Bug 679373

Summary: intel [915GM] Minor text rendering corruption
Product: [openSUSE] openSUSE 11.4 Reporter: Neil Darlow <neil>
Component: X.OrgAssignee: E-mail List <xorg-maintainer-bugs>
Status: RESOLVED WONTFIX QA Contact: E-mail List <xorg-maintainer-bugs>
Severity: Critical    
Priority: P3 - Medium CC: Martin.Seidler, neil
Version: Final   
Target Milestone: ---   
Hardware: i586   
OS: Other   
Whiteboard:
Found By: --- Services Priority:
Business Priority: Blocker: ---
Marketing QA Status: --- IT Deployment: ---
Attachments: Corrupted text with "intel" driver
Intel 915GM 1400x1050 driver i915: Brocken fonts with horizonal lines in firefox

Description Neil Darlow 2011-03-14 13:24:04 UTC
User-Agent:       Mozilla/5.0 (X11; Linux i686; rv:2.0b12) Gecko/20110222 Firefox/4.0b12

The release notes for openSUSE-11.4 direct you to use "intellegacy" in the case that the "intel" Xorg driver causes problems.

For Intel 915G(M) graphics (in my case an EeePC 900) the intellegacy driver fixes a problem with broken text rendering but there are issues with openSUSE-11.4:

1) In GNOME it causes GDM to endlessly crash and respawn. You cannot login.
2) In KDE it causes some applications e.g. MozillaFirefox and Chromium to crash the session. You are returned to the KDM login.

The regular intel driver does not exhibit the behaviour of (1) and (2) above but character rendering is broken.

Another issue with 915G(M) graphics and openSUSE-11.4 is that performance is considerably worse than with openSUSE-11.3. Desktop Effects are no longer possible on my EeePC 900 because you have to wait > 5 seconds for any GIU action to happen after initiating it. This does not seem to be the case with GNOME Desktop Effects (Compiz) which work rather well. I assume this to be a problem with the new Plasma of KDE-4.6.0.

Reproducible: Always

Steps to Reproduce:
1. Install openSUSE-11.4 KDE i586 on a 915G(M) equipped system
2. Edit /etc/X11/xorg.conf.d/50-device.conf to add -> Driver "intellegacy"
3. Restart X and login to KDE
4. Execute MozillaFirefox-4-Beta12
Actual Results:  
The session crashes and you are returned to KDM login after executing firefox.

Expected Results:  
Firefox should execute normally - which it does with the "intel" driver.

This is a fresh install of openSUSE-11.4 with the only repositories added being:
openSUSE BuildService - KDE:Extra for akonadi-googledata and libgcal0
Main Repository (Contrib) for the crash test with Chromium
Comment 1 Stefan Dirsch 2011-03-14 14:07:34 UTC
What I'm wondering is how could you figure out that compiz on top of GNOME works when you can't login in into GNOME?
Comment 2 Neil Darlow 2011-03-14 14:14:06 UTC
(In reply to comment #1)
> What I'm wondering is how could you figure out that compiz on top of GNOME
> works when you can't login in into GNOME?

You can login to GNOME with the standard "intel" driver. Read the report again, it is in there :-;
Comment 3 Stefan Dirsch 2011-03-14 14:23:07 UTC
Ok. So this behaviour was with intel driver. That wasn't obvious to me at all. Remains the broken text rendering. Could you attach a screenshot for that?
Comment 4 Neil Darlow 2011-03-14 15:03:44 UTC
No!

The Firefox/Chromium crash are with KDE and intellegacy.
The GDM login failure is with GNOME and intellegacy.

The text curruption is with intel, which I af forced to run at the moment.
Comment 5 Stefan Dirsch 2011-03-14 15:28:26 UTC
Ok, but please attach a screenshot for the text rendering corruption with intel driver.
Comment 6 Neil Darlow 2011-03-14 17:23:24 UTC
Created attachment 419193 [details]
Corrupted text with "intel" driver

YaST Repositories listing with "intel" driver.

Notice the corruption of the title text at the top of the image. This corruption can happen in any application where text is rendered.
Comment 7 Neil Darlow 2011-03-14 18:04:18 UTC
So, a bug reported for crashes in KDE and GNOME with "intellegacy" has been transformed into a text rendering one with "intel".

Does that imply that you prefer to support "intel" over "intellegacy"?

What about the Release Notes advice to use "intellegacy" where "intel" causes problems?

Ignoring the crashes could cause problems for other users.
Comment 8 Stefan Dirsch 2011-03-14 18:11:58 UTC
You need to know that intellegacy is no longer in development. So nobody is fixing
any issues in this driver any longer. Still it can be useful for users, if intel driver doesn't work, but intellegacy does.
Comment 9 Martin Seidler 2011-03-28 18:25:55 UTC
Created attachment 421740 [details]
Intel 915GM 1400x1050 driver i915: Brocken fonts with horizonal lines in firefox

Seems to me my laptop/system has the same/a simular bug (sometimes).
If the fonts are broken I could make the horizontal with lines disappear by  enlarging the fonts (Contr and + in firefox).

openSUSE 11.4 (factory->zypper dup) 2.6.37.1-1.2-default (mainly) GNOME 2.32.1;
Samsung X20 Pentium M 740 Intel 915GM 1400x1050


# lspci -vnn
[...]

00:02.0 VGA compatible controller [0300]: Intel Corporation Mobile 915GM/GMS/910GML Express Graphics Controller [8086:2592] (rev 03) (prog-if 00 [VGA controller])
	Subsystem: Samsung Electronics Co Ltd Device [144d:c01a]
	Flags: bus master, fast devsel, latency 0, IRQ 16
	Memory at b0080000 (32-bit, non-prefetchable) [size=512K]
	I/O ports at 1800 [size=8]
	Memory at c0000000 (32-bit, prefetchable) [size=256M]
	Memory at b0000000 (32-bit, non-prefetchable) [size=256K]
	Expansion ROM at <unassigned> [disabled]
	Capabilities: [d0] Power Management version 2
	Kernel driver in use: i915

00:02.1 Display controller [0380]: Intel Corporation Mobile 915GM/GMS/910GML Express Graphics Controller [8086:2792] (rev 03)
	Subsystem: Samsung Electronics Co Ltd Device [144d:c01a]
	Flags: fast devsel
	Memory at 86000000 (32-bit, non-prefetchable) [disabled] [size=512K]
	Capabilities: [d0] Power Management version 2
[...]
Comment 10 Stefan Dirsch 2011-09-01 14:21:22 UTC
Unfortunately we do not have the ressources to address that issue
still for openSUSE 11.4. Could you please test again with
a current Milestone of openSUSE 12.1, whether the issue still
exists and give me feedback about the result? Thanks.
Comment 11 Neil Darlow 2011-09-01 19:32:42 UTC
I don't currently have the hardware to test an additional milestone.

I have found a workaround for the problem I experienced. Adding the following to /etc/X11/xorg.conf.d/50-device does the trick:

Section "Device"
    ...
    Option "DebugWait"
    ...
EndSection
Comment 12 Stefan Dirsch 2011-09-01 20:39:38 UTC
Thanks!
Comment 13 Stefan Dirsch 2013-03-05 10:25:53 UTC
openSUSE 11.4 is no longer supported. Feel free to reopen, if the issue still exists with openSUSE 12.3. Thanks.