The PGI compiler unnecessarily complains (issues a warning) when 0
(which it should be automatically converting to the null pointer
constant) is an operand in a conditional expression. For example,
there are two instances (lines 444 and 459) in opal/mca/memory/
ptmalloc2/hooks.c:
444 and
The PGI C compiler complains (issues a warning) for the #pragma ident
in opal/runtime/opal_init.c:
PGC-W-0281-Pragma ignored - string expected after #pragma ident (../
opal/util/sys_limits.h: 58)
This is because the PCI C compiler does not (whereas, curiously, the
PGI C++ compiler does) s
Both the Intel and PGI compilers complain (issue a warning) about the
arguments to the LT_STRLEN() macro in opal/libltdl/ltdl.c:
ltdl.c(1268): warning #279: controlling expression is constant
archive_name = MALLOC (char, LT_STRLEN (name) + LT_STRLEN
(libext) + 2);
Both PGI C and C++ have issues with the inline assembly in OpenMPI.
configure includes the code below to disable the inline assembly test
for PGI C. The same code needs to be inserted for the C++ inline
assembly test as well.
{ $as_echo "$as_me:${as_lineno-$LINENO}: checking if $CC su
The PGI compiler version number parsing in the libtool.m4 files is
incorrect. It should be like the parser in configure, i.e., there
should be a period between "]]" and "*"
[baker@hydra openmpi-1.5rc5]$ grep '1-5' configure
*pgCC\ [1-5].* | *pgcpp\ [1-5].*)
[baker@hydra openmpi-
Yes - the selection logic now only picks ONE module to run.
On Thu, Aug 19, 2010 at 12:01 PM, Wesley Bland wrote:
> So just to clarify, this means that we don't need to worry about having
> more than one errmgr module handling a single failure and therefore don't
> have to set the stack_state (w
So just to clarify, this means that we don't need to worry about having more
than one errmgr module handling a single failure and therefore don't have to
set the stack_state (which is now gone anyway). Am I reading this
correctly?
Thanks,
Wesley
On Thu, Aug 19, 2010 at 9:09 AM, wrote:
> Author
