diff options
| author | Seppo Tomperi <seppo.tomperi@vtt.fi> | 2017-04-27 21:05:29 +0200 |
|---|---|---|
| committer | Martin Storsjö <martin@martin.st> | 2017-04-27 23:05:27 +0300 |
| commit | 0d4d43513786f1df4d561e1fac924fb0722c6700 (patch) | |
| tree | a4c0a68a55c233b0bcdbe04d7175f641ea99e30b /libavcodec/arm/hevc_idct.S | |
| parent | 970c76f32283bddf3a5afd24fe52db7a96186244 (diff) | |
| download | ffmpeg-0d4d43513786f1df4d561e1fac924fb0722c6700.tar.gz | |
hevc: Add NEON add_residual for bitdepth 8
Optimized by Alexandra Hájková.
Signed-off-by: Martin Storsjö <martin@martin.st>
Diffstat (limited to 'libavcodec/arm/hevc_idct.S')
| -rw-r--r-- | libavcodec/arm/hevc_idct.S | 88 |
1 files changed, 88 insertions, 0 deletions
diff --git a/libavcodec/arm/hevc_idct.S b/libavcodec/arm/hevc_idct.S index 41b1b290bf..3966e93d85 100644 --- a/libavcodec/arm/hevc_idct.S +++ b/libavcodec/arm/hevc_idct.S @@ -30,6 +30,94 @@ const trans, align=4 .short 57, 43, 25, 9 endconst +function ff_hevc_add_residual_4x4_8_neon, export=1 + vld1.16 {q0-q1}, [r1, :128] + vld1.32 d4[0], [r0, :32], r2 + vld1.32 d4[1], [r0, :32], r2 + vld1.32 d5[0], [r0, :32], r2 + vld1.32 d5[1], [r0, :32], r2 + sub r0, r0, r2, lsl #2 + vmovl.u8 q8, d4 + vmovl.u8 q9, d5 + vqadd.s16 q0, q0, q8 + vqadd.s16 q1, q1, q9 + vqmovun.s16 d0, q0 + vqmovun.s16 d1, q1 + vst1.32 d0[0], [r0, :32], r2 + vst1.32 d0[1], [r0, :32], r2 + vst1.32 d1[0], [r0, :32], r2 + vst1.32 d1[1], [r0, :32], r2 + bx lr +endfunc + +function ff_hevc_add_residual_8x8_8_neon, export=1 + mov r3, #8 +1: subs r3, #2 + vld1.16 {q0-q1}, [r1, :128]! + vld1.8 {d16}, [r0, :64] + add r12, r0, r2 + vld1.8 {d17}, [r12, :64] + vmovl.u8 q9, d16 + vmovl.u8 q8, d17 + vqadd.s16 q0, q9 + vqadd.s16 q1, q8 + vqmovun.s16 d0, q0 + vqmovun.s16 d1, q1 + vst1.8 d0, [r0, :64], r2 + vst1.8 d1, [r0, :64], r2 + bne 1b + bx lr +endfunc + +function ff_hevc_add_residual_16x16_8_neon, export=1 + mov r3, #16 + add r12, r0, r2 + add r2, r2, r2 +1: subs r3, #2 + vld1.8 {q8}, [r0, :128] + vld1.16 {q0, q1}, [r1, :128]! + vld1.8 {q11}, [r12, :128] + vld1.16 {q2, q3}, [r1, :128]! + vmovl.u8 q9, d16 + vmovl.u8 q10, d17 + vmovl.u8 q12, d22 + vmovl.u8 q13, d23 + vqadd.s16 q0, q9 + vqadd.s16 q1, q10 + vqadd.s16 q2, q12 + vqadd.s16 q3, q13 + vqmovun.s16 d0, q0 + vqmovun.s16 d1, q1 + vqmovun.s16 d2, q2 + vqmovun.s16 d3, q3 + vst1.8 {q0}, [r0, :128], r2 + vst1.8 {q1}, [r12, :128], r2 + bne 1b + bx lr +endfunc + +function ff_hevc_add_residual_32x32_8_neon, export=1 + mov r3, #32 +1: subs r3, #1 + vldm r1!, {q0-q3} + vld1.8 {q8, q9}, [r0, :128] + vmovl.u8 q10, d16 + vmovl.u8 q11, d17 + vmovl.u8 q12, d18 + vmovl.u8 q13, d19 + vqadd.s16 q0, q10 + vqadd.s16 q1, q11 + vqadd.s16 q2, q12 + vqadd.s16 q3, q13 + vqmovun.s16 d0, q0 + vqmovun.s16 d1, q1 + vqmovun.s16 d2, q2 + vqmovun.s16 d3, q3 + vst1.8 {q0, q1}, [r0, :128], r2 + bne 1b + bx lr +endfunc + .macro idct_4x4_dc bitdepth function ff_hevc_idct_4x4_dc_\bitdepth\()_neon, export=1 ldrsh r1, [r0] |