diff options
| author | Wu Jianhua <jianhua.wu@intel.com> | 2021-09-16 15:34:08 +0800 |
|---|---|---|
| committer | Paul B Mahol <onemda@gmail.com> | 2021-09-18 12:38:01 +0200 |
| commit | 7bbad32d5ab69cb52bc92a5ec30c7b9838daa08a (patch) | |
| tree | 9260a90a0962514b7a456ff929aaeeb6be709733 | |
| parent | fcf10c925dc4a4ab883b4ad14ea61bbdaa6f8521 (diff) | |
libavfilter/x86/vf_gblur: correct the order of loop step
The problem was caused by if the width of the processed block
minus 1 is a multiple of the aligned number the instruction
jle .bscale_scalar would skip the Optimized Loop Step, which
will lead to an incorrect sampling when specifying steps more
than 1. Move the Optimized Loop Step after .bscale_scalar to
ensure the loop step is enabled.
Signed-off-by: Wu Jianhua <jianhua.wu@intel.com>
| -rw-r--r-- | libavfilter/x86/vf_gblur.asm | 3 |
1 files changed, 1 insertions, 2 deletions
diff --git a/libavfilter/x86/vf_gblur.asm b/libavfilter/x86/vf_gblur.asm index 64c067538a..16e802e002 100644 --- a/libavfilter/x86/vf_gblur.asm +++ b/libavfilter/x86/vf_gblur.asm @@ -524,9 +524,8 @@ cglobal horiz_slice, 4, 9, 9, ptr, width, height, steps, nu, bscale, x, y, step, cmp xq, 0 jg .loop_x_scalar - OPTIMIZED_LOOP_STEP - .bscale_scalar: + OPTIMIZED_LOOP_STEP sub ptrq, 4 sub localbufq, mmsize mulps m3, m1 |