- 05 Jan, 2014 1 commit
-
-
Rob Latham authored
This reverts commit 3146e104.
-
- 03 Jan, 2014 3 commits
-
-
Sameh Sharkawi authored
Signed-off-by:
Michael Blocksome <blocksom@us.ibm.com>
-
Rob Latham authored
In romio-standalone fixes, I zapped the inclusion of mpitypedefs.h, but the "ensure this large value fits into pointer without overflowing" check relied on some MPICH-internal macros/routines. Only Blue Gene driver uses those routines, so nightly tests didn't flag it. No reviewer.
-
This tests for the functionality of MPI_Dims_create when the number of processes is very large (1 billion in this case). Signed-off-by:
Ken Raffenetti <raffenet@mcs.anl.gov>
-
- 02 Jan, 2014 3 commits
-
-
Fixes #1992. Signed-off-by:
Ken Raffenetti <raffenet@mcs.anl.gov>
-
Pavan Balaji authored
We were setting the default eager threshold inside nemesis, instead of in ch3. This was causing other ch3 channels, notably sock, to get a garbage default value for the eager threshold. Signed-off-by:
Wesley Bland <wbland@mcs.anl.gov>
-
Pavan Balaji authored
For non rank-0 processes, store the stdin socket in a dummy variable instead of passing NULL. Passing NULL will cause the create_process function to close the STDIN socket, allowing the process to reuse that socket. However, if an application reopens stdin, it causes an incorrect socket (which is not STDIN) to be closed. This is technically a user application bug, but this is a safe-guard to workaround that. Fixes #1622. Signed-off-by:
Ken Raffenetti <raffenet@mcs.anl.gov>
-
- 01 Jan, 2014 1 commit
-
-
Junchao Zhang authored
Type checking for MPI_2INT is commented out because Clang 3.3 flags struct {int i1; int i2;} as different from int[2]. But actually these two types are of the same layout. Clang gives a type mismatch warning for a definitely correct code like the following: int in[2], out[2]; MPI_Reduce(in, out, 1, MPI_2INT, MPI_MAXLOC, 0, MPI_COMM_WORLD); So, we disable type checking for MPI_2INT until Clang fixes this bug. Fixes #1993 Signed off by Rajeev Thakur <thakur@mcs.anl.gov>
-
- 31 Dec, 2013 5 commits
-
-
(ibm) F189037 Signed-off-by:
Michael Blocksome <blocksom@us.ibm.com>
-
Signed-off-by:
Pavan Balaji <balaji@mcs.anl.gov>
-
Signed-off-by:
Pavan Balaji <balaji@mcs.anl.gov>
-
Signed-off-by:
Pavan Balaji <balaji@mcs.anl.gov>
-
- add check epoch - changes after intergrating Sameer's get_accumulate - update to remove dependency on mpid data structure (ibm) F189038 Signed-off-by:
Michael Blocksome <blocksom@us.ibm.com> Minor modifications by Pavan to minimize the changes in mpir_request.c, and to reduce duplication. Collective requests and RMA requests are handled the same way. Signed-off-by:
Pavan Balaji <balaji@mcs.anl.gov>
-
- 30 Dec, 2013 9 commits
-
-
Antonio J. Pena authored
A couple of unused variables when assert does nothing were causing compiler warnings. Fixed by declaring them as potentially unused. Signed-off-by:
Pavan Balaji <balaji@mcs.anl.gov>
-
Antonio J. Pena authored
Signed-off-by:
Pavan Balaji <balaji@mcs.anl.gov>
-
Antonio J. Pena authored
Signed-off-by:
Pavan Balaji <balaji@mcs.anl.gov>
-
Antonio J. Pena authored
Mark those variables exclusively being used for assertions as potentially unused to avoid compiler warnings when the assertion macro does nothing. These show up with --enable-fast. Signed-off-by:
Pavan Balaji <balaji@mcs.anl.gov>
-
Antonio J. Pena authored
These only show up with --enable-fast Signed-off-by:
Pavan Balaji <balaji@mcs.anl.gov>
-
Antonio J. Pena authored
Fixes the following warnings: "src/binding/cxx/initcxx.cxx", line 444: warning: missing return statement at end of non-void function "MPI::MPIR_Comm_delete_attr_cxx_proxy" "src/binding/cxx/initcxx.cxx", line 496: warning: missing return statement at end of non-void function "MPI::MPIR_Comm_copy_attr_cxx_proxy" Signed-off-by:
Pavan Balaji <balaji@mcs.anl.gov>
-
Antonio J. Pena authored
Fixes the following warnings: "src/binding/cxx/initcxx.cxx", line 721: warning: variable "err" was set but never used "src/binding/cxx/initcxx.cxx", line 745: warning: variable "err" was set but never used "src/binding/cxx/initcxx.cxx", line 770: warning: variable "err" was set but never used Signed-off-by:
Pavan Balaji <balaji@mcs.anl.gov>
-
Antonio J. Pena authored
Fixes the following warnings when compiling socksm.c with --enable-strict: src/mpid/ch3/channels/nemesis/netmod/tcp/socksm.c: In function 'alloc_sc_plfd_tbls': src/mpid/ch3/channels/nemesis/netmod/tcp/socksm.c:183:5: warning: the comparison will always evaluate as 'true' for the address of 'MPID_nem_tcp_g_lstn_sc' will never be NULL [-Waddress] src/mpid/ch3/channels/nemesis/netmod/tcp/socksm.c:184:5: warning: the comparison will always evaluate as 'true' for the address of 'MPID_nem_tcp_g_lstn_plfd' will never be NULL [-Waddress] Reported in ticket #1966 Signed-off-by:
Pavan Balaji <balaji@mcs.anl.gov>
-
Antonio J. Pena authored
Fixes the following warnings (with --enable-strict): src/mpid/ch3/src/ch3u_rma_acc_ops.c: In function 'MPIDI_Get_accumulate': src/mpid/ch3/src/ch3u_rma_acc_ops.c:31:5: warning: unused variable 'mpiu_chklmem_stk_sz_' [-Wunused-variable] src/mpid/ch3/src/ch3u_rma_ops.c: In function 'MPIDI_Accumulate': src/mpid/ch3/src/ch3u_rma_ops.c:350:5: warning: unused variable 'mpiu_chklmem_stk_sz_' [-Wunused-variable] See ticket #1966 Signed-off-by:
Pavan Balaji <balaji@mcs.anl.gov>
-
- 29 Dec, 2013 3 commits
-
-
Test all compilers, not just gcc. Use nesting safe push/pop flag macros. Signed-off-by:
Pavan Balaji <balaji@mcs.anl.gov>
-
Signed-off-by:
Charles Archer <archerc@us.ibm.com> Signed-off-by:
Michael Blocksome <blocksom@us.ibm.com> Signed-off-by:
Pavan Balaji <balaji@mcs.anl.gov>
-
(ibm) D192180: Changing MPIU_Malloc size from unsigned to size_t Signed-off-by:
Charles Archer <archerc@us.ibm.com> Signed-off-by:
Michael Blocksome <blocksom@us.ibm.com> Signed-off-by:
Pavan Balaji <balaji@mcs.anl.gov>
-
- 25 Dec, 2013 1 commit
-
-
Pavan Balaji authored
This patch tests the code when we are duping overlapping communicators (with two processes each) essentially forming a ring that should deadlock with blocking Comm_dup. With Comm_idup, this should work correctly as per the MPI standard.
-
- 20 Dec, 2013 5 commits
-
-
Pavan Balaji authored
-
Antonio J. Pena authored
Signed-off-by:
Pavan Balaji <balaji@mcs.anl.gov>
-
Antonio J. Pena authored
Fixes #1765. Patch contributed to mpich-discuss by Ian Hutchinson with minor style fixes. Signed-off-by:
Pavan Balaji <balaji@mcs.anl.gov>
-
Antonio J. Pena authored
Signed-off-by:
Pavan Balaji <balaji@mcs.anl.gov>
-
Rob Latham authored
as fallout from deferred open, we were doing the "scalable stat" from the first aggregator, but broadcasting the result from rank 0. now we broadcast from the process that actually has a meaningful value.
-
- 19 Dec, 2013 9 commits
-
-
Junchao Zhang authored
Old perl (e.g., Perl 5.10) does not support $_ in readdir. The fix is using an explicit variable instead. Fixes #1991 Signed-off-by:
Ken Raffenetti <raffenet@mcs.anl.gov>
-
Junchao Zhang authored
When printing Fortran subroutine prototypes to the f90 module files, use dummy argument (i.e., C formal argument) names defined by MPI, instead of short names like v0, v1. The side-effect is that longer names may make a line exceed the length limit set by Fortran. So we also need to properly wrap long lines. The revise is to support Fortran keyword arguments. Signed-off-by:
Pavan Balaji <balaji@mcs.anl.gov>
-
Junchao Zhang authored
Signed-off-by:
Pavan Balaji <balaji@mcs.anl.gov>
-
Junchao Zhang authored
Revised Perl subroutine ReadInterface(), which is used to extract function prototypes from a file. Previously, ReadInterface only extracted argument types of C functions. With this revise, it also extracts argument names. This is needed to support Fortran keyword arguments. Signed-off-by:
Pavan Balaji <balaji@mcs.anl.gov>
-
Junchao Zhang authored
Signed-off-by:
Pavan Balaji <balaji@mcs.anl.gov>
-
Signed-off-by:
Pavan Balaji <balaji@mcs.anl.gov>
-
Fixes #1928 Signed-off-by:
Pavan Balaji <balaji@mcs.anl.gov>
-
Fixes #1963 Signed-off-by:
Pavan Balaji <balaji@mcs.anl.gov>
-
MPIR_CVAR_PROCTABLE_PRINT related code was originally commented out. Revert the code so that MPIR_CVAR_PROCTABLE_PRINT does what it means. Signed-off-by:
Pavan Balaji <balaji@mcs.anl.gov>
-