[Bug 28048] New: winmm wave tests failing

http://bugs.winehq.org/show_bug.cgi?id=28048

Summary: winmm wave tests failing
           Product: Wine
           Version: 1.3.26
          Platform: x86
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: -unknown
        AssignedTo: [e-mail]
        ReportedBy: [e-mail]

On my Ubuntu 11.04 laptop, "make test" reliably fails in winmm
in the wave tests.  Judging by
http://test.winehq.org/data/tests/winmm:wave.html
there has usually been one or two machines similarly affected.

trace:mmdevapi:MMDevice_Activate (0x12d110)->(0x686ee9a0,1,(nil),0x141714)
trace:alsa:AUDDRV_GetAudioEndpoint "hw:0,0" 0x12d110 0 0x141714
warn:alsa:AUDDRV_GetAudioEndpoint Unable to open PCM "hw:0,0": -16 (Device or
resource busy)
...
wave.c:659: Test failed: waveOutOpen(1): format=22050x 8x1
flags=50000(CALLBACK_EVENT) rc=MMSYSERR_ERROR(Undefined external error.)

I'll attach a +alsa,+winmm,+mmdevapi,+module log.

[Bug 28048] winmm wave tests failing

http://bugs.winehq.org/show_bug.cgi?id=28048

--- Comment #1 from Dan Kegel  2011-08-10 15:55:20 CDT ---
Created an attachment (id=35915)
 --> (http://bugs.winehq.org/attachment.cgi?id=35915)
output of WINEDEBUG=+alsa,+winmm,+mmdevapi,+module make wave.ok

[Bug 28048] winmm wave tests failing

http://bugs.winehq.org/show_bug.cgi?id=28048

Andrew Eikum  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |[e-mail]

[Bug 28048] winmm wave tests failing

http://bugs.winehq.org/show_bug.cgi?id=28048

Sylvain Petreolle  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |[e-mail]

[Bug 28048] winmm wave tests failing

http://bugs.winehq.org/show_bug.cgi?id=28048

Jörg Höhle  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |[e-mail]
                   |                            |t

--- Comment #2 from Jörg Höhle  2011-08-18 03:02:45 CDT ---
Could somebody please analyse whether this has to do with device enumeration
and the fact that some (e.g. pulse) may hang onto the underlying HW device
(e.g. hw:0) for a few seconds prior to releasing it?

E.g. open "hw:0" succeeds (speaker-test -D hw:0 -c2 -r<match HW parameters>)
open "pulse" succeeds
open hw:0 immediately thereafter fails in non-blocking mode
wait a few seconds, then open hw:0 succeeds.

If yes, please change the subject line to reflect that.

[Bug 28048] winmm wave tests failing

http://bugs.winehq.org/show_bug.cgi?id=28048

Raymond  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |[e-mail]

--- Comment #3 from Raymond  2011-08-26 18:31:02 CDT ---
This is quite normal when your sound card does not support hardware mixing

The first device tested by wine is "default" ,

either "dmix" or "pulse" act as "audio engine" in mmdevapi term

so there may be other application still using the "audio engine" to access
"hw:0" and open "hw:0" can be regarded as exclusive mode if your sound card
does not support hardware mixing

The point is PA does not release the alsa device immediately ,
module-suspend-on-idle has a default timeout of 5 seconds

[Bug 28048] winmm wave tests failing (device in use)

http://bugs.winehq.org/show_bug.cgi?id=28048

Jörg Höhle  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
          Component|-unknown                    |mmdevapi
            Summary|winmm wave tests failing    |winmm wave tests failing
                   |                            |(device in use)

[Bug 28048] winmm wave tests failing (device in use)

http://bugs.winehq.org/show_bug.cgi?id=28048

Jörg Höhle  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |source

--- Comment #4 from Jörg Höhle  2012-11-09 18:14:45 CST ---
Are my error mapping patches in 1.5.17 enough to silence this?
If the cause is the same, shall we make it a duplicate of bug #28109?

[Bug 28048] winmm wave tests failing (device in use)

https://bugs.winehq.org/show_bug.cgi?id=28048

Austin English  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|---                         |ABANDONED

--- Comment #5 from Austin English  ---
(In reply to Jörg Höhle from comment #4)
> Are my error mapping patches in 1.5.17 enough to silence this?
> If the cause is the same, shall we make it a duplicate of bug #28109?

Abandoned.

[Bug 28048] winmm wave tests failing (device in use)

https://bugs.winehq.org/show_bug.cgi?id=28048

Austin English  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|RESOLVED                    |CLOSED

--- Comment #6 from Austin English  ---
Closing.

[Bug 28048] winmm wave tests failing (device in use)

https://bugs.winehq.org/show_bug.cgi?id=28048

Sylvain Petreolle  changed:

What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|[e-mail]         |