Skip to contentSkip to conversation

3.15-rc5: Regression in i915 driver?

Intel-gfx Thomas Meyer 1 participant 1 message May 17, 2014

lore
Share & tools
lore
#0 lore raw

3.15-rc2 seems to be good, 3.15-rc5 seems to be bad. Bisecting this the next days.

Am 12.05.2014 18:54 schrieb Daniel Vetter <[email protected]>:

Quoted 36 lines · Jani Nikula
Quoted 21 lines · Daniel Vetter
Quoted 11 lines · Dave Airlie
Quoted 9 lines · Thomas Meyer
Quoted 8 lines

Hi,

3.14.3 works as expected.
3.15-rc5 shows a strange behaviour: When resuming from ram the X server seems to be disfunctional.

I see this WARNING in the kernel log before suspend to ram in the early boot process:

Quoted 8 lines

Doesn't ring a bell really.
- Is there anything in dmesg after resume?
- How exactly does X misbehave? Is fbcon still working? Does X behaviour get restored if you restart X?
- Can you please try to bisect this issue? Note that the backlight issue might be unrelated to the issues with X misbehaving after resume. You might need to do a bisect for both if the symptoms don't agree.

Quoted 2 lines

I agree the WARNING from i965_enable_backlight() is unrelated. Focus on the other symptoms first.

The backlight backtrace seems to be a genuine new bug. Bisecting it would be highly appreciated. Bisecting on the other issue and appending the result to the bug Chris quoated should also really be useful - atm we don't have any reported who can reproduce this clearly enough to do a bisect.
-Daniel

Signature · 3 lines
Daniel Vetter 
Software Engineer, Intel Corporation 
+41 (0) 79 365 57 48 - http://blog.ffwll.ch