Bug 29709
| Summary: | last sound driver MFC introduced problems with various applications | ||
|---|---|---|---|
| Product: | Base System | Reporter: | Andre Albsmeier <Andre.Albsmeier> |
| Component: | kern | Assignee: | sound |
| Status: | Closed FIXED | ||
| Severity: | Affects Only Me | ||
| Priority: | Normal | ||
| Version: | 4.4-PRERELEASE | ||
| Hardware: | Any | ||
| OS: | Any | ||
Responsible Changed From-To: freebsd-bugs->sound Sound issue >pcm0: <SB16 DSP 4.16> at io 0x220 irq 5 drq 1:5 (1p/1r/0v channels duplex) > - mtv (linux) audio doesn't work, the console says: > pcm0: play interrupt timeout, channel dead > - in mxaudio the progress slider runs like hell; a 4 minute > song is processed in 4 seconds. No audio appears.. these are both consistent with the channel not starting. does 'systat -vm' indicate interrupts happening on irq5 while playing? > - sox produces a short blip as though the sample is playing too fast is the length of the blip proportional to the length of the sample? -cg >> - sox produces a short blip as though the sample is playing too fast
>
>is the length of the blip proportional to the length of the sample?
Yes, roughly.
FYI, xmcd works fine, too.
On Tue, 14-Aug-2001 at 17:29:22 +0100, Cameron Grant wrote: > >pcm0: <SB16 DSP 4.16> at io 0x220 irq 5 drq 1:5 (1p/1r/0v channels duplex) > > - mtv (linux) audio doesn't work, the console says: > > pcm0: play interrupt timeout, channel dead > > - in mxaudio the progress slider runs like hell; a 4 minute > > song is processed in 4 seconds. No audio appears.. > > these are both consistent with the channel not starting. does 'systat -vm' > indicate interrupts happening on irq5 while playing? No. The 'sbc0 irq5' output of 'systat -vm' didn't even appear when I first started mtv. Then I launched mpg123 and sound could be heard and I got about 86 irqs. Then I started mtv and fired systat -vm again. Now the 'sbc0 irq5' output was available but no irqs were shown: 8 users Load 0.16 0.20 0.15 Tue Aug 14 19:47 Mem:KB REAL VIRTUAL VN PAGER SWAP PAGER Tot Share Tot Share Free in out in out Act 25072 3548 49048 5308 75172 count All 450324 13444 2937476 26224 pages zfod Interrupts Proc:r p d s w Csw Trp Sys Int Sof Flt cow 269 total 3 1 21 522 200 1785 269 216 1 80204 wire intpm0 irq 62944 act ahc0 irq14 1.7%Sys 0.0%Intr 21.8%User 0.0%Nice 76.4%Idl 294316 inact 8 mux irq15 | | | | | | | | | | 12860 cache ahc1 irq10 =>>>>>>>>>>> 62312 free 7 atkbd0 irq daefr 26 psm0 irq12 Namei Name-cache Dir-cache prcfr fdc0 irq6 Calls hits % hits % react ppc0 irq7 pdwak sbc0 irq5 pdpgs 100 clk irq0 Disks ccd0 da0 da1 da2 da3 da4 da10 intrn 128 rtc irq8 KB/t 0.00 0.00 8.00 0.00 0.00 0.00 0.00 62048 buf tps 0 0 0 0 0 0 0 4 dirtybuf MB/s 0.00 0.00 0.00 0.00 0.00 0.00 0.00 32315 desiredvnodes % busy 0 0 0 0 0 0 0 13095 numvnodes 26 freevnodes Here is the contents of the pop up windows that appears shortly after starting mtv: Warning! Cannot play audio: Invalid argument (muting) With GNOME, use "esddsp mtv". Use option -aw to debug. I tried the -aw option. Here the the part that seems interesting: (I can send the whole log file which is about 36 KB...) [4442,1026] mtvp: audio: audiolib.c:2657: t=997811378.6596: looking for next audio syncword [4442,1026] mtvp: audio: audiolib.c:2664: t=997811378.6596: found audio syncword [4442,1026] mtvp: audio: audioioctl.c:1106: t=997811378.6597: writing 4608 bytes to device [4442,1026] mtvp: audio: audioioctl.c:1138: t=997811378.6597: 4608 bytes written [4442,1026] mtvp: audio: audiolib.c:2657: t=997811378.6602: looking for next audio syncword [4442,1026] mtvp: audio: audiolib.c:2664: t=997811378.6604: found audio syncword [4442,1026] mtvp: audio: audioioctl.c:1106: t=997811378.6604: writing 4608 bytes to device [4442,1026] mtvp: audio: audioioctl.c:1138: t=997811378.6604: 4608 bytes written [4442,1026] mtvp: audio: audiolib.c:2657: t=997811378.6610: looking for next audio syncword [4442,1026] mtvp: audio: audiolib.c:2664: t=997811378.6610: found audio syncword [4442,1026] mtvp: audio: audioioctl.c:1106: t=997811378.6610: writing 4608 bytes to device [4442,1026] mtvp: audio: audioioctl.c:1138: t=997811378.6610: 4608 bytes written [4442,1026] mtvp: audio: audiolib.c:2657: t=997811378.6615: looking for next audio syncword [4442,1026] mtvp: audio: audiolib.c:2664: t=997811378.6616: found audio syncword [4442,1026] mtvp: audio: audioioctl.c:1106: t=997811378.6616: writing 4608 bytes to device [4442,1026] mtvp: audio: audioioctl.c:1138: t=997811378.6616: 4608 bytes written [4442,1026] mtvp: audio: audiolib.c:2657: t=997811378.6621: looking for next audio syncword [4442,1026] mtvp: audio: audiolib.c:2664: t=997811378.6621: found audio syncword [4442,1026] mtvp: audio: audioioctl.c:1106: t=997811378.6622: writing 4608 bytes to device [4442,1026] mtvp: audio: audioioctl.c:1138: t=997811378.6622: 4608 bytes written [4442,1026] mtvp: audio: audiolib.c:2657: t=997811378.6627: looking for next audio syncword [4442,1026] mtvp: audio: audiolib.c:2664: t=997811378.6627: found audio syncword [4442,1026] mtvp: audio: audioioctl.c:1106: t=997811378.6627: writing 4608 bytes to device [4442,1026] mtvp: audio: audioioctl.c:1138: t=997811379.6529: 1536 bytes written [4442,1026] mtvp: audio: audioioctl.c:1106: t=997811379.6530: writing 3072 bytes to device [4442,1026] mtvp: audio: audioioctl.c:1134: t=997811379.6530: error: write to device failed: Invalid argument (status=22) [4442,1026] mtvp: audio: audiolib.c:6077: t=997811379.6531: error: couldn't write 4608 bytes to devi ce: Invalid argument (status=22) [4442,1026] mtvp: audio: audiolib.c:6532: t=997811379.6531: error: in mal_decode_something(): mal_de code_frame() returned -1: Invalid argument (status=22) [4442,1026] mtvp: audio: audio_fork.c:971: t=997811379.6532: error: cannot play audio: Invalid argum ent (status=22) [4442,1026] mtvp: audio: audio_fork.c:105: t=997811379.6532: sending msg: type: 84, sz: 4, val: 22 [4442,1026] mtvp: audio: audio_fork.c:254: t=997811379.6532: flushing all pending commands until A2S _COM_RESTART [4442,1026] mtvp: audio: audiolib.c:7039: t=997811379.6532: flushing buffer data [4442,1026] mtvp: audio: audiolib.c:2317: t=997811379.6533: flushing pts fifo [4442,1026] mtvp: audio: audiolib.c:2718: t=997811379.6533: closing audio device [4442,1026] mtvp: audio: audioioctl.c:267: t=997811379.6533: closing audio device... [4442,1026] mtvp: audio: audioioctl.c:253: t=997811379.6533: ioctl(24, SNDCTL_DSP_RESET, ...) = 0 [4442,1026] mtvp: audio: audioioctl.c:342: t=997811380.6555: audio device closed [4442,1026] mtvp: audio: audiolib.c:2722: t=997811380.6555: audio device closed > > > - sox produces a short blip as though the sample is playing too fast > > is the length of the blip proportional to the length of the sample? Yes, I tried an mp3 file with a duration of 30 minutes and 45 seconds with mxaudio. The progess indicator went from start to end in about 30 seconds. Seems to be factor 60 ?!? -Andre please test this patch.
-cg
Index: channel.c
===================================================================
RCS file: /home/ncvs/src/sys/dev/sound/pcm/channel.c,v
retrieving revision 1.19.2.13
diff -u -b -r1.19.2.13 channel.c
--- channel.c 2001/08/01 03:41:03 1.19.2.13
+++ channel.c 2001/08/14 17:55:32
@@ -24,7 +24,7 @@
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*
- * $FreeBSD: src/sys/dev/sound/pcm/channel.c,v 1.19.2.13 2001/08/01
03:41:03 cg Exp $
+ * $FreeBSD: src/sys/dev//sound/pcm/channel.c,v 1.19.2.13 2001/08/01
03:41:03 cg Exp $
*/
#include <dev/sound/pcm/sound.h>
@@ -422,7 +422,7 @@
CHN_LOCKASSERT(c);
/* if we're running, or if we're prevented from triggering, bail */
- if ((c->flags & CHN_F_TRIGGERED) || (c->flags & CHN_F_NOTRIGGER))
+ if ((c->flags & CHN_F_TRIGGERED) || ((c->flags & CHN_F_NOTRIGGER) &&
!force))
return EINVAL;
i = (c->direction == PCMDIR_PLAY)? sndbuf_getready(bs) :
sndbuf_getfree(bs);
Index: dsp.c
===================================================================
RCS file: /home/ncvs/src/sys/dev/sound/pcm/dsp.c,v
retrieving revision 1.15.2.8
diff -u -b -r1.15.2.8 dsp.c
--- dsp.c 2001/08/01 03:41:03 1.15.2.8
+++ dsp.c 2001/08/14 18:35:56
@@ -23,7 +23,7 @@
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*
- * $FreeBSD: src/sys/dev/sound/pcm/dsp.c,v 1.15.2.8 2001/08/01 03:41:03 cg
Exp $
+ * $FreeBSD: src/sys/dev//sound/pcm/dsp.c,v 1.15.2.8 2001/08/01 03:41:03 cg
Exp $
*/
#include <sys/param.h>
@@ -854,20 +854,18 @@
if (rdch) {
CHN_LOCK(rdch);
rdch->flags &= ~(CHN_F_TRIGGERED | CHN_F_NOTRIGGER);
- if (*arg_i & PCM_ENABLE_INPUT) {
- rdch->flags |= CHN_F_TRIGGERED;
+ if (*arg_i & PCM_ENABLE_INPUT)
chn_start(rdch, 1);
- } else
+ else
rdch->flags |= CHN_F_NOTRIGGER;
CHN_UNLOCK(rdch);
}
if (wrch) {
CHN_LOCK(wrch);
wrch->flags &= ~(CHN_F_TRIGGERED | CHN_F_NOTRIGGER);
- if (*arg_i & PCM_ENABLE_OUTPUT) {
- wrch->flags |= CHN_F_TRIGGERED;
+ if (*arg_i & PCM_ENABLE_OUTPUT)
chn_start(wrch, 1);
- } else
+ else
wrch->flags |= CHN_F_NOTRIGGER;
CHN_UNLOCK(wrch);
}
On Tue, 14-Aug-2001 at 20:49:22 +0100, Cameron Grant wrote:
> please test this patch.
>
> -cg
>
I had to apply it by hand due to some line wrap problems
but it seems I did it right since it fixes the problem :-)
At least it works here on the <SB16 DSP 4.16> with mtv
and mxaudio. Maybe others can try theur stuff as well..
Thanks for the quick help,
-Andre
On Tue, Aug 14, 2001 at 10:36:09PM +0200, Andre Albsmeier wrote:
> The PR (29709) contains a patch from Cameron which fixes the problem
> here (SB16 with mxaudio and mtv)!
>
> Maybe you want to try it out as well...
I can't get to the original machine until later, but I build the
PRERELEASE on my laptop and it exhibits the same problems.
arvon:~$ cat /dev/sndstat
FreeBSD Audio Driver (newpcm) Aug 14 2001 15:55:37
Installed devices:
pcm0: <Yamaha DS-1E (YMF744)> at memory 0xfedf8000 irq 9 (4p/2r/0v channels duplex)
I applied the patch, and mtv works. Once. A second invocation can't
open /dev/dsp. aviplay has similar problems, although it may run more
than once, eventually the sound device gets jammed up. (mtv may run
more than once, too, but the time I tried it failed on the second
attempt).
play from the sox package still fails, as does catting the file,
although the failure mode is a little different: a pop then the sound
played too fast and then a pop.
I'll try this all again on my soundblaster at home, but I don't think
this should close yet.
This PR can be closed. At least regarding my problem :-) Thanks, -Andre State Changed From-To: open->closed Closed at originator's request. |
Since the pcm MFC on August 1st various applications have problems. Here is a summary of the observations taken from -multimedia: pcm0: <SB16 DSP 4.16> at io 0x220 irq 5 drq 1:5 (1p/1r/0v channels duplex) - mtv (linux) audio doesn't work, the console says: pcm0: play interrupt timeout, channel dead - in mxaudio the progress slider runs like hell; a 4 minute song is processed in 4 seconds. No audio appears.. - sox produces a short blip as though the sample is playing too fast - mpg123 works! pcm0: <Aureal Vortex 8820> at memory 0xf4000000 irq 10 (4p/1r/0v channels duplex) - same with mtv I assume the list isn't complete; it just lists the problems seen until now. Fix: I replaced /sys/dev/sound completely with the version before august 1st and created empty pci/ich.c, pcm/vchan.c, pcm/sndstat.c and pcm/feeder_rate.c files so I didn't have to mess around with sys/conf/files :-) How-To-Repeat: Update to a recent 4.4-PRERELEASE and try some of the above combinations.