From: Marton Balint <cus@passwd.hu>
To: FFmpeg development discussions and patches <ffmpeg-devel@ffmpeg.org>
Subject: Re: [FFmpeg-devel] [PATCH] libRIST: allow setting fifo size and fail on overflow.
Date: Fri, 10 Jun 2022 00:05:41 +0200 (CEST)
Message-ID: <366220ee-ccee-286b-522f-53f1d2fc6c1b@passwd.hu> (raw)
In-Reply-To: <8eaadf2e-6c71-581a-677a-fa46a7f049ca@passwd.hu>
On Tue, 1 Feb 2022, Marton Balint wrote:
>
>
> On Tue, 11 Jan 2022, Gijs Peskens wrote:
>
>> Introduce fifo_size and overrun_nonfatal params to configure fifo buffer
>> behavior.
>>
>> Use newly introduced RIST_DATA_FLAGS_OVERFLOW flag to check for overrun
>> and error out in that case.
>> ---
>> doc/protocols.texi | 9 +++++++++
>> libavformat/librist.c | 37 +++++++++++++++++++++++++++++++++++++
>> 2 files changed, 46 insertions(+)
Applied with some fixes.
Regards,
Marton
>>
>> diff --git a/doc/protocols.texi b/doc/protocols.texi
>> index d207df0b52..f1acf0cc77 100644
>> --- a/doc/protocols.texi
>> +++ b/doc/protocols.texi
>> @@ -745,6 +745,15 @@ Set internal RIST buffer size in milliseconds for
>> @@ retransmission of data.
>> Default value is 0 which means the librist default (1 sec). Maximum value
>> is 30
>> seconds.
>>
>> +@item fifo_size
>> +Size of the librist receiver output fifo in number of packets. This must
>> be a
>> +power of 2.
>> +Defaults to 8192 (vs the libRIST default of 1024).
>> +
>> +@item overrun_nonfatal=@var{1|0}
>> +Survive in case of libRIST fifo buffer overrun. Default
>> +value is 0.
>
> Please use "librist" (all lowercase) consistently.
>
>> +
>> @ item pkt_size
>> Set maximum packet size for sending data. 1316 by default.
>>
>> diff --git a/libavformat/librist.c b/libavformat/librist.c
>> index 378b635ea7..87c14eb265 100644
>> --- a/libavformat/librist.c
>> +++ b/libavformat/librist.c
>> @@ -43,6 +43,9 @@
>> ((patch) + ((minor)* 0x100) + ((major) *0x10000))
>> #define FF_LIBRIST_VERSION
>> FF_LIBRIST_MAKE_VERSION(LIBRIST_API_VERSION_MAJOR,
>> LIBRIST_API_VERSION_MINOR, LIBRIST_API_VERSION_PATCH)
>> #define FF_LIBRIST_VERSION_41 FF_LIBRIST_MAKE_VERSION(4, 1, 0)
>> +#define FF_LIBRIST_VERSION_42 FF_LIBRIST_MAKE_VERSION(4, 2, 0)
>> +
>> +#define FF_LIBRIST_FIFO_SIZE_DEFAULT 8192
>
> FF_LIBRIST prefix is not really needed, as this is not in a header.
>
>>
>> typedef struct RISTContext {
>> const AVClass *class;
>> @@ -52,6 +55,8 @@ typedef struct RISTContext {
>> int packet_size;
>> int log_level;
>> int encryption;
>> + int fifo_size;
>> + int overrun_nonfatal;
>> char *secret;
>>
>> struct rist_logging_settings logging_settings;
>> @@ -70,6 +75,8 @@ static const AVOption librist_options[] = {
>> { "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 in ms", OFFSET(buffer_size),
>> AV_OPT_TYPE_INT, {.i64=0}, 0, 30000, .flags = D|E },
>> + { "fifo_size", "Set libRIST fifo buffer. Size must be power of 2",
>> OFFSET(fifo_size), AV_OPT_TYPE_INT, {.i64=FF_LIBRIST_FIFO_SIZE_DEFAULT}, 2
>> << 9, 2 << 15, .flags = D|E },
>
> Minimum and maximum binary notation is very strange. Use 1<<power, or maybe
> I'd just write 1024 and 65536. Much more readable. Are these arbitrary
> limits, or the implementation only supports these? Maybe you can be a bit
> more lax, e.g. 256 minimum and 262144 max.
>
>> + { "overrun_nonfatal", "survive in case of libRIST receiving circular
>> buffer overrun", OFFSET(overrun_nonfatal), AV_OPT_TYPE_BOOL, {.i64 = 0},
>> 0, 1, D },
>> { "pkt_size", "set packet size", OFFSET(packet_size),
>> AV_OPT_TYPE_INT, {.i64=1316}, 1, MAX_PAYLOAD_SIZE,
>> .flags = D|E },
>> { "log_level", "set loglevel", OFFSET(log_level),
>> AV_OPT_TYPE_INT, {.i64=RIST_LOG_INFO}, -1, INT_MAX, .flags =
>> D|E },
>> { "secret", "set encryption secret",OFFSET(secret),
>> AV_OPT_TYPE_STRING,{.str=NULL}, 0, 0, .flags =
>> D|E },
>> @@ -161,6 +168,20 @@ static int librist_open(URLContext *h, const char
>> @@ *uri, int flags)
>> if (ret < 0)
>> goto err;
>>
>> + //Prior to 4.2.0 there was a bug in libRIST which made this call
>> always fail.
>> +#if FF_LIBRIST_VERSION >= FF_LIBRIST_VERSION_42
>> + if (flags & AVIO_FLAG_READ) {
>> + ret = rist_receiver_set_output_fifo_size(s->ctx, s->fifo_size);
>> + if (ret != 0) {
>> + goto err;
>> + }
>> + }
>> +#else
>> + if (s->fifo_size != FF_LIBRIST_FIFO_SIZE_DEFAULT) {
>
> Isn't this supposed to be if (s->fifo_size != 1024)? Considering that we are
> aiming for a bigger default in ffmpeg compared to what is in the library as
> default, maybe the user should be aware that we cannot set it?
>
>> + av_log(h, AV_LOG_ERROR, "libRIST prior to 0.2.7 has a bug which
>> fails setting the fifo buffer size\n");
>> + }
>> +#endif
>> +
>
> Thanks,
> Marton
> _______________________________________________
> ffmpeg-devel mailing list
> ffmpeg-devel@ffmpeg.org
> https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
>
> To unsubscribe, visit link above, or email
> ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
>
_______________________________________________
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-request@ffmpeg.org with subject "unsubscribe".
next prev parent reply other threads:[~2022-06-09 22:05 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-01-11 15:07 Gijs Peskens
2022-01-31 8:12 ` Gijs Peskens
2022-01-31 23:03 ` Marton Balint
2022-06-09 22:05 ` Marton Balint [this message]
-- strict thread matches above, loose matches on Subject: below --
2022-01-11 9:23 Gijs Peskens
2022-01-11 9:43 ` "zhilizhao(赵志立)"
2022-01-11 9:45 ` Gijs Peskens
2022-01-11 10:34 ` Gijs Peskens
2022-01-11 10:35 ` Gijs Peskens
2022-01-11 11:20 ` "zhilizhao(赵志立)"
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=366220ee-ccee-286b-522f-53f1d2fc6c1b@passwd.hu \
--to=cus@passwd.hu \
--cc=ffmpeg-devel@ffmpeg.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Git Inbox Mirror of the ffmpeg-devel mailing list - see https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
This inbox may be cloned and mirrored by anyone:
git clone --mirror https://master.gitmailbox.com/ffmpegdev/0 ffmpegdev/git/0.git
# If you have public-inbox 1.1+ installed, you may
# initialize and index your mirror using the following commands:
public-inbox-init -V2 ffmpegdev ffmpegdev/ https://master.gitmailbox.com/ffmpegdev \
ffmpegdev@gitmailbox.com
public-inbox-index ffmpegdev
Example config snippet for mirrors.
AGPL code for this site: git clone https://public-inbox.org/public-inbox.git