Bug 115281 - fglrx segfaults if too many ModeLines are present
Summary: fglrx segfaults if too many ModeLines are present
Status: RESOLVED FIXED
Alias: None
Product: SUSE LINUX 10.0
Classification: openSUSE
Component: X11 3rd Party (show other bugs)
Version: Beta 4
Hardware: Other All
: P5 - None : Normal
Target Milestone: ---
Assignee: Jammy Zhou
QA Contact: Stefan Dirsch
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2005-09-05 10:24 UTC by Matthias Hopf
Modified: 2006-12-05 20:02 UTC (History)
1 user (show)

See Also:
Found By: Other
Services Priority:
Business Priority:
Blocker: ---
Marketing QA Status: ---
IT Deployment: ---

Attachments
Working xorg.conf (6.77 KB, text/plain)
2005-09-05 10:24 UTC, Matthias Hopf 		Details
Logfile for working config (42.06 KB, text/plain)
2005-09-05 10:25 UTC, Matthias Hopf 		Details
Non-working xorg.conf (7.68 KB, text/plain)
2005-09-05 10:26 UTC, Matthias Hopf 		Details
Logfile for non-working config (30.05 KB, text/plain)
2005-09-05 10:27 UTC, Matthias Hopf 		Details
Output of hwinfo --gfxcard (926 bytes, text/plain)
2005-09-05 10:27 UTC, Matthias Hopf 		Details
Output of hwinfo --monitor (387 bytes, text/plain)
2005-09-05 10:27 UTC, Matthias Hopf 		Details
Output of lspci -vv (10.41 KB, text/plain)
2005-09-05 10:28 UTC, Matthias Hopf 		Details
View All Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.
Description Matthias Hopf 2005-09-05 10:24:15 UTC 
fglrx 8.16.20 does not work with openSuSE 10.0 when too many ModeLines are
present. Additionally it behaves strange, if ModeLines are set at all.
I tested everything on an IBM T42p, a fairly common laptop.

I'm not exactly sure what's going on, but when the driver fails

1st) it does not load the xaa, fglrxdrm, fb, vbe and ddc modules any
     more (getting loads of unresolved symbols, which vanish if I add
     these to the 'Module' section by hand).

2nd) it segfaults, leaving the screen in an unusable state (even if the
     missing modules have been added by hand).

This all has to do with the number of ModeLines.
Find two configurations and log files attached, one working and one non
working. Also find the output of lspci -vv, hwinfo --gfxcard, and hwinfo
--monitor attached..

3rd) With both configs I do *not* get the native resolution of 1600x1200 on
     this display! The radeon driver works flawlessly BTW.
Comment 1 Matthias Hopf 2005-09-05 10:24:57 UTC 
Created attachment 48768 [details]
Working xorg.conf
Comment 2 Matthias Hopf 2005-09-05 10:25:26 UTC 
Created attachment 48769 [details]
Logfile for working config
Comment 3 Matthias Hopf 2005-09-05 10:26:33 UTC 
Created attachment 48770 [details]
Non-working xorg.conf
Comment 4 Matthias Hopf 2005-09-05 10:27:12 UTC 
Created attachment 48771 [details]
Logfile for non-working config
Comment 5 Matthias Hopf 2005-09-05 10:27:40 UTC 
Created attachment 48772 [details]
Output of hwinfo --gfxcard
Comment 6 Matthias Hopf 2005-09-05 10:27:58 UTC 
Created attachment 48773 [details]
Output of hwinfo --monitor
Comment 7 Matthias Hopf 2005-09-05 10:28:23 UTC 
Created attachment 48774 [details]
Output of lspci -vv
Comment 8 Michael Gross 2005-09-05 16:11:38 UTC 
Please check the status of this bug (`new' vs. `assigned').
Comment 9 Matthias Hopf 2005-09-05 16:15:19 UTC 
Matthew is no member of SuSE.

This bug is for tracking 3rd party related driver bugs only. I don't really know
about ATI's submission to fixing these bugs.
Comment 10 Matthias Hopf 2005-09-05 16:16:19 UTC 
Ok, in order to reduce conflicts, setting this to normal, because we do have a
workaround in SL 10b4.