On Mon, Sep 17, 2018 at 05:20:28PM +0800, Shiyou Yin wrote:
> Combined 1st and 2nd loop into one inline asm in function
> ff_vc1_inv_trans_8x8_mmi to
> reduce memory operation, and made some small optimization in
> ff_vc1_inv_trans_4x8_mmi.
> ---
> libavcodec/mips/vc1dsp_mmi.c | 227
> +
Combined 1st and 2nd loop into one inline asm in function
ff_vc1_inv_trans_8x8_mmi to
reduce memory operation, and made some small optimization in
ff_vc1_inv_trans_4x8_mmi.
---
libavcodec/mips/vc1dsp_mmi.c | 227 ++-
1 file changed, 94 insertions(+), 133 d
