On Tue, 19 Jun 2012, Måns Rullgård wrote:
Martin Storsjö <[email protected]> writes:
On Tue, 19 Jun 2012, Måns Rullgård wrote:
Martin Storsjö <[email protected]> writes:
Also use ff_neterrno() instead of errno directly (which doesn't work
on windows), for getting the error code.
---
libavformat/mmst.c | 20 +++++++++++++++-----
libavformat/tcp.c | 6 ++++--
libavformat/udp.c | 31 +++++++++++++++++++------------
3 files changed, 38 insertions(+), 19 deletions(-)
diff --git a/libavformat/mmst.c b/libavformat/mmst.c
index 4b96f5d..1e89f7f 100644
--- a/libavformat/mmst.c
+++ b/libavformat/mmst.c
@@ -141,10 +141,12 @@ static int send_command_packet(MMSTContext *mmst)
// write it out.
write_result= ffurl_write(mms->mms_hd, mms->out_buffer, exact_length);
if(write_result != exact_length) {
+ char errbuf[100];
+ av_strerror(write_result, errbuf, sizeof(errbuf));
Can ffurl_write return a partial length?
If I read the code correctly, it could do that at EOF (likewise for
ffurl_read_complete).
Then a lot of error handling in that file is broken.
Yes, it looks a bit wonky.
Any opinions on the rest of the patch, for tcp/udp?
// Martin
_______________________________________________
libav-devel mailing list
[email protected]
https://lists.libav.org/mailman/listinfo/libav-devel
