[FFmpeg-devel] [PATCH] avcodec/ffv1dec: fix some unsupported pix_fmt

[Jerome Martinez]

Le 13/06/2016 à 19:55, Michael Niedermayer a écrit :
> On Mon, Jun 13, 2016 at 07:31:15PM +0200, Jerome Martinez wrote:
>> FFV1 decoder:
>>
>> When checking pix_fmt mapping, some bitstreams are mapped to an
>> incorrect pix_fmt instead of being rejected (ENOSYS).
>> Actually, such bitstreams are not supported (FFmpeg encoder does not
>> produce such bitstream, such bitstream may come only from another
>> encoder for the moment).
>>
>> - JPEG 2000 RCT 11/13/15/16 bit depths are mapped to a 8-bit FFmpeg
>> pix_fmt (e.g. bgr0), which is not expected.
>> - JPEG 2000 RCT 9/10/12/14 bit depths with alpha are mapped to a
>> FFmpeg pix_fmt without alpha (e.g. AV_PIX_FMT_GBRP9 for 9-bit with
>> alpha), which is not expected.
>>
>> The order for choosing the pix_fmt is changed to the one used by
>> YCbCr selection (<=8 bit first).
>> " && !f->transparency" is added to the other lines.
>>
>>   ffv1dec.c |   15 ++++++++-------
>>   1 file changed, 8 insertions(+), 7 deletions(-)
>> 27efc1a9821576a2a5bd8d1feaaf6c584fc69a62  0001-avcodec-ffv1dec-fix-some-unsupported-pix_fmt.patch
>>  From 90bfd748b0e25d7a0be037280f4a0a40242f8d27 Mon Sep 17 00:00:00 2001
>> From: =?UTF-8?q?J=C3=A9r=C3=B4me=20Martinez?= <jerome at mediaarea.net>
>> Date: Mon, 13 Jun 2016 19:18:22 +0200
>> Subject: [PATCH] avcodec/ffv1dec: fix some unsupported pix_fmt
>> MIME-Version: 1.0
>> Content-Type: text/plain; charset=UTF-8
>> Content-Transfer-Encoding: 8bit
>>
>> Signed-off-by: Jérôme Martinez <jerome at mediaarea.net>
>> ---
>>   libavcodec/ffv1dec.c | 15 ++++++++-------
>>   1 file changed, 8 insertions(+), 7 deletions(-)
>>
>> diff --git a/libavcodec/ffv1dec.c b/libavcodec/ffv1dec.c
>> index d2bf3a8..6a932b2 100644
>> --- a/libavcodec/ffv1dec.c
>> +++ b/libavcodec/ffv1dec.c
>> @@ -768,17 +768,18 @@ static int read_header(FFV1Context *f)
>>                      "chroma subsampling not supported in this colorspace\n");
>>               return AVERROR(ENOSYS);
>>           }
>> -        if (     f->avctx->bits_per_raw_sample ==  9)
>> +        if (     f->avctx->bits_per_raw_sample <=  8 && !f->transparency)
>> +            f->avctx->pix_fmt = AV_PIX_FMT_0RGB32;
>> +        else if (f->avctx->bits_per_raw_sample <=  8 && f->transparency)
>> +            f->avctx->pix_fmt = AV_PIX_FMT_RGB32;
>> +        else if (f->avctx->bits_per_raw_sample ==  9 && !f->transparency)
>>               f->avctx->pix_fmt = AV_PIX_FMT_GBRP9;
>> -        else if (f->avctx->bits_per_raw_sample == 10)
>> +        else if (f->avctx->bits_per_raw_sample == 10 && !f->transparency)
>>               f->avctx->pix_fmt = AV_PIX_FMT_GBRP10;
>> -        else if (f->avctx->bits_per_raw_sample == 12)
>> +        else if (f->avctx->bits_per_raw_sample == 12 && !f->transparency)
>>               f->avctx->pix_fmt = AV_PIX_FMT_GBRP12;
>> -        else if (f->avctx->bits_per_raw_sample == 14)
>> +        else if (f->avctx->bits_per_raw_sample == 14 && !f->transparency)
>>               f->avctx->pix_fmt = AV_PIX_FMT_GBRP14;
>> -        else
>> -        if (f->transparency) f->avctx->pix_fmt = AV_PIX_FMT_RGB32;
>> -        else                 f->avctx->pix_fmt = AV_PIX_FMT_0RGB32;
> an else "error" feels missing

I did exactly as the code few lines above: if f->colorspace == 0 and 
f->avctx->bits_per_raw_sample >=17, there is also no else:
         } else if (f->avctx->bits_per_raw_sample == 16 && f->transparency){
             switch(16 * f->chroma_h_shift + f->chroma_v_shift) {
             case 0x00: f->avctx->pix_fmt = AV_PIX_FMT_YUVA444P16; break;
             case 0x10: f->avctx->pix_fmt = AV_PIX_FMT_YUVA422P16; break;
             case 0x11: f->avctx->pix_fmt = AV_PIX_FMT_YUVA420P16; break;
             }
         } // **** an else "error" feels missing here too. ****



> pix_fmt would not be initialized after the patch

When I test the code, pix_fmt is initialized to -1 (AV_PIX_FMT_NONE) 
before running this part, and the error is caught by the "format not 
supported" message, which looks like a good message for that.


I have no problem with adding such else in the patch, but the code would 
not be coherent (an else "error" feels missing if f->colorspace == 0, 
and else "error" does not feel missing if f->colorspace == 1), so I 
would like to have the confirmation that this is what you really want.

Note: BTW about missing things, there is no av_log() message when 
f->colorspace == 0 && f->transparency && !f->chroma_plane && 
->avctx->bits_per_raw_sample >8 (=gray+alpha with a bit depth of more 
than 8), just a "else return AVERROR(ENOSYS);", I was actually willing 
to remove these 2 lines in order to have more coherency (no "else" so 
"format not supported" message a bit later in the code) but I did not 
want to add different fixes in the same patch.