Samuel,
I didn't examine the code to see what the exact issue was when the "=r"
version failed.
I only know that "make check" had a large number of failures (SEGVs if
IIRC).
I don't have infinite time, but might be able to check if an early-clobber
fixes the problem (assuming first that I can recall which platforms had the
issue).
However, in this case I don't see why an early-clobber should make a
difference. Am I missing something?
-Paul
On Tue, Jul 28, 2015 at 7:33 AM, Samuel Thibault <samuel.thiba...@inria.fr>
wrote:
> Paul Hargrove, le Tue 21 Jul 2015 16:15:24 -0700, a écrit :
> > I am glad you asked me to test widely, because I did find 2 compilers
> that
> > rejected my version with "=r" and one that generated bad code for that
> case.
>
> What kind of bad code was it generating? Perhaps it was due to not using
> an early clobber ("=&r" instead of just "=r")?
>
> Samuel
> _______________________________________________
> hwloc-devel mailing list
> hwloc-de...@open-mpi.org
> Subscription: http://www.open-mpi.org/mailman/listinfo.cgi/hwloc-devel
> Link to this post:
> http://www.open-mpi.org/community/lists/hwloc-devel/2015/07/4521.php
>
--
Paul H. Hargrove phhargr...@lbl.gov
Computer Languages & Systems Software (CLaSS) Group
Computer Science Department Tel: +1-510-495-2352
Lawrence Berkeley National Laboratory Fax: +1-510-486-6900
