On Thu, Aug 16, 2012 at 6:28 PM, Luca Barbato <[email protected]> wrote:
> On 08/16/2012 02:16 PM, Martin Storsjö wrote:
>> On Thu, 16 Aug 2012, Samuel Pitoiset wrote:
>>
>>> Some strict servers like crtmpserver returns an error when the specified
>>> stream is not found in call to releaseStream.
>>> ---
>>> libavformat/rtmpproto.c | 7 ++++---
>>> 1 file changed, 4 insertions(+), 3 deletions(-)
>>>
>>> diff --git a/libavformat/rtmpproto.c b/libavformat/rtmpproto.c
>>> index db11501..49b7e19 100644
>>> --- a/libavformat/rtmpproto.c
>>> +++ b/libavformat/rtmpproto.c
>>> @@ -398,7 +398,7 @@ static int gen_release_stream(URLContext *s,
>>> RTMPContext *rt)
>>>     ff_amf_write_null(&p);
>>>     ff_amf_write_string(&p, rt->playpath);
>>>
>>> -    return rtmp_send_packet(rt, &pkt, 0);
>>> +    return rtmp_send_packet(rt, &pkt, 1);
>>> }
>>>
>>> /**
>>> @@ -1257,8 +1257,9 @@ static int handle_invoke_error(URLContext *s,
>>> RTMPPacket *pkt)
>>>
>>>     if (!ff_amf_get_field_value(pkt->data + 9, data_end,
>>>                                 "description", tmpstr, sizeof(tmpstr))) {
>>> -        if (tracked_method && !strcmp(tracked_method, "_checkbw")) {
>>> -            /* Ignore _checkbw errors. */
>>> +        if (tracked_method && (!strcmp(tracked_method, "_checkbw") ||
>>> +                               !strcmp(tracked_method,
>>> "releaseStream"))) {
>>> +            /* Ignore _checkbw and releaseStream errors. */
>>>             level = AV_LOG_WARNING;
>>>             ret = 0;
>>>         } else
>>> --
>>> 1.7.11.3
>>
>> Looks good to me. Luca mentioned that there were some other methods as
>> well that we should ignore errors for (FCSubscribe, or was it
>> FCPublish?), it might be good to fix them all at once.
>
> both, you can have rtmp servers with clustering unimplemented.

Okay, but do you have a test case for that? I mean a rtmp link because
I never seen that's issue...

-- 
Best regards,
Samuel Pitoiset.
_______________________________________________
libav-devel mailing list
[email protected]
https://lists.libav.org/mailman/listinfo/libav-devel

Reply via email to