Created attachment 94213 [details] kernel config Vanilla kernel 3.13.3 (config attached) on x86_64 (Core(TM) i5-2410M) no longe resumes from both sleep and hibernate. Since there were no relevant ACPI changes between 3.13.2 -> 3.13.3 but quite a load of i915 drm ones, I presume (without backing) that these may be the culprit. I can't git bissect because this machine will be busy for the next days, so I'm simply reverting to 3.13.2 but I found worthy to file this bug report here, if someone else with spare time happens to fall on the same hole. Thanks.
Created attachment 94214 [details] lspci -vvv output
Created attachment 94215 [details] dmesg from clean boot
At the moment there is no indication that it is i915.ko related - not even a clue from dmesg.
Could you try if the issue can be reproduced without i915 being loaded?: Boot with the i915.modeset=0 kernel parameter and # echo mem > /sys/power/state
found this when searching, not sure if relevant but - B89M-G43 motherboard - i3 haswell cpu Suspend and resume works fine with kernel 3.12.x. Kernel 3.13.x (all versions I tried, anyway) and 3.14.2 lets me suspend and resume (to ram or disk or both) ONCE then the box hangs on suspend (when doing it for a second time). May or may be related to this bug since it's such same kernel version. Note that I do not use IGP or intel graphics but a radeon 7850. I suspect this is not GPU-related.
Alexandre, can you please bisect which commit exactly in 3.13.2 -> 3.13.3 broke resume for you? Also please boot with drm.debug=0xe and attach a new dmesg (so we know what's in your box). Öyvind since you're regression is 3.12->3.13 you likely have a different problem. Please try to bisect where exactly this regression was introduced (there are execellent howtos for almost all distros) and file a new bug report. Please make it clear in that report that it's a regression.
I'll be away from the affected machine for quite a while, besides, I believe the problem was fixed later on (IIRC I'd ran 3.14.0 - didn't try intermediate versions - and it worked fine).
(In reply to comment #7) > I'll be away from the affected machine for quite a while, besides, I believe > the problem was fixed later on (IIRC I'd ran 3.14.0 - didn't try > intermediate versions - and it worked fine). Confirmation for the fix would be great. Otherwise we'll timeout this bug in about a month or so.
(In reply to comment #8) > (In reply to comment #7) > > I'll be away from the affected machine for quite a while, besides, I believe > > the problem was fixed later on (IIRC I'd ran 3.14.0 - didn't try > > intermediate versions - and it worked fine). > > Confirmation for the fix would be great. Otherwise we'll timeout this bug in > about a month or so. Timeout, please reopen if the problem persists. Thanks for the report.
Closing resolved+fixed. Reporter verified that failure does not happen on 3.14.0 anymore.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.