Recherche avancée

Sur d’autres sites (11037)

• Including objects to a shared library from a C++ archive (.a)

  1er septembre 2021, par El Sampsa

  I am trying to include some object files into a shared library I am building. Take the following command (things in [ETC] have been omitted for brevity) :

  



  


  /usr/bin/c++ -fPIC -std=c++14 -pthread -Iinclude/ext/liveMedia -Iinclude/ext/groupsock [ETC] -g -shared -Wl,-soname,libValkka.so -o lib/libValkka.so CMakeFiles/Valkka.dir/src/avthread.cpp.o CMakeFiles/Valkka.dir/src/opengl.cpp.o [ETC] CMakeFiles/Valkka.dir/src/decoders.cpp.o -lX11 -lGLEW -lGLU -lGL -Wl,—whole-archive lib/libavcodec.a -Wl,—no-whole-archive

  


  



  So basically I am just creating a shared library where most of the objects come from my own source code (i.e. CMakeFiles/Valkka.dir/src/*.o), but some of them come from an external static library, located at "lib/libavcodec.a". I get the following error :

  



  


  /usr/bin/ld : lib/libavcodec.a(h264_cabac.o) : relocation R_X86_64_PC32 against symbol 'ff_h264_cabac_tables' can not be used when making a shared object ; recompile with -fPIC
 /usr/bin/ld : final link failed : Bad value
 collect2 : error : ld returned 1 exit status

  


  



  But that is so untrue ! I can extract "libavcodec.a" with

  



  ar x libavcodec.a


  



  And after that check that

  



  readelf --relocs h264_cabac.o | egrep '(GOT|PLT|JU?MP_SLOT)' 


  



  does give some **it :

  



  


  00000000175d 003100000004 R_X86_64_PLT32 0000000000000000 __stack_chk_fail - 4
 000000001926 003100000004 R_X86_64_PLT32 0000000000000000 __stack_chk_fail - 4

  
 


  ...

  


  



  As does

  



  objdump -r h264_cabac.o | grep -i "relocation"


  



  So, indeed, the object files in "libavcodec.a" have been compiled to get PIC (position independent code).

  



  Why does the linker believe otherwise !?

  



  Related links :

  



  How to include all objects of an archive in a shared object ?

  



  Linking archives (.a) into shared object (.so)

  



  Is there a way to determine that a .a or .so library has been compiled as position indepenent code ?

  



  How can I tell, with something like objdump, if an object file has been built with -fPIC ?

  


• FFMPEG MKV -> MP4 Batch Conversion

  15 juillet 2024, par blaziken386

  I'm trying to write a program that lets me convert a series of .mkv files with subtitle files into .mp4 files with the subs hardcoded.

  


  Right now, the script I use is

  


  ffmpeg -i input.mkv -vf subtitles=input.mkv output.mp4



  


  This is fine, but it means I can only convert them one at a time, and it's kind of a hassle because it means I have to fiddle with it every few minutes to set up the next one.

  


  I have another script I use for converting .flac files to .mp3 files, which is

  


  @ECHO OFF

FOR %%f IN (*.flac) DO (
echo Converting: %%f
ffmpeg -i "%%f" -ab 320k -map_metadata 0 "%%~nf.mp3"
)

echo Finished

PAUSE



  


  Running that converts every single .flac folder into an .mp3 equivalent, with the same filename and everything.

  


  I've tried to combine the above scripts into something like this :

  


  @ECHO OFF

FOR %%f IN (*.mkv) DO (
echo Converting: %%f
ffmpeg -i "%%f" -vf subtitles=%%f "%%~nf.mp4"
)

echo Finished

PAUSE


  


  but every time I do so, it returns errors like "invalid argument" or "unable to find a suitable output type", or "error initializing filters", or "2 frames left in the queue on closing" or something along those lines. I've swapped out subtitles=%%f for "subtitles-%%f" or subtitles="%%f.mkv" and so on and so forth, and none of those give me what I want either. Sometimes it creates Empty .mp4 file containers with nothing in them, sometimes it does nothing at all.

  


  I don't really understand what exactly is happening under the hood in that flac->mp3 code, because I grabbed it from a different stackoverflow post years ago. All I know is that trying to copy that code and repurpose it into something else doesn't work. Is this just an issue where I've fucked up the formatting of the code and not realized it, or is this a "ffmpeg can't actually do that because of a weird technical issue" thing ?

  


  I also tried the code listed here, when Stackoverflow listed that as a possible duplicate, but that gave me similar errors, and I don't really understand why !

  


  Also, if it's relevant, I'm running windows.

  


• Best way to take a live-video-stream and split the image into thirds to broadcast out ?

  3 avril 2020, par user1258530

  Can anyone tell me the best tools to make this work ? I basically need to pass a video-feed from an Arduino device, stream it into the cloud, and then have the cloud-video-server "slice" that video feed into thirds (top third, middle third, and bottom-third), and simultaneously stream each of those third-video-slices out to the world via m3u8 (HLS) endpoint.

  



  I know I can pass video up to the cloud via ffmpeg or via RTMP - - - but I haven't found the method yet for the cloud-video-server-based "slicing" process. I haven't found an ffmpeg way to do it, and I haven't found a Wowza server way to do it, nor an OpenCV way to do it. Lots of examples of cropping, but I haven't found examples of cropping multiple segments at the same time and streaming out simultaneously.

  



  Any advice is greatly appreciated.