[FFmpeg-devel] [PATCH 1/3] avcodec/mpeg12dec: Check input size

[FFmpeg-devel] [PATCH 1/3] avcodec/mpeg12dec: Check input size

[Michael Niedermayer]

Fixes: Timeout
Fixes: 53599/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_IPU_fuzzer-4950102511058944

Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
---
 libavcodec/mpeg12dec.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/libavcodec/mpeg12dec.c b/libavcodec/mpeg12dec.c
index 914516bbd9..c93368e255 100644
--- a/libavcodec/mpeg12dec.c
+++ b/libavcodec/mpeg12dec.c
@@ -2969,6 +2969,9 @@ static int ipu_decode_frame(AVCodecContext *avctx, AVFrame *frame,
     GetBitContext *gb = &m->gb;
     int ret;
 
+    if (avpkt->size*8LL < (avctx->width+15)/16 * ((avctx->height+15)/16) * 2 * 7)
+        return AVERROR_INVALIDDATA;
+
     ret = ff_get_buffer(avctx, frame, 0);
     if (ret < 0)
         return ret;
-- 
2.17.1

_______________________________________________
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] [PATCH 2/3] avcodec/vqcdec: Check width & 15

[Michael Niedermayer]

Various parts of the code assume that width can be divided by various powers of 2
without rounding

Fixes: out of array access
Fixes: 53623/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_VQC_fuzzer-6209269924233216

Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
---
 libavcodec/vqcdec.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/libavcodec/vqcdec.c b/libavcodec/vqcdec.c
index 18cd99462e..c3bce87974 100644
--- a/libavcodec/vqcdec.c
+++ b/libavcodec/vqcdec.c
@@ -71,6 +71,9 @@ static av_cold int vqc_decode_init(AVCodecContext * avctx)
     static AVOnce init_static_once = AV_ONCE_INIT;
     VqcContext *s = avctx->priv_data;
 
+    if (avctx->width & 15)
+        return AVERROR_PATCHWELCOME;
+
     s->vectors = av_malloc((avctx->width * avctx->height * 3) / 2);
     if (!s->vectors)
         return AVERROR(ENOMEM);
-- 
2.17.1

_______________________________________________
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] [PATCH 3/3] tools/target_dec_fuzzer: Adjust threshold for RSCC

[Michael Niedermayer]

Fixes: Timeout
Fixes: 53681/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_RSCC_fuzzer-6627455180079104

Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
---
 tools/target_dec_fuzzer.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/tools/target_dec_fuzzer.c b/tools/target_dec_fuzzer.c
index e64a362fb6..a20345db5c 100644
--- a/tools/target_dec_fuzzer.c
+++ b/tools/target_dec_fuzzer.c
@@ -267,6 +267,7 @@ int LLVMFuzzerTestOneInput(const uint8_t *data, size_t size) {
     case AV_CODEC_ID_QTRLE:       maxpixels  /= 16;    break;
     case AV_CODEC_ID_PAF_VIDEO:   maxpixels  /= 16;    break;
     case AV_CODEC_ID_PRORES:      maxpixels  /= 256;   break;
+    case AV_CODEC_ID_RSCC:        maxpixels  /= 256;   break;
     case AV_CODEC_ID_RASC:        maxpixels  /= 16;    break;
     case AV_CODEC_ID_SANM:        maxpixels  /= 16;    break;
     case AV_CODEC_ID_SCPR:        maxpixels  /= 32;    break;
-- 
2.17.1

_______________________________________________
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".

Re: [FFmpeg-devel] [PATCH 1/3] avcodec/mpeg12dec: Check input size

[Kieran Kunhya]

On Sun, 27 Nov 2022 at 22:34, Michael Niedermayer <michael@niedermayer.cc>
wrote:

> Fixes: Timeout
> Fixes:
> 53599/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_IPU_fuzzer-4950102511058944
>
> Found-by: continuous fuzzing process
> https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
> Signed-off-by
> <https://github.com/google/oss-fuzz/tree/master/projects/ffmpegSigned-off-by>:
> Michael Niedermayer <michael@niedermayer.cc>
> ---
>  libavcodec/mpeg12dec.c | 3 +++
>  1 file changed, 3 insertions(+)
>
> diff --git a/libavcodec/mpeg12dec.c b/libavcodec/mpeg12dec.c
> index 914516bbd9..c93368e255 100644
> --- a/libavcodec/mpeg12dec.c
> +++ b/libavcodec/mpeg12dec.c
> @@ -2969,6 +2969,9 @@ static int ipu_decode_frame(AVCodecContext *avctx,
> AVFrame *frame,
>      GetBitContext *gb = &m->gb;
>      int ret;
>
> +    if (avpkt->size*8LL < (avctx->width+15)/16 * ((avctx->height+15)/16)
> * 2 * 7)
> +        return AVERROR_INVALIDDATA;
> +
>      ret = ff_get_buffer(avctx, frame, 0);
>      if (ret < 0)
>          return ret;
>

Where does this AVPacket limitation come from?
Are you able to explain in a comment where these numbers come from? In
particular the "2 * 7".

Kieran
_______________________________________________
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".

Re: [FFmpeg-devel] [PATCH 1/3] avcodec/mpeg12dec: Check input size

[Michael Niedermayer]

[-- Attachment #1.1: Type: text/plain, Size: 2251 bytes --]

On Mon, Nov 28, 2022 at 12:47:32AM +0000, Kieran Kunhya wrote:
> On Sun, 27 Nov 2022 at 22:34, Michael Niedermayer <michael@niedermayer.cc>
> wrote:
> 
> > Fixes: Timeout
> > Fixes:
> > 53599/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_IPU_fuzzer-4950102511058944
> >
> > Found-by: continuous fuzzing process
> > https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
> > Signed-off-by
> > <https://github.com/google/oss-fuzz/tree/master/projects/ffmpegSigned-off-by>:
> > Michael Niedermayer <michael@niedermayer.cc>
> > ---
> >  libavcodec/mpeg12dec.c | 3 +++
> >  1 file changed, 3 insertions(+)
> >
> > diff --git a/libavcodec/mpeg12dec.c b/libavcodec/mpeg12dec.c
> > index 914516bbd9..c93368e255 100644
> > --- a/libavcodec/mpeg12dec.c
> > +++ b/libavcodec/mpeg12dec.c
> > @@ -2969,6 +2969,9 @@ static int ipu_decode_frame(AVCodecContext *avctx,
> > AVFrame *frame,
> >      GetBitContext *gb = &m->gb;
> >      int ret;
> >
> > +    if (avpkt->size*8LL < (avctx->width+15)/16 * ((avctx->height+15)/16)
> > * 2 * 7)
> > +        return AVERROR_INVALIDDATA;
> > +
> >      ret = ff_get_buffer(avctx, frame, 0);
> >      if (ret < 0)
> >          return ret;
> >
> 
> Where does this AVPacket limitation come from?
> Are you able to explain in a comment where these numbers come from? In
> particular the "2 * 7".

the ipu format has 2 bits minimum per MB except the first but theres more than
1 bit before so that can be simplified to 2 bits minimum per MB "header"
each MB has 6 blocks
blocks can be mpeg1 intra or mpeg2 intra these will at least contain one DC VLC
and the shortest is 2 bits. We could consider that the luma one is min 3 but i
didnt also we could consider teh AC EOB again i didnt
6 blocks 2 bits + 2 for MB is 7*2
(3*4 + 2*2 + 2) would consider the luma dc seperatly.
we could add a 6*2 for the AC EOBs
so 2 + 3*4 + 2*2 + 6*2 would be closer to the theoretic minimum.
i will change the code to that and add a comment that this is the minimum
MB size considering MB, DC and AC-EOB

thx

[...]
-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

In a rich man's house there is no place to spit but his face.
-- Diogenes of Sinope

[-- Attachment #1.2: signature.asc --]
[-- Type: application/pgp-signature, Size: 195 bytes --]

[-- Attachment #2: Type: text/plain, Size: 251 bytes --]

_______________________________________________
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".

Re: [FFmpeg-devel] [PATCH 3/3] tools/target_dec_fuzzer: Adjust threshold for RSCC

[Michael Niedermayer]

[-- Attachment #1.1: Type: text/plain, Size: 618 bytes --]

On Sun, Nov 27, 2022 at 11:34:35PM +0100, Michael Niedermayer wrote:
> Fixes: Timeout
> Fixes: 53681/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_RSCC_fuzzer-6627455180079104
> 
> Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
> Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
> ---
>  tools/target_dec_fuzzer.c | 1 +
>  1 file changed, 1 insertion(+)

will apply

[...]
-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

The worst form of inequality is to try to make unequal things equal.
-- Aristotle

[-- Attachment #1.2: signature.asc --]
[-- Type: application/pgp-signature, Size: 195 bytes --]

[-- Attachment #2: Type: text/plain, Size: 251 bytes --]

_______________________________________________
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".

Re: [FFmpeg-devel] [PATCH 2/3] avcodec/vqcdec: Check width & 15

[Peter Ross]

[-- Attachment #1.1: Type: text/plain, Size: 1202 bytes --]

On Sun, Nov 27, 2022 at 11:34:34PM +0100, Michael Niedermayer wrote:
> Various parts of the code assume that width can be divided by various powers of 2
> without rounding
> 
> Fixes: out of array access
> Fixes: 53623/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_VQC_fuzzer-6209269924233216
> 
> Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
> Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
> ---
>  libavcodec/vqcdec.c | 3 +++
>  1 file changed, 3 insertions(+)
> 
> diff --git a/libavcodec/vqcdec.c b/libavcodec/vqcdec.c
> index 18cd99462e..c3bce87974 100644
> --- a/libavcodec/vqcdec.c
> +++ b/libavcodec/vqcdec.c
> @@ -71,6 +71,9 @@ static av_cold int vqc_decode_init(AVCodecContext * avctx)
>      static AVOnce init_static_once = AV_ONCE_INIT;
>      VqcContext *s = avctx->priv_data;
>  
> +    if (avctx->width & 15)
> +        return AVERROR_PATCHWELCOME;
> +
>      s->vectors = av_malloc((avctx->width * avctx->height * 3) / 2);
>      if (!s->vectors)
>          return AVERROR(ENOMEM);
> -- 
> 2.17.1

ok. please apply.

-- Peter
(A907 E02F A6E5 0CD2 34CD 20D2 6760 79C5 AC40 DD6B)

[-- Attachment #1.2: signature.asc --]
[-- Type: application/pgp-signature, Size: 195 bytes --]

[-- Attachment #2: Type: text/plain, Size: 251 bytes --]

_______________________________________________
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".

Re: [FFmpeg-devel] [PATCH 2/3] avcodec/vqcdec: Check width & 15

[Michael Niedermayer]

[-- Attachment #1.1: Type: text/plain, Size: 1517 bytes --]

On Thu, Dec 08, 2022 at 06:32:40PM +1100, Peter Ross wrote:
> On Sun, Nov 27, 2022 at 11:34:34PM +0100, Michael Niedermayer wrote:
> > Various parts of the code assume that width can be divided by various powers of 2
> > without rounding
> > 
> > Fixes: out of array access
> > Fixes: 53623/clusterfuzz-testcase-minimized-ffmpeg_AV_CODEC_ID_VQC_fuzzer-6209269924233216
> > 
> > Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
> > Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
> > ---
> >  libavcodec/vqcdec.c | 3 +++
> >  1 file changed, 3 insertions(+)
> > 
> > diff --git a/libavcodec/vqcdec.c b/libavcodec/vqcdec.c
> > index 18cd99462e..c3bce87974 100644
> > --- a/libavcodec/vqcdec.c
> > +++ b/libavcodec/vqcdec.c
> > @@ -71,6 +71,9 @@ static av_cold int vqc_decode_init(AVCodecContext * avctx)
> >      static AVOnce init_static_once = AV_ONCE_INIT;
> >      VqcContext *s = avctx->priv_data;
> >  
> > +    if (avctx->width & 15)
> > +        return AVERROR_PATCHWELCOME;
> > +
> >      s->vectors = av_malloc((avctx->width * avctx->height * 3) / 2);
> >      if (!s->vectors)
> >          return AVERROR(ENOMEM);
> > -- 
> > 2.17.1
> 
> ok. please apply.

will apply

thx

[...]

-- 
Michael     GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB

Complexity theory is the science of finding the exact solution to an
approximation. Benchmarking OTOH is finding an approximation of the exact

[-- Attachment #1.2: signature.asc --]
[-- Type: application/pgp-signature, Size: 195 bytes --]

[-- Attachment #2: Type: text/plain, Size: 251 bytes --]

_______________________________________________
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".