https://bugzilla.kernel.org/show_bug.cgi?id=71051 Thomas J. Moore <darktjm@xxxxxxxxx> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |darktjm@xxxxxxxxx --- Comment #16 from Thomas J. Moore <darktjm@xxxxxxxxx> --- I have a Dell Inspiron with AMD A6 (Advanced Micro Devices, Inc. [AMD/ATI] Mullins [Radeon R4/R5 Graphics]), and stopped upgrading my kernel at 3.16.3 because it would no longer resume from disk about 90% or more of the time. I have finally given in and looked for the cause a bit, and have traced it to this patch. Without this patch, 3.18.3 (and probably all other versions I skipped before it) works fine for me. That said, I never suspend to memory, and so I tried with and without, and indeed the patch seems to be necessary for resume-from-memory on my system as well. So, while this patch does fix a bug, it introduces a regression for me. Is it possible the value of nb_dpm_enabled is not restored or initialized properly on resume from disk, due to the video mode change? The symptoms when resuming from disk are similar to comment 9 (ring test failures, which do not occur without the patch), with X crashing (and, due to the fact that I start it in inittab, retrying with "couldn't schedule ib" error until I reboot or disable the inittab entry). -- You are receiving this mail because: You are watching the assignee of the bug. _______________________________________________ dri-devel mailing list dri-devel@xxxxxxxxxxxxxxxxxxxxx http://lists.freedesktop.org/mailman/listinfo/dri-devel
