<html> <head> <base href="https://bugs.freedesktop.org/" /> </head> <body> <div> <a class="bz_bug_link bz_status_NEW " title="NEW - Xorg unresponsive under heavy system-load" href="https://bugs.freedesktop.org/show_bug.cgi?id=90492#c10">Comment # 10</a> on <a class="bz_bug_link bz_status_NEW " title="NEW - Xorg unresponsive under heavy system-load" href="https://bugs.freedesktop.org/show_bug.cgi?id=90492">bug 90492</a> from <a class="email" href="mailto:alexander.maznev@gmail.com" title="alexander.maznev@gmail.com">alexander.maznev@gmail.com</a> <pre>Created <a href="attachment.cgi?id=115863" name="attach_115863" title="sudo perf top, xorg log, vmstat">attachment 115863</a> <a href="attachment.cgi?id=115863&action=edit" title="sudo perf top, xorg log, vmstat">[details]</a> sudo perf top, xorg log, vmstat > Looking at the page, firefox allocates over 3.6GiB of Pixmaps on my 32bit system. (Quite an impressive feat!) Yeah there's always some subtle bug browsers screw the user with, but I finally got one good for diagnosing other stuff! > Since that is beyond the address space limits for Xorg, we will be trying to hide allocation and access failures. Trying again without swap, resulted in oom and firefox dying (as expected) - but managed to trigger a few i915.ko warnings. So with oom_kill_allocating_task 0, xserver freezes, and whether firefox is actually ever attempted to be killed I still don't know. This is with a 1 minute scheduled shutdown (x-server froze, than system powered off correctly). Running with renice priorities for vmstat seems to help, sudo perf top output included. oomscore is from watch cat /proc/<firefoxpid>/oom_score > oomscore.txt, with renice -5.</pre> </div> <hr> You are receiving this mail because: <ul> <li>You are the QA Contact for the bug.</li> </ul> </body> </html>