Bug 50891

Summary: Noisy lines on screen, Xorg unusable on Radeon HD 5670
Product: xorg Reporter: Goulou <fredericg_99>
Component: Driver/RadeonAssignee: xf86-video-ati maintainers <xorg-driver-ati>
Status: RESOLVED DUPLICATE QA Contact: Xorg Project Team <xorg-team>
Severity: critical    
Priority: medium    
Version: unspecified   
Hardware: x86-64 (AMD64)   
OS: Linux (All)   
Whiteboard:
i915 platform: i915 features:
Attachments:
Description Flags
screenshot of the bug
none
lspci
none
Xorg.log with normal kernel options
none
Xorg.log with nomodeset kernel option none

Description Goulou 2012-06-08 11:02:03 UTC
Created attachment 62806 [details]
screenshot of the bug

When Xorg starts, instead of a normal display, I get only blank and noisy lines (see screenshot).

I tried several things :
-proprietary driver : works without problems
-start with nomodeset : instead of noisy lines, I get a dark screen (see attachement Xorg_log_nomodeset)
-normal start, I get the noisy lines, see Xorg_log

The Xorg logs were obtained by stopping Xorg (systemctl isolate multi-user.target), and then using directly "Xorg  -verbose 10 > log_file 2>&1"

When the bug occured, I got some "GPU lockup" errors on the terminal (not every time) :
[  131.344500] radeon 0000:01:00.0: GPU lockup CP stall for more than 10000msec
[  131.344646] GPU lockup (waiting for 0x00000014 last fence id 0x00000012)
[  131.345898] radeon 0000:01:00.0: GPU softreset 
[  131.345908] radeon 0000:01:00.0:   GRBM_STATUS=0xB1403828
[  131.345919] radeon 0000:01:00.0:   GRBM_STATUS_SE0=0x28000007
[  131.345930] radeon 0000:01:00.0:   GRBM_STATUS_SE1=0x00000007
[  131.345942] radeon 0000:01:00.0:   SRBM_STATUS=0x200000C0
[  131.345963] radeon 0000:01:00.0:   GRBM_SOFT_RESET=0x00007F6B
[  131.346076] radeon 0000:01:00.0:   GRBM_STATUS=0x00003828
[  131.346081] radeon 0000:01:00.0:   GRBM_STATUS_SE0=0x00000007
[  131.346087] radeon 0000:01:00.0:   GRBM_STATUS_SE1=0x00000007
[  131.346092] radeon 0000:01:00.0:   SRBM_STATUS=0x200000C0
[  131.347097] radeon 0000:01:00.0: GPU reset succeed
[  131.428073] [drm] PCIE GART of 512M enabled (table at 0x0000000000040000).
[  131.428244] radeon 0000:01:00.0: WB enabled
[  131.428255] [drm] fence driver on ring 0 use gpu addr 0x40000c00 and cpu addr 0xffff8804238e8c00
[  131.444522] [drm] ring test on 0 succeeded in 1 usecs
[  131.444565] [drm] ib test on ring 0 succeeded in 1 usecs

Versions :
Fedora 17, up to date
Kernel : 3.3.7, but I also tried 3.4.0 (recently released) and even 3.5-rc1 (from rahwhide) : the bug was exactly similar. It has been present in all the kernel versions that I tested with that card (bought in september 2011)
#Xorg -version
X.Org X Server 1.12.0
Release Date: 2012-03-04
X Protocol Version 11, Revision 0
Build Operating System: x86-14 2.6.32-220.4.1.el6.x86_64 
Current Operating System: Linux vmserver.grelot.net 3.4.0-1.fc17.x86_64 #1 SMP Sun Jun 3 06:35:17 UTC 2012 x86_64

If necessary, I may use more recent kernels, bug I may need some help if it is necessary to apply a patch :-)

I hope this will help!
Comment 1 Goulou 2012-06-08 11:02:41 UTC
Created attachment 62807 [details]
lspci
Comment 2 Goulou 2012-06-08 11:04:24 UTC
Created attachment 62808 [details]
Xorg.log with normal kernel options
Comment 3 Goulou 2012-06-08 11:04:48 UTC
Created attachment 62809 [details]
Xorg.log with nomodeset kernel option
Comment 4 Alex Deucher 2012-06-08 12:30:04 UTC

*** This bug has been marked as a duplicate of bug 49792 ***

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.