Bug 3127 - Savage at 24 bit color depth shows "banding"
Summary: Savage at 24 bit color depth shows "banding"
Status: RESOLVED FIXED
Alias: None
Product: xorg
Classification: Unclassified
Component: Driver/savage (show other bugs)
Version: 6.8.2
Hardware: x86 (IA32) Linux (All)
: high normal
Assignee: Xorg Project Team
QA Contact:
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2005-04-26 03:03 UTC by Jaime
Modified: 2005-05-12 21:46 UTC (History)
2 users (show)

See Also:
i915 platform:
i915 features:


Attachments
Output from xdpyinfo (3.94 KB, text/plain)
2005-04-26 03:11 UTC, Jaime
no flags Details
Xorg log file (34.40 KB, text/plain)
2005-04-26 03:11 UTC, Jaime
no flags Details
Patch referred to in "additional comment #6" (1.52 KB, patch)
2005-05-13 14:46 UTC, Jaime
re2824: 6.8-branch?
Details | Splinter Review

Description Jaime 2005-04-26 03:03:16 UTC
Hi.

I have a problem which only shows up when using 24-bit depth on my 86C270-294
Savage/IX-MV. Basically, in 24-bit mode, smooth gradients appear "banded", as
though least significant bits of colour are being discarded (but I'm not an xorg
dev, so I don't know how these things work).

Firstly, the problem does not show up at all in screenshots (they look perfect
when viewed on another machine). Here are a few pictures of screens with the
problem taken with a digital camera (excuse the quality):

https://bugzilla.ubuntu.com/attachment.cgi?id=1293
https://bugzilla.ubuntu.com/attachment.cgi?id=1294
https://bugzilla.redhat.com/beta/attachment.cgi?id=100722

This problem has been reported in different places in the past, so there may be
additional information that might help here:

https://bugzilla.ubuntu.com/show_bug.cgi?id=6127
https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=88264
http://forums.gentoo.org/viewtopic.php?t=104293

I'll post an xdpyinfo output and an xorg log now.

Jaime
Comment 1 Jaime 2005-04-26 03:11:01 UTC
Created attachment 2557 [details]
Output from xdpyinfo

This is output from xdpyinfo running at 24 bit colour depth
Comment 2 Jaime 2005-04-26 03:11:59 UTC
Created attachment 2558 [details]
Xorg log file

Xorg log file
Comment 3 Alex Deucher 2005-05-04 06:40:11 UTC
I fixed this in xorg cvs.  The CLUT was left in 6 bit mode.  Binary snapshots
are available here:
http://dri.freedesktop.org/wiki/Download#head-55420c59a1c2e9a70f07a6fa02f0d228ffb87b76

I'm going to close this bug, if you still have problems, please re-open.
Comment 4 Jaime 2005-05-05 19:31:41 UTC
Alex,

24-bit colour now works perfectly. I extracted savage_drv.o from the binary
snapshot available at:
http://dri.freedesktop.org/snapshots/savage-20050504-linux.i386.tar.bz2
and swapped it for the file:
/usr/X11R6/lib/modules/drivers/savage_drv.o
on my "ubuntu hoary" (which uses x.org 6.8.2) and the problem has now disappeared.

Thank you very much, Jaime :-)
Comment 5 Mike A. Harris 2005-05-06 14:47:24 UTC
Will this be backported to 6.8.x?  If not, Savage users wont get this
fix until X11R7 (unless they apply the patch themselves and rebuild
X 6.8.x).
Comment 6 Alex Deucher 2005-05-06 15:57:15 UTC
(In reply to comment #5)
> Will this be backported to 6.8.x?  If not, Savage users wont get this
> fix until X11R7 (unless they apply the patch themselves and rebuild
> X 6.8.x).

feel free to nominate it if I don't get around to it.  trivial patch:
http://cvs.freedesktop.org/xorg/xc/programs/Xserver/hw/xfree86/drivers/savage/savage_driver.c?r1=1.27&r2=1.28
Comment 7 Jaime 2005-05-13 14:46:35 UTC
Created attachment 2677 [details] [review]
Patch referred to in "additional comment #6"

Hi. I don't understand the "nomination" process, but I'm going to attempt to
get this patch nominated...

Jaime :-)


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.