RV515 card failed to output video on VGA port
Dear Radeonhd developer, I have met a problem on my RV515 card, below is the steps to reproduce it: 1) connect a DFP monitor to DVI port and a CRT monitor to VGA port 2) boot ubuntu 7.10 3) startx with radeonhd 4) DFP monitor start alright, while my CRT monitor failed to put out video. (attach Xorg.0.log-CRT-fail for detailed information) Here is another test that will bring up video on CRT monitor 1) only connect CRT monitor to VGA port 2) boot ubuntu 7.10 3) startx with radeonhd 4) CRT monitor starts alright (attach Xorg.0.log-CRT-OK for detailed information) Thanks as always for your great help and hard work! Best Regards, Lisa Wu
Your logs indicate:
RADEONHD: version 0.0.2, built from
Hm. We don't know which git version you built the driver from, which is
crucial for debugging. Be sure that there is git installed and you're
building from the source directory where you actually did the checkout.
Matthias
--
Matthias Hopf
Matthias Hopf wrote:
Your logs indicate:
RADEONHD: version 0.0.2, built from
Hm. We don't know which git version you built the driver from, which is crucial for debugging. Be sure that there is git installed and you're building from the source directory where you actually did the checkout.
I guess I screwed up there in some way. I hope the following makes the git version detection more robust and, in case of failure, easier to debug by sending along the generated git_version.h: http://radeonhd.lauft.net/patches/ndim-git-version/ -- Hans Ulrich Niedermann
On Nov 15, 07 16:10:09 +0100, Hans Ulrich Niedermann wrote:
Matthias Hopf wrote:
Your logs indicate:
RADEONHD: version 0.0.2, built from
Hm. We don't know which git version you built the driver from, which is crucial for debugging. Be sure that there is git installed and you're building from the source directory where you actually did the checkout.
I guess I screwed up there in some way.
I hope the following makes the git version detection more robust and, in case of failure, easier to debug by sending along the generated git_version.h:
Did you think of packaging git_version.h when doing 'make dist'?
CU
Matthias
--
Matthias Hopf
Matthias Hopf wrote:
On Nov 15, 07 16:10:09 +0100, Hans Ulrich Niedermann wrote:
I hope the following makes the git version detection more robust and, in case of failure, easier to debug by sending along the generated git_version.h:
Did you think of packaging git_version.h when doing 'make dist'?
Yes, but we will want to distinguish a) a build directly from a git worktree of commit 12345678 b) a build from a dist tarball generated from commit 12345678 In case a), we can programmatically determine whether the local source contains changes in addition to 12345678 and note them. In case b), we have no way of finding out whether/what has changed locally, so we need to note that in the message somehow. I do have code for b), but it is too ugly, too fragile and too large to commit just yet. -- Hans Ulrich Niedermann
Hans Ulrich Niedermann wrote:
Matthias Hopf wrote:
Did you think of packaging git_version.h when doing 'make dist'?
Yes, but we will want to distinguish
a) a build directly from a git worktree of commit 12345678 b) a build from a dist tarball generated from commit 12345678
In case a), we can programmatically determine whether the local source contains changes in addition to 12345678 and note them.
In case b), we have no way of finding out whether/what has changed locally, so we need to note that in the message somehow.
I do have code for b), but it is too ugly, too fragile and too large to commit just yet.
I think the following addresses all that: http://radeonhd.lauft.net/patches/ndim-git-version/ or $ git pull \ http://radeonhd.lauft.net/xf86-video-radeonhd.git/ \ ndim-git-version 1. Ship git_version.h in dist tarball, marked as being part of dist tarball. 2. If git_version.h exists and if git_version.sh finds no git repo (.git dir), keep existing git_version.h. 3. Have git_version.h print special message when building from dist tarball. Please pull from ndim-git-version and push to master and/or initial-randr if you like it. -- Hans Ulrich Niedermann
Matthias Hopf wrote: +AD4- Your logs indicate: +AD4- +AD4APg- RADEONHD: version 0.0.2, built from +AD4- +AD4- Hm. We don't know which git version you built the driver from, which +AD4- is crucial for debugging. Be sure that there is git installed and +AD4- you're building from the source directory where you actually did the checkout. I get the code by using git clone: git://anongit.freedesktop.org/git/xorg/driver/xf86-video-radeonhd and I am building from the source directory where I actually did the checkout. However I didn't find a src/git+AF8-version.h in the source code.
From git log, the latest commit of my source code is: +ACI-commit 8ac590882a822a02fd15d7671d6796f6a1a54e13 Author: Luc Verhaegen +ADw-libv+AEA-skynet.be+AD4- Date: Wed Nov 14 17:11:23 2007 +-0100 TMDSA: create a table for electrical adjustments.
Create a pci device id based table with electrical values for all currently known devices with TMDSA. Warn verbosely about unknown devices. TMDSB now also includes support for the RV530 on the Dell low profile card.+ACI- Best Regards, Lisa Wu -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org
Hi Lisa, There seems to be an issue with the DAC code for your specific graphics card. Can you provide me with a copy of your ROM? Thanks, Luc Verhaegen. SUSE/Novell X Driver Developer. -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org
Just a note on a successfull installation and the xorg log in case it adds value anywhere... Compared to fglrx which I use - colors are a bit on the cold side and the speed is somewhat not there yet, but that is understandable. I'm not sure what this line means though: TMDSBVoltageControl: unhandled chipset: 0x7142 (apart from not being recognized at an earlier stage). If these drivers won't make system freeze occasionally as fglrx does, it's for keeps. :)
It's probably to soon to report on this, but since I'm not an expert, here I go. One thing that isn't working, though, is testing video in skype. I'm not sure what the problem is. This is from syslog: Nov 19 23:50:07 slash kernel: ioctl32(skype:5825): Unknown cmd fd(31) cmd(c0cc5616){t:'V';sz:204} arg(f417ae40) on /dev/video0 Nov 19 23:50:07 slash kernel: compat_ioctl32: v4l2 ioctl VIDIOC_S_PARM, dir=rw (0xc0cc5616) Using the latest skype 2.0 beta (2.0.0.13) on 64bit OS. It runs fine under fglrx. -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org
On Nov 19, 07 23:56:04 +0100, Aljaž Prusnik wrote:
One thing that isn't working, though, is testing video in skype. I'm not sure what the problem is. This is from syslog:
I assume skype relies on XVideo. We don't support that yet.
Nov 19 23:50:07 slash kernel: ioctl32(skype:5825): Unknown cmd fd(31) cmd(c0cc5616){t:'V';sz:204} arg(f417ae40) on /dev/video0 Nov 19 23:50:07 slash kernel: compat_ioctl32: v4l2 ioctl VIDIOC_S_PARM, dir=rw (0xc0cc5616)
I have *no* idea what it wants to do there, but I assume this is V4L
(video input).
Matthias
--
Matthias Hopf
Hi, > On Nov 19, 07 23:56:04 +0100, Aljaž Prusnik wrote: >> One thing that isn't working, though, is testing video in skype. I'm not >> sure what the problem is. This is from syslog: > > I assume skype relies on XVideo. We don't support that yet. I just hit the same problem. And you are right looking at http://www.skype.com/intl/en/download/skype/linux/beta/ states "Hardware requirements: - Video card driver with Xv support." Do you have any idea, when this will be implemented? Is there any way to work around this requirement? Thanks for your work so far. Felix Möller -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org
On Wed, Dec 12, 2007 at 01:04:18AM +0100, Felix M??ller wrote: > Hi, > > > On Nov 19, 07 23:56:04 +0100, Alja?? Prusnik wrote: > >> One thing that isn't working, though, is testing video in skype. I'm not > >> sure what the problem is. This is from syslog: > > > > I assume skype relies on XVideo. We don't support that yet. > I just hit the same problem. And you are right looking at > http://www.skype.com/intl/en/download/skype/linux/beta/ states > "Hardware requirements: > - Video card driver with Xv support." > > Do you have any idea, when this will be implemented? Is there any way to > work around this requirement? > > Thanks for your work so far. > > Felix M??ller The capabilities of the hardware video overlay of the avivo style hardware was severely reduced and can never do what you will expect it to do: * scaling * accept the most important FOURCCs directly. So do not wait up for this, we will need partly working 3d to make up for this. And TBH, if an application has a hard dependency on Xv, then that application could be considered rather clueless and broken. Luc Verhaegen. SUSE/Novell X Driver Developer. -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org
On Mon, Nov 19, 2007 at 11:46:42PM +0100, Alja?? Prusnik wrote:
Just a note on a successfull installation and the xorg log in case it adds value anywhere...
Compared to fglrx which I use - colors are a bit on the cold side and the speed is somewhat not there yet, but that is understandable. I'm not sure what this line means though:
TMDSBVoltageControl: unhandled chipset: 0x7142 (apart from not being recognized at an earlier stage).
If these drivers won't make system freeze occasionally as fglrx does, it's for keeps. :)
Should be fixed with the latest git. Thanks, Luc Verhaegen. SUSE/Novell X Driver Developer. -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org
On čet, 2007-11-29 at 01:54 +0100, Luc Verhaegen wrote:
TMDSBVoltageControl: unhandled chipset: 0x7142 (apart from not being recognized at an earlier stage).
If these drivers won't make system freeze occasionally as fglrx does, it's for keeps. :)
Should be fixed with the latest git.
It is indeed. Thanks. -- To unsubscribe, e-mail: radeonhd+unsubscribe@opensuse.org For additional commands, e-mail: radeonhd+help@opensuse.org
participants (6)
-
Aljaž Prusnik
-
Felix Möller
-
Hans Ulrich Niedermann
-
Luc Verhaegen
-
Matthias Hopf
-
Wu, Lisa