[FFmpeg-devel] [PATCH] Observe Set-Cookie headers in HLS streams

wm4 nfxjfg at googlemail.com
Fri Apr 7 09:48:29 EEST 2017


On Thu,  6 Apr 2017 22:48:59 -0400
Micah Galizia <micahgalizia at gmail.com> wrote:

> Signed-off-by: Micah Galizia <micahgalizia at gmail.com>
> ---
>  libavformat/hls.c | 10 ++++++++--
>  1 file changed, 8 insertions(+), 2 deletions(-)
> 
> diff --git a/libavformat/hls.c b/libavformat/hls.c
> index bac53a4..ab81863 100644
> --- a/libavformat/hls.c
> +++ b/libavformat/hls.c
> @@ -630,8 +630,14 @@ static int open_url(AVFormatContext *s, AVIOContext **pb, const char *url,
>      ret = s->io_open(s, pb, url, AVIO_FLAG_READ, &tmp);
>      if (ret >= 0) {
>          // update cookies on http response with setcookies.
> -        void *u = (s->flags & AVFMT_FLAG_CUSTOM_IO) ? NULL : s->pb;
> -        update_options(&c->cookies, "cookies", u);
> +        char *new_cookies = NULL;
> +
> +        av_opt_get(*pb, "cookies", AV_OPT_SEARCH_CHILDREN, (uint8_t**)&new_cookies);
> +        if (new_cookies) {
> +            av_free(c->cookies);
> +            c->cookies = new_cookies;
> +        }
> +
>          av_dict_set(&opts, "cookies", c->cookies, 0);
>      }
>  

So far, all code is doing the same thing by calling update_options()
on pb, or NULL if it's a custom IO context.

What you seem to change is always using the pb (even if it's a custom
context), duplicating the update_options() code (subtly changing some
corner case behavior), and, this is probably the important one, you use
*pb instead of s->pb.

I suspect the latter is the only change that matters, and you probably
want:

void *u = (s->flags & AVFMT_FLAG_CUSTOM_IO) ? NULL : *pb;

Now I have no idea why it checks "s->flags & AVFMT_FLAG_CUSTOM_IO", or
what it means, but it feels very wrong. A common case is probably that
the main playlist is accessed through a custom AVIOContext, but further
nested playlists or actual .ts data probably use libavformat's code.

I'm thinking that all uses of AVFMT_FLAG_CUSTOM_IO anywhere are
probably bugs.


More information about the ffmpeg-devel mailing list