[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Monitors requiring interlaced video modes with older Nvidia video cards are out of options in wheezy



On 2011-04-24 22:19 +0200, Stephen Powell wrote:

> Actually, now that I think about it, this bug report has probably been
> sitting on the wrong queue all this time.  I reported the bug against
> package xserver-xorg-video-nouveau, which is a user-space X driver.  But
> the driver uses kernel mode setting (KMS) to set the actual video mode.
> The same bug can be reproduced by using the following kernel boot parameter:
>
>    video=VGA-1:1024x768@87i
>
> Except that in this case the blank screen is produced on the frame-buffer
> text consoles (vt1-vt6) instead of the X server console (vt7).  This tells
> me that the bug is on the kernel-space side rather than the user-space
> side.  If I were the maintainer, that would probably have been the first
> thing I would have tried to determine.  The bug could be in the nouveau
> kernel module, or perhaps in more-generic KMS code.  To determine if the
> problem is nouveau-specific, I would need a non-Nvidia video card whose
> driver also uses KMS.

Please add these findings to the bug report.

> But that brings up another issue, which is reporting bugs upstream.
> The official Debian policy is "Don't report bugs upstream".  I quote from
> http://www.debian.org/Bugs/Reporting:
>
>    Don't file bugs upstream
>
>    If you file a bug in Debian, don't send a copy to the upstream software
>    maintainers yourself, as it is possible that the bug exists only in Debian.
>    If necessary, the maintainer of the package will forward the bug upstream.

This opinion is certainly not universally shared among Debian
developers.  Maintainers of popular packages are usually swamped with
bug reports, and if the bug is hardware specific (like the majority of
kernel and video driver bugs are) playing proxy between the user and
upstream is a very inefficient work flow -- better ask the user to
report the bug upstream himself and answer any questions that upstream
has directly.

> But if the bug is clearly in upstream code, and the maintainer is ignoring
> the bug report, what else can one do?

Like Boyd and Ron recommended, report the bug upstream and mark it as
forwarded.  Bonus points for you if you can first try kernel 2.6.39-rc4
which has just been uploaded to experimental.

Sven


Reply to: