[FFmpeg-devel] libossupport status

Luca Abeni lucabe72
Sun Dec 23 13:23:33 CET 2007

Hi Vadim,

On Sat, 2007-12-22 at 23:42 +0100, Vadim Lebedev wrote:
> But it seems that original message in this thread suggest different 
> approach:  The patch for example proposes
> to remove #include "os_support.h" and
> to replace:
>     ff_socket_nonblock(server_fd, 1);
> by:
>     fcntl(server_fd, F_SETFL, fcntl(server_fd, F_GETFL) | O_NONBLOCK);
> This will no work on win32 (there is no F_SETFL on WIN32)

If you look at libossupport/netinet/in.h, you will see:
static inline int fcntl(int fd, int cmd, ...)
    if(cmd == F_GETFL){
        return 0;

    if(cmd == F_SETFL){
        va_list vl;
        long arg;
        u_long enable;

        va_start(vl, cmd);
        arg = va_arg(vl, long);
        enable= !!(arg & O_NONBLOCK);
        return ioctlsocket(fd, FIONBIO, &enable);

    return -1;

I agree that this is the wrong header for containing this code (as I
wrote in the original email, I just put all the code il netinet/in.h,
and I still have to split it in the correct header files).

> The best way IMO will be to keep evrywhere
> #include "os_support.h" 
> and have functions like 'ff_socket_nonblock' defined in os_support.h as 
> macros or noops
> this way there will be no runtime and linktime deps on most platforms
This is the idea I tried in my previous patch. But this solution will
either put some OS-dependent code in libav* or create a libav*
dependency on libossupport... So, I switched to the new proposal...
But now, reading the last emails from Michael I suspect I misunderstood
something (maybe the trick is in defining "depends on"... ;-)

I'll have to re-think about this...


More information about the ffmpeg-devel mailing list