GitLab

The ongoing march towards 64-bit clean continues.  Address areas where large
product of two ints might have overflowed.

Ref: #1767
Signed-off-by: Pavan Balaji <balaji@anl.gov>

Wesley Bland authored Feb 25, 2015 and Rob Latham committed Feb 26, 2015

Signed-off-by: Rob Latham <robl@mcs.anl.gov>

When an operation is completed in an MPIC function, it should no longer
have the error bits set since that information has now been captured in
the errflag. This prevents unnecessary assert failures.
Signed-off-by: Huiwei Lu <huiweilu@mcs.anl.gov>

This issue was added in commit [54362c00

].
Signed-off-by: Igor Ivanov <Igor.Ivanov@itseez.com>
Signed-off-by: Wesley Bland <wbland@anl.gov>

Requests that aren't for receive operations don't have anything in them,
so trying to process them only generates valgrind warnings with
potentially unsafe behavior.
Signed-off-by: Huiwei Lu <huiweilu@mcs.anl.gov>

The previous commits didn't take into account empty requests when
extracting the status. It also introduced a dumb bug that didn't get
tested first about a null pointer check.
Signed-off-by: Junchao Zhang <jczhang@mcs.anl.gov>

Part of converting the NBC code to use the MPIC_* functions requires
an MPIC_Issend function to exist. This adds it.
Signed-off-by: Huiwei Lu <huiweilu@mcs.anl.gov>

The MPIC helper functions have been using MPI_Comm and MPI_Request
objects instead of their MPID_* counterparts. This leads to a bunch of
unnecessary conversions back and forth between the two types of objects
and makes the work incompatible with other parts of the codebase
(non-blocking collectives for instance).

This patch converts all of the MPIC_* functions to use MPID_Comm and
MPID_Request and changes all of the collective calls to use them now
too.
Signed-off-by: Huiwei Lu <huiweilu@mcs.anl.gov>

The collective helper functions generally have an errflag that is used
when a failure is detected to allow the collective to continue while
also communicating that a failure occurred. That flag is now included
as a parameter for MPIC_Wait.

The rest of this commit is the refactoring necessary in the rest of the
helper functions to support the change.
Signed-off-by: Huiwei Lu <huiweilu@mcs.anl.gov>

The tag for send was ignored and recvtag incorrectly used in its place.
Signed-off-by: Wesley Bland <wbland@anl.gov>

When process fails, fault tolerance scheme takes a different path to
deal with MPI object reference counts than the existing one. Some
reference counts were not properly set in FT path so when configured
with --enable-g=all, some ft tests will show leaked context id, dirty
COMM, GROUP and REQUEST objects and so on when exit.

This patch fixes ft/shrink and ft/agree with "--enable-g=all". Stack
allocated objects of requests, communicators and groups will be freed by
FT.
Signed-off-by: Wesley Bland <wbland@anl.gov>

The MPI collectives get and set the errflag used by the collective
helper functions (MPIC_*). The possible values of the errflag changed,
so the collective functions need to appropriately set this value using
either MPIR_ERR_NONE (MPI_SUCCESS), MPIR_ERR_PROC_FAILED
(MPIX_ERR_PROC_FAILED), or MPIR_ERR_OTHER (MPI_ERR_OTHER).

This should allow collectives to correctly report process failures when
they occur now, fixing the FT tests that use collectives (see #1945).
Signed-off-by: Huiwei Lu <huiweilu@mcs.anl.gov>

The errflag value being used in the MPIC helper functions only
propagated whether or not an error occurred. It did not contain any
information about what kind of error occurred, which made returning the
correct error code after a process failure impossible.

This patch converts the binary value to an enum with three options:
MPIR_ERR_NONE
MPIR_ERR_PROC_FAILED
MPIR_ERR_OTHER

The original use of TRUE and false maps to MPIR_ERR_NONE and
MPIR_ERR_OTHER.

MPIR_ERR_PROC_FAILED indicates that the error occurred
because of a process failure. It uses the new bit set aside from the tag
space to track such information between processes.

This change required modifying lots of function signatures and type
declarations to use the new enum type, but these are actually not very
intrusive changes and shouldn't be a problem going forward.
Signed-off-by: Huiwei Lu <huiweilu@mcs.anl.gov>

We need to take another bit from the tag space to specify the difference
between a generic failure and a process failure. This patch modifies the
macros to handle this situation.
Signed-off-by: Huiwei Lu <huiweilu@mcs.anl.gov>

Back in the 3.1 series, we made the FT versions of all of the MPIC functions
default. However, we never changed the names of all of the states. This
removes the extra state names.

No reviewer.

The macro that called the bcast function left out an underscore in the
mpi_errno return value. This caused the test to always return MPI_ERR_OTHER
instead of the value being returned by the underlying bcast function.
Signed-off-by: Huiwei Lu <huiweilu@mcs.anl.gov>

The previous commit only adjusted the tag checking proceedure for MPIC_Recv,
but it should be the same for MPIC_Sendrecv and MPIC_Sendrecv_replace.
Signed-off-by: Huiwei Lu <huiweilu@mcs.anl.gov>

Tags were sometimes being tested when the communication call had already
failed, leading to bad asserts later due to uninitialized values. Sometimes,
these results would have ended up with an error anyway, but sometimes not.
Signed-off-by: Huiwei Lu <huiweilu@mcs.anl.gov>

This CVAR was deprecated in MPICH 3.1 and will now be removed for MPICH 3.2.
Signed-off-by: Junchao Zhang <jczhang@mcs.anl.gov>

Igor Ivanov authored Oct 10, 2014 and Pavan Balaji committed Oct 10, 2014

Fixed wrong parameter check condition for MPI_Iallgather and MPI_Iallgatherv
-1 is valid value for sendcount in case MPI_IN_PLACE
MPI spec says:
The in place option for intracommunicators is specified by passing the value
MPI_IN_PLACE to the argument sendbuf at all processes. In such a case, sendcount and
sendtype are ignored, and the input data of each process is assumed to be in the area where
that process would receive its own contribution to the receive buffer.
Signed-off-by: Igor Ivanov <Igor.Ivanov@itseez.com>
Signed-off-by: Pavan Balaji <balaji@anl.gov>

The warning message is:
    warning: ISO C90 forbids mixed declarations and code [-Wdeclaration-after-statement]

Fixes #2167
Signed-off-by: Antonio J. Pena <apenya@mcs.anl.gov>

If a process is dead, collectives still do all of the communictaions to
prevent a deadlock. However, if we just skip the part where the data is
updated in the allreduce_group function, we can let it be slightly more
resilient to failures and possibly even produce a correct answer in the
presence of a failure.
Signed-off-by: Junchao Zhang <jczhang@mcs.anl.gov>

Signed-off-by: Junchao Zhang <jczhang@mcs.anl.gov>

Adds a parameter to MPID_Comm_valid_ptr to take a second parameter that will
either cause the macro to ignore the revoke flag or not.
Signed-off-by: Junchao Zhang <jczhang@mcs.anl.gov>

Added the check on count and datatype on the following functions:

  - MPI_Iallreduce
  - MPI_Ialltoall
  - MPI_Ibcast
  - MPI_Igather
  - MPI_Ireduce
  - MPI_Iscan

(ibm) D198793

Fixes #2139
Signed-off-by: Michael Blocksome <blocksom@us.ibm.com>
Signed-off-by: Ken Raffenetti <raffenet@mcs.anl.gov>

Antonio J. Pena authored Jul 03, 2014 and Kenneth Raffenetti committed Jul 07, 2014

Moved the weak,alias attribute declarations from header files to the
implementation. Complies with the requirement that alias targets are
defined in the same compilation unit.

Fixes #2002
Signed-off-by: Ken Raffenetti <raffenet@mcs.anl.gov>

(ibm) D198243
Signed-off-by: Michael Blocksome <blocksom@us.ibm.com>
Signed-off-by: Wesley Bland <wbland@anl.gov>

If the size of the input buffers is 0, don't bother to check if they are
aliased.

Fixes #2124
Signed-off-by: Antonio J. Pena <apenya@mcs.anl.gov>

Lisandro Dalcin authored Jun 26, 2014 and Sangmin Seo committed Jun 26, 2014

The patch provides a better implementation where each group does a
intra-reduce concurrently, exchanges the reduce result, and broadcasts
in the group. The current implementation had a problem of serializing
intra-reduces of each group.

Fixes ticket #2074.
Signed-off-by: Sangmin Seo <sseo@anl.gov>

If the user isn't using MPI_IN_PLACE when they should, this check will do a
better job of warning them about it.

See #2049
Signed-off-by: Antonio J. Pena <apenya@mcs.anl.gov>

In the MPIC_Sendrecv functions, the status object should always be defined
since we use it internally. This won't have any impact on performance since
the default is always to have FT collectives turned on anyway, but it will
prevent a crash when someone overwrites that default.

Fixes #2026
Signed-off-by: Sangmin Seo <sseo@anl.gov>

Fixes the following warning when --enable-fast is enabled:

src/mpi/coll/bcast.c:1012:23: warning: suggest braces around empty body
in an 'if' statement [-Wempty-body]

See #1966.
Signed-off-by: Pavan Balaji <balaji@mcs.anl.gov>

Pavan Balaji authored Mar 31, 2014 and Kenneth Raffenetti committed Apr 01, 2014

Maintain a list of files that go into each library.  If a particular
binding is not enabled, the list variable still exists, but will just
be empty.  This simplifies the management of which files/symbols go
into which library.

Move all MPI_ symbols to the libmpi library and all other symbols to
the libpmpi library.  All Fortran 77 symbols go into libmpif77.so,
while C symbols go into libmpi.so.  There are some exceptions, such as
status_f2c, which are handled by the Fortran code but used in C.  Our
Fortran 90 build only creates a few symbols and uses the f77 symbols
for everything else.  These few symbols go into libmpifort.so.

Also update compiler wrappers to link to correct libraries.  mpif77
should now link with libmpif77.  mpif90 links with both libmpifort and
libmpif77, since our F90 build still keeps the core Fortran library
symbols in libf77.

We completely ignored the F77 library earlier.  This was OK because
all of the Fortran symbols were ending up in libmpi.  Now that we have
separated out the symbols to the right library, we now need to link to
libmpif77 as well.

Also added inter-library dependencies.

libmpi has a dependency on several internal libraries: libmpl, libopa.
libmpicxx did not have a dependency on libmpi, added.
libmpif77 did not have a dependency on libmpi, added.
libmpifort did not have a dependency on libmpi, added.

This dependency model is sufficient for C and F77, but not for C++ and
F90.  The C and F77 libraries contain all the symbols the application
relies on, but the F90 and C++ libraries don't.  In the case of F90,
symbols such as mpi_bcast are missing and are borrowed from the F77
library.  In the case of C++, mpicxx.h contains calls directly to C
functions (such as MPI_Reduce_local), which get embedded into the
application.

Fixes #2023.
Signed-off-by: Ken Raffenetti <raffenet@mcs.anl.gov>

Simply ran the new ./maint/check_copyright.bash script.

Fixes #2032.
Signed-off-by: Ken Raffenetti <raffenet@mcs.anl.gov>

The current implementation of SMP-awareness in MPI_Barrier was in
MPIR_Barrier_impl.  This makes the default implementation of barrier
SMP-aware.  However, if a device overrides barrier and then calls back
the default implementation through MPIR_Barrier, it can no longer take
advantage of SMP-awareness.  This patch moves the SMP-aware
implementation to MPIR_Barrier_intra, which is called by
MPIR_Barrier_impl through MPIR_Barrier, in the default implementation.
See #1957.
Signed-off-by: Michael Blocksome <blocksom@us.ibm.com>

Use MPIR_Bcast as an inline function, and let the compiler do its job,
instead of duplicating the code.
Signed-off-by: Michael Blocksome <blocksom@us.ibm.com>

Junchao Zhang authored Dec 29, 2013 and Pavan Balaji committed Dec 31, 2013

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

Junchao Zhang authored Dec 29, 2013 and Pavan Balaji committed Dec 31, 2013

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

Junchao Zhang authored Dec 29, 2013 and Pavan Balaji committed Dec 31, 2013

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