[Ffmpeg-cvslog] r7674 - trunk/libavformat/mpegts.c[PATCH]

[Michel Bardiaux]

Michel Bardiaux wrote:
> Diego Biurrun wrote:
>> On Wed, Jan 24, 2007 at 02:17:34PM +0100, Michel Bardiaux wrote:
>>> Michael Niedermayer wrote:
>>>> On Wed, Jan 24, 2007 at 12:12:45PM +0100, Michel Bardiaux wrote:
>>>>> michael wrote:
>>>>>> Modified: trunk/libavformat/mpegts.c
>>>>>>
>>>>>> Log: "converts printf's to avlog's" patch by Daniel Kristjansson
>>>>>> / danielk mrl.nyu edu another 1 year old patch forgoten by mans 
>>>>>> trailing whitespace removial by me
>>>>>>
>>>>>> --- trunk/libavformat/mpegts.c    (original) +++
>>>>>> trunk/libavformat/mpegts.c    Tue Jan 23 22:41:30 2007 @@ -162,7
>>>>>> +162,7 @@ MpegTSSectionFilter *sec;
>>>>>>
>>>>>> #ifdef DEBUG_SI -    printf("Filter: pid=0x%x\n", pid); +
>>>>>> av_log(NULL, AV_LOG_DEBUG, "Filter: pid=0x%x\n", pid); #endif
>>>>> [snip]
>>>>>
>>>>> Why NULL? Could be ts->stream, or add an AVClass at the top of 
>>>>> MpegTSContext. (I dont propose a patch because it would depend on
>>>>> the choice made)
>>>> it should be AVFormatContext of course instead of NULL, patch welcome
>>>>
>>> Here it is. Compiles of course (with DEBUG_SI defined), and passes 
>>> regression (up to the point where it says "This test requires FFmpeg 
>>> to be compiled with --enable-gpl") but I dont know how much that 
>>> means, and I dont have MPEGTS files handy for a more targetted test.
>>
>> Recompile with --enable-gpl and rerun the regression tests.
>>
>> Diego
> 
> Just did, alas it does not go well:
> 
> Waiting for feeds to startup...
> Launch commandline: ../ffmpeg -loop_input -flags +bitexact -dct fastint 
> -idct simple -y -f pgmyuv -i vsynth1/%02d.pgm 
> http://127.0.0.1:9999/feed1.ffm
> ffserver started.
> New connection: GET /feed1.ffm
> 127.0.0.1 - - [Thu Jan 25 11:16:35 2007] "GET /feed1.ffm HTTP/1.1" 200 4165
> New connection: POST /feed1.ffm
> New connection: GET /test_l.avi?date=19700101T000000Z
> New connection: GET /test.swf?date=19700101T000000Z
> New connection: GET /test_h.avi?date=19700101T000000Z
> New connection: GET /test_h.asf?date=19700101T000000Z
> New connection: GET /test_l.asf?date=19700101T000000Z
> New connection: GET /test_h.rm?date=19700101T000000Z
> New connection: GET /test_l.rm?date=19700101T000000Z
> New connection: GET /test.jpg?date=19700101T000000Z
> New connection: GET /test_small.jpg?date=19700101T000000Z
> New connection: GET /test.mjpg?date=19700101T000000Z
> /besdev01/home/people/michel/internet/ffmpeg-svn-avlog/tests/server-regression.sh: 
> line 39:  3209 Segmentation fault      ../ffserver -d -f data/test.conf 
> 2>/dev/null
> /besdev01/home/people/michel/internet/ffmpeg-svn-avlog/tests/server-regression.sh: 
> line 40: kill: (3209) - No such process
> --- data/ffserver.regression    2007-01-25 11:16:38.000000000 +0100
> +++ 
> /besdev01/home/people/michel/internet/ffmpeg-svn-avlog/tests/ffserver.regression.ref 
>        2006-10-09 15:12:24.000000000 +0200
> @@ -1,10 +1,10 @@
> -fb2d22ac836e84ed8099615f9d1a8b31  ff-test_h.avi
> -6a4a11f3e8f3bad93ace41864efaa112  ff-test_l.avi
> -3d44dbc12b743720661b262d8b14f41e  ff-test.swf
> -435b70f495bfb69aa98bb7eaf3c636fa  ff-test_h.asf
> -bd89ebc6a511a03324ab18b5d7804a81  ff-test_l.asf
> -ad84891b0f4138675adde2819ec5dc00  ff-test_h.rm
> -3c98590122cb7c7decd0f3e59d9f9583  ff-test_l.rm
> -d41d8cd98f00b204e9800998ecf8427e  ff-test.jpg
> -d41d8cd98f00b204e9800998ecf8427e  ff-test_small.jpg
> -6ce0af834ee0abf4753df419f67648c5  ff-test.mjpg
> +418d0c843a6605a7d48f285804a69df4  ff-test_h.avi
> +ad4605187044d5e169f32a2ea205f937  ff-test_l.avi
> +34f4f0c160efcdb37c0999680e07c41c  ff-test.swf
> +fe6de6234a8bb4ace52b86763741cecd  ff-test_h.asf
> +68e7c8a8ff64c4016a8d4e5ea27353f3  ff-test_l.asf
> +cba8663681bf3ef89c95c09bb90204db  ff-test_h.rm
> +0a14086c708a7b8f4604e705a2bb7087  ff-test_l.rm
> +353a9a8fb3d5bb2334ac991b5c60f6d3  ff-test.jpg
> +7cbd46a6174d8203c26be333980595b5  ff-test_small.jpg
> +aaa459e638cd1dc2c120760fa4e1e384  ff-test.mjpg
> 
> Being wholly unfamiliar with ffserver, I dont have a clue what to do...
> 
> Greetings,

But the pristine svn 7707 fails too, exactly the same way. So I guess 
someone should commit the mpegts patch (which Michael has okayed anyway) 
and someone should check why the regression fails. If some nonexpert 
work like blindly checking past revs is needed, I'm willing to have a 
go, unless someone else is already on it?

-- 
Michel Bardiaux
R&D Director
T +32 [0] 2 790 29 41
F +32 [0] 2 790 29 02
E mailto:mbardiaux at mediaxim.be

Mediaxim NV/SA
Vorstlaan 191 Boulevard du Souverain
Brussel 1160 Bruxelles
http://www.mediaxim.com/