Bug#493185: v86d doesn't work from initramfs
On 2009-04-23 23:13 +0200, Evgeni Golov wrote: I've just uploaded 0.1.9 and I think I've found the solution for our problem :) Instead of adding anything to /etc/initramfs-tools/modules, just append video=uvesafb:mode_option=MODE to your kernel boot parameters. With this and a recent kernel (I tried 2.6.29 from Debian and a self-build 2.6.30-rc3), uvesafb works just fine (even without building v86d with x86emu or klibc). Could you test this on your setups and tell me whether it works or not? It does indeed seem to work. Incidentally, the way documented in Documentation/fb/uvesafb.txt: video=uvesafb:1280x1024-32 does not quite work right, it only gave me a 640x480 screen, but video=uvesafb:mode_option=1280x1024-32 worked fine. Anyway, I had stopped using uvesafb on my PC anyway, because - my crappy graphics card BIOS ignores refresh rate settings, always using 60 Hz like with vesafb; - v86d does not work on i386 systems with a 64-bit kernel because of #492470. Regards, Sven -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#493185: v86d doesn't work from initramfs
On Fri, Apr 24, 2009 at 09:57:19AM +0200, Sven Joachim wrote: On 2009-04-23 23:13 +0200, Evgeni Golov wrote: I've just uploaded 0.1.9 and I think I've found the solution for our problem :) Instead of adding anything to /etc/initramfs-tools/modules, just append video=uvesafb:mode_option=MODE to your kernel boot parameters. With this and a recent kernel (I tried 2.6.29 from Debian and a self-build 2.6.30-rc3), uvesafb works just fine (even without building v86d with x86emu or klibc). Could you test this on your setups and tell me whether it works or not? It does indeed seem to work. Incidentally, the way documented in Documentation/fb/uvesafb.txt: video=uvesafb:1280x1024-32 does not quite work right, it only gave me a 640x480 screen, but video=uvesafb:mode_option=1280x1024-32 worked fine. Yeah, this is #439846 :( Anyway, I had stopped using uvesafb on my PC anyway, because - my crappy graphics card BIOS ignores refresh rate settings, always using 60 Hz like with vesafb; - v86d does not work on i386 systems with a 64-bit kernel because of #492470. Mh, in the meantime I dont need to rebuild with x86emu for my Radeon X1400. But I dont have amd64 either, maybe Corsac want's to test this? :) Regards -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#493185: v86d doesn't work from initramfs
On ven, 2009-04-24 at 10:19 +0200, Evgeni Golov wrote: - v86d does not work on i386 systems with a 64-bit kernel because of #492470. Mh, in the meantime I dont need to rebuild with x86emu for my Radeon X1400. But I dont have amd64 either, maybe Corsac want's to test this? :) I don't have i386 userlands :) Cheers, -- Yves-Alexis -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#493185: v86d doesn't work from initramfs
Hey, I've just uploaded 0.1.9 and I think I've found the solution for our problem :) Instead of adding anything to /etc/initramfs-tools/modules, just append video=uvesafb:mode_option=MODE to your kernel boot parameters. With this and a recent kernel (I tried 2.6.29 from Debian and a self-build 2.6.30-rc3), uvesafb works just fine (even without building v86d with x86emu or klibc). Could you test this on your setups and tell me whether it works or not? Regards Evgeni Golov -- Bruce Schneier Fact Number 433: Bruce Schneier is chock full of midichlorians. pgp0Lsqg0DzCe.pgp Description: PGP signature
Bug#493185: v86d doesn't work from initramfs
I workedaround this issue by removing the vga=... from the kernel command line . Having it with the video= switch made it fails this way under my setup (could be a conflict with vesafb legacy). Is this the same issue ? Cheers, Alban -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Bug#493185: v86d doesn't work from initramfs
On 2008-08-01 09:31 +0200, Evgeni Golov wrote:
Hi Corsac,
On Fri, 01 Aug 2008 08:31:14 +0200 Yves-Alexis Perez wrote:
I tried to use the new v86d from the initramfs so I could have a framebuffer
in early stages, but it doesn't work.
[...]
Aug 1 08:11:06 hidalgo kernel: agpgart: Detected an Intel 965GM Chipset.
Aug 1 08:11:06 hidalgo kernel: agpgart: Detected 7676K stolen memory.
Aug 1 08:11:06 hidalgo kernel: agpgart: AGP aperture is 256M @ 0xe000
Aug 1 08:11:06 hidalgo kernel: uvesafb: Getting VBE info block failed
(eax=0x4f00, err=-3)
Aug 1 08:11:06 hidalgo kernel: uvesafb: vbe_init() failed with -22
Aug 1 08:11:06 hidalgo kernel: uvesafb: probe of uvesafb.0 failed with
error -22
I have the same behaviour on my Z61 (but not my X31). I have to load
v86d after the real system is up (and before Xorg, as the radeon driver
does something which lets uvesafb segfault).
There seems to be some timing problem. I followed the instructions on [1]
to get uvesafb working on my laptop which has a 1280x800 resolution,
and it worked out perfectly. When I tried to adapt it to my desktop PC
which has a 1280x1024 resolution and does not need to run 915resolution
before loading the uvesafb module, it failed:
[/etc/initramfs-tools/scripts/init-top/framebuffer, modified from [1]]
#!/bin/sh
PREREQ=
prereqs()
{
echo $PREREQ
}
case $1 in
# get pre-requisites
prereqs)
prereqs
exit 0
;;
esac
echo fbhack
for x in $(cat /proc/cmdline); do
case $x in
fbhack|video=*)
echo creating /dev/mem file
mknod /dev/mem c 1 1
if [ -e /dev/zero ] ; then
echo Dev Zero Exists
else
mknod /dev/zero c 1 5
fi
if [ -e /dev/tty1 ] ; then
echo tty1 exists
else
mknod /dev/tty1 c 4 1
fi
if [ -e /dev/console ] ; then
echo console exists
else
mknod /dev/console c 5 1
fi
depmod -a
echo modprobe uvesafb
chk=`cat /proc/modules | grep uvesafb`
if [ x$chk != x ] ; then
rmmod uvesafb
fi
/sbin/modprobe uvesafb mode=1280x1024-16
echo Done!
;;
esac
done
This failed with uvesafb: probe of uvesafb.0 failed with error -22.
However, inserting a sleep 1 before the
/sbin/modprobe uvesafb mode=1280x1024-16 line does the trick.
Regards,
Sven
[1] http://wiki.debian.org/DebianEeePC/HowTo/SplashyWithDmcryptAndStandardGrub
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#493185: v86d doesn't work from initramfs
Package: v86d Version: 0.1.5.2-1 Severity: normal Hey, I tried to use the new v86d from the initramfs so I could have a framebuffer in early stages, but it doesn't work. Logs say: Aug 1 08:11:06 hidalgo kernel: agpgart: Detected an Intel 965GM Chipset. Aug 1 08:11:06 hidalgo kernel: agpgart: Detected 7676K stolen memory. Aug 1 08:11:06 hidalgo kernel: agpgart: AGP aperture is 256M @ 0xe000 Aug 1 08:11:06 hidalgo kernel: uvesafb: Getting VBE info block failed (eax=0x4f00, err=-3) Aug 1 08:11:06 hidalgo kernel: uvesafb: vbe_init() failed with -22 Aug 1 08:11:06 hidalgo kernel: uvesafb: probe of uvesafb.0 failed with error -22 As you may know I run a Thinkpad T61 full intel and can provide other stuff. I didn't specify a mode in the /etc/initramfs-tools/modules line because anyway bios doesn't send 1400x1050 as a valid resolution for my screen. If you need more info, please ask :) Cheers, -- Yves-Alexis -- System Information: Debian Release: lenny/sid APT prefers unstable APT policy: (500, 'unstable'), (500, 'testing'), (1, 'experimental') Architecture: amd64 (x86_64) Kernel: Linux 2.6.26-1-amd64 (SMP w/2 CPU cores) Locale: LANG=fr_FR.UTF-8, LC_CTYPE=fr_FR.UTF-8 (charmap=UTF-8) Shell: /bin/sh linked to /bin/bash Versions of packages v86d depends on: ii libc6 2.7-13 GNU C Library: Shared libraries ii libx86-1 1.1+ds1-2 x86 real-mode library Versions of packages v86d recommends: ii initramfs-tools 0.92e tools for generating an initramfs v86d suggests no packages. -- no debconf information -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#493185: v86d doesn't work from initramfs
Hi Corsac, On Fri, 01 Aug 2008 08:31:14 +0200 Yves-Alexis Perez wrote: I tried to use the new v86d from the initramfs so I could have a framebuffer in early stages, but it doesn't work. [...] Aug 1 08:11:06 hidalgo kernel: agpgart: Detected an Intel 965GM Chipset. Aug 1 08:11:06 hidalgo kernel: agpgart: Detected 7676K stolen memory. Aug 1 08:11:06 hidalgo kernel: agpgart: AGP aperture is 256M @ 0xe000 Aug 1 08:11:06 hidalgo kernel: uvesafb: Getting VBE info block failed (eax=0x4f00, err=-3) Aug 1 08:11:06 hidalgo kernel: uvesafb: vbe_init() failed with -22 Aug 1 08:11:06 hidalgo kernel: uvesafb: probe of uvesafb.0 failed with error -22 I have the same behaviour on my Z61 (but not my X31). I have to load v86d after the real system is up (and before Xorg, as the radeon driver does something which lets uvesafb segfault). Could you please try the following: 1. Load uvesafb from /etc/modules and not from initrd 2. Rebuild v86d without 01_use-external-libx86.patch and test in initrd again. 3. And outside of initrd too. When no of the three will help, I'll prepare you an unpatched 0.1.6 from upstream for testing. Regards Evgeni pgpllz3DNwcmw.pgp Description: PGP signature
