job is now going to abort; sorry.
On Mon, Feb 2, 2015 at 7:01 PM, Paul Hargrove wrote:
> Howard,
>
> This was seen on NERSC's Carver.
>
> -Paul
>
> On Mon, Feb 2, 2015 at 6:49 PM, Howard Pritchard
> wrote:
>
>> Hi Paul,
>>
>> Thanks for check
On a Linux/PPC64 system I see the failure below from a build of the current
master tarball.
This build was configured with
--prefix=... --enable-debug \
CFLAGS=-m64 --with-wrapper-cflags=-m64 \
CXXFLAGS=-m64 --with-wrapper-cxxflags=-m64 \
FCFLAGS=-m64 --with-wrapper-fcflags=-m64
I am not
On a Linux/x86-64 system I am using the Solaris Studio 12.3 compilers.
I have configured the current master tarball as follows:
--prefix=... --enable-debug \
CC=cc CXX=CC FC=f90 \
CXXFLAGS='-L/lib/x86_64-linux-gnu -L/usr/lib/x86_64-linux-gnu
-library=stlport4' \
--with-wrapper-cxxflags='-L/
On Mon, Feb 2, 2015 at 5:47 PM, Paul Hargrove wrote:
> I'll report my test results more completely later, but all 4 PGI-based
> builds I have results for so far have failed with libtool replacing
> "-lltdl" in link command line with "/usr/lib/libltdl.so" rat
On Mon, Feb 2, 2015 at 9:26 PM, Paul Hargrove wrote:
> I am now going to see about a PGI compiler on a system at another center
> (or two?) in order to see how universal the problem is.
That was a dead-end.
Of the many non-NERSC non-Cray institutions where I have accounts, I could
onl
I found another failure mode for non-embedded libltdl.
On a system with libltdl.so on the login node but NOT the compute nodes I
encountered the following, once per rank, at job launch:
/home/phhargrove/OMPI/openmpi-libltdl-linux-x86_64 psm/INST/bin/orted:
error while loading shared libraries: li
I have seen opal_fifo hang on 2 distinct systems
+ Linux/ppc32 with xlc-11.1
+ Linux/x86-64 with icc-14.0.1.106
I have no explanation to offer for either hang.
No "weird" configure options were passed to either.
-Paul
--
Paul H. Hargrove phhargr...@lbl.gov
Computer La
On a Linux/x86_64 system with PGI-14.3 I have configured a current master
tarball with the following:
--prefix=... --enable-debug CC=pgcc CXX=pgCC FC=pgfortran
I see "make V=1" fail as shown below.
This does NOT occur with GNU or Intel compilers on the same system.
Initial guess is mis-ordered
CORRECTION:
It is the opal_lifo (not fifo) test which hung on both systems.
-Paul
On Mon, Feb 2, 2015 at 11:03 PM, Paul Hargrove wrote:
> I have seen opal_fifo hang on 2 distinct systems
> + Linux/ppc32 with xlc-11.1
> + Linux/x86-64 with icc-14.0.1.106
>
> I have no expla
2/04 4:17, Nathan Hjelm wrote:
>
> Thats the second report involving icc 14. I will dig into this later
> this week.
>
> -Nathan
>
> On Mon, Feb 02, 2015 at 11:03:41PM -0800, Paul Hargrove wrote:
>
> I have seen opal_fifo hang on 2 distinct systems
> + Linux/ppc32 with xlc
bric/libfabric/src/enosys.c:
50)
PGC/x86-64 Linux 10.9-0: compilation completed with severe errors
make[2]: *** [libfabric/src/libmca_common_libfabric_la-enosys.lo] Error 1
make[2]: Leaving directory
`/global/scratch2/sd/hargrove/OMPI/openmpi-master-linux-x86_64-pgi-10.9/BLD/opal/mca/common/lib
Yes, this time I really mean "fifo", not "lifo". ;-)
With last night's master tarball (Open MPI dev-845-ga3275aa) configured
with only --prefix and --enable-debug
A Linux-86-64 system running debian Wheezy and compiler = "gcc (Debian
4.7.2-5) 4.7.2"
Failure from "make check":
/home/phargrov/OMP
The following in orte/mca/ess/alps/Makefile.am assumes a GNU (or GNU-like)
compiler:
mca_ess_alps_la_CPPFLAGS = $(ess_alps_CPPFLAGS) -fno-ident
If building with PGI, the result is
pgcc-Error-Unknown switch: -fno-ident
when compiling orte/mca/ess/alps/ess_alps_component.c
This is last night's
With last night's master tarball (openmpi-dev-845-ga3275aa) on a
Linux/x86-64 system, I am seeing a crash (below) from ring_c run on a login
node. Other than CC/CXX/FC settings I've configured with only --prefix=...
--enable-debug --with-tm=...
This is occurring with at least the Gnu, Intel, Path
(LWP 19685)]
0x00401417 in opal_fifo_pop_atomic (fifo=0x7fffe130)
at
/home/phargrov/OMPI/openmpi-master-linux-x86_64-sl7x/openmpi-dev-904-g08dceda/opal/class/opal_fifo.h:127
127 next = (opal_list_item_t *) item->opal_list_next;
-Paul
On Fri, Feb 6, 2015 at 4:22 PM, Paul Hargrove
jelm
> wrote:
> >
> > I think I see the issue. Looks like there is a missing memory
> barrier
> > after the head consistency code. I will add one and see if that
> fixes
> > your problem.
> >
> > BTW, I can't repro
but in both cases only with Portland Group compilers
(Gnu, Intel, PathScale and Open64 compilers worked fine). So, this is not
a SUSE-specific issue.
-Paul
On Thu, Feb 12, 2015 at 11:49 PM, Philipp Thomas wrote:
> Hi Paul,
> sorry for chiming in so late, but this list is on low priority for m
Jeff,
It might be helpful to define "PR" before first use on this page and others.
While PR = "Pull Request" is common usage among folks familiar with Git, it
stands for "Problem Report" in other s/w development contexts (e.g. the
gcc-bugs mailing list and associated bugzilla). So, the phrase "f
INTEGER JEFF(3)
DATA JEFF/4HMR. ,4HFORT,3HRAN/
If you can understand that, you should probably pretend you can't :-)
-Paul [who has actually used Computed GO TO and Arithmetic IF]
On Fri, Feb 20, 2015 at 11:28 AM, Larry Baker wrote:
> Excellent, Mr. Fortran. Thank you.
>
> By th
Jeff,
+0.95
Read the new PR yesterday and agree it makes sense to bypass libltdl where
it would add little or nothing to a "dlopen-lovin' platform".
Forgive me for asking a question I am sure I could answer by reading the
.m4:
How are you planning to distinguish which platforms support dlopen()?
See two responses inline below.
On Tue, Feb 24, 2015 at 1:08 PM, Jeff Squyres (jsquyres) wrote:
> On Feb 24, 2015, at 1:55 PM, Paul Hargrove wrote:
> >
> > Forgive me for asking a question I am sure I could answer by reading the
> .m4:
> > How are you planning to dis
On Tue, Feb 24, 2015 at 1:45 PM, Paul Hargrove wrote:
[...]
>
> Smoke testing will begin momentarily...
>
[...]
I am choking on all the smoke.
Somebody call the fire marshall!
It looks like with Jeff's tarball all the BSDs are failing i
platforms when configured with "--enable-static
--disable-shared"
All my x86, x86-64, ppc and ppc64 Linux platforms
I did not test Linux on IA64, MIPS or ARM
-Paul
On Tue, Feb 24, 2015 at 2:44 PM, Paul Hargrove wrote:
>
>
> On Tue, Feb 24, 2015 at 1:45 PM, Paul Hargrove
On Wed, Feb 25, 2015 at 8:45 AM, Jeff Squyres (jsquyres) wrote:
> > SECOND:
> > On {Free,Net,Open}BSD dlopen() appears in libc, not in libdl.
> > So, I suspect one *should* be able to compile dl:dlopen on all these
> systems with the proper configure tests.
>
> Cool; I'll fix this. ...done.
L
On Wed, Feb 25, 2015 at 9:56 AM, Jeff Squyres (jsquyres) wrote:
> On Feb 25, 2015, at 11:51 AM, Dave Goodell (dgoodell)
> wrote:
> >
> >> This is a good question: what should we do here?
> >>
> >> 1. Abort the configure (e.g., insist that the user install libltdl or
> --disable-dlopen)
> >
> > I
On Wed, Feb 25, 2015 at 10:17 AM, Paul Hargrove wrote:
> I did that and just shipped a tarball to get Hargroved.
>>
>
> Tests have been dispatched... I will report complete results later today.
> The first of the BSD results should be in soon, and I'll plan to report
FYI:
On several systems where Jeff's tarball for pr410 ran fine yesterday, I am
seeing errors in today's tarball related to either libverbs or mca_oob_ud.
Issue #1:
On Solaris verbs support is now rejected at configure time.
Configure output appears below as "1)"
Issue #2:
On Linux I get undefi
, 2015, at 3:01 PM, Jeff Squyres (jsquyres)
> wrote:
> >
> > Per my prior mail, m4 typo fixed -- could you release the hounds again?
> >
> >> On Feb 25, 2015, at 2:13 PM, Paul Hargrove wrote:
> >>
> >>
> >> On Wed, Feb 25, 2015 at 10:1
On Wed, Feb 25, 2015 at 12:29 PM, Jeff Squyres (jsquyres) <
jsquy...@cisco.com> wrote:
> /me wishes yet again that shell scripting had a "strict" mode that would
> yell at you when you use "$foop" instead of "$foo" (and $foop doesn't
> exist/was never set).
See http://redsymbol.net/articles/unof
; (and $foop doesn't
> exist/was never set).
>
>
>
> > On Feb 25, 2015, at 3:04 PM, Paul Hargrove wrote:
> >
> > I've queued new tests for the platforms w/ verbs-related failures.
> > Is there any point retesting the BSDs as well?
> >
> >
On Wed, Feb 25, 2015 at 4:14 PM, Jeff Squyres (jsquyres) wrote:
> On Feb 25, 2015, at 4:17 PM, Paul Hargrove wrote:
> >
> > The Linux and Solaris verbs issues are resolved.
>
> Good.
>
> > The BSD results are unchanged.
>
> That means this, right:
>
Clang noted the following on FreeBSD-10/amd64 using the current master
tarball:
Making check in threads
make opal_thread opal_condition
CC opal_thread.o
CCLD opal_thread
CC opal_condition.o
/home/phargrov/OMPI/openmpi-master-freebsd10-amd64/openmpi-dev-1118-gdc80863/test/thr
I have been testing mtl:psm on a very old system.
Sometime pretty recently (this week I think), this started to break at
configure time:
--- MCA component mtl:psm (m4 configuration macro)
checking for MCA component mtl:psm compile mode... dso
checking --with-psm value... sanity check ok (/usr/loca
your lap, Jeff.
-Paul
On Thu, Feb 26, 2015 at 4:12 PM, Paul Hargrove wrote:
> I have been testing mtl:psm on a very old system.
> Sometime pretty recently (this week I think), this started to break at
> configure time:
>
> --- MCA component mtl:psm (m4 configuration macro)
> checki
The warning below comes from pgi-14.7 on the latest master tarball (output
from "make V=1").
-Paul
libtool: compile: pgcc -DHAVE_CONFIG_H -I.
-I/scratch/scratchdirs/hargrove/OMPI/openmpi-master-linux-x86_64-pgi-14.7/openmpi-dev-1118-gdc80863/orte/mca/oob/ud
-I../../../../opal/include -I../../../
Initially I was testing Jeff's tarball for PR 410, on Mac OS X 10.8 where
cc is clang, I have configured with
--prefix=[...] --enable-debug --enable-osx-builtin-atomics CC=cc CXX=c++
I passed "make check", but when I try to run ring_c I get the first failure
shown (far) below.
HOWEVER, I tried
I am using Oracle's Studio 12.4 compilers for Linux/x86-64 to build the
current master tarball.
However, I am passing "-m32" to generate x86 (ILP32 ABI) executables.
The full configure mess is:
--prefix=[...] --enable-debug \
CC=cc CFLAGS="-m32" --with-wrapper-cflags="-m32" \
CXX=CC CXXFLAGS="
I assume I am not the only one seeing the following with
openmpi-dev-1184-gbb22d26.tar.bz2
make[2]: *** No rule to make target
`/scratch/scratchdirs/hargrove/OMPI/openmpi-master-linux-x86_64-icc-12.1/BLD/opal/mca/common/verbs/
libmca_common_verbs.la', needed by `mca_oob_ud.la'. Stop.
make[2]: Lea
I have a system with InifniPath HCAs, where I've historically tested
mtl:psm.
For some reason, that appears to have ceased working some time in the past
4 months.
However, this report is about something else.
I am using the current master tarball: openmpi-dev-1203-g171d674.tar.bz2
When I ran confi
> fail and no one cares.
>
> Can you tell us why ibv_fork_init() would fail?
>
>
>
> > On Mar 4, 2015, at 9:56 AM, Paul Hargrove wrote:
> >
> > I have a system with InifniPath HCAs, where I've historically tested
> mtl:psm.
> > For some reason, that ap
The problem reported below is still present in openmpi-dev-1203-g171d674
-Paul
On Wed, Mar 4, 2015 at 4:47 AM, Jeff Squyres (jsquyres)
wrote:
> I think Mike just recently committed the fix. I'll go kick off a master
> tarball creation manually.
>
>
> > On Mar 3,
On Wed, Mar 4, 2015 at 1:04 PM, Dave Goodell (dgoodell)
wrote:
[...]
> > libibverbs: Warning: couldn't open config directory '/etc/libibverbs.d'.
> > libibverbs: Warning: no userspace device-specific driver found for
> /sys/class/infiniband_verbs/uverbs0
>
> I think that warning is printed by lib
I don't know if anybody *cares*, but I find that the common:libfabric code
compiles with pgi-{9,12,13,14} but not with pgi-{10,11}.
These are versions 10.9 and 11.9 of pgi that built Open MPI 1.8.4 just fine.
If somebody does care, let me know who and I'll send logs off-list.
However, this can be
Mike,
Mike, I think in this case "version does not match" really means "not
installed". I am pretty sure, as Dave G. and I discussed earlier on this
thread, those two lines are because the head node has an HCA and kernel
modules, but only the "devel" libraries have been installed (because the
HCA
On Thu, Mar 5, 2015 at 6:51 AM, Jeff Squyres (jsquyres)
wrote:
> These are the things that need to be figured out.
If somebody has a set of mca params to recommend, I am prepared to collect
as much debug/trace output as you need to understand in what order
components are initializing.
Also, I
rity since the libfabric embedding
> within
> open mpi should hopefully soon be a thing of the past.
>
> Howard
>
>
> 2015-03-04 14:28 GMT-07:00 Paul Hargrove :
>
>> I don't know if anybody *cares*, but I find that the common:libfabric
>> code compiles with pgi-{9,12
In "work for you" does "you" == @PHHargrove?
If YES:
1) are the changes to be tested reflected in the master tarball yet?
2) other than presence/absence of the warning how am I testing if the
support "works" when the param is set to 1?
-Paul
On Thu, Mar 5, 2015 at 6:29 PM, Jeff Squyres (jsquyres
Seen earlier today with last night's master tarball:
$ mpirun -mca btl sm,self,verbs -np 2 -host pcp-j-31,pcp-j-35
examples/ring_c'
[pcp-j-35:01400]
[/shared/OMPI/openmpi-master-solaris11-x64-ib-ss12u3/openmpi-dev-1351-gccba8ce/orte/mca/oob/tcp/oob_tcp_common.c:103]
setsockopt(TCP_KEEPALIVE) faile
twork routing requirements).
--
On Fri, Mar 20, 2015 at 7:13 AM, Ralph Castain wrote:
> Hi Paul
>
> It should have kept running, albeit with that warning - did the program
> actually fail?
>
>
> On Mar 1
--
>
>
>
>
> On Fri, Mar 20, 2015 at 7:13 AM, Ralph Castain wrote:
>
>> Hi Paul
>>
>> It should have kept running, albeit with that warning - did the program
>> actually fail?
>>
>>
>> On Mar 19, 2015, at
In case anybody cares:
In GASNet we have atomics for "add", "sub", "incr", "decr", and
"dec-and-test".
On some platforms all five are implemented in terms of a single
inline-atomic for "add".
There are platforms on which one or more of "incr", "decr" and
"decr-and-test" have custom implementatio
Nathan,
I test sparcv8+, sparcv9, ia64 and mips in release candidates.
That isn't the same as *using* any of those platforms in production.
I just mean to say that the implementations are known to pass "make check".
-Paul
On Thu, Mar 26, 2015 at 8:48 AM, Nathan Hjelm wrote:
>
> As a follow-on.
> On Thu, Mar 26, 2015 at 09:40:06AM -0700, Paul Hargrove wrote:
> >Nathan,
> >I test sparcv8+, sparcv9, ia64 and mips in release candidates.
> >That isn't the same as *using* any of those platforms in production.
> >I just mean to say that the impleme
Two comments on the proposed NEWS:
- Don't use inline functions with Clang compiler
>
Without motivation/explanation that sounds like a bad thing.
> - Improved support for Cray
>
Cray's compilers, networks or the programming environment in general?
-Paul
--
Paul H. Hargrove
I am trying to build Open MPI 1.8.5rc1 wit AMDs Open64 4.5.2.1.
I see the following failure (from "make V=1") while building VT:
openCC -DHAVE_CONFIG_H -I.
-I/scratch/phargrov/OMPI/openmpi-1.8.5rc1-linux-x86_64-open64/openmpi-1.8.5rc1/ompi/contrib/vt/vt/tools/vtfilter
-I../..
-I/scratch/phargro
On MacOS X 10.8 (where cc and c++ are clang and clang++, and the default
ABI is LP64) I've configured the release candidate with
--prefix= --enable-debug CC=cc CXX=c++
The build fails linking opal_wrapper as shown in the "make V=1" output
below.
Based on the undefined symbol (_lt_libltdlc_
On Mon, Apr 6, 2015 at 5:13 PM, Ralph Castain wrote:
[...]
> I believe we have seen this before, and it was an issue caused by a change
> in libtool itself. The Mac automatically updated to the new version, which
> triggers the problem.
>
> This is why we embedded ltdl directly into the OMPI mast
My testing of real hardware is almost complete and I've reported the only
two issues[*] I encountered:
http://www.open-mpi.org/community/lists/devel/2015/04/17183.php
http://www.open-mpi.org/community/lists/devel/2015/04/17184.php
There were roughly 5 failing configurations out of about 70.
ot; ABIs on a 64-bit (MIPS 5Kc) system.
I believe that completely covers the atomics implementations for those
architectures.
-Paul
On Mon, Apr 6, 2015 at 6:38 PM, Paul Hargrove wrote:
> My testing of real hardware is almost complete and I've reported the only
> two issues[*] I encountered
wrote:
> Thanks Paul!!
>
> The VT fix is in the queue, if/when you can check it. I'm going to look at
> the other issue today.
>
>
> On Apr 8, 2015, at 8:10 AM, Paul Hargrove wrote:
>
> My 5 ARM and 3 MIPS testers completed without any problems.
>
> The ARM test
Ralph and Bert,
I found a couple issues with the VT fix.
I will document them in the PR when I've finished collecting the outputs..
-Paul
On Wed, Apr 8, 2015 at 8:40 AM, Paul Hargrove wrote:
> Ralph,
>
> According to his comments in the PR, Bert already tested with the same
>
Paul Hargrove wrote:
> Ralph and Bert,
>
> I found a couple issues with the VT fix.
> I will document them in the PR when I've finished collecting the outputs..
>
> -Paul
>
> On Wed, Apr 8, 2015 at 8:40 AM, Paul Hargrove wrote:
>
>> Ralph,
>>
>> Accord
ou don't enable CUDA, then you should be able to test.
>
> On Thu, Apr 9, 2015 at 10:03 AM, Paul Hargrove wrote:
>
>> Ralph,
>>
>> It appears that Bert and I have iterated to a correct fix for the
>> VT+Open64 issue.
>>
>> Did you make any progre
Ok, looks good now on MacOS 10.8 (at v1.8.4-189-ga6169b4)
-Paul
On Thu, Apr 9, 2015 at 10:15 AM, Paul Hargrove wrote:
> Ok, Ralph.
> I will test the v1.8 nightly and report.
>
> -Paul
>
> On Thu, Apr 9, 2015 at 10:06 AM, Ralph Castain wrote:
>
>> I committed
On Fri, Apr 17, 2015 at 5:57 AM, Jeff Squyres (jsquyres) wrote:
> - OS X (10.6, 10.7, 10.8, 10.9), 32 and 64 bit (x86_64), with gcc and
> Absoft compilers (*)
>
Since about 10.7 (depending which XCode you installed), cc and c++ have
been Clang and Clang++ on Mac OS X.
The "gcc" is optional
On Fri, Apr 17, 2015 at 1:02 PM, Ralph Castain wrote:
[...regarding Cray XC...]
> If you don't want to support it, I understand - but we should be clear
> that it may well work anyway.
Ralph,
Do you really want to enumerate all of the "it may well work anyway"
platforms?
If so, I have quite a
Ralph,
I think David's concern is that because Slurm has changed their default
behavior, Open MPI's default behavior has changed as well.
The request (on which I have no opinion) appears to be that ORTE make an
explicit request for the behavior that was the previous default in Slurm.
That would en
ng on Intel Xeon Phi coprocessors.
> - Opportunistically switch away from using GNU Libtool's libltdl
> library when possible (by default).
> - Fix some VampirTrace errors. Thanks to Paul Hargrove for reporting
> the issues.
> - Correct default binding patterns when --use-hw
ctly/completely
from master to v1.8.
-Paul [a.k.a. bot:hargrove]
On Tue, Apr 21, 2015 at 4:22 PM, Paul Hargrove wrote:
> Is the following configure-fails-by-default behavior really the desired
> one in 1.8.5?
> I thought this was more of a 1.9 change than a mid-series change.
>
&g
, 2015, at 7:38 PM, Paul Hargrove wrote:
>
> Sorry the output in the previous email left out some relevant detail.
> See here that BOTH dl components were unable to compile with the 1.8.5rc2
> tarball:
>
> +++ Configuring MCA framework dl
> checking for no configure componen
rocessors.
> - Opportunistically switch away from using GNU Libtool's libltdl
> library when possible (by default).
> - Fix some VampirTrace errors. Thanks to Paul Hargrove for reporting
> the issues.
> - Correct default binding patterns when --use-hwthread-cpus was
>
processes, but perhaps mpirun propogates the
rlimits.
So, on NERSC's Carver I can reproduce the problem (in my build of 1.8.5rc2)
quite easily (below).
I have configured with --enable-debug, which probably explains why I see an
assertion failure rather than the reported SEGV.
-Paul
{hargrove@
, 2015 at 9:41 AM, Paul Hargrove wrote:
> Howard,
>
> Unless there is some reason the settings must be global, you should be
> able to set the limits w/o root privs:
>
> Bourne shells:
> $ ulimit -l 64
> C shells:
> % limit -h memorylocked 64
>
> I would have
When building from a git clone of master I encountered the following:
checking for flex... no
checking for lex... no
configure: WARNING: *** Could not find GNU Flex on your system.
configure: WARNING: *** GNU Flex required for developer builds of Open MPI.
configure: WARNING: *** Other versions of
I had reason to look at the linux timer code today and noticed what I
believe to be a subtle error.
This is in both 'master' and v1.8.5rc2
Since casts bind tighter than multiplication in C, I believe that the
1-line patch below is required to produce the desired result of conversion
to an integer
On Wed, Apr 22, 2015 at 2:02 PM, Jeff Squyres (jsquyres) wrote:
> FWIW, it *did* print a list of files for me on my Mac when I faked it out
> and forced it to *not* find flex.
>
A quick look and the commit shows
for lfile in `find . -name \*.l -print`; do
Notice the find is rooted at ".".
I had an opportunity to try the 1.8.5rc2 tarball on a little-endian POWER8
(aka ppc64el or powerpc64le).
The good news is that things "just worked" as they did when I tried ARMv8
(aka aarch64).
However, I see a little room for improvement with almost no work at all.
I noticed:
checking for __syn
Well, I tried rc2 on just about everything except my phone and my linksys.
For me the configure failure (dlopen() not found) on {Free,Net,Open}BSD is
the only problem.
Since it works on 'master' I am confident Jeff will sort this out.
-Paul
--
Paul H. Hargrove phhargr..
On Wed, Apr 22, 2015 at 2:43 PM, Jeff Squyres (jsquyres) wrote:
> > In addition to the one-line patch below, I needed to run autogen.pl
> with a new enough config/config.{guess,sub}.
> > Along the way I noticed
> > opal/mca/common/libfabric/libfabric/config/config.guess
> > opal/m
past that point and making updates to the PR instead of this email
thread.
-Paul
>
>
> > On Apr 21, 2015, at 8:40 PM, Paul Hargrove wrote:
> >
> >
> >
> > On Tue, Apr 21, 2015 at 5:33 PM, Jeff Squyres (jsquyres) <
> jsquy...@cisco.com> wrote:
&g
When running "make dist" (several times) today, I noticed the following:
- WARNING: Got bad config.guess from ftp.gnu.org (non-existent or empty)
- WARNING: using included versions for both config.sub and config.guess
Nevermind for the moment that the wget is NOT from ftp.gnu.org any longer.
Th
I have attached a patch (against master) that fixes some typos and makes an
update.
It applies *almost* cleanly to v1.8, requiring "-C2" if applying with "git
apply" due to context changes.
I also noted the following which I believe is just plain false, but don't
have an alternative for.
Portals
Applied -- thank you!
>
> -Jeff [accepting patches from Paul since 2002]
>
> ;-p
>
> > On Apr 23, 2015, at 2:29 PM, Paul Hargrove wrote:
> >
> > I have attached a patch (against master) that fixes some typos and makes
> an update.
> > It applies almost clea
I have a system w/ xlc-11.1.
It has essentially always failed "make check" in a LP64 build due to xlc
botching the atomics.
So, when it failed with 1.8.5.rc2 I didn't look closely.
Today it has failed with rc3 and I *did* look closely and here is what I
see:
PASS: predefined_gap_test
/bin/sh: lin
/atomic.h:158
So, this is a new symptom of the known inability of this compiler to get
the inline asm right.
Sorry for the false alarm,
-Paul
On Thu, Apr 23, 2015 at 4:09 PM, Paul Hargrove wrote:
> I have a system w/ xlc-11.1.
> It has essentially always failed "make check" in a LP64
Sorry my ARMv6, ARMv8 and PowerPC64LE systems were dedicated to other
purposes today.
So, I was only able to test 1.8.5rc3 against 71 distinct configurations.
;-)
There was 1 known-failure (xlc-11.1 LP64) that scared me by changing
failure modes between rc1 and rc2.
Several slow tests (Linux on M
I gave "make dist" a try on NetBSD (with its own /bin/sh) and Ubuntu Trusty
(w/ /bin/sh symlinked to dash).
Both generated the tarballs, but dash spewed some warnings on the unalias
commands.
So here is my code review (roughly as long as the scipt itself).
1)
#!/usr/bin/env sh
The presence of
There is a block of code near the start of the OPAL_CONF_ASM which begins:
# OS X Leopard ld bus errors if ...
However, Leopard is OS X 10.5 and the minimum supported by Open MPI is 10.6.
So, that code should be unreachable at this time (and since Jan 2014
http://www.open-mpi.org/community/lists
5 of the 6 MIPS and ARM testers that were still running last night have
completed successfully.
No reason to think the last one won't pass on rc3 as it did on rc2, if
given another 2 or 3 hours to complete.
-Paul
On Fri, Apr 24, 2015 at 9:52 AM, Ralph Castain wrote:
> Any last minute issues peo
All done! All good!
Summary:
3 Unavailable
1 Known bad[*] configuration
70 PASS
-Paul
[*] Compiler bug confirmed by Nysal
On Thu, Apr 23, 2015 at 7:29 PM, Paul Hargrove wrote:
> Sorry my ARMv6, ARMv8 and PowerPC64LE systems were dedicated to other
> purposes today.
&
reloading of firmware is
involved.
-Paul
On Fri, Apr 24, 2015 at 3:04 PM, Troy Benjegerdes wrote:
> On Wed, Apr 22, 2015 at 02:19:07PM -0700, Paul Hargrove wrote:
> > I had an opportunity to try the 1.8.5rc2 tarball on a little-endian
> POWER8
> > (aka ppc64el or powerpc64le).
AIX, Solaris and {Free,Open,Net}BSD results are also not consistent with
regards to units used for reporting:
AIX$ no -o tcp_keepidle -o tcp_keepintvl
tcp_keepidle = 14400
tcp_keepintvl = 150
{phargrov@solaris11-amd64 ~}$ ndd -get /dev/tcp tcp_keepalive_interval
720
[phargrov@freebsd10-amd64
On Tue, May 19, 2015 at 9:37 PM, Howard Pritchard
wrote:
> Pretty soon the developer will get trained to use the PR process, unless
> they are that engineer I've yet to meet who always writes flawless code.
I've never met that developer, either.
However, I have met one (and thankfully only one)
I see a failure on OpenBSD when building ROMIO:
libtool: compile: gcc -std=gnu99 -DHAVE_CONFIG_H -I.
-I/home/phargrov/OMPI/openmpi-1.8.6rc1-openbsd5-amd64/openmpi-1.8.6rc1/ompi/mca/io/romio/romio
-I./adio/include -DOMPI_BUILDING=1
-I/home/phargrov/OMPI/openmpi-1.8.6rc1-openbsd5-amd64/openmpi-1.8.
ng code assumes that if that configure test
passed, then struct statfs contains a f_type field.
I do notice that Open MPI's configure actually checks rather than assuming
this:
checking for struct statfs.f_type... no
-Paul
On Sun, May 24, 2015 at 5:57 PM, Paul Hargrove wrote:
> I see a fa
I still have some slow emulated ARM and MIPS testers running (though some
positive ARM results are in).
I will report the MIPS and remaining ARM results when they are all ready
(probably Tue AM).
However, with 70 configurations tested the only issue I've seen is the
ROMIO build failure on OpenBSD-
On Sun, May 24, 2015 at 10:24 PM, Paul Hargrove wrote:
> I will report the MIPS and remaining ARM results when they are all ready
> (probably Tue AM).
>
They completed normally.
-Paul
--
Paul H. Hargrove phhargr...@lbl.gov
Computer Languages & Sys
My testing report for 1.8.6rc2 is nearly unchanged relative to rc1:
Around 70 configurations have completed, but slow qemu-emulated ARM and
MIPS systems will continue overnight.
At this point everything is successful except for some known issues (also
present in 1.8.5 and thus not regressions):
I can see cloning of existing component's source as a starting point for a
new one as a common occurrence (at least relative to creating new
components from zero).
So, this is probably not the last time this will ever occur.
Would a build with --disable-dlopen have detected this problem (by failin
On Thu, Jun 25, 2015 at 4:59 PM, Gilles Gouaillardet
wrote:
> In this case, mca_coll_hcoll module is linked with the proprietary
> libhcoll.so.
> the ml symbols are defined in both mca_coll_ml.so and libhcoll.so
> i am not sure (i blame my poor understanding of linkers) this is an error
> if
> Op
701 - 800 of 925 matches
Mail list logo