

	* [r2872] Add call to calculate_xc_2nd_deriv_atom

2004-09-24 15:20  tchassai

	* [r2871] Further work on GAPW/TDDFT xc calls.

2004-09-24 15:19  tchassai

	* [r2870] Adapt to new allocate_vtriple routine.

2004-09-24 09:04  fawzi

	* [r2869] goto?!? make that an exit (Joost)

2004-09-24 08:44  fawzi

	* [r2868] one last single to double change

2004-09-24 00:42  fawzi

	* [r2866] * new input update (parsing activated!!)
	  * minor xc fixes, switching to integer flags
	  * minor global fixes
	  * iteration info object

2004-09-23 23:57  fawzi

	* [r2865] removing some forall

2004-09-23 23:51  fawzi

	* [r2864] x.d0 -> _dp, cleanup

2004-09-23 22:12  mcgrath

	* [r2862] slowly making it more autonomous

2004-09-23 21:58  marcella

	* [r2861] GAPW : cleanup

2004-09-23 21:31  mcgrath

	* [r2860] some more output formatting

2004-09-23 20:54  fawzi

	* [r2858] removing old style output method

2004-09-23 15:30  tchassai

	* [r2857] Refactoring: extracted the calculation of rho and drho.
	  Added a routine to calculate the atomic contributions to xc in
	  GAPW.

2004-09-22 23:31  ikuo

	* [r2856] Bug Fix...

2004-09-22 16:11  marcella

	* [r2855] GAPW : fixing

2004-09-22 14:50  marcella

	* [r2854] GAPW: maxlgto takes into account the compensation density
	  rho0

2004-09-22 11:50  marcella

	* [r2853] GAPW: some changes in the initialization routines

2004-09-21 09:29  marcella

	* [r2852] GAPW : fixing of some memory allocations

2004-09-20 13:08  mcgrath

	* [r2850] changed some formatting, still trying to debug

2004-09-19 12:19  vondele

	* [r2849] switched back to old value

2004-09-17 10:25  marcella

	* [r2847] GAPW : changes for mpi parallelization

2004-09-15 21:01  vondele

	* [r2845] format typo

2004-09-15 20:14  marcella

	* [r2844] GAPW : fix for all_electron and PBC

2004-09-15 17:50  tlaino

	* [r2843] More changes on qm/mm... all results on qm/mm in regtest
	  will be different!!!
	  that's ok... teo

2004-09-15 16:42  marcella

	* [r2841] GAPW : fix for PBC

2004-09-14 15:25  vondele

	* [r2839] check stat with failure_level

2004-09-14 15:15  vondele

	* [r2838] cp_fm_create: check stat

2004-09-14 14:48  marcella

	* [r2837] GAPW and TDDFT: hartree complete

2004-09-14 13:53  marcella

	* [r2836] ops

2004-09-14 12:48  marcella

	* [r2835] GAPW and TDDFT

2004-09-14 07:59  marcella

	* [r2834] GAPW : cutoff and tddft

2004-09-14 07:45  vondele

	* [r2833] free extended type stuff for md runs

2004-09-13 20:50  vondele

	* [r2832] keep virial_create in force_env_create only, fix init of
	  pv_scalar in md_energies

2004-09-12 15:32  vondele

	* [r2831] write format fix

2004-09-12 15:24  vondele

	* [r2830] innocent input typo fix

2004-09-12 09:21  vondele

	* [r2829] format fix

2004-09-10 19:33  marcella

	* [r2828] GAPW : timing

2004-09-10 19:31  marcella

	* [r2827] GAPW: again about cutoff-radii

2004-09-10 11:30  marcella

	* [r2826] GAPW : work-around for problems with the cutoff-radii

2004-09-10 07:52  tchassai

	* [r2825] Moved the GAPW variables to p_env as they fit better
	  there.

2004-09-09 21:18  mcgrath

	* [r2824] trying to make things a little more autonomous

2004-09-09 19:27  ikuo

	* [r2823] Here you go Matt...

2004-09-09 14:56  vondele

	* [r2822] OMP problem fix

2004-09-09 14:48  tchassai

	* [r2821] Little change for easier debugging.

2004-09-09 09:00  tchassai

	* [r2820] Bugfix: add variable initialization

2004-09-08 15:48  marcella

	* [r2818] GAPW : fix for LSD

2004-09-08 15:29  vondele

	* [r2817] bug fix

2004-09-08 15:08  tchassai

	* [r2816] Reverting a premature change.

2004-09-08 15:04  tchassai

	* [r2815] Update subroutine calls for GAPW TDDFT.

2004-09-08 15:02  tchassai

	* [r2814] Prepare init_rho0 for GAPW TDDFT

2004-09-08 14:56  tchassai

	* [r2813] Prepare rho0_s_grid_create for GAPW TDDFT.

2004-09-08 14:52  tchassai

	* [r2812] Changed the prepare_gapw_den call so it works with future
	  versions of the
	  subroutine.

2004-09-08 11:19  marcella

	* [r2811] GAPW: use new xc routines to calculate
	  the local (atomic) contributions to the xc potential

2004-09-07 21:42  fawzi

	* [r2810] * using same names to call new functionals (so that for
	  example LDA works).
	  * fixing test routine for non GGA
	  * fixing intent in old routines to test them against new using
	  nag

2004-09-07 14:29  fawzi

	* [r2809] added a note about the existence of the maple worksheet
	  to generate the code.

2004-09-07 14:21  fawzi

	* [r2806] added pbe functional (lda 3. deriv /lsd 2. deriv)

2004-09-07 14:16  fawzi

	* [r2805] correcting test routine.

2004-09-06 15:43  vondele

	* [r2804] fix build (ionode -> logical)

2004-09-06 12:28  tchassai

	* [r2803] Add initializations for GAPW linear response densities.

2004-09-06 12:28  tchassai

	* [r2802] Cleanup and add fields for GAPW linear response
	  densities.

2004-09-06 12:26  tchassai

	* [r2801] Remove unused variable.

2004-09-06 12:26  tchassai

	* [r2800] Some cleanup of old code.

2004-09-06 12:25  tchassai

	* [r2799] Add possibility to switch off the multipole moment of the
	  nucleus for TDDFT

2004-09-06 12:23  tchassai

	* [r2798] Refactoring for GAPW TDDFT.

2004-09-02 13:00  marcella

	* [r2797] GAPW: fixing for isolated projectors

2004-09-02 11:25  marcella

	* [r2796] GAPW : all-electrons forces

2004-09-02 09:24  vondele

	* [r2795] fix build

2004-09-02 08:33  marcella

	* [r2794] GAPW : forces on ions

2004-09-01 11:05  vondele

	* [r2793] added error check

2004-08-29 14:54  vondele

	* [r2792] add last_used_block_node heuristic to speed up
	  find/add_real_block_node ten times.

2004-08-29 12:47  vondele

	* [r2791] ! further assert reduction

2004-08-28 19:59  vondele

	* [r2790] not quite useless...

2004-08-28 19:48  vondele

	* [r2789] save a few asserts and useless init

2004-08-28 17:30  vondele

	* [r2788] speed up cp_sm_mix

2004-08-28 17:06  vondele

	* [r2787] added comment about quadratic scaling

2004-08-28 16:47  vondele

	* [r2786] avoid using deltap if not needed

2004-08-28 15:18  vondele

	* [r2785] speed up transfer_real_matrix

2004-08-28 15:12  vondele

	* [r2784] p -> matrix_p

2004-08-28 12:00  vondele

	* [r2783] remove unused USE

2004-08-27 12:32  vondele

	* [r2782] fix undefined qmmm_gaussian_fns

2004-08-26 21:55  marcella

	* [r2779] GAPW: density gradient correction ok

2004-08-25 15:17  vondele

	* [r2778] fix intent(inOUT)

2004-08-25 12:35  tchassai

	* [r2777] qs_ks_build_kohn_sham_matrix: refactoring a bit to make
	  more readable

2004-08-24 13:35  tlaino

	* [r2776] Cleaning qm/mm (only!) memory leaks.. ;)

2004-08-24 10:44  vondele

	* [r2775] change name of standardize_dervitative_desc

2004-08-24 07:36  vondele

	* [r2773] full functionality now

2004-08-23 20:20  vondele

	* [r2772] Fix cp2k_units leak.

2004-08-23 07:33  tlaino

	* [r2770] MaxErr on debug derivatives less tight. This need to be
	  fixed later on..

2004-08-20 17:19  vondele

	* [r2768] unused USE

2004-08-20 11:30  vondele

	* [r2767] more leaks removed

2004-08-20 11:05  vondele

	* [r2766] further deallocates

2004-08-20 10:52  vondele

	* [r2765] further deallocates

2004-08-20 09:24  vondele

	* [r2764] deallocate virial in virial_release

2004-08-20 09:13  marcella

	* [r2763] GAPW : derivatives of the sherical harmonics wrt theta
	  and phi

2004-08-20 09:03  vondele

	* [r2762] free at a better place

2004-08-20 08:56  vondele

	* [r2761] fix memory leak

2004-08-20 08:52  vondele

	* [r2760] fix leak

2004-08-20 08:25  vondele

	* [r2759] fixing memory leaks

2004-08-20 00:22  cjmundy

	* [r2758] OMP fixes

2004-08-19 20:39  vondele

	* [r2757] further target stuff

2004-08-19 20:03  vondele

	* [r2756] further TARGET related stuff

2004-08-19 17:30  vondele

	* [r2755] further TARGET mess

2004-08-19 17:24  vondele

	* [r2754] more target mess: qs_env made a pointer in get_qs_env ...

2004-08-19 16:31  vondele

	* [r2753] TARGET mess fix 1

2004-08-19 16:30  vondele

	* [r2752] comment typo

2004-08-19 09:42  tlaino

	* [r2750] Bug Fix for non gaussian QM/MM coupling..

2004-08-18 20:03  vondele

	* [r2749] serious bug fix ;-) continue after doing nothing

2004-08-18 16:52  fawzi

	* [r2748] adding enumeration and cp2k input

2004-08-18 16:52  fawzi

	* [r2747] adding commented out test on new parser

2004-08-18 16:51  fawzi

	* [r2746] making new input work, adding enumeration, and cp2k input

2004-08-18 16:48  fawzi

	* [r2745] fixing bug in to_array

2004-08-18 08:28  jgh

	* [r2744] Update of semi-empirical code (input only at the moment)

2004-08-18 07:51  marcella

	* [r2742] GAPW : some more fixing

2004-08-17 07:13  tchassai

	* [r2741] - 2nd derivatives of GGA's now work.

2004-08-15 14:48  mcgrath

	* [r2740] possibly a reversed negative sign

2004-08-14 08:45  krack

	* [r2739] temperature control again enabled

2004-08-13 17:12  mcgrath

	* [r2738] trying again...

2004-08-13 17:08  mcgrath

	* [r2737] writing out more coordinates for diagnostics

2004-08-13 16:43  mcgrath

	* [r2736] diagnostic stuff

2004-08-13 16:17  mcgrath

	* [r2735] forgot something

2004-08-13 15:46  mcgrath

	* [r2734] Can't build locally, so I may have to do a couple of
	  these...

2004-08-13 15:44  mcgrath

	* [r2733] jumble random numbers to prevent problems in short runs

2004-08-12 18:04  tlaino

	* [r2732] 1) Switching to Spline interpolator (Default) (N.B.: This
	  will change the result of the
	  H2O-qmmm-gauss.inp in QS/regtest.. sorry)
	  2) Bug Fix in GAussian LongRange Part
	  3) Optimizing the collocation of the LongRange Part.

2004-08-12 07:43  tlaino

	* [r2731] RealSpace Spline Interpolator to evaluate QM/MM
	  derivatives. tlaino.

2004-08-11 21:51  marcella

	* [r2729] GAPW : analytic integrals for the local, one center,
	  hartree terms

2004-08-11 15:21  fawzi

	* [r2728] * removed methods with more than 2 readstreams
	  * fixed pw_project_sp3 (still wrong boundary)

2004-08-11 13:33  fawzi

	* [r2727] moving declaration of n before using it in dimension(n).

2004-08-11 12:48  fawzi

	* [r2726] adding forgotten use

2004-08-11 11:55  jgh

	* [r2725] Analytic integrals for Coulomb potential

2004-08-11 10:00  fawzi

	* [r2724] call to new spline interpolator.

2004-08-11 09:58  fawzi

	* [r2723] spline interpolation (with wrong border).

2004-08-11 09:55  fawzi

	* [r2722] removing duplicate duplicate ;)

2004-08-11 09:52  fawzi

	* [r2721] adding scale to sumup

2004-08-11 09:51  fawzi

	* [r2720] adding first (non parallel) version of pw_grid_duplicate

2004-08-10 00:00  ikuo

	* [r2718] Cleanup for DEC

2004-08-09 21:54  marcella

	* [r2717] GAPW : all electron potential

2004-08-09 15:14  vondele

	* [r2715] further updating and cleaning for OT and occupation

2004-08-09 13:10  tlaino

	* [r2713] Bug fix for intel ifort compiler. teo.

2004-08-09 08:55  tlaino

	* [r2712] Bug Fix for Ar-qmmm.inp in regtest. teo.

2004-08-08 17:26  jgh

	* [r2711] Allow for more defaults in the input for potential and
	  basis, plus some
	  trivial changes

2004-08-07 14:42  tlaino

	* [r2709] QM/MM: Fixed Derivatives. Analytical QM/MM derivatives
	  are equal to the numericals
	  both using GRID or GAUSS option. Changed the two tests on QM/MM
	  into the regtest
	  directory. Some cosmetics around. teo.

2004-08-06 15:09  vondele

	* [r2708] Initial commit for OT with non-equivalent occupied orbs.
	  Still experimental

2004-08-05 21:35  marcella

	* [r2707] GAPW : fix

2004-08-05 07:29  vondele

	* [r2705] bug fix

2004-08-04 19:23  marcella

	* [r2704] GAPW : some error fixing

2004-08-04 14:55  vondele

	* [r2703] fix

2004-08-04 13:14  vondele

	* [r2702] write molecular dipoles

2004-08-03 20:46  tlaino

	* [r2701] Deleting the "newline" kind, getting compilation problems
	  with NAG on G5
	  (Fawzi suggestion)

2004-08-03 20:09  tlaino

	* [r2700] More on QM/MM stuff. Using pw_type instead of
	  realspace_grid_type in QM/MM.

2004-08-03 15:06  tlaino

	* [r2699] Introducing the routine pw_grid_duplicate. It is not
	  general and it will
	  need some more work to generalize it further.

2004-08-02 17:53  jgh

	* [r2698] new files for KG

2004-08-02 17:31  jgh

	* [r2697] Intermediate update for KG_GPW method

2004-08-02 11:16  vondele

	* [r2696] remove unused variables

2004-08-01 12:03  vondele

	* [r2695] renamed qs_env variables c,k,s,h in mos,
	  matrix_ks,matrix_s,matrix_h to make finding these in the code
	  easier

2004-07-31 18:33  vondele

	* [r2693] add (unoccupied) MOs to a calculation (use with care for
	  now)

2004-07-30 16:48  vondele

	* [r2692] added ssquare

2004-07-30 16:09  vondele

	* [r2691] minor bug fix

2004-07-29 07:41  vondele

	* [r2690] bug fix : only write on the node that opened the file

2004-07-28 21:43  vondele

	* [r2689] minor stuff (copy only needed part)

2004-07-28 16:19  ikuo

	* [r2688] Remove extra USE statement and missing openmp variable.

2004-07-28 12:48  vondele

	* [r2687] small fix

2004-07-27 15:48  tlaino

	* [r2686] Bug Fix.

2004-07-27 11:55  tlaino

	* [r2685] Introducing Long range QM/MM coupling.

2004-07-27 06:31  tlaino

	* [r2684] Compilation bug fix.

2004-07-27 00:24  fawzi

	* [r2683] spline3 multigrid interpolation.

2004-07-25 16:41  tlaino

	* [r2682] Joost's suggestion to treat the carriage return
	  character... It should work now!

2004-07-25 15:30  tlaino

	* [r2681] Checking bounds of collocation. Don't waste any time if
	  the collocation is out of the QM cube. teo

2004-07-25 14:13  vondele

	* [r2680] added safer_diis as an OT option

2004-07-25 14:12  tlaino

	* [r2679] fixing routine name. teo.

2004-07-25 14:06  tlaino

	* [r2678] fixing timeset and timestop. teo.

2004-07-25 13:41  tlaino

	* [r2677] bug fix. teo.

2004-07-25 13:25  tlaino

	* [r2675] More Stuff on QM/MM. eheheh... ;-)

2004-07-25 13:19  tlaino

	* [r2674] Relating OMP: Ok I checked this routine but it's not my
	  stuff.
	  I guess it should be correct.

2004-07-25 13:17  tlaino

	* [r2673] Introduced keyword to dump the qm/mm electrostatic
	  potential on file with a cube format.
	  Introduced a new input file MM_POTENTIAL to read the gaussian
	  expansion of the
	  MM electrostatic potential. teo.

2004-07-25 13:13  tlaino

	* [r2672] New routines to interpolate functions and densities
	  between multigrds. First
	  Order algorithm without PBC. teo.

2004-07-25 13:12  tlaino

	* [r2671] New routines to collocate and/or integrate gaussians
	  without PBC. teo.

2004-07-23 07:49  vondele

	* [r2670] another OMP fix for intel 7.1

2004-07-22 22:03  ikuo

	* [r2669] Some more OpenMP stuff to get intel 8.0 compiler to work.

2004-07-22 00:02  cjmundy

	* [r2668] Fixing OMP declarations.
	  Teo, please do in src
	  grep "CHECK TEO" *.F
	  to make sure all is okay

2004-07-21 22:56  cjmundy

	* [r2667] bug fix

2004-07-21 17:22  cjmundy

	* [r2666] cleanup

2004-07-21 17:20  cjmundy

	* [r2665] Parrallel Thermostats with Parallel restarts.

2004-07-21 14:45  jgh

	* [r2664] More routines for STO basis set handling

2004-07-20 11:49  vondele

	* [r2662] added an integrator for a simple isokinetic ensemble

2004-07-19 18:54  vondele

	* [r2661] back to a pgf90 compatible state

2004-07-19 18:32  ikuo

	* [r2660] Fix for nve div 0 problem.

2004-07-19 14:06  vondele

	* [r2659] shorten name: restraint_functional_deriv_check

2004-07-18 23:14  ikuo

	* [r2658] blah blah...

2004-07-18 05:53  cjmundy

	* [r2657] Blah

2004-07-18 04:54  cjmundy

	* [r2655] Towards working parallel thermostats with restart. Almost
	  there

2004-07-16 08:08  vondele

	* [r2654] fix : NULLIFY before use

2004-07-16 00:38  cjmundy

	* [r2653] Parallel thermostats work. What a pain.
	  No restart capability. This will be
	  done tomorrow. I think I messed up
	  Joost's formating in md_energies.F
	  ONLY for the E-E0. The rest is
	  fine.

2004-07-15 18:08  ikuo

	* [r2652] Bug Fixes...

2004-07-15 17:06  ikuo

	* [r2651] Bug fix...

2004-07-15 16:52  vondele

	* [r2650] cosmetics for printout

2004-07-15 08:23  vondele

	* [r2649] added numerical routine to check the derivative.

2004-07-12 22:04  mcgrath

	* [r2648] one last commit to make life easy for Will

2004-07-10 15:26  vondele

	* [r2647] added a way to give additional input

2004-07-09 21:17  mcgrath

	* [r2646] making sure everything is up to date before I leave

2004-07-09 19:37  fawzi

	* [r2645] sample interface for random number generator.

2004-07-08 16:06  fawzi

	* [r2642] last bugfix before going home ;)

2004-07-08 15:38  tchassai

	* [r2641] - cleanup of the code
	  - little bugfix for GGA

2004-07-08 15:25  tchassai

	* [r2640] bugfix

2004-07-08 14:15  tchassai

	* [r2639] - Running version (on pgi) of the 2nd derivative with the
	  new functionals.

2004-07-08 12:28  tchassai

	* [r2638] - New functionals for 2nd derivatives

2004-07-08 08:21  tchassai

	* [r2636] - further development towards new functionals

2004-07-08 08:21  tchassai

	* [r2635] - res_etype change

2004-07-08 08:19  tchassai

	* [r2634] - shortened the field "restricted_excitations_type" to
	  res_etype as it
	  was really too long.

2004-07-08 07:25  tchassai

	* [r2633] - Further work to use the new functionals in the 2nd
	  derivative.

2004-07-08 01:15  mcgrath

	* [r2632] had a random number call in the wrong place...amazing how
	  much that can screw things up

2004-07-07 06:52  tchassai

	* [r2631] Fix minor bug.

2004-07-07 01:26  mcgrath

	* [r2630] added a couple checking lines

2004-07-06 23:17  mcgrath

	* [r2629] change the USE statements so it doesn't complain

2004-07-06 23:15  mcgrath

	* [r2628] some bug fixes, some different ways of doing things

2004-07-06 20:59  marcella

	* [r2627] GAPW : local hartree 2-centers integrals

2004-07-06 15:22  tchassai

	* [r2626] - xc_rho_set_types: changed xc_rho_set_update to be able
	  to compute
	  the necessary rho's independently of dft_control%nspins (we need
	  this for singlet and triplet excitations)
	  - qs_kpp1_env_methods.F: Added the computation of the epsilon_xc
	  derivatives. Need to put them together to create the pot now.

2004-07-06 14:01  vondele

	* [r2625] rewrite of the mulliken charge routines, increased
	  efficiency
	  some cleaning
	  added the possibility for restraints

2004-07-06 11:53  vondele

	* [r2624] a few more digits for the output

2004-07-05 15:36  tchassai

	* [r2623] Working towards getting TDDFT to work with GGA's.

2004-07-02 20:27  mcgrath

	* [r2622] whoops...forgot a stop statement

2004-07-02 18:21  mcgrath

	* [r2621] caught a small bug

2004-06-30 16:17  marcella

	* [r2620] GAPW: some fixing

2004-06-30 09:22  marcella

	* [r2619] GAPW : hartree local terms : 3 centers integrals

2004-06-29 16:24  ikuo

	* [r2618] Yet another bug fix...

2004-06-29 13:07  tchassai

	* [r2617] - Fixed a bug in the orthoganalization
	  - Implemented preconditioner for the Davdison method. Doesn't
	  seem to improve
	  much.

2004-06-29 08:55  vondele

	* [r2616] change format for printout

2004-06-29 01:00  ikuo

	* [r2615] Bug Fix...

2004-06-27 09:42  vondele

	* [r2613] change output to avoid very long lines

2004-06-25 14:41  vondele

	* [r2612] fix compilation

2004-06-25 14:35  fawzi

	* [r2611] fixing character length mismatch for pgi

2004-06-25 13:54  fawzi

	* [r2609] * reflective wall
	  * new structure for collective variables: only the derivative
	  with respect
	  * to atomic positions is now computed in colvar.F
	  [alessandro & fawzi]

2004-06-25 13:03  tchassai

	* [r2607] Major rehaul of the code:
	  - Fixed memory leaks
	  - support for Davidson/Lanczos method
	  - now works with or without computing the inverse of S
	  - Some cleanup

2004-06-25 13:01  tchassai

	* [r2606] Added new keys to the tddfpt control
	  - diag_method : davidson|lanczos
	  - invert_s : compute the inverse of the overlap matrix to
	  get faster convergence in tddfpt

2004-06-25 11:14  tlaino

	* [r2605] 1) Some cleaning and reorganization of depencies
	  2) Preliminary version of a new QM/MM interface: Collocating
	  Gaussian Functions Fitting Coulomb Potential. Energy and
	  derivatives. Still to debug.

2004-06-25 10:36  fawzi

	* [r2604] adding linked lists, and input types

2004-06-25 10:35  fawzi

	* [r2603] adding new input types.

2004-06-25 10:34  fawzi

	* [r2602] adding MPI_COMM_SELF, cosmetics (unused vars)

2004-06-25 10:33  fawzi

	* [r2601] removing write_cputime to diminish the dependencies

2004-06-25 10:32  fawzi

	* [r2600] removing unused var

2004-06-25 10:32  fawzi

	* [r2599] adding parameter to variables

2004-06-25 10:31  fawzi

	* [r2598] removed obsolete types from set_qs_env

2004-06-25 10:29  fawzi

	* [r2597] fixed leak, bugfix (influence_function)

2004-06-25 10:27  fawzi

	* [r2596] moved print_keys here from global_types to diminish the
	  dependencies.

2004-06-25 10:26  fawzi

	* [r2595] moved print_keys to print_key_types, added
	  cp_para_env_from_globenv, prepared to add input_file

2004-06-25 10:23  fawzi

	* [r2594] removed unused variables

2004-06-25 10:23  fawzi

	* [r2593] new parser based on qs_parser, but without global
	  variables, and able to fail without stopping the program.

2004-06-25 10:21  fawzi

	* [r2592] diminished dependency, moved cp_para_env_from_globenv to
	  global_types

2004-06-25 10:19  fawzi

	* [r2591] diminished dependency of cp_error_handling,
	  cp_synchronize_error

2004-06-25 10:17  fawzi

	* [r2590] dimished dependency of logger, added open_file,
	  close_file

2004-06-25 10:15  fawzi

	* [r2589] * removed %initialized
	  * generic cp_to_array
	  * some cleaning
	  * added char,logical,real and val lists
	  * added *_p_type

2004-06-25 10:11  fawzi

	* [r2588] removed some unused vars, synopsys, not yet clean :(

2004-06-25 10:09  fawzi

	* [r2587] cleaning away unused vars.

2004-06-25 10:05  fawzi

	* [r2586] moved write cputime to thi module (to diminish the
	  dependency of the timings module)

2004-06-24 16:17  gtb

	* [r2584] Rewritten to use the new exclusion lists, now works also
	  for rigid molecules
	  (only simple kg!)

2004-06-24 15:08  jgh

	* [r2581] STO-nG coefficients, for later use

2004-06-24 10:28  vondele

	* [r2578] bug fix (qmmm) : only ask components needed (II)

2004-06-24 09:54  vondele

	* [r2577] bug fix (qmmm) : only ask the components needed

2004-06-24 09:03  vondele

	* [r2576] INTENT(INOUT) fix

2004-06-23 16:36  cjmundy

	* [r2574] PME now works in parallel.

2004-06-23 14:36  mcgrath

	* [r2573] things should be working in serial with Intel 8.0, but
	  not parallel yet

2004-06-22 22:42  mcgrath

	* [r2572] another bug...not the one I'm looking for, though

2004-06-22 21:44  ikuo

	* [r2571] Bug fixes...

2004-06-22 21:06  cjmundy

	* [r2569] Juergs's final bug fix for parallel SPME

2004-06-22 20:38  mcgrath

	* [r2568] heh...forgot to remove a write statement

2004-06-22 18:50  mcgrath

	* [r2567] minor bug fix...doesn't really affect anything, but it
	  makes me feel better

2004-06-22 17:27  cjmundy

	* [r2566] Redundant declaration

2004-06-22 17:08  cjmundy

	* [r2565] Bug fix

2004-06-22 14:27  krack

	* [r2563] fixed atoms constraint for MD

2004-06-22 10:07  tlaino

	* [r2561] QM/MM Interface with Fist Completed! Now charges and core
	  radius are set in the proper way.
	  Splitted and rewritten with the Joost fashion the qmmm_input
	  section..
	  Modified the QM/MM tests in QS/regtest:
	  1) Ar-qmmm.inp : 2 Ar atoms around an hundread of Ar MM atoms
	  (JUST MECHANICAL COUPLING)
	  2) H2O-qmmm.inp: 1 QM water molecule + 2 MM water molecule
	  (ELECTROSTATIC COUPLING)
	  Some cleaning around. (teo)

2004-06-22 04:01  cjmundy

	* [r2560] bug fix

2004-06-22 00:36  ikuo

	* [r2559] Ready to try out alpha helix with Fist...

2004-06-21 20:44  cjmundy

	* [r2557] SPME/PME works in parallel

2004-06-21 04:28  cjmundy

	* [r2556] Juerg's bug fix for parallel spme/pme.
	  However, still needs some work. The
	  forces are not being added up correctly.
	  rattle_3x3 was buggy. Now fixed

2004-06-20 17:32  krack

	* [r2555] fix atom constraint for GEO_OPT and MD

2004-06-19 17:27  cjmundy

	* [r2553] spme works again. Fawzi, make sure
	  what I did was okay.

2004-06-19 06:01  ikuo

	* [r2552] Last set of utils needed for force field...

2004-06-19 04:24  cjmundy

	* [r2551] Slow Ewald works in parallel.
	  neighbor lists updates for FIST
	  only printed when print_level > medium.
	  Cleaned up parallel output.

2004-06-18 16:20  tlaino

	* [r2550] Bug fix for NAG compiler. water.inp now works on NAG!

2004-06-18 12:39  tlaino

	* [r2549] Fixing bug! Argon108 didn't work anymore..

2004-06-18 09:28  fawzi

	* [r2548] restoring old versions of files I wasn't supposed to
	  check in :(

2004-06-18 09:20  fawzi

	* [r2547] using the correct bound in parallel

2004-06-18 09:15  fawzi

	* [r2546] cosmetics, setting the ks matrix name correctly.

2004-06-17 23:53  ikuo

	* [r2545] Cleaning up of some outputs.

2004-06-17 22:22  cjmundy

	* [r2544] Fist works in parallel now.
	  Need to test more on both
	  ewald routines (SPME and EWALD).
	  Gotta go now.

2004-06-17 19:18  marcella

	* [r2543] GAPW: some more integrals of rho0_soft on the global grid

2004-06-16 22:42  cjmundy

	* [r2542] Really, really stupid bug. Now slow ewald works

2004-06-16 17:47  tlaino

	* [r2541] Preliminary Tuckerman (to debug!) and some fawzi's cell
	  cosmetics

2004-06-16 17:18  ikuo

	* [r2540] Nullify(elp_potential)

2004-06-16 16:47  cjmundy

	* [r2539] distributed particles works

2004-06-15 21:33  ikuo

	* [r2534] effective charges for fist to test water example

2004-06-15 19:21  cjmundy

	* [r2532] bug fix. r_last_update is replicated now

2004-06-15 18:24  ikuo

	* [r2531] Bug fix for mass.

2004-06-15 17:38  cjmundy

	* [r2530] debugging exclusion lists. Now works.

2004-06-15 17:20  ikuo

	* [r2529] Commit to get openmp working on ia32 and ia64 with Intel
	  8.0 compiler

2004-06-15 06:32  cjmundy

	* [r2528] Some bug fixes

2004-06-14 18:00  ikuo

	* [r2525] Potparm bug fix...

2004-06-14 16:08  tlaino

	* [r2522] QM/MM coupling Fist to QS.. At the moment just mecanical
	  coupling.. waiting for the mapping of the charges into the
	  topology

2004-06-14 13:25  tlaino

	* [r2521] fawzi's suggestion..

2004-06-14 12:09  tlaino

	* [r2520] Some Bug Fixes for the Argon108 Test. There were
	  uninitialized pointers and unzeroed variables.
	  The problems were discovered under OS/X NAG compiler.

2004-06-14 10:19  marcella

	* [r2519] GAPW: rho0 soft on the grid

2004-06-12 19:08  cjmundy

	* [r2517] Intra routines in, bonds, bends, and torsions. Need
	  to make sure all is mapped in topology.

2004-06-12 00:20  cjmundy

	* [r2516] Better Nose Defaults

2004-06-10 23:12  cjmundy

	* [r2513] Argon 108 works!

2004-06-10 17:56  ikuo

	* [r2512] Bug fix for FIST mass problem.
	  No effect on QS because mass for QS readin at read_atomic_kind

2004-06-10 14:33  ikuo

	* [r2511] Bug Fix...

2004-06-10 00:45  cjmundy

	* [r2510] closer

2004-06-09 21:17  ikuo

	* [r2509] deallocate exclusion list...

2004-06-09 18:45  cjmundy

	* [r2508] One step closer to Bernie Alders dream.

2004-06-09 18:34  ikuo

	* [r2507] Exclusion list to work for water...

2004-06-09 17:43  mcgrath

	* [r2506] MC certified leak free by HMD

2004-06-09 16:23  ikuo

	* [r2505] Exculsion list added.

2004-06-09 15:07  fawzi

	* [r2504] reactivated velocity rescale (nve + temp_tol xx in the md
	  section)

2004-06-09 01:44  mcgrath

	* [r2503] the results of a day spend debugging...and I'm not done
	  yet

2004-06-08 17:57  cjmundy

	* [r2501] One step closer to 1968:

2004-06-08 16:48  ikuo

	* [r2500] Bug Fix.

2004-06-08 11:37  vondele

	* [r2499] cosmetic typo

2004-06-08 02:13  mcgrath

	* [r2498] had some more array indices screwed up

2004-06-07 23:24  cjmundy

	* [r2496] Ewald "NONE" now an option

2004-06-07 20:17  mcgrath

	* [r2494] made mc_par elements private

2004-06-07 20:07  cjmundy

	* [r2493] Fawzi and Teos bug fixes... Thanks guys and sorry
	  for the stupid errors.

2004-06-07 16:29  hforbert

	* [r2491] Fixed (hopefully) deallocating force_env_p in qs_mol_dyn1
	  twice.
	  (gets already deallocated by md_env_release...)

2004-06-07 14:39  vondele

	* [r2490] bug fix, use defined constant

2004-06-06 21:22  mcgrath

	* [r2489] cleaning up a bunch of things...most notably, I'm in the
	  proccess of making mc_par private
	  and use get_ and set_ routines to access it...I've also stopped
	  accessing cell
	  directly

2004-06-05 21:31  mcgrath

	* [r2487] fixed some out of bounds bugs

2004-06-05 20:40  mcgrath

	* [r2486] inserted a few more DEALLOCATES to fix a leak in the MC
	  keyword

2004-06-04 21:30  marcella

	* [r2485] GAPW : local Eh: one center contributions, and 1 center
	  terms in the KS H

2004-06-04 13:33  tlaino

	* [r2484] Introduced the nuclear QM/MM contribution to the energy.
	  Some clean-up...

2004-06-04 11:25  tlaino

	* [r2483] Introducing the forces into the QM/MM. Just for the
	  potential computed on the
	  GRID.

2004-06-04 11:06  fawzi

	* [r2482] fixing bug wrt. mass of kinds.

2004-06-03 16:53  tlaino

	* [r2481] Cleaning some structure for evaluation of energy and
	  forces in QM/MM.

2004-06-03 16:11  marcella

	* [r2480] gapw : hartree local, one center integrals

2004-06-03 00:01  ikuo

	* [r2479] Cleanup for DEC.

2004-06-02 19:06  ikuo

	* [r2478] MD-virial patch.

2004-06-02 17:57  tlaino

	* [r2477] Changes:
	  1) Cube -> coherent with coordinates (origin of the cube file in
	  0,0,0)
	  2) Clean-up and fixing bugs in QM/MM

2004-06-02 16:08  jgh

	* [r2476] initialize last_atom in molecule type

2004-06-02 16:05  jgh

	* [r2475] cosmetics

2004-06-02 16:05  jgh

	* [r2474] added Last_atom to the molecule type

2004-06-02 14:59  hforbert

	* [r2472] Merge with PIMD branch (classical code hopefully working
	  now) (path integrals
	  themselves will have to wait a bit)
	  There were a couple of files removed, and even more files are not
	  part of the
	  PIMD branch and don't seem to be in use anywhere atm, I left them
	  in, but
	  we should review these.

2004-06-02 12:13  marcella

	* [r2468] gapw : hartree local, one center integrals

2004-06-02 09:45  marcella

	* [r2467] gapw : hartree local, one center integrals

2004-06-02 09:22  marcella

	* [r2466] gapw : hartree local, one center integrals

2004-06-01 15:10  tlaino

	* [r2464] few more modification on qmmm:
	  1) translation of the system into the center of the qm cell
	  2) deleted qmmm_coulomb_potential.F -> inserted into
	  qmmm_electrostatic_methods.F

2004-06-01 00:57  mcgrath

	* [r2463] Heh...found a small bug already

2004-05-31 22:53  mcgrath

	* [r2462] some more MC stuff...I split a file into two in order to
	  distinguish between subroutines a little better

2004-05-30 16:36  jgh

	* [r2461] Correct bug in set_qs_env introduced with qmmm

2004-05-28 22:08  mcgrath

	* [r2459] a new topology keyword, specific for MC...and then some
	  cleaning up

2004-05-28 17:37  fawzi

	* [r2455] * lsd version of tpss
	  * ldaCore on rho and ndrho

2004-05-28 17:36  fawzi

	* [r2454] modifications to test ldaCore (rho or ndrho)

2004-05-28 14:46  vondele

	* [r2453] converted &QS input section to the augmented format.

2004-05-28 11:25  vondele

	* [r2450] Added a searchable list of keywords for a section,
	  introducing a short description.

2004-05-28 00:25  mcgrath

	* [r2449] a couple of minor bug fixes

2004-05-27 16:39  tlaino

	* [r2447] Bug fix on the QM/MM potential expression.

2004-05-27 09:22  tlaino

	* [r2445] Fixing an unitialized vector. tlaino.

2004-05-26 21:33  mcgrath

	* [r2444] bug fix...changed a 1 to an i

2004-05-26 21:22  tlaino

	* [r2443] Added more features on QM/MM :
	  -) ks_qmmm_env
	  -) implemented the analytic electrostatic coupling on the grid
	  -) now you can do an SCF+QM/MM : just energy

2004-05-26 18:40  tlaino

	* [r2441] Fix a bug for occupancy and beta.

2004-05-24 13:50  krack

	* [r2435] type components torsion_list and ntorsion are now
	  considered

2004-05-24 12:47  vondele

	* [r2434] dealloc globenv at the end.

2004-05-24 09:36  krack

	* [r2433] QS memory leaks removed (missing release of subsys).
	  Explicit deallocate of ftable enabled.

2004-05-23 20:21  vondele

	* [r2426] this year also

2004-05-23 16:44  krack

	* [r2425] Do not call init_orbital_pointers with negative l values
	  any longer.

2004-05-23 09:58  tlaino

	* [r2424] Bug fix for qmmm_small_cell on IFC.

2004-05-22 13:42  jgh

	* [r2423] Correct typo and change comments

2004-05-21 16:57  tlaino

	* [r2421] Enabling two different cell in QM/MM calculations.

2004-05-21 08:46  krack

	* [r2420] Allow for clean up of the orbital pointers and
	  transformation matrices
	  in order to find the real memory leaks

2004-05-20 16:07  tlaino

	* [r2418] topology_pdb.F : bug fix in PDB format. TER keyword
	  Added!
	  qmmm_init.F : more work to do on qmmm input. Added a Namelist
	  &QMMM &END
	  not properly working yet!

2004-05-19 18:31  fawzi

	* [r2407] * cutoff on tau
	  * better debugging for functional_routine DEBUG

2004-05-19 18:29  fawzi

	* [r2406] fixed numerical problem in TPSS

2004-05-19 01:42  mcgrath

	* [r2404] made cell_ref cleaner...actual bug fix

2004-05-18 21:26  mcgrath

	* [r2403] an ugly fix to get H2O_tddft.inp working...looking into
	  making things nicer

2004-05-18 20:20  mcgrath

	* [r2402] some more MC stuff...bug fixes, mostly

2004-05-18 19:49  mcgrath

	* [r2401] testing to see if I can commit

2004-05-14 09:05  vondele

	* [r2390] remove useless printing

2004-05-13 21:00  marcella

	* [r2389] gapw: construction of multipoles and compensation charge
	  densities

2004-05-13 20:57  marcella

	* [r2388] gapw : construction of multipoles and compensation charge
	  densities

2004-05-12 15:18  fawzi

	* [r2383] making deallocation of coeff more clean

2004-05-12 13:53  fawzi

	* [r2381] fixing bug with metadynamics and farming

2004-05-11 08:17  fawzi

	* [r2375] while waiting for the classical code... introducing the
	  fantatic QM/QM method ;)

2004-05-11 08:05  fawzi

	* [r2374] new version of tpss functional

2004-05-11 08:04  fawzi

	* [r2373] release cell instead of dealloc

2004-05-11 08:01  fawzi

	* [r2372] * better handling of files
	  * preparing for parallel runs

2004-05-07 17:18  vondele

	* [r2366] correct output of commensurate_mgrids

2004-05-06 10:55  vondele

	* [r2363] init some variables

2004-05-06 10:22  vondele

	* [r2362] only copy needed part

2004-05-06 09:21  vondele

	* [r2361] bug fixes : avoid undefined variables

2004-05-05 22:00  ikuo

	* [r2360] Reorganize topology_parameter_type so its easier to put
	  in force field next.

2004-05-04 19:13  mcgrath

	* [r2359] made some changes Fawzi suggested...looks like it's
	  working

2004-05-04 16:14  vondele

	* [r2358] minor cleaning

2004-05-04 14:03  fawzi

	* [r2357] fixing bug for grad_deriv < 1

2004-05-04 13:22  fawzi

	* [r2356] fixing bug in xc_exc_calc

2004-05-04 12:49  fawzi

	* [r2355] qs initialization with a different subsys.

2004-05-04 11:59  fawzi

	* [r2354] xc update
	  * activated converted functionals (untested)
	  * metafunctionals (not yet really working)

2004-05-04 11:51  fawzi

	* [r2353] nag mac with optimizations fixes.

2004-05-04 11:48  fawzi

	* [r2352] xlf on mac with optimizations needs this.

2004-05-04 00:01  ikuo

	* [r2351] Place holder for routine to read force field

2004-05-03 22:06  ikuo

	* [r2350] More non debuged force field stuff.

2004-05-03 21:47  mcgrath

	* [r2349] quick fix to a problem noted by Fawzi and Joost...I'll
	  keep working on it

2004-05-03 09:57  marcella

	* [r2348] gapw: initialization of paw-atom

2004-05-02 19:27  mcgrath

	* [r2347] hopefully fixed a memory problem with cell_ref

2004-05-01 22:57  mcgrath

	* [r2346] some bug fixes...and a new type of cell for an array of
	  cell types

2004-04-30 14:25  fawzi

	* [r2344] higher max_cache for lsd

2004-04-30 14:16  fawzi

	* [r2343] the complex3d vs real3d bug was more done consistently
	  also in create_cr3d :(.

2004-04-29 18:38  mcgrath

	* [r2342] added a reference cell, so we can use a cell of a certain
	  size with the number of grid points of a different sized cell

2004-04-29 18:27  mcgrath

	* [r2341] bug fixes? yup

2004-04-29 17:27  ikuo

	* [r2340] 1.Change syntax for constraints.
	  
	  2.Can enter intra molecular force field parameters in the input
	  file.

2004-04-29 10:06  fawzi

	* [r2338] coordination number collective variable, smaller fixes.

2004-04-28 20:05  marcella

	* [r2337] gapw

2004-04-28 17:42  marcella

	* [r2336] gapw

2004-04-28 15:45  vondele

	* [r2335] fixes needed to compile on SUN / NAG

2004-04-28 14:39  marcella

	* [r2334] gapw

2004-04-28 12:34  marcella

	* [r2333] *** empty log message ***

2004-04-28 11:48  marcella

	* [r2332] gapw: a couple of routines public

2004-04-28 11:45  marcella

	* [r2331] USE force_env_types instead than force_env_methods when
	  required.

2004-04-28 11:42  marcella

	* [r2330] gapw : routines that have been modified for the set up of
	  the gapw variables and algorithms

2004-04-28 11:37  marcella

	* [r2329] gapw: routines for the calculations of the local
	  densities and the XCX contributions
	  only partially tested

2004-04-28 08:42  fawzi

	* [r2328] splitted force_env, adding the forgotten file :(.

2004-04-27 21:32  fawzi

	* [r2327] * first version of metodynamics

2004-04-27 21:24  fawzi

	* [r2326] * removed synopsis
	  * first version of metadynamics
	  * force_env_types splitted in force_env_types and
	  force_env_methods (you will likely want to use the latter one)

2004-04-27 20:48  fawzi

	* [r2324] bond_kind_set was not nullified.

2004-04-27 18:56  ikuo

	* [r2323] Another bug fix...

2004-04-27 15:47  ikuo

	* [r2322] Small bug fix for force field routines...

2004-04-26 21:49  ikuo

	* [r2321] Utils needed for force field...

2004-04-25 19:47  mcgrath

	* [r2317] bug fixes...added a file to hold all types to avoid
	  circular dependencies

2004-04-23 16:09  mcgrath

	* [r2316] added a couple types, fixed some bugs

2004-04-21 22:26  ikuo

	* [r2313] More small changes for constraints

2004-04-21 16:04  cjmundy

	* [r2312] Should work

2004-04-20 22:26  ikuo

	* [r2310] Another small change for constraint

2004-04-20 22:11  ikuo

	* [r2309] Almost 3x3

2004-04-20 18:36  cjmundy

	* [r2308] Constraints are almost there

2004-04-20 13:59  fawzi

	* [r2307] ldaCoreFixes

2004-04-20 12:04  krack

	* [r2306] bug fix for scaled_coordinates (no unit conversion
	  necessary, since the cell
	  vectors are already defined in Bohr)

2004-04-19 18:08  fawzi

	* [r2303] reactivating fd deriv

2004-04-19 15:02  fawzi

	* [r2302] added the new version of all the xc functionals.

2004-04-19 15:01  fawzi

	* [r2301] omp bug fix

2004-04-19 14:21  fawzi

	* [r2300] - first version of ldaCore (still very much in
	  development)
	  - cosmetic changes (kinetic -> tau)
	  - first preparations for energy only xc, and tpss

2004-04-19 14:17  fawzi

	* [r2299] cosmetic changes (auto update synopsis)

2004-04-19 14:10  fawzi

	* [r2298] converted all the functionals to the new interface.

2004-04-16 08:45  vondele

	* [r2297] fix out of bounds (ih) of g-vector not found

2004-04-15 16:20  vondele

	* [r2296] further cleaning

2004-04-15 15:34  vondele

	* [r2295] further cleaning

2004-04-15 13:19  vondele

	* [r2294] clean some of the never dereferenced pointers away

2004-04-15 13:17  fawzi

	* [r2293] added debug routine for xc that outputs x=0 yz plane of
	  vcx,...
	  Changed smooth cutoff making it inconsistent, but nuerically
	  better (but probably still not useful)

2004-04-14 17:52  fawzi

	* [r2292] fixing stupid bug in pw_pool (COMPLEXDATA3D /=
	  REALDATA3D), and making bugtracking of those bugs easier

2004-04-13 22:16  ikuo

	* [r2291] Arghhh.. Can't quite kill these files yet...

2004-04-13 21:50  ikuo

	* [r2290] Replaced with new molecule structures

2004-04-13 21:48  ikuo

	* [r2289] Removed mol_kind_list_types.o mol_struct_list_types.o
	  molecule_types.o

2004-04-13 15:34  vondele

	* [r2287] fix a bug with the plan being incorrectly freed twice.

2004-04-13 08:13  vondele

	* [r2286] bug fix : initialize cell%p3m

2004-04-12 15:52  vondele

	* [r2285] bug fix : initialize

2004-04-12 15:51  vondele

	* [r2284] bug fix : initialize return_t_derivatives

2004-04-09 14:41  fawzi

	* [r2283] adding max_cache.

2004-04-09 14:41  fawzi

	* [r2282] fixing bug in *_get_length

2004-04-09 13:59  fawzi

	* [r2281] removing character pointer because of intel compiler 8.0
	  bug.

2004-04-08 15:42  fawzi

	* [r2280] better cutoffing for lsd.

2004-04-08 14:00  fawzi

	* [r2279] fixing small leack

2004-04-08 11:20  vondele

	* [r2278] (mulliken_charges) : fix memory leak

2004-04-07 15:14  vondele

	* [r2274] save 20% time by multiplying instead of adding for
	  reallocate buffer space

2004-04-07 11:09  fawzi

	* [r2273] fixing nn50_smooth deriv.

2004-04-06 14:30  vondele

	* [r2271] triage of the OMP directives

2004-04-06 03:01  fawzi

	* [r2267] removing duplicate declaration

2004-04-06 03:00  fawzi

	* [r2266] wrote interface routines.

2004-04-06 02:59  fawzi

	* [r2265] fixing comments

2004-04-06 02:58  fawzi

	* [r2264] Commenting out the call to m_abort as I find annying that
	  it always generates a core dump disregarding the shell limits.

2004-04-05 23:32  fawzi

	* [r2263] fixing hcth120 (broken by the too greedy delete of test
	  code).

2004-04-05 19:09  fawzi

	* [r2262] added nn50, cosmetic changes.

2004-04-05 19:08  fawzi

	* [r2261] removing HCTH testing code

2004-04-02 00:00  fawzi

	* [r2260] tries to use better file names for local logs

2004-04-01 23:58  fawzi

	* [r2259] fixed lsd when there is a deriv wrt. norm_drho.

2004-04-01 23:57  fawzi

	* [r2258] don't calculate derivs that arent needed if you ask for 0
	  and 1. deriv.

2004-04-01 23:56  fawzi

	* [r2257] addend function to set the filenames (so farming tries to
	  generate more aptly named log files.

2004-04-01 23:54  fawzi

	* [r2256] making farming use the project or output_file input
	  strings to generate output files.

2004-04-01 15:13  vondele

	* [r2253] more explicit S condition number estimate if CHOLESKY_OFF

2004-03-27 04:15  mcgrath

	* [r2244] got all the bugs...maybe? hopefully?

2004-03-24 09:42  fawzi

	* [r2243] added the kinetic labels to the documentation.

2004-03-23 12:23  fawzi

	* [r2242] deactivating jump smoothing by default as it gives more
	  problems than it solves

2004-03-23 10:27  vondele

	* [r2240] update to help ifc and pgf90

2004-03-23 09:58  fawzi

	* [r2239] more consisten cutoffing for lsd.

2004-03-23 09:57  fawzi

	* [r2238] removed obligatory dependency on pw_pool.

2004-03-23 09:54  fawzi

	* [r2237] adding initialization of the diis buffer.

2004-03-23 09:53  fawzi

	* [r2236] fixing omp

2004-03-23 09:53  fawzi

	* [r2235] omp update

2004-03-19 15:47  vondele

	* [r2232] lsd optx

2004-03-19 01:51  ikuo

	* [r2229] Make distinction between residue and molecule

2004-03-18 15:20  fawzi

	* [r2228] making it work with functional=NONE

2004-03-17 14:12  fawzi

	* [r2226] fixing bug in rho_smoothing

2004-03-17 11:09  fawzi

	* [r2225] some compiler voodoo to be fast with all of them...

2004-03-16 14:32  fawzi

	* [r2224] hand inlined version to be faster with xlf (ugly
	  transitory solution until I have again access to other compiler),
	  thanks Joost.

2004-03-15 22:04  mcgrath

	* [r2223] added an IF(globenv%ionode) statement to get farming
	  working

2004-03-14 16:29  vondele

	* [r2219] fix pot INTENT(INOUT) for old routines (works now, but
	  it's too late :-)

2004-03-14 15:53  jgh

	* [r2218] Basic routines for TPSS functional

2004-03-14 12:43  jgh

	* [r2217] Set up QS subsystems for KG theory

2004-03-14 09:58  jgh

	* [r2216] New type atomic_kind_p_type
	  Attach/detach routines

2004-03-14 09:57  jgh

	* [r2215] Correct error checking

2004-03-14 03:48  krack

	* [r2214] Sign for dihedral angle

2004-03-14 00:19  fawzi

	* [r2213] performance update (mostly
	  pw_spine_utils:pw_nn_compose_r), bug fixes

2004-03-14 00:16  fawzi

	* [r2212] fix for nag 4.2 O3

2004-03-13 18:36  krack

	* [r2211] - BFGS optimizer uses now force_env instead of qs_env,
	  but needs further improvement
	  - mathlib updated and cleaned
	  - simulation_cell updated w.r.t. the changes in mathlib
	  - print%angles removed
	  - optional structure data section introduced mainly for the
	  use with GEO_OPT:
	  &STRUCTURE_DATA
	  1 # position vector of atom 1 is printed
	  1 2 # distance vector from atom 1 to 2
	  1 2 3 # angle defined by atom 1 2 3
	  1 2 3 4 # dihedral angle defined by the atoms 1 2 3 4
	  &END

2004-03-13 00:29  mcgrath

	* [r2210] small bug

2004-03-12 13:07  fawzi

	* [r2209] fixing a bug for lsd

2004-03-12 09:55  fawzi

	* [r2208] fixed various bugs, fft free version for smooth xc, LSD
	  GGAS.

2004-03-12 09:53  fawzi

	* [r2207] added the no_cache option.

2004-03-12 09:52  fawzi

	* [r2206] fixed various bugs in the realspace compose, added
	  fd_deriv

2004-03-12 09:51  fawzi

	* [r2205] added nn10_deriv and fd_deriv

2004-03-12 01:45  mcgrath

	* [r2204] whoops...another bug

2004-03-12 00:07  mcgrath

	* [r2203] another bug fix

2004-03-11 23:50  mcgrath

	* [r2202] bug fixes

2004-03-11 12:56  krack

	* [r2201] Allow atom index range for NFIXED to shorten the input

2004-03-11 12:55  krack

	* [r2200] Allow for scaled coordinates

2004-03-10 16:34  vondele

	* [r2199] (deallocate_farming_env) : also dealloc the type itself

2004-03-10 16:29  vondele

	* [r2198] refactor farming a bit, and introduce a input setable
	  partitioning of CPUs (i.e. 128=64+32+32)

2004-03-10 16:07  fawzi

	* [r2196] do the right check to decide if the functional has no
	  info.

2004-03-10 14:48  fawzi

	* [r2195] moved initialization of performance conters before call
	  to mp_sum.

2004-03-10 14:45  fawzi

	* [r2194] cosmetic changes (auto synopsys generation)

2004-03-10 14:38  fawzi

	* [r2193] added m_loc_r and m_loc_c to get the adderss of real and
	  complex arrays (to keep track of them. (actually works only on
	  aix, and nag, placeholder on other platforms, if you need it
	  modify them).

2004-03-10 11:28  vondele

	* [r2191] remove unused label (1)

2004-03-10 09:02  krack

	* [r2190] Process list printing recovered (only ionode is printing
	  now)

2004-03-08 09:13  krack

	* [r2184] Bug fix for upper loop limit

2004-03-07 20:48  vondele

	* [r2183] bug fix : init rat

2004-03-07 16:39  vondele

	* [r2182] added optx exchange functional

2004-03-05 19:59  vondele

	* [r2178] add walltime: xxx to &MD. Stop MD after approx walltime
	  seconds

2004-03-05 15:53  vondele

	* [r2177] bug fix:(mp_world_finalize) use mpi call directly

2004-03-05 08:30  vondele

	* [r2175] update output handling for farming

2004-03-05 07:47  fawzi

	* [r2174] don't request unit_nr on non ionode

2004-03-04 22:49  fawzi

	* [r2173] don't use globenv%src as global unit on non ionode.

2004-03-04 22:47  fawzi

	* [r2172] don't accept unit 0 as valid.

2004-03-04 20:58  vondele

	* [r2171] i,j,k=1,2,3

2004-03-04 15:15  fawzi

	* [r2168] removed unused var

2004-03-04 15:14  fawzi

	* [r2167] cosmetic change (auto update synopsis)

2004-03-04 15:11  fawzi

	* [r2166] cosmetic changes (auto clean)

2004-03-04 15:06  fawzi

	* [r2165] added hcth functional

2004-03-04 15:04  fawzi

	* [r2163] minor numerical changes

2004-03-04 15:03  fawzi

	* [r2162] use pw_pool in release

2004-03-04 15:01  fawzi

	* [r2161] added hcth functional (LDA only)

2004-03-04 14:59  fawzi

	* [r2160] added smooth cutoff

2004-03-04 12:08  vondele

	* [r2159] only get default unit nr if ionode

2004-03-03 16:56  ikuo

	* [r2158] Routine to dump out the PDBs

2004-03-01 17:32  gtb

	* [r2154] added some logic (aux_basis_set, orb_basis_set might not
	  be associated for some atomic kinds)

2004-03-01 17:30  gtb

	* [r2153] minor bug fix in ELP potential type

2004-03-01 17:28  gtb

	* [r2152] small fix in normalization

2004-03-01 17:26  gtb

	* [r2151] bug fix for ELP potential type

2004-03-01 09:57  jgh

	* [r2150] Updates to KGmodels (setup phase)

2004-03-01 09:56  jgh

	* [r2149] Right adjusted (80) output

2004-03-01 07:56  vondele

	* [r2148] remove duplicate definition of comm

2004-02-29 16:13  jgh

	* [r2147] More changes for switch from kg_control_type to
	  qs_control_type

2004-02-29 16:07  jgh

	* [r2146] delete kg_control_type, use qs_control_type also for
	  KGmodels

2004-02-29 14:59  jgh

	* [r2145] Removed Robodoc SOURCE statement, some lines rearranged

2004-02-29 14:23  jgh

	* [r2144] Header for KGmodel
	  some error checking in kg_main

2004-02-29 13:54  jgh

	* [r2143] Flexible scheme for normalization of basis functions
	  (not yet a perfect solution)

2004-02-27 17:45  ikuo

	* [r2142] Update to readin a properly formated PDB file

2004-02-26 11:04  fawzi

	* [r2129] correcting cutoff handling

2004-02-26 08:46  vondele

	* [r2126] fix : the unit number can be different for all tasks, in
	  particular negative (not-associated) for non-ionode tasks. Hence
	  no communication within the IF (info<=0) part, as not all tasks
	  will be there.

2004-02-26 08:44  vondele

	* [r2125] further fine tuning, add to local log instead of
	  overwrite

2004-02-26 08:40  vondele

	* [r2124] write for all nodes

2004-02-25 20:30  fawzi

	* [r2123] xc fixes, lsd test, performance update: getting there

2004-02-25 20:28  fawzi

	* [r2122] fixing intent.

2004-02-25 20:28  fawzi

	* [r2121] using m_abort instead of abort (the machine module is
	  there to insulate from compiler specific functions)

2004-02-25 20:15  fawzi

	* [r2120] added flad to test lsd functional against lda

2004-02-23 23:29  ikuo

	* [r2119] Remove extra return in the title part for gaussian cube
	  files

2004-02-23 20:12  vondele

	* [r2118] (read_md_restart): do not write to globenv%scr if not
	  globenv%ionode

2004-02-23 18:02  vondele

	* [r2117] revert till tested better

2004-02-23 16:25  vondele

	* [r2116] (farming_run) : do not create files for non-ionodes

2004-02-23 15:59  fawzi

	* [r2115] don't initialize global logger upon creation (creates
	  lots of local logger).

2004-02-23 15:43  fawzi

	* [r2114] better handling of global_logger.

2004-02-23 15:35  vondele

	* [r2113] try only to request an output_unit on to ionode

2004-02-23 07:50  tchassai

	* [r2112] * build_core_hamiltonian_matrix: Workaround in the knode
	  loop for SUN ssmp.
	  The < and > operators don't work correctly on arrays in the ssmp
	  version,
	  changed it to a for loop with < and > acting on scalars now.

2004-02-23 02:22  mcgrath

	* [r2111] bug fixes

2004-02-22 14:42  vondele

	* [r2110] real parallel farming arrived !

2004-02-21 19:13  vondele

	* [r2108] improve cp_logger_get_default_unit_nr, and use it

2004-02-21 15:40  vondele

	* [r2107] fix to write to default unit nr

2004-02-21 13:44  vondele

	* [r2106] bug fix : (minor) make clear that gaussian_gridlevel
	  changes the gridlevel_info (keeps a count), and correct
	  assignment fix.

2004-02-21 12:48  vondele

	* [r2105] introduced a stack of timing_envs

2004-02-21 10:07  vondele

	* [r2104] made a stack of message passing performance indicators,
	  removed parallel.F

2004-02-20 18:17  vondele

	* [r2103] avoid double public declaration

2004-02-20 17:43  vondele

	* [r2102] farming changes : have a stack of default loggers, avoid
	  writing directly to unit 6

2004-02-20 14:19  vondele

	* [r2100] needed new files

2004-02-20 14:18  vondele

	* [r2099] first 'draft' farming version. Many things still need
	  improvement, but seems to be a working concept. At least the
	  following things need to be addressed :
	  - timings / message_passing / default_logger
	  - actual farming in parallel

2004-02-20 10:20  vondele

	* [r2098] tentative default subroutine and module headers ...

2004-02-19 20:06  ikuo

	* [r2097] Updated the instantiation and template file to get rid of
	  the duplicate
	  USE statement in cp_linked_list_3d_r

2004-02-19 12:39  fawzi

	* [r2096] * omp fixes
	  * xc fixes (ugly, will change)

2004-02-19 00:44  ikuo

	* [r2095] Initial test version of atom reordering

2004-02-18 21:33  ikuo

	* [r2094] Help DEC compile: Remove duplicate USE.

2004-02-18 15:59  vondele

	* [r2093] add m_chdir to our 'non-standard' tools (not tested on
	  all machines, in particular t3e/absoft might be wrong)

2004-02-17 23:17  fawzi

	* [r2090] * new BLYP implementation
	  * xc fixes
	  * possibility to easily switch between old and new routined from
	  input file

2004-02-17 19:57  vondele

	* [r2089] bug_fix : (init_qs_env) nullify(qs_env%blacs_env)

2004-02-17 19:25  vondele

	* [r2088] 1) generate a blacs_env for a given para_env
	  2) creation of the blacs_env moved to within the creation of a
	  qs_env

2004-02-17 19:22  vondele

	* [r2087] bug fix : (topology_generate_bond) fix memory leak

2004-02-17 18:24  vondele

	* [r2086] bug fix : (calculate_rho_elec, integrate_v_rspace) fix
	  memory leak

2004-02-17 16:16  vondele

	* [r2085]
	  (cp_sm_fm_multiply_general,cp_sm_plus_fm_fm_t_general,cp_fm_block_jacobi_classic)
	  : move away from cp_blacs_pinfo / cp_blacs_get as they are linked
	  to MPI_COMM_WORLD

2004-02-14 17:44  gtb

	* [r2083] 1) introduced the simpar%output_format option in
	  simulation
	  parameters type for specifying the output format of coordinates
	  and velocities in MD runs. In the routine
	  write_particle_coordinate (module particle_types)
	  the output format 'ATOMIC' has been added, in order to enable
	  also printing
	  of coordinates in atomic units. (the default format is XMOL).
	  2) integrator,md_run, force_env_type md_energies has been
	  modified in order to allow
	  also propagation in time of the dynamical_coeffs (i.e. for KG)
	  3) removed old components (related to the old coefficient types)
	  from the md_environment
	  4)thermostats for the dynamical_coefficients have been added
	  (changes in: nhc, extended_system_dynamics extended_system_types,
	  initialize_extended_types)
	  5) In initialize_nhc_forces (module initialize_extended_types),
	  the arguments nhc_part, nhc_baro, nhc_coefs are now optional

2004-02-11 22:50  ikuo

	* [r2082] Added optional error argument.

2004-02-11 22:45  ikuo

	* [r2081] Remove duplicate USE to help DEC

2004-02-11 12:36  vondele

	* [r2080] added test_mpi (MPI X) option to the performance testing

2004-02-10 16:47  krack

	* [r2078] max_dr,max_force,rms_force,and rms_dr keywords added

2004-02-10 16:21  vondele

	* [r2077] bug fix : avoid taking square root of negative numbers :)

2004-02-10 16:02  vondele

	* [r2076] bug fix : uninitialized error (pointer)

2004-02-10 15:51  vondele

	* [r2075] bug fix : read_dft_section fix case construct (but what
	  is it doing here anyway !?)

2004-02-09 15:14  krack

	* [r2074] wf initialization removed, since it is already done
	  (qs_environment)

2004-02-08 16:20  vondele

	* [r2073] introduced the option map_consistent

2004-02-06 23:47  ikuo

	* [r2072] Error checking code and bug fix.

2004-02-06 20:21  vondele

	* [r2071] help ifort and dec.

2004-02-06 20:05  vondele

	* [r2070] cleaning out old debugging

2004-02-06 17:28  vondele

	* [r2069] (pw_inject) INTENT(INOUT) needed would be logical.

2004-02-06 15:21  vondele

	* [r2068] experimentation with realspace multgridding

2004-02-05 11:06  vondele

	* [r2067] fix problem with workspace of DSYEVD on IRIX with a given
	  lapack library

2004-02-04 16:37  vondele

	* [r2066] bug fix : (set_mc_env) INOUT not OUT, (qs_mon_car)
	  improve comment

2004-02-03 19:45  mcgrath

	* [r2065] bug fix?

2004-02-03 17:33  vondele

	* [r2064] (pw_grid_setup) : change needed to make sure the grids
	  are the same for the default case

2004-02-03 17:01  vondele

	* [r2063] Introduced an option to have grids that are also
	  commensurate in real space (simplest case only 2/1)

2004-02-02 15:20  vondele

	* [r2062] fix for long expansions

2004-02-02 09:53  krack

	* [r2061] _GE_ changed back to _ge_ to avoid troubling make

2004-02-01 21:27  mcgrath

	* [r2060] bug fixes

2004-02-01 20:24  vondele

	* [r2059] bug fix : (get_subgrid_size) : fix Ar-3.inp by calling
	  init_orbital_pointers

2004-02-01 19:31  mcgrath

	* [r2057] bug fixes

2004-02-01 18:37  vondele

	* [r2056] bug fix : (pw_grid_distribute) only mp_sum if
	  PW_MODE_DISTRIBUTED.
	  (others) introduce named constants PW_MODE_DISTRIBUTED /
	  PW_MODE_LOCAL

2004-02-01 15:25  krack

	* [r2055] Bug fix for Sdiag /= 1

2004-02-01 13:53  vondele

	* [r2054] bug fix : KIND=8 doesn't exist.

2004-02-01 10:25  gtb

	* [r2053] 1) new type of dyn coefs implemented (md with coefs not
	  yet done)
	  2) energy calculation with new coefs workinng
	  3) removed the corresponding outdated modules
	  (kg_energy_optimize, kg_pol_coefs)

2004-02-01 10:22  gtb

	* [r2052] removed subroutine calc_kxc_pot_kg, now the module is
	  used only as
	  an interface to
	  calculate xc derivatives. MOre cleaning.

2004-02-01 10:19  gtb

	* [r2051] New type of coefs implemented, but still not tested.
	  Cleaned.

2004-02-01 10:18  gtb

	* [r2050] kg energy calculation (with coefs) implemented. New type
	  of coefs. Cleaned.

2004-02-01 10:14  gtb

	* [r2049] Rewritten to work with the new dyn_coefs. only diis
	  optimization method is supported

2004-02-01 10:12  gtb

	* [r2048] works also with the new dyn coeffs. Cleaned.

2004-02-01 10:10  gtb

	* [r2047] Energy calculation for KG_pol (with coeffs) implemented

2004-02-01 10:07  gtb

	* [r2046] energy calculation for kg rewritten

2004-02-01 10:06  gtb

	* [r2045] new type of dynamical coeffs for KG_pol. not finished
	  yet, works
	  only for energy calculation not for MD

2004-02-01 09:32  gtb

	* [r2044] outdated. To be rewritten

2004-02-01 09:31  gtb

	* [r2043] outdated. To be rewritten when needed

2004-02-01 09:28  gtb

	* [r2042] outdated. Replaced by kg_energy.F

2004-02-01 09:19  gtb

	* [r2041] changed type of coefs: routine
	  replaced by dynamical_coeff_type.F

2004-01-31 19:46  mcgrath

	* [r2040] whoops...forget the object file

2004-01-31 19:45  mcgrath

	* [r2039] bug fixes

2004-01-31 10:09  vondele

	* [r2036] (qs_energy_type) : remove unused variable total_old

2004-01-29 22:07  gtb

	* [r2034] replaced INTENT(OUT) with INTENT(INOUT) for pot(:,:)

2004-01-29 15:20  vondele

	* [r2029] looking at the effect of main/branch changes

2004-01-29 14:18  hforbert

	* [r2028] Author: hforbert
	  removed PUBLIC declaration of convert_to_cp2k_units which is not
	  in
	  this file anymore.

2004-01-28 20:40  vondele

	* [r2025] bug fix : (init_cube_utils) : dangerous pasto for
	  non-cubic boxes

2004-01-28 15:59  tchassai

	* [r2019] * refactoring tddfpt_es

2004-01-28 15:58  tchassai

	* [r2018] removed unnecessary comments

2004-01-27 16:59  vondele

	* [r2013] (integrate_v_rspace) : save one useless fft on the high
	  cutoff grid

2004-01-27 15:58  vondele

	* [r2012]
	  (calculate_rho,integrate_v_rspace,calculate_wavefunction,integrate_rho_core,calculate_rho_core)
	  : change interface, do not require already available information
	  as arguments (pw multigrids, cube_utils, l_utils, ..)

2004-01-27 10:19  vondele

	* [r2011] (fft_wrap_pw1pw2)(rs_pw_transfer) : add distinct timings
	  for the various pw_grid levels (identifiers).

2004-01-26 21:31  vondele

	* [r2010] bugfixes (tddft_es) : nullify(T%matrix),
	  nullify(Tev%matrix)
	  (calc_initial_guess) : initialize energies

2004-01-26 17:04  vondele

	* [r2009] renamed p_new into p_mix_new. Do not allocate p_mix_new
	  if not ao_mixing_nr (saves one matrix for OT). Changes to store
	  directly in qs_env%rho instead of scf_env%p_mix_new if possible.

2004-01-26 15:08  tchassai

	* [r2008] * revert to xc_calculate_pw_new

2004-01-26 15:07  tchassai

	* [r2007] * very small optimization in calc_initial_guess

2004-01-26 15:00  tchassai

	* [r2006] * bugfix in calculate_initial_guess for the case that
	  fewer lumos than
	  requested eigenvalues are provided.

2004-01-26 14:57  vondele

	* [r2005] (qs_ot_allocate) : zero ls_diis

2004-01-26 14:41  vondele

	* [r2004] init_cube_utils : smarten error checking (avoid 0)

2004-01-26 12:44  tchassai

	* [r2003] * simplified the sorter section in calc_initial_guess
	  routine

2004-01-26 09:45  tchassai

	* [r2002] bugfix

2004-01-25 20:34  vondele

	* [r2001] fix the fix ;-)

2004-01-25 20:09  fawzi

	* [r2000] fixing my sloppy cleaning.

2004-01-25 19:24  vondele

	* [r1999] decided to guard against the common case, so that results
	  between sopt and sdbg are more likely to be similar. This will
	  affect energies on all platforms (see previous message). We're
	  trying to guard against CEILING of 7.99999999999999 or
	  8.000000000000001 leading to significantly different results (8
	  or 9).

2004-01-25 19:16  vondele

	* [r1998] bug fix to guarantee a consistent evaluation of
	  CEILING(-dxi*sqrt(max(radius2-kg2*dz2-jg2*dy2,0.0_dp))) on PC.
	  The code as is should prevent a different number of elements in
	  the counting and execution loop of init_cube_utils (where the
	  value of the expression depends on its location in the code,
	  presumably due to 80 bit fp registers). However, the above
	  expression can still lead to a result +-1 depending on
	  optimization level, and hence optimization level can influence
	  the QS energy by as much as EPS_DEFAULT (one gridpoint more or
	  less in mapping the density).

2004-01-25 15:40  fawzi

	* [r1997] fixes related to cleaning of omp directives.

2004-01-25 14:06  vondele

	* [r1996] removed rho ao new from update_rho. Remarkable 5-10%
	  faster on H2O.inp. Still needs to get rid of p_new in qs_scf

2004-01-25 12:58  vondele

	* [r1995] clean away rho_ao_with_h_struct which wasn't really
	  existing anymore

2004-01-25 12:01  vondele

	* [r1994] First pass of cleaning the rebuild feature away. Some
	  unrelated cleaning in the same files. Initial results 10% in
	  memory and speed for H2O.inp. Not all possible cleaning done in
	  kpp1 related stuff.

2004-01-24 17:55  krack

	* [r1993] more checks to avoid a crash

2004-01-24 13:02  vondele

	* [r1992] fft performance per cpu

2004-01-24 09:40  vondele

	* [r1989] restore OMP

2004-01-24 08:41  vondele

	* [r1988] some further hand cleaning

2004-01-23 22:30  fawzi

	* [r1987] cleaned the whole source with
	  tools/normalizeFortranFile.py.
	  removed 2305 use and 1816 vars.

2004-01-23 17:04  vondele

	* [r1985] added pw_derive_fd

2004-01-23 16:46  tchassai

	* [r1984] Enabled the lsd pade functional with the new xc
	  interface.

2004-01-23 16:41  krack

	* [r1983] GHOST potential keyword for BSSE computation added (CP
	  correction)

2004-01-23 08:22  tchassai

	* [r1982] Bugfix in pade (new).

2004-01-23 02:30  mcgrath

	* [r1981] bug fixes

2004-01-22 08:30  vondele

	* [r1980] no mo_ortho printing for medium

2004-01-21 10:34  krack

	* [r1978] reallocate_r5 added

2004-01-21 09:45  vondele

	* [r1977] fix : correct subroutine name in stop

2004-01-21 09:19  krack

	* [r1976] OMP update

2004-01-20 20:23  vondele

	* [r1975] bug fix : double precision constants

2004-01-20 20:06  vondele

	* [r1974] bug fix : use double precision constants

2004-01-20 16:30  tchassai

	* [r1973] Bugfix. The pade functional now works with the new xc
	  interface (no
	  copying of rho anymore).

2004-01-19 17:33  vondele

	* [r1972] bug fix : use all needed subroutines for the include

2004-01-19 10:17  tchassai

	* [r1969] Implemented pade with the new interface. Gives wrong
	  results unfortunately.
	  On to debugging...

2004-01-19 09:45  krack

	* [r1967] invalid array elements are not accessed any longer,
	  allocation of ai_* work
	  array moved to qs_core_hamiltonian, more explicit integrals

2004-01-19 09:45  fawzi

	* [r1966] cleaning preparations

2004-01-18 01:11  mcgrath

	* [r1960] bug fixes

2004-01-16 07:57  tchassai

	* [r1959] More refactoring for the new xc facilities.

2004-01-15 14:19  tchassai

	* [r1958] Workaround for circular dependency.

2004-01-15 13:38  tchassai

	* [r1957] Separated dft_types into cp_control_types and
	  cp_control_utils (read routines)
	  to remove a problem with circular dependencies and the new xc
	  facilities.

2004-01-13 12:58  vondele

	* [r1956] return_nan not for absoft...

2004-01-13 10:20  tchassai

	* [r1955] Example for new function call naming scheme in
	  xc_functional_eval.

2004-01-12 16:27  fawzi

	* [r1954] * some bugfixes wrt. xc.
	  * unfortunately not yet all
	  * declaration normalizer took much longer than expected to debug,
	  will apply it tomorrow to these files

2004-01-11 00:13  vondele

	* [r1953] to debug, uncommenting the return_nan stuff,
	  qs_core_hamiltonian returns, whereas it should not. Possibly
	  several other instances of the same problem lying around (in
	  particular whatever leads to forces gth_ppl).

2004-01-11 00:02  vondele

	* [r1952] added function to return nan. Useful for debugging (?!)

2004-01-10 17:07  krack

	* [r1951] stars for printing of atomic number Z,
	  default_string_length introduced,
	  declarations "beautified" (big indentation due to KIND =)

2004-01-10 16:53  vondele

	* [r1950] timing different for forces and non-forces

2004-01-10 16:32  krack

	* [r1949] string_utilities (minor cleaning)

2004-01-10 15:27  krack

	* [r1948] improved VPPNL structure and search

2004-01-10 15:14  vondele

	* [r1947] cosmetics

2004-01-10 14:54  vondele

	* [r1946] added timing in reading and writing MO restarts

2004-01-09 18:34  jgh

	* [r1945] improved list search

2004-01-09 16:52  tchassai

	* [r1944] Refactoring xc for the new interface.

2004-01-09 15:44  vondele

	* [r1943] xc_calc_derivatives : add timing

2004-01-09 13:18  vondele

	* [r1942] bug fix : OMP : explicitely declate my_compute_tau shared

2004-01-09 11:06  vondele

	* [r1941] Improved version of OT using taylor expansion whenever
	  useful

2004-01-08 11:33  vondele

	* [r1940] remove useless use

2004-01-08 11:06  vondele

	* [r1939] Added a print key for HPM so that it can be used more
	  flexibly (default is false)

2004-01-07 17:11  vondele

	* [r1938] added __HPM section to use IBM's hardware performance
	  monitor

2004-01-07 15:45  krack

	* [r1937] Additional printing

2004-01-06 16:25  vondele

	* [r1936] added cp_fm_maxabsrownorm (||A||_inf)

2004-01-06 16:00  tchassai

	* [r1935] Refactoring the xc routines.

2004-01-06 08:51  vondele

	* [r1933] Comment non-fortran stuff.

2004-01-05 16:30  vondele

	* [r1932] rename cp_fm_maxval in cp_fm_maxabsval to describe more
	  accurately what the routine does

2003-12-30 15:17  tchassai

	* [r1931] Forgot these comments.

2003-12-30 15:14  tchassai

	* [r1930] - A little cleanup of the tddfpt files.
	  - Cleaned (and corrected) comments of xc_rho_set_types.F. Author
	  and
	  Modification History are really useless in the sourcecode since
	  you can
	  easily get them from cvs.
	  - Added a precondition to allocate_rho_set

2003-12-23 08:28  vondele

	* [r1927] - atomic forces for tau dependent functionals
	  - changed definition of tau to be consistent with TPSS paper (1/2
	  * ... )

2003-12-22 18:28  vondele

	* [r1926] small fix in tau_xc for parallel runs

2003-12-22 15:05  vondele

	* [r1925] added kinetic energy density (tau) to integrate an
	  collocate.

2003-12-22 09:58  vondele

	* [r1924] basic changes to allow for the kinetic energy density
	  (tau) to be used for xc

2003-12-21 19:14  vondele

	* [r1923] bug fix : define dfac,gamma0,gamma1

2003-12-19 17:32  vondele

	* [r1921] improved error message if the kind section for a given
	  atom is missing, and a default is not present

2003-12-19 16:19  gtb

	* [r1918] added NULLIFY(elec_conf)

2003-12-17 17:10  ikuo

	* [r1915] appended "_dbl" to topology%bondparm_factor

2003-12-17 13:09  krack

	* [r1913] All-electron configuration removed from the GTH-* data
	  sets

2003-12-17 00:39  ikuo

	* [r1906] New &COORD section format...
	  atom_name1 X1 Y1 Z1 [mol_name1]
	  atom_name2 X2 Y2 Z2 [mol_name2]
	  atom_name3 X3 Y3 Z3 [mol_name3]

2003-12-17 00:35  ikuo

	* [r1905] Remove duplicate "USE kinds" for DEC

2003-12-16 18:12  vondele

	* [r1904] allow for system calls

2003-12-16 15:03  vondele

	* [r1903] explicitly use the new module

2003-12-16 12:20  gtb

	* [r1901] converted to a module to be nag-friendly

2003-12-16 11:34  vondele

	* [r1900] remove ';'

2003-12-16 07:32  krack

	* [r1899] INTENT(OUT) -> INTENT(INOUT) for pot

2003-12-15 12:54  krack

	* [r1898] more NAG digestable; some cleaning in ai_*

2003-12-13 22:10  vondele

	* [r1897] a first fix for the -nan options. Only copy the part of
	  the array that is really needed (and defined)

2003-12-13 21:17  vondele

	* [r1896] no ; on an empty line

2003-12-13 20:32  fawzi

	* [r1895] changed linked list template to allow for private
	  functions (and thus make xc_deriv compare a function and respect
	  maximum line length).

2003-12-12 23:14  fawzi

	* [r1893] new xc structure

2003-12-12 19:27  vondele

	* [r1892] fix an intent(out) bugfix

2003-12-12 16:27  jgh

	* [r1890] KIND default matches all

2003-12-12 15:41  vondele

	* [r1889] bug fix : use double precision constants

2003-12-12 10:33  vondele

	* [r1887] bugfix : use double precision constants

2003-12-11 15:34  vondele

	* [r1884] made a wrapper for all blacs calls

2003-12-11 10:59  vondele

	* [r1879] bug fix : nullify all needed pointers

2003-12-10 13:46  vondele

	* [r1877] correct typo

2003-12-10 13:45  vondele

	* [r1876] sun bugfix

2003-12-10 07:16  krack

	* [r1875] pointer initialized

2003-12-09 22:01  vondele

	* [r1874] Bug fix : another INTENT(out) bug

2003-12-09 19:53  vondele

	* [r1872] machine file for the nag linux f95 compiler

2003-12-09 14:21  vondele

	* [r1869] a try to add the nag compiler here.

2003-12-09 13:38  vondele

	* [r1865] move all non-fortran stuff to the machine files. More
	  fixes for nag coming.

2003-12-09 12:07  vondele

	* [r1864] remove unused OMP lock.

2003-12-09 09:29  tchassai

	* [r1863] Added the derivative set in the function calls to xc_lsd
	  and xc_lda.

2003-12-08 10:04  vondele

	* [r1862] bug fix : CMPLX(a_dbl,b_dbl) needs third argument to give
	  a double complex result

2003-12-07 16:36  vondele

	* [r1861] small changes to compile on IBM (a variable should appear
	  only once in a private clause, and a named critical section
	  should have different names somehow (if in different subroutines
	  ?). I also gave different critical sections for all components of
	  the force, so they will not block each other)

2003-12-07 14:43  jgh

	* [r1860] OpenMP version

2003-12-06 17:35  krack

	* [r1858] bug fix for trps2=0

2003-12-05 17:15  ikuo

	* [r1857] Reverting back to version 1.11

2003-12-05 16:25  vondele

	* [r1856] more INTENT(INOUT)

2003-12-05 13:36  vondele

	* [r1854] bug fix : various instances of incorrect INTENT(OUT)

2003-12-05 13:08  vondele

	* [r1853] remove hard coded kind(8)

2003-12-05 13:07  vondele

	* [r1852] avoid possible bugs if calling message passing from a
	  serial code

2003-12-05 13:06  vondele

	* [r1851] bug fix : ; not allowed if f95

2003-12-05 13:05  vondele

	* [r1850] help absoft compile (is this right ?)

2003-12-05 13:04  vondele

	* [r1849] fix warning about the string length

2003-12-05 13:01  vondele

	* [r1848] bug fix : added OPTIONAL to error

2003-12-05 12:59  vondele

	* [r1847] bug fix : wrong arguments of DCOPY

2003-12-04 08:10  tchassai

	* [r1846] Bugfix in GGA 2nd deriv

2003-12-04 00:29  ikuo

	* [r1845] Fixes so that it runs on DEC with "-check bounds"

2003-12-02 16:03  vondele

	* [r1844] bug fix : initialize variable

2003-12-02 15:34  gtb

	* [r1843] 1) changed epc_potential_type to elp_potential_type
	  2) new neigbor lists for KG

2003-12-02 15:05  vondele

	* [r1842] bug fix : add nullify (of an otherwise unused variable ?)

2003-12-02 14:27  vondele

	* [r1841] provide explicit initial values for a few saved variables
	  (at least helps the intel compiler in debug mode, but makes sense
	  in general)

2003-12-02 11:59  vondele

	* [r1840] make double precision real

2003-12-01 17:33  vondele

	* [r1839] change to allocate instead of the likely incorrect use of
	  target (the targetted variables are local to the subroutine,I
	  guess they are non-existent after the call)

2003-12-01 17:18  vondele

	* [r1838] added #if defined(__HMD) so that _dump_allocated will be
	  used more often

2003-12-01 16:47  vondele

	* [r1837] release what you retain ...

2003-12-01 13:49  krack

	* [r1836] lpi handling started

2003-12-01 05:20  cjmundy

	* [r1835] samples nhc velocitiest

2003-11-30 15:43  vondele

	* [r1834] trivial change to input keywords

2003-11-29 17:48  gtb

	* [r1833] parallel version works

2003-11-27 18:19  ikuo

	* [r1830] Cleanup of the bug found by Gloria...

2003-11-27 13:01  gtb

	* [r1828] KG with the new structures. Still much to do: only
	  simple KG works (no coefficents yet). Only serial.

2003-11-27 12:48  gtb

	* [r1827] small fix (please check!)

2003-11-27 05:20  cjmundy

	* [r1826] 1) DEC bugs (Viva DEC compiler)
	  2) Parallel thermostats work

2003-11-26 17:37  fawzi

	* [r1825] adding back some needed USE statements

2003-11-26 09:25  jgh

	* [r1824] Remove test code for OpenMP

2003-11-26 06:24  krack

	* [r1823] update of FIST neighbor lists (not tested)

2003-11-26 02:00  cjmundy

	* [r1822] Silly parallel bugs.
	  I am sure I will find
	  more.

2003-11-25 16:24  jgh

	* [r1821] Another npme=0 bug fixed

2003-11-25 15:16  krack

	* [r1820] forgotten USE added to make popt compile

2003-11-25 05:16  cjmundy

	* [r1819] 1) Thermostats with new structures
	  2) Fist commented out. Matthias, please update
	  3) md_env rewritten
	  4) md_conserved_quantity works with new structures
	  5) Fixed initialization of velocity bug
	  6) structure_type out out

2003-11-24 15:48  jgh

	* [r1818] Bug fix for case with npme=0

2003-11-23 16:48  jgh

	* [r1817] New version of grid collocation/integration routines -
	  removes all redundencies

2003-11-23 16:46  jgh

	* [r1816] Nullify all pointers in grid_construct

2003-11-22 11:56  fawzi

	* [r1815] removed unneeded USEs, minor cleanup and fixes

2003-11-22 11:53  fawzi

	* [r1814] * new xc interface (using xc_rho_set)
	  * pw_pool now caches also simple cr3d arrays (for the new xc
	  interface)

2003-11-21 18:53  ikuo

	* [r1812] Allows for choice of vdw radius to be used to generate
	  bonds

2003-11-21 16:56  mcgrath

	* [r1810] more up to date version to look for the bug

2003-11-21 02:41  ikuo

	* [r1809] A cleaner version of molecular states... Hopefully.

2003-11-19 17:00  fawzi

	* [r1805] added forgotten USE

2003-11-19 16:10  fawzi

	* [r1804] smaller fixes, just before the big changes.

2003-11-18 18:27  vondele

	* [r1802] output on one CPU only

2003-11-18 18:21  ikuo

	* [r1801] Try this again...

2003-11-18 18:14  ikuo

	* [r1798] Fix bug to for H2 dimer example to work...

2003-11-18 11:41  vondele

	* [r1797] do not include pbc twice

2003-11-18 10:58  vondele

	* [r1796] improved format descriptor

2003-11-17 16:28  vondele

	* [r1795] bug fix : check the status of the right pointer

2003-11-17 16:03  vondele

	* [r1794] bug fix : deallocate fully

2003-11-17 15:53  vondele

	* [r1793] bug fix : deallocate fully

2003-11-17 15:52  vondele

	* [r1792] bug fix, give back to the pools

2003-11-17 15:16  vondele

	* [r1791] useless USE

2003-11-17 13:51  krack

	* [r1790] - possible fix for the IBM "IF leak"
	  - updated and some cleaning

2003-11-17 09:10  vondele

	* [r1789] bug fix (leak) : if obtained from a pool, give back to
	  the pool !

2003-11-16 21:57  vondele

	* [r1788] useless use

2003-11-16 21:45  vondele

	* [r1787] bug fix (leak) : move deallocate in the loop

2003-11-16 21:38  vondele

	* [r1786] bug fix : deallocate tot_rho_*

2003-11-16 21:24  vondele

	* [r1785] bug fix : release the created scf_env

2003-11-16 19:02  vondele

	* [r1784] useless use

2003-11-16 17:35  vondele

	* [r1783] possible bug fix : deallocate lmi

2003-11-16 17:30  vondele

	* [r1782] spurious stat check

2003-11-14 19:15  ikuo

	* [r1781] More cleaning...

2003-11-14 15:43  krack

	* [r1780] alloc/dealloc updated

2003-11-14 12:42  tchassai

	* [r1778] Bugfix.

2003-11-14 10:46  fawzi

	* [r1777] fixing the bugfix ;(

2003-11-14 10:42  fawzi

	* [r1776] fixed a small bug (size vs ubound), but probably
	  atom_kind info should be used

2003-11-14 10:41  tchassai

	* [r1775] Added xc_rho_set_type for functional calculations.

2003-11-13 23:54  ikuo

	* [r1774] 1. Clean up topology_pack routines
	  2. Started using default_string_length in topology section

2003-11-13 19:28  vondele

	* [r1773] remove unneeded code

2003-11-13 19:21  vondele

	* [r1772] bug fix deallocate stuff

2003-11-13 15:46  vondele

	* [r1771] bugfix release all variables

2003-11-13 15:39  vondele

	* [r1770] fix leakage in distribution_2d_release

2003-11-13 14:57  vondele

	* [r1769] deallocate gidx

2003-11-13 14:14  vondele

	* [r1768] fix leak due to fft_get_length

2003-11-13 13:58  vondele

	* [r1767] first try at cleaning up a bit of the greens_fn leaks

2003-11-13 11:43  vondele

	* [r1766] gf : explicit private

2003-11-13 11:31  vondele

	* [r1765] remove useless use

2003-11-13 10:52  vondele

	* [r1764] added allocated_XXX_count

2003-11-13 10:32  vondele

	* [r1763] useless USE

2003-11-13 10:31  vondele

	* [r1762] further experimenting.

2003-11-13 10:21  fawzi

	* [r1761] calculate_rho_core now uses pw_pools

2003-11-13 10:18  fawzi

	* [r1760] fixing release of rs_grids (leak, thanks Joost)

2003-11-12 21:11  vondele

	* [r1759] remove useless uses

2003-11-12 21:08  vondele

	* [r1758] remove useless use

2003-11-12 21:04  vondele

	* [r1757] remove useless uses

2003-11-12 20:19  vondele

	* [r1756] also dealloc rho_core itself

2003-11-12 19:50  vondele

	* [r1755] do not deallocate rho_core here (only its coeff)

2003-11-12 15:36  fawzi

	* [r1754] fixing more leaks (mol_setup), the old mol_struct and
	  mol_type are still leaking away.

2003-11-12 14:04  vondele

	* [r1753] use fixed deallocation routines

2003-11-12 14:04  vondele

	* [r1752] bug fix ? only deallocate associated members

2003-11-12 09:20  fawzi

	* [r1751] fixing some leaks

2003-11-12 09:17  fawzi

	* [r1750] better test for matrix compatibility.

2003-11-12 09:15  fawzi

	* [r1749] replacig wp -> dp in templates

2003-11-12 08:54  fawzi

	* [r1748] cosmetic diffs between reinstantiating the template and
	  doing the substitution wp -> dp in the instances.

2003-11-11 16:47  ikuo

	* [r1746] 2nd try at hacking away the cell stuff from the setup
	  section...

2003-11-10 23:11  ikuo

	* [r1745] Remove cell stuff from the setup section...

2003-11-10 21:24  mcgrath

	* [r1743] Alpha version of Monte Carlo stuff

2003-11-10 15:02  vondele

	* [r1741] added a very early version of a trajectory analysis tool.
	  Right now not functional

2003-11-10 00:05  ikuo

	* [r1739] Fix to work for pyridine.inp

2003-11-09 18:21  vondele

	* [r1738] Added some small additional functionality for OT

2003-11-09 17:41  vondele

	* [r1737] step around maxmo pools for the time being

2003-11-09 16:49  vondele

	* [r1736] Added a little extra timing and iteration information
	  scf_iteration count
	  md time / step

2003-11-07 17:32  vondele

	* [r1735] removed USEless USEs

2003-11-07 16:28  vondele

	* [r1734] unspecified.ener is easier to use than '(not
	  specified).ener' :-)

2003-11-07 15:00  krack

	* [r1733] just a more reasonable name and print also the forgotten
	  density matrix for spin 2

2003-11-07 13:24  vondele

	* [r1732] do not fold particles in the box when writing

2003-11-07 12:04  vondele

	* [r1731] added stop_parser for unmatched keywords in &SCF

2003-11-07 11:02  krack

	* [r1730] - improved atomic guess for spin polarized systems
	  - no full matrices are used any longer

2003-11-07 09:15  tchassai

	* [r1729] Bugfix for derivative_set.

2003-11-07 07:19  tchassai

	* [r1728] Fixed the uninitialized pointer to derivative_set in xc.F

2003-11-06 21:53  ikuo

	* [r1727] molecule_input.F got split up...
	  
	  most of topology_control.F is in topology.F now...

2003-11-06 21:46  ikuo

	* [r1726] Fix some leaks...

2003-11-06 17:42  gtb

	* [r1725] allows also order=3

2003-11-06 13:38  vondele

	* [r1724] bug fix : do not redefine nmol incorrectly

2003-11-06 13:20  krack

	* [r1723] bug for ntheta=0 fixed, but there may be more instances
	  with similar bugs

2003-11-06 09:18  vondele

	* [r1721] Implemented Nth order PS extrapolation

2003-11-06 08:30  tchassai

	* [r1720] Reanimated the allocation of data arrays in the
	  xc_derivative_type.

2003-11-06 01:37  ikuo

	* [r1718] New input for QS:
	  1) No more &MOLECULE (will
	  add better version later)
	  2) Use only one call to
	  topology control.
	  3) use NEW molecule structures in
	  molecular states code.
	  4) try with examples:
	  
	  H2O.inp
	  H20-32.inp
	  molecular_states.inp

2003-11-05 16:32  tchassai

	* [r1717] Introduced the derivative_set into the different
	  functional calculation
	  modules.

2003-11-04 13:36  vondele

	* [r1716] Implemented linear PS extrapolation

2003-11-04 13:35  vondele

	* [r1715] USEless, added check

2003-11-04 13:04  vondele

	* [r1714] but keep beta=0 if matrix_b is not present

2003-11-04 12:49  vondele

	* [r1713] cp_fm_scale_and_add : changed the default for beta from
	  0.0 to 1.0 (i.e. if present, add it, not ignore it)

2003-11-03 13:59  vondele

	* [r1712] bug fix : derivate_set%data starts at zero.

2003-10-31 15:58  vondele

	* [r1711] allocate zero-sized arrays instead of having a nullified
	  pointer.

2003-10-31 14:58  tchassai

	* [r1710] Revert to old state regarding xc_control in
	  derivative_set to make
	  cp2k build again.

2003-10-31 14:19  fawzi

	* [r1708] renamed fragement->subsys force_control->force_env_types
	  removed qs from generic files.
	  see tools/replacer.py for the replace script

2003-10-31 10:47  tchassai

	* [r1707] Integrated xc_control into xc_derivative_set_type for a
	  more compact code.

2003-10-30 16:56  tchassai

	* [r1706] Working towards the new derivative_types. Introduced the
	  derivative_set,
	  which will replaces the current pot.

2003-10-28 09:35  vondele

	* [r1704] wp -> dp change, script in cp2k/tools

2003-10-27 17:29  vondele

	* [r1703] remove unused USE statements

2003-10-26 21:16  vondele

	* [r1702] experimental routine for level shifting

2003-10-26 16:31  vondele

	* [r1701] centers become an argument of the wannier computation

2003-10-26 16:11  gtb

	* [r1699] removed the redundant component run_type from
	  setup_parameter_types,
	  now also FIST and KG use the globenv%run_type component

2003-10-26 15:51  vondele

	* [r1698] started splitting in qs_scf_post and added a few comments

2003-10-26 13:14  vondele

	* [r1697] Moved the post scf calculations to a new module
	  (qs_scf_post).

2003-10-25 21:22  cjmundy

	* [r1696] Found some wasps

2003-10-25 20:27  cjmundy

	* [r1695] More mosquitoesi

2003-10-25 20:10  cjmundy

	* [r1694] Bug fixes for constraints

2003-10-23 23:50  cjmundy

	* [r1693] Cleaning and 4x6 constraints (untested).
	  Don't worry Joost, we will have NH3
	  in no time.

2003-10-23 19:52  vondele

	* [r1692] give more workspace to zheevd (bug in Scalapack)

2003-10-20 19:07  vondele

	* [r1691] small stuff : replace constant by HUGE, split memory
	  check

2003-10-20 07:40  gtb

	* [r1690] corrected stupid missprint

2003-10-19 01:34  cjmundy

	* [r1689] NPT fix. Performace still seems OK..
	  Will test more.

2003-10-18 18:54  gtb

	* [r1688] some bugs removed, a little cleaning

2003-10-17 13:23  tchassai

	* [r1687] introduced a new type: xc_control_type, to have the
	  functional related
	  control parameters in one place.

2003-10-17 12:45  krack

	* [r1686] accessing only pointers which are present

2003-10-17 08:56  krack

	* [r1685] - usage of the obsolete particle_type components avoided
	  in many locations
	  - obsolete components are still present to avoid trouble, because
	  of possibly
	  indirect access via pnode

2003-10-16 08:45  tchassai

	* [r1684] finished replacing ftags with gradient_functionals and
	  crossterms

2003-10-16 07:10  tchassai

	* [r1683] eliminated the ftags

2003-10-15 15:14  tchassai

	* [r1682] Working to make xc clearer. ftags is being split into two
	  separate
	  arrays: gradient_functionals and crossterms.

2003-10-15 13:34  tchassai

	* [r1681] Cleaning up xc a bit.

2003-10-15 13:16  tchassai

	* [r1680] Update on the new xc derivatives interface. Nothing
	  working yet.

2003-10-14 11:28  tchassai

	* [r1679] - implemented xc_derivative_create
	  - added xc_derivative_set_create

2003-10-14 08:46  tchassai

	* [r1678] - split the module xc_derivative_desc from
	  xc_derivative_types

2003-10-10 15:02  krack

	* [r1677] to get popt working

2003-10-10 13:39  krack

	* [r1676] now using force_env instead of qs_env

2003-10-10 13:38  krack

	* [r1675] init of POINTERs to get PGI sopt working

2003-10-10 09:19  krack

	* [r1674] lost nullify recovered

2003-10-09 17:11  krack

	* [r1673] some bugs fixed, but still not working properly

2003-10-09 17:08  krack

	* [r1672] bug fix

2003-10-09 15:59  vondele

	* [r1671] use FFTSG if the user given library does not exist

2003-10-09 15:41  krack

	* [r1670] PDB format considered as special case

2003-10-09 10:51  krack

	* [r1665] missing calls recovered

2003-10-08 16:57  krack

	* [r1664] BFGS/PRFO optimizer from Allchem/MSINDO
	  adapted to CP2K by Nisanth Nair

2003-10-08 16:55  krack

	* [r1663] - new molecule_kind and molecule structures elaborated
	  - parallelization via molecules
	  - new read_particle_set routine for general format input
	  - QS parser updated to digest also horizontal tabs
	  => the missing feature in the parser from Joost wish list
	  - QS MD for NVE should work

2003-10-08 12:35  vondele

	* [r1662] added NULLIFY to keep pgf happy

2003-10-08 10:41  vondele

	* [r1661] remove unused USEd subroutine

2003-10-08 10:31  vondele

	* [r1660] avoid the use of dp as a variable name

2003-10-08 08:37  vondele

	* [r1659] change needed to help pgf90 compile the file

2003-10-07 14:37  fawzi

	* [r1658] * removed private to ease conversion, but do as if it
	  were there

2003-10-07 14:18  vondele

	* [r1657] added wannier none

2003-10-07 13:35  fawzi

	* [r1656] * began fully using force_env as interface for the
	  various run types
	  - only in QS for the moment, but never uses qs_env directly,
	  should work straight away even for fist and kg)
	  - cell changes not yet considered
	  - compatibility release: uses many old calls to give the old
	  behaviour, but many things could be done much better and in a
	  more clean way (especially wrt. to restart, input output)
	  * updated fragment
	  - added Matthias's new moelcule structure
	  - enforced private in existing code (but left non private)
	  - added pnode (for the time being, but to be removed)
	  * qs_env now directly uses a fragment for its particles,
	  molecules,...
	  - thighter integration with force_env, in future the
	  initialization of the fragment should be separated from the one
	  of qs_env
	  * minor fixes
	  - better logger handling in cp_error
	  - minor stuff

2003-10-07 13:22  fawzi

	* [r1655] added lists for Matthias's new molecule structures.

2003-10-07 13:21  fawzi

	* [r1654] changed comment that tells that the .F files are
	  generated automatically through a template.

2003-10-07 13:16  fawzi

	* [r1653] *added more visible comment to template instances, that
	  tells that they are automatically generated via a template.

2003-10-07 13:11  fawzi

	* [r1652] - list type for Matthias's new molecule structures.
	  - better deallocation handling of lists.

2003-10-07 10:28  vondele

	* [r1651] nice physcons print at medium

2003-10-07 10:22  vondele

	* [r1650] stop on unknown keyword, allow for UNIT{S}

2003-10-07 10:06  vondele

	* [r1649] new setting for printout. Medium and OT now compatible
	  (i.e. mo_eigenvalues moved), smaller changes

2003-10-07 10:05  vondele

	* [r1648] intendation

2003-10-02 11:16  vondele

	* [r1647] added DIRECT to the wannier mini options
	  {DIRECT,JACOBI,CRAZY}

2003-10-02 10:20  vondele

	* [r1646] print cubes centered around approx 0

2003-10-02 08:24  vondele

	* [r1645] remove unused use statements

2003-10-01 17:06  vondele

	* [r1644] included a good initial guess, and fixed a bug

2003-10-01 15:01  vondele

	* [r1643] made crazy localization the default

2003-09-30 17:59  vondele

	* [r1642] introduction of the crazy_rotations

2003-09-29 14:29  vondele

	* [r1641] moving towards 3pnt mini

2003-09-29 12:48  vondele

	* [r1640] Started using the preconditioner

2003-09-26 13:47  tchassai

	* [r1638] New types for functional calculations (including
	  derivatives).

2003-09-26 08:32  tchassai

	* [r1637] Bugfix for GGA

2003-09-24 22:42  cjmundy

	* [r1635] Added routine for preconditioner. Not used in CG yet

2003-09-24 14:17  fawzi

	* [r1634] * added flat_local_rows/cols
	  * fixed bug related to mixing different kinds (OHHOHH vs OOHHHHH)
	  in parallel

2003-09-24 14:12  fawzi

	* [r1633] offset begins at 0 ;-)

2003-09-24 14:10  fawzi

	* [r1632] better comment to para_env%group

2003-09-24 14:09  fawzi

	* [r1631] better cp_sm_write (but also check out cp_sm_output)

2003-09-24 14:08  fawzi

	* [r1630] nicer error message

2003-09-24 12:29  tchassai

	* [r1629] forgot to delete the old line

2003-09-24 12:21  tchassai

	* [r1628] * changed a pot=0 into a BLAS call

2003-09-24 09:45  jgh

	* [r1627] Robodoc related changes by Thierry Deutsch

2003-09-23 09:00  vondele

	* [r1626] more line search options

2003-09-20 00:30  cjmundy

	* [r1625] DEC compiler fix

2003-09-19 13:56  vondele

	* [r1624] took care of the overlap matrix

2003-09-19 11:40  jgh

	* [r1623] small changes for ifc compiler by Thierry Deutsch

2003-09-18 15:40  tchassai

	* [r1622] Bugfix

2003-09-18 14:38  vondele

	* [r1621] added extra features and output for a molecular state
	  analysis

2003-09-18 13:51  tchassai

	* [r1620] * added an option to use different functionals in the dft
	  and tddfpt
	  calculations respectively. Use the same keywords (for funtionals)
	  in
	  the &tddfpt ... &end section as in the &dft ... &end section.

2003-09-17 16:00  vondele

	* [r1618] first try at direct minimization for wannier orbitals

2003-09-17 07:56  vondele

	* [r1617] Additional complex full matrix functionality added.

2003-09-16 17:03  krack

	* [r1616] Simple MD for QS removed

2003-09-16 16:58  krack

	* [r1615] Restart of QS MD improved and temperature scaling
	  introduced

2003-09-16 16:29  cjmundy

	* [r1613] Cleaned up time convert
	  to make use of MK's new
	  additions. Now, convert_units
	  is very modular and general.
	  Please use this instead of
	  physcon.F

2003-09-16 01:15  ikuo

	* [r1608] Separte the functionality of localization and molecular
	  states.

2003-09-15 23:40  ikuo

	* [r1607] Forgot this

2003-09-15 23:02  cjmundy

	* [r1606] Some cleaning.
	  added &EWALD

2003-09-15 16:53  krack

	* [r1605] Append the new MD data to the existing MD file for a
	  restarted QS MD
	  (seems to work basically, but needs more tests)

2003-09-15 16:51  krack

	* [r1604] optional arg added to open_file utility

2003-09-15 16:50  krack

	* [r1603] enable restart of QS MD via dump file

2003-09-15 11:41  vondele

	* [r1602] correct syntax is a good thing

2003-09-15 11:11  vondele

	* [r1601] some more digits in iterate.dat

2003-09-15 09:45  krack

	* [r1600] temperature scaling

2003-09-12 09:15  krack

	* [r1597] better scaling for first step

2003-09-11 22:03  fawzi

	* [r1596] * refactored lbfgs optimizer
	  * rescaling of optimizer (now first line search is OK)
	  * better interface for force_env
	  * minor enhancements

2003-09-11 09:51  krack

	* [r1595] Print interatomic distances, if requested

2003-09-11 09:44  tchassai

	* [r1594] Bugfix for SUN/Solaris. The reallocate_str1 subroutine in
	  memory_utilities
	  doesn't work on SUN. The problem lies in the parameter list. If
	  the parameter
	  remains character(*) it works for arbitrary length strings (up to
	  20 before
	  because 20 was hardcoded for the temp array) but deallocate can't
	  determine
	  how many bytes it should free.

2003-09-10 16:25  krack

	* [r1591] Simple MD module for QS for intermediate use.

2003-09-10 16:22  krack

	* [r1590] Updated to CODATA 1998 values of the fundamental physical
	  constants

2003-09-10 12:40  tchassai

	* [r1589] Bugfix

2003-09-10 08:11  krack

	* [r1588] bug fix

2003-09-09 12:59  jgh

	* [r1587] fixed typo

2003-09-09 09:39  tchassai

	* [r1586] Bugfix: Shortened some identifiers that were longer than
	  31 characters.

2003-09-09 09:15  fawzi

	* [r1585] usability enhancements: touch EXIT to stop a calculation
	  in progress, restart writte each 10 iterations if no convergence
	  is reached (can be changed).

2003-09-09 09:13  fawzi

	* [r1584] fixed parallel runs, needs refactoring

2003-09-09 07:24  fawzi

	* [r1583] first geo_opt version.

2003-09-08 15:08  fawzi

	* [r1582] commented out check of sm_plus_fm_fm_t and sm_fm_matrix
	  multiply

2003-09-08 14:46  fawzi

	* [r1581] the long due junkification, that both Joost and me kept
	  forgetting...

2003-09-08 14:09  fawzi

	* [r1580] removed debug output

2003-09-08 12:00  fawzi

	* [r1579] first geo_opt update.

2003-09-08 12:00  fawzi

	* [r1578] * cleanup of distribution 1d/2d
	  * minor fixes

2003-09-05 18:05  gtb

	* [r1577] new potential type for KG, some bug fixed

2003-09-05 12:09  jgh

	* [r1576] Faster pw_grid_distribute (may change results slightly)

2003-09-05 04:45  cjmundy

	* [r1574] tau_nhc in 1/cm seems to work now

2003-09-04 17:04  krack

	* [r1572] Temporary hack to have PBE using the old routines

2003-09-04 10:25  vondele

	* [r1571] bug fix

2003-09-03 00:20  ikuo

	* [r1570] Need more RAID!!!

2003-09-02 19:03  ikuo

	* [r1569] Change so setup%natom_type check will go through for
	  different molecules
	  that share a common atom label.

2003-09-02 19:00  ikuo

	* [r1568] Add reallocate_str1 for strings

2003-09-02 17:48  ikuo

	* [r1567] BUG FIX in determining the num mol for each mol_setup

2003-09-02 08:24  vondele

	* [r1566] faster algorithm for rs_pw_transfer based on mp_shift

2003-09-01 17:49  fawzi

	* [r1565] fixed bug with 0 sized arrays in isendrecv

2003-09-01 17:09  krack

	* [r1564] Bug fix

2003-09-01 16:37  krack

	* [r1563] molecule_kind_type and new molecule_type (stage 1, just
	  its definition)

2003-08-30 13:13  vondele

	* [r1562] do not replicate the distribution_2d for these matrices
	  (not respected in rs_get_my_tasks)

2003-08-30 13:11  fawzi

	* [r1561] * new matrix_multiply_2d that takes advantege of the
	  sparse matrix distribution and scales better with a large nr of
	  cpus
	  * smaller fixes

2003-08-30 08:22  ikuo

	* [r1560] *** empty log message ***

2003-08-30 08:12  ikuo

	* [r1559] parallel molecular states, works for water examples...

2003-08-30 08:11  ikuo

	* [r1558] Add another routine to mp_allgather

2003-08-30 08:10  ikuo

	* [r1557] Remove a write statement that was used in debugging

2003-08-30 02:05  ikuo

	* [r1554] add molecular_states entry

2003-08-30 00:39  cjmundy

	* [r1553] Reverted to Fawzi and Joost's fix

2003-08-29 23:42  cjmundy

	* [r1552] Lots of new stuff:
	  1) Topology with QS
	  2) Molecular states
	  3) units

2003-08-29 13:09  vondele

	* [r1551] fixed a bug related to passing a non nullified argument
	  to replicate_matrix

2003-08-29 11:52  jgh

	* [r1550] A test

2003-08-20 09:08  krack

	* [r1545] stop_program in deallocate_neighbor_list_* routines
	  removed

2003-08-20 07:27  vondele

	* [r1544] started number theory @ cp2k

2003-08-19 15:56  krack

	* [r1543] print only a warning instead of program termination in
	  order to allow
	  for dirty restart tricks.

2003-08-19 15:53  krack

	* [r1542] pointer status checking removed, since the status should
	  be ok (if not then
	  we will learn hopefully more from the next crash :-)

2003-08-19 15:51  krack

	* [r1541] new abstol definition avoids crash of the debug version

2003-08-19 05:17  ikuo

	* [r1540] Fix for AIX, DEC, and INTEL compiler

2003-08-19 00:44  ikuo

	* [r1539] Change so it'll compile on AIX

2003-08-19 00:08  ikuo

	* [r1538] Reorganizing and moving things around.

2003-08-15 17:34  ikuo

	* [r1537] Cleanup the file just a little.

2003-08-15 17:31  ikuo

	* [r1536] Remove unused USE statement.

2003-08-15 06:48  ikuo

	* [r1535] Some cleanup.

2003-08-15 04:47  ikuo

	* [r1534] Some cleanup and checks for memory allocation and
	  deallocation
	  
	  File topology.F moved to topology_gromos.F and will be called
	  from topology_control.F in the near future.

2003-08-15 00:46  ikuo

	* [r1533] Add topology_*

2003-08-15 00:35  ikuo

	* [r1531] Add PDB and PSF read.

2003-08-13 11:48  vondele

	* [r1530] safer version till testing can begin

2003-08-13 11:36  vondele

	* [r1529] change name of distribution_2D in distribution_2d to
	  avoid recompilation

2003-08-13 10:02  vondele

	* [r1528] initial introduction of distribution_2D in the overlap
	  matrix

2003-08-13 09:01  vondele

	* [r1527] remove unused USE statements

2003-08-13 07:30  vondele

	* [r1526] Addition of distribution_2D_types, and little cleaning
	  removing unused USE statements

2003-08-12 15:17  vondele

	* [r1525] bug fix, release of the wrong object

2003-08-12 13:52  jgh

	* [r1524] Remove unneeded dependency

2003-08-11 09:00  vondele

	* [r1523] increased buffer size as this is slightly faster

2003-08-10 11:03  vondele

	* [r1521] quite (50%, 256 H2O, 32 CPUs SP4) faster
	  apply_preconditioner by going from
	  cp_fm_symm to cp_fm_gemm

2003-08-09 18:50  vondele

	* [r1520] further documentation for cp_fm_types

2003-08-09 18:25  vondele

	* [r1519] added some documentation to cp_fm_basic_linalg

2003-08-09 16:08  vondele

	* [r1518] also add qs_mo_methods.F

2003-08-09 16:06  vondele

	* [r1517] move subroutines from qs_mo_types to qs_mo_methods

2003-08-09 14:43  vondele

	* [r1516] qs_blacs.F is gone (after three days of cleaning) ! added
	  qs_mo_methods for make_basis (needs cleaning)

2003-08-09 14:12  vondele

	* [r1514] modifications so that ifc compiles with -C (but ifc -C
	  turns out to be not bug free ...) and some general cleaning

2003-08-08 19:11  vondele

	* [r1513] removed write_blacs_matrix (cp_fm_write,
	  write_fm_with_basis_info ).
	  cp_fm_write might need some additional work

2003-08-08 18:27  vondele

	* [r1512] remove blacs_column_copy (cp_fm_to_fm)

2003-08-08 15:47  vondele

	* [r1511] remove cp_fm_utils also here

2003-08-08 15:44  vondele

	* [r1510] removed cp_fm_utils.F

2003-08-08 15:32  vondele

	* [r1509] got rid of copy_blacs_to_ionode_matrix in favour of
	  cp_fm_get_submatrix

2003-08-08 14:51  vondele

	* [r1508] removed blacs_get_element (cp_fm_get_element)

2003-08-08 14:02  vondele

	* [r1506] more qs_blacs cleaning (remove unused subroutines)

2003-08-08 13:16  vondele

	* [r1505] removed arpack from qs

2003-08-08 13:02  vondele

	* [r1504] remove blacs_scale_matrix (cp_fm_column_scale)

2003-08-08 12:06  tchassai

	* [r1503] Removed references to arpack in tddfpt and bugfix.

2003-08-08 11:31  vondele

	* [r1502] remove blacs_replicated_copy (cp_fm_get_submatrix,
	  cp_fm_set_submatrix)

2003-08-08 08:50  vondele

	* [r1501] further changes to remove qs_blacs
	  cp_fm_transpose, cp_fm_upper_to_full, blacs_add_to_element,
	  blacs_scale_and_d, ..

2003-08-07 17:14  vondele

	* [r1500] ! and again further cleaning of qs_blacs
	  * replicate_blacs_matrix (gone use cp_fm_create /cp_fm_to_fm)
	  * set_local_block (fully gone)
	  * copy_blacs_to_blacs_matrix (cp_fm_to_fm)

2003-08-07 15:02  vondele

	* [r1499] added cp_fm_diag as to collect the diagonalization
	  routines for cp_fm_types

2003-08-07 12:52  vondele

	* [r1498] encountered an in parallel very strange routine. Now
	  named : copy_local_sm_to_replicated_fm but this still misses the
	  point.

2003-08-07 12:26  vondele

	* [r1497] - cp_sm_fm_interactions.F : new file that contains the
	  operations that act on sparse matrices and full matrices at the
	  same time.
	  - further cleaning

2003-08-07 09:55  vondele

	* [r1496] first step towards removing qs_blacs (more to follow).
	  - cleaned USE statements
	  - consistently renamed towards cp_fm_*

2003-08-06 12:01  vondele

	* [r1495] remove bad ideas in comments

2003-08-05 19:07  vondele

	* [r1494] small prep change for a new version of cp_sm_fm_multiply
	  and cp_sm_plus_fm_fm_t that should be written one of these days

2003-08-05 17:39  vondele

	* [r1493] small rename for clarity

2003-08-05 13:08  vondele

	* [r1492] cosmetics

2003-08-05 10:52  vondele

	* [r1491] Added initialization for the molecule_type

2003-08-05 09:37  vondele

	* [r1490] bug fix : new liwork calculation (following the
	  instructions in pdsyevx.f)

2003-08-05 08:37  vondele

	* [r1489] possible bug fix for parallel runs (chris, please check)

2003-08-05 07:42  vondele

	* [r1488] again the xlf problem with len_trim giving back -1 for
	  empty strings.... (not reproducable with small code ...)

2003-08-05 01:32  cjmundy

	* [r1487] Got rid of uppercase function.
	  Lets the user define labels
	  z.B. Ar NOT AR for argon

2003-08-04 20:22  vondele

	* [r1486] bug fix ? unclear if these might be not associated at
	  that point, will need double checking (Matthias)

2003-08-04 15:05  vondele

	* [r1485] bug fix : avoid using a non-present optional argument

2003-08-04 14:07  fawzi

	* [r1484] fixed bug in init of fragment.

2003-08-04 14:02  vondele

	* [r1483] rename nose.F extended_system_dynamics.F to avoid
	  problems with the dependency generation process that requires
	  file names = module names

2003-08-04 11:44  vondele

	* [r1481] new real_matrix_to_full_matrix code (signigicant parallel
	  speedup)
	  rename of mp_transpose to mp_alltoall
	  modified comments

2003-08-04 11:19  fawzi

	* [r1480] added particle list instance

2003-08-04 11:04  fawzi

	* [r1479] added list of molecult kinds

2003-08-04 10:45  vondele

	* [r1478] bugfix : added initial zeroing

2003-08-04 09:47  fawzi

	* [r1477] * new interface for force_env
	  * fixed a bug with homo/lumo when nmo=nao (well probably it
	  should check and avoid the calculation if it is already
	  calculated,...)
	  * minor fixes

2003-08-04 00:45  cjmundy

	* [r1475] First step towards input merge of qs/fist/kg. I

2003-08-03 08:21  vondele

	* [r1474] remove globenv from calculate_w_matrix

2003-08-03 07:57  vondele

	* [r1473] faster algorithm for calculate_w_matrix_1

2003-08-01 17:40  vondele

	* [r1472] updated comment

2003-08-01 17:40  vondele

	* [r1471] faster parallel MO-ortho scheme

2003-07-31 17:43  fawzi

	* [r1470] fragment based force interface proposal

2003-07-31 11:35  vondele

	* [r1469] further bug fix, and reduced temporary array size to
	  provoke bugs...

2003-07-25 07:38  krack

	* [r1468] migrating to new stop_*

2003-07-24 08:15  tchassai

	* [r1467] Bugfix for the intel compiler.

2003-07-23 14:52  krack

	* [r1466] cleaning of qs_neighbor_list_types.F and let's give a try
	  to the new
	  versions of stop_memory and stop_program

2003-07-23 11:48  krack

	* [r1465] stack printing comes now also with stop_program and
	  stop_memory (work around)

2003-07-22 19:52  fawzi

	* [r1464] adding nn10 smoothing method.

2003-07-22 11:54  krack

	* [r1463] Parallelisation of the neighbor lists and 1e integrals
	  over rows and columns,
	  i.e. P>=1 and Q>=1 using the distribution for BLACS

2003-07-18 16:53  jgh

	* [r1462] Further bug fixes for distributed grids

2003-07-18 16:43  vondele

	* [r1461] added timing

2003-07-18 13:57  vondele

	* [r1460] deTab

2003-07-18 13:20  vondele

	* [r1459] added more output for the trace command

2003-07-18 06:52  krack

	* [r1458] distribution of rows and columns (QS, not activated, yet)

2003-07-17 17:54  jgh

	* [r1457] Bug fix for distributed grid code

2003-07-17 16:40  krack

	* [r1456] prefactor for diagonal elements corrected

2003-07-17 14:45  vondele

	* [r1454] remove tabs

2003-07-17 14:44  vondele

	* [r1453] remove tab

2003-07-16 07:21  vondele

	* [r1452] added RS_GRID {DISTRIBUTED,REPLICATED} in &QS block,
	  default REPLICATED for the time being

2003-07-15 23:27  ikuo

	* [r1451] DEC serial compiler fixes

2003-07-15 15:25  vondele

	* [r1450] comment away flush on SUN, as this causes crashes

2003-07-15 15:23  vondele

	* [r1449] added initialization

2003-07-15 14:24  vondele

	* [r1448] changed maximum allowed fft lengths for ESSL and FFTW

2003-07-15 13:29  krack

	* [r1447] consideration of limited numerical accuracy

2003-07-15 08:36  vondele

	* [r1446] back to the old style, but with the bug fix

2003-07-15 07:53  krack

	* [r1445] bug fix

2003-07-14 09:27  krack

	* [r1444] pointer status check

2003-07-14 09:20  krack

	* [r1443] pointer status check

2003-07-14 07:27  vondele

	* [r1442] added additional bug-catching

2003-07-12 15:50  cjmundy

	* [r1441] Bug fix in parallel setup found by Joost

2003-07-11 12:58  krack

	* [r1439] bug fix

2003-07-11 12:49  krack

	* [r1438] bug fix

2003-07-11 09:55  vondele

	* [r1437] bug fix for OT-MD and added a few more timing routines

2003-07-10 12:05  jgh

	* [r1436] Main code update for distributed realspace grids.
	  This code is currently only working for replicated grids.
	  Distributed grids are disabled.

2003-07-10 12:02  jgh

	* [r1435] Further updates for distributed realspace grids

2003-07-10 10:59  vondele

	* [r1434] bug fix in the sign of the 'forces' ?

2003-07-10 10:59  vondele

	* [r1433] small changes in format

2003-07-09 15:28  krack

	* [r1431] integrate_v_core parallelized and cleaned

2003-07-09 13:59  vondele

	* [r1430] More work on MD, something like a trajectory, energies
	  are printed out.
	  Implementation and format likely to change.

2003-07-09 13:31  tchassai

	* [r1429] Little improvements in the TDDFPT methods.

2003-07-09 12:48  krack

	* [r1428] cleaning

2003-07-09 10:03  krack

	* [r1427] qs_core_overlap parallelized; now only a single call for
	  each configuration
	  even if forces are requested.

2003-07-09 10:02  krack

	* [r1426] parsing of &QS section updated (DO WHILE replaced, has to
	  be done
	  also in other places)

2003-07-09 09:59  krack

	* [r1425] small bug fix

2003-07-08 16:05  krack

	* [r1424] work-around for LEN_TRIM problem using xlf 8.1.1

2003-07-08 11:10  vondele

	* [r1423] OMP fix, only reduction operations allowed on reduction
	  variables (and help not so smart compilers to see this is the
	  case)

2003-07-08 08:19  vondele

	* [r1422] removed debugging statement :-(

2003-07-07 10:51  vondele

	* [r1421] bugfix added initialization

2003-07-07 07:53  vondele

	* [r1419] broadcast input file name (appears needed in cntl_input)

2003-07-06 18:51  vondele

	* [r1417] changes to get qs_md parallel, and small clean-up in some
	  other files, a major cleaning of how the setup of fist/qs/kg is
	  performed is still needed.

2003-07-04 12:29  vondele

	* [r1416] modification to remove the blacs_env from the globenv
	  includes removal of cp_fm_create2
	  unused code in preconditioner.F
	  unused code in qs_blacs.F

2003-07-04 09:26  krack

	* [r1415] bug fix

2003-07-02 12:08  krack

	* [r1412] separate timing, if force calculation is requested

2003-07-02 12:08  krack

	* [r1411] cleaning, separate timing, if force calculation requested

2003-07-02 11:50  vondele

	* [r1409] replace mp_sum by a blacs-broadcast in get_element

2003-07-02 10:55  vondele

	* [r1408] wanniers working in parallel now

2003-07-02 09:27  vondele

	* [r1407] remove useless use

2003-07-02 09:26  vondele

	* [r1406] bug fix related to change in cp_cfm_get_element

2003-07-02 09:18  krack

	* [r1405] printing of sub-matrices enabled

2003-07-01 18:45  vondele

	* [r1404] Rewrote functions to make consistent use of complex
	  matrices. A few hundred times faster for the 32 H2O in serial :-)
	  Initial improvements to get the parallel version working, but not
	  so far yet.

2003-07-01 18:44  vondele

	* [r1403] cosmetics

2003-07-01 18:43  vondele

	* [r1402] added a few functions, trying to keep it in sync with
	  cp_fm_types.F

2003-07-01 14:17  vondele

	* [r1401] small bug-fixes, deallocate

2003-07-01 09:24  vondele

	* [r1400] added writing of wannier cube files, small fix for D Eq.
	  11, write the number of iterations

2003-07-01 00:23  cjmundy

	* [r1399] pzrot now works in serial

2003-06-30 12:09  vondele

	* [r1398] Id again

2003-06-30 12:08  vondele

	* [r1397] small update, check of id keyword

2003-06-30 10:43  krack

	* [r1396] check agreement of matrix structures

2003-06-30 09:05  vondele

	* [r1394] added a reference to the last CVS entry in the output

2003-06-28 18:39  vondele

	* [r1393] remove references to (.NOT.) new_core completely

2003-06-28 15:28  jgh

	* [r1392] Some changes for parallel real space grids (not
	  activated)

2003-06-28 15:10  jgh

	* [r1391] New changes for parallel real space grids

2003-06-28 15:09  jgh

	* [r1390] Bug fix for 1 processor MPI runs

2003-06-27 17:35  krack

	* [r1389] Pointer initialized

2003-06-27 17:02  krack

	* [r1388] missing declaration for OMP

2003-06-27 16:46  krack

	* [r1387] new row-wise parallelization scheme (first step: only QS)

2003-06-27 12:45  vondele

	* [r1386] added a TARGET attribute to make things compile on pgf90

2003-06-27 06:31  tchassai

	* [r1385] Started a cleanup of the code.

2003-06-26 10:54  vondele

	* [r1384] changed the way max radius is computed for the cube
	  utils. Now it should make sense for all cutoffs / eps_*
	  It is not trivial (impossible ?) to compute this, hence a
	  safety_factor is included.
	  The fact that we map the core charges on the highest grid only
	  prevents a small radius for some systems, as these core charges
	  should be done on a lower cutoff grid

2003-06-25 20:36  vondele

	* [r1383] fix write format problem for large systems

2003-06-25 14:34  tchassai

	* [r1382] Changes to tddfpt. Educated guesses for the initial guess
	  are now implemented.
	  To use the 'clever' initial guess rather than the ramdom guess,
	  specify the
	  LUMOS X option in the PRINT section of the input file. TDDFPT
	  will then auto-
	  matically use the computed virtual orbitals to find an initial
	  guess.

2003-06-25 14:30  tchassai

	* [r1381] Added a parameter for cp_fm_init_random. One can now
	  specify which cols
	  to initialize with random values.

2003-06-25 13:44  vondele

	* [r1380] bugfix for the optical conductivity needed to set the
	  derivatives of S

2003-06-25 09:14  fawzi

	* [r1379] * printing of local_lists
	  * fixed round-rbin bug when natom<num_pe

2003-06-24 12:59  fawzi

	* [r1378] fixed some bugs wrt. particle distribution.

2003-06-24 09:06  fawzi

	* [r1377] fixed bug in mol_local_particles_create

2003-06-23 13:50  fawzi

	* [r1375] new structure for the distribution of particles

2003-06-23 13:03  fawzi

	* [r1374] * fixed many bugs in xc splining (it reproduces the old,
	  non optimized values)
	  * new structure to describe the distribution of particles between
	  processors
	  * double give_back in pools now catched if cp_debug==.true.
	  (-DFD_DEBUG in the precompiler defines)
	  * fixed a bug in linked lists insert_ordered

2003-06-23 08:40  tchassai

	* [r1373] Bugfix for LSD.

2003-06-23 08:38  tchassai

	* [r1372] Removed some debugging code.

2003-06-20 16:02  tchassai

	* [r1371] Fixes for TDDFT. Singlets and Triplets now work. Still
	  need to clean
	  the code a bit though.

2003-06-20 10:13  vondele

	* [r1370] removed a two timing calls, as they appeared thousands of
	  times, making it effectively useless

2003-06-19 15:49  jgh

	* [r1369] parallel real space grids: generalization and
	  simplifications

2003-06-17 15:27  tchassai

	* [r1367] Moved the tddfpt call to a better place.

2003-06-17 15:26  tchassai

	* [r1366] Changed the "zeroing" of pot to a call to dcopy in two
	  instances.

2003-06-17 15:24  tchassai

	* [r1365] Update on tddfpt. Unfortunately triplets still don't
	  work.

2003-06-17 15:23  tchassai

	* [r1364] Cleanup of the module along with an merge of the tddfpt
	  functionality
	  into the original kernel.

2003-06-17 15:22  tchassai

	* [r1363] Simplified the calls to kpp1_calc_k_p_p1.

2003-06-17 15:21  tchassai

	* [r1362] Changed the eigensolver routine to dynamically choose the
	  number of
	  krylov space vectors through an iterative procedure. Thus the
	  N_KV_FAC
	  keyword of the TDDFTP section in the configuration file is
	  obsolete
	  and this option has been removed. Also a nice convergence factor
	  is
	  now displayed, such that the progress of the calculation can be
	  followed (this is for you Joost :-) .

2003-06-17 14:17  krack

	* [r1361] bug fix in PPNL neighbor lists

2003-06-17 14:16  krack

	* [r1360] typo corrected

2003-06-16 17:35  fawzi

	* [r1356] * cleanup of splining in xc
	  * minor fixes

2003-06-11 16:55  fawzi

	* [r1349] * splining using 1Dcomplex pw
	  * quadratic splining
	  * parallel splining
	  tested, not yet fully connected to xc.
	  smaller fixes

2003-06-11 13:48  vondele

	* [r1348] replaced MOD by MODULO to avoid possible bugs for
	  negative arguments

2003-06-11 08:27  jgh

	* [r1345] cosmetics

2003-06-11 08:24  jgh

	* [r1344] new file real_space_test

2003-06-11 08:23  jgh

	* [r1343] Changes related to new rs_grid_setup (only unused code)

2003-06-11 08:22  jgh

	* [r1342] Towards distributed real space grids

2003-06-11 08:06  vondele

	* [r1341] bugfix : check stat only if allocate has been used

2003-06-10 19:34  vondele

	* [r1340] remove real_matrix_to_full_matrix copy for the new_core
	  option.
	  don't even allocate the full matrix if ot and new_core !

2003-06-10 14:48  krack

	* [r1338] new core Hamiltonian (available via the keyword NEW_CORE
	  in the QS section)
	  which is built without any redundancy.

2003-06-06 16:17  fawzi

	* [r1336] smoothed rho + smoothed deriv (78) now works.

2003-06-06 11:25  vondele

	* [r1335] remove SUN warning (real*4 const out of range)

2003-06-06 09:11  tchassai

	* [r1334] Changes to the eigensolver. Invalidated the -k option for
	  tddfpt. Now
	  the number of used vectors is determined by a convergence
	  criteria.

2003-06-06 09:09  tchassai

	* [r1333] Bugfix.

2003-06-06 07:58  vondele

	* [r1332] little change needed to compile on SUN (intent(INOUT))

2003-06-05 21:55  fawzi

	* [r1331] added vxc spline smoothing for non gga functionals,
	  changes to compile on absoft (respect intent(in)).

2003-06-05 17:09  fawzi

	* [r1330] added test chek of vxc

2003-06-05 12:25  fawzi

	* [r1329] some bug fixes and tests for smooth and sharp.

2003-06-05 09:09  vondele

	* [r1328] bugfix for the OMP part

2003-06-05 09:09  vondele

	* [r1327] removed double declaration of n

2003-06-04 16:36  vondele

	* [r1326] bugfix : flush istat is needed on IRIX

2003-06-04 16:17  vondele

	* [r1325] added smooth_rho (a deus ex machina)

2003-06-04 15:22  vondele

	* [r1324] removed the intent in (sun complains)

2003-06-04 10:01  fawzi

	* [r1323] fast spline deriv (in g space)

2003-06-03 13:57  fawzi

	* [r1322] * old (working) version of spline interpolation (just to
	  have a reference for the new wersion)

2003-05-30 07:45  vondele

	* [r1314] more optimism (but way too early)

2003-05-29 10:01  vondele

	* [r1313] added the derive_function_rho option to the DFT section.

2003-05-28 11:17  fawzi

	* [r1312] * added qs_matrix_pools
	  * added spline derivative of xc

2003-05-22 16:38  vondele

	* [r1311] bugfix :? Moved the NULLIFY before the assignment

2003-05-22 15:31  vondele

	* [r1310] added timestop to cp_fm_lu_decompose and
	  cp_cfm_lu_decompose (+corrected name)

2003-05-22 13:42  krack

	* [r1309] variable precision for force output

2003-05-22 08:34  krack

	* [r1308] Missing declaration added.

2003-05-22 05:32  cjmundy

	* [r1307] dumb bug fix

2003-05-22 04:45  cjmundy

	* [r1306] Serial Wannier functionality.

2003-05-22 04:44  cjmundy

	* [r1305] Now has serial Wannier functionality. Needs some work on
	  a nice input interface, but all seems well.

2003-05-21 07:33  krack

	* [r1304] init of delta moved after the jacobi threshold check

2003-05-14 09:15  vondele

	* [r1303] Juergs temporary hack to get the grids sorted for large
	  (4000Ry) cutoffs

2003-05-14 07:45  vondele

	* [r1302] removed hard-coded eps. Could influence the value for the
	  energy.

2003-05-13 09:01  vondele

	* [r1301] small output fix (extrapolation method)

2003-05-07 21:03  vondele

	* [r1300] bug fix for the eigensolver if cholesky_off is used and
	  nmo.ne.nao

2003-05-07 15:38  vondele

	* [r1298] bug fix : nullify pointer

2003-05-07 11:17  vondele

	* [r1297] changed the meaning of cp_fm_get_element. Now the value
	  of the of the matrix
	  element is always returned.

2003-05-06 21:04  vondele

	* [r1296] cp_fm_set_all has a second scalar element that allows to
	  set the diagonal. This is an easy way to obtain the unit matrix

2003-05-06 13:08  vondele

	* [r1295] ORTHO / MO_ORTHONORMALITY print key added, checks the
	  numerical orthonormality of the MOs.

2003-05-05 13:02  vondele

	* [r1294] remove omp_get_wtime

2003-05-05 11:13  gtb

	* [r1293] removed the replicated _pgf_ routines frm kg_density

2003-05-05 11:12  gtb

	* [r1292] removed the replicated _pgf routines in KG_density

2003-05-02 18:03  fawzi

	* [r1290] * md update (interpolation)
	  * prints stack on cp_* warnigs and failures
	  * removed globenv from some routines
	  * adapted some sparse matrix routines to multiples matrix
	  structures

2003-04-29 11:24  vondele

	* [r1289] improve calling style :-)

2003-04-29 09:57  krack

	* [r1288] bug fix for empty sac_ppl neighbor list etc.

2003-04-29 09:55  vondele

	* [r1287] call to timestop needed, changed bad calling behaviour

2003-04-29 09:53  krack

	* [r1286] bug fix for reading alias names in database files
	  original potential name is kept

2003-04-29 09:33  tchassai

	* [r1285] Fixed a bug in cp_fm_scale_and_add, that appeared, when
	  matrix_a and
	  matrix_b were the same. Now cp_fm_scale is used instead but a
	  warning
	  is given, since the code should call cp_fm_scale in the first
	  place.

2003-04-28 14:04  vondele

	* [r1283] fix problem related to cp_fm_scale_and_add

2003-04-28 08:41  vondele

	* [r1282] avoid out of bounds for if the if is guaranteed to be
	  false

2003-04-24 15:58  vondele

	* [r1281] restore original version

2003-04-24 15:33  vondele

	* [r1280] h and p redundant..... I appears that, using the gth, the
	  h_core might have non-zero blocks that are not present in s, even
	  on a single cpu. This hack fixes the problem.

2003-04-24 14:50  vondele

	* [r1279] copy and paste bug fixed

2003-04-24 14:49  vondele

	* [r1278] minor bug correction & init of print key

2003-04-23 16:47  vondele

	* [r1277] Implemented LSD forces. Seems to yield energy conserving
	  dynamics :-)

2003-04-22 09:17  vondele

	* [r1276] Fawzi's trick for the MD, OMP forces bug fix

2003-04-22 07:13  tchassai

	* [r1275] - Changes to time-dependent DFT. Singlet excitations in
	  Tamm-Dancoff
	  approximation now work.

2003-04-22 07:01  tchassai

	* [r1274] - Modified cp_fm_set_all to use BLAS.
	  - Fixed comment of cp_fm_get_element.

2003-04-17 14:27  fawzi

	* [r1273] corrected lda (bug when matrix size /= n_row_global)

2003-04-17 14:25  fawzi

	* [r1272] corrected check for leading dimension (thanks TC)

2003-04-16 19:41  vondele

	* [r1271] changed the reading of the potential name (e.g. GTH-BLYP)
	  so that not
	  just the first part is used (leading to wrong potentials read)

2003-04-16 19:01  vondele

	* [r1270] shorter lines for OMP directives needed on SUN

2003-04-16 14:08  vondele

	* [r1269] s

2003-04-16 12:14  tchassai

	* [r1268] - Changed cp_fm_scale_and_add to use BLACS.
	  - Bugfix for bug introduced in 1.24.
	  - A little cleanup.

2003-04-16 08:09  vondele

	* [r1267] more work on optical conductivity

2003-04-16 07:19  fawzi

	* [r1266] corrected stupid bug (grad vs forces)

2003-04-15 08:20  vondele

	* [r1265] added the rs_pools for integrate_potential

2003-04-12 20:55  fawzi

	* [r1264] * introduced force_env
	  * added a type for a complete neighbor list (neighbor_group)

2003-04-12 17:15  vondele

	* [r1263] cleaned mulliken charge routine

2003-04-12 16:49  vondele

	* [r1262] Initial modifications to compute the optical
	  conductivity, NYI

2003-04-12 09:49  vondele

	* [r1261] added a reference in banner

2003-04-10 08:46  fawzi

	* [r1260] activated rs_grid pools

2003-04-09 08:05  fawzi

	* [r1259] - template for pools
	  - more retain/release for pw stuff
	  - fm_struct now with local_leading_dimension
	  - minor ep changes

2003-04-08 08:56  tchassai

	* [r1257] Changed xc_prep_2nd_deriv and xc_calc_2nd_deriv

2003-04-08 08:33  tchassai

	* [r1256] Modifications to time-dependent calculations to compute
	  the correct singlet
	  and triplet excitation energies. Not yet working.

2003-04-08 08:31  tchassai

	* [r1255] Added subroutine xc_lsd.

2003-04-08 08:28  tchassai

	* [r1254] Integrated fields for a time-dependent perturbation
	  calculation.

2003-04-04 14:41  fawzi

	* [r1253] removing pippo :).

2003-04-04 12:48  vondele

	* [r1252] removed globenv and cp_fm_create2

2003-04-04 12:38  vondele

	* [r1251] slightly modified blocking behaviour

2003-04-04 07:34  vondele

	* [r1250] The data in random initial matrices is no longer
	  dependent on # cpus or nrow_block / ncol_block

2003-04-03 17:22  fawzi

	* [r1249] * some md updates
	  * timeset,timestop checks for mismatched calls
	  * wf interpolation (very preliminary, not yet activated)
	  * fixes for dealloc of target argument

2003-04-03 14:58  vondele

	* [r1248] bug fix for small matrices (yes, Fawzi, you were right,
	  it's dangerous)

2003-04-03 13:54  vondele

	* [r1247] small bug fix for rs_pw_transfer

2003-04-01 17:32  fawzi

	* [r1246] fixed deallocation bug (thanks Joost)

2003-04-01 16:06  krack

	* [r1245] Suppress redundant error output

2003-03-31 16:56  vondele

	* [r1243] hack needed for the SUN optimizer

2003-03-31 16:38  krack

	* [r1242] bug fix

2003-03-31 15:34  vondele

	* [r1241] changed timing routines

2003-03-31 13:40  krack

	* [r1240] duplicated error messages in parallel runs for
	  "SECTION_NOT_FOUND" are now avoided

2003-03-31 13:26  krack

	* [r1239] Improved error handling for undefined or wrongly defined
	  atomic kinds.

2003-03-31 11:15  vondele

	* [r1236] clarify

2003-03-31 10:30  vondele

	* [r1233] OMP update of the grid-related routines.
	  Good scaling of integrate_v_rspace_omp
	  and of calculate_rho_elec_omp
	  OMP version now a little faster than MPI version on IBM.
	  Still some work to do to get it really fast
	  (I have the feeling lots of time is lost in copying big arrays)
	  Locking technique used for the real_space_grid_types.
	  Needed new calling for the grid routines to avoid problem with
	  the xlf compiler (a factor of 1000 (sic) in speed)
	  we need to avoid data(na1:na2,nb1:nb2) and other array
	  expressions (unhappily).
	  
	  Gloria, I noticed you replicated the _pgf_ routines, it would be
	  great if you could
	  try not to duplicate them, in this way, changes could be much
	  more local. Please,
	  also check that the changes I made do not influence your results
	  (numbers should be identical)
	  (and if you would manage to use the original integrate_potential
	  and so on routines kim gordon
	  would be automatically OMP-parallel !)

2003-03-31 10:21  vondele

	* [r1232] cpu_time to wall_time, that is what we care about
	  (especially omp runs)

2003-03-31 10:20  vondele

	* [r1231] Now sort on elapsed time, use CPU_TIME and omp_get_wtime
	  for timings

2003-03-31 09:48  vondele

	* [r1230] init print_key%potential_energy_matrix

2003-03-31 09:34  vondele

	* [r1229] added some OMP

2003-03-31 08:54  krack

	* [r1228] bug fix

2003-03-28 20:48  fawzi

	* [r1227] added create, release and retain (to support pointer
	  based handling).

2003-03-28 20:46  fawzi

	* [r1226] cosmetic

2003-03-28 20:45  fawzi

	* [r1225] added localizazion forcing

2003-03-28 20:44  fawzi

	* [r1224] fixed wrong col number in matrix mutiplies.

2003-03-28 20:43  fawzi

	* [r1223] added localization forcing

2003-03-28 20:42  fawzi

	* [r1222] fixed compare

2003-03-28 20:42  fawzi

	* [r1221] corrected realspace_grid linked list

2003-03-28 20:41  fawzi

	* [r1220] cp_fm_get_info calls cp_fm_struct_get

2003-03-28 20:40  fawzi

	* [r1219] made get function compatible with cp_fm_get_info
	  (nrow_local,ncol_local are numbers).

2003-03-27 14:01  krack

	* [r1218] cleaning of QS neighbor_lists no longer needed

2003-03-27 08:32  fawzi

	* [r1216] fixed a bug in the get function, moved allocation of row
	  and col indices to the get function.

2003-03-26 21:53  vondele

	* [r1215] added a few more OpenMP comments

2003-03-26 19:13  vondele

	* [r1214] hidden bug removed ...

2003-03-26 16:43  krack

	* [r1212] variable name corrected in scalar version

2003-03-26 16:37  krack

	* [r1211] core_hamiltonian revisited, minor improvements

2003-03-26 13:51  vondele

	* [r1210] additional argument checks for sm_fm_mult

2003-03-25 16:14  krack

	* [r1209] Optional args trouble PGI compiler - fixed

2003-03-25 15:33  tchassai

	* [r1208] There were undeclared variables.

2003-03-25 12:51  tchassai

	* [r1207] Fixed small typo.

2003-03-25 12:39  tchassai

	* [r1206] Made the execution of the kernel a configurable
	  parameter.

2003-03-25 10:49  fawzi

	* [r1205] more arg checks

2003-03-25 10:25  fawzi

	* [r1204] smaller fixes.

2003-03-21 21:17  vondele

	* [r1203] cp_sm_fm_multiply & cp_sm_plus_fm_fm_t rewritten. Now up
	  to 5 times faster
	  (large parallel runs, total quickstep x 1.5). The old code is
	  still present for debugging.
	  New communication shifts the mos from cpu to cpu instead of
	  mp_summing them.
	  Uses a badly documented scalapack call to convert a given
	  scalapack format
	  on a given blacs grid to a different scalapack format on a
	  different grid :-)
	  This turns out to be rather fast.

2003-03-21 21:13  vondele

	* [r1202] correctly init total_charge

2003-03-21 21:04  vondele

	* [r1201] MULLIKEN print key added

2003-03-21 20:03  vondele

	* [r1200] Added very basic support for Mulliken analysis. In
	  Particular, core charges
	  need still to be substracted. Doesn't use full matrices to
	  compute SP (might
	  be useful elsewhere). Needs a fix in the case of smearing.

2003-03-21 18:01  fawzi

	* [r1199] fixed scf initialization bugs

2003-03-21 13:56  krack

	* [r1198] (p)dsyevd is now used for the full diag. instead of
	  (p)dsyevx as recommended
	  by Joost

2003-03-21 13:05  fawzi

	* [r1197] fixed release bug (thanks joost)

2003-03-21 12:42  tchassai

	* [r1196] Small bugfix.

2003-03-21 12:17  tchassai

	* [r1195] Fixes for parallel version.

2003-03-21 11:07  fawzi

	* [r1194] hopefully fixed S_sm_pools

2003-03-21 10:32  fawzi

	* [r1193] quick fix of sm_pools errors (sorry)

2003-03-20 22:29  fawzi

	* [r1192] fixed stupid bug.

2003-03-20 21:36  fawzi

	* [r1191] smaller md related fixes, and cleanup.
	  ("to few blocks"? bugs still there.

2003-03-20 19:22  fawzi

	* [r1190] fixed minor bug

2003-03-20 18:45  fawzi

	* [r1189] fixed bug in sm_struct add_block

2003-03-20 16:20  krack

	* [r1188] Minor bugs removed.

2003-03-19 17:14  fawzi

	* [r1187] * removed p from qs_env
	  * cleaned scf_env_do_scf (scf is also smarter wrt. changes)
	  - moved ot initialization
	  - fixed diis
	  * consolidated the changes when the structure of s changes
	  (now one has just to call qs_env_update_s_mstruct)
	  * removed energy_update and forces_update
	  (I just made the non update routine smarter).
	  * qs_env%rho%rho_ao is not replicated anymore
	  * the H_replicated_sm_pools are a little smarter in their
	  construction
	  * qs_rho now can be updated by just changing %rho_ao directly
	  (but this forces a rebuild)
	  * renamed cp_fm_add to cp_fm_scale_and_add
	  * many smaller changes

2003-03-19 13:51  tchassai

	* [r1186] Lanczos eigensolver implemented. Tam-Dancoff implmented.
	  Still needs testing and improvements.

2003-03-19 09:07  vondele

	* [r1185] improved mp_shift (compiles now on pc, does what comment
	  says, better order of arguments)

2003-03-19 08:16  vondele

	* [r1184] Added a shift operation. Corrected a possible bug with
	  the status variable in MPI

2003-03-17 14:53  vondele

	* [r1179] hopefully improved handling of very small matrices in
	  parallel.
	  nrow_local/ncol_local are now zero if there is no data on the cpu
	  (i.e. no element of the matrix is local).
	  However, local_data needs to be at least 1 element. Hence,
	  nrow_local * ncol_local != size(local_data).
	  In fact, it might be needed (in some more or less near future)to
	  have local_data allocated in such a way that its size is
	  identical on all cpus. (Either by having an adjusted leading
	  dimension, or by having more cols in the matrix).

2003-03-12 11:00  vondele

	* [r1178] trying to get the block size used in scalapack consistent

2003-03-12 10:45  vondele

	* [r1177] initial cleaning for new sm_fm_mult routine

2003-03-12 09:26  tchassai

	* [r1176] Eigensolvers for TD-DFT (not yet working).

2003-03-12 09:25  tchassai

	* [r1175] Further work on TD-DFT.

2003-03-12 09:23  tchassai

	* [r1174] Added a cp_fm_scale subroutine (to be able to scale a fm
	  without
	  providing a second dummymatrix)

2003-03-11 12:29  vondele

	* [r1173] made shared OMP variables explicit

2003-03-10 16:40  vondele

	* [r1172] some blacs grid flexibility

2003-03-10 13:38  vondele

	* [r1171] changed the preconditioner to pointer in ot_eigensolver
	  added timing

2003-03-10 12:57  vondele

	* [r1170] added timing and bugfix ?

2003-03-08 16:16  fawzi

	* [r1169] big scf update with many, many changes: changed sequence
	  of ops, splitted functions, removed module variables,...
	  tested
	  qs_env => handed around more as pointer
	  ep fixed Smo_inv
	  introduced pools for sparse matrixes

2003-03-07 09:07  fawzi

	* [r1166] fixed bug with redundant structure of rho (wrong forces
	  with 3term pseudo) (thanks Matthias).

2003-03-06 17:12  krack

	* [r1165] Some bug fixes. More flexible restart (hopefully).

2003-03-06 10:59  fawzi

	* [r1164] fixed an insidious bug in the parallel initialization
	  (thanks Matthias).

2003-03-06 10:58  fawzi

	* [r1163] cosmetic fix (print ndep)

2003-03-03 13:18  fawzi

	* [r1162] corrected factor -2 in ep, qs_scf streamlining coming
	  along.

2003-03-03 11:34  vondele

	* [r1161] 1 more digit for ASD

2003-03-03 10:07  vondele

	* [r1160] print out ASD (Average Stack Depth) of the routine in the
	  timings section
	  this tells you which are the serving routines, and if something
	  is missing in
	  the timings. Added additional timings and reference for OT

2003-02-28 18:21  fawzi

	* [r1159] added Smo_inv to p_env, more doc, ep seems to work (up to
	  a factor -2 ;)

2003-02-27 17:32  jgh

	* [r1157] Check for out of bound factorials

2003-02-27 14:27  jgh

	* [r1156] Test for time=0

2003-02-27 10:33  vondele

	* [r1154] little bug fix (scf_work1 not correctly associated for
	  OT)

2003-02-27 09:41  fawzi

	* [r1153] - first setp for scf cleanup
	  - first step for md wf extrapolation
	  - hacked in rescaling of ep operators in ep, pi H_rho pi v still
	  without S^-1
	  - commented out automatic tpdft calc
	  - minor fixes

2003-02-27 07:57  vondele

	* [r1152] OMP sm_fm_multiply, make_density_matrix

2003-02-26 16:25  jgh

	* [r1151] Updated the OpenMP directives, some private variables
	  were not declared

2003-02-26 15:59  vondele

	* [r1150] possible bug corrected ? needed to compile on IBM

2003-02-26 13:15  jgh

	* [r1149] OpenMP directives added

2003-02-25 21:11  vondele

	* [r1148] help the SUN optimizer, preallocate the array nco so that
	  the lower bound
	  is known before entering the subroutine that needs to use it
	  (even though
	  this one is correctly allocating it mmmmh a bit strange)

2003-02-24 16:19  vondele

	* [r1147] added printout of OMP threads (not yet used, looks good)

2003-02-24 16:18  vondele

	* [r1146] Help OMP compile

2003-02-24 16:17  vondele

	* [r1145] added comment

2003-02-24 16:08  vondele

	* [r1144] changed defaults for QS runs, changed EPS_DEFAULT

2003-02-24 16:07  vondele

	* [r1143] commented trace_debug (PS TRACE can be put in CP2K
	  section)

2003-02-24 14:41  jgh

	* [r1142] Call to trace_debug has to be after initialization of MPI
	  (at least
	  for parallel runs)

2003-02-24 14:40  jgh

	* [r1141] Add TARGET attribut to Lebedev grids

2003-02-22 13:33  jgh

	* [r1140] Replaced call to coeff_add by coeff_sumup to avoid
	  conflicts if
	  coeff_add is called with replicated arguments.

2003-02-22 13:32  jgh

	* [r1139] Extended reference grid concept to more types of pw
	  grids.
	  Pw grid type can be chosen in Quickstep and KG modules.
	  Extended some functions that apply to pw to automatically
	  recognize reference grids.

2003-02-22 13:29  jgh

	* [r1138] Trivial changes to comments and use statements

2003-02-21 15:56  vondele

	* [r1137] trace gives (allocated bytes/heap size). machine needs
	  mallinfo. ONLY machine_aix changed

2003-02-19 19:13  gtb

	* [r1135] some bugs fixed

2003-02-19 16:01  jgh

	* [r1134] pw_grid_setup now works with negative cutoff, this
	  results in a non-spherical
	  grid that can host at least pw with the cutoff.
	  Sorting of all grid types is now performed (locally), this allows
	  now to
	  have multi-grids with all kinds of pw setups.

2003-02-19 15:53  jgh

	* [r1133] Deleted unused "use" statement

2003-02-19 15:53  jgh

	* [r1132] Default for run_type is now method specific

2003-02-19 15:52  jgh

	* [r1131] just changed some comments

2003-02-19 15:32  tchassai

	* [r1130] - Separated p_op_l2 into two routines with the same name
	  (interface).
	  One takes a matrix of orbits as input the other takes a density
	  matrix.
	  - Fixed memory leak.

2003-02-17 12:38  jgh

	* [r1129] New version of rs_pw_transfer: Reduces scratch space
	  requirements
	  drastically (especially for jobs with many cpus).

2003-02-17 12:36  jgh

	* [r1128] Added vector support to mp_sum_root

2003-02-17 08:49  fawzi

	* [r1127] output of kpp1_env internals (to debug)
	  added structure for real mateixes (compiled, not yet tested)
	  smaller changes.

2003-02-13 15:16  krack

	* [r1126] typo corrected

2003-02-12 16:03  krack

	* [r1125] local BLACS context released

2003-02-12 14:06  krack

	* [r1124] Atomic mass converted to u for printing

2003-02-12 14:00  krack

	* [r1123] smear and/or restart works again

2003-02-12 10:21  krack

	* [r1122] some cleaning

2003-02-11 13:30  fawzi

	* [r1121] wf interpolation object (not yet activated)
	  added array pointer types to array utils

2003-02-10 10:32  vondele

	* [r1120] Added basic support for complex full matrices; added a
	  function to compute
	  determinants

2003-02-10 09:30  fawzi

	* [r1119] fixed stupid bug when deallocating pw_pool.

2003-02-10 09:01  gtb

	* [r1118] int_group defined

2003-02-07 18:38  gtb

	* [r1117] module for calculating conserved quantity; cleanup
	  of the FIST code (no more particle_prop_type).
	  general unit conversion still to be fixed

2003-02-06 14:47  fawzi

	* [r1116] cleaned xc module
	  cleaned ep
	  some print keys to control ep output
	  finite diff l2 calculation
	  scale and add of sparse matrixes
	  minor cleanup

2003-02-05 13:25  tchassai

	* [r1115] Some work on tddft and on the get_real_matrix_occupation
	  routine.

2003-02-04 17:02  fawzi

	* [r1114] eat_muchmemory--;

2003-02-01 11:37  vondele

	* [r1112] OT and LSD made compatible (?!)

2003-02-01 11:35  vondele

	* [r1111] Added a little check (problems with e.g. triplet H2)

2003-01-31 18:54  vondele

	* [r1110] init correctly

2003-01-30 16:37  fawzi

	* [r1109] fixed p_env l2 (now it seems to be correct), minor fixes.

2003-01-30 16:12  vondele

	* [r1108] PGI is a bit sloppy

2003-01-30 16:01  vondele

	* [r1107] minor changes for output

2003-01-30 14:33  vondele

	* [r1105] Modified allocation / Compute kinetic energy

2003-01-30 14:33  vondele

	* [r1104] Changed sign of the force

2003-01-29 17:54  krack

	* [r1103] jacobi now more efficient

2003-01-29 14:05  vondele

	* [r1102] init ndep

2003-01-27 15:46  vondele

	* [r1101] Para bug fix

2003-01-27 15:09  vondele

	* [r1100] provided a light interface to integrate_v_rspace

2003-01-27 15:08  vondele

	* [r1099] add cp_fm_get_element to public

2003-01-27 15:07  vondele

	* [r1098] initial work on localization

2003-01-27 15:06  vondele

	* [r1097] Initial work on localization

2003-01-27 13:52  fawzi

	* [r1096] fixed 1x1 parallel matrix and minor output bugs.

2003-01-24 18:16  vondele

	* [r1094] The universal librarian

2003-01-24 16:25  vondele

	* [r1093] Help beginning chemists

2003-01-24 16:14  vondele

	* [r1092] ! Interesting bug. Don't try to modify a literal string !

2003-01-24 14:04  vondele

	* [r1091] parallel bug fix

2003-01-24 12:27  krack

	* [r1089] Multip(licity) keyword added in QS section

2003-01-24 11:33  vondele

	* [r1087] LSD : spin density cube / integrated spin density

2003-01-23 16:34  fawzi

	* [r1086] - ep with Joost's preconditioner
	  - corrected bug in full matrix filename generation

2003-01-23 12:08  vondele

	* [r1085] q1 + q2 = bug found

2003-01-23 12:06  vondele

	* [r1084] More LSD-ization

2003-01-22 13:29  fawzi

	* [r1083] fixed minor bug for the GC LDA with new routines.
	  activated new vxc potential (with LSD).

2003-01-21 16:11  gtb

	* [r1082] removed pgicompiler bug

2003-01-21 14:48  gtb

	* [r1081] KG-polarizable MD (version0)

2003-01-21 14:43  gtb

	* [r1080] KG-polarizable MD (version0, not fully debugged)

2003-01-21 14:11  gtb

	* [r1079] adapted to kg-pol MD

2003-01-21 14:10  gtb

	* [r1078] added nhc_coef_setup for kgpol

2003-01-21 14:06  gtb

	* [r1077] adapted for pol-kg

2003-01-21 14:02  gtb

	* [r1076] adapted for kgpol md

2003-01-21 10:30  tchassai

	* [r1075] Changed the subroutine name of
	  copy_blacs_to_sparse_matrix of qs_blacs.F to
	  copy_full_matrix_to_real_matrix. Added a routine to determine the
	  occupancy
	  of the real_matrix in sparse_matrix_types.F

2003-01-21 08:17  krack

	* [r1074] cholesky is no longer automatically disabled for jacobi

2003-01-20 16:27  fawzi

	* [r1073] oops, re-commented new pot calculation and its check.

2003-01-20 15:18  fawzi

	* [r1072] compares old and new pot (useful only for LDA)

2003-01-20 15:17  fawzi

	* [r1071] minor fixes (some undefined var on first pass).

2003-01-20 11:49  vondele

	* [r1070] Minor side-fixes

2003-01-20 10:40  tchassai

	* [r1069] Renamed the subroutine copy_sparse_to_blacs_matrix from
	  qs_blacs.F to
	  copy_real_matrix_to_full_matrix.

2003-01-17 18:52  fawzi

	* [r1068] LSD functionals, works on mac, not yet activated.

2003-01-14 14:35  krack

	* [r1067] cholesky is no longer automatically disabled for jacobi

2003-01-14 14:33  krack

	* [r1066] timer added

2003-01-14 09:57  vondele

	* [r1065] tried to clean a little

2003-01-13 17:52  vondele

	* [r1064] More LSD (write wavefunctions)

2003-01-13 15:05  vondele

	* [r1063] DIIS and LSD made (hopefully?) compatible

2003-01-10 10:56  tchassai

	* [r1062] Reduce software entropy

2003-01-09 22:40  fawzi

	* [r1061] - better fm matrix output
	  - fixed various ep bugs (compiled with pgi)
	  - fixed pao-lsd thing that did not compile on absoft ppc

2003-01-09 16:21  vondele

	* [r1060] initial LSD-isation : only yielding LDA results at twice
	  the cost :-)

2003-01-09 08:05  fawzi

	* [r1058] corrected scaling problem.

2003-01-09 08:00  fawzi

	* [r1057] temporary hack for LSD

2003-01-09 07:59  fawzi

	* [r1056] better interface to cp_fm_add (now scaling only or
	  copying supported).
	  Should be renamed to something like cp_fm_scale_sum or
	  cp_fm_lin_comb.

2003-01-08 11:54  vondele

	* [r1055] weird change needed to get things compiled on PGI
	  No idea why this is working and the old version not
	  this was the error message I got before ....
	  
	  PGF90-S-0084-Illegal use of symbol wp - KIND parameter
	  (kg_intra.f90: 121)

2003-01-08 10:59  tchassai

	* [r1054] Bugfix.

2003-01-06 16:42  gtb

	* [r1047] temporary fix for working units

2003-01-06 16:41  gtb

	* [r1046] a little cleaning

2003-01-06 16:40  gtb

	* [r1045] kg code for molecules

2003-01-06 16:34  gtb

	* [r1044] a little cleaning

2003-01-06 16:32  gtb

	* [r1043] small changes in reading KG potentials

2003-01-06 16:31  gtb

	* [r1042] kg code for molecules - a little cleaning

2003-01-06 16:29  gtb

	* [r1041] intra-molecular correction for gaussian charges on the
	  same molecule

2002-12-23 19:35  fawzi

	* [r1039] minor fixes

2002-12-22 10:19  fawzi

	* [r1038] - kpp1_env now caches pot
	  - various ep bugfixes (now it begins to converge)

2002-12-19 13:40  fawzi

	* [r1037] - reanmed (qs_p_build_kernel.F qs_p_types.F
	  qs_p_utils.F)->(qs_kpp1_env_methods.F qs_kpp1_env_types.F
	  qs_p_env_methods.F qs_p_env_types.F)
	  - first changes to cache 2nd deriv in kpp1

2002-12-18 14:44  fawzi

	* [r1036] - fixed bugs in second xc deriv related to ii
	  - added stupid_solve (compiled, not tested)
	  - minor bug fixes

2002-12-17 10:13  fawzi

	* [r1035] - restart adapted for ep (ugly, should be rethougth)
	  - added kpp1 to ep
	  - names and printing of matrices made more nice

2002-12-17 09:59  krack

	* [r1034] virial defined

2002-12-13 17:35  krack

	* [r1033] cleanup_pw_poisson_solver call added

2002-12-12 08:10  tchassai

	* [r1032] Minor changes.

2002-12-11 12:17  fawzi

	* [r1031] bugfixes, ep produces numbers with l1 (l2 not active).

2002-12-10 19:08  fawzi

	* [r1029] - enhanced a little the output of matrixes
	  - cholesky now can have size of the matrix as parameter
	  - p_env write
	  - many ep-* fixes

2002-12-10 09:16  fawzi

	* [r1028] - now preconditions,... have the line number even if
	  -DFD_LONG_EXPANSIONS is not passed to the precompiler (now that I
	  develop on a platform that does not support long lines, it wass
	  really annoying not to have the line number).
	  - ep_* coming along.

2002-12-07 10:00  fawzi

	* [r1027] - began serious debugging of perturbation calls
	  - added a print_key

2002-12-06 16:14  tchassai

	* [r1026] Bugfix for SUN: Zero rep matrices

2002-12-06 11:42  fawzi

	* [r1025] fixed robodoc documentation

2002-12-06 07:52  fawzi

	* [r1023] - qs_env now supports reference counting
	  - ep_methods diagonalizes again the subsystems.

2002-12-05 14:25  tchassai

	* [r1022] Changed the names of debug_control and force_control of
	  the modules
	  kg_debug and kg_force to kg_debug_control and kg_force_control,
	  to get
	  the debugger working again on SUN.

2002-12-05 13:05  fawzi

	* [r1021] - renamed qs_build_ks_matrix to qs_ks_methods
	  - qs_env_type now uses ref_counting and is indipendent from
	  qs_env (i.e. in the future it could be part of qs_env)
	  - introduced ep_env, big chenges in ep_*, code not yet working,
	  smaller fixes in second xc deriv.

2002-12-04 15:11  krack

	* [r1020] Syntax error removed

2002-12-04 13:38  tchassai

	* [r1019] Some additional functionality and cosmetical changes.

2002-12-03 15:45  fawzi

	* [r1018] added %in_space.

2002-12-03 15:26  fawzi

	* [r1017] diago of subsystems works.

2002-12-03 13:43  tchassai

	* [r1016] Fix so that arpack also works with smaller basis sets.

2002-12-03 10:43  tchassai

	* [r1015] Small bugfixes

2002-12-03 10:28  tchassai

	* [r1014] upcased keywords and added missing initialization of
	  n_spins

2002-12-03 08:35  tchassai

	* [r1013] Trying to debug on sun.

2002-12-02 13:38  fawzi

	* [r1012] - ks_env now contained in qs_env, and using reference
	  counting
	  - fixed initialization bug in perdew86
	  - smaller changes to qs_ep_methods

2002-11-29 11:46  krack

	* [r1011] bug fix (correct maxl value for ncoset)

2002-11-29 08:11  tchassai

	* [r1010] Small changes.

2002-11-28 10:11  fawzi

	* [r1007] - removed globenv from build KS matrix
	  - first set-ups for ep
	  - changes following from the fact that now prettify does not
	  upcase string constants

2002-11-26 15:51  fawzi

	* [r1006] - added skeleton for ep calculations
	  - made poisson solver initialization more safe
	  - made drho in xc_functionals calls optional
	  - smaller bug fixes (especially in xc 2nd deriv)

2002-11-26 08:05  tchassai

	* [r1003] Workaround for bug with interfaces and PGI compiler
	  (V3.2.4)

2002-11-25 10:04  krack

	* [r1001] sum_ener_forces inlined and some flops saved.

2002-11-25 09:10  fawzi

	* [r1000] xc_calc_2nd_deriv: qs_rho_type => coeff_type.

2002-11-24 21:51  fawzi

	* [r999] moved calculation of the second xc derivative to module
	  xc, commented out old code in xc.

2002-11-24 21:49  fawzi

	* [r998] stupid change (sequence of keys in a comment), now the
	  python instantiation script sorts them to avoid this in the
	  future.

2002-11-22 11:28  fawzi

	* [r993] prettied cp_* pao_*

2002-11-22 10:52  krack

	* [r992] More MD stuff for QS

2002-11-21 17:13  krack

	* [r991] FIST nonbonded neighbor lists (modified version)

2002-11-21 08:32  fawzi

	* [r989] - updated qs_p_build_kernel, various bug fixes, better
	  flow, made vxc derif more efficent
	  - moved charge of rho_r grid into rho
	  - smaller fixes

2002-11-20 10:51  tchassai

	* [r988] Minor bugfixes.

2002-11-18 15:04  hforbert

	* [r986] Author: hforbert
	  
	  Fixed redefinition of output_unit in linklist_verlet_list.F
	  Fixed initialization of %energy_cutoff in the pair potentials
	  to be symmetric (a,b) = (b,a)
	  Fixed bug in splines.F for x values at or slightly above
	  maximum x. (as far as I know, nobody really encountered that one,
	  but it had to be fixed)

2002-11-18 09:44  fawzi

	* [r985] - fixed a bug in qs_rho rebuilding (fully rewrote rebbuild
	  process)
	  - implemented 2nd deriv of xc, just written, and compiled, many
	  changes in qs_p_build_kernel
	  - smaller bug fixes

2002-11-18 09:01  tchassai

	* [r984] Unified the naming of the full matrices. f_matrix and
	  fmatrix in names have
	  been renamed to fm.

2002-11-17 18:56  cjmundy

	* [r983] Re-instated some tchassi changes that were lost

2002-11-17 18:50  cjmundy

	* [r982] Re-instated some gtb fixes that somehow became lost

2002-11-17 18:40  cjmundy

	* [r981] Re-instated tchassi's fixes. Somehow were lost.

2002-11-17 17:22  cjmundy

	* [r980] Cleaning

2002-11-17 14:41  jgh

	* [r979] Some routines for spherical gaussian calculations

2002-11-16 21:50  cjmundy

	* [r978] Now calls potential_f

2002-11-16 19:24  cjmundy

	* [r973] QS neighbor list compatible

2002-11-13 08:58  krack

	* [r972] moved to qs_force.F

2002-11-11 16:24  fawzi

	* [r968] normailzed use.

2002-11-11 16:08  fawzi

	* [r967] - added fm_pools array
	  - changed ao_mo_fm_pools to an array of pools
	  - two phases initialization of MOs
	  - description of qs_env attributes
	  - p_utils adopted pools, some rewrites.

2002-11-07 16:31  krack

	* [r965] run type for QS

2002-11-07 16:21  krack

	* [r964] some reordering in cp2k.F, environment.F and qs_main.F;
	  RUN_TYPE introduced in &CP2K

2002-11-07 16:17  krack

	* [r963] cp2k_input.F merged with environment.F; qs_init.F moved to
	  qs_environment.F

2002-11-06 13:44  fawzi

	* [r959] qs_blacs:
	  - renamed modules cp_full_matrix -> cp_fm_types,
	  cp_f_matrix_struct ->
	  cp_fm_struct, cp_fm_pool -> cp_fm_pool_types
	  - separated cholesky into cp_fm_cholesky
	  - cp_fm_triangular_multiply
	  qs_p_types: extended for non orthogonal orbitals, used pools
	  (compiled not tested)
	  minor fixes and cleanup.

2002-11-04 13:43  tchassai

	* [r958] fixed flags

2002-10-31 18:26  fawzi

	* [r957] corrected bug (use of uninitialized is_master).

2002-10-31 17:58  fawzi

	* [r956] Undo of stupid namechange (rho_r -> rho_elec_r) I didn't
	  mean to chack in.

2002-10-31 17:47  fawzi

	* [r955] fixed a bug in qs_rho_did_change.

2002-10-31 17:34  fawzi

	* [r954] - converted ks matrix calculation to pools
	  - removed most temp object from ks_env
	  - prepared for LSD
	  - changed initialization/destruction point of ks_env

2002-10-28 09:04  tchassai

	* [r953] More perturbation stuff.

2002-10-25 22:04  cjmundy

	* [r952] PGI compiler fixes. Wierd stuff

2002-10-25 13:41  hforbert

	* [r948] Author: hforbert
	  Added path integral test case, fixed minimal support of path
	  integrals in
	  code (part was missing from my side, part was broken by other
	  people :) )

2002-10-23 17:23  krack

	* [r947] prelim. version of a block Jacobi pseudo diagonalisation
	  controlled by eps_jacobi

2002-10-23 17:19  krack

	* [r946] initial print* substituted by a more useful write

2002-10-23 15:15  gtb

	* [r926] initial version

2002-10-23 15:13  gtb

	* [r925] added md_kg_energies

2002-10-23 15:10  gtb

	* [r924] adapted for kgmd

2002-10-23 15:06  gtb

	* [r923] kim-gordon md now acrive

2002-10-23 15:04  gtb

	* [r922] adapted to new xc functionals

2002-10-23 15:02  gtb

	* [r921] adapted to new xc functionasl

2002-10-23 15:00  gtb

	* [r920] minor change

2002-10-23 14:58  gtb

	* [r919] adapted for kim-gordon md

2002-10-23 13:28  tchassai

	* [r918] Some more tddfpt stuff. Still not working.

2002-10-18 10:15  jgh

	* [r917] Correct last commit

2002-10-18 08:42  jgh

	* [r916] Activate new functional codes for Quickstep.
	  Only spin unpolarized works. Not all functionals are tested.
	  qs_rho_did_change "hack" for SGI needed.

2002-10-16 15:31  fawzi

	* [r915] fixed a bug (nullified pointers) in pw_env_rebuild.

2002-10-16 13:42  vondele

	* [r914] added USE calculate_rho_core

2002-10-16 13:19  fawzi

	* [r911] added pw_env and qs_charges.

2002-10-16 13:18  fawzi

	* [r910] use rho%rho_ao in place of p_old.

2002-10-16 13:16  fawzi

	* [r909] corrected a couple of small bugs.CVS:
	  ----------------------------------------------------------------------

2002-10-16 13:13  fawzi

	* [r908] added call to update qs_env%rho.

2002-10-16 13:11  fawzi

	* [r907] rebuild pw_pools => rebuild pw_env, undo of double bug
	  fix.CVS:
	  ----------------------------------------------------------------------

2002-10-16 13:07  fawzi

	* [r906] uses rho%rho_ao as p_old.

2002-10-16 13:06  fawzi

	* [r905] added qs_charges, rho_core, pw_env, moved pw_pools to
	  pw_env.

2002-10-16 13:04  fawzi

	* [r904] added methods to rebuild or update pw_env, rho, rho_core.

2002-10-16 13:02  fawzi

	* [r903] moved from globenv to cp_error based logging

2002-10-16 13:00  fawzi

	* [r902] added type for charges on the grids.

2002-10-16 12:56  fawzi

	* [r901] renamed pw_pool module to pw_pool_types

2002-10-16 12:55  fawzi

	* [r900] added pw environement.

2002-10-16 12:51  fawzi

	* [r899] chaged logging from globenv to error, upcasing.

2002-10-16 12:48  fawzi

	* [r898] bcast of global unitnr

2002-10-16 12:11  fawzi

	* [r897] cosmetic changes.

2002-10-15 07:02  tchassai

	* [r896] Changed the calls to the intrinsic function size to the
	  standard syntax
	  to help with compiling a debug version. To get the size of the
	  second
	  dimension of a 3 dimensional array use size(array,2) instead of
	  size(array(1,:,1)).

2002-10-14 16:06  gtb

	* [r888] initial revision: rspace energy/forces of kg core charges

2002-10-14 16:05  gtb

	* [r887] initial revision: kg debug routine

2002-10-14 16:03  gtb

	* [r886] initial revision: rspace densitities for kg

2002-10-14 16:00  gtb

	* [r885] initial revision: kg force drive

2002-10-14 15:58  gtb

	* [r884] initial revision: kg numerical forces (debugging)

2002-10-14 15:57  gtb

	* [r883] initial revision: force types for kim-gordon

2002-10-14 15:55  gtb

	* [r882] temporary interface for kg tf+xc calculation

2002-10-14 15:52  gtb

	* [r881] kg is activated

2002-10-14 15:51  gtb

	* [r880] debug is now working, md not yet completed

2002-10-14 15:49  gtb

	* [r879] changed some variables names

2002-10-14 15:46  gtb

	* [r878] kg is activated

2002-10-14 15:45  gtb

	* [r877] some trivial changes, i.e.
	  CALL set_force_env ( fist_env_type = fist_env )

2002-10-14 15:42  gtb

	* [r876] added some new KG modules

2002-10-11 11:36  vondele

	* [r875] detabification

2002-10-10 16:06  fawzi

	* [r874] corrected release point of scf_control (thanks Thomas)

2002-10-10 14:16  krack

	* [r873] r4 and i4 added

2002-10-10 13:44  tchassai

	* [r872] Left out in previous commit.

2002-10-10 13:43  tchassai

	* [r871] Full skeleton for time-dependent dfpt now stands.

2002-10-10 13:37  tchassai

	* [r870] Fixed crash on SUN.

2002-10-10 09:27  fawzi

	* [r868] * activated pools
	  * changed calls that used a pointer returning function as
	  argument (mostly logging and par_env), not to use it directly.
	  * minor cleanups of logging
	  * normalized usestatements and upcased some more files.

2002-10-09 09:37  jgh

	* [r866] Variable ks_env%initialized has to be initialized before
	  the first
	  call to qs_ks_init!

2002-10-07 07:13  vondele

	* [r862] Memory now doing OK

2002-10-07 07:12  vondele

	* [r861] corrected format specifier

2002-10-07 06:47  vondele

	* [r860] more ot

2002-10-06 17:55  jgh

	* [r859] check early for rho > eps

2002-10-06 17:55  jgh

	* [r858] print memory activated

2002-10-03 15:30  krack

	* [r857] Memory leak removed

2002-10-03 15:01  krack

	* [r856] repeated QS wavefunction opt. feasible without memory
	  leaks (at least with AIX)

2002-10-03 14:58  krack

	* [r855] r_skin introduced and update_* changed to rebuild_*

2002-10-02 13:20  hforbert

	* [r853] Author: hforbert
	  
	  redefined some arguments in linklists.F to be pointers to arrays
	  instead
	  of arrays, as that's what we call these functions with and this
	  can
	  produce runtime errors on some machines if the pointers are null.
	  Also fixed a small typo in format string of constraint output in
	  force_fields.F (this produced binary garbage in the output).

2002-09-30 21:38  cjmundy

	* [r852] Removed -e in echo

2002-09-30 15:36  krack

	* [r851] sec added

2002-09-27 12:07  krack

	* [r850] Substituted by machine.F

2002-09-27 12:02  krack

	* [r849] - timesl substituted by machine
	  - Optional output of checkpoint information in QS branch

2002-09-27 11:58  krack

	* [r848] print_memory activated

2002-09-27 11:57  krack

	* [r847] Output checkpoint information

2002-09-27 09:06  krack

	* [r846] npt%v changed to npt(:,:)%v due to IBM compiler problems

2002-09-26 22:42  cjmundy

	* [r845] A little cleaning

2002-09-26 18:46  gtb

	* [r844] cleaning (rep_env, ewald_param and pimd removed)

2002-09-26 18:44  gtb

	* [r843] restructuring (rep_env removed)

2002-09-26 18:33  cjmundy

	* [r842] Moved pstat allocation to atoms_input.F

2002-09-26 10:08  krack

	* [r841] obsolete

2002-09-26 10:03  krack

	* [r840] kg_environment_type.o changed to kg_environment_types.o

2002-09-26 10:01  krack

	* [r839] Name of module and file are now equal to avoid continous
	  recompilation of kg_main.F

2002-09-25 15:46  krack

	* [r838] -e flag added to echo command for Linux

2002-09-25 15:17  krack

	* [r837] Display of memory usage (try keyword MEMORY in section
	  &PRINT also for cp2k.popt)

2002-09-25 12:22  vondele

	* [r836] coredumping tdddft removed

2002-09-25 12:22  vondele

	* [r835] more OT

2002-09-25 09:41  gtb

	* [r833] adapted for KG use

2002-09-25 09:41  gtb

	* [r832] globenv passed in CALL
	  init_atomic_kind_set(atomic_kind_set,globenv)

2002-09-25 09:38  gtb

	* [r831] added Kim-Gordon option

2002-09-25 09:37  gtb

	* [r830] kg_control_type added

2002-09-25 09:36  gtb

	* [r829] adapted for Pol/KG Use

2002-09-25 09:35  gtb

	* [r828] adapted for Pol/Kim-Gordon use

2002-09-25 09:34  gtb

	* [r827] adapted for Kim-Gordon/Pol use

2002-09-25 09:33  gtb

	* [r826] added kg_main and kg_environment_type

2002-09-25 09:32  gtb

	* [r825] initial version (not complete)

2002-09-25 07:27  krack

	* [r824] print_memory updated and activated for memory leak search

2002-09-25 07:25  krack

	* [r823] Memory leak removed.

2002-09-24 14:54  vondele

	* [r822] more OT

2002-09-24 07:58  gtb

	* [r821] fixed minor bug, line 871 (md_env % itimes)

2002-09-24 04:32  cjmundy

	* [r820] A little clean-up.

2002-09-23 14:59  krack

	* [r819] Direct usage of cp_para_types, some redundant USE
	  statements removed

2002-09-23 14:56  krack

	* [r818] Full QS neighbor list update (new) and accelerated build
	  (GTH-PPNL)

2002-09-23 13:19  gtb

	* [r817] moved allocation/initialization of part pointer

2002-09-23 13:18  gtb

	* [r816] added atomic_kind_set to replica_environment_type

2002-09-23 13:17  gtb

	* [r815] 1)pointer part is allocated and initialized here
	  2)atom coordinates can also be read from &COORD section in input

2002-09-22 03:15  cjmundy

	* [r807] fixed DEC compiler bugs

2002-09-22 01:13  cjmundy

	* [r806] Fixed old typo

2002-09-22 01:12  cjmundy

	* [r805] Clean-up for future FIST/QS merge

2002-09-22 00:44  cjmundy

	* [r804] Slimmed done the code to FIST and QS

2002-09-22 00:43  cjmundy

	* [r803] Clean-up for the future QS/FIST merge

2002-09-22 00:24  cjmundy

	* [r802] Clean-up for future QS/FIST merge

2002-09-19 12:36  vondele

	* [r799] restructuring of OT files

2002-09-19 12:36  vondele

	* [r798] flexible timings

2002-09-19 07:56  tchassai

	* [r797] Minor stuff.

2002-09-17 10:43  krack

	* [r795] GTH-PPNL neighbor list generation accelerated

2002-09-17 08:16  tchassai

	* [r793] - Added arpack routines for the non-symmetric standard (no
	  B-matrix)
	  eigenvalue problem.
	  - Added skeleton for tddfpt eigenvalue calculation
	  (implementation of
	  the operator still needs some work.

2002-09-16 15:23  vondele

	* [r792] some more cleaning. Can we get qs_scf below 500 lines ?

2002-09-16 12:37  vondele

	* [r791] YACC : yet another compiler correction

2002-09-16 08:53  vondele

	* [r790] improved format on ouput

2002-09-14 13:23  vondele

	* [r789] adapted to present day computers :-)

2002-09-14 13:14  vondele

	* [r788] fixed various compiler problems and bugs

2002-09-14 01:34  fawzi

	* [r787] added fm_pool, linked_list_fm, scf_control.

2002-09-14 01:33  fawzi

	* [r786] moved input parameters to scf_control_type.

2002-09-14 01:31  fawzi

	* [r785] created a separated scf_control_type (got rid of many
	  module variables in qs_scf).

2002-09-14 01:30  fawzi

	* [r784] brought back most basic methods for the rho type.

2002-09-14 01:29  fawzi

	* [r783] upcased keywords.

2002-09-14 01:27  fawzi

	* [r782] created scf_control, began to plan for pools
	  initialization.

2002-09-14 01:25  fawzi

	* [r781] addet scf_control, para_env, pw_pools and mo_ao_fm_pool

2002-09-14 01:24  fawzi

	* [r780] created scf_control type, smaller fixes.

2002-09-14 01:23  fawzi

	* [r779] moved optimal block size to cp_f_matrix_struct, smaller
	  cosmetic fixes.

2002-09-14 01:21  fawzi

	* [r778] added flush cache, various interface cleanups, smaller
	  fixes.

2002-09-14 01:20  fawzi

	* [r777] updated logger attributes description.

2002-09-14 01:19  fawzi

	* [r776] added new full matrix linked list, smaller fixes.

2002-09-14 01:17  fawzi

	* [r775] moved optimal block sizes to cp_f_matrix_struct, added
	  cp_fm_write.

2002-09-14 01:15  fawzi

	* [r774] mofed optimal_blacs_block_size to cp_f_matrix_struct.

2002-09-13 17:17  fawzi

	* [r773] added pool for full matrixes.

2002-09-13 17:16  fawzi

	* [r772] added optimal blocksize (moved from full_matrix), smaller
	  changes.

2002-09-13 10:58  vondele

	* [r771] small changes

2002-09-13 10:58  vondele

	* [r770] kinetic energy preconditioner

2002-09-12 13:12  vondele

	* [r769] bug fix

2002-09-12 09:43  tchassai

	* [r768] small changes

2002-09-12 09:34  tchassai

	* [r767] new setup subroutine

2002-09-12 09:32  tchassai

	* [r766] small fix

2002-09-12 09:10  vondele

	* [r765] more preconditioner + small changes

2002-09-12 09:10  vondele

	* [r764] extended preconditioner story

2002-09-12 09:09  vondele

	* [r763] improved trace debug (! try it)

2002-09-12 08:53  tchassai

	* [r762] Quick fix of the arpack stuff and changed qs_scf back so
	  it doesn't
	  require arpack.

2002-09-12 07:00  tchassai

	* [r761] Splitted qs_arpack.F into several files to make it more
	  flexible and
	  corrected routine names.

2002-09-11 12:31  krack

	* [r759] neighbor list rebuild

2002-09-11 12:07  krack

	* [r758] Bug fix in finish_blacs

2002-09-11 06:41  tchassai

	* [r757] n_spins now extracted from qs_env%dft_control%nspins.

2002-09-10 15:28  krack

	* [r756] QS overlap neighbor list update (sab_orb)

2002-09-10 13:00  tchassai

	* [r755] Removed variable n_spins, as it was already there (in
	  dft_control).

2002-09-09 11:09  tchassai

	* [r754] Corrections. Module doesn't crash anymore.

2002-09-09 05:55  vondele

	* [r753] added a method to get the pointer to the local data

2002-09-09 05:55  vondele

	* [r752] improved cholesky restore function, some small changes

2002-09-09 05:54  vondele

	* [r751] added functionality

2002-09-09 05:53  vondele

	* [r750] more preconditioner

2002-09-07 15:11  vondele

	* [r749] more work on preconditioner

2002-09-06 19:19  fawzi

	* [r748] corrected syntax errors.

2002-09-06 16:13  krack

	* [r747] Bug fix in cp_fm_maxval

2002-09-06 11:06  vondele

	* [r743] preconditioner for ot

2002-09-05 17:33  fawzi

	* [r742] first split up of blacs, not yet finished.

2002-09-05 17:30  fawzi

	* [r740] began renaming of blacs and splitting in more modules. see
	  cp2k/tools/renamer.py to help the conversion of older files.
	  smaller fixes.
	  pw_pool completed.

2002-09-05 15:09  tchassai

	* [r739] - Added n_spins variable to qs_environment_type. This
	  variable
	  is pre-initialized to 1 in set_qs_env. Alternatively a value
	  (preferably 1 or 2) can be specified. It has no impact on
	  the interfaces as they are used now (n_spins is always optional).
	  - More time-dependent perturbation stuff. Doesn't work yet.

2002-09-05 15:06  krack

	* [r738] Bug fix for PBC atoms at -l/2

2002-09-05 06:11  vondele

	* [r736] added timing

2002-09-04 11:48  tchassai

	* [r735] Added temporary matrizes. Should eventually be replaced by
	  objects
	  from a pool.

2002-09-04 09:24  fawzi

	* [r734] always alloc %pw.

2002-09-03 19:12  fawzi

	* [r733] renamed pw_grid_pool to pw_pool, pool implementation
	  finished, compiled, not tested.

2002-09-03 19:11  fawzi

	* [r732] nullifies also the %mass_c* upon initialization.

2002-09-03 19:10  fawzi

	* [r731] make %pw a pointer to better use the pw_pool.

2002-09-03 09:22  tchassai

	* [r729] added qs_p_env_dealloc_ref and qs_p_env_psi0_changed

2002-09-02 17:38  fawzi

	* [r727] upcased keywords, but CP* macros in one line.

2002-09-02 15:23  tchassai

	* [r726] Switched to the new qs_p_env_type. Doesn't work yet.

2002-09-02 09:33  tchassai

	* [r725] Added a routine to make it compile on the SUN.

2002-09-02 09:20  fawzi

	* [r724] added pw linked list and (begin of) pw pool

2002-09-02 09:19  fawzi

	* [r723] upcased keywords, smaller changes.

2002-09-02 09:00  fawzi

	* [r722] began definition of the new perturbation structure.

2002-09-02 08:59  fawzi

	* [r721] upcased keywords, smaller changes.

2002-09-02 08:54  fawzi

	* [r720] upcased keywords.

2002-09-02 08:51  fawzi

	* [r719] added write routine, id_nr and pw_p_type.

2002-08-30 06:59  jgh

	* [r718] Multi-grids independent from density and potential
	  store old density in g space

2002-08-30 06:58  jgh

	* [r717] revised input for QS section

2002-08-30 06:52  jgh

	* [r716] Many changes to format specifiers, now compiles on SGI
	  Added some use statements

2002-08-29 15:59  fawzi

	* [r715] log and error handling: removed generic functions, smaller
	  changes

2002-08-29 07:25  vondele

	* [r714] X->1X,

2002-08-28 06:45  vondele

	* [r713] Added clean(er) HOMO/LUMO/E_DENSITY/V_HARTREE capabilities

2002-08-27 14:01  vondele

	* [r712] small changes (bugs/speed)

2002-08-27 11:49  vondele

	* [r710] bug fix for nrow_local, ncol_local

2002-08-27 07:25  vondele

	* [r709] added unoccupied subspace calculation

2002-08-26 20:23  fawzi

	* [r708] corrected some smaller bugs, updated synopsis

2002-08-26 20:22  fawzi

	* [r707] updated synopsis

2002-08-26 20:17  fawzi

	* [r706] updated synopsis.

2002-08-26 12:18  tchassai

	* [r705] New files for perturbation theory. Doesn't work yet.

2002-08-25 13:31  vondele

	* [r703] rs_grid & pw -> cube file

2002-08-25 13:30  vondele

	* [r702] added cube files for density / hartree / homo

2002-08-25 13:29  vondele

	* [r701] added routines for eigenvalues of a given subspace

2002-08-25 13:29  vondele

	* [r700] new routine for adding a wavefunction to the grid

2002-08-25 13:27  vondele

	* [r699] dscal -> dcopy for safe zero

2002-08-25 08:26  vondele

	* [r698] temporary bug fix

2002-08-25 07:45  vondele

	* [r697] small bug fix

2002-08-25 07:19  vondele

	* [r696] bug fix for uneducated compilers

2002-08-24 18:06  fawzi

	* [r695] new logging in cp_log_handling... (parallel aware), not
	  yet fully finished.

2002-08-22 22:01  fawzi

	* [r694] changes to blacs, matrix structure, reference counting,
	  removed globenv, normalized blacs env.
	  Renaming and splitting of blacs routines not yet done.
	  Tested single and parallel versions.

2002-08-21 14:58  tchassai

	* [r693] Added hooks for tddfpt

2002-08-21 14:44  tchassai

	* [r692] Added qs_tddfpt_module.o

2002-08-21 14:41  tchassai

	* [r691] Just started work

2002-08-21 14:35  tchassai

	* [r690] Changes for tddfpt

2002-08-14 11:17  fawzi

	* [r688] renamed set_rho_ao

2002-08-14 10:54  fawzi

	* [r687] Added qs_rho_types.F, structure to hold the different
	  representations of the density.

2002-08-14 09:40  fawzi

	* [r686] underlying real_matrix (sparse matrix or sm) of
	  cp_block_matrix renamed from matrix to sm.

2002-08-12 15:14  jgh

	* [r685] small changes

2002-08-12 14:01  fawzi

	* [r684] introduced the cp_para_env type to store the parallel
	  informations (replaces cp_mpi_env).

2002-08-11 16:48  jgh

	* [r683] Minor changes

2002-08-09 17:43  fawzi

	* [r682] many changes, now most of the code is there, and the
	  structure should be more stable.

2002-08-09 17:42  fawzi

	* [r681] added blacs_triangular_multiply (still unimplemented).

2002-08-09 17:41  fawzi

	* [r680] splitted lines with CPSourceFileRef, so that it remains
	  alone on one line.

2002-08-09 14:02  fawzi

	* [r679] corrected bug in sparse_times_blacs, optimized
	  blacs_scale_and_d for scale==0.0.

2002-08-04 09:10  fawzi

	* [r678] better mpi_env concept in cp_block_matrix and
	  cp_b_matrix_structure-

2002-08-04 09:02  fawzi

	* [r677] added blacs_scale_and_d (rescaling and digonal update).

2002-08-04 09:00  fawzi

	* [r676] added s_struct and update_s_struct.

2002-08-04 08:59  fawzi

	* [r675] removed rhs, added setup routine p_env_psi0_changed.

2002-08-04 08:56  fawzi

	* [r674] parentized format (for Absoft-macOSX)

2002-07-31 10:08  fawzi

	* [r673] added preo and post ortho.

2002-07-31 10:07  fawzi

	* [r672] cp_sp_dealloc => cp_sp_release
	  commented out untested paralleization.

2002-07-31 10:04  fawzi

	* [r671] cp_sp_dealloc => cp_sp_release

2002-07-31 10:03  fawzi

	* [r670] made cp_sp_dealloc private (you should use cp_sp_release)

2002-07-31 08:23  fawzi

	* [r669] first add of a lightweight mpi enironment.

2002-07-30 19:13  fawzi

	* [r667] cp_sparse_matrix => cp_block_matrix

2002-07-30 19:10  fawzi

	* [r666] renamed cp_sparse_matrix to cp_block_matrix, introduction
	  of mpi_env, various changes.

2002-07-30 19:08  fawzi

	* [r665] introduction of cp_mpi_env, consequent usage of sm for
	  sparse (real_) matrix methods.

2002-07-30 19:05  fawzi

	* [r664] added to/from array, add ordered and contains methods.

2002-07-30 19:02  fawzi

	* [r663] added silent automatic restart if task begins with
	  restart.

2002-07-30 18:58  fawzi

	* [r662] removed f>f_old from covergence criteria.

2002-07-30 18:56  fawzi

	* [r661] added one small check.

2002-07-30 18:55  fawzi

	* [r660] added cp_mpi_env, qs_p_env, qs_p_build_kernel

2002-07-30 18:52  fawzi

	* [r659] introduction of mpi_env, cp_sparse_matrix =>
	  cp_block_matrix
	  guarantee_owned_NUi for parallelization.

2002-07-30 18:50  fawzi

	* [r658] introduction of mpi_env
	  cp_sparse_matrix => cp_block_matrix.

2002-07-30 18:44  fawzi

	* [r657] glob_env => mpi_env (in some places)
	  cp_sparse_matrix => cp_block_matrix
	  stupid parallelization of atomic_ortho.

2002-07-30 18:41  fawzi

	* [r656] cp_sparse_matrix => cp_block_matrix.

2002-07-30 18:40  fawzi

	* [r655] cp_sparse_matrix => cp_block_matrix

2002-07-30 18:39  fawzi

	* [r654] cosmetic change to a comment.

2002-07-30 18:37  fawzi

	* [r653] added methods to allocate/deallocate vectors of blacs
	  matrixes.
	  extended sparse_times_blacs: nonsquare non symm sparse,
	  alpha,beta scaling, use of mpi_env.
	  methods to get and set a submatrix of a blacs matrix (inspired by
	  blacs_replicated_copy).

2002-07-30 18:32  fawzi

	* [r652] skeleton of the code to build the second order
	  perturbation matrix.

2002-07-30 18:31  fawzi

	* [r651] skeleton the code for the superenvironment of qs_env to do
	  perturbation.

2002-07-30 18:29  fawzi

	* [r650] glob_env => mpi_env changes, pao now updates p_new

2002-07-30 18:26  fawzi

	* [r649] sparse_times_replicated now uses mpi_env in place of
	  glob_env, added optional block_owner.

2002-07-23 07:02  vondele

	* [r648] added lots of (1,1)

2002-07-22 08:43  vondele

	* [r646] grid core routines added

2002-07-22 08:42  vondele

	* [r645] made the core routines F77 style and put them in lib/
	  notice the F77 style of passing the pointers (i.e. grid(1,1,1)
	  instead of just grid)

2002-07-19 13:58  krack

	* [r644] Strange write statement removed

2002-07-19 13:56  krack

	* [r643] Missing initialization added

2002-07-19 11:52  krack

	* [r642] Missing TARGET attribute added

2002-07-19 09:08  jgh

	* [r641] Bug fix for 3d FFT parallel: case with processors without
	  work

2002-07-19 08:16  krack

	* [r640] PGI hack also used for AIX (xlf 8.1)

2002-07-19 07:59  krack

	* [r639] bug fix forces

2002-07-18 14:26  krack

	* [r638] reallocate changed from FUNCTION to SUBROUTINE, reallocate
	  usage reduced

2002-07-18 11:43  vondele

	* [r637] generalised w_matrix, forces with OT

2002-07-17 15:44  vondele

	* [r636] fixed pbc/unroll bug

2002-07-17 13:25  krack

	* [r635] level_shift /= 0 implies cholesky_off

2002-07-17 11:11  krack

	* [r634] mo_set -> sparse p,w matrix works with smear and forces

2002-07-17 09:36  vondele

	* [r633] removed tabs and double declarations

2002-07-17 08:33  vondele

	* [r632] reduced memory usage for OT

2002-07-17 08:32  vondele

	* [r631] added sparse matrix DM for p_old + OT stuff

2002-07-16 14:07  krack

	* [r629] force calc. parallelized

2002-07-16 07:13  vondele

	* [r628] bug fix : NULLIFY pointer

2002-07-15 14:50  vondele

	* [r627] added magnitude analysis of s and p

2002-07-15 14:49  vondele

	* [r626] new definition of delta

2002-07-15 13:57  vondele

	* [r625] identical numbers for nrebuild in parallel

2002-07-15 06:43  vondele

	* [r624] OT modifications

2002-07-12 15:10  jgh

	* [r623] Bug fix for mpi groups: the PW grid group was used instead
	  of the real space
	  group (of the plane waves) = rs%group

2002-07-12 07:44  vondele

	* [r622] bug fix (?) in work allocation

2002-07-09 15:56  vondele

	* [r621] MODIFIED QS RESTART. Single file, independent of # cpus
	  preliminary : should be able to use a different basis set, and
	  still give,
	  from time to time, a resonable restart (i.e. DZV -> DZVP restart)

2002-07-07 12:13  vondele

	* [r620] minor changes

2002-07-06 10:52  vondele

	* [r619] added sparse density matrix construction from blacs mo

2002-07-06 08:02  vondele

	* [r618] added sparse matrix times blacs matrix

2002-07-05 15:22  fawzi

	* [r617] added remove_block_node

2002-07-05 15:17  fawzi

	* [r616] added test for next on non-associated block

2002-07-05 14:37  tchassai

	* [r614] Changed add_real_matrix_block to add the new matrix block
	  ordered wrt the block column. The matrix blocks in will now be
	  ordered with ascending column numbers.

2002-07-05 14:31  fawzi

	* [r613] added abs() to delta f stopping test (i.e. does not stop
	  if f_old< f).

2002-07-02 11:08  gtb

	* [r612] first version: full calculation of the first-order kernel
	  K1
	  (it replaces pol_thk_pw)

2002-07-02 11:06  gtb

	* [r611] added routines to calculate numer forces due to the new
	  thk term

2002-07-02 11:05  gtb

	* [r610] CALL force_thk is replaced by CALL force_fo_kernel;
	  rho0 is allocated/deallocated in this module

2002-07-02 11:03  gtb

	* [r609] replaced calls to part_thk_numer WITH
	  part_fo_kernek_numer,
	  the same is done for coefs.

2002-07-02 11:01  gtb

	* [r608] rho0 is now passed in i/o

2002-07-02 10:59  gtb

	* [r607] added object pol_fo_kernel_pw.o

2002-07-01 09:47  vondele

	* [r606] added ot more cleanly

2002-07-01 09:10  vondele

	* [r605] Example commit

2002-06-27 12:37  fawzi

	* [r604] fixed some bugs with more than one atom, sucessfully
	  tested on H2O.

2002-06-27 12:35  fawzi

	* [r603] logging also atomic ortho.

2002-06-27 07:14  fawzi

	* [r602] updated synopsis.

2002-06-26 17:03  fawzi

	* [r600] many changes and bug fixes - now single atom (Ar) works,
	  H2O still has problems.

2002-06-26 17:00  fawzi

	* [r599] pao: outputs pao angles, transf_p_to_f in place of
	  transf_to_f (because of atomic ortho).

2002-06-25 07:51  fawzi

	* [r598] moved charge to dft_control, renamed s_changed to
	  s_struct_changed.

2002-06-24 16:34  fawzi

	* [r597] removed natoms from qs_env.
	  atomic ortho implementation completed, not tested.

2002-06-24 16:28  fawzi

	* [r596] made sum of matrices add_real_matrixes faster if alpha and
	  beta are 1.0

2002-06-24 16:26  fawzi

	* [r595] adapted to new build KS matrix.

2002-06-24 15:11  fawzi

	* [r594] removed natoms and n_electrons (as requested by krack)

2002-06-24 15:09  fawzi

	* [r593] added charge to dft_types.

2002-06-24 13:25  tchassai

	* [r592] atom_parameters_type.F is no longer needed due to rewrite
	  of atom_module.F

2002-06-24 13:24  tchassai

	* [r591] removed atom_parameters_type.F from the list

2002-06-24 13:22  tchassai

	* [r590] atom_type.F is no longer needed due to rewrite of
	  atom_module.F

2002-06-24 13:22  tchassai

	* [r589] atom_calculations.F is no longer needed due to rewrite of
	  atom_module.F

2002-06-24 13:20  tchassai

	* [r588] atom_types.F replaces atom_type.F

2002-06-24 13:19  tchassai

	* [r587] atom_util.F replaces atom_calculations.F

2002-06-21 13:30  tchassai

	* [r586] added radial_integrate

2002-06-20 18:11  fawzi

	* [r585] Introduction of atomic orthogonalization almost complete.

2002-06-20 18:10  fawzi

	* [r584] cp_release now works even with disassociated pointers
	  (does nothing).

2002-06-20 15:13  vondele

	* [r583] added cg to ot

2002-06-20 08:11  fawzi

	* [r576] splitted //CPSourceFileRef on two lines (some c
	  preprocessors don't make substitutions in C++ style comments)

2002-06-20 08:09  fawzi

	* [r575] more matrix structures.
	  began introduction of atomic orthogonalization.
	  splitted //CPSourceRef in two lines.

2002-06-20 06:17  vondele

	* [r574] removed multiple declarations

2002-06-19 08:34  fawzi

	* [r573] fixed more bugs ( QS P is twice pao P)

2002-06-19 08:32  fawzi

	* [r572] fixed bug in sm_transfer_to_f

2002-06-19 08:30  fawzi

	* [r571] full matrix m_struct now not symmetric

2002-06-19 08:27  fawzi

	* [r570] cosmetic changes in add_blacs_to_block_diag_sm

2002-06-19 08:24  fawzi

	* [r569] fixed smaller bugs in pao

2002-06-19 08:22  fawzi

	* [r568] flushes every log.
	  slightly higher precision in real (double) to string
	  transformation.

2002-06-19 08:20  fawzi

	* [r567] step really stops after each iteration.
	  tries autorestart after errors

2002-06-19 08:07  fawzi

	* [r566] higher default output precision for reals

2002-06-18 08:39  vondele

	* [r564] putting ot in scf

2002-06-18 02:27  cjmundy

	* [r563] EWALD/PME option with POL

2002-06-14 17:05  fawzi

	* [r562] some cosmetic and bugfixes, removed context from blacs
	  calls.

2002-06-14 17:01  fawzi

	* [r561] added more debugging info.

2002-06-14 17:00  fawzi

	* [r560] nullified cexp_ppl(i)%c (for pgi compiler)

2002-06-14 12:46  vondele

	* [r559] cosmetics

2002-06-14 10:06  vondele

	* [r558] added qs_ot.F, initial orbital transformation package

2002-06-14 10:05  vondele

	* [r557] minor modifications for context

2002-06-13 18:43  krack

	* [r555] icol_global declared

2002-06-13 18:38  krack

	* [r554] IMPLICIT NONE moved

2002-06-13 18:18  krack

	* [r553] args of build_core_hamiltonian changed

2002-06-13 17:27  krack

	* [r552] a) context included in globenv
	  b) force driver
	  c) m_flush added (new machine routine)
	  d) parallel_include changed to free format
	  e) timestamp for exe-file
	  f) makefile update due to d) and e)

2002-06-13 14:06  vondele

	* [r551] :wq!xxxdds/xxx/yy/g

2002-06-12 17:43  fawzi

	* [r550] added "none" to the symmetries of a matrix (in
	  write_sparse_matrix)

2002-06-12 17:42  fawzi

	* [r549] did some serious testing, fixed bugs in
	  pao_l_angles_methods, now seems bug free.

2002-06-12 17:38  fawzi

	* [r548] reactivated get_qs_env (now s is an array of matrices).

2002-06-12 17:35  fawzi

	* [r547] reactivated get_qs_env (now s is an array of matrices).
	  better debug output using cp_sm_output.

2002-06-12 17:33  fawzi

	* [r546] corrected indexing of the angles

2002-06-12 17:32  fawzi

	* [r545] initialization of angles to 0.0 at the beginning

2002-06-12 17:31  fawzi

	* [r544] removed p from initialization (not yet initialized in
	  qs_env)

2002-06-12 17:30  fawzi

	* [r543] downcased the matrix symmetry string.

2002-06-12 17:28  fawzi

	* [r542] added function to output a sparse matrix to a file.

2002-06-12 17:27  fawzi

	* [r541] corrected sparsity allocation bug.

2002-06-12 09:27  vondele

	* [r540] small bug fix

2002-06-12 08:15  krack

	* [r539] allocate/deallocate status check

2002-06-11 18:01  gtb

	* [r538] allows the calculation of distinct e_ovl_ii and e_ovl_ij
	  terms

2002-06-11 17:56  gtb

	* [r537] calculated distinct e_ovlp_ii and e_ovlp_ij terms

2002-06-11 17:55  gtb

	* [r536] e_ovlp_ii and e_ovlp_ij terms calculated

2002-06-11 17:54  gtb

	* [r535] e_ovl_ii and e_ovlp_ij terms calculated

2002-06-11 17:53  gtb

	* [r534] e_ovlp_ii and e_ovlp_ij are calculated

2002-06-11 17:52  gtb

	* [r533] added e_pol, e_overlap_ii, e_overlap_ij in
	  thermodynamic_type

2002-06-11 17:51  gtb

	* [r532] calculated distinct e_ovlp_ii and e_ovl_ji contributions

2002-06-11 17:50  gtb

	* [r531] calculated distinct e_ovlp_ii and e_ovlp_ij terms

2002-06-11 14:19  vondele

	* [r530] SUN problem with pointers : fix it with intermediate
	  pointer

2002-06-11 14:01  krack

	* [r529] Bug fix: Missing pointer initialisation

2002-06-11 10:19  krack

	* [r528] TARGET attribute added for pw

2002-06-10 18:06  gtb

	* [r526] initial version

2002-06-10 18:05  gtb

	* [r525] added module fit_pair_potential for fit code

2002-06-10 18:01  gtb

	* [r524] new fit code

2002-06-10 18:00  gtb

	* [r523] deleted calls to old fit_module

2002-06-10 17:57  gtb

	* [r522] fit_module is now distinct from control_module

2002-06-10 17:55  gtb

	* [r521] modified mininfo type

2002-06-10 17:54  gtb

	* [r520] added routines needed for forces fit.

2002-06-10 17:53  gtb

	* [r519] some modification to print statments

2002-06-10 17:52  gtb

	* [r518] new fit code: works also for pol_pw

2002-06-10 17:50  gtb

	* [r517] new fit code

2002-06-10 17:49  gtb

	* [r516] main module for the new fit code.

2002-06-10 17:46  gtb

	* [r515] modified thermodinamic_type

2002-06-10 17:41  gtb

	* [r514] modified to calculate distinct electrostatics
	  contributions

2002-06-10 17:41  gtb

	* [r513] added get_pol_charge to calculate polarization charge

2002-06-10 17:39  gtb

	* [r512] modified to calculate distinct rho0rho0 and pol
	  electrostatic energies

2002-06-10 17:37  gtb

	* [r511] modified to calculate rho0-rho0 and pol electorstatic
	  contributions

2002-06-10 17:30  gtb

	* [r510] adapted for new fit code

2002-06-10 15:45  krack

	* [r509] Forces for GPW

2002-06-10 14:56  jgh

	* [r508] Rename of XC functionals
	  redo of input for XC functionals
	  QS calculation is hardwired to PADE for the moment

2002-06-10 14:45  jgh

	* [r507] Cosmetic changes, removed tabs, output formats and
	  copyright

2002-06-10 12:46  fawzi

	* [r506] more debugging, removed private from iter.

2002-06-10 10:13  fawzi

	* [r505] added much debug output.

2002-06-10 10:13  fawzi

	* [r504] corrected bug (size vs block size)

2002-06-10 10:12  fawzi

	* [r503] output to separate file, better 2d r output

2002-06-06 14:11  fawzi

	* [r494] renamed qs_build_KS_matrix to qs_build_ks_matrix to avoid
	  uneeded compilation.

2002-06-06 08:46  fawzi

	* [r491] moved the routines to build the KS matrix from qs_scf to
	  qs_build_KS_matrix.
	  compiled and tested.
	  still ugly.

2002-06-06 08:45  fawzi

	* [r490] added p_old and n_electrons

2002-06-06 08:43  fawzi

	* [r489] removed double declaration, added intent(in)

2002-06-06 08:40  fawzi

	* [r488] added cp_unimplemented_error

2002-06-06 08:38  fawzi

	* [r487] added qs_build_KS_matrix module.

2002-06-05 08:06  fawzi

	* [r486] used specific function for cp_guarantee_size

2002-06-04 15:04  vondele

	* [r485] added integrate_v_core_rspace

2002-06-04 15:04  vondele

	* [r484] added call to integrate_v_core_rspace

2002-06-03 16:18  cjmundy

	* [r483] more cleaning

2002-06-03 14:32  vondele

	* [r482] corrected copy & paste bugs

2002-06-03 09:55  vondele

	* [r481] removed test_it

2002-06-02 18:18  cjmundy

	* [r480] Global thermostat for POL bug fix

2002-06-02 17:32  cjmundy

	* [r479] Bug fix for DEC_OS.

2002-06-02 16:55  cjmundy

	* [r478] cleaned

2002-06-01 17:44  cjmundy

	* [r477] bug fixes for PW integrator

2002-05-31 14:35  vondele

	* [r476] added additional function
	  davidson diag

2002-05-30 11:20  gtb

	* [r475] for COMPLEXDATA3D replaced pw%cr3d with pw%cc3d (copy_rc,
	  copy_cr)

2002-05-29 17:31  gtb

	* [r474] mapping bug fixed

2002-05-29 09:24  fawzi

	* [r466] higher max number of f eval per iteration

2002-05-28 16:22  jgh

	* [r465] remove files

2002-05-28 16:15  fawzi

	* [r464] just fixed more bugs, compiles but does not work (yet).

2002-05-28 14:28  jgh

	* [r463] Change of file names (module name = file name)

2002-05-27 12:47  vondele

	* [r462] included arpack as diagonalisation routine,
	  reordered/merged some calls in qs_scf

2002-05-27 11:12  vondele

	* [r461] variables declared once : test_it: experiment

2002-05-26 16:47  vondele

	* [r460] dcopy

2002-05-26 16:29  vondele

	* [r459] smaller modifications

2002-05-26 12:21  jgh

	* [r458] added copy_rc routine: copy real to complex

2002-05-26 02:10  cjmundy

	* [r457] silly bug

2002-05-25 18:16  jgh

	* [r456] Calculate core density only once per SCF calculation

2002-05-25 18:15  jgh

	* [r455] removed tab

2002-05-25 17:25  cjmundy

	* [r454] clean-up the DEC bug-fix

2002-05-25 16:37  jgh

	* [r453] Optimised rs_pw_transfer, replaced mp_sum

2002-05-25 04:13  cjmundy

	* [r452] Fixed readin bug for DEC

2002-05-25 04:13  cjmundy

	* [r451] add thk contribution to particle force

2002-05-23 18:27  jgh

	* [r450] Activate parallel FFT's in Quickstep

2002-05-23 16:03  fawzi

	* [r449] updated domments (doc)

2002-05-23 15:21  fawzi

	* [r447] corrected various bugs, now a pao run works (but gives
	  wrong values)

2002-05-23 15:20  fawzi

	* [r446] corrected single-cpu version of add_blacs_to_block_diag_sm

2002-05-21 14:56  jgh

	* [r443] Clean up of PW FFT interfaces, preparing the parallel FFTs
	  for Quickstep

2002-05-21 13:16  vondele

	* [r442] Distributed grids -> 1) real space grid types introduced

2002-05-21 13:15  fawzi

	* [r441] changed cp_insert interface (gave problems with intel
	  compiler)

2002-05-21 13:14  fawzi

	* [r440] reduced line length of some comments.

2002-05-21 06:12  vondele

	* [r438] nloc declared once

2002-05-20 18:10  gtb

	* [r437] added a write statement (npoints pw grid)

2002-05-18 15:37  vondele

	* [r436] qs_arpack.o added

2002-05-18 15:35  vondele

	* [r434] Interface to (P)ARPACK : Lanczos diagonalization
	  (Paralllel)

2002-05-18 15:33  vondele

	* [r433] Added new routines for distributed vector x blacs matrix
	  multiplication
	  Needed for the (p)arpack routines

2002-05-16 15:52  gtb

	* [r432] bug fix: ao%grid_id replaced by ao%grid_type_id

2002-05-16 12:53  fawzi

	* [r429] added pao to the scf, single processor compiled, not
	  tested.

2002-05-16 12:52  fawzi

	* [r428] added pao.

2002-05-13 17:19  gtb

	* [r418] added call to new routine get_proj_scale_factors

2002-05-13 17:17  gtb

	* [r417] simpar passed to pol_force_pw

2002-05-13 17:17  gtb

	* [r416] added some logic for initialization of pw's

2002-05-13 17:15  gtb

	* [r415] bug fix + projector's scale factor

2002-05-13 17:09  gtb

	* [r414] projector's scale factors have been introduced

2002-05-13 17:08  gtb

	* [r413] projectors' scale factors are introduced

2002-05-13 17:07  gtb

	* [r412] passed simpar to pol_force_pw

2002-05-13 17:06  gtb

	* [r411] added option pol_init_type for pw's initialization

2002-05-13 17:05  gtb

	* [r410] added in ao_type a new pointer (norm) for scale factors

2002-05-13 17:03  gtb

	* [r409] simpar is passed to debug_control_pw

2002-05-13 17:01  gtb

	* [r408] initial version

2002-05-13 16:59  gtb

	* [r407] added module pol_projectors_scale

2002-05-08 18:10  fawzi

	* [r389] changes to compile with the intel compiler.

2002-05-08 18:09  fawzi

	* [r388] changes to compile with intel compiler.

2002-05-08 18:08  fawzi

	* [r387] changes to compile with intel.

2002-05-08 18:06  fawzi

	* [r386] changed intent(in) to intent(inout) to conform to the
	  interface of amoeba_evaluate.

2002-05-08 11:51  vondele

	* [r384] coded away dygvx since it's not in all lapack
	  distributions

2002-05-07 14:54  vondele

	* [r383] Added Cholesky based diagonalisation

2002-05-07 08:36  vondele

	* [r382] choose short expansions at comp

2002-05-07 07:46  fawzi

	* [r381] changed from target to pointer arg (target cannot be
	  null).

2002-05-07 07:44  fawzi

	* [r380] corrected bug (not checking present)

2002-05-07 07:41  fawzi

	* [r379] updated synopsis.

2002-05-06 15:02  gtb

	* [r378] 1/gsq preconditioning

2002-05-06 11:48  fawzi

	* [r377] renamed pao_projection to pao_bas_projection_types.

2002-05-06 11:34  fawzi

	* [r376] Initialization of projection in atomic_kind_info, now only
	  checks it.
	  Initialization actually performed, and should work.

2002-05-06 11:29  fawzi

	* [r375] added initialization with linked lists (instread of
	  arrays)

2002-05-06 11:27  fawzi

	* [r374] fixed use (changed to use,only:)

2002-05-06 11:25  fawzi

	* [r373] changed forall to simple do.

2002-05-06 11:24  fawzi

	* [r372] removed superfluos use statement from sll_tests.

2002-05-06 11:20  fawzi

	* [r371] moved initialization of pao_env after the initialization
	  of qs_env

2002-05-06 11:18  fawzi

	* [r370] Changed pointer to target in get_atomic_kind.
	  added initialization of pao_bas_projection.

2002-05-06 11:15  fawzi

	* [r369] removed int_ptr linked list

2002-05-03 16:01  gtb

	* [r367] removed a bug in diis

2002-05-02 17:35  gtb

	* [r365] fixed a memory leak

2002-05-02 16:15  fawzi

	* [r364] I had checked in the wrong version, this one actually
	  compiles.

2002-05-02 14:51  vondele

	* [r363] mp_allgather : calculate_xc improved parallel performance
	  introduced mp_allgather / mp_range
	  this should replace mp_sum when appropriate

2002-05-02 08:16  tchassai

	* [r362] Cosmetic changes

2002-05-02 07:40  fawzi

	* [r361] added cp_sparse_matrix and pao_qs_env_methods.

2002-05-02 07:38  fawzi

	* [r360] replaced pao_matrix with cp_sparse matrix.
	  General implementation and algebra not yet done, compiled, not
	  tested.

2002-05-02 07:35  fawzi

	* [r359] added pao environment initialization.

2002-05-02 07:34  fawzi

	* [r358] added pao projection

2002-05-02 07:33  fawzi

	* [r357] adapted to the new qs_* structures (compiled, not yet
	  functional)

2002-04-30 18:28  gtb

	* [r355] fixed an initialization bug

2002-04-30 17:24  gtb

	* [r349] modified for pw

2002-04-30 17:20  gtb

	* [r348] we treat explicitly 1/v factor

2002-04-30 17:18  gtb

	* [r347] modified for debugging

2002-04-30 17:17  gtb

	* [r346] *** empty log message ***

2002-04-30 17:13  gtb

	* [r345] adapted tochanges of energy_optimize

2002-04-30 17:08  gtb

	* [r344] adapted for pw optimization

2002-04-29 13:40  vondele

	* [r342] bug fix : removed second allocation

2002-04-29 09:22  fawzi

	* [r341] added natoms declaration for get method.

2002-04-29 07:17  tchassai

	* [r339] The routines extracted in going from revision 1.5 to 1.6
	  of mltfftsg.F.

2002-04-29 07:16  tchassai

	* [r338] Extracted the embedded routines to mltfftsg_tools.F and
	  worked some of the code over to allow for OpenMP parallelization.
	  Added OpenMP parallelization.

2002-04-29 07:13  tchassai

	* [r337] Added OpenMP directives.

2002-04-28 08:07  fawzi

	* [r336] added natoms to qs_env

2002-04-26 15:12  tchassai

	* [r335] Added OpenMP directives and cosmetic changes.

2002-04-25 12:52  fawzi

	* [r334] commented out the use,only (then with pgi you should put
	  the use, only evrywhere.)

2002-04-25 07:03  fawzi

	* [r333] remove the forgotten public::cp_set, changed use cp_log
	  and error handling to use,only:

2002-04-24 11:46  vondele

	* [r327] sun update, sun bug fix in dsyevx workspace for big
	  systems

2002-04-24 11:45  vondele

	* [r326] sun update

2002-04-24 09:40  vondele

	* [r322] completed derivatives of matrix elements <a|V|b> with
	  respect to Ra and Rb
	  dhab(.,.,1:6) contains the derivates corresponding to hab(.,.)
	  based upon explicit derivation of a and b, not on FFT
	  overhead +- 15%

2002-04-24 09:37  vondele

	* [r321] ipzyx added

2002-04-23 15:58  krack

	* [r320] - MO level shifting implemented
	  - Pointer to the MO set data structure added to the QS
	  environment
	  - The components of the TYPE mo_set_type are now PRIVATE

2002-04-23 07:59  fawzi

	* [r319] added intent inout to compile on mac

2002-04-20 05:11  cjmundy

	* [r317] Zero pw_grid%bounds before call to pw_grid_setup

2002-04-20 05:10  cjmundy

	* [r316] Bug fix. maxlcgf now gone.

2002-04-19 15:37  gtb

	* [r315] forces on coefs debugged

2002-04-19 11:56  krack

	* [r314] Deallocation of pcon moved.

2002-04-18 17:58  fawzi

	* [r312] renamed cp_pao_* to pao_*, added reading of the pao
	  section to qs_init.

2002-04-18 13:33  vondele

	* [r311] forces : initial modifications

2002-04-18 13:32  vondele

	* [r310] l_init : lmax-> lmax+1, needed for grid forces

2002-04-18 13:31  vondele

	* [r309] exp_radius, added optional argument, bissection search

2002-04-18 09:08  krack

	* [r306] Smearing of MO occupation numbers.

2002-04-17 12:58  fawzi

	* [r301] added linked lists, added back cp_*, not yet integrated
	  with new qs_* structures.

2002-04-17 12:56  fawzi

	* [r300] corrected use of cp_error and log handling, removed now
	  superfluos init_default_loggers and error.

2002-04-17 11:55  fawzi

	* [r299] made find_block_node public

2002-04-17 11:54  fawzi

	* [r298] removed the only: operator(.over.) from use mathlib
	  because it didn't compile with abosoft on macintosh.

2002-04-17 07:33  krack

	* [r297] init_orbital_pointers called with maxlcgf+1 instead of
	  maxlcgf (maxder=1).

2002-04-16 19:38  cjmundy

	* [r296] ngrids not defined

2002-04-16 18:21  cjmundy

	* [r295] MPI bug fix

2002-04-16 09:50  krack

	* [r293] Unterminated string in comment line removed.

2002-04-16 08:54  krack

	* [r292] Dirty fix: line 1387 of parallel version commented, since
	  thermo%kin is not
	  defined.

2002-04-15 23:15  cjmundy

	* [r290] Bug fix for coef forces

2002-04-15 15:34  krack

	* [r289] Renamed to sparse_matrix_types.

2002-04-15 15:33  krack

	* [r288] *** empty log message ***

2002-04-15 15:27  krack

	* [r287] *** empty log message ***

2002-04-15 15:10  krack

	* [r286] Sorted.

2002-04-15 14:55  krack

	* [r285] Sorted.

2002-04-15 14:49  gtb

	* [r284] again volumes

2002-04-15 14:48  gtb

	* [r283] initial verzion

2002-04-15 14:46  gtb

	* [r281] added initialization of pw's

2002-04-15 14:45  gtb

	* [r280] *** empty log message ***

2002-04-15 14:41  krack

	* [r279] New Quickstep data structures.

2002-04-15 14:30  krack

	* [r278] New Quickstep (QS) data structures (old or renamed
	  modules)

2002-04-15 00:22  cjmundy

	* [r277] Bug fixes for ION forces

2002-04-10 17:58  gtb

	* [r276] *** empty log message ***

2002-04-10 17:00  gtb

	* [r275] initializes coeffs with a gaussian

2002-04-10 16:58  gtb

	* [r274] initialization is done in pol_force_pw

2002-04-10 16:57  gtb

	* [r273] ncoeff = ncoeff/100

2002-04-10 16:53  gtb

	* [r272] forces on ions

2002-04-10 16:50  gtb

	* [r271] fixed volume factors

2002-04-10 16:47  gtb

	* [r270] increase max_stack to 10000 for debugging

2002-04-09 15:12  tchassai

	* [r268] Removed incorrect OpenMP comments.

2002-04-09 15:09  tchassai

	* [r267] Refactored calculate_rho_elec() code.

2002-04-08 15:51  fawzi

	* [r266] removed tabs to indent, transformed CPSourceFileRef in a
	  macro vadiable (removed empty arg list.

2002-04-05 21:39  fawzi

	* [r265] added cp_file_ref to cp_error_handling imports

2002-04-05 21:25  fawzi

	* [r264] added cp_file_ref function ro compile on sgi

2002-04-05 21:25  fawzi

	* [r263] added short expansions to compile on sgi

2002-04-05 16:31  fawzi

	* [r262] removed the SOURCE robodoc keyword from routines.
	  put CPSourcefileRef() always on short lines.

2002-04-05 12:39  fawzi

	* [r261] compiled cp_matrix_structure.F

2002-04-05 09:07  fawzi

	* [r260] renamed cp_matrix_structure.F to cp_b_matrix_structure.F

2002-04-05 08:02  fawzi

	* [r259] cleaning up some compilation problems: forgot linked list
	  templates, problems with the '' in the comments if cpp is used
	  without -traditional.

2002-04-04 19:11  fawzi

	* [r258] renamed qs_* to cp_* and QS* to CP*

2002-04-04 19:09  fawzi

	* [r257] added cp_* files, compiled with pgi and power4.

2002-04-04 19:04  fawzi

	* [r254] removed circulare ref to qs_scf with a rip off from it.

2002-04-04 18:47  fawzi

	* [r253] made find_block_node public (to efficiently iterate on
	  them).

2002-04-04 18:45  fawzi

	* [r252] renamed qs_* to cp_*

2002-04-04 18:44  fawzi

	* [r251] added routines to work with sparse block diag matrices.

2002-04-04 18:42  fawzi

	* [r250] first add of template files instatiation info

2002-04-04 18:41  fawzi

	* [r249] first add of template files

2002-04-04 18:41  fawzi

	* [r248] renamed qs_* to cp_*, many many changes.

2002-04-03 14:59  tchassai

	* [r247] Added init and info to the energy_density interface.

2002-04-03 13:29  tchassai

	* [r246] Added subroutine to compute the per particle energy
	  density.

2002-04-03 13:27  tchassai

	* [r245] Added subroutine to compute per particle correlation
	  energy

2002-04-01 17:37  cjmundy

	* [r244] Volumes and bugfixes

2002-03-28 13:35  jgh

	* [r243] Correct version, passt first set of tests

2002-03-26 08:25  jgh

	* [r242] Some further updates of functionals.

2002-03-26 05:51  cjmundy

	* [r241] bug fix

2002-03-25 22:54  cjmundy

	* [r240] included pw_grid in call to pol_grids_initialize

2002-03-24 20:37  cjmundy

	* [r239] bug fixes

2002-03-24 03:56  cjmundy

	* [r238] Signs and def. of ngt

2002-03-21 01:04  cjmundy

	* [r236] AIX 604 compiler bug fix

2002-03-21 00:31  cjmundy

	* [r235] Silly stuff

2002-03-21 00:30  cjmundy

	* [r234] silly stuff

2002-03-21 00:30  cjmundy

	* [r233] Fixed sign on coef force. Fixed logic in electrostatics
	  calculation

2002-03-18 18:40  gtb

	* [r232] trivial stuff

2002-03-18 18:19  fawzi

	* [r231] removed trigraphs "??'" that the precompiler substitued
	  with "^"

2002-03-17 17:52  jgh

	* [r230] More functionals

2002-03-15 23:02  cjmundy

	* [r229] Silly stuff

2002-03-15 22:49  cjmundy

	* [r228] Made good on some volume factors and factor of 2's for
	  forces. So, now we only worry about signs!

2002-03-15 19:27  gtb

	* [r227] *** empty log message ***

2002-03-15 19:26  gtb

	* [r226] eigr was already nullified

2002-03-15 19:25  gtb

	* [r225] eigr_flag TRUE for proj

2002-03-15 15:11  jgh

	* [r224] Update of functionals, still not yet a final version

2002-03-14 19:15  gtb

	* [r223] fcoef multipled by dvols

2002-03-14 19:14  gtb

	* [r222] * drpot1 has to be passed to force_evaluete_part
	  * in_space for coeff_dummy defined

2002-03-14 19:12  gtb

	* [r221] initialized coefs to be 0 at g=0
	  added definition in_space = PW_RECIPROCALSPACE for pw in coefs

2002-03-14 19:10  gtb

	* [r220] 1) added line
	  struc ( 1 ) % pw_grid ( 1 ) % grid_span = HALFSPACE
	  2) pw_grids is now passed in the initialization of coefs

2002-03-14 19:08  gtb

	* [r219] added statement
	  ELSE IF ( PRESENT ( ehartree ) .AND. .NOT. PRESENT ( vhartree ) )
	  THEN
	  to avoid crashing

2002-03-14 19:06  gtb

	* [r218] iw is now assigned

2002-03-14 19:05  gtb

	* [r217] the pointers in the intra_param structure are nullified

2002-03-14 19:04  gtb

	* [r216] checks if the pointers in the intra_param structure
	  are associated

2002-03-14 16:48  jgh

	* [r215] Function definitions checked wrt to definitions in
	  orbital_transformations

2002-03-14 16:47  jgh

	* [r214] ao_structure_factor_type has to be public

2002-03-13 17:31  cjmundy

	* [r212] Modifications for PLANEWAVE manifestation of polarizable
	  model.

2002-03-13 17:29  cjmundy

	* [r211] Added pw_grid and projector ( of type ao ).

2002-03-13 17:28  cjmundy

	* [r210] Contains logic for PLANEWAVE vs ATOMICORBITAL for
	  polarizable model

2002-03-13 17:27  cjmundy

	* [r209] Charge transfer model in

2002-03-13 17:26  cjmundy

	* [r208] Atomic orbital routines for polarizable model

2002-03-13 17:25  cjmundy

	* [r207] Plane wave routines for polarizable model

2002-03-11 17:07  gtb

	* [r206] Faster version

2002-03-09 14:44  jgh

	* [r205] ome updates on functionals and test routines

2002-03-07 10:22  vondele

	* [r202] round-off bug fixed in cube_utils.F

2002-03-04 16:01  jgh

	* [r201] Cleaned up versions of Quickstep grid routines

2002-03-04 09:20  jgh

	* [r200] Clebsch-Gordon Coefficients added to module
	  spherical_harmonics
	  New modules for testing : cg_test and lebedev (from Quickstep)

2002-03-01 09:03  jgh

	* [r199] Correct size estimate for work array in syevx (Joost
	  VandeVondele)

2002-02-28 15:22  jgh

	* [r198] Joost VandeVondele: Multi-Grid code for Quickstep

2002-02-28 15:21  jgh

	* [r197] LYP functional (not yet working)

2002-02-28 15:20  jgh

	* [r196] Update of functionals, some work some not.

2002-02-22 09:59  jgh

	* [r194] Update of QS grid code: optimizations for different
	  compilers
	  Add isolated system option to QS poisson solver

2002-02-22 09:58  jgh

	* [r193] Reformat a comment: some compilers have a default max line
	  length

2002-02-21 19:18  jgh

	* [r192] Kinetic energy GGA functionals

2002-02-21 14:34  jgh

	* [r191] Initial version of Perdew-Zunger LDA functional

2002-02-20 10:41  jgh

	* [r190] Joost VandeVondele's fast grid routines for Quickstep

2002-02-20 10:40  jgh

	* [r189] Intermediate update for new functional code

2002-02-18 09:49  jgh

	* [r187] New code for XC functionals

2002-02-18 09:48  jgh

	* [r186] Update of atomic code

2002-02-11 18:16  fawzi

	* [r185] now the error is changed only if the previous is less
	  important.

2002-02-11 16:40  fawzi

	* [r184] corrected big bug in precondition,... macros without
	  filename.
	  new macros with less arguments.

2002-02-11 16:37  fawzi

	* [r183] corrected a bug with old failure in qs_assert, better
	  handling of the case of lighter error after the first.

2002-02-07 09:36  fawzi

	* [r180] added qs_error_type to the nedeed use in the description.

2002-02-07 09:24  fawzi

	* [r179] added _type to all the types
	  qs_error_handling fully robodoc-tized
	  qs_pao_types in the flow (and grown way too big)
	  compiled and (small) tests.

2002-01-30 16:27  fawzi

	* [r177] never do a small change without compiling before checkin!
	  the inlining of the tests was not done.

2002-01-30 09:47  jgh

	* [r176] Rename variable topology to topo, avoid name conflict with
	  module

2002-01-30 09:43  jgh

	* [r175] Update of atomic code

2002-01-29 22:33  fawzi

	* [r174] deactivating filename by default (so that assertion macros
	  have costant length and don't give problems if the file path is
	  too long)

2002-01-29 17:45  fawzi

	* [r173] fixed a small bug that would lock the print_message if
	  there is a word without spaces that is longer than a line. Ugly
	  fix, I should rewrite the code.

2002-01-29 17:40  fawzi

	* [r171] first add of the pao files.
	  qs_obj_function.F must still be defined (I hadn't realized that
	  there were two kinds matrix in the code of cp2k).
	  lbfgs must still go in a module.
	  Compiled with pgf90, and passed all the (minimal) tests.
	  Still run as standalone (connection with cp2k not yet written).
	  Matthias is changing many names, will soon need changes to be
	  compiled.

2002-01-28 12:39  jgh

	* [r170] Changes related to new input through Gromos topology file
	  Input definition of force field parameter has changed

2001-12-12 16:52  jgh

	* [r169] Optional arguments to parser_init : comment_character,
	  continuation line
	  character and end of section string

2001-12-09 14:33  jgh

	* [r168] Split parser_init routine into two - new routine
	  search_label is called
	  after parser_init (that only opens the input files and stops if
	  the file
	  was not found) to position the file at the beginning of the
	  section
	  label.
	  
	  Added CELL to the setup section. At the moment this is compulsory
	  input.

2001-12-09 14:31  jgh

	* [r167] Update on atomic code, nothing special

2001-12-01 17:56  jgh

	* [r163] New style of input in molecule_input.F
	  - number of molecules is automatically detected (no more keyword)
	  - simplified input for "atoms"
	  - defaults for thermostats (also as input option)
	  - moved some setup code from force_fields to molecule_input.F

2001-12-01 17:53  jgh

	* [r162] Input of BASISSET file, Pseudopotental library, topology
	  file etc.

2001-12-01 17:51  jgh

	* [r161] only a new comment

2001-12-01 17:50  jgh

	* [r160] some updates, nothing finished yet

2001-11-30 13:18  krack

	* [r159] Tabs removed (not accepted by the Intel compiler)

2001-11-30 13:16  krack

	* [r158] Source files changes caused by the Intel compiler (ifc)

2001-11-30 13:11  krack

	* [r157] Files required by the Intel compiler (ifc)

2001-11-28 11:27  jgh

	* [r156] added information on electronic structure to the input
	  type

2001-11-28 11:26  jgh

	* [r155] add some new routines, disabled some for the moment

2001-11-28 11:25  jgh

	* [r154] disable DFT force for the moment

2001-11-28 11:24  jgh

	* [r153] trivial changes

2001-11-28 11:23  jgh

	* [r152] new control structure

2001-11-28 11:22  jgh

	* [r151] New Parser routine get_next, overloaded function of
	  get_int get_real and cfield

2001-11-28 11:20  jgh

	* [r150] control module for TBMD

2001-11-28 11:19  jgh

	* [r149] Towards a simplified input for atomic based codes
	  also update of the parser routines

2001-11-26 16:42  gtb

	* [r148] 905,906c905,906
	  > COMPLEX ( dbl ), DIMENSION ( : ), ALLOCATABLE, SAVE :: zs
	  > COMPLEX ( dbl ), DIMENSION ( :,:,: ), ALLOCATABLE, SAVE :: cd

2001-11-26 16:23  gtb

	* [r147] *** empty log message ***

2001-11-26 16:11  gtb

	* [r146] fixed a meory leak

2001-11-26 13:37  jgh

	* [r145] Deallocate all temporary arrays at the end of the
	  force_control routine.
	  This fixes a memory leak.

2001-11-26 10:33  hforbert

	* [r144] Author: hforbert
	  
	  New splines.F with additional functionality and comments
	  Changed pair_potential.F (the only file to use init_spline) since
	  the init_spline() call changed its arguments.

2001-11-21 11:01  jgh

	* [r143] *** empty log message ***

2001-11-21 10:23  jgh

	* [r142] A bunch of changes:
	  - new attribut path_integrals in setup
	  - some new atomic code routines
	  - separated the wave control module

2001-11-08 23:48  cjmundy

	* [r140] POSVEL option now works. Before, the masses
	  were NOT initialized

2001-11-08 19:49  cjmundy

	* [r139] Better, general method for counting degrees of
	  freedom in nhc_pol

2001-11-08 19:44  cjmundy

	* [r138] Changed maxiter to 500 instead of 50 in exp_radius

2001-11-08 19:40  cjmundy

	* [r137] Reordered ALLOCATE and DEALLOCATE statements
	  to fit better with the standard "the first deallocate
	  should be the last allocate." Hopefully this will
	  cause less memory "fracture." As if I know what I
	  am talking about!

2001-11-01 12:49  gtb

	* [r136] fixed a bug in initialization

2001-11-01 12:49  gtb

	* [r135] nullified pointer pw%pw_grid

2001-11-01 12:48  gtb

	* [r134] fixed a bug in calculation of overlap forces
	  on ions

2001-11-01 12:47  gtb

	* [r133] deallocated 2 pointers

2001-11-01 12:46  gtb

	* [r132] added routine initialize_coef_velocities

2001-11-01 12:45  gtb

	* [r131] fixed some bugs in thermostats on coefficents

2001-11-01 12:44  gtb

	* [r130] reads additional input variables needed for pol MD

2001-11-01 12:43  gtb

	* [r129] modified routine energy_coef for polarization MD

2001-11-01 12:42  gtb

	* [r128] nullified pointer vh

2001-11-01 12:41  gtb

	* [r127] added conversion for polarizaition variables

2001-11-01 12:40  gtb

	* [r126] passed globenv to initialize_pol_coefs

2001-10-29 21:20  cjmundy

	* [r125] Added _dbl on line 153. DEC compiler complained

2001-10-29 21:19  cjmundy

	* [r124] NULLIFIED atom_list pointer. Without this crashed on DEC

2001-10-23 19:32  gtb

	* [r123] added objects pol_diis and energy_optimize

2001-10-23 19:32  gtb

	* [r122] initlal version

2001-10-23 19:31  gtb

	* [r121] initial version. diis minimizer

2001-10-23 19:29  gtb

	* [r120] added interface lapack_sgesv

2001-10-23 19:28  gtb

	* [r119] removed a bug

2001-10-23 19:27  gtb

	* [r118] energy minimization is now in and working

2001-10-23 19:25  gtb

	* [r117] added subtoutine solve_system (calls library routine
	  dgesv)

2001-10-23 19:23  gtb

	* [r116] modified update_image_list in order to adapt it to the
	  overlap list

2001-10-23 19:22  gtb

	* [r115] added the overlap contribution to forces

2001-10-23 19:21  gtb

	* [r114] forces are initialized to 0

2001-10-23 19:20  gtb

	* [r113] added in thermodynamic_type the variables (real)
	  e_overlap, e_thk, e_class, e_electro

2001-10-19 00:04  cjmundy

	* [r112] coef is now REAL in dg_rho0_s

2001-10-19 00:02  cjmundy

	* [r111] Computes cutoff_radius using exp_radius as
	  is done in polarization

2001-10-19 00:00  cjmundy

	* [r110] Modeled PME setup after POL setup. Now
	  much leaner. Uses exp_radius to find
	  cut-off ( in interactions.F )

2001-10-10 08:41  jgh

	* [r109] new method diff6p, code clean up

2001-10-05 16:21  cjmundy

	* [r108] Fixed minor bugs in nve_ao and nvt_ao

2001-10-02 00:16  cjmundy

	* [r107] Added logic to allow one easily to take small grid = big
	  grid limit
	  in the case of polarization

2001-08-31 09:40  hforbert

	* [r106] Author: hforbert
	  Fixed a lone ' in a comment that gave my preprocessor a hickup

2001-08-27 17:42  cjmundy

	* [r105] Clean-up

2001-08-25 21:46  cjmundy

	* [r104] Change thermo % pv_kin to pv_kin in get_veps. Works

2001-08-23 21:21  cjmundy

	* [r103] dbg statements for isotropic limit

2001-08-23 21:20  cjmundy

	* [r102] NPT_F bug fix. Left dbg statements
	  for Isotropic limit

2001-08-14 20:47  cjmundy

	* [r101] Mass_flag is only deployed when allocating c_vel

2001-08-10 14:08  krack

	* [r100] HCTH functionals (HCTH/93, HCTH/120,HCTH/147,HCTH/407)
	  added in functionals.F;
	  Quickstep (GPW) works now also for GGA functionals

2001-08-10 14:03  krack

	* [r99] scratch vector for eigenvalues added in blacs_syevx

2001-08-10 14:01  krack

	* [r98] string variable "apname" initialised

2001-08-09 16:34  gtb

	* [r97] adapted to the changes in linlist_utilities (for images)

2001-08-08 20:37  cjmundy

	* [r96] Clean up!

2001-08-08 20:36  cjmundy

	* [r95] Clean up.

2001-08-08 20:36  cjmundy

	* [r94] Clean up. Fixed bug in pme_force_numer

2001-07-27 17:23  hforbert

	* [r93] Author: hforbert
	  
	  some additional code cleanup after the cutoff>boxsize/2 bugfix

2001-07-27 04:05  cjmundy

	* [r92] Cleaned up

2001-07-27 03:48  cjmundy

	* [r91] Fixed imaging bug when cut > box/2

2001-07-13 17:15  gtb

	* [r90] added some routines for
	  energy optimization

2001-07-13 17:13  gtb

	* [r89] added framework to treat energy optimization

2001-07-13 17:05  gtb

	* [r88] replaced simpar%ensemble mit box_change

2001-07-13 17:03  gtb

	* [r87] added routines for debugging overlap forces.
	  logical box_change is defined and passed

2001-07-13 17:02  gtb

	* [r86] logical box_change is now passed

2001-07-13 17:01  gtb

	* [r85] added logic for the debugging of overlap forces

2001-07-13 16:50  gtb

	* [r84] added new routine dump_pol_coefs to print
	  coefficents on file

2001-07-13 16:49  gtb

	* [r83] only minor changes

2001-07-13 16:47  gtb

	* [r82] Initial Version
	  
	  Prints to screen information about
	  energy minimization runs

2001-07-13 16:47  gtb

	* [r81] Initial Revision
	  
	  Reads in the set_file the section &ENERGY
	  in which are contained the information for the
	  minimization of the energy functional

2001-07-13 01:07  cjmundy

	* [r80] New debug routines

2001-07-13 01:06  cjmundy

	* [r79] New debug version

2001-07-13 01:05  cjmundy

	* [r78] Made debug friendly

2001-07-13 01:04  cjmundy

	* [r77] General electrostatics debugger.

2001-07-13 01:04  cjmundy

	* [r76] Added get_ewald_grid routine for debugging

2001-07-13 01:03  cjmundy

	* [r75] General electrostatics debugger ( Ewald, pme, etc )

2001-07-12 05:57  cjmundy

	* [r74] More cleaning

2001-07-12 02:19  cjmundy

	* [r73] Cleaned up and made debug worthy

2001-07-12 02:18  cjmundy

	* [r72] Added routine to obtain grid info

2001-07-12 02:18  cjmundy

	* [r71] Debug driver for polarization

2001-07-12 02:17  cjmundy

	* [r70] Added logic for pol_debug.F

2001-07-12 02:16  cjmundy

	* [r69] Numerical force routine for polarization

2001-07-12 02:14  cjmundy

	* [r68] Added debug routines for polarization

2001-07-12 02:11  cjmundy

	* [r67] Zeroed energy

2001-07-10 20:53  cjmundy

	* [r66] Nullified kind_info_type

2001-07-10 16:06  cjmundy

	* [r65] BUG FIX for extended systems integration. Units were
	  previously not being converted.

2001-07-08 18:07  gtb

	* [r64] removed a lot of bugs!

2001-07-08 18:06  gtb

	* [r63] initializes the variables relative to polarization

2001-07-08 18:05  gtb

	* [r62] added pol_thk

2001-07-08 18:05  gtb

	* [r61] initial revision

2001-07-08 18:04  gtb

	* [r60] removed a bug

2001-07-08 18:03  gtb

	* [r59] removed some bugs

2001-07-08 18:03  gtb

	* [r58] minor changes in routine get_kind_and_part_index

2001-07-08 18:02  gtb

	* [r57] minor changes; added call to routine that calculates
	  first order contribution to polaRization

2001-07-08 17:40  gtb

	* [r56] the integer pointer coef_to_basis_set has been replaced
	  by 2 integer pointers coef_to_basis and coef_to_set

2001-07-08 17:38  gtb

	* [r55] Minor change: added a select case for polarization

2001-07-05 12:02  jgh

	* [r51] not yet working version for g95 compiler

2001-07-05 12:00  jgh

	* [r50] added G95 interface

2001-07-05 11:59  jgh

	* [r49] Restrict comments to 72 cols (F77-fixed form standard)

2001-07-03 13:00  jgh

	* [r44] file removed (new in makefiles)

2001-07-03 12:59  jgh

	* [r43] files removed (new in makefiles)

2001-07-02 17:22  cjmundy

	* [r41] NPT_I bug fix

2001-07-02 15:18  krack

	* [r40] - Line 1195 corrected

2001-07-02 15:02  jgh

	* [r39] Workaround for PGI compiler bug

2001-07-02 14:43  krack

	* [r38] - Three-center overlap integrals [a|c|b] with a,b
	  contracted and c primitive.

2001-07-02 14:33  krack

	* [r37] - Parallelization of the QS SCF using ScaLAPACK/PBLAS/BLACS
	  (-> blacs.F)
	  - CRAY and Linux architecture files updated
	  - Some subroutine names changed to lower case (third try)

2001-07-02 09:08  jgh

	* [r36] Removed some TABs

2001-07-01 17:46  gtb

	* [r35] added some integer pointer for the mapping to
	  the basis sets

2001-07-01 17:45  gtb

	* [r34] removed the integer pointer node in part

2001-07-01 17:44  gtb

	* [r33] *** empty log message ***

2001-07-01 17:44  gtb

	* [r32] fixed a bug
	  modified the loop structure

2001-07-01 17:42  gtb

	* [r31] added empparm in interactions

2001-07-01 17:42  gtb

	* [r30] small chenges for addition of empirical parameters

2001-07-01 17:41  gtb

	* [r29] added changes for the reading of the emoirical parameters
	  in polarization code

2001-07-01 17:40  gtb

	* [r28] added subroutine init_cphi

2001-07-01 17:39  gtb

	* [r27] removed a bug in subroitune get_kind_and_part_index

2001-07-01 17:37  gtb

	* [r26] *** empty log message ***

2001-06-30 18:50  cjmundy

	* [r25] NPT_I bug fix

2001-06-30 18:47  cjmundy

	* [r24] NPT_I bug fix. vol factors removed from
	  stress calculation in pw_poisson_solver.
	  The volume factors were present in
	  pw_integral_**.

2001-06-30 15:54  jgh

	* [r23] Cleaning up after last major update

2001-06-29 06:58  cjmundy

	* [r19] New version:
	  
	  Contains
	  
	  1) fit modules
	  2) polarization modules
	  
	  Includes many bug fixes to SPME. There still seems to be a
	  problem
	  with pw_integral_ab in regards to SPME/PME. This version will
	  work for SPME not for PME. Needs to be examined.

2001-06-29 06:49  cjmundy

	* [r18] Makefile for DEC_OS

2001-06-27 10:03  jgh

	* [r17] some small changes

2001-06-26 18:44  jgh

	* [r16] update to version /06/25/2001

2001-06-26 18:34  jgh

	* [r14] update to version /03/20/2001

2001-06-26 18:15  jgh

	* [r13] update to version 02/01/2001

2001-06-26 18:06  jgh

	* [r11] update to version 01/01/2001

2001-06-26 17:51  jgh

	* [r9] update to version 01/12/2000

2001-06-26 17:41  jgh

	* [r7] update to version 11/1/2000

2001-06-26 16:46  jgh

	* [r4] Initial revision

