[Bug 110732] New: GPU hangs on render ring (coincides with CPU clock throttling).

bugzilla-daemon at freedesktop.org bugzilla-daemon at freedesktop.org
Wed May 22 17:12:50 UTC 2019


https://bugs.freedesktop.org/show_bug.cgi?id=110732

            Bug ID: 110732
           Summary: GPU hangs on render ring (coincides with CPU clock
                    throttling).
           Product: DRI
           Version: unspecified
          Hardware: x86-64 (AMD64)
                OS: Linux (All)
            Status: NEW
          Severity: normal
          Priority: medium
         Component: DRM/Intel
          Assignee: intel-gfx-bugs at lists.freedesktop.org
          Reporter: ben.e.whitney at post.harvard.edu
        QA Contact: intel-gfx-bugs at lists.freedesktop.org
                CC: intel-gfx-bugs at lists.freedesktop.org

Created attachment 144321
  --> https://bugs.freedesktop.org/attachment.cgi?id=144321&action=edit
Excerpt of `/var/log/syslog`.

My computer sometimes turns off when I max out* all its CPUs for more than a
few minutes. I would estimate that it turns off something like 1 out of 10
times when the CPUs are maxed out for more than five minutes, with the
probability increasing with the amount of time under load. It happens pretty
reliably (maybe 2 out of 3 times? more often?) when they're working for more
than an hour. Normally when it turns off I just rerun my job with fewer CPUs,
but this time I opened up the system log and found a report of a GPU hang. I'm
attaching part of the system log, the GPU crash dump, and some system
information. Please let me know if there's more information I can provide.

* By this I mean that `top` reports CPU usage at or near 100%. I doubt that
much of what I'm doing is CPU-bound.

-- 
You are receiving this mail because:
You are the QA Contact for the bug.
You are on the CC list for the bug.
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/intel-gfx-bugs/attachments/20190522/1372b6e1/attachment.html>


More information about the intel-gfx-bugs mailing list