Interesting - still, I see no reason for OMPI to fail just because of that. We
can run just fine with the uid, so I'll make things a little more flexible.
Thanks for tracking it down!
On Jan 22, 2014, at 7:54 PM, Paul Hargrove wrote:
> Not lacking getpwuid():
>
>
Hi,
I'm trying to port openmpi-1.6.5 in l4/fiasco. I checked the libmpi.a. I
did the " ar t libmpi.a " in my terminal. I can't find the source file (.c)
of some object files created in libmpi.a, such as:
ompi_bitmap.o
op_predefined.o
convertor.o
copy_functions.o
copy_functions_heterogeneous.o
The reason appears to be:
checking if Fortran compiler supports BIND(C) with LOGICAL params... no
The requested files are attached.
-Paul
On Wed, Jan 22, 2014 at 7:46 PM, Jeff Squyres (jsquyres) wrote:
> Can you send me the configure output and config.log from this
Not lacking getpwuid():
[phh1@biou2 BLD]$ grep HAVE_GETPWUID */include/*_config.h
opal/include/opal_config.h:#define HAVE_GETPWUID 1
I also can't see why the quoted code could fail.
The following is working fine:
[phh1@biou2 BLD]$ cat q.c
#include
#include
#include
#include
int main(void) {
Can you send me the configure output and config.log from this build? I'd like
to see why it chose not to build the mpi_f08 module.
On Jan 22, 2014, at 10:08 PM, Paul Hargrove wrote:
>
> On Wed, Jan 22, 2014 at 6:31 PM, Jeff Squyres (jsquyres)
>
Here is the offending code:
/* get the name of the user */
uid = getuid();
#ifdef HAVE_GETPWUID
pwdent = getpwuid(uid);
#else
pwdent = NULL;
#endif
if (NULL != pwdent) {
user = strdup(pwdent->pw_name);
} else {
orte_show_help("help-orte-runtime.txt",
On Wed, Jan 22, 2014 at 6:31 PM, Jeff Squyres (jsquyres) wrote:
> But just to confirm: you said that your pathscale compilers *do* compile
> 1.7.3 -- including the mpi_f08 module -- with no problems? That would be a
> little surprising, because those same >=32 character
On yet another test platform I see the following:
$ mpirun -mca btl sm,self -np 1 examples/ring_c
--
Open MPI was unable to obtain the username in order to create a path
for its required temporary directories. This type of
Ok, here's my update: I fixed a bunch of issues in the Fortran support today;
most are minor, but they took a while to verify (and some are slated for v1.7.5
because they aren't critical). I also added the ability to disable building
the mpi_f08 module.
Here's what's on the trunk / v1.7, and
Yep, this is 'zactly what I was contemplating today. Chris from Pathscale is
checking into what their symbol length limits are for me and promises to get
back to me shortly. So I'm holding off on the symbol length issues until then.
But just to confirm: you said that your pathscale compilers
On Mon, Jan 20, 2014 at 12:24 PM, Barrett, Brian W wrote:
> Ugh, this is a 32 bit RISC problem; we don't have a 64 bit atomic on a 32
> bit platform. People are supposed to check to see if there's 64 bit atomic
> support, but that clearly hasn't been happening. I've fixed
On Wed, Jan 22, 2014 at 8:50 AM, Jeff Squyres (jsquyres) wrote:
> Can you do me a favor and cd into ompi/mpi/fortran/use-mpi-f08 and try to
> manually "make type_create_indexed_block_f08.lo" and see if it also
> complains? That's a 32 character name -- let's see if the
Creating nightly hwloc snapshot git tarball was a success.
Snapshot: hwloc dev-40-g183a7da
Start time: Wed Jan 22 21:01:01 EST 2014
End time: Wed Jan 22 21:03:37 EST 2014
Your friendly daemon,
Cyrador
On Wed, Jan 22, 2014 at 2:22 PM, Paul Hargrove wrote:
> My ia64 asm is a bit rusty, but I'll give a quick look if/when I can.
I had a look (in v1.7) and this is what I see:
$cat -n IA64.asm | grep -A14 opal_atomic_cmpset_acq_64:
70 opal_atomic_cmpset_acq_64:
71
On Wed, Jan 22, 2014 at 2:40 PM, Paul Hargrove wrote:
>
> + PathScale and Open64 which fail building in ompi/mpi/fortran/use-mpi-f08/
>
I implied, but forgot to state the following explicitly:
Both PathScale and Open64 can build the Fortran support present in 1.7.3
(verified
Update: I've been working all day on Fortran issues (pulling on one
Paul-Fortran--sweater-thread revealed several other issues :-( ).
I'll be sending an update soon...
On Jan 22, 2014, at 5:40 PM, Paul Hargrove wrote:
>
> On Wed, Jan 22, 2014 at 1:33 PM, Ralph Castain
On Wed, Jan 22, 2014 at 1:33 PM, Ralph Castain wrote:
> My main concern with 1.7.4 at the moment stems from all the Fortran
> changes we pushed into that release - this occurred *after* 1.7.3, and so
> those problems represent a regression in the 1.7 series.
Unless I am
On Wed, Jan 22, 2014 at 1:59 PM, Ralph Castain wrote:
> Huh - afraid I can't see anything wrong so far. All looks normal and then
> it just hangs. Any chance you can "gdb" to the proc and see where it is
> stuck?
>
Ralph,
The gstack output below looks like one thread is
Huh - afraid I can't see anything wrong so far. All looks normal and then it
just hangs. Any chance you can "gdb" to the proc and see where it is stuck?
On Jan 22, 2014, at 11:39 AM, Paul Hargrove wrote:
> Ralph,
>
> Attached is the requested output with the addition of
I appreciate those points, Paul. My main concern with 1.7.4 at the moment stems
from all the Fortran changes we pushed into that release - this occurred
*after* 1.7.3, and so those problems represent a regression in the 1.7 series.
We obviously appreciate all your testing since you have far
My $0.02USD:
I agree that "just keep the bar high" for 1.7.4 is the right approach.
In other words: just because I found all these issues doesn't mean they
should delay 1.7.4.
Considering most (possibly all) were in 1.7.3 and nobody noticed, what harm
in leaving the issue unresolved in 1.7.4?
If
On Jan 22, 2014, at 12:57 PM, Rolf vandeVaart wrote:
> Hi Ralph:
> In my opinion, we still try to get to a stable 1.7.4. I think we can just
> keep the bar high (as you said in the meeting) about what types of fixes need
> to get into 1.7.4. I have been telling folks
On Wed, Jan 22, 2014 at 8:50 AM, Jeff Squyres (jsquyres) wrote:
> Wow. Pulling on this thread turned up a whole pile of bugs :-\, including
> several other names that are >=32 characters:
>
> Found long name: ompi_type_create_indexed_block_f (32)
> Found long name:
Hi Ralph:
In my opinion, we still try to get to a stable 1.7.4. I think we can just keep
the bar high (as you said in the meeting) about what types of fixes need to get
into 1.7.4. I have been telling folks 1.7.4 would be ready "really soon" so
the idea of folding in 1.7.5 CMRs and delaying
Ralph,
It took all night to run, but I CAN confirm "make check" passed on my
64-bit MIPS platform with last night's v1.7 tarball (1.7.4rc2r30361).
This was with "-mabi=64" in {C,CXX,FC}FLAGS and the corresponding wrapper
flags. I would guess from Brian's response to my failures reported for
Ralph,
Attached is the requested output with the addition of "-mca
grpcomm_base_verbose 5".
I have also attached a 2nd output with the further addition of "-mca
oob_tcp_if_include lo" to ensure that this is not related to the firewall
issues I've seen on other hosts.
I have use of this host
My copy of the Fortran 2003 Standard (Adams, et al., The Fortran 203 Handbook),
says Fortran Names (incl. procedures, section 3.2.2) are permitted to be up to
63 characters. This is not phrased as a requirement, though. It could be that
a conforming processor could restrict this to fewer
On Jan 21, 2014, at 11:49 PM, Paul Hargrove wrote:
> Looks like we may be getting closer, but are not quite there:
>
> PPFC mpi-f08.lo
>BIND(C, name="ompi_type_create_hindexed_block_f")
> ^
> pathf95-1690 pathf95: ERROR
Following-up as promised:
Output from an --enable-debug build is attached.
-Paul
On Tue, Jan 21, 2014 at 11:25 PM, Paul Hargrove wrote:
> Yes, this is familiar. See:
> http://www.open-mpi.org/community/lists/devel/2013/11/13182.php
>
> If I understand correctly, the
Yes, this is familiar. See:
http://www.open-mpi.org/community/lists/devel/2013/11/13182.php
If I understand correctly, the thread ended with:
On 03 December 2013, Sylvestre Ledru wrote:
> FYI, Debian has stopped supporting ia64 for its next release
> So, I stopped working on that issue.
30 matches
Mail list logo