Timeline
07/25/09:
- 13:58 Ticket #689 (Duplicate flags) reopened by
- As my comment above states, there is still a problem with duplicate flags. …
- 13:09 Ticket #691 (parallel make support in MPE) reopened by
- MPE's parallel make support needs to work when it is built standalone. …
- 13:02 Ticket #747 (mpich2 x64 bug (shm AV)) created by
- […]
- 11:41 Ticket #691 (parallel make support in MPE) closed by
- duplicate: Parallel builds for the entire MPICH2 stack are being handled in ticket …
- 11:40 Ticket #689 (Duplicate flags) closed by
- fixed: This has been fixed in r4998. We are still running into problems with …
07/24/09:
- 17:55 Ticket #746 (pthread barrier not always available) created by
- The test test/mpi/threads/pt2pt/threads.c makes use of pthread_barrier_xxx …
- 17:30 Changeset [5052] by
- Add MPITEST_PROGRAM_WRAPPER functionality to the test suite (for …
- 17:26 Ticket #197 (rreq->dev.recv_pending_count uninitialized) closed by
- fixed: Fixed in r5051. FWIW, this was much easier to track down with the new …
- 17:24 Changeset [5051] by
- Fix for ticket #197, uninitialized usage of rreq->dev.recv_pending_count. …
- 17:01 Changeset [5049] by
- Make sure NUMVERSION never has leading zeroes. Otherwise C will interpret …
- 16:23 Changeset [5048] by
- Process-core mapping was broken in r5047 due to uninitialized variables.
- 15:39 Changeset [5047] by
- Accidentally inter-changed library and symbol names in Hydra configure.
- 15:26 Changeset [5046] by
- Separate out process binding capabilities to basic and topology aware. …
- 12:56 Changeset [5044] by
- executable name has changed to th_taskmaster
- 11:39 Ticket #745 (Files checked into SVN are also in the svn:ignore property) closed by
- fixed: Fixed in r5043. Resolving.
- 11:39 Changeset [5043] by
- Remove files that are checked into the svn, but accidentally made it into …
- 11:08 Ticket #745 (Files checked into SVN are also in the svn:ignore property) created by
- […] untitled-part.html
- 10:21 Changeset [5042] by
- If the runtime system detects that process-core binding is not possible, …
- 03:58 Ticket #721 (mpich2, gfortran, F90 files and MPE) reopened by
- […]
- 01:07 Changeset [5041] by
- commented out AC_PREREQ in all non-standalone configure.in, i.e. except …
- 00:24 Ticket #721 (mpich2, gfortran, F90 files and MPE) closed by
- worksforme: resolving this for now until we hear more from the user.
07/23/09:
- 19:54 Ticket #551 (MacOSX & jumpshot) closed by
- fixed: verified the right-click works on MacOSX, resolving…
- 17:14 Ticket #744 (Cannot run a simple program with ch3:nemesis:mx) created by
- […]
- 16:15 Changeset [5039] by
- merge trunk to kumudb r4870:5038
- 15:58 Changeset [5038] by
- recording merge info from merge commit r4870.
- 15:50 Changeset [5037] by
- changes to mpif77
07/22/09:
- 19:37 Ticket #743 (mpich2-1.1.1 : FreeBSD-7.2-Release mpiexec gives collective abort) closed by
- wontfix
- 18:08 Ticket #743 (mpich2-1.1.1 : FreeBSD-7.2-Release mpiexec gives collective abort) created by
- […] untitled-part.html
- 17:07 Changeset [5036] by
- Merge r5032 mpich2-1.1.1 --> trunk. Fixes MPICH2LIB_xFLAGS broken in …
- 17:02 Changeset [5035] by
- Merge r5022 mpich2-1.1.1 --> trunk. Fixes duplicated CFLAGS with …
- 16:31 Changeset [5033] by
- A simple dynamic throttle for wintcp poll/select. Improves latency on …
- 16:27 Changeset [5032] by
- Restore MPICH2LIB_xFLAGS functionality accidentally disabled in r4998. …
- 11:42 Changeset [5031] by
- sync the bytecode with changes up to r5030.
- 11:16 Ticket #742 (pb with configure) closed by
- wontfix
- 11:06 Changeset [5030] by
- cleaned up of the java code so it can be compiled with java6 compiler …
- 09:40 Ticket #742 (pb with configure) created by
- […] c.txt
07/21/09:
- 23:30 Milestone mpich2-1.1.1 completed
- 17:39 Changeset [5027] by
- 16:44 Changeset [5025] by
- Re-tag the mpich2-1.1.1 release (for real this time) to include the fix in …
- 16:43 Changeset [5024] by
- Undo the bad copy in r5032.
- 16:42 Changeset [5023] by
- Re-tag the mpich2-1.1.1 release to include the fix in r5022.
- 16:32 Changeset [5022] by
- Fix many repeated CFLAGS when --enable-strict is specified. The …
- 14:36 Changeset [5021] by
- Make changes corresponding to r5005 in windows configure script. 1) Since …
- 14:03 Ticket #733 (Duplicate names for tests) closed by
- fixed: Fixed in r5020. Randomly picked test names, since I couldn't come up with …
- 14:01 Changeset [5020] by
- Rename tests to make them globally unique.
- 13:42 Changeset [5019] by
- Turn off browse information in VS projs - release configs. Also turn off …
- 11:36 Ticket #733 (Duplicate names for tests) created by
- From: mpich2-dev-bounces@… …
- 09:37 Changeset [5016] by
- smpd fixed.
07/20/09:
- 17:50 Changeset [5015] by
- fflush on stderr should have been on stdout in MPIU_Msg_printf Thanks to …
- 17:21 Changeset [5014] by
- Move the struct alignment check code to a compile-time autoconf macro.
- 16:27 Changeset [5012] by
- Update confdb and mpl externals.
- 16:17 Changeset [5011] by
- Use the 1.1.1 release tag for the nightly test.
- 15:27 Changeset [5007] by
- tagging the mpich2-1.1.1 release
- 15:23 Changeset [5006] by
- updating the Version file to 1.1.1 for the release
- 15:21 Changeset [5005] by
- Get rid of many of the cross-compile variables that are no longer needed …
- 15:19 Changeset [5004] by
- creating a branch for any fixes/changes to mpich2-1.1.1
- 15:17 Changeset [5003] by
- update externals: confdb in romio and mpl in hydra
- 15:16 Changeset [5002] by
- update confdb external in mpl
- 14:56 Changeset [5001] by
- romio works with parallel build.
- 12:50 Changeset [5000] by
- Fixing misplaced brackets - added in r4982 - in *NEEDS_DECL() macro. …
- 11:54 Changeset [4999] by
- Handling unhandled defns in windows configure script. lseek64 is not used …
- 11:35 Changeset [4998] by
- The build system was checking if the C compiler needs to link to specific …
07/19/09:
- 23:55 Changeset [4996] by
- Adding some confdb macros to manage flags, though they are currently …
- 22:55 Changeset [4995] by
- this is needed for MAKE to work in test/mpi, otherwise it fails. didn't …
- 22:51 Changeset [4994] by
- this is needed for MAKE to work in test/mpi, otherwise it fails. didn't …
- 18:54 Changeset [4993] by
- updated thread_multiple comment
07/18/09:
- 13:13 Ticket #297 (make -j N support) closed by
- duplicate: This is being handled in ticket #711. Resolving this as a duplicate.
- 09:54 Ticket #706 (MPICH2 errors) closed by
- worksforme: Resolving until we hear back from the user.
07/17/09:
- 14:44 Changeset [4990] by
- Setting version number to 1.1.1rc1.
- 12:28 Changeset [4989] by
- Merged 4938:4988
- 11:21 Changeset [4988] by
- Output commented line in case of thread-level mismatch.
- 11:19 Changeset [4987] by
- We don't return a failure when the MPI implementation does not support …
07/16/09:
- 17:10 Changeset [4985] by
- Altered rank calculation to be done only once and stored into an array on …
- 16:54 Changeset [4984] by
- Updating the GUID for windows installers in preparation for the 1.1.1 rc …
- 16:48 Changeset [4983] by
- Adding user specified proc binding scheme for SMPD on windows. Also added …
- 16:08 Ticket #720 (lseek64 in FreeBSD) closed by
- fixed: RobL has fixed this in r4982.
- 15:51 Changeset [4982] by
- configure-time test for lseek64, though it's only used on a few rare file …
- 15:16 Changeset [4981] by
- Fixed bug with running multiple jobs. Corrected logic error in callback …
- 15:14 Ticket #704 (Problems running romio test, coll_perf.c, on windows) closed by
- worksforme
- 13:43 Changeset [4980] by
- Update confdb and mp.
- 12:56 Ticket #724 (MPIR_Scatterv typo) closed by
- fixed: This has been fixed in r4978, including a regression test. If anyone …
- 12:50 Ticket #732 (mpd -gdb error with centos) created by
- This ticket has been split off from #697. Here are the relevant bits: > …
- 12:47 Ticket #697 ("mpirun -gdb" produces a time out) closed by
- fixed: This bug has been fixed in r4979. I'll open a new ticket for the centos …
- 12:46 Changeset [4979] by
- increased the timeout for waiting for debugged app to respond to mpd. …
- 12:42 Changeset [4978] by
- Regression test and fix for ticket #724, in-place MPI_Scatterv. Reviewed …
- 12:42 Changeset [4977] by
- Whitespace and bracing fixes for the scatterv.c test. Normally I wouldn't …
- 12:35 Ticket #717 (Please Help Me!) closed by
- worksforme
- 11:09 Ticket #731 (32-bit Intel C++ compile failures in the mpich2 test suite) created by
- When building with: […] Many of the C++ tests in the MPICH2 test suite …
- 10:43 Ticket #730 (processes "nap" with --enable_thread=multiple and process affinity) created by
- A user reported that his processes took a ~1 minute nap and took a long …
07/15/09:
- 22:12 Changeset [4976] by
- Run nightly tests on all 48 nodes.
- 20:05 Changeset [4975] by
- ch3:sctp minor updates
- 17:44 Changeset [4974] by
- Checking fields collectively was accidentally removed in r4915. Adding it …
- 17:30 Changeset [4973] by
- Workaround for ticket #729 (strict compiles break opa). This is not a …
- 17:01 Changeset [4972] by
- Added O_TRUNC for file opening.
- 13:55 Changeset [4970] by
- Handling unhandled defns in windows configure - No *GCC*, No *APPLE* on …
- 11:39 Ticket #729 (opa configure fails in nightlies) created by
- There are some configure failures in opa in the new nightly tests that ran …
- 08:48 Ticket #724 (MPIR_Scatterv typo) created by
- […]
07/14/09:
- 18:49 Changeset [4968] by
- source code modification to use list of default colors when the requested …
- 18:38 Changeset [4967] by
- source code modification to add support of Cntl-Click as RightMouseClick? …
- 16:47 Ticket #719 (spawnmanyarg fails with gforker, Intel compilers, 64 bit) closed by
- fixed: Solutions 1 & 3 are implemented in r4966. Resolving.
- 16:46 Changeset [4966] by
- Fix for ticket #719 - gforker+spawnmanyarg test failures. This change …
- 16:13 Changeset [4965] by
- f90 fixed in buildiface
- 16:05 Changeset [4964] by
- fixes to make install in src/binding/f90 directory
- 14:06 Changeset [4961] by
- commented
- 13:26 Ticket #684 (a trouble with 64-bit system, Alex Mazheika) closed by
- worksforme: I'm not sure why there would be any significant performance discrepancy in …
- 09:55 Changeset [4960] by
- Adding mtestthread.o to the linker dependencies for multisend . Without …
07/13/09:
- 17:35 Ticket #722 (shared libs not updated on a rebuild) created by
- I noticed that even with "make dependencies" the .so and .la files weren't …
- 17:32 Ticket #690 (Nemesis' MPIU_Memcpy broken) closed by
- fixed: Fixed in r4959. We were missing "memory" in the clobber list of the …
- 17:30 Changeset [4959] by
- adding 'memory' to clobber list in asm memcpy implementations
- 17:29 Changeset [4958] by
- added svn ignore to src/util/ex
- 16:29 Ticket #721 (mpich2, gfortran, F90 files and MPE) created by
- […]
- 16:18 Changeset [4957] by
- the f90 works. everhything works except mpe and romio.
- 16:02 Changeset [4956] by
- Fix bad file descriptor macro names in confdb/fortran90.m4. This actually …
- 16:00 Changeset [4955] by
- # Modifying multisend test to use MTest wrappers for threads # Adding …
- 15:26 Changeset [4954] by
- Update openpa to r83 (v1.0.1) to pick up bug fixes.
- 11:51 Changeset [4953] by
- updated the configure help output in the installation manual and tweaked …
- 11:37 Changeset [4952] by
- reformatted and clarified some configure help text, namely that the …
- 11:30 Changeset [4951] by
- Fix the check for net/if.h in the ch3 configure.in. Checking for net/if.h …
- 11:08 Ticket #720 (lseek64 in FreeBSD) created by
- The ROMIO code uses lseek64, which is not supported in FreeBSD. We need a …
- 10:54 Ticket #719 (spawnmanyarg fails with gforker, Intel compilers, 64 bit) created by
- […]
- 08:42 Changeset [4950] by
- clarified wording of MPI_Init scalability fix
- 08:40 Changeset [4949] by
- this test fails every once in a while. clarified the print statement that …
- 03:09 Changeset [4948] by
- fixed error message.
- 02:59 Ticket #617 (Fwd: [mpich2-dev] Null MPE log routines) closed by
- fixed: fixed by r4947.
- 02:57 Changeset [4947] by
- patched for req#617.
- 01:20 Ticket #717 (Please Help Me!) created by
- […]
07/12/09:
- 23:51 Ticket #343 (Request to enhance jumpshot script for portability) closed by
- fixed: Use env variables JAVA_HOME, JRE_HOME, JVM to set java runtime and env …
- 23:48 Changeset [4946] by
- added the support to use env variables JAVA_HOME, JRE_HOME, JVM to set …
- 17:35 Ticket #715 ([help] failed to run mpi programs) closed by
- worksforme: You should always use mpicc on all platforms. Different platforms use …
- 13:59 Ticket #715 ([help] failed to run mpi programs) created by
- […]
07/11/09:
- 11:23 Ticket #710 (Intel abort tests failing with smpd-nemesis) closed by
- fixed: Works. Resolving.
07/10/09:
- 17:27 Ticket #694 (Fedora 11, MPICH2-1.1 collective abort of all ranks, killed by signal 4) closed by
- fixed: This has been fixed in openpa revision 81 and in mpich2 r4945. -d
- 17:23 Changeset [4945] by
- updated openpa external
- 17:22 Changeset [4944] by
- the parallel building of dirs works. f90 still does not work
- 17:05 Ticket #712 (Strange problem with pi running on opensuse11.1) closed by
- wontfix
- 15:02 Changeset [4943] by
- Adapt PVFS2 listio code for zoidfs. compiles and links; still needs …
- 13:06 Changeset [4942] by
- Update openpa to r80 in order to pick up the all-executable target. No …
- 12:35 Changeset [4941] by
- Correct a typo introduced in a prior change in this branch. No reviewer.
- 12:31 Ticket #714 (GUI binaries missing in Windows XP64) created by
- […]
- 12:22 Changeset [4940] by
- Return an error code when conn is closed while trying to read on a socket. …
- 11:52 Ticket #712 (Strange problem with pi running on opensuse11.1) created by
- […]
- 10:44 Changeset [4939] by
- merged 4906:4938
- 10:07 Ticket #711 (Parallel builds in MPICH2) created by
- […]
- 09:55 Ticket #710 (Intel abort tests failing with smpd-nemesis) created by
- […]
- 07:27 Ticket #684 (a trouble with 64-bit system, Alex Mazheika) reopened by
- Hi Alexej, I suspect that your remaining problem with F90 was fixed in …
- 03:04 Ticket #708 (mpich2-1 on an IBM AIX 5.2 machine) created by
- […]
07/09/09:
- 18:06 Changeset [4938] by
- reduced nloops and max message size since the test times out on some …
- 16:04 Changeset [4937] by
- # Integrating the Executive progress engine from MSMPI code base - r3692 - …
- 15:43 Changeset [4936] by
- More fixes to the binding auto-detection.
- 15:29 Changeset [4935] by
- More fixes to the binding auto-detection.
- 15:15 Changeset [4934] by
- working with j 2 and j 4
- 15:06 Changeset [4933] by
- Change -version to -info.
- 15:01 Changeset [4932] by
- Added command-line support to pick a binding library.
- 15:00 Changeset [4931] by
- Code cleanup.
- 14:45 Changeset [4930] by
- The binding API is available irrespective of whether a binding library is …
- 14:35 Changeset [4929] by
- Move out plpa into the bind directory.
- 14:27 Changeset [4928] by
- More support for allowing multiple binding libraries.
- 13:22 Changeset [4927] by
- all except romio and mpe works. the libstamp business cleaned. now …
- 12:53 Changeset [4926] by
- Each directory can now potentially include itself in the include path. The …
- 12:46 Changeset [4925] by
- Generalized process binding detection a bit to allow for alternate …
- 12:35 Changeset [4924] by
- Error in the configure test was causing process bindingto not get detected …
- 11:26 Ticket #695 (FW: Python 2.6 (from Fedora 11 distro): 'popen2' isdeprecated) closed by
- fixed: These deprecation warning messages were suppressed in r4923.
- 11:24 Changeset [4923] by
- Suppress more deprecation warnings in mpd (tickets #362,#695,#697) …
- 09:07 Changeset [4922] by
- needs printf no errors
07/08/09:
- 16:37 Changeset [4921] by
- Replaced instances of pid with rank for clarity.
- 16:10 Changeset [4920] by
- Fixed bug with pid generation for filenames in mpiexec.c Removed …
- 15:34 Changeset [4919] by
- Added a simplified version of the thread message rate measurement …
- 14:31 Ticket #706 (MPICH2 errors) created by
- […]
- 11:30 Ticket #688 (how to install mpif90 command) closed by
- worksforme
- 11:30 Ticket #699 (the problem about installation of mpif90) closed by
- worksforme
- 11:26 Ticket #692 (help) closed by
- wontfix
- 11:26 Ticket #702 (please help me!) closed by
- invalid
- 11:17 Ticket #281 (closesocket failed error when running an MinGWcompiled executable) closed by
- fixed: This bug should be fixed by r4906 (The MPICH2 1.1.1 release will include …
- 09:36 Ticket #704 (Problems running romio test, coll_perf.c, on windows) created by
- Hi all, I am using coll_perf.c in order to evaluate some optimizations at …
- 01:49 Ticket #702 (please help me!) created by
- […]
- 00:59 Ticket #668 (PWD and getcwd() coflict in Hydra) closed by
- fixed: Fixed as far as possible in r4918. We don't catch all possible cases, but …
- 00:57 Changeset [4918] by
- Check for the similarity of the PWD environment and getcwd(). If they are …
- 00:33 Changeset [4917] by
- Casting of integers to enums is not allowed; we need to cast it to …
- 00:01 Ticket #681 (Hydra Warnings with Intel Compiler) closed by
- fixed: This should be fixed in r4914-r4916. Resolving.
- 00:00 Changeset [4916] by
- Update confdb and MPL versions across the board. This, together with r4914 …
07/07/09:
- 23:57 Changeset [4915] by
- While checking if the compiler accepts strict flags, do a stricter test …
- 23:54 Changeset [4914] by
- Warning stomp for the Intel compiler (ticket #681).
- 23:34 Changeset [4913] by
- Nemesis' implementation of memcpy seems to be broken for small unaligned …
- 16:58 Changeset [4912] by
- Modified make_filename to add in leading zeroes for pid. Added if-checking …
- 16:52 Changeset [4911] by
- profile library, static, and shared library works, mpe and romio unknown
- 14:12 Ticket #701 (mpich2 *Fortran* debugging with MS Vis Stud) closed by
- fixed: Hi, Let us know if you need further help. (PS: You can use the …
- 13:11 Changeset [4910] by
- Finished implementing redirection of stdout/stderr to files based on …
- 12:03 Ticket #701 (mpich2 *Fortran* debugging with MS Vis Stud) created by
- […]
- 11:34 Ticket #700 (Bug in src/pm/smpd/smpd_launch_process.c) closed by
- fixed: Applied patch from Edric in r4909 . -Jayesh
- 11:34 Changeset [4909] by
- Don't indefinitely wait for an exit from child processes in the SIGCHLD …
- 10:14 Ticket #622 (Intel/ANL MPI test suite release) reopened by
- Nothing is planned right now for the Intel/ANL test suite release. …
- 08:09 Ticket #700 (Bug in src/pm/smpd/smpd_launch_process.c) created by
- […]
- 04:21 Ticket #699 (the problem about installation of mpif90) created by
- […]
07/06/09:
- 17:22 Changeset [4908] by
- merged 4747:4906... again.
- 15:29 Ticket #655 (Tests with Intel compilers leaving behind processes) closed by
- fixed: Hopefully the fix for ticket #633 (SMPD+Nemesis+Intel_compilers fail in …
- 15:27 Changeset [4907] by
- Everything (ncluding the shared library works). The profile library does …
- 15:25 Ticket #633 (SMPD nightlies with Intel compilers fail) closed by
- fixed: This is fixed in r4906. See the check-in comment for details. (PS: The …
- 15:09 Ticket #696 (Possible problem in persistent requests) closed by
- fixed: Bill fixed the Intel tests.
- 14:43 Changeset [4906] by
- 1) When stdin is closed and there is no data to read don't send any data …
- 11:22 Ticket #688 (how to install mpif90 command) reopened by
- After each failed inpath configure, you need to do a "make distclean" …
07/05/09:
- 18:50 Ticket #697 ("mpirun -gdb" produces a time out) created by
- […]
- 12:13 Changeset [4905] by
- Correct the use of MPITEST_AINT (should be integer) and also increase size …
- 12:13 Changeset [4904] by
- Use the correct file for fortran tests for the byroutinef target
- 12:12 Changeset [4903] by
- Fixed missing ranlib - needed in OSX
07/04/09:
- 12:58 Changeset [4902] by
- Increased size of testname character variable to fit the test name
- 12:55 Changeset [4901] by
- Corrected some buffer addresses
- 12:53 Changeset [4900] by
- increased character buffer for test name
07/03/09:
- 14:23 Changeset [4899] by
- all executable works including cpi in examples and mpiexec in pm
- 12:49 Changeset [4898] by
- executable works except mpiexec
07/02/09:
- 17:15 Ticket #696 (Possible problem in persistent requests) created by
- We routinely fail one of the Fortran tests in the Intel suite when using …
- 16:34 Changeset [4897] by
- Adding nemesis & removing stale channels from mpiexec wrapper utils for …
- 16:22 Changeset [4896] by
- Add missing 'eval' for the f90 sub-configure invocation. Thanks to Steve …
- 15:12 Ticket #615 (mpich2.module) closed by
- fixed: Fixed mpich2.module.in so that it is self-contained and usable by modules. …
- 14:54 Changeset [4895] by
- bugfix for PATH which should point to bindir not anywhere in datadir.
- 14:19 Changeset [4894] by
- removed the installation of mpich2.module until Deji has a complete …
- 14:19 Ticket #695 (FW: Python 2.6 (from Fedora 11 distro): 'popen2' isdeprecated) created by
- […]
- 14:15 Ticket #694 (Fedora 11, MPICH2-1.1 collective abort of all ranks, killed by signal 4) created by
- […]
- 14:02 Ticket #693 (Porting the Executive from MSMPI code to unix) created by
- This is a placeholder to track our work on porting the Executive (A …
- 13:29 Changeset [4891] by
- squashed float == warnings in opminloc and opmaxloc, and refactored code
- 10:44 Changeset [4890] by
- removed debugging code that shouldn't have been checked in
- 10:19 Changeset [4889] by
- From Pascal Deveze <pascal.deveze@…>: Update lustre for ROMIO's …
- 09:58 Changeset [4888] by
- squashed more warnings
07/01/09:
- 23:23 Changeset [4887] by
- Warning stomp.
- 17:39 Changeset [4886] by
- make envmods more reasonable. But mpich2.module still lacks a version ID …
- 16:18 Ticket #692 (help) created by
- […]
- 15:37 Ticket #622 (Intel/ANL MPI test suite release) closed by
- invalid: Just to clarify my earlier comment, all the lines from ln 20-28 on pg 247 …
- 15:01 Ticket #521 (Availability of fortran lib with lowercase symbols with MPICH2 64-bit on ...) closed by
- fixed: Added fortran lib with lower case symbols for 64-bit platforms in r4883 . …
- 14:58 Changeset [4883] by
- Adding fortran library with lower case symbols for 64-bit windows …
- 14:34 Ticket #691 (parallel make support in MPE) created by
- reminder to add parallel make support in MPE
- 14:15 Changeset [4882] by
- Created a separate directory for random projects related to MPICH2, but …
- 09:04 Changeset [4881] by
- Forgot to add a dependency on mtest for typecommit.
- 07:25 Changeset [4880] by
- Update confdb and mpl version numbers. We got to automate this!
06/30/09:
- 19:09 Changeset [4879] by
- comented out debugging statements.
- 18:23 Changeset [4878] by
- Separate warning flags onto separate lines for easier VCS tracking. No …
- 18:23 Changeset [4877] by
- Add -Wno-format-zero-length to the strict arguments. No reviewer.
- 17:12 Changeset [4876] by
- tweaked warning options some more
- 16:46 Ticket #500 (Fwd: Problem with MPE with IBM MPI) closed by
- fixed: mpich2's configure.in has updated to use autoconf-2.63 as of #4861
- 16:20 Ticket #664 (FW: Problem with MPI_Type_commit() and assert insegment_ops.c) closed by
- fixed: Fixed by RobR in r4874 and r4875
- 16:14 Changeset [4875] by
- fix for bug similar to #664 in block indexed flatten code.
- 16:12 Changeset [4874] by
- fix for bug #664.
- 16:02 Changeset [4873] by
- minor cleanup to help with readability while debugging.
- 15:50 Changeset [4872] by
- Got rid of incorrect "unused" attributes on input variables that are used.
- 15:35 Ticket #354 (smpd + mpiexec_rsh.c) closed by
- wontfix: It is not trivial to handle the failure case (if path not existent back …
- 15:30 Changeset [4871] by
- Adding doc regarding usage of relative paths when running jobs with -rsh …
- 15:27 Changeset [4870] by
- Merge from trunk to kumudb r4748:r4869
- 15:06 Changeset [4869] by
- simplemake.in modified so that all-executable target is added in every …
- 14:53 Ticket #636 (MPICH2 and .NET issues) closed by
- wontfix: Hi, Sorry for the delayed response. Since MPICH2 is compiled with VS …
- 14:53 Changeset [4868] by
- Added the datatype failure reported by IBM to the test suite.
- 14:52 Ticket #615 (mpich2.module) reopened by
- This file can't be used. I have been using "module" even now. I don't …
- 14:44 Changeset [4867] by
- --enable-strict now works just like --enable-posix=strict, so we don't …
- 14:43 Changeset [4866] by
- --enable-strict=posix is really a superset of --enable-strict=all, so …
- 14:41 Changeset [4865] by
- stomped some warnings
- 14:39 Changeset [4864] by
- added and removed warning options to reduce the number of nonessential …
- 13:59 Ticket #179 (Inconsistent and incorrect use of MPIR_Nest_incr and MPIR_Nest_decr) closed by
- fixed: RobL thinks this is fixed.
- 13:40 Ticket #615 (mpich2.module) closed by
- wontfix: The Fedora guys gave us this file to include in MPICH2. They use it for …
- 13:36 Ticket #605 (Error compiling Fortran programs on Windows) closed by
- fixed: Resolving until we hear further
- 13:26 Ticket #683 (转发: problems with installing mpich2 under cygwin) closed by
- wontfix: Resolving until we hear further.
- 13:24 Ticket #684 (a trouble with 64-bit system, Alex Mazheika) closed by
- wontfix: Resolving until we hear further
- 13:24 Ticket #688 (how to install mpif90 command) closed by
- fixed: Resolving until we hear further.
- 11:18 Changeset [4863] by
- Mostly BillG: introduce quiet-make from MPICH2 to ROMIO
- 11:03 Changeset [4862] by
- Get rid of the *hackish* version of import file generator for windows dlls …
06/29/09:
- 17:08 Ticket #363 (Re: MPI_IN_PLACE bug in Allgatherv in MPE's collchk) closed by
- fixed: The libmpe_f2cmpi.a works with inplacef.f, statuses.f as well as FLASH3 …
- 16:49 Changeset [4861] by
- Use AC_CHECK_LIB for detecting -lsocket and -lnsl, instead of …
- 16:32 Ticket #686 (MPICH2-1.1 : make fails with undefined reference to 'pthread_create') closed by
- fixed: Great! Thanks for the logs. This problem has been fixed in the svn as of …
- 16:26 Changeset [4860] by
- Re-enable auto-detection of process binding in Hydra by default. Also, …
- 15:15 Ticket #680 (Warnings on 64-bit systems re MPIUI_Memcpy) closed by
- fixed: should now be fixed
- 15:14 Changeset [4859] by
- updating mpl
- 15:13 Changeset [4858] by
- updating confdb
- 15:09 Changeset [4857] by
- updated install-sh to newer version from autoconf 1.11
- 14:59 Ticket #665 (MPICH2 with Boost) closed by
- fixed: After some discussion here, I've finally added flags that only get used …
- 14:55 Changeset [4856] by
- Added a new set of environment variables (MPICH2_MPICC_FLAGS, etc.), that …
- 14:54 Changeset [4855] by
- deleting generated file
- 14:53 Changeset [4854] by
- removed error on implicit declaration warning option which broke …
- 14:51 Changeset [4853] by
- Update confdb and mpl versions. Waiting for svn relative path externals to …
- 14:34 Changeset [4852] by
- Disabling a default selecn for the fortran symbol name convention in …
- 14:23 Changeset [4851] by
- removed some overly strict warning options
- 14:22 Changeset [4850] by
- Remove restrict from memcpy implementation to quiet compiler warnings
- 13:54 Changeset [4849] by
- Adding log_mpi_util.c to windows proj and makefiles
- 13:38 Ticket #690 (Nemesis' MPIU_Memcpy broken) created by
- […]
- 05:34 Changeset [4848] by
- Fix for the init
- 01:23 Changeset [4847] by
- Update confdb and mpl versions everywhere.
- 01:17 Ticket #689 (Duplicate flags) created by
- In MPICH2's top-level configure.in, when arguments such as --enable-strict …
- 01:09 Changeset [4846] by
- Disable process binding by default, till we figure out why this isn't …
- 01:08 Changeset [4845] by
- This patch is really a hack to get around some of the issues with the …
06/28/09:
- 21:43 Changeset [4844] by
- Cleaned up --enable-strict options added some additional options that were …
- 21:39 Ticket #688 (how to install mpif90 command) created by
- […]
- 20:55 Ticket #687 (which library should be used in windows with cvf) created by
- […]
- 18:48 Changeset [4843] by
- It looks like just checking for the existence of pthread.h is not …
- 17:45 Changeset [4842] by
- Fix for Hydra's alt-install. "pmi_proxy" was not getting installed in the …
- 17:35 Changeset [4841] by
- Typo in r4840 in the test to include -lpthread.
- 17:32 Changeset [4840] by
- Don't link with libpthread.a if threading is not requested for. This is …
- 12:58 Ticket #686 (MPICH2-1.1 : make fails with undefined reference to 'pthread_create') created by
- […]
06/27/09:
- 15:08 Changeset [4839] by
- Moved most of the README content to the wiki.
- 10:00 Ticket #680 (Warnings on 64-bit systems re MPIUI_Memcpy) reopened by
- This doesn't seem to have fixed it. There are as many warnings as before. …
- 01:05 Changeset [4838] by
- Error in the configure for platforms that don't support PLPA.
- 00:50 Changeset [4837] by
- Auto-detect if PLPA works on the system. If it does, turn on process …
- 00:04 Ticket #682 (Hydra: Environment overflows static buffers) closed by
- fixed: This should be fixed in r4836. Resolving.
- 00:04 Changeset [4836] by
- Calculate an upper bound of the number of executable arguments we expect …
06/26/09:
- 23:34 Changeset [4835] by
- The proxy arguments are the same for all partitions, so we don't need to …
- 20:59 Changeset [4834] by
- Update MPL to the latest version.
- 20:58 Changeset [4833] by
- Update confdb version to the latest.
- 20:57 Changeset [4832] by
- Update confdb version to the latest.
- 20:56 Changeset [4831] by
- Use confdb macros instead of redoing the same thing in configure.in
- 20:37 Ticket #685 (test/mpi/perf/commcreatep fails with Hydra) closed by
- duplicate: This failure is occurring because the number of temporary arguments Hydra …
- 19:55 Ticket #677 (Hydra processes left behind) closed by
- fixed
- 19:09 Ticket #665 (MPICH2 with Boost) reopened by
- No, there isn't any flag that is only used by mpicc and friends, but not …
- 17:09 Ticket #665 (MPICH2 with Boost) closed by
- fixed
- 16:45 Ticket #671 (taskmaster test in the wrong directory) closed by
- fixed: The problem was that the test still called MPI_Init_thread with …
- 16:44 Changeset [4830] by
- fixes build with single-threaded implementations, ticket #671
- 16:37 Ticket #645 (MPICH2 MPI_Get_elements bug and proposed fix) closed by
- fixed: Thanks. I have applied your patch in r4829. I agree with Bill that the …
- 16:33 Changeset [4829] by
- applied patch to MPI_Get_elements provided by Cray, ticket #645
- 16:23 Changeset [4828] by
- Reverting to 4747 from merge 4826
- 16:06 Ticket #679 (can I translate the content of mpich2 wiki to chinese?) closed by
- wontfix
- 14:58 Changeset [4827] by
- the static library gets generated conditionally.
- 14:49 Changeset [4826] by
- Merged from trunk to dkim 4747:4825
- 13:19 Ticket #685 (test/mpi/perf/commcreatep fails with Hydra) created by
- […]
- 10:22 Changeset [4825] by
- the static libraries: libmpich.a, libfmpich.a, libmpichcxx.a, …
- 10:17 Ticket #489 (Fallback Singleton Init Warning) closed by
- wontfix: I'm not sure this "bug" is solvable this way. I'm resolving this for now, …
- 09:47 Changeset [4824] by
- Added -help, usage, error info for -stdout-file and -stderr-file options. …
- 03:35 Ticket #684 (a trouble with 64-bit system, Alex Mazheika) created by
- […]
- 00:47 Ticket #683 (转发: problems with installing mpich2 under cygwin) created by
- […]
06/25/09:
- 20:30 Changeset [4823] by
- Protect fn_fail in HAVE_PROC_BINDING to avoid warnings.
- 18:10 Ticket #629 (FW: New/modified tests in mpich2-1.1b1 not following MTest convention) closed by
- fixed: Applied this patch in r4822
- 18:10 Changeset [4822] by
- applied Jim Hoekstra's patch to fix missing use of MTest convention in …
- 18:06 Ticket #682 (Hydra: Environment overflows static buffers) created by
- […]
- 17:52 Ticket #681 (Hydra Warnings with Intel Compiler) created by
- […]
- 16:13 Ticket #680 (Warnings on 64-bit systems re MPIUI_Memcpy) closed by
- fixed: This should be fixed in r4821.
- 16:13 Changeset [4821] by
- removing restrict from parameters. Should fix #680
- 15:50 Ticket #467 (MPICH2 does not work with Boost.MPI) closed by
- fixed: I've corrected the symbolic link in r4819 and verified that the latest …
- 15:21 Ticket #680 (Warnings on 64-bit systems re MPIUI_Memcpy) created by
- […]
- 14:27 Changeset [4820] by
- Oops, the Lustre fixes are not tested yet.
- 13:26 Changeset [4819] by
- The symbolic link for mpic++ was incorrectly setup. This should fix ticket …
- 13:10 Changeset [4818] by
- Fix to the MPMD launch problem.
- 12:51 Changeset [4817] by
- added contig datatype optimization to pack and unpack. Reviewed by …
- 02:55 Changeset [4816] by
- Updated version number in preparation for the 1.1.1 release.
- 02:51 Changeset [4815] by
- Revert the PWD usage introduced in r4782. Perl's chdir() was not updating …
Note: See TracTimeline
for information about the timeline view.
