Re: [FFmpeg-devel] [PATCH] swresample/swresample: check for invalid sample rates
On 5/25/19, Michael Niedermayer wrote:
> On Fri, May 24, 2019 at 06:05:42PM +0200, Paul B Mahol wrote:
>> Signed-off-by: Paul B Mahol
>> ---
>> libswresample/swresample.c | 8
>> 1 file changed, 8 insertions(+)
>>
>> diff --git a/libswresample/swresample.c b/libswresample/swresample.c
>> index 6d28e6a798..1ac5ef9a30 100644
>> --- a/libswresample/swresample.c
>> +++ b/libswresample/swresample.c
>> @@ -164,6 +164,14 @@ av_cold int swr_init(struct SwrContext *s){
>> return AVERROR(EINVAL);
>> }
>>
>> +if(s-> in_sample_rate <= 0){
>> +av_log(s, AV_LOG_ERROR, "Requested input sample rate %d is
>> invalid\n", s->in_sample_rate);
>> +return AVERROR(EINVAL);
>> +}
>> +if(s->out_sample_rate <= 0){
>> +av_log(s, AV_LOG_ERROR, "Requested output sample rate %d is
>> invalid\n", s->out_sample_rate);
>> +return AVERROR(EINVAL);
>> +}
>
> probably ok
>
> Only hypothetical issue i could see is convering sample types or channel
> layout with unspecified sample rate, that is both 0.
> dont know if that works currently but it at least semantically would make
> sense
It does not work, it reach floating-point exception and crashes.
(outlink sample rate set to 0 and input sample rate set to 48000)
>
> Thanks
>
> [...]
> --
> Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB
>
> I know you won't believe me, but the highest form of Human Excellence is
> to question oneself and others. -- Socrates
>
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".
Re: [FFmpeg-devel] [PATCH] swresample/swresample: check for invalid sample rates
On Sat, May 25, 2019 at 06:45:01PM +0200, Hendrik Leppkes wrote:
> On Sat, May 25, 2019 at 5:58 PM Michael Niedermayer
> wrote:
> >
> > On Fri, May 24, 2019 at 06:05:42PM +0200, Paul B Mahol wrote:
> > > Signed-off-by: Paul B Mahol
> > > ---
> > > libswresample/swresample.c | 8
> > > 1 file changed, 8 insertions(+)
> > >
> > > diff --git a/libswresample/swresample.c b/libswresample/swresample.c
> > > index 6d28e6a798..1ac5ef9a30 100644
> > > --- a/libswresample/swresample.c
> > > +++ b/libswresample/swresample.c
> > > @@ -164,6 +164,14 @@ av_cold int swr_init(struct SwrContext *s){
> > > return AVERROR(EINVAL);
> > > }
> > >
> > > +if(s-> in_sample_rate <= 0){
> > > +av_log(s, AV_LOG_ERROR, "Requested input sample rate %d is
> > > invalid\n", s->in_sample_rate);
> > > +return AVERROR(EINVAL);
> > > +}
> > > +if(s->out_sample_rate <= 0){
> > > +av_log(s, AV_LOG_ERROR, "Requested output sample rate %d is
> > > invalid\n", s->out_sample_rate);
> > > +return AVERROR(EINVAL);
> > > +}
> >
> > probably ok
> >
> > Only hypothetical issue i could see is convering sample types or channel
> > layout with unspecified sample rate, that is both 0.
> > dont know if that works currently but it at least semantically would make
> > sense
> >
>
> Unspecified channel layout can at least make some sense, since you
> still have the channel count to properly handle the audio data, but
> unspecified sample rate? How would that ever work?
you sure can convert 0.5 0.0 1.0 to 128 0 255 you do not need to know
the sample rate.
For a real life example, that would be a raw pcm file, to convert it
from 32bit float to 16bit signed int only that needs to be known
neither the number of channels nor the sample rate is needed
or said differently, for this example, you could set the sample rate
to anything it would not affect the output. So it is basically not
needed for the process
not saying this is worth supporting. Just that it could be supported
and the case at least to me makes sense ...
thanks
[...]
--
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB
Opposition brings concord. Out of discord comes the fairest harmony.
-- Heraclitus
signature.asc
Description: PGP signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".
Re: [FFmpeg-devel] [PATCH] swresample/swresample: check for invalid sample rates
On Sat, May 25, 2019 at 5:58 PM Michael Niedermayer
wrote:
>
> On Fri, May 24, 2019 at 06:05:42PM +0200, Paul B Mahol wrote:
> > Signed-off-by: Paul B Mahol
> > ---
> > libswresample/swresample.c | 8
> > 1 file changed, 8 insertions(+)
> >
> > diff --git a/libswresample/swresample.c b/libswresample/swresample.c
> > index 6d28e6a798..1ac5ef9a30 100644
> > --- a/libswresample/swresample.c
> > +++ b/libswresample/swresample.c
> > @@ -164,6 +164,14 @@ av_cold int swr_init(struct SwrContext *s){
> > return AVERROR(EINVAL);
> > }
> >
> > +if(s-> in_sample_rate <= 0){
> > +av_log(s, AV_LOG_ERROR, "Requested input sample rate %d is
> > invalid\n", s->in_sample_rate);
> > +return AVERROR(EINVAL);
> > +}
> > +if(s->out_sample_rate <= 0){
> > +av_log(s, AV_LOG_ERROR, "Requested output sample rate %d is
> > invalid\n", s->out_sample_rate);
> > +return AVERROR(EINVAL);
> > +}
>
> probably ok
>
> Only hypothetical issue i could see is convering sample types or channel
> layout with unspecified sample rate, that is both 0.
> dont know if that works currently but it at least semantically would make
> sense
>
Unspecified channel layout can at least make some sense, since you
still have the channel count to properly handle the audio data, but
unspecified sample rate? How would that ever work?
- Hendrik
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".
Re: [FFmpeg-devel] [PATCH] swresample/swresample: check for invalid sample rates
On Fri, May 24, 2019 at 06:05:42PM +0200, Paul B Mahol wrote:
> Signed-off-by: Paul B Mahol
> ---
> libswresample/swresample.c | 8
> 1 file changed, 8 insertions(+)
>
> diff --git a/libswresample/swresample.c b/libswresample/swresample.c
> index 6d28e6a798..1ac5ef9a30 100644
> --- a/libswresample/swresample.c
> +++ b/libswresample/swresample.c
> @@ -164,6 +164,14 @@ av_cold int swr_init(struct SwrContext *s){
> return AVERROR(EINVAL);
> }
>
> +if(s-> in_sample_rate <= 0){
> +av_log(s, AV_LOG_ERROR, "Requested input sample rate %d is
> invalid\n", s->in_sample_rate);
> +return AVERROR(EINVAL);
> +}
> +if(s->out_sample_rate <= 0){
> +av_log(s, AV_LOG_ERROR, "Requested output sample rate %d is
> invalid\n", s->out_sample_rate);
> +return AVERROR(EINVAL);
> +}
probably ok
Only hypothetical issue i could see is convering sample types or channel
layout with unspecified sample rate, that is both 0.
dont know if that works currently but it at least semantically would make
sense
Thanks
[...]
--
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB
I know you won't believe me, but the highest form of Human Excellence is
to question oneself and others. -- Socrates
signature.asc
Description: PGP signature
___
ffmpeg-devel mailing list
ffmpeg-devel@ffmpeg.org
https://ffmpeg.org/mailman/listinfo/ffmpeg-devel
To unsubscribe, visit link above, or email
ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".
Re: [FFmpeg-devel] [PATCH] swresample/swresample: check for invalid sample rates
On 5/24/19, Derek Buitenhuis wrote: > On 24/05/2019 17:05, Paul B Mahol wrote: >> Signed-off-by: Paul B Mahol >> --- >> libswresample/swresample.c | 8 >> 1 file changed, 8 insertions(+) > > Seems reasonable. What happens if these aren't in place? It may divide by zero and crash with floating-point exception. ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or email ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".
Re: [FFmpeg-devel] [PATCH] swresample/swresample: check for invalid sample rates
On 24/05/2019 17:05, Paul B Mahol wrote: > Signed-off-by: Paul B Mahol > --- > libswresample/swresample.c | 8 > 1 file changed, 8 insertions(+) Seems reasonable. What happens if these aren't in place? - Derek ___ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org https://ffmpeg.org/mailman/listinfo/ffmpeg-devel To unsubscribe, visit link above, or email ffmpeg-devel-requ...@ffmpeg.org with subject "unsubscribe".
