[PATCH 0/4] openacc: Async fixes
Thomas Schwinge
thomas@codesourcery.com
Wed Jun 30 08:28:00 GMT 2021
Hi Julian!
On 2021-06-29T16:42:00-0700, Julian Brown <julian@codesourcery.com> wrote:
> This patch series contains fixes for various problems with async support
> for OpenACC at present:
Thanks, I shall be looking into these in detail, "soonish".
Some quick comments.
> - Asynchonous host-to-device copies invoked from within libgomp
> (target.c) could copy bad data to the target -- and the workaround
> for that currently used in the AMD GCN target plugin could lead to
> a different problem (a race condition).
As per discussion on Monday, I like to you moved (back?) the "ephemeral"
handling into 'libgomp/target.c:gomp_copy_host2dev'.
> - The OpenACC profiling-interface implementation did not measure
> asynchronous operations properly.
We'll need to be careful: (possibly, an older version of) that one we
internally had identified to be causing some issues; see the
"acc_prof-parallel-1.c intermittent failure on og10 branch" emails,
2020-07.
> - Several test cases misuse OpenACC asynchronous support (more race
> conditions).
Mostly ACK, but some more changes may be necessary; please see
<http://mid.mail-archive.com/87sg1s9s9l.fsf@euler.schwinge.homeip.net>
(you were CCed).
> .../libgomp.oacc-c-c++-common/deep-copy-10.c | 14 +-
Please provide some detail about that one ("Fix async behaviour"). It's
not obvious to me what's wrong with the current version (but I haven't
really spent time on that yet).
Grüße
Thomas
-----------------
Mentor Graphics (Deutschland) GmbH, Arnulfstrasse 201, 80634 München Registergericht München HRB 106955, Geschäftsführer: Thomas Heurung, Frank Thürauf
More information about the Gcc-patches
mailing list