Re: [OMPI devel] 1.10.0rc2

2015-07-22 Thread Orion Poplawski
On 07/21/2015 07:59 PM, Jeff Squyres (jsquyres) wrote: > On Jul 21, 2015, at 4:38 PM, Orion Poplawski wrote: >> >> I see that 1.10.0 can make use of libfabric, but I don't see much >> documentation on it and certainly nothing in the README file. > > Oops; shame on us. We'll update README. > >

Re: [OMPI devel] 1.10.0rc2

2015-07-22 Thread Paul Hargrove
Yohann, Things run fine with those additional flags. In fact, adding just "--mca pml cm" is sufficient to eliminate the SEGV. -Paul On Wed, Jul 22, 2015 at 8:49 AM, Burette, Yohann wrote: > Hi Paul, > > > > Thank you for doing all this testing! > > > > About 1), it’s hard for me to see whethe

Re: [OMPI devel] anyone built master on qlogic system today?

2015-07-22 Thread Howard Pritchard
Hi Folks, Found the problem, had to do a hard reset to origin/master for some reason to get missing files back. Howard 2015-07-22 12:17 GMT-06:00 Jeff Squyres (jsquyres) : > On Jul 22, 2015, at 1:46 PM, Howard Pritchard wrote: > > > > Hello Folks, > > > > I"m investigating a psm/ofi mtl probl

Re: [OMPI devel] anyone built master on qlogic system today?

2015-07-22 Thread Jeff Squyres (jsquyres)
On Jul 22, 2015, at 1:46 PM, Howard Pritchard wrote: > > Hello Folks, > > I"m investigating a psm/ofi mtl problem on one of our qlogic systems and > ended up investigating something else. > > There seem to be a bunch of missing config.h.in files if I of current master > head. > If I go back t

[OMPI devel] anyone built master on qlogic system today?

2015-07-22 Thread Howard Pritchard
Hello Folks, I"m investigating a psm/ofi mtl problem on one of our qlogic systems and ended up investigating something else. There seem to be a bunch of missing config.h.in files if I of current master head. If I go back to bd60ce16 things seem to be okay. The upshot is that one doesn't get past

Re: [OMPI devel] 1.10.0rc2

2015-07-22 Thread Burette, Yohann
Hi Paul, Thank you for doing all this testing! About 1), it’s hard for me to see whether it’s a problem with mtl:ofi or with how OMPI selects the components to use. Could you please run your test again with “--mca mtl ofi --mca mtl_ofi_provider sockets --mca pml cm”? The idea is that if it stil

Re: [OMPI devel] race condition in finalize

2015-07-22 Thread Ralph Castain
Thanks Gilles! The “clean” event doesn’t have to go last - any messages that arrive after all recvs have been removed will simply be dropped upon termination. This commit only ensures that the list of posted recvs is cleanly destructed, which will prevent the segfault. > On Jul 22, 2015, at 1

Re: [OMPI devel] 1.10.0rc2

2015-07-22 Thread Paul Hargrove
1.10.0rc2 looks mostly good to me, but I still found some issues. 1) New to this round of testing, I have built mtl:ofi with gcc, pgi, icc, clang, open64 and studio compilers. I have only the sockets provider in libfaric (v1.0.0 and 1.1.0rc2). However, unless I pass "-mca mtl ^ofi" to mpirun I ge

Re: [OMPI devel] race condition in finalize

2015-07-22 Thread Gilles Gouaillardet
Thanks Ralph, i was unable to reproduce any crash with this fix :-) i checked the code that is invoked in the progress thread, and it might queue other events. bottom line, i am not 100% convinced the "clean" event will be executed at the very last. that being said, and once again, i was una

Re: [OMPI devel] 1.10.0rc2

2015-07-22 Thread Paul Hargrove
Jeff, Sorry if you already noticed: It appears that there is no mention of --with-libfabric or --with-libfabric-libdir in the README for 1.10.0rc2 -Paul On Tue, Jul 21, 2015 at 6:59 PM, Jeff Squyres (jsquyres) wrote: > On Jul 21, 2015, at 4:38 PM, Orion Poplawski wrote: > > > > I see that 1.1