When I am trying to record something using rawrec-0.9.991: /usr/local/bin/rawrec myfile.raw I get an error: rawrec: read syscall of dsp device failed: Invalid argument Same with cp: cp /dev/dsp myfile.raw cp: /dev/dsp: Invalid argument It used to work in 6.x. Yuri Fix: n/a How-To-Repeat: 1. Have EMU10K1 based card 2. cp /dev/dsp myfile.raw 3. observe an error
Responsible Changed From-To: freebsd-bugs->freebsd-multimedia reassign to multimedia team
Now recording works for me with this kernel: FreeBSD 7.0-PRERELEASE #1: Thu Jan 24 19:51:52 PST 2008 Not sure if the original issue was/is intermittent or it is really gone. Yuri
Hi. I have been experiencing this problem as well with emu10k1 since moving from 6.x to 7.x. I've been watching this PR and saw Yuri's latest message that an updated kernel allowed him to record again. I just updated my system to RELENG_7 as of yesterday however the issue still persists for me. I get the same errors when using rawrec/cp and Audacity does not record properly either. Audacity outputs lots of these to the terminal: PortAudio: read interrupted! PortAudio: read interrupted! PortAudio: read interrupted! Thanks in advance, -Mark FreeBSD amd645200.localhost 7.0-PRERELEASE FreeBSD 7.0-PRERELEASE #1: Fri Feb 15 16:29:55 CST 2008 mixx941@amd645200.localhost:/usr/obj/usr/src/sys/AMD645200 amd64 FreeBSD Audio Driver (newpcm: 64bit 2007061600/amd64) Installed devices: pcm0: <Creative Audigy 2 (EMU10K2)> at io 0xdc00 irq 16 [MPSAFE] (8p:1v/2r:1v channels duplex default)
Now I tried again and it doesn't work again for me. Kernel: 7.0-PRERELEASE FreeBSD 7.0-PRERELEASE #4: Tue Feb 19 08:18:11 PST 2008 Seems the problem is intermittent or dependent on the version of 7.0. Relevant dmesg messages: pcm0: <TriTech TR28602 AC97 Codec> pcm0: [ITHREAD] at the time of failure: pcm0:record:dsp0.r1: record interrupt timeout, channel dead Sound recording is a major function and emu10k1 is probably a very (if not most) common card. Is there any more debug messaging that could be turned on for this driver?
On Tue, Feb 19, 2008, at 20:20:03 GMT, Yuri wrote: > Now I tried again and it doesn't work again for me. > Kernel: 7.0-PRERELEASE FreeBSD 7.0-PRERELEASE #4: Tue Feb 19 > 08:18:11 PST 2008 > Seems the problem is intermittent or dependent on the version of 7.0. > > Relevant dmesg messages: > pcm0: <TriTech TR28602 AC97 Codec> > pcm0: [ITHREAD] > at the time of failure: > pcm0:record:dsp0.r1: record interrupt timeout, channel dead > > Sound recording is a major function and emu10k1 is probably a very > (if not most) common card. > > Is there any more debug messaging that could be turned on for this > driver? I was just writing another reply to the PR with some additional information (mainly the dmesg messages and a workaround) but I will answer in reply to you here instead. Yesterday I came across kern/118546 and tried the suggestion of setting "dev.pcm.0.rec.vchans" to 0 which appears to have fixed the recording problem for me here on emu10k1. When set to 1 (the default) it does not record properly. I tried to adapt the patch from snd_ich in kern/118546 and in the commit that fixed the problem for that driver to emu10k1 but I am not very experienced in C yet so I did not have much luck. So for now, "sysctl dev.pcm.0.rec.vchans=0" allows recording to work again for me on an Audigy 2 Platinum with emu10k1 but a proper fix would be great too. Thanks! -Mark
7.0 is not really supported anymore. Does this work in supported releases?
batch change: For bugs that match the following - Status Is In progress AND - Untouched since 2018-01-01. AND - Affects Base System OR Documentation DO: Reset to open status. Note: I did a quick pass but if you are getting this email it might be worthwhile to double check to see if this bug ought to be closed.
This is probably considered to be an ancient card now. I still have it somewhere. Should it even still be supported? Not sure if it's worth to retest it.
Closing as this is no longer relevant.