[FFmpeg-devel] [PATCH] libavformat: add librist protocol

[Paul B Mahol]

On Fri, Feb 26, 2021 at 9:54 PM Marton Balint <cus at passwd.hu> wrote:

>
>
> On Fri, 26 Feb 2021, Paul B Mahol wrote:
>
> > This work is sponsored by Open Broadcast Systems.
> >
> > Signed-off-by: Paul B Mahol <onemda at gmail.com>
> > ---
> > configure               |   5 +
> > doc/protocols.texi      |  29 +++++
> > libavformat/Makefile    |   1 +
> > libavformat/librist.c   | 248 ++++++++++++++++++++++++++++++++++++++++
> > libavformat/protocols.c |   1 +
> > 5 files changed, 284 insertions(+)
> > create mode 100644 libavformat/librist.c
> >
>
> > +#define D AV_OPT_FLAG_DECODING_PARAM
> > +#define E AV_OPT_FLAG_ENCODING_PARAM
> > +#define OFFSET(x) offsetof(RISTContext, x)
> > +static const AVOption librist_options[] = {
> > +    { "rist_profile","set profile",     OFFSET(profile),
>  AV_OPT_TYPE_INT,   {.i64=RIST_PROFILE_MAIN},     0, 2, .flags = D|E,
> "profile" },
> > +    { "simple",      NULL,              0,
>  AV_OPT_TYPE_CONST, {.i64=RIST_PROFILE_SIMPLE},   0, 0, .flags = D|E,
> "profile" },
> > +    { "main",        NULL,              0,
>  AV_OPT_TYPE_CONST, {.i64=RIST_PROFILE_MAIN},     0, 0, .flags = D|E,
> "profile" },
> > +    { "advanced",    NULL,              0,
>  AV_OPT_TYPE_CONST, {.i64=RIST_PROFILE_ADVANCED}, 0, 0, .flags = D|E,
> "profile" },
> > +    { "buffer_size", "set buffer_size", OFFSET(buffer_size),
> AV_OPT_TYPE_INT,   {.i64=0},                     0, INT_MAX, .flags = D|E },
> > +    { "log_level",   "set loglevel",    OFFSET(log_level),
>  AV_OPT_TYPE_INT,   {.i64=-1},                   -1, INT_MAX, .flags = D|E
> },
> > +    { "secret", "set encryption secret",OFFSET(secret),
> AV_OPT_TYPE_STRING,{.str=NULL},                  0, 0,       .flags = D|E },
> > +    { "encryption","set encryption type",OFFSET(encryption),
> AV_OPT_TYPE_INT   ,{.i64=0},                     0, INT_MAX, .flags = D|E },
>
> Why have you removed pkt_size? For writing, it is surely useful to be
> configurable, please put it back with -1 as default, so different default
> can be used for read and write. And for writing, the default of 1316
> should be used, as we typically want to pump mpegts into it, split on
> packet boundaries and want no fragmentation.
>

This options makes no sense for read/receive end.


>
> > +static int librist_open(URLContext *h, const char *uri, int flags)
> > +{
> > +    RISTContext *s = h->priv_data;
> > +    struct rist_logging_settings *logging_settings =
> &s->logging_settings;
> > +    struct rist_peer_config *peer_config = &s->peer_config;
> > +    int ret;
> > +
> > +    if ((flags & (AVIO_FLAG_WRITE | AVIO_FLAG_READ)) ==
> (AVIO_FLAG_WRITE | AVIO_FLAG_READ))
> > +        return AVERROR(EINVAL);
>
> AVIO_FLAG_READ_WRITE
>
> > +
> > +    return risterr2ret(ret);
> > +}
> > +
> > +static int librist_read(URLContext *h, uint8_t *buf, int size)
> > +{
> > +    RISTContext *s = h->priv_data;
> > +    int available_size = size;
> > +    int queue_size = 0;
> > +    int offset = 0;
> > +
> > +    do {
> > +        const struct rist_data_block *data_block;
> > +        int ret;
> > +
> > +        ret = rist_receiver_data_read(s->ctx, &data_block, offset == 0
> ? POLLING_TIME : 0);
> > +        if (ret < 0)
> > +            return risterr2ret(ret);
> > +
> > +        if (ret == 0 || data_block->payload_len <= 0)
> > +            break;
> > +
> > +        queue_size = ret;
> > +        av_assert0(data_block->payload_len <= available_size);
> > +
> > +        size = FFMIN(available_size, data_block->payload_len);
> > +        memcpy(buf + offset, data_block->payload, size);
> > +        offset += size;
> > +        available_size -= size;
> > +        rist_receiver_data_block_free((struct
> rist_data_block**)&data_block);
> > +        queue_size--;
>
> No, read one packet in one read call, do not merge packets. Protocols
> don't merge packets. UDP does not merge packets. If this is some
> workaround for librist dropping packets then figure out the bug causing
> that. If UDP works, librist should work too, even without merging packets.
>
> Regards,
> Marton
> _______________________________________________
> ffmpeg-devel mailing list
> ffmpeg-devel at ffmpeg.org
> https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
>
> To unsubscribe, visit link above, or email
> ffmpeg-devel-request at ffmpeg.org with subject "unsubscribe".