David van der Spoel [Thu, 19 Jan 2012 19:20:46 +0000 (20:20 +0100)]
Added two new bonded functions.
The first one implements a potential for linear angles without any
discontinuities at 180 degrees. A paper is being written about this. The
potential has [ angles ] type 9 or F_LINEAR_ANGLES internally.
A free energy option has been implemented as well for the linear angles.
The second is an anharmonic polarization function, based on the normal harmonic
term plus a forth order term, as introduced by the Roux group in
J. Chem. Theor. Comput. 6 (2010) p. 774-786. The potential has [ polarization ]
type 2 or F_ANHARM_POL internally.
Change-Id: I0497ed20baf1c1c0b2a6c2354eed77bc9ed75c6d
Roland Schulz [Wed, 18 Jan 2012 08:05:54 +0000 (09:05 +0100)]
Revert "Replace stderr by stdout in editconf terminal output for easier scripting."
This reverts commit
bc2fbe9b50a8a39a185cb6a448dea43c8ad9ed7b
David van der Spoel [Tue, 17 Jan 2012 10:56:28 +0000 (11:56 +0100)]
Replace stderr by stdout in editconf terminal output for easier scripting.
Change-Id: If477f6148d347fbbed16d37db62e2e12f1444ece
Mark Abraham [Sat, 14 Jan 2012 05:37:18 +0000 (06:37 +0100)]
Merge "Remove use of FindMPI.cmake" into release-4-6
Mark Abraham [Sat, 14 Jan 2012 05:18:14 +0000 (06:18 +0100)]
Merge "Removing outdated vsite comment" into release-4-6
Mark Abraham [Mon, 9 Jan 2012 09:38:33 +0000 (20:38 +1100)]
Remove use of FindMPI.cmake
The functionality of FindMPI.cmake should not be needed for GROMACS,
since the wrapper compiler will do the complete job. This already
worked for Cray and BlueGene, and so far this new version is known
to work for OpenMPI wrapping icc and gcc.
CMake now warns the user about possible unsuitable versions of
OpenMPI and MVAPICH2.
The process of managing MPI is more modular than it used to be.
To use:
cmake .. -DGMX_MPI=ON -DCMAKE_C_COMPILER=`which mpicc`
or with bash
CC=mpicc cmake .. -DGMX_MPI=ON
and whatever else suits your setup.
Fixes #851 and #636.
Change-Id: Ibe41206bed8b70b83a25da1e4e29dd87b61ea17d
Christoph Junghans [Fri, 13 Jan 2012 18:00:13 +0000 (11:00 -0700)]
CMake improvements
* make CMake quiet again, when running for the second time
* remember that VMD was found before
* removed duplicate definition of HAVE_FSEEKO
Change-Id: If2a2cc600fd14e67c6169e12172e7c5aa8ebedfc
Roland Schulz [Fri, 13 Jan 2012 22:08:04 +0000 (23:08 +0100)]
Merge "Fixed MSVC compiling with plug-in support" into release-4-6
Roland Schulz [Fri, 13 Jan 2012 18:54:42 +0000 (13:54 -0500)]
Removing outdated vsite comment
Vsite parameters were replace by Berk in commit
14bc4ad9 by
calculated parameters. Since then the comment was outdated.
Change-Id: I6ddacfbdcf40e11db9d2d3091d9c34f4872816bc
Christoph Junghans [Thu, 9 Jun 2011 09:16:58 +0000 (11:16 +0200)]
Adds AdResS feature to gromacs
This is actually a rebase of the following commits:
* backport adress to 4.5, was ported from adress branch, which is based on the master
* update init_forcerec in gmx_membed.c due to adress
* energy terms in AdreS only caclulated for meaningful properties.
* adress variables only printed if adress is used
* grompp checks to prevent adress being used with lr coulomb or pressure coupling
* Windows/C90 fixes
* put inputrec variables into adress struct
* increases tpx_version from 74 to 75 (as enforced rotation was 74)
* rebase to recent 4.6 branch
* bAdress to false for tpx < 75 (disabled by default)
Change-Id: Ica1d121f8edb9086ac2d5d456b1037a3d57a24e4
Mark Abraham [Mon, 9 Jan 2012 13:55:58 +0000 (00:55 +1100)]
Fixed MSVC compiling with plug-in support
Fixes #598
Change-Id: I8fe274a83c4965466f1241d2237ca3e77b93b038
Mark Abraham [Mon, 9 Jan 2012 13:50:56 +0000 (00:50 +1100)]
Fixed use of _snprintf_s with MSVC
Change-Id: I807ef17e8b4eb07861600c6a1d9d03f330f17218
Roland Schulz [Sat, 7 Jan 2012 16:52:32 +0000 (11:52 -0500)]
Check for VMD plugins that they are at least abi-version 10
All plugins with VMD 1.8 are abi-version 10. Plugin support is
untested for older versions.
Change-Id: I98fe11313d7115d476e44bff0bc729c7380c52f0
Mark Abraham [Thu, 5 Jan 2012 06:21:14 +0000 (17:21 +1100)]
Made use of VMD plugins more robust
Attempt to use VMD trajectory-reading plugins only if plugin
loading functionality exists and shared libraries are being used
(else dlopen can be unreliable).
Also, for helping to find those plugins, the user can supply the plug-in
path in a CMake cache variable. If not, the code can now fall back
on the run-time value of the environment variable VMDDIR, or its
value at configuration time. Previously only an explicit run-time
environment variable or a hard-coded path was available on non-Windows
platforms.
Also printed diagnostic text when GROMACS cannot tell if the
trajectory being read by a plug-in might have atom velocities.
Fixes #598
Change-Id: I44267e115267772dae2c1d754895f8ff25c98a6a
Roland Schulz [Mon, 26 Dec 2011 14:06:22 +0000 (09:06 -0500)]
Merge remote-tracking branch 'gerrit/release-4-5-patches' into release-4-6
Conflicts:
src/tools/gmx_tune_pme.c
Change-Id: I0cb93d965cd30a6d0e158eac595fac7c8b891492
Roland Schulz [Tue, 27 Dec 2011 09:29:19 +0000 (10:29 +0100)]
Merge "Added -dhdl to tpr file list, fixes #855" into release-4-5-patches
Carsten Kutzner [Mon, 19 Dec 2011 14:42:29 +0000 (15:42 +0100)]
Added -dhdl to tpr file list, fixes #855
Change-Id: I2621c7bb44daa01e25831e2d44c22cdcf47f58e7
Roland Schulz [Tue, 20 Dec 2011 12:28:48 +0000 (07:28 -0500)]
Merge release-4-5-patches into release-4-6
Conflicts:
admin/GerritBuild.bat
include/gstat.h
src/tools/anadih.c
src/tools/gmx_chi.c
Change-Id: If70475e8f38978aeb52652afce2ca2a62d355987
Roland Schulz [Tue, 20 Dec 2011 07:52:33 +0000 (08:52 +0100)]
Merge "Using MinSizeRel configuration and Release target on Windows" into release-4-5-patches
Roland Schulz [Tue, 20 Dec 2011 07:52:06 +0000 (08:52 +0100)]
Merge changes Iaf4c56b7,Ibabed87f into release-4-5-patches
* changes:
Testing external BLAS/LAPACK
Using 2 cores per build on Windows
Christoph Junghans [Mon, 19 Dec 2011 18:23:25 +0000 (11:23 -0700)]
Set GMX_*_ASM true if GMX_ACCELERATION != auto
In case GMX_ACCELERATION was set to SSE manually (like in Gentoo Linux)
GMX_*_ASM would never be set (false), so sse kernel will not be build.
Change-Id: Icf04707c6174a0182fe4472536b568e1d2d91166
Carsten Kutzner [Wed, 23 Nov 2011 17:37:38 +0000 (18:37 +0100)]
Enforced rotation
Change-Id: Ia2d3b409c197fd2a6bd5922b636bd40a47046918
Roland Schulz [Thu, 15 Dec 2011 12:01:55 +0000 (07:01 -0500)]
Using MinSizeRel configuration and Release target on Windows
Faster build:
no need to compute debug symbols
dynamic library
low optimization level
Also fixing MSVC warnings with this configuration
Change-Id: Ic87da40f1e78e2ac70ef543525b49f4cf1a3f639
Roland Schulz [Thu, 15 Dec 2011 08:20:17 +0000 (03:20 -0500)]
Testing external BLAS/LAPACK
Change-Id: Iaf4c56b7e62654425801cd8d3272c4028b016a2a
Roland Schulz [Thu, 15 Dec 2011 06:27:08 +0000 (01:27 -0500)]
Using 2 cores per build on Windows
Windows build is much slower than Linux build. Don't understand why
but hopefully using 2 cores will help narrow that gap.
Change-Id: Ibabed87f2f4d6c931fbb1ebd25c571be9a2cb9ee
Roland Schulz [Wed, 14 Dec 2011 01:49:28 +0000 (02:49 +0100)]
Merge "Re-fixed static linking of maths libraries under UNIX" into release-4-5-patches
Mark Abraham [Wed, 14 Dec 2011 01:10:26 +0000 (12:10 +1100)]
Re-fixed static linking of maths libraries under UNIX
In commit
5e51ffa9 I submitted a patch for the above case, and
suggested the same to CMake. The CMake developers have refined
this and I'm back-porting their fix here.
See http://public.kitware.com/Bug/view.php?id=12477 for details
Change-Id: Ib48bc6b5a87b66b51c893e24aa42e4cd4db32a38
Roland Schulz [Tue, 13 Dec 2011 07:29:42 +0000 (08:29 +0100)]
Merge "Converted macro usage for readability" into release-4-6
Mark Abraham [Sun, 4 Dec 2011 23:58:54 +0000 (10:58 +1100)]
Converted macro usage for readability
Replaced some helper macros with inline functions, and expanded
the macro that was implementing three nested loops.
Change-Id: I48e1d6a5401e21cd99b51158e3ddcb1729f842b3
Roland Schulz [Tue, 13 Dec 2011 02:04:29 +0000 (03:04 +0100)]
Merge "Relaxed some assumptions about trajectory times" into release-4-5-patches
Mark Abraham [Tue, 13 Dec 2011 00:04:04 +0000 (11:04 +1100)]
Relaxed some assumptions about trajectory times
Counting of dihedral transitions used to assume that each frame was
equally spaced in time. This is only necessary for making a
subsequent histogram of those transition times (unless more
elaborate machinery was written for that histogram).
Also documented what g_angle -ot and -oh do.
Change-Id: I587dba94c40a1a69e5aa227d09271166db055595
Mark Abraham [Sun, 4 Dec 2011 22:50:25 +0000 (09:50 +1100)]
Fixed macros that ended in semicolons
These now don't end in semicolons, which makes life much simpler
when using parsing tools that are not (fully) macro-aware.
Also replaced some compound-statement macros with inline functions.
Change-Id: Ic83d9419a67d82665ff250e0568a85cf9a8f1c32
Roland Schulz [Sun, 11 Dec 2011 20:05:19 +0000 (15:05 -0500)]
Merge release-4-5-patches into release-4-6
Conflicts:
admin/GerritBuild.bat
Change-Id: Ib62e8f5949cc8e65342cd1826888901d11032474
Roland Schulz [Sun, 11 Dec 2011 19:47:54 +0000 (20:47 +0100)]
Merge "Set -DMPI only if compiled for FAH" into release-4-5-patches
Roland Schulz [Sun, 11 Dec 2011 18:51:07 +0000 (13:51 -0500)]
Set -DMPI only if compiled for FAH
MPI conflicts with C++ MPI namespace.
This allows to compile with FAH or with C++ and MPI.
Compiling with FAH and C++ and MPI will still have a name conflict.
Resolves #849
Change-Id: I73ccb7b0e682ae6d12352c76e26c6d79e93ea798
Roland Schulz [Sun, 11 Dec 2011 02:35:47 +0000 (21:35 -0500)]
Fixing MSVC unsigned/signed warnings
Change-Id: I110b591cfd8c8078f3b7853bdf2db6a705e9e453
Roland Schulz [Sat, 10 Dec 2011 00:53:12 +0000 (19:53 -0500)]
Run tests also on Windows
Change-Id: Iab166b5cc8e394e3d330afffd86aced00aa7e0a2
Roland Schulz [Tue, 11 Oct 2011 21:32:42 +0000 (17:32 -0400)]
Only check those atoms for PBC which are part of selection (Issue #809)
Change-Id: I2afccb82c3b761890f4d1f53d842606bb8404c98
Roland Schulz [Thu, 8 Dec 2011 19:59:20 +0000 (20:59 +0100)]
Merge "Added an option to easily compile contrib programs with cmake" into release-4-6
Roland Schulz [Thu, 8 Dec 2011 19:57:05 +0000 (20:57 +0100)]
Merge "Fix selection parsing in double precision." into release-4-5-patches
Roland Schulz [Thu, 8 Dec 2011 19:53:08 +0000 (20:53 +0100)]
Merge "Avoid possible warnings if HUGE_VAL != infinity." into release-4-5-patches
Roland Schulz [Fri, 25 Nov 2011 11:34:06 +0000 (06:34 -0500)]
Remove RETSIGTYPE. C89 defines return value to be void.
Because it is defined in C89 no sane system should require this anymore.
Checked for Windows.
See e.g. http://lists.gnu.org/archive/html/emacs-devel/2011-03/msg01076.html
Change-Id: If24fa3d47a1ddec189784f49e3e59e1936fe263d
Roland Schulz [Fri, 2 Dec 2011 21:48:17 +0000 (16:48 -0500)]
Adding support for Windows build and different CMake Versions
Ant doesn't work very well - trying shell script instead
Change-Id: Ia08e6979c0e0734c33183840c9d34c3a7367af4a
Roland Schulz [Fri, 2 Dec 2011 21:48:17 +0000 (16:48 -0500)]
Adding support for Windows build and different CMake Versions
Ant doesn't work very well - trying shell script instead
Change-Id: Ia08e6979c0e0734c33183840c9d34c3a7367af4a
Teemu Murtola [Thu, 1 Dec 2011 16:00:09 +0000 (18:00 +0200)]
Fix selection parsing in double precision.
Fixes #850.
Change-Id: Ifd0bf1f5401ca4d62571c8c2fc39e67cac537ecf
Teemu Murtola [Thu, 1 Dec 2011 15:52:28 +0000 (17:52 +0200)]
Avoid possible warnings if HUGE_VAL != infinity.
Replaced HUGE_VAL by GMX_REAL_MAX in one source file for trajectory
analysis library source file, since it seemed to cause compiler warnings
on some systems.
Reported by Chris Neale on gmx-users.
Change-Id: Ic52a01c5bcedb244d1453117f626ef4f0d46df78
Roland Schulz [Wed, 30 Nov 2011 10:10:25 +0000 (05:10 -0500)]
Adding Gerrit Build description
Change-Id: Ib9e88333967178c8f0962a72e5f8432067094e77
Roland Schulz [Wed, 30 Nov 2011 10:10:25 +0000 (05:10 -0500)]
Adding Gerrit Build description
Change-Id: Ib9e88333967178c8f0962a72e5f8432067094e77
Roland Schulz [Sun, 27 Nov 2011 21:30:25 +0000 (16:30 -0500)]
Removing TestBuildAll test
No need for it because ctest automtically tests building anyhow
It does not work under Windows (using msbuild not make)
Change-Id: I02c7152d61cbe5b536c86c406f37ed78db94ee11
Roland Schulz [Wed, 30 Nov 2011 05:37:37 +0000 (06:37 +0100)]
Merge "Merge remote-tracking branch 'gerrit/release-4-5-patches' into release-4-6" into release-4-6
Roland Schulz [Mon, 28 Nov 2011 02:47:22 +0000 (21:47 -0500)]
Fix FFTW API usage
FFTW API uses the fftw/fftwf prefix to select the precision.
To make the code indepent Gromacs uses the FFTW macro.
In this case the marco wasn't used but it had no effect
because for iodim both version are identical.
Submitted by Ake Sandgren (ake.sandgren@hpc2n.umu.se) to gmx-developers
Change-Id: I8ad037b90caef5952b7b49379b81b2c8e90b0daa
Roland Schulz [Sun, 27 Nov 2011 23:57:57 +0000 (00:57 +0100)]
Merge "Fixes inconsistent spelling of Git cmake variables" into release-4-6
Roland Schulz [Sun, 27 Nov 2011 21:52:42 +0000 (16:52 -0500)]
Merge remote-tracking branch 'gerrit/release-4-5-patches' into release-4-6
Change-Id: I9f12fee0e49e1f7d758a2080337269753f31f3bf
Roland Schulz [Sun, 27 Nov 2011 21:51:48 +0000 (22:51 +0100)]
Merge "Replace '_' in all mdp options to '-'" into release-4-6
Roland Schulz [Sun, 27 Nov 2011 21:48:50 +0000 (22:48 +0100)]
Merge "Cleaning up some bad coding practices" into release-4-6
Roland Schulz [Sun, 27 Nov 2011 21:48:33 +0000 (22:48 +0100)]
Merge "Initialize variables to make GCC 4.6 not complain" into release-4-5-patches
Roland Schulz [Sun, 27 Nov 2011 21:30:25 +0000 (16:30 -0500)]
Removing TestBuildAll test
No need for it because ctest automtically tests building anyhow
It does not work under Windows (using msbuild not make)
Change-Id: I02c7152d61cbe5b536c86c406f37ed78db94ee11
Berk Hess [Sun, 27 Nov 2011 21:29:35 +0000 (22:29 +0100)]
fixed pull constraint periodicity consistency check, fixes #844
Change-Id: Ifb8e37c5ca3bb5c36631c0b3db65928f27887bdb
Roland Schulz [Sun, 27 Nov 2011 21:24:43 +0000 (22:24 +0100)]
Merge "Added support for submission to cdash servers." into release-4-6
Roland Schulz [Fri, 25 Nov 2011 12:21:10 +0000 (13:21 +0100)]
Merge "Changed cryptic variable names" into release-4-6
Roland Schulz [Wed, 23 Nov 2011 07:34:07 +0000 (23:34 -0800)]
Cleaning up some bad coding practices
- missing return statement
- recommended parenthesis
- uneeded declarations
- statement without effect
- missing switch/break or undocumented fall-through
- unused static variables
Unless I'm mistaken none of the changes has any effect.
Change-Id: I060ce5b50d8e7df3a97917633fc77d7d2e8481a9
Rossen Apostolov [Fri, 25 Nov 2011 10:44:09 +0000 (11:44 +0100)]
Added support for submission to cdash servers.
With the default setup ctest will send the results to
the Gromacs project on cdash.gromacs.org
Change-Id: Ieb096b7c20734ec7a641523aba06775eb1c2ccc6
Roland Schulz [Wed, 23 Nov 2011 01:45:51 +0000 (17:45 -0800)]
Fixes inconsistent spelling of Git cmake variables
The cmake Git script uses "GIT" but the gmxGenerateVersionInfo
was still using "Git".
Re-introduced cmake/FindGit.cmake copied from CMake 2.8.2,
because CMake 2.8.0 did not include this module. This branch
now builds with CMake 2.8.0 and higher.
Change-Id: I06ee2f7478c16923746591ff5171430454d55549
Mark Abraham [Wed, 23 Nov 2011 03:03:24 +0000 (14:03 +1100)]
Changed cryptic variable names
'multiplicity' is much more readable than 'xity'
Change-Id: I4c070a66c4863414cddfbaaa9c099736f6c98e3d
Roland Schulz [Thu, 24 Nov 2011 06:03:50 +0000 (01:03 -0500)]
Initialize variables to make GCC 4.6 not complain
The variables are only used when initialized but GCC doesn't detect that.
Change-Id: Ib79383e14d5de90f99f0f267e5039f811980bdb8
Mark Abraham [Wed, 23 Nov 2011 03:19:17 +0000 (14:19 +1100)]
Only analyse angle transitions for dihedral angles
Change-Id: Ia973d5b31c38f94a7ca2e3c2d8265b3552253103
Roland Schulz [Thu, 24 Nov 2011 00:20:03 +0000 (01:20 +0100)]
Merge "Fixed g_angle -f -or to copy times from -f to -or" into release-4-5-patches
Mark Abraham [Tue, 24 Nov 2009 02:10:17 +0000 (15:10 +1300)]
Added protective preprocessor directives
Avoids code compiled with -O0 dereferencing null pointers. Duplicates
and expands a fix suggested by Valentin Pavlov <x.pavlov@gmail.com>.
Change-Id: I41f27bce385c5f960270ac616abd6fdd038312fa
Roland Schulz [Wed, 23 Nov 2011 07:27:51 +0000 (23:27 -0800)]
Added an option to easily compile contrib programs with cmake
Change-Id: Ib3fa2368c74c85bb733ba0d52e13e560134cfc5c
Mark Abraham [Wed, 23 Nov 2011 01:46:05 +0000 (12:46 +1100)]
Fixed g_angle -f -or to copy times from -f to -or
Change-Id: Ie3d45c7454880cf65d583980cf6639f5527b22c9
Carsten Kutzner [Tue, 22 Nov 2011 14:14:48 +0000 (15:14 +0100)]
Actually -cpi and -bcpo must not be identical
Change-Id: Iac44c2dc7d7add1f951ff6018bbfaabe8160af25
Berk Hess [Tue, 15 Nov 2011 14:14:45 +0000 (15:14 +0100)]
fixed dd load print frequency with nsteps=-1, fixes #829
Change-Id: I66736af49665e6cc90875acbf52a45b5a9727915
Alexey Shvetsov [Thu, 20 Oct 2011 19:08:31 +0000 (23:08 +0400)]
mdp options in *.html also uses '-' instead of '_'
Change-Id: I856d64badfe26598650d65feae1354fb8d688999
Signed-off-by: Alexey Shvetsov <alexxy@omrb.pnpi.spb.ru>
Christoph Junghans [Tue, 8 Nov 2011 03:18:08 +0000 (04:18 +0100)]
Merge "Fixed bug #830 (segv in do_dssp)" into release-4-5-patches
Kasson [Tue, 8 Nov 2011 02:47:36 +0000 (03:47 +0100)]
Merge "Fix for GB, double, SSE2 and optimized kernels" into release-4-5-patches
Mark Abraham [Wed, 2 Nov 2011 05:00:18 +0000 (16:00 +1100)]
Fix for GB, double, SSE2 and optimized kernels
As suggested by Julius Su in
http://lists.gromacs.org/pipermail/gmx-developers/2011-November/005432.html
Change-Id: Id50d63958e8b3fee810a8d4ed6e021510c98d9aa
Mark Abraham [Tue, 1 Nov 2011 15:41:04 +0000 (02:41 +1100)]
Trigger correct kernels under Cygwin
CMAKE_SYSTEM_PROCESSOR calls uname -p, which returns "unknown" on
Cygwin, so we need to use another mechanism to trigger the use of
SSE/SSE2 optimization.
Change-Id: Ic6406bb36b853a2bd3b22345e36e4914eb04fd8d
Carsten Kutzner [Mon, 31 Oct 2011 10:52:33 +0000 (11:52 +0100)]
Fixed bug #830 (segv in do_dssp)
Change-Id: Ibaa2cf25586d879fce62e9086fee1b510e034e40
Mark Abraham [Thu, 27 Oct 2011 06:34:37 +0000 (17:34 +1100)]
Made reading of binary matrices more robust
The routine to read the binary matrix should check that the storage is
available before using it. Also, the documentation in the header
should be explicit that valid storage locations need to be passed in.
Change-Id: I95e9120bceac02031a7f39412ba226ef2644a444
Alexey Shvetsov [Thu, 20 Oct 2011 19:21:34 +0000 (23:21 +0400)]
Replace '_' to '-' in mdp options for tutorials
Change-Id: I918e50f4c48369ec5a303b2475d89275808ca654
Signed-off-by: Alexey Shvetsov <alexxy@omrb.pnpi.spb.ru>
Alexey Shvetsov [Thu, 20 Oct 2011 19:03:13 +0000 (23:03 +0400)]
Replace '_' in all mdp options to '-'
This changeset replaces '_' to '-' in all mdp options and in
*.log output of mdrun. So options like
init_step -> init-step and so on
Change-Id: If04e4408e7f9f27ee37cbdaa624e14aab46cddfe
Signed-off-by: Alexey Shvetsov <alexxy@omrb.pnpi.spb.ru>
Christoph Junghans [Thu, 20 Oct 2011 12:17:21 +0000 (14:17 +0200)]
Merge "Changed default value for g_anaeig -last to -1" into release-4-6
Christoph Junghans [Thu, 20 Oct 2011 12:14:49 +0000 (14:14 +0200)]
Merge "Fixes for GB kernel" into release-4-5-patches
Christoph Junghans [Thu, 20 Oct 2011 12:10:18 +0000 (14:10 +0200)]
Merge "Add CMakeLists.txt.user to gitignore" into release-4-6
Teemu Murtola [Wed, 19 Oct 2011 15:30:38 +0000 (17:30 +0200)]
Merge "fixed pull constraining doing nothing, fixes #825" into release-4-5-patches
Mark Abraham [Wed, 19 Oct 2011 11:59:48 +0000 (22:59 +1100)]
Corrected charge of CHARMM zinc ion to +2
Change-Id: Id64c3f5bbd9e0993e4b598dc971a780d3d7337dc
Christoph Junghans [Sat, 15 Oct 2011 13:35:37 +0000 (15:35 +0200)]
Merge "Fixes MPI_GROUP_EMPTY for tMPI" into release-4-6
Roland Schulz [Fri, 14 Oct 2011 03:28:10 +0000 (23:28 -0400)]
call to gmx_mtop_generate_local_top is only required once
fixes memory leak
Change-Id: I1ef9860e4b5c24a15acd1b2de88d89721771d840
Mark Abraham [Thu, 13 Oct 2011 03:14:54 +0000 (14:14 +1100)]
Changed default value for g_anaeig -last to -1
The old default for -last was 8, which is pretty arbitrary. There are so many
different operations for g_anaeig that requiring the user to make an active choice
about any eigenvector subset seems wise. The new default is to do the operation
on all of the eigenvectors.
Change-Id: I52ba7539d70e559712b52e5ec46dff6731c99094
Berk Hess [Thu, 13 Oct 2011 08:16:11 +0000 (10:16 +0200)]
fixed pull constraining doing nothing, fixes #825
Change-Id: Iefadc0bb34241f91998e6d9f974150da54397045
Alexey Shvetsov [Tue, 11 Oct 2011 22:34:58 +0000 (02:34 +0400)]
Add CMakeLists.txt.user to gitignore
This will add CMakeLists.txt.user [file generated by qtcreator] to
gitignore
Change-Id: I5511a03dbbeaa4f376e8b1fd3218c82d77272167
Signed-off-by: Alexey Shvetsov <alexxy@omrb.pnpi.spb.ru>
Alexey Shvetsov [Tue, 11 Oct 2011 22:09:49 +0000 (02:09 +0400)]
Updated grompp.mdp for tutor/water
Updated grompp.mdp for tutor water that works with current gromacs
Issue #802
Change-Id: I3a80fa15ea08b25dd817b743331fa278aefb0d15
Signed-off-by: Alexey Shvetsov <alexxy@omrb.pnpi.spb.ru>
Roland Schulz [Tue, 11 Oct 2011 22:14:25 +0000 (18:14 -0400)]
Fixes MPI_GROUP_EMPTY for tMPI
Fixes breakege of tMPI version
introduced by commit
8cd25f0d45567b53027c7c4ab9bb2e8d5606932d
Change-Id: I8a79b863f5c1a25efebde55a597f6194889b4d2f
Szilárd Páll [Thu, 6 Oct 2011 23:30:50 +0000 (01:30 +0200)]
Merge "Removed duplicate CMake functionality" into release-4-6
Berk Hess [Thu, 6 Oct 2011 14:55:30 +0000 (16:55 +0200)]
fixed segv in parallel replica exchange
The last fix and commit of mdrun.c, ea3111, caused a segv with multisim
when the simulations were in parallel. Fixes #821
Change-Id: Ie3936b224d81dc3a304357bcd3438d2b4f5192e3
Mark Abraham [Thu, 6 Oct 2011 12:20:13 +0000 (23:20 +1100)]
Removed erroneous use of MASSV on BlueGene
MASSV does not provide BLAS or LAPACK (indeed, there is no standard
version for BlueGene), and is not useful otherwise. So it was erroneous
to require BLAS and LAPACK, and not helpful to link to MASSV.
Change-Id: I9b82875e9c6b02800e1ec6330f2a9006a563719d
Mark Abraham [Mon, 23 Nov 2009 02:17:06 +0000 (15:17 +1300)]
Fixes for GB kernel
Change-Id: I9108b02d032b6c7779c1d88b60e84a6e35e2025e
Roland Schulz [Tue, 4 Oct 2011 07:56:22 +0000 (03:56 -0400)]
Only create group if part of load communicator and thus the group is not empty
Unnessary groups are a problem if maximum number of group is limited (e.g. scalasca)
Change-Id: I2daca6414e82cca8d64d3dd5c08c6ae0060d07dc
Mark Abraham [Fri, 30 Sep 2011 08:01:04 +0000 (18:01 +1000)]
Removed duplicate CMake functionality
The FindXXX modules that I have removed are found in standard CMake
2.8.2, so we don't need to duplicate them if we choose to have a
similar minimum requirement for CMake. So far, I've arbitrarily chosen
2.8.0 as a compromise between modernity and widespread adoption.
See discussion in Redmine issue #815.
The logic for the work-around for Cray MPI is now in CMakeLists.txt
The test for the Git version has also moved into CMakeLists.txt
We had a minor change in the LibXml2 module that catered to some
weird/broken CMake environment Teemu had, but it looks to be
unnecessary for modern CMake.
The LAPACK find module is nearly the same as standard, apart from a
bug fix that we have found and reported upstream. So we have to
keep the duplicated code.
Change-Id: If312671392f9b77c0007c761ccb37688bcfc13b3
Mark Abraham [Tue, 27 Sep 2011 23:47:36 +0000 (09:47 +1000)]
Added tip for converting .xpm to something useful
Originally provided by Justin Lemkul on gmx-users.
Change-Id: I92e0d17095dca5a738dbb3188067d4394742e653