diff options
| author | Martin Storsjö <martin@martin.st> | 2024-05-21 14:03:07 +0300 |
|---|---|---|
| committer | Martin Storsjö <martin@martin.st> | 2024-05-21 19:20:06 +0300 |
| commit | 6093367147b72c0f3c78a61597a3452f7f0e510e (patch) | |
| tree | e8474ae3993fb5b055efd9dc03fcc705daa7cda1 | |
| parent | d43e1238374e77788977b2427612b526d6840a1e (diff) | |
| download | ffmpeg-6093367147b72c0f3c78a61597a3452f7f0e510e.tar.gz | |
checkasm: h264dsp: Avoid out of buffer writes when benchmarking
The loop filters can write before the pointer given to them;
the actual test invocations correctly used an offset, while
the benchmark calls were lacking an offset. Therefore, when
running with benchmarking, these tests could have spurious
failures.
Signed-off-by: Martin Storsjö <martin@martin.st>
| -rw-r--r-- | tests/checkasm/h264dsp.c | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/tests/checkasm/h264dsp.c b/tests/checkasm/h264dsp.c index 0cc1f32740..67b8dce53c 100644 --- a/tests/checkasm/h264dsp.c +++ b/tests/checkasm/h264dsp.c @@ -362,7 +362,7 @@ static void check_loop_filter(void) tc0[j][0], tc0[j][1], tc0[j][2], tc0[j][3]); \ fail(); \ } \ - bench_new(dst1, 32, alphas[j], betas[j], tc0[j]); \ + bench_new(dst1 + off, 32, alphas[j], betas[j], tc0[j]);\ } \ } \ } while (0) @@ -421,7 +421,7 @@ static void check_loop_filter_intra(void) j, alphas[j], betas[j]); \ fail(); \ } \ - bench_new(dst1, 32, alphas[j], betas[j]); \ + bench_new(dst1 + off, 32, alphas[j], betas[j]); \ } \ } \ } while (0) |