Bug 238141

Summary: i810 that used to work with older x11 fails to be configured properly
Product: [Fedora] Fedora Reporter: Lubomir Kundrak <lkundrak>
Component: xorg-x11-drv-i810Assignee: Adam Jackson <ajax>
Status: CLOSED RAWHIDE QA Contact:
Severity: high Docs Contact:
Priority: high    
Version: 6CC: mcepl, xgl-maint
Target Milestone: ---Keywords: HardwareEnablement, Regression
Target Release: ---   
Hardware: All   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2007-09-21 13:03:45 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Attachments:
Description Flags
Differencies between the logs of old and updated i810 driver
none
Config files and logs from i810 on FC4 and FC6
none
Xorg.0.log from FC6
none
Xorg.0.log from FC4
none
Xorg.0.log.rpath-fc6
none
Xorg.0.log.rpath-rpath
none
xorg.conf.fc6
none
xorg.conf.rpath none

Description Lubomir Kundrak 2007-04-27 13:56:18 UTC
Description of problem:

Symptom of the problem is that screen of a Daewoo 1509B monitor connected to
a Intel 82845 just gets blank and power led blinks with configuration
generated by anaconda (xorg.conf.fc6, Xorg.0.log.fc6-fc6). The same
applies for no configuration file, and also for one generated by
X -configure.

I've also run rPath Linux 1.0.5 (which is, I believe, derived from FC5),
and there things work fine. It has XOrg 6.8.2, and there everything
worked (Xorg.0.log.rpath-rpath) with anaconda-generated configuration
(xorg.conf.rpath). The same configuration did not work in FC6
(Xorg.0.log.rpath-fc6). Also, the configuration generated by FC6
anaconda did not work in rPath (Xorg.0.log.fc6-rpath).

Version-Release number of selected component (if applicable):

Although the log files are not from latest update, the same thing
happens with everything updated.

Additional info:

Everything here applies for "1024x768" resolution. When I add Mode
"800x600" to SubSection "Display", things work also in FC6. BPP seem
not to make any difference.

Comment 1 Lubomir Kundrak 2007-04-27 14:10:37 UTC
What is in FC6 log and not in rPath one:

No idea what's this:
(WW) I810(0): Extended BIOS function 0x5f05 failed.
(II) I810(0): BIOS call 0x5f05 not supported, setting refresh with VBE 3 method.

This is certainly incorrect:
(==) I810(0): VideoRAM: 65536 kByte
rPath's x11 reported correctly:
(==) I810(0): VideoRAM: 32768 kByte

No idea, either.
(WW) EDID preferred timing clock 94.50MHz exceeds claimed max 90MHz, fixing

Whis is weird. I do not have that much video memory:
(II) I810(0): Allocated 52992 kB for textures at 0x620000
Versus old x11:
(II) I810(0): Allocated 32 kB for the logical context at 0x73f8000.

And here... Certainly not right too. IRQ 177...
(II) I810(0): [drm] Initialized kernel agp heap manager, 54263808
(II) I810(0): [drm] dma control initialized, using IRQ 177
versus
(II) I810(0): [drm] dma control initialized, using IRQ 5
(II) I810(0): [drm] Initialized kernel agp heap manager, 21757952

Comment 2 Lubomir Kundrak 2007-04-28 07:05:19 UTC
I lied. rPath is not derived from FC5.
And FC5's 7.0.0 also has the bug.

Comment 3 Lubomir Kundrak 2007-04-28 11:30:14 UTC
Two more things:

The smolt profile of the machine is here:
http://smolt.fedoraproject.org/show?UUID=548e727f-f164-4e32-8e91-c194a060ddc0

When using "intel" driver instead of "i810" nothing changes, the amount ov
VideoRAM detected by the X11 server is still incorrect (64M).

Comment 4 Lubomir Kundrak 2007-04-28 15:55:31 UTC
After I updated the machine's firmware, X11 server no longer complains about
invalid V_BIOS checksum, as can be seen in the attached logs. Nothing else
changed.

Comment 5 Lubomir Kundrak 2007-04-29 17:53:11 UTC
Created attachment 153738 [details]
Differencies between the logs of old and updated i810 driver

The problem doesn't appear in xorg-x11-6.8.2-31.FC4, but it does in
xorg-x11-6.8.2-37.FC4.49.4. The difference that is suspect to be guilty is the
patch xorg-x11-6.8.2-add-i945-support.patch.

I attach the differencies between the logs of the two.

Comment 6 Matěj Cepl 2007-05-02 21:53:15 UTC
Hi, Lubomir, could we get full /var/log/Xorg.*.log and /etc/X11/xorg.conf, please?

Comment 7 Lubomir Kundrak 2007-05-03 08:38:11 UTC
Created attachment 154012 [details]
Config files and logs from i810 on FC4 and FC6

Sure. Pardon me for not providing those before, I have even collected and
packed them, but forgot to attach.

In the tarball attached, "rPath" basically corresponds to FC4 without the
patch.

Comment 8 Matěj Cepl 2007-05-28 21:38:09 UTC
Created attachment 155560 [details]
Xorg.0.log from FC6

Comment 9 Matěj Cepl 2007-05-28 21:38:49 UTC
Created attachment 155561 [details]
Xorg.0.log from FC4

Comment 10 Matěj Cepl 2007-05-28 21:40:32 UTC
Created attachment 155562 [details]
Xorg.0.log.rpath-fc6

Comment 11 Matěj Cepl 2007-05-28 21:41:23 UTC
Created attachment 155563 [details]
Xorg.0.log.rpath-rpath

Comment 12 Matěj Cepl 2007-05-28 21:41:59 UTC
Created attachment 155564 [details]
xorg.conf.fc6

Comment 13 Matěj Cepl 2007-05-28 21:42:46 UTC
Created attachment 155565 [details]
xorg.conf.rpath

Comment 14 Lubomir Kundrak 2007-09-21 11:03:37 UTC
Section "Device"
        Identifier      "Videocard0"
        Driver          "i810"
        Option          "NoDDC"
EndSection

Section "Monitor"
     Identifier         "Monitor0"
     VertRefresh        56 - 75
     HorizSync          30 - 69
EndSection

Fixed that then (NoDCC was needed). The problem was only with that particular
monitor, that I'll send for a recyclation in a week or so :)

Comment 15 Matěj Cepl 2007-09-21 11:44:14 UTC
OK, so should I close this as INSUFFICIENT_DATA?

Comment 16 Lubomir Kundrak 2007-09-21 13:03:45 UTC
Works alright with rawhide.