From Bugzilla Helper: User-Agent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.4) Gecko/20030703 Description of problem: If I abort a preview of a WAV file -- RIFF little-endian, PCM, if it matters -- mid-stream, and then try and preview any WAV file (whether the same file or not), the system locks up. CapsLock+ScrollLock flash, and no user input is accepted, including mouse, virtual terminal change key sequences, and SysRq. The system must be manually restarted. Sometimes (depending on the length of the WAV apparently) the WAV file will play endlessly in a loop until the system is reset by hand. I believe all the files I tried were 16-bit, some 22kHz and some 44kHz. Version-Release number of selected component (if applicable): nautilus-media-0.3.1-1 How reproducible: Always Steps to Reproduce: 1. Hover mouse pointer over a WAV file in Nautilus browser in order to hear preview. 2. Abort the preview mid-stream, by moving mouse off the file icon. 3. Hover mouse pointer over any WAV file, the same one or a different one. Actual Results: The system locks hard, requiring a manual reset or power cycle. Expected Results: The preview should have stopped, and the second preview should start normally without anomalies. Additional info: Fedora Core test2, have upgraded to nautilus 2.4.0-4 from rawhide. Other possibly related culprits' versions on my system: gstreamer-0.6.3-1 gstreamer-plugins-0.6.3-3 I am not sure that nautilus-media is the exact place to file this bug; I took my best guess.
Possibly related bug? http://bugzilla.gnome.org/show_bug.cgi?id=120202
If capslock and scrolllock are flashing, your kernel is oopsing.
Just an FYI, same behavior under 2.4.22-1.2082.nptl from rawhide.
Behavior the same under kernel-2.4.22-1.2087.nptl from rawhide. All rawhide updates in place as of this writing (except fedora-release). This is an Athlon T-bird 1.4GHz, KT133A (Iwill KK266+) mobo, CMI8738 (13f6/0111) sound, FWIW. Can I provide something more helpful to help chase this down? I am not a kernel hacker, but I am RHCE and willing to help if someone will shove me in the right direction with a map and flashlight.
Just a quick "breaks for me too" comment. Would love this fixed as it's pretty easy to trip oneself up with this kind of problem.
More information available now. This problem has persisted into Fedora Core 1 (kernel-2.4.22-1.2115.nptl), but appears to be hardware specific and perhaps should be moved to the kernel component. The problem does *NOT* occur on the following hardware that I had available for FC1 installs: - i815 based laptop (Gateway Solo 9500), ESS Maestro ES1988 sound - VIA KT400 (Soyo Dragon Ultra) desktop, VIA 8233/A/8235 AC97 sound Nevertheless, the KT133a/CMI8738 is common enough (and of an appropriate vintage) that this bug should be addressed.
Oops, it is in kernel. Sorry about that.
I think I've fixed this. It'll be in the forthcoming errata.
It is indeed. Thank you, sir! I'm very late on closing this bug, but certainly it's been fixed since at least before 2.4.22-1.2138.nptl, and probably much earlier.