Bug 195806

Summary: Wrong pci ids in via.xinf
Product: [Fedora] Fedora Reporter: Xavier Bachelot <xavier>
Component: xorg-x11-drv-viaAssignee: X/OpenGL Maintenance List <xgl-maint>
Status: CLOSED RAWHIDE QA Contact:
Severity: low Docs Contact:
Priority: medium    
Version: 5   
Target Milestone: ---   
Target Release: ---   
Hardware: All   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2006-06-19 21:55:23 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:

Description Xavier Bachelot 2006-06-17 22:48:48 UTC
If I understand correctly the meaning of via.xinf, the file contains 3 wrong
chipset ids (3204, 3205 and 7204), which are not unichrome graphic controllers,
and must probably be removed. Also 3022 is known to the driver but disabled.

It also lacks some others ids (3108, 3118, 3344), but these ones (unichrome pro)
are not yet supported by the driver in xorg, and thus should not be added.

See
http://webcvs.freedesktop.org/xorg/driver/xf86-video-via/src/via_id.h?view=markup
for reference.

Comment 1 Mike A. Harris 2006-06-19 19:18:13 UTC
(In reply to comment #0)
> If I understand correctly the meaning of via.xinf, the file contains 3 wrong
> chipset ids (3204, 3205 and 7204), which are not unichrome graphic controllers,
> and must probably be removed.

Good catch!

Reviewed and confirmed.  All of the *.xinf files were autogenerated
by a script, which took our old pcitable from the hwdata package and
extracted all video devices, putting each in their own file using the
new syntax.  It seems that the pcitable had some incorrect entries
as well, leading to this wonkiness.  ;o)

I've removed the invalid PCI IDs now.

> Also 3022 is known to the driver but disabled.

Left in the file, but commented out, with a description above it, and
a link to this bug report.

> It also lacks some others ids (3108, 3118, 3344), but these ones (unichrome pro)
> are not yet supported by the driver in xorg, and thus should not be added.

It appears that the 3108 and 3118 are supported, as far as the chip IDs
being present, and being used in the via_driver.c detection logic in the
0.2.1 driver, so I have added them.  If these really are broken, or should
be disabled however though, I can comment them out again if someone is
sure they should be disabled.

> See
> http://webcvs.freedesktop.org/xorg/driver/xf86-video-via/src/via_id.h?view=markup
> for reference.

The 0x3344 however is not listed in the ViaChipsets array, and it appears
from the link you provided that this is intentional, so I have added it
but left it disabled with a comment for now.  I also filed a bug upstream
as it should either be listed in the array and commented out for documentive
purposes, or it should be enabled.

https://bugs.freedesktop.org/show_bug.cgi?id=7272

The current via.xinf is now:


------8<--------------------------------------------------------------------
# NOTE: This file was last updated June 19, 2006.  It has been synchronized
# to match the support present and enabled in the upstream X.Org X11R7.1
# via driver release "xf86-video-via-0.2.1.tar.bz2", and to resolve
# the following issue:  https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=195806

# NOTE: Comments in via 0.2.1 driver source indicate this device does not
# exist in the wild, so it has been disabled for now.
#alias pcivideo:v00001106d00003022sv*sd*bc*sc*i* via

# 1106:3108 - K8M800 (PCI_CHIP_VT3204)
alias pcivideo:v00001106d00003108sv*sd*bc*sc*i* via

# 1106:3118 - PM800/PM880/CN400 (PCI_CHIP_VT3259)
alias pcivideo:v00001106d00003118sv*sd*bc*sc*i* via

# 1106:3122 - CLE266 (PCI_CHIP_CLE3122)
alias pcivideo:v00001106d00003122sv*sd*bc*sc*i* via

# 1106:7205 - KM400/KN400 (PCI_CHIP_VT3205)
alias pcivideo:v00001106d00007205sv*sd*bc*sc*i* via

# NOTE: This PCI ID is present in the 0.2.1 driver, but not part of the
# ViaChipset array, so we disable it.  Further investigation yields a commit
# message:  "Xv VM800 support (Ivor Hewitt). Not enabled yet pending more testing."
# http://webcvs.freedesktop.org/xorg/driver/xf86-video-via/src/via_id.h?view=markup
# 1106:3344 - VM800 (PCI_CHIP_VT3314)
#alias pcivideo:v00001106d00003344sv*sd*bc*sc*i* via
------8<--------------------------------------------------------------------


I'm commiting the changes to CVS now, and building a new package in
rawhide.  We are probably going to release X11R7.1 as an update for FC5
in the future as well.  Thanks again for reporting this inaccuracy.


Comment 2 Mike A. Harris 2006-06-19 21:55:23 UTC
Fixed in 0.2.1-4, thanks again.