[x264-devel] [PATCH 1/2] fix arm asm for clang in 'arm: Implement x264_mbtree_propagate_{cost, list}_neon'

[Martin Storsjö]

On Wed, 30 Sep 2015, Janne Grunau wrote:

> ---
> common/arm/mc-a.S | 6 +++---
> 1 file changed, 3 insertions(+), 3 deletions(-)
>
> diff --git a/common/arm/mc-a.S b/common/arm/mc-a.S
> index b06b957..30d1c1a 100644
> --- a/common/arm/mc-a.S
> +++ b/common/arm/mc-a.S
> @@ -1777,7 +1777,7 @@ function x264_mbtree_propagate_cost_neon
>     vld1.16         {q9},  [r2]!
>     vld1.16         {q10}, [r3]!
>     vld1.16         {q11}, [r4]!
> -    vbic.u16        q10, q10, #0xc000
> +    vbic.u16        q10, #0xc000
>     vmin.u16        q10, q9,  q10
>     vmull.u16       q12, d18, d22           @ propagate_intra
>     vmull.u16       q13, d19, d23           @ propagate_intra
> @@ -1847,9 +1847,9 @@ function x264_mbtree_propagate_list_internal_neon
>     vmovn.i16       d16, q8
>     vmovn.i16       d17, q9
>     vadd.s16        q13, q13, q0
> -    vbic.i16        q8,  q8,  #128+64+32
> +    vbic.i16        q8,  #128+64+32
>     vadd.s16        q0,  q0,  q11
> -    vbic.i16        q8,  q8,  #(128+64+32)<<8
> +    vbic.i16        q8,  #(128+64+32)<<8
>     vst1.16         {q12, q13},  [r3, :128]!
>     vsub.i8         q9,  q3,  q8
>     vmull.u8        q12, d17, d16           @ idx3weight = y*x
> -- 
> 2.6.0

Ok

// Martin