After an evaluation, GNOME has moved from Bugzilla to GitLab. Learn more about GitLab.
No new issues can be reported in GNOME Bugzilla anymore.
To report an issue in a GNOME project, go to GNOME GitLab.
Do not go to GNOME Gitlab for: Bluefish, Doxygen, GnuCash, GStreamer, java-gnome, LDTP, NetworkManager, Tomboy.
Bug 140994 - [mplex] configure failure
[mplex] configure failure
Status: RESOLVED FIXED
Product: GStreamer
Classification: Platform
Component: gst-plugins
0.8.1
Other Linux
: Normal normal
: 0.8.6
Assigned To: GStreamer Maintainers
GStreamer Maintainers
: 140969 151287 (view as bug list)
Depends on:
Blocks:
 
 
Reported: 2004-04-24 02:04 UTC by Pierre
Modified: 2004-12-22 21:47 UTC
See Also:
GNOME target: ---
GNOME version: ---


Attachments
patch fixing the configure check (544 bytes, patch)
2004-11-13 13:27 UTC, Arwed v. Merkatz
none Details | Review

Description Pierre 2004-04-24 02:04:05 UTC
[...]
checking for mjpegtools >= 1.6.1.93... yes
checking MPLEX_CFLAGS... -I/usr/local/include/mjpegtools
-I/usr/local/include/mjpegtools/mpeg2enc -I/usr/local/include/mjpegtools/mplex
checking MPLEX_LIBS... -L/usr/local/lib -lmjpegutils
checking interact.hpp usability... yes
checking interact.hpp presence... yes
checking for interact.hpp... yes
checking for valid mplex objects... no
configure: *** These plugins will not be built: mplex
[...]

In the config.log file :

[...]
configure:41475: checking for mjpegtools >= 1.6.1.93
configure:41479: result: yes
configure:41483: checking MPLEX_CFLAGS
configure:41486: result: -I/usr/local/include/mjpegtools
-I/usr/local/include/mjpegtools/mpeg2enc -I/usr/local/include/mjpegtools/mplex  
configure:41489: checking MPLEX_LIBS
configure:41492: result: -L/usr/local/lib -lmjpegutils  
configure:41531: checking interact.hpp usability
configure:41544: g++ -c -s -O3 -march=i686  -I/usr/local/include/mjpegtools
-I/usr/local/include/mjpegtools/mpeg2enc -I/usr/local/include/mjpegtools/mplex 
 conftest.cc >&5
configure:41547: $? = 0
configure:41550: test -s conftest.o
configure:41553: $? = 0
configure:41563: result: yes
configure:41567: checking interact.hpp presence
configure:41578: g++ -E  -I/usr/local/include/mjpegtools
-I/usr/local/include/mjpegtools/mpeg2enc -I/usr/local/include/mjpegtools/mplex 
 conftest.cc
configure:41584: $? = 0
configure:41603: result: yes
configure:41639: checking for interact.hpp
configure:41646: result: yes
configure:41655: checking for valid mplex objects
configure:41699: g++ -o conftest -s -O3 -march=i686 
-I/usr/local/include/mjpegtools -I/usr/local/include/mjpegtools/mpeg2enc
-I/usr/local/include/mjpegtools/mplex    conftest.cc  -L/usr/local/lib
-lmjpegutils   -lmplex2 -lm >&5
configure:41702: $? = 0
configure:41704: ./conftest
   INFO: [conftest] Found 0 audio streams and 0 video streams
   INFO: [conftest] Selecting generic MPEG1 output profile
   INFO: [conftest] Multiplexing video program stream!
./configure: line 1: 17938 Segmentation fault      ./conftest$ac_exeext
configure:41707: $? = 139
configure: program exited with status 139
configure: failed program was:
[...]

I have mjpegtools 1.6.2 installed.
Comment 1 Ronald Bultje 2004-04-25 01:18:27 UTC
Does mplex itself run?
Comment 2 Pierre 2004-04-25 03:11:29 UTC
Yes :

# mplex
mjpegtools mplex-2 version 1.6.2 (2.2.3)
Usage: mplex [params] -o <output filename pattern> <input file>...
         %d in the output file name is by segment count
[...]
Comment 3 Ronald Bultje 2004-04-25 04:05:56 UTC
Also if you try to mux an audio and video file into a system stream? ;).
Comment 4 Pierre 2004-04-25 04:58:58 UTC
# mplex test.mp2 /usr/local/share/xine/skins/xine_logo.mpv -o test.mpg
   INFO: [mplex] mplex version 1.6.2 (2.2.3 $Date: 2004/01/13 20:45:26 $)
   INFO: [mplex] File test.mp2 looks like an MPEG Audio stream.
   INFO: [mplex] File /usr/local/share/xine/skins/xine_logo.mpv looks like an
MPEG Video stream.
   INFO: [mplex] Found 1 audio streams and 1 video streams
   INFO: [mplex] Selecting generic MPEG1 output profile
   INFO: [mplex] Multiplexing video program stream!
[...]
Comment 5 Benjamin Otte (Company) 2004-04-25 11:59:50 UTC
*** Bug 140969 has been marked as a duplicate of this bug. ***
Comment 6 Pierre 2004-04-25 16:31:36 UTC
Are you sure it is the same problem ?
Comment 7 Ronald Bultje 2004-04-29 00:45:10 UTC
It seems like the same problem. Can you give some details on architecture, OS
etc.? Maybe glibc & kernel versions?
Comment 8 Pierre 2004-04-29 01:16:38 UTC
LFS on i686
kernel 2.6.5
glibc 2.2.5
Comment 9 Ronald Bultje 2004-04-29 01:30:43 UTC
Could you retry with glibc-2.3.x? I know there's a memory alignment bug in
several glibc-2.2.x versions that affected mjpegtools. If 2.3.x works, the bug
might not have been fully fixed after all, in which case I'd have to forward
this to the mjpegtools team (which includes me as well, but it's not GStreamer).
Comment 10 Pierre 2004-04-29 02:04:45 UTC
Ok I will try to update my glibc.
Comment 11 Ronald Bultje 2004-05-16 20:26:41 UTC
Oh, and something else - I noticed your bug report in mjpegtools and continued
on their mailinglist, can you provide a backtrace of the segfault?
Comment 12 Pierre 2004-06-01 06:49:37 UTC
How can I execute the conftest without the configure ?
Comment 13 Ronald Bultje 2004-06-01 14:17:45 UTC
Copy the code belonging to it from configure.ac (search for "mplex") to a
separate .cpp file, compile it using "g++ -o test test.cpp $(pkg-config --cflags
--libs mjpegtools)", then run it.
Comment 14 Pierre 2004-06-03 09:43:24 UTC
(gdb) run
Starting program: /tmp/conftest
   INFO: [conftest] Found 0 audio streams and 0 video streams
   INFO: [conftest] Selecting generic MPEG1 output profile
   INFO: [conftest] Multiplexing video program stream!
 
Program received signal SIGSEGV, Segmentation fault.
0x401da2fc in free () from /lib/libc.so.6
(gdb) bt
  • #0 free
    from /lib/libc.so.6
  • #1 operator delete(void*)
    from /usr/local/lib/libstdc++.so.6
  • #2 __gnu_cxx::new_allocator<ElementaryStream*>::deallocate(ElementaryStream**, unsigned)
    at new_allocator.h line 86
  • #3 std::_Vector_base<ElementaryStream*, std::allocator<ElementaryStream*> >::_M_deallocate(ElementaryStream**, unsigned)
    at stl_vector.h line 117
  • #4 ~_Vector_base
    at stl_vector.h line 106
  • #5 ~vector
    at stl_vector.h line 256
  • #6 ~Multiplexor
    at conftest.cpp line 24
  • #7 main
    at conftest.cpp line 25

Comment 15 Arwed v. Merkatz 2004-11-13 13:27:16 UTC
Created attachment 33729 [details] [review]
patch fixing the configure check

This patch fixes it for me, should I apply it Ronald?
Comment 16 Ronald Bultje 2004-11-13 14:13:34 UTC
Thanks for fixing, although I still don't understand it. ;).
Comment 17 Stephane Loeuillet 2004-11-14 01:07:42 UTC
*** Bug 151287 has been marked as a duplicate of this bug. ***
Comment 18 Stephane Loeuillet 2004-11-14 01:09:17 UTC
i had the same problem twice, for mplex and its twin (which test is close to
this one)

shouldn't the 2 case be fixed ? (i see only one affected by this patch)