date | project | content | link |
|---|
| 01:13 today | OMPI | Commit by rusraink :: r22587 /trunk/opal/mca/carto/ (2 files in 2 dirs): - We don't configure-check for errno.h and don't need errno.h here...
| # |
| 00:31 today | OMPI | Commit by jjhursey :: r22586 /trunk/orte/tools/orte-restart/orte-restart.c: The ordering of the appfile matters. The ranks must be in increasing rank order, so that they get assigned the same ORTE name on restart.
cmr:v1.4.2 cmr:v1.5 | # |
| 23:07 yesterday | OMPI | Commit by jsquyres :: r22585 /vendor/libevent2/libevent-2.0.3-alpha/: Tag libevent 2.0.3-alpha | # |
| 22:57 yesterday | OMPI | Commit by jsquyres :: r22583 /vendor/libevent2/: Make a new directory for libevent 2.x so that we can play with it without disturbing the /vendor/libevent branch (since that is still used in all current OMPI releases). | # |
| 22:06 yesterday | OMPI | Commit by jjhursey :: r22580 /trunk/README: undo r22579 | # |
| 22:02 yesterday | OMPI | Commit by jjhursey :: r22579 /trunk/README: This is a test commit to test the auto-CMR functionality...
cmr:v1.4 cmr:v1.5.1:jjhursey | # |
| 20:56 yesterday | OMPI | Commit by bosilca :: r22578 /branches/v1.5/ompi/mca/coll/tuned/coll_tuned_module.c: From r22510: We enable the dynamic decision if the user force it via an MCA argument or set it in the decision file. In addition do a fine grain activation, i.e. per collective function.
This commit closes #2205. | # |
| 20:55 yesterday | OMPI | Commit by jsquyres :: r22577 /trunk/opal/config/opal_setup_libevent.m4: Add a check for strtoll for libevent.
Not having this check was causing distcheck errors on the OMPI tarball-build machine because it's still a 32-bit-default machine, so the evutil.c code was failing some #if conditionals (since it didn't think it had strtoll available). | # |
| 20:39 yesterday | OMPI | Commit by bosilca :: r22576 /branches/v1.5/configure.ac: From r22535: Fix CPPFLAGS issues with wrapper compilers. CPPFLAGS should be set for all projects and don't push CPPFLAGS into WRAPPER_FLAGS (for now, removes -D_REENTRANT).
This commit closes #2211. | # |
| 20:36 yesterday | OMPI | Commit by bosilca :: r22575 /branches/v1.5/ (5 files in 5 dirs): From r22547: AC_CHECK_FUNCS. Removes an annoying warning during application link on Catamount.
This commit closes #2228. | # |
| 20:34 yesterday | OMPI | Commit by bosilca :: r22574 /branches/v1.5/ (4 files in 4 dirs): From r22542: This fixes the failing mpic++/mpiCC MTT tests, bailing due to not finding symbol pthread_atfork, e.g. cxx-test-suite.
This commit closes #2217. | # |
| 20:33 yesterday | OMPI | Commit by bosilca :: r22573 /branches/v1.5/ (7 files in 7 dirs): From r22536:
Updates to make trunk run on Catamount again:
- Don't build the pstat component if all defines needed aren't there.
- Update platform file to work better
- Work around two places that depended on modex being operational
This commit closes #2215. | # |
| 20:32 yesterday | OMPI | Commit by bosilca :: r22572 /branches/v1.5/ (6 files in 6 dirs): From r22513: Move the user arguments up to be the first set of argv after the compiler argv tokens.
This commit closes #2207. | # |
| 20:28 yesterday | OMPI | Commit by bosilca :: r22571 /branches/v1.5/ (6 files in 4 dirs): From r22483: This bites us with make check (read MTT) on static builds (read jaguar) Only test using dlopen, when we do have it...
From 22534: Do an actual sh-test.
This commit closes #2197. | # |
| 20:24 yesterday | OMPI | Commit by bosilca :: r22570 /branches/v1.5/ (6 files in 5 dirs): From r22544: In multiple places in the code base, we expect opal_bitmap_is_set_bit to return either true or false...not a negative error code. If the index is out of range, this is effectively a "false" condition as the bit was clearly not set by the program. This revision is consistent with how the function was called in the OMPI code base. Will fix the test to match.
From r22545: Update the opal_bitmap test
From r22546: Grrr...remove debug
This commit closes #2226. | # |
| 20:07 yesterday | OMPI | Commit by bosilca :: r22569 /branches/v1.5/ (10 files in 8 dirs): From r22479: Originally the patch was to improve the error message, but when digging into the code I found a subtle bug. If the daemon does not tell the HNP what CRS component it used, then the HNP tries to figure it out from the metadata (this is an uncommon case). The path the HNP used was not complete, so it was unable to find the metadata information. This patch fixes this by adding the 'snapshot_reference' to the 'snapshot_location' which completes the path for this search.
This commit closes #2196. | # |
| 23:16 Sunday | OMPI | Commit by shiqing :: r22568 /trunk/ (4 files in 2 dirs): Check the Fortran compiler base on the build type (32 or 64 bit build), and set corresponding VC environment. | # |
| 06:22 Saturday | OMPI | Commit by eugene :: r22567 /trunk/orte/tools/orterun/orterun.1in: Add description of the sequential mapper (-mca rmaps seq) to the mpirun man page. | # |
| 22:27 Friday | OMPI | Commit by bosilca :: r22566 /trunk/ompi/mca/coll/tuned/coll_tuned_module.c: Remove an unused local variable. | # |
| 17:56 Friday | OMPI | Commit by shiqing :: r22565 /trunk/contrib/platform/win32/CMakeModules/setup_f77.cmake: Add the executable extension. | # |