Bug 120451
| Summary: | snd_hda_* crash after resuming laptop from suspend | ||
|---|---|---|---|
| Product: | Other | Reporter: | Paolo Stivanin (paolo.stivanin) |
| Component: | Modules | Assignee: | other_modules |
| Status: | NEW --- | ||
| Severity: | normal | CC: | jwrdegoede, peter |
| Priority: | P1 | ||
| Hardware: | All | ||
| OS: | Linux | ||
| Kernel Version: | 4.5.x, 4.6.x | Subsystem: | |
| Regression: | No | Bisected commit-id: | |
| Attachments: |
full dmesg output
Patch for 4.6.4 reverting the commit causing this problem. dmesg after patch |
||
|
Description
Paolo Stivanin
2016-06-16 08:37:13 UTC
Please post a full dmesg? The stacktrace is missing essential details. How reproducible is this? Your laptop looks like a Clevo P6xxRE, maybe I can reproduce it on a P651RA. Created attachment 222751 [details]
full dmesg output
Hello Peter,
yes my laptop (Santech T67) is based on a Clevo P640RE.
Attached you can find a full dmesg output after the laptop has been resumed.
PS: I had this problem also with kernel 4.5.x and on Ubuntu 16.04.
Oh, it is just a warning. Not a crash as you mentioned before: [ 82.571396] WARNING: CPU: 7 PID: 899 at drivers/gpu/drm/i915/intel_uncore.c:649 __unclaimed_reg_debug+0x7b/0x90 [i915] [ 82.571398] Unclaimed register detected after reading register 0x65f10 I've seen this happen without even headphones plugged in (with no nvidia/nouveau, but bbswitch and the Plasma desktop) and if I remember correctly, without even having to suspend. Sorry, I thought it was a crash. BTW, I do have this problem, every time I resume my laptop headphones don't work any more. PS I'm not using bbswitch or Plasma. Only the official nvidia driver with Gnome. And I'm using only the intel card (xorg.conf setting) because I don't need the nvidia one :) I mentioned my environment for completeness (forgot to mention that the power_save module option is on). It looks like the bug is known and a patch is available: https://bugs.freedesktop.org/show_bug.cgi?id=96214 https://patchwork.freedesktop.org/patch/97995/ I've seen that the patch hasn't been merged on 4.7-rc7. Do you know it if will be merged before the final 4.7 release? Or do I have to wait for the 4.8-rc1 one? Thank you very much Peter! The patch is currently still under review. I see no indication of a backport (CC: stable@..), but I guess you could request it and add that you seem to lose sound. Are you able to test this patch and report upstream that it has been tested? I have patched the kernel 4.6.3 but nothing changed. Same warning on dmesg and same result: can't use my headphones after the laptop has been resumed. Cheers @Paolo can you double-check that the patch was applied? @Peter: double-checked, the patch is there :( Actually, using "init-headphone" my problem is solved. Better than nothing :-) Created attachment 223831 [details] Patch for 4.6.4 reverting the commit causing this problem. As discussed in: https://bugs.freedesktop.org/show_bug.cgi?id=96214 The original patch indeed did not work. Attached is another patch (suggested by Chris Wilson) which does fix the WARN_ON for me. @Hans thank you! I'm rebuilding 4.6.4 kernel right now. How do I know if the patch has worked? Will the "unclaimed reg blabla" disappear or will my headphone start working after the laptop has been resumed? Created attachment 223841 [details]
dmesg after patch
Hi,
I have applied the patch against the 4.6.4 source but still nothing changed :(
Attached you can find the dmesg output before and after suspending the laptop.
Cheers
|