On April 30, 2004 06:00 am, Branden Robinson wrote: > On Thu, Apr 29, 2004 at 09:14:58AM +0000, slaven peles wrote: > > Apr 29 08:25:01 ivor kernel: atkbd.c: Unknown key released (translated > > set 2, code 0x7a on isa0060/serio0). > > Apr 29 08:25:01 ivor kernel: atkbd.c: This is an XFree86 bug. It > > shouldn't access hardware directly. > > Apr 29 08:25:01 ivor kernel: atkbd.c: Unknown key released (translated > > set 2, code 0x7a on isa0060/serio0). > > Apr 29 08:25:01 ivor kernel: atkbd.c: This is an XFree86 bug. It > > shouldn't access hardware directly. > > On Thu, Apr 29, 2004 at 10:54:24AM -0500, Diego Enrique Rodriguez wrote: > > I have same problem look my log in > > http://uvirtual.ean.edu.co/~derodriguez/kernelprob/XFree86.0.log > > On Thu, Apr 29, 2004 at 03:52:18PM -0400, Josh Metzler wrote: > > This happens to me, too - Radeon 7000, same kernel. I filed bug #246587 > > against kernel-image-2.6.5-1-686. > > Hi guys, > > This should be fixed in the version of XFree86 that entered sid on 29 > April. > > xfree86 (4.3.0.dfsg.1-1) unstable; urgency=low > [...] > * Fix AT keyboard rate I/O controls by operating on the actual console > file descriptor, not on file descriptor zero (thanks, Keith Packard). > Suppresses warning messages from Linux 2.6. (Closes: #224909) [...] > -- Fabio M. Di Nitto <[EMAIL PROTECTED]> Wed, 28 Apr 2004 18:55:17 > +0200
I checked this again and it's definitely a firmware issue. Thanks to Michael Daenzer there is an easy fix for this. Download and install drm-trunk-module-src package from http://people.debian.org/~daenzer/dri-trunk-sid/ as well as appropriate kernel-headers-2.6.5 package. Then follow standard procedure (as a root): cd /usr/src tar xvfz drm-trunk.tar.gz export KVERS=2.6.5-XXX export KSRC=/usr/src/kernel-headers-2.6.5-XXX cd modules/drm-trunk debian/rules kdist_image This creates drm-trunk-module...deb package in /usr/src directory, which can be easily installed with dpkg -i (no --force-overwrite necessary). At least, this worked for me ;-). Cheers, Slaven
