[FFmpeg-user] new configure issue re libopencore

[Jim Worrall]

On Oct 27, 2012, at 2:19 PM, Jim Worrall <coniophora at gmail.com> wrote:

> 
> On Oct 27, 2012, at 2:02 PM, Jim Worrall <coniophora at gmail.com> wrote:
> 
>> 
>> On Oct 27, 2012, at 1:23 PM, Reindl Harald <h.reindl at thelounge.net> wrote:
>> 
>>> --enable-nonfree --enable-gpl --enable-version3
>> 
>> Thank you, that was very helpful.  
>> But now I run into another error while configuring ffmpeg:
>> 
>> ERROR: libvpx encoder version must be >=0.9.7
> 
> I traced that back to an error with the make command on libvpx:
> 
> Jims-MacBook-Pro:libvpx-v1.1.0 jim$ make -j 4
>    [DEP] vpx_config.c.d
>    [CC] vpx_config.c.o
>    [AS] vp8/encoder/x86/quantize_sse2.asm.o
>    [AS] vp8/encoder/x86/temporal_filter_apply_sse2.asm.o
>    [AS] vp8/encoder/x86/quantize_ssse3.asm.o
> yasm: FATAL: yasm: FATAL: unable to open include file `asm_enc_offsets.asm'unable to open include file `asm_enc_offsets.asm'
> 

In the continuing saga, I seem to have worked past the libvpx problem.  
Now I'm back to ffmpeg itself not compiling, but in a new and different way.  
This is MacOS 10.8.2, using Xcode 4.5.1 and the corresponding command-line tools.
Here are the beginning and end of config.log, trying a very simple configuration:

# ./configure --prefix=/Volumes/Ramdisk/sw --enable-gpl --enable-version3 --cc=clang
ALL_COMPONENTS='
    aac_adtstoasc_bsf
chomp_bsf
dump_extradata_bsf
h264_mp4toannexb_bsf
. . . .
check_ld cc
check_cc
BEGIN /var/folders/sy/2d072p1s4sdc95lpnwp6s3jh0000gn/T//ffconf.M9D61Tgp.c
    1	int main(void){ return 0; }
END /var/folders/sy/2d072p1s4sdc95lpnwp6s3jh0000gn/T//ffconf.M9D61Tgp.c
clang -I/Volumes/Ramdisk/sw/include 
 -c -o /var/folders/sy/2d072p1s4sdc95lpnwp6s3jh0000gn/T//ffconf.CcIJ2L0K.o /var/folders/sy/2d072p1s4sdc95lpnwp6s3jh0000gn/T//ffconf.M9D61Tgp.c
clang: error: no such file or directory: '\n'
C compiler test failed.

I've tried it both with both --cc=clang (which worked under the previous operating system and tools) 
and the default compiler (which it refers to as gcc).  
I'm not completely convinced it is a compiler problem.  
Thirteen other packages have all compiled without problem now.  
Also, the compiler seems to be getting tripped up by a newline character somewhere 
in the code (see the end of the log)?