xserver-i810 killed by xrandr
Hello, this is my first attempt to fix a bug in a big programm like the x system is. Currently I'm feeling a little bit overextended by the size of the source code and the dependencies of different packets. I hope you can give me some hints, where to begin my search. 'xrandr -s 1' kills my xserver. I wrote a bug report (371144), providing further information about the problem. I wanted to localize what happens. Today I downgraded to xserver-xorg-video-i810=1:1.4.1.3-3, everything is working fine with this version. Upgrading to 1:1.5.1.0-1 or 1:1.5.1.0-2 brings the problem back. Diffing the sources it gives me about 33000 lines of difference: | $ diff -r xserver-xorg-video-i810-1.4.1.3/ xserver-xorg-video-i810-1.5.1.0/ | wc -l | 33664 I wanted to localize the part of the driver, that causes the crash. Is there any possibility to compile and install the packet exactely the same way like the original debian packet is but including debug symbols and run the xserver manually in a debugger? Is there a better way to find the relevant part? How should I begin my work? Thanks for comments! -daniel- -- http://www.blackamp.de GPG-Key: 0x53BE81EC Fingerprint: A854 4C0A FF1A 09DF 5433 0EF8 1CF0 0213 53BE 81EC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Patch cplay to use mplayer with speed-control
Hi everyone, I tried patching cplay to use mplayer(>=1.0pre7) which can also do speed-control on audio-files. My patch creates and uses a fifo to control mplayer and adds some keybindings to cplay (decr: [, incr: ], reset: =). I tried to write the patch as clean and stable as possible and provide my first trial on my webpage. I would be glad to receive some reviews and comments how to improve my code quality and some ideas what to do with my patch: - Should I mail it to the author of cplay? - Should I mail it to the maintainer of the debian package? - Should I fill a bugreport "feature-request" providing my patch? You can find further information, the patch and a patched version of cplay on my webpage: http://www.argafal.de/cgi-bin/index.pl?page=projcplay Thank you! -daniel- -- http://www.blackamp.de GPG-Key: 0x53BE81EC Fingerprint: A854 4C0A FF1A 09DF 5433 0EF8 1CF0 0213 53BE 81EC -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Patch cplay to use mplayer with speed-control
On 2006-06-23, Kevin B. McCarty <[EMAIL PROTECTED]> wrote:
> If it is optional (that is, if you improve your patch to test for an mplayer
> binary at runtime), you could certainly consider submitting it to the Debian
> maintainer or to upstream.
I worked on my patch. Now it only extents cplay of the possibility to use
mplayer, but all players of the original version still work. So there is no
_need_ of mplayer, it also works without.
> - It would be more readable if you use unified diff format ("diff -u")
> to show some lines of context.
> - The code where you create the fifo makes it with a predictable
> filename; also, it assumes that if it (or another file of the same name)
> already exists, it may be deleted by the current cplay user. Hence it
> is susceptible to a very trivial DOS attack. You ought to fix this
> before submitting the patch.
Thank you, I did like you suggested. The code for the fifo from my first patch
was copied from the control-fifo of the original cplay-1.49. I fixed both in
different ways according to the ideas of the fifo use. I also added equalizer
support and tried fixing
http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=375060.
I sent my patch to Ulf, the author of cplay and will wait for his reactions.
The cplay-dev mailinglist seems dead, but I will announce it there, too.
Thanks for further comments, see
http://www.argafal.de/cgi-bin/index.pl?page=projcplay for the latest version.
Best regards,
-daniel-
--
http://www.blackamp.de
GPG-Key: 0x53BE81EC
Fingerprint: A854 4C0A FF1A 09DF 5433 0EF8 1CF0 0213 53BE 81EC
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
