http://bugs.winehq.org/show_bug.cgi?id=25148
Summary: Regression: GOM Player gets stuck at the start of a stream Product: Wine Version: unspecified Platform: x86 OS/Version: Linux Status: UNCONFIRMED Severity: normal Priority: P2 Component: -unknown AssignedTo: wine-bugs@winehq.org ReportedBy: florian@fkoeberle.de
Since the following commit the GOM Player gets stuck in a infinite short audio loop when it tries to show a stream from GomTV. Note that at the point this bug was introduced there is no visual output due to bug 25147. It was earlier possible that the audio/video got stuck (bug 25146) now pausing doesn't help anymore.
95897b6819ca7d0f2a7d687746990e186640b4e7 is the first bad commit commit 95897b6819ca7d0f2a7d687746990e186640b4e7 Author: Maarten Lankhorst m.b.lankhorst@gmail.com Date: Tue Nov 9 23:42:49 2010 +0100
strmbase: Add support for rendering algorithms to quality control.
:040000 040000 8efe76e82c494d901007a6f5eb3298bbe715cf45 f843b4586cce0e7f92609a8eede5ba7296bd0a57 M dlls :040000 040000 9d1926952c9ae6ffd96088ac262261293fea14e7 8d2e46216b6ac5f8150cc7000664590bae815b1b M include
http://bugs.winehq.org/show_bug.cgi?id=25148
Florian florian@fkoeberle.de changed:
What |Removed |Added ---------------------------------------------------------------------------- Keywords| |regression CC| |m.b.lankhorst@gmail.com
http://bugs.winehq.org/show_bug.cgi?id=25148
--- Comment #1 from Florian florian@fkoeberle.de 2010-11-13 07:36:07 CST --- Created an attachment (id=31896) --> (http://bugs.winehq.org/attachment.cgi?id=31896) Log with wine 1.3.7 (aborted with [ctrl] + c)
Since the commit 79f16188289eb84732313e36c0d88db234596f28 it spams fixmes with strange values. The same fixmes occur when using wine 1.3.7:
fixme:quartz:DSoundRender_GetWritePos Delta too big 90548/1, too far ahead fixme:quartz:DSoundRender_GetWritePos Advancing 2591922420 bytes
http://bugs.winehq.org/show_bug.cgi?id=25148
Dmitry Timoshkov dmitry@codeweavers.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Version|unspecified |1.3.7 Summary|Regression: GOM Player gets |GOM Player gets stuck at |stuck at the start of a |the start of a stream |stream |
http://bugs.winehq.org/show_bug.cgi?id=25148
Austin English austinenglish@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Component|-unknown |winegstreamer
http://bugs.winehq.org/show_bug.cgi?id=25148
--- Comment #2 from Florian florian@fkoeberle.de 2010-11-13 17:00:00 CST --- Created an attachment (id=31908) --> (http://bugs.winehq.org/attachment.cgi?id=31908) Log made with wine 1.3.7 + patch (attachment 31898)
I tried the patch Maarten Lankhorst posted as attachment 25146 but the behavior is still like described in this bug report.
http://bugs.winehq.org/show_bug.cgi?id=25148
--- Comment #3 from Florian florian@fkoeberle.de 2010-11-27 04:34:12 CST --- Created an attachment (id=32161) --> (http://bugs.winehq.org/attachment.cgi?id=32161) Log with wine 1.3.8 (aborted with [ctrl] + c)
With wine 1.3.8 GOM player still gets stuck at start even when the stream is fast enough. It also still prints massive amounts of strange fixmes, but they are a little different now:
fixme:quartz:DSoundRender_GetWritePos Delta too big -187297/1, too far ahead fixme:quartz:DSoundRender_GetWritePos Advancing 421540124 bytes
Notice the minus sign before the delta which isn't in the log made with wine 1.3.7 (attachment 31896) but in the log made with the patch (attachment 31898).
The log made with wine 1.3.8 also contains the same error like the log made with wine 1.3.7:
err:quartz:CheckCircularConnection Checking filtergraph returned 80040231, something's not right!
http://bugs.winehq.org/show_bug.cgi?id=25148
Maarten Lankhorst m.b.lankhorst@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|UNCONFIRMED |ASSIGNED AssignedTo|wine-bugs@winehq.org |m.b.lankhorst@gmail.com Ever Confirmed|0 |1
--- Comment #4 from Maarten Lankhorst m.b.lankhorst@gmail.com 2010-11-27 05:19:43 CST --- Created an attachment (id=32163) --> (http://bugs.winehq.org/attachment.cgi?id=32163) attempt at fixing sound
http://bugs.winehq.org/show_bug.cgi?id=25148
--- Comment #5 from Florian florian@fkoeberle.de 2010-11-27 14:35:07 CST --- I accidental attached the log when applying the sound patch to another bug as attachment 32182. The sound still stops playing after about 0.5 seconds.
http://bugs.winehq.org/show_bug.cgi?id=25148
--- Comment #6 from Maarten Lankhorst m.b.lankhorst@gmail.com 2010-11-27 16:36:53 CST --- Created an attachment (id=32192) --> (http://bugs.winehq.org/attachment.cgi?id=32192) Select sane sync source
Revert the filtergraph.c hacks and try this patch
http://bugs.winehq.org/show_bug.cgi?id=25148
Maarten Lankhorst m.b.lankhorst@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Attachment #32192|0 |1 is obsolete| |
--- Comment #7 from Maarten Lankhorst m.b.lankhorst@gmail.com 2010-11-27 16:39:45 CST --- Created an attachment (id=32193) --> (http://bugs.winehq.org/attachment.cgi?id=32193) woops, try 2
Last one was buggy xD
http://bugs.winehq.org/show_bug.cgi?id=25148
--- Comment #8 from Florian florian@fkoeberle.de 2010-11-28 01:39:39 CST --- Created an attachment (id=32199) --> (http://bugs.winehq.org/attachment.cgi?id=32199) log (+tid,+gstreamer,+quartz,+strmbase,+strmbase_qc) with wine 1.3.8 + filtergraph patch (attachment 32193)
It still gets stuck at the start of the stream: I hear no sound and the display area stays black. It are now however 3 lines which get spamed instead of two. This part looks interesting:
0026:trace:quartz:DSoundRender_UpdatePositions Moving from 172268 to 175664: clearing 3396 bytes 0026:fixme:quartz:DSoundRender_GetWritePos Delta too big -192513/1, too far ahead 0026:fixme:quartz:DSoundRender_GetWritePos Advancing 420620032 bytes 0026:trace:quartz:DSoundRender_UpdatePositions Moving from 175664 to 1056: clearing 1792 bytes
It first moves forward and shortly afterwards backwards.
http://bugs.winehq.org/show_bug.cgi?id=25148
--- Comment #9 from Florian florian@fkoeberle.de 2010-11-28 04:13:12 CST --- Created an attachment (id=32204) --> (http://bugs.winehq.org/attachment.cgi?id=32204) log (+tid,+gstreamer,+quartz,+strmbase,+strmbase_qc) with wine 1.3.8 + time trace
Log with time traces in DSoundRender_Run and ReferenceClock_GetTime
http://bugs.winehq.org/show_bug.cgi?id=25148
--- Comment #10 from Maarten Lankhorst m.b.lankhorst@gmail.com 2010-11-28 08:58:11 CST --- *** Bug 25147 has been marked as a duplicate of this bug. ***
http://bugs.winehq.org/show_bug.cgi?id=25148
--- Comment #11 from Maarten Lankhorst m.b.lankhorst@gmail.com 2010-11-29 02:27:19 CST --- Ok so here is my technical analysis of what is going wrong. This might be something you could pass on to someone knowledgeable in support, probably most persons won't have a clue what I'm talking about.
The root cause is all the same, I think wine is correct here, and that if they didn't force the directsound renderer to provide the reference clock it would fail on windows too.
The time as passed by the gom source filter and retried by GetTime is media time, not stream time. Since it is confusing the 2, it will cause the whole thing to hang. I assume that the only reason it works in windows is probably that the DirectSound renderer provides the clock and notices that its far behind and has to catch up. This will reset the clock and force a unhang.
I've done a lot of work in wine to separate the 2 times.
http://msdn.microsoft.com/en-us/library/dd374626%28v=VS.85%29.aspx explains the stream time. I suspect that '# After a seek operation, stream time resets to zero.' is causing the issue here. In the GOM filter SetTime is set to media time, not stream time. SetMediaTime sets the media time.
http://msdn.microsoft.com/en-us/library/dd758085%28VS.85%29.aspx is how it's done for a source filter.
Of course if my analysis is wrong, can you tell the responsible gretech guys to contact me directly? I'd love to figure this one out.
http://bugs.winehq.org/show_bug.cgi?id=25148
Henri Verbeet hverbeet@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Regression SHA1| |95897b6819ca7d0f2a7d6877469 | |90e186640b4e7
http://bugs.winehq.org/show_bug.cgi?id=25148
--- Comment #12 from Maarten Lankhorst m.b.lankhorst@gmail.com 2011-09-26 04:31:00 CDT --- I fear it happens because GOM tv is a live source, wine doesn't handle that case currently, and is missing support to get it working right. As a workaround, you could remove the blocking from videorenderer.c by changing hr = QualityControlRender_WaitFor(&This->qcimpl, pSample, This->blocked); to hr = S_OK;
and from dsoundrenderer.c by changing DSoundRender_GetWritePos
} else write_at = -1;
to } write_at = -1;
http://bugs.winehq.org/show_bug.cgi?id=25148
--- Comment #13 from Florian florian@fkoeberle.de 2011-10-03 05:16:00 CDT --- Confirmed bug in wine.1.3.29. With the changes described in comment 12 video and audio work, but aren't in sync. The visual output is about as twice as fast as the audio output. I tried also wine 1.3.6 again and it still works there without any problems.
http://bugs.winehq.org/show_bug.cgi?id=25148
Rafal Stanilewicz washuu@eastnews.com.pl changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |washuu@eastnews.com.pl
--- Comment #14 from Rafal Stanilewicz washuu@eastnews.com.pl 2012-01-26 06:25:30 CST --- Wine 1.3.37.
When I try to open working URL to see media stream, GOM player still crashes, but in my case perhaps it is something about missing DLLs?
err:ole:CoGetClassObject class {cd8743a1-3736-11d0-9e69-00c04fd7c15b} not registered err:ole:CoGetClassObject no class object {cd8743a1-3736-11d0-9e69-00c04fd7c15b} could be created for context 0x1 fixme:quartz:FilterGraphInner_QueryInterface unknown interface {fc4801a3-2ba9-11cf-a229-00aa003d7352} err:ole:CoGetClassObject class {cd8743a1-3736-11d0-9e69-00c04fd7c15b} not registered err:ole:CoGetClassObject no class object {cd8743a1-3736-11d0-9e69-00c04fd7c15b} could be created for context 0x1 fixme:quartz:FilterGraphInner_QueryInterface unknown interface {fc4801a3-2ba9-11cf-a229-00aa003d7352} err:ole:CoGetClassObject class {cd8743a1-3736-11d0-9e69-00c04fd7c15b} not registered err:ole:CoGetClassObject no class object {cd8743a1-3736-11d0-9e69-00c04fd7c15b} could be created for context 0x1 wine: Unhandled page fault on read access to 0x00000000 at address 0x3eab820 (thread 002b), starting debugger... Unhandled exception: page fault on read access to 0x00000000 in 32-bit code (0x03eab820).
http://bugs.winehq.org/show_bug.cgi?id=25148
Rafal Stanilewicz washuu@eastnews.com.pl changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |washuu@eastnews.com.pl
Aric Stewart aric@codeweavers.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |aric@codeweavers.com
--- Comment #14 from Rafal Stanilewicz washuu@eastnews.com.pl 2012-01-26 06:25:30 CST --- Wine 1.3.37.
When I try to open working URL to see media stream, GOM player still crashes, but in my case perhaps it is something about missing DLLs?
err:ole:CoGetClassObject class {cd8743a1-3736-11d0-9e69-00c04fd7c15b} not registered err:ole:CoGetClassObject no class object {cd8743a1-3736-11d0-9e69-00c04fd7c15b} could be created for context 0x1 fixme:quartz:FilterGraphInner_QueryInterface unknown interface {fc4801a3-2ba9-11cf-a229-00aa003d7352} err:ole:CoGetClassObject class {cd8743a1-3736-11d0-9e69-00c04fd7c15b} not registered err:ole:CoGetClassObject no class object {cd8743a1-3736-11d0-9e69-00c04fd7c15b} could be created for context 0x1 fixme:quartz:FilterGraphInner_QueryInterface unknown interface {fc4801a3-2ba9-11cf-a229-00aa003d7352} err:ole:CoGetClassObject class {cd8743a1-3736-11d0-9e69-00c04fd7c15b} not registered err:ole:CoGetClassObject no class object {cd8743a1-3736-11d0-9e69-00c04fd7c15b} could be created for context 0x1 wine: Unhandled page fault on read access to 0x00000000 at address 0x3eab820 (thread 002b), starting debugger... Unhandled exception: page fault on read access to 0x00000000 in 32-bit code (0x03eab820).
--- Comment #15 from Aric Stewart aric@codeweavers.com 2012-01-31 07:33:22 CST --- Could you retest this in 1.4 and get me a +quartz,+gstreamer,+strmbase,+tid log?
http://bugs.winehq.org/show_bug.cgi?id=25148
diego style_v@yahoo.com changed:
What |Removed |Added ---------------------------------------------------------------------------- CC| |style_v@yahoo.com
https://bugs.winehq.org/show_bug.cgi?id=25148
Austin English austinenglish@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|ASSIGNED |RESOLVED Resolution|--- |ABANDONED
--- Comment #16 from Austin English austinenglish@gmail.com --- No reply in over 2 years, abandoned.
https://bugs.winehq.org/show_bug.cgi?id=25148
Béla Gyebrószki gyebro69@gmail.com changed:
What |Removed |Added ---------------------------------------------------------------------------- Status|RESOLVED |CLOSED
--- Comment #17 from Béla Gyebrószki gyebro69@gmail.com --- Closing all abandoned/duplicated/invalid bugs.