avcodec/dirac_dwt_template: Fix integer overflow in vertical_compose53iL0()
authorMichael Niedermayer <michael@niedermayer.cc>
Fri, 18 Aug 2017 14:42:58 +0000 (16:42 +0200)
committerMichael Niedermayer <michael@niedermayer.cc>
Mon, 21 Aug 2017 21:27:09 +0000 (23:27 +0200)
Fixes: runtime error: signed integer overflow: 2147483646 + 2 cannot be represented in type 'int'
Fixes: 3013/clusterfuzz-testcase-minimized-4644084197097472

Found-by: continuous fuzzing process https://github.com/google/oss-fuzz/tree/master/projects/ffmpeg
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
(cherry picked from commit a165b53daa8a3a526d2328ca72c4aa9e7f163045)
Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
libavcodec/dirac_dwt_template.c

index 972c711..e436c24 100644 (file)
@@ -49,7 +49,7 @@ static void RENAME(vertical_compose53iL0)(uint8_t *_b0, uint8_t *_b1, uint8_t *_
     TYPE *b1 = (TYPE *)_b1;
     TYPE *b2 = (TYPE *)_b2;
     for (i = 0; i < width; i++)
-        b1[i] -= (b0[i] + b2[i] + 2) >> 2;
+        b1[i] -= (int)(b0[i] + (unsigned)b2[i] + 2) >> 2;
 }
 
 static av_always_inline void RENAME(interleave)(TYPE *dst, TYPE *src0, TYPE *src1, int w2,