[Ffmpeg-devel] cabac.h compile error on MinGW

[Plotky, Jon]

I pulled ffmpeg from SVN yesterday and tried to compile.  Got the error
shown below.  My MinGW setup compiles an old CVS version of ffmpeg
correctly.

I did:
1) svn checkout svn://svn.mplayerhq.hu/ffmpeg/trunk ffmpeg
2) ./configure  --disable-static --enable-shared --enable-amr_nb
--enable-memalign-hack --enable-mingw32
3) make

GCC version: $ gcc --version
gcc.exe (GCC) 3.2.3 (mingw special 20030504-1)

The output of configure is down below.  The last lines of the make output
are:

gcc -I/c/ffmpeg/libswscale   -DHAVE_AV_CONFIG_H -D_FILE_OFFSET_BITS=64
-D_LARGEFILE_SOURCE -D_ISOC9X_SOURCE -I/c/ffmpeg -I/c/ffmpeg
-I/c/ffmpeg/libavutil -g -Wall -Wno-switch -Wdisabled-optimization
-Wpointer-arith -Wredundant-decls -Winline -O3  -c -o h261.o h261.c
gcc -I/c/ffmpeg/libswscale   -DHAVE_AV_CONFIG_H -D_FILE_OFFSET_BITS=64
-D_LARGEFILE_SOURCE -D_ISOC9X_SOURCE -I/c/ffmpeg -I/c/ffmpeg
-I/c/ffmpeg/libavutil -g -Wall -Wno-switch -Wdisabled-optimization
-Wpointer-arith -Wredundant-decls -Winline -O3  -c -o h264.o h264.c
h264.c: In function `hl_decode_mb':
h264.c:3661: warning: suggest parentheses around arithmetic in operand of ^
h264.c:3667: warning: suggest parentheses around arithmetic in operand of ^
h264.c: At top level:
h264.c:6037: warning: `used' attribute ignored
h264.c: In function `decode_cabac_residual':
h264.c:6143: warning: passing arg 4 of `decode_significance_8x8_x86'
discards qualifiers from pointer target type
h264.c:6070: warning: unused variable `last'
svq3.c: In function `decode_mb_cabac':
cabac.h:522: can't find a register in class `GENERAL_REGS' while reloading
`asm'
cabac.h:522: can't find a register in class `GENERAL_REGS' while reloading
`asm'
make[1]: *** [h264.o] Error 1
make[1]: Leaving directory `/c/advdev/videomail/ffmpegnew/ffmpeg/libavcodec'
make: *** [lib] Error 2


Output from configure:

$ ./configure  --disable-static --enable-shared --enable-amr_nb
--enable-memali
gn-hack --enable-mingw32
install prefix   C:\Program Files/FFmpeg
source path      /c/advdev/videomail/ffmpegnew/ffmpeg
C compiler       gcc
make             make
ARCH             x86_32 (generic)
big-endian       no
inttypes.h       yes
broken inttypes.h no
MMX enabled      yes
CMOV enabled     no
CMOV is fast     no
gprof enabled    no
zlib enabled     no
libgsm enabled   no
mp3lame enabled  no
libogg enabled   no
Vorbis enabled   no
FAAD enabled     no
faadbin enabled  no
FAAC enabled     no
XviD enabled     no
x264 enabled     no
a52 support      no
a52 dlopened     no
DTS support      no
pp support       no
Software Scaler enabled no
AVISynth enabled no
debug symbols    yes
strip symbols    yes
optimize         yes
static           no
shared           yes
video hooking    no
SDL support      no
Sun medialib support no
pthreads support no
AMR-NB float support yes
AMR-NB fixed support no
AMR-WB float support no
AMR-WB IF2 support no
network support      no
.align is power-of-two no
License: LGPL
Creating config.mak and config.h...

AMR NB FLOAT NOTICE ! Make sure you have downloaded TS26.104
REL-5 V5.1.0 from 
http://www.3gpp.org/ftp/Specs/archive/26_series/26.104/26104-510.zip
and extracted the source to libavcodec/amr_float
If you try this on alpha, you may need to change Word32 to int in
amr/typedef.h