Open MPI logo

Open MPI Development Mailing List Archives

  |   Home   |   Support   |   FAQ   |   all Development mailing list

From: Ferris McCormick (fmccor_at_[hidden])
Date: 2005-09-16 08:44:05


On Thu, 2005-09-15 at 20:25 -0500, Brian Barrett wrote:
> On Sep 15, 2005, at 8:17 PM, Ferris McCormick wrote:
>
> > On Thu, 15 Sep 2005, Ferris McCormick wrote:
> >

> >
> > A little experimentation suggests that instead of "mov %tick, ..." we
> > need
> > "rd %tick,%o4". I'll verify tomorrow when I am on a system with a
> > build
> > on it, but at least "rd %tick,o4" assembles properly but "mov
> > %tick,%o4"
> > gives an error.
> >
>
> Yeah, ok, that makes sense. Damn Solaris for letting me be lazy ;).
>
> Let me know if that works and I'll commit the change. I already
> committed the fixes so that the type wouldn't be #if 0'ed and the
> header would be in the dist tarball.
>
> Thanks!
>
> Brian
With current svn and with the 'mov %tick...' changed to 'rd %tick...'
everything builds and the "Can't load ..." message is gone. the
opal_timer test gives (on SB1000 (2x900)):
==========================================
fmccor_at_polylepis util [235]% ./opal_timer
--> frequency: 900000000
--> cycle count
    Slept approximately 903151189 cycles, or 1003501 us
--> usecs
    Slept approximately 18446744073289684648 us
==========================================

This just leaves the SegFault in opal/util/output.c on the
OBJ_DESTRUCT(&verbose); (in opal_output_close() at about 283)
unless either (1) OMPI_ENABLE_DEBUG is set or (2) I do something like
if(OMPI_ENABLE_DEBUG) {OBJ_DESTRUCT(&verbose);}

Without OMPI_ENABLE_DEBUG, the run_destructors loop goes insane (it dies
with i = something huge), but I can't tell why.
It's consistent across all my systems, though, and if no one else sees
it, I suppose it must be sparc-specific.

Thanks for your help,
Regards,

-- 
Ferris McCormick (P44646, MI) <fmccor_at_[hidden]>
Developer, Gentoo Linux (Sparc, Devrel)