On Wed, 25 May 2022, Swinney, Jonathan wrote:
Signed-off-by: Jonathan Swinney <jswin...@amazon.com> --- tests/checkasm/sw_scale.c | 38 ++++++++++++++++++++++---------------- 1 file changed, 22 insertions(+), 16 deletions(-)diff --git a/tests/checkasm/sw_scale.c b/tests/checkasm/sw_scale.c index 3c0a083b42..6c223c48f9 100644 --- a/tests/checkasm/sw_scale.c +++ b/tests/checkasm/sw_scale.c @@ -148,7 +148,11 @@ static void check_hscale(void) { 8, 18 }, }; - int i, j, fsi, hpi, width; +#define LARGEST_INPUT_SIZE 512 +#define INPUT_SIZES 6 + static const int input_sizes[INPUT_SIZES] = {8, 24, 128, 144, 256, 512}; +
While these are all nice even values, would it make sense to test things with uneven sizes too?
This patch does seem good to me, however it does break the checkasm test on x86. Offhand I don't know where it differs between how swscale is used in reality (where I presume it does work just fine) and how it's set up in checkasm.
// Martin _______________________________________________ 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".
