On Sat, Apr 30, 2011 at 01:54:57PM -0400, Brad wrote:
> On 30/04/11 1:03 PM, Diogo Franco wrote:
>> FreeBSD, unlike (whatever OS the default -D are trying to fix),
>> defines every symbol by default; conformance defines are supposed to
>> restrict FreeBSD's definitions to only the requested specification.
>>
>> Fix this by defining conformance flags only when we are not building
>> against FreeBSD's libc.
>> ---
>>   configure |    5 ++++-
>>   1 files changed, 4 insertions(+), 1 deletions(-)
>>
>> diff --git a/configure b/configure
>> index 2b5aeab..0d39f07 100755
>> --- a/configure
>> +++ b/configure
>> @@ -2310,7 +2310,6 @@ if test "$?" != 0; then
>>       die "C compiler test failed."
>>   fi
>>
>> -add_cppflags -D_ISOC99_SOURCE -D_POSIX_C_SOURCE=200112
>>   check_cflags -std=c99
>>   check_cc -D_FILE_OFFSET_BITS=64<<EOF&&  add_cppflags -D_FILE_OFFSET_BITS=64
>>   #include<stdlib.h>
>> @@ -2376,9 +2375,11 @@ case $target_os in
>>       dragonfly)
>>           enable malloc_aligned
>>           disable symver
>> +        fbsd_libc=1
>>           ;;
>>       freebsd)
>>           enable malloc_aligned
>> +        fbsd_libc=1
>>           ;;
>>       bsd/os)
>>           add_extralibs -lpoll -lgnugetopt
>> @@ -2499,6 +2500,8 @@ case $target_os in
>>           ;;
>>   esac
>>
>> +test -z "$fbsd_libc"&&  add_cppflags -D_ISOC99_SOURCE 
>> -D_POSIX_C_SOURCE=200112
>> +
>>   echo 
>> "config:$arch:$subarch:$cpu:$target_os:$cc_ident:$LIBAV_CONFIGURATION">config.fate
>>
>>   check_cpp_condition stdlib.h "defined(__PIC__) || defined(__pic__) || 
>> defined(PIC)"&&  enable pic
>
> This isn't specific to just FreeBSD but all BSD and derived OS's. .e.g.  
> FreeBSD, OpenBSD, NetBSD, DragonFly and OS X. Also _ISOC99_SOURCE isn't  
> an issue, just defining _POSIX_C_SOURCE. Then the hacks in  
> libavdevice/bktr.c and libavutil/ppc/cpu.c can be removed and
> the -D_XOPEN_SOURCE=600 for NetBSD which didn't actually fix
> anything at the time it was added.

What about the -D__EXTENSIONS__ employed by Solaris?

Diego
_______________________________________________
libav-devel mailing list
[email protected]
https://lists.libav.org/mailman/listinfo/libav-devel

Reply via email to