Bug 13867 - [regression] Xorg crashes after hibernate/resume to disk
Summary: [regression] Xorg crashes after hibernate/resume to disk
Status: RESOLVED NOTABUG
Alias: None
Product: xorg
Classification: Unclassified
Component: Driver/intel (show other bugs)
Version: git
Hardware: x86 (IA32) All
: high blocker
Assignee: CIJOML CIJOMLovic CIJOMLov
QA Contact: Xorg Project Team
URL:
Whiteboard:
Keywords: NEEDINFO
Depends on:
Blocks: 13493
  Show dependency treegraph
 
Reported: 2007-12-30 12:50 UTC by CIJOML CIJOMLovic CIJOMLov
Modified: 2008-01-27 16:51 UTC (History)
1 user (show)

See Also:
i915 platform:
i915 features:


Attachments
Xorg.0.log (40.63 KB, text/plain)
2007-12-30 12:52 UTC, CIJOML CIJOMLovic CIJOMLov
no flags Details
Restore all VGA regs early (949 bytes, patch)
2008-01-03 14:34 UTC, Jesse Barnes
no flags Details | Splinter Review

Description CIJOML CIJOMLovic CIJOMLov 2007-12-30 12:50:34 UTC
Hi there,

few last git changes I see regression - When I hibernate to disk and resume again, after loading data from swap, everythink starts, but Xorg imeditially crashes!

Marking blocker, because users lost their data and also regression

Michal
Comment 1 CIJOML CIJOMLovic CIJOMLov 2007-12-30 12:52:02 UTC
Created attachment 13419 [details]
Xorg.0.log

saved /var/log/Xorg.0.log
Comment 2 Gordon Jin 2007-12-31 05:52:03 UTC
cijoml, could you git-bisect to find which git commit causes this regression?
Comment 3 Jesse Barnes 2008-01-02 10:19:48 UTC
Michal, this looks like either 13196 or 13609 (probably the former since I just pushed a change for that), can you confirm?
Comment 4 Jesse Barnes 2008-01-02 11:13:12 UTC
Nevermind, I closed 13196 since that was a regression caused by the fix for this bug (which was to leave the pipes disabled).  Now to figure out how we can restore the pipe configuration across 9xx *and* 8xx chips...
Comment 5 Jesse Barnes 2008-01-02 11:24:13 UTC
Michal, what model of laptop are you using?  Are you using the intelfb driver (lsmod or your kernel config could tell you)?
Comment 6 CIJOML CIJOMLovic CIJOMLov 2008-01-02 12:08:41 UTC
Hi there,

My setup is Acer TravelMate 242, Intel 852GME.

00:02.0 VGA compatible controller: Intel Corporation 82852/855GM Integrated Graphics Device (rev 02)
00:02.1 Display controller: Intel Corporation 82852/855GM Integrated Graphics Device (rev 02)

I use clean vanilla 2.6.23.11 kernel, with intel fb compiled into kernel (not as module)

Linux agpgart interface v0.102
agpgart: Detected an Intel 855GM Chipset.
agpgart: Detected 16252K stolen memory.
agpgart: AGP aperture is 128M @ 0xe8000000
[drm] Initialized drm 1.1.0 20060810
ACPI: PCI Interrupt Link [LNKA] enabled at IRQ 11
PCI: setting IRQ 11 as level-triggered
ACPI: PCI Interrupt 0000:00:02.0[A] -> Link [LNKA] -> GSI 11 (level, low) -> IRQ 11
[drm] Initialized i915 1.6.0 20060119 on minor 0
[drm] Initialized i915 1.6.0 20060119 on minor 1
intelfb: Framebuffer driver for Intel(R) 830M/845G/852GM/855GM/865G/915G/915GM/945G/945GM chipsets
intelfb: Version 0.9.4
intelfb: 00:02.0: Intel(R) 852GME, aperture size 128MB, stolen memory 16252kB
intelfb: Non-CRT device is enabled ( LVDS port ).  Disabling mode switching.
intelfb: Video mode must be programmed at boot time.

Comment 7 Jesse Barnes 2008-01-02 12:11:06 UTC
Can you try reproducing the problem w/o intelfb (or any other fb drivers) in use?  They tend to cause hw programming conflicts...
Comment 8 Jesse Barnes 2008-01-03 14:34:25 UTC
Created attachment 13497 [details] [review]
Restore all VGA regs early

On the theory that VGA modes require the VGA misc. regs to be restored to work correctly, this patch restores them early before re-enabling the PLLs into VGA mode.  Can you give it a try?
Comment 9 CIJOML CIJOMLovic CIJOMLov 2008-01-04 08:34:01 UTC
with this patch my Xorg crashes whole laptop imeditially after start.
Aplied to:
drwxr-sr-x  6 root     src           4096 2007-12-27 19:43 xf86-video-intel
Comment 10 Jesse Barnes 2008-01-10 11:20:38 UTC
What about if you run w/o the fb layer & fb drivers?
Comment 11 Benjamin Close 2008-01-11 02:36:50 UTC
Bugzilla Upgrade Mass Bug Change

NEEDSINFO state was removed in Bugzilla 3.x, reopening any bugs previously listed as NEEDSINFO.

  - benjsc
    fd.o Wrangler
Comment 12 CIJOML CIJOMLovic CIJOMLov 2008-01-11 12:34:03 UTC
Hi,

I haven't checked yet, this requires kernel recompilation and it takes long time on my laptop. I will compile when 2.6.24 is released.

Michal
Comment 13 Michael Fu 2008-01-18 05:36:21 UTC
cijoml, this is very likely to be caused by intel-fb which was compiled into kernel.. please remove it and re-test.
Comment 14 CIJOML CIJOMLovic CIJOMLov 2008-01-26 14:24:31 UTC
Hi there,

I just tested 2.6.24 without fb it works like a charm, with it, it is worse than with 2.6.23 - kernel completely freezes after resume.
Comment 15 Michael Fu 2008-01-27 16:51:56 UTC
removing intel-fb is required... mark this bug as resolved. thanks.


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.