GitLab

Masamichi Takagi authored Dec 04, 2013 and Pavan Balaji committed Dec 05, 2013

Add PAC_SET_HEADER_LIB_PATH(dcfa) in
src/mpid/ch3/channels/nemesis/subconfigure.m4
so that --with-dcfa, --with-dcfa-lib and --with-dcfa-include
is enabled.
Signed-off-by: Pavan Balaji <balaji@mcs.anl.gov>

Masamichi Takagi authored Dec 04, 2013 and Pavan Balaji committed Dec 05, 2013

Signed-off-by: Pavan Balaji <balaji@mcs.anl.gov>

Masamichi Takagi authored Dec 02, 2013 and Pavan Balaji committed Dec 03, 2013

Signed-off-by: Pavan Balaji <balaji@mcs.anl.gov>

Signed-off-by: Masamichi Takagi <masamichi.takagi@gmail.com>

Signed-off-by: Masamichi Takagi <masamichi.takagi@gmail.com>

Signed-off-by: Masamichi Takagi <masamichi.takagi@gmail.com>

Masamichi Takagi authored Nov 29, 2013 and Pavan Balaji committed Dec 03, 2013

dcfa_poll.c and dcfa_send.c are modified.
Signed-off-by: Pavan Balaji <balaji@mcs.anl.gov>

Masamichi Takagi authored Nov 29, 2013 and Pavan Balaji committed Dec 03, 2013

Signed-off-by: Pavan Balaji <balaji@mcs.anl.gov>

Masamichi Takagi authored Nov 29, 2013 and Pavan Balaji committed Dec 03, 2013

Replace manipulating req->status.count with MPIR_STATUS_SET_COUNT.
Signed-off-by: Pavan Balaji <balaji@mcs.anl.gov>

Masamichi Takagi authored Nov 27, 2013 and Pavan Balaji committed Dec 03, 2013

DCFA is IB drivers for Xeon Phi running on McKernel. McKernel is an OS
developed by the University of Tokyo.
Signed-off-by: Pavan Balaji <balaji@mcs.anl.gov>

Signed-off-by: Wesley Bland <wbland@mcs.anl.gov>

Warnings related to unused variables intended to be used for debugging, but
always declared.
Signed-off-by: Wesley Bland <wbland@mcs.anl.gov>

Fixes the following warnings:

PGC-W-0095-Type cast required for this conversion
(./src/mpid/ch3/include/mpidrma.h: 703)

PGC-W-0095-Type cast required for this conversion
(./src/mpid/ch3/include/mpidrma.h: 864)
Signed-off-by: Wesley Bland <wbland@mcs.anl.gov>

Fixes the following warning in ch3u_handle_connection.c whith --enable-strict:

src/mpid/ch3/src/ch3u_handle_connection.c:317:194: warning: suggest braces
around empty body in an 'if' statement [-Wempty-body]
Signed-off-by: Wesley Bland <wbland@mcs.anl.gov>

Fixes the following warning when compiling tcp_init.c with --enable-strict:

src/mpid/ch3/channels/nemesis/netmod/tcp/tcp_init.c: In function
'GetSockInterfaceAddr':
src/mpid/ch3/channels/nemesis/netmod/tcp/tcp_init.c:248:9: warning:
dereferencing type-punned pointer will break strict-aliasing rules
[-Wstrict-aliasing]

Reported in ticket #1966
Signed-off-by: Wesley Bland <wbland@mcs.anl.gov>

Added support for Ibsend and persistent sends, and fixed all other cases by
clearing out the dgb-next field of send requests. Closes #1932.
Signed-off-by: Ken Raffenetti <raffenet@mcs.anl.gov>

Delete code for zero-size data transfer in packet handlers
of Put/Accumulate/Accumulate_Immed/Get_AccumulateResp/GetResp/
LockPutUnlock/LockAccumUnlock, because they are redundant.

(Note that packet handlers of LockPutUnlock and LockAccumUnlock
are for single operation optimization in passive RMA)

Zero-size data transfer has already been handled when issuing
RMA operations (L146, L258, L369 in src/mpid/ch3/src/ch3u_rma_ops.c
and L50 in src/mpid/ch3/src/ch3u_rma_acc_ops.c). RMA operation
routines will directly exit if data size is zero.
Signed-off-by: Wesley Bland <wbland@mcs.anl.gov>

This reverts commit 676c29f9.

Addresses #1932. Includes:
  - MPI_Bsend/MPI_Ibsend
  - Several collectives
  - Some RMA operations
  - MPI_Dist_graph_create
Signed-off-by: Ken Raffenetti <raffenet@mcs.anl.gov>

MPIU_Assert at L2311 checks if rma_ops_list is empty before exiting
MPIDI_Win_flush. It causes /test/mpi/threads/rma/multirma to fail
because while one thread is executing the loop of poking progress
engine at L2293 ~ L2302, another thread may enqueue new RMA operations
to rma_ops_list.

rma_ops_list has already been checked for empty before exiting
MPIDI_CH3I_Do_passive_target_rma (L2724) to ensure that all enqueued
operations are issued out, therefore it does not need to be checked
again here.
Signed-off-by: Wesley Bland <wbland@mcs.anl.gov>

Fixes #1781
Signed-off-by: Wesley Bland <wbland@mcs.anl.gov>

Fixed a set of compiler warnings related to MPI_T code when compiled
with -Wall, including "assignment discards 'const' qualifier from pointer
target type", "unused variable", "shadows a global declaration" etc.
Signed-off-by: Antonio J. Pena <apenya@mcs.anl.gov>

Drop payload into a union paired with a double, and payload will end up
double-aligned.  Can't use an anonymous union, though: that's c11.

named-union requires updating any place that used payload.  There aren't too
many, though.
Signed-off-by: Pavan Balaji <balaji@mcs.anl.gov>

Rob Latham authored Oct 18, 2013 and Pavan Balaji committed Oct 31, 2013

clang carped about too many parens around (( a == b))

Fixes #1929
Signed-off-by: Pavan Balaji <balaji@mcs.anl.gov>

Jeff Hammond authored Oct 15, 2013 and Pavan Balaji committed Oct 31, 2013

goal: make MPICH count-safe
problem: casting to int when value exceeds INT_MAX
solution: remove all "(int)" in front of values assigned to the count
field
Signed-off-by: Rob Latham <robl@mcs.anl.gov>

Dancing around with various schemes to detect/deal-with compiler padding
in structs so the 'payload' field ends up double-aligned.  The new
"large type , large count" work introduced a pointer into the nemisis
packet, so we have to be a bit more dynamic about how or if we pad out
the structure.

- put the members of MPID_nem_pkt_header inside a struct, so subsequent
  payload does not "fall into" compiler-introduced padding
- use this updated datatype to determine the size, and if we need to pad
  or not.
- use the result of the configure test to introduce padding, or omit the
  padding from the definition if the amount of padding is '0'
Signed-off-by: Pavan Balaji <balaji@mcs.anl.gov>

Dave Goodell authored Apr 04, 2013 and Rob Latham committed Oct 31, 2013

Also includes random fixes to `-Wshorten-64-to-32` warnings which might
need to be teased out.
Signed-off-by: Pavan Balaji <balaji@mcs.anl.gov>

Based on an Intel contributed patch.  The idea is to use the bits from
the cancelled field to extend the count, rather than increasing the
count datatype itself.
Signed-off-by: Ken Raffenetti <raffenet@mcs.anl.gov>

Fixes to the bit manipulation based on feedback from Artem Yalozo @
Intel.

Fixes to the naming convention based on feedback from Bill Gropp.
Signed-off-by: William Gropp <wgropp@illinois.edu>

Junchao Zhang authored Oct 25, 2013 and Pavan Balaji committed Oct 26, 2013

To adapt to naming for control variables in MPI_T.
Signed-off-by: Pavan Balaji <balaji@mcs.anl.gov>

Junchao Zhang authored Oct 18, 2013 and Pavan Balaji committed Oct 26, 2013

Signed-off-by: Pavan Balaji <balaji@mcs.anl.gov>

Junchao Zhang authored Oct 18, 2013 and Pavan Balaji committed Oct 26, 2013

Signed-off-by: Pavan Balaji <balaji@mcs.anl.gov>

Junchao Zhang authored Oct 18, 2013 and Pavan Balaji committed Oct 26, 2013

New code should use interfaces provided by new MPI_T impl.
Signed-off-by: Pavan Balaji <balaji@mcs.anl.gov>

This was changed to 1024 to give space for larger error strings.  But
this caused an ABI breakage for a number of MPICH partners.  This
patch reverts the size back the size to what we had in older mpich
releases.

An initial version of this patch was contributed by Intel.
Signed-off-by: Ken Raffenetti <raffenet@mcs.anl.gov>

Optimize the case where the origin and target both use basic
datatypes.  In this case, we assume that the data is aligned correctly
for the appropriate datatype and perform a direct assignment instead
of a memory copy.
Signed-off-by: Xin Zhao <xinzhao3@illinois.edu>

The check was originally in the ch3 layer, but doesn't seem to use any
ch3 specific information.  This macro will be useful at the upper
layers for optimizations, e.g., in the localcopy routine.
Signed-off-by: Xin Zhao <xinzhao3@illinois.edu>

We already do a check for shared memory before calling the
shared-memory specific functions.  This patch simplifies some of those
redundant checks.
Signed-off-by: Xin Zhao <xinzhao3@illinois.edu>

The memory barrier ensures that all load/store operations issued
directly to shared memory are complete.
Signed-off-by: Xin Zhao <xinzhao3@illinois.edu>

Xin Zhao authored Sep 26, 2013 and Pavan Balaji committed Sep 26, 2013

Because when SHM is allocated, it is possible that orig rank and target rank
are on different nodes, in such situation operations are not done yet and
win_flush cannot exit.
Signed-off-by: Pavan Balaji <balaji@mcs.anl.gov>

Xin Zhao authored Sep 26, 2013 and Pavan Balaji committed Sep 26, 2013

Check shm_allocated flag in win_flush to determine if do full
memory barrier or not.
Signed-off-by: Pavan Balaji <balaji@mcs.anl.gov>

Xin Zhao authored Sep 26, 2013 and Pavan Balaji committed Sep 26, 2013

Check both shm_allocated flag and VC's node_id for
req-based operations.
Signed-off-by: Pavan Balaji <balaji@mcs.anl.gov>