number of recv system call by caching the data. Each endpoint has a buffer
(the size is an MCA parameter) that can be use as a cache. Before each receive
operation this buffer is added at the end of the iovec list. All data that are
not expected by the fragment will go in this cache. If the cache contain data
all subsequent receive will just memcpy the data into the BTL buffers.
The only drawback is that we will spin around the receive_handle until all the
cached data is readed by the PML layer. This limitation come from the fact that
the event library is unable to call us if there is no events on the socket.
Therefore we are unable to keep the data in the cache until the next loop
into the progress engine.
This commit was SVN r8398.
x86_64 as most opterons do. So add test to see how many bytes a long is,
and if it's 8 bytes, assume we're on an AMD64 instead of an IA32 for the
purposes of assembly.
This commit was SVN r8384.
both mmap and munmap), adjusting the configure script so that the
component will only be activated on systems that use ptmalloc2 in the
first place -- ie, Linux
* Remove the malloc_hooks component - it became an unworkable solution
once threads and such were considered.
* Remove malloc_interpose component - it never worked quite right and
was not going to be able to intercept malloc, so it wasn't going to
be useful for OMPI's purposes.
* Update tests a little bit to match recent memory hooks api
issues - still needs a bit of work.
This commit was SVN r8381.
the script did not complete successfully
- ROMIO likes to default to using the system compiler if no compiler is
specified. This can lead to using a different compiler for ROMIO as
for Open MPI, which is not always a good thing. Reset the default to
behave the same way Open MPI does.
This commit was SVN r8380.
- since we now have three places where we need to add a -l<foo> if
and only if it isn't already in something in LIBS that provides
symbol bar...
This commit was SVN r8370.
tarballs: one with greek and one without -- without running "svn up"
in between. This guarantees that we have two tarballs of the same SVN
R number (e.g., vA.B.Crc7 and vA.B.C if rc7 turns out to be perfect).
Add option --greekonly if you know that there's no chance that this
tarball will be perfect, and therefore there's no point in making the
non-greek tarball.
This commit was SVN r8367.
- Need to make sure that SIZE_MAX exists as a constant if stdint.h
doesn't exist
- struct timeval is defined in unistd.h on IRIX, so need to include
that headerfile where ever struct timeval is used.
This commit was SVN r8361.
- when processing out of order list - reset match to null on each iteration
- check matched request type and if probe - complete probe and queue fragment
on unexpected list
This commit was SVN r8339.
ability at this point to roll back and build without libltdl if we thought we
we going to have libltdl (we can configure without libltdl if we know from
the start it won't be there, from --disable-dlopen).
This commit was SVN r8336.
following for the rationale:
http://www.open-mpi.org/community/lists/users/2005/11/0391.php
Note that mpi.h only receives AC_DEFINE's (not AC_SUBST's), so we
still have to AC_DEFINE to get the value in there -- so we AC_DEFINE
something that we give to a typedef.
This commit was SVN r8326.
