From a1c1d3c003b0ec16fdb6574913781313fb2c7ab6 Mon Sep 17 00:00:00 2001 From: Mans Rullgard Date: Wed, 2 Feb 2011 16:26:20 +0000 Subject: [PATCH] VP8: ARM NEON optimisations for dsp functions This adds NEON optimised versions of all functions in VP8DSPContext. Based on initial work by Rob Clark. Signed-off-by: Mans Rullgard --- libavcodec/arm/Makefile | 3 + libavcodec/arm/vp8dsp_init_arm.c | 163 +++ libavcodec/arm/vp8dsp_neon.S | 1910 ++++++++++++++++++++++++++++++ libavcodec/vp8dsp.c | 2 + libavcodec/vp8dsp.h | 1 + 5 files changed, 2079 insertions(+) create mode 100644 libavcodec/arm/vp8dsp_init_arm.c create mode 100644 libavcodec/arm/vp8dsp_neon.S diff --git a/libavcodec/arm/Makefile b/libavcodec/arm/Makefile index 014456ee32..15269ea676 100644 --- a/libavcodec/arm/Makefile +++ b/libavcodec/arm/Makefile @@ -2,6 +2,7 @@ OBJS-$(CONFIG_DCA_DECODER) += arm/dcadsp_init_arm.o \ OBJS-$(CONFIG_VP5_DECODER) += arm/vp56dsp_init_arm.o OBJS-$(CONFIG_VP6_DECODER) += arm/vp56dsp_init_arm.o +OBJS-$(CONFIG_VP8_DECODER) += arm/vp8dsp_init_arm.o OBJS-$(CONFIG_H264DSP) += arm/h264dsp_init_arm.o OBJS-$(CONFIG_H264PRED) += arm/h264pred_init_arm.o @@ -54,6 +55,8 @@ NEON-OBJS-$(CONFIG_VP5_DECODER) += arm/vp56dsp_neon.o \ NEON-OBJS-$(CONFIG_VP6_DECODER) += arm/vp56dsp_neon.o \ arm/vp3dsp_neon.o \ +NEON-OBJS-$(CONFIG_VP8_DECODER) += arm/vp8dsp_neon.o + OBJS-$(HAVE_NEON) += arm/dsputil_init_neon.o \ arm/dsputil_neon.o \ arm/fmtconvert_neon.o \ diff --git a/libavcodec/arm/vp8dsp_init_arm.c b/libavcodec/arm/vp8dsp_init_arm.c new file mode 100644 index 0000000000..c970ca548c --- /dev/null +++ b/libavcodec/arm/vp8dsp_init_arm.c @@ -0,0 +1,163 @@ +/** + * This file is part of FFmpeg. + * + * FFmpeg is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; either + * version 2.1 of the License, or (at your option) any later version. + * + * FFmpeg is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with FFmpeg; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA + */ + +#include +#include "libavcodec/vp8dsp.h" + +void ff_vp8_luma_dc_wht_neon(DCTELEM block[4][4][16], DCTELEM dc[16]); +void ff_vp8_luma_dc_wht_dc_neon(DCTELEM block[4][4][16], DCTELEM dc[16]); + +void ff_vp8_idct_add_neon(uint8_t *dst, DCTELEM block[16], int stride); +void ff_vp8_idct_dc_add_neon(uint8_t *dst, DCTELEM block[16], int stride); +void ff_vp8_idct_dc_add4y_neon(uint8_t *dst, DCTELEM block[4][16], int stride); +void ff_vp8_idct_dc_add4uv_neon(uint8_t *dst, DCTELEM block[4][16], int stride); + +void ff_vp8_v_loop_filter16_neon(uint8_t *dst, int stride, + int flim_E, int flim_I, int hev_thresh); +void ff_vp8_h_loop_filter16_neon(uint8_t *dst, int stride, + int flim_E, int flim_I, int hev_thresh); +void ff_vp8_v_loop_filter8uv_neon(uint8_t *dstU, uint8_t *dstV, int stride, + int flim_E, int flim_I, int hev_thresh); +void ff_vp8_h_loop_filter8uv_neon(uint8_t *dstU, uint8_t *dstV, int stride, + int flim_E, int flim_I, int hev_thresh); + +void ff_vp8_v_loop_filter16_inner_neon(uint8_t *dst, int stride, + int flim_E, int flim_I, int hev_thresh); +void ff_vp8_h_loop_filter16_inner_neon(uint8_t *dst, int stride, + int flim_E, int flim_I, int hev_thresh); +void ff_vp8_v_loop_filter8uv_inner_neon(uint8_t *dstU, uint8_t *dstV, + int stride, int flim_E, int flim_I, + int hev_thresh); +void ff_vp8_h_loop_filter8uv_inner_neon(uint8_t *dstU, uint8_t *dstV, + int stride, int flim_E, int flim_I, + int hev_thresh); + +void ff_vp8_v_loop_filter16_simple_neon(uint8_t *dst, int stride, int flim); +void ff_vp8_h_loop_filter16_simple_neon(uint8_t *dst, int stride, int flim); + + +#define VP8_MC(n) \ + void ff_put_vp8_##n##_neon(uint8_t *dst, int dststride, \ + uint8_t *src, int srcstride, \ + int h, int x, int y) + +#define VP8_EPEL(w) \ + VP8_MC(pixels ## w); \ + VP8_MC(epel ## w ## _h4); \ + VP8_MC(epel ## w ## _h6); \ + VP8_MC(epel ## w ## _v4); \ + VP8_MC(epel ## w ## _h4v4); \ + VP8_MC(epel ## w ## _h6v4); \ + VP8_MC(epel ## w ## _v6); \ + VP8_MC(epel ## w ## _h4v6); \ + VP8_MC(epel ## w ## _h6v6) + +VP8_EPEL(16); +VP8_EPEL(8); +VP8_EPEL(4); + +VP8_MC(bilin16_h); +VP8_MC(bilin16_v); +VP8_MC(bilin16_hv); +VP8_MC(bilin8_h); +VP8_MC(bilin8_v); +VP8_MC(bilin8_hv); +VP8_MC(bilin4_h); +VP8_MC(bilin4_v); +VP8_MC(bilin4_hv); + +av_cold void ff_vp8dsp_init_arm(VP8DSPContext *dsp) +{ + if (HAVE_NEON) { + dsp->vp8_luma_dc_wht = ff_vp8_luma_dc_wht_neon; + dsp->vp8_luma_dc_wht_dc = ff_vp8_luma_dc_wht_dc_neon; + + dsp->vp8_idct_add = ff_vp8_idct_add_neon; + dsp->vp8_idct_dc_add = ff_vp8_idct_dc_add_neon; + dsp->vp8_idct_dc_add4y = ff_vp8_idct_dc_add4y_neon; + dsp->vp8_idct_dc_add4uv = ff_vp8_idct_dc_add4uv_neon; + + dsp->vp8_v_loop_filter16y = ff_vp8_v_loop_filter16_neon; + dsp->vp8_h_loop_filter16y = ff_vp8_h_loop_filter16_neon; + dsp->vp8_v_loop_filter8uv = ff_vp8_v_loop_filter8uv_neon; + dsp->vp8_h_loop_filter8uv = ff_vp8_h_loop_filter8uv_neon; + + dsp->vp8_v_loop_filter16y_inner = ff_vp8_v_loop_filter16_inner_neon; + dsp->vp8_h_loop_filter16y_inner = ff_vp8_h_loop_filter16_inner_neon; + dsp->vp8_v_loop_filter8uv_inner = ff_vp8_v_loop_filter8uv_inner_neon; + dsp->vp8_h_loop_filter8uv_inner = ff_vp8_h_loop_filter8uv_inner_neon; + + dsp->vp8_v_loop_filter_simple = ff_vp8_v_loop_filter16_simple_neon; + dsp->vp8_h_loop_filter_simple = ff_vp8_h_loop_filter16_simple_neon; + + dsp->put_vp8_epel_pixels_tab[0][0][0] = ff_put_vp8_pixels16_neon; + dsp->put_vp8_epel_pixels_tab[0][0][2] = ff_put_vp8_epel16_h6_neon; + dsp->put_vp8_epel_pixels_tab[0][2][0] = ff_put_vp8_epel16_v6_neon; + dsp->put_vp8_epel_pixels_tab[0][2][2] = ff_put_vp8_epel16_h6v6_neon; + + dsp->put_vp8_epel_pixels_tab[1][0][0] = ff_put_vp8_pixels8_neon; + dsp->put_vp8_epel_pixels_tab[1][0][1] = ff_put_vp8_epel8_h4_neon; + dsp->put_vp8_epel_pixels_tab[1][0][2] = ff_put_vp8_epel8_h6_neon; + dsp->put_vp8_epel_pixels_tab[1][1][0] = ff_put_vp8_epel8_v4_neon; + dsp->put_vp8_epel_pixels_tab[1][1][1] = ff_put_vp8_epel8_h4v4_neon; + dsp->put_vp8_epel_pixels_tab[1][1][2] = ff_put_vp8_epel8_h6v4_neon; + dsp->put_vp8_epel_pixels_tab[1][2][0] = ff_put_vp8_epel8_v6_neon; + dsp->put_vp8_epel_pixels_tab[1][2][1] = ff_put_vp8_epel8_h4v6_neon; + dsp->put_vp8_epel_pixels_tab[1][2][2] = ff_put_vp8_epel8_h6v6_neon; + + dsp->put_vp8_epel_pixels_tab[2][0][0] = ff_put_vp8_pixels4_neon; + dsp->put_vp8_epel_pixels_tab[2][0][1] = ff_put_vp8_epel4_h4_neon; + dsp->put_vp8_epel_pixels_tab[2][0][2] = ff_put_vp8_epel4_h6_neon; + dsp->put_vp8_epel_pixels_tab[2][1][0] = ff_put_vp8_epel4_v4_neon; + dsp->put_vp8_epel_pixels_tab[2][1][1] = ff_put_vp8_epel4_h4v4_neon; + dsp->put_vp8_epel_pixels_tab[2][1][2] = ff_put_vp8_epel4_h6v4_neon; + dsp->put_vp8_epel_pixels_tab[2][2][0] = ff_put_vp8_epel4_v6_neon; + dsp->put_vp8_epel_pixels_tab[2][2][1] = ff_put_vp8_epel4_h4v6_neon; + dsp->put_vp8_epel_pixels_tab[2][2][2] = ff_put_vp8_epel4_h6v6_neon; + + dsp->put_vp8_bilinear_pixels_tab[0][0][0] = ff_put_vp8_pixels16_neon; + dsp->put_vp8_bilinear_pixels_tab[0][0][1] = ff_put_vp8_bilin16_h_neon; + dsp->put_vp8_bilinear_pixels_tab[0][0][2] = ff_put_vp8_bilin16_h_neon; + dsp->put_vp8_bilinear_pixels_tab[0][1][0] = ff_put_vp8_bilin16_v_neon; + dsp->put_vp8_bilinear_pixels_tab[0][1][1] = ff_put_vp8_bilin16_hv_neon; + dsp->put_vp8_bilinear_pixels_tab[0][1][2] = ff_put_vp8_bilin16_hv_neon; + dsp->put_vp8_bilinear_pixels_tab[0][2][0] = ff_put_vp8_bilin16_v_neon; + dsp->put_vp8_bilinear_pixels_tab[0][2][1] = ff_put_vp8_bilin16_hv_neon; + dsp->put_vp8_bilinear_pixels_tab[0][2][2] = ff_put_vp8_bilin16_hv_neon; + + dsp->put_vp8_bilinear_pixels_tab[1][0][0] = ff_put_vp8_pixels8_neon; + dsp->put_vp8_bilinear_pixels_tab[1][0][1] = ff_put_vp8_bilin8_h_neon; + dsp->put_vp8_bilinear_pixels_tab[1][0][2] = ff_put_vp8_bilin8_h_neon; + dsp->put_vp8_bilinear_pixels_tab[1][1][0] = ff_put_vp8_bilin8_v_neon; + dsp->put_vp8_bilinear_pixels_tab[1][1][1] = ff_put_vp8_bilin8_hv_neon; + dsp->put_vp8_bilinear_pixels_tab[1][1][2] = ff_put_vp8_bilin8_hv_neon; + dsp->put_vp8_bilinear_pixels_tab[1][2][0] = ff_put_vp8_bilin8_v_neon; + dsp->put_vp8_bilinear_pixels_tab[1][2][1] = ff_put_vp8_bilin8_hv_neon; + dsp->put_vp8_bilinear_pixels_tab[1][2][2] = ff_put_vp8_bilin8_hv_neon; + + dsp->put_vp8_bilinear_pixels_tab[2][0][0] = ff_put_vp8_pixels4_neon; + dsp->put_vp8_bilinear_pixels_tab[2][0][1] = ff_put_vp8_bilin4_h_neon; + dsp->put_vp8_bilinear_pixels_tab[2][0][2] = ff_put_vp8_bilin4_h_neon; + dsp->put_vp8_bilinear_pixels_tab[2][1][0] = ff_put_vp8_bilin4_v_neon; + dsp->put_vp8_bilinear_pixels_tab[2][1][1] = ff_put_vp8_bilin4_hv_neon; + dsp->put_vp8_bilinear_pixels_tab[2][1][2] = ff_put_vp8_bilin4_hv_neon; + dsp->put_vp8_bilinear_pixels_tab[2][2][0] = ff_put_vp8_bilin4_v_neon; + dsp->put_vp8_bilinear_pixels_tab[2][2][1] = ff_put_vp8_bilin4_hv_neon; + dsp->put_vp8_bilinear_pixels_tab[2][2][2] = ff_put_vp8_bilin4_hv_neon; + } +} diff --git a/libavcodec/arm/vp8dsp_neon.S b/libavcodec/arm/vp8dsp_neon.S new file mode 100644 index 0000000000..01c39593a0 --- /dev/null +++ b/libavcodec/arm/vp8dsp_neon.S @@ -0,0 +1,1910 @@ +/** + * VP8 NEON optimisations + * + * Copyright (c) 2010 Rob Clark + * Copyright (c) 2011 Mans Rullgard + * + * This file is part of FFmpeg. + * + * FFmpeg is free software; you can redistribute it and/or + * modify it under the terms of the GNU Lesser General Public + * License as published by the Free Software Foundation; either + * version 2.1 of the License, or (at your option) any later version. + * + * FFmpeg is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * Lesser General Public License for more details. + * + * You should have received a copy of the GNU Lesser General Public + * License along with FFmpeg; if not, write to the Free Software + * Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA + */ + +#include "asm.S" + +function ff_vp8_luma_dc_wht_neon, export=1 + vld1.16 {q0-q1}, [r1,:128] + vmov.i16 q15, #0 + + vadd.i16 d4, d0, d3 + vadd.i16 d6, d1, d2 + vst1.16 {q15}, [r1,:128]! + vsub.i16 d7, d1, d2 + vsub.i16 d5, d0, d3 + vst1.16 {q15}, [r1,:128] + vadd.i16 q0, q2, q3 + vsub.i16 q1, q2, q3 + + vmov.i16 q8, #3 + + vtrn.32 d0, d2 + vtrn.32 d1, d3 + vtrn.16 d0, d1 + vtrn.16 d2, d3 + + vadd.i16 d0, d0, d16 + + vadd.i16 d4, d0, d3 + vadd.i16 d6, d1, d2 + vsub.i16 d7, d1, d2 + vsub.i16 d5, d0, d3 + vadd.i16 q0, q2, q3 + vsub.i16 q1, q2, q3 + + vshr.s16 q0, q0, #3 + vshr.s16 q1, q1, #3 + + mov r3, #32 + vst1.16 {d0[0]}, [r0,:16], r3 + vst1.16 {d1[0]}, [r0,:16], r3 + vst1.16 {d2[0]}, [r0,:16], r3 + vst1.16 {d3[0]}, [r0,:16], r3 + vst1.16 {d0[1]}, [r0,:16], r3 + vst1.16 {d1[1]}, [r0,:16], r3 + vst1.16 {d2[1]}, [r0,:16], r3 + vst1.16 {d3[1]}, [r0,:16], r3 + vst1.16 {d0[2]}, [r0,:16], r3 + vst1.16 {d1[2]}, [r0,:16], r3 + vst1.16 {d2[2]}, [r0,:16], r3 + vst1.16 {d3[2]}, [r0,:16], r3 + vst1.16 {d0[3]}, [r0,:16], r3 + vst1.16 {d1[3]}, [r0,:16], r3 + vst1.16 {d2[3]}, [r0,:16], r3 + vst1.16 {d3[3]}, [r0,:16], r3 + + bx lr +endfunc + +function ff_vp8_luma_dc_wht_dc_neon, export=1 + ldrsh r2, [r1] + mov r3, #0 + add r2, r2, #3 + strh r3, [r1] + asr r2, r2, #3 + .rept 16 + strh r2, [r0], #32 + .endr + bx lr +endfunc + +function ff_vp8_idct_add_neon, export=1 + vld1.16 {q0-q1}, [r1,:128] + movw r3, #20091 + movt r3, #35468/2 + vdup.32 d4, r3 + + vmull.s16 q12, d1, d4[0] + vmull.s16 q13, d3, d4[0] + vqdmulh.s16 d20, d1, d4[1] + vqdmulh.s16 d23, d3, d4[1] + vshrn.s32 d21, q12, #16 + vshrn.s32 d22, q13, #16 + vadd.s16 d21, d21, d1 + vadd.s16 d22, d22, d3 + + vadd.s16 d16, d0, d2 + vsub.s16 d17, d0, d2 + vadd.s16 d18, d21, d23 + vsub.s16 d19, d20, d22 + vadd.s16 q0, q8, q9 + vsub.s16 q1, q8, q9 + + vtrn.32 d0, d3 + vtrn.32 d1, d2 + vtrn.16 d0, d1 + vtrn.16 d3, d2 + + vmov.i16 q15, #0 + vmull.s16 q12, d1, d4[0] + vst1.16 {q15}, [r1,:128]! + vmull.s16 q13, d2, d4[0] + vst1.16 {q15}, [r1,:128] + vqdmulh.s16 d21, d1, d4[1] + vqdmulh.s16 d23, d2, d4[1] + vshrn.s32 d20, q12, #16 + vshrn.s32 d22, q13, #16 + vadd.i16 d20, d20, d1 + vadd.i16 d22, d22, d2 + + vadd.i16 d16, d0, d3 + vsub.i16 d17, d0, d3 + vadd.i16 d18, d20, d23 + vld1.32 {d20[]}, [r0,:32], r2 + vsub.i16 d19, d21, d22 + vld1.32 {d22[]}, [r0,:32], r2 + vadd.s16 q0, q8, q9 + vld1.32 {d23[]}, [r0,:32], r2 + vsub.s16 q1, q8, q9 + vld1.32 {d21[]}, [r0,:32], r2 + vrshr.s16 q0, q0, #3 + vtrn.32 q10, q11 + vrshr.s16 q1, q1, #3 + + sub r0, r0, r2, lsl #2 + + vtrn.32 d0, d3 + vtrn.32 d1, d2 + vtrn.16 d0, d1 + vtrn.16 d3, d2 + + vaddw.u8 q0, q0, d20 + vaddw.u8 q1, q1, d21 + 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[1]}, [r0,:32], r2 + vst1.32 {d1[0]}, [r0,:32], r2 + + bx lr +endfunc + +function ff_vp8_idct_dc_add_neon, export=1 + mov r3, #0 + ldrsh r12, [r1] + strh r3, [r1] + vdup.16 q1, r12 + vrshr.s16 q1, q1, #3 + vld1.32 {d0[]}, [r0,:32], r2 + vld1.32 {d1[]}, [r0,:32], r2 + vld1.32 {d0[1]}, [r0,:32], r2 + vld1.32 {d1[1]}, [r0,:32], r2 + vaddw.u8 q2, q1, d0 + vaddw.u8 q3, q1, d1 + sub r0, r0, r2, lsl #2 + vqmovun.s16 d0, q2 + vqmovun.s16 d1, q3 + vst1.32 {d0[0]}, [r0,:32], r2 + vst1.32 {d1[0]}, [r0,:32], r2 + vst1.32 {d0[1]}, [r0,:32], r2 + vst1.32 {d1[1]}, [r0,:32], r2 + bx lr +endfunc + +function ff_vp8_idct_dc_add4uv_neon, export=1 + vmov.i16 d0, #0 + mov r3, #32 + vld1.16 {d16[]}, [r1,:16] + vst1.16 {d0[0]}, [r1,:16], r3 + vld1.16 {d17[]}, [r1,:16] + vst1.16 {d0[0]}, [r1,:16], r3 + vld1.16 {d18[]}, [r1,:16] + vst1.16 {d0[0]}, [r1,:16], r3 + vld1.16 {d19[]}, [r1,:16] + vst1.16 {d0[0]}, [r1,:16], r3 + mov r3, r0 + vrshr.s16 q8, q8, #3 @ dc >>= 3 + vld1.8 {d0}, [r0,:64], r2 + vrshr.s16 q9, q9, #3 + vld1.8 {d1}, [r0,:64], r2 + vaddw.u8 q10, q8, d0 + vld1.8 {d2}, [r0,:64], r2 + vaddw.u8 q0, q8, d1 + vld1.8 {d3}, [r0,:64], r2 + vaddw.u8 q11, q8, d2 + vld1.8 {d4}, [r0,:64], r2 + vaddw.u8 q1, q8, d3 + vld1.8 {d5}, [r0,:64], r2 + vaddw.u8 q12, q9, d4 + vld1.8 {d6}, [r0,:64], r2 + vaddw.u8 q2, q9, d5 + vld1.8 {d7}, [r0,:64], r2 + vaddw.u8 q13, q9, d6 + vqmovun.s16 d20, q10 + vaddw.u8 q3, q9, d7 + vqmovun.s16 d21, q0 + vqmovun.s16 d22, q11 + vst1.8 {d20}, [r3,:64], r2 + vqmovun.s16 d23, q1 + vst1.8 {d21}, [r3,:64], r2 + vqmovun.s16 d24, q12 + vst1.8 {d22}, [r3,:64], r2 + vqmovun.s16 d25, q2 + vst1.8 {d23}, [r3,:64], r2 + vqmovun.s16 d26, q13 + vst1.8 {d24}, [r3,:64], r2 + vqmovun.s16 d27, q3 + vst1.8 {d25}, [r3,:64], r2 + vst1.8 {d26}, [r3,:64], r2 + vst1.8 {d27}, [r3,:64], r2 + + bx lr +endfunc + +function ff_vp8_idct_dc_add4y_neon, export=1 + vmov.i16 d0, #0 + mov r3, #32 + vld1.16 {d16[]}, [r1,:16] + vst1.16 {d0[0]}, [r1,:16], r3 + vld1.16 {d17[]}, [r1,:16] + vst1.16 {d0[0]}, [r1,:16], r3 + vld1.16 {d18[]}, [r1,:16] + vst1.16 {d0[0]}, [r1,:16], r3 + vld1.16 {d19[]}, [r1,:16] + vst1.16 {d0[0]}, [r1,:16], r3 + vrshr.s16 q8, q8, #3 @ dc >>= 3 + vld1.8 {q0}, [r0,:128], r2 + vrshr.s16 q9, q9, #3 + vld1.8 {q1}, [r0,:128], r2 + vaddw.u8 q10, q8, d0 + vld1.8 {q2}, [r0,:128], r2 + vaddw.u8 q0, q9, d1 + vld1.8 {q3}, [r0,:128], r2 + vaddw.u8 q11, q8, d2 + vaddw.u8 q1, q9, d3 + vaddw.u8 q12, q8, d4 + vaddw.u8 q2, q9, d5 + vaddw.u8 q13, q8, d6 + vaddw.u8 q3, q9, d7 + sub r0, r0, r2, lsl #2 + vqmovun.s16 d20, q10 + vqmovun.s16 d21, q0 + vqmovun.s16 d22, q11 + vqmovun.s16 d23, q1 + vqmovun.s16 d24, q12 + vst1.8 {q10}, [r0,:128], r2 + vqmovun.s16 d25, q2 + vst1.8 {q11}, [r0,:128], r2 + vqmovun.s16 d26, q13 + vst1.8 {q12}, [r0,:128], r2 + vqmovun.s16 d27, q3 + vst1.8 {q13}, [r0,:128], r2 + + bx lr +endfunc + +@ Register layout: +@ P3..Q3 -> q0..q7 +@ flim_E -> q14 +@ flim_I -> q15 +@ hev_thresh -> r12 +@ +.macro vp8_loop_filter, inner=0, simple=0 + .if \simple + vabd.u8 q9, q3, q4 @ abs(P0-Q0) + vabd.u8 q15, q2, q5 @ abs(P1-Q1) + vqadd.u8 q9, q9, q9 @ abs(P0-Q0) * 2 + vshr.u8 q10, q15, #1 @ abs(P1-Q1) / 2 + vqadd.u8 q11, q9, q10 @ (abs(P0-Q0)*2) + (abs(P1-Q1)/2) + vmov.i8 q13, #0x80 + vcle.u8 q8, q11, q14 @ (abs(P0-Q0)*2) + (abs(P1-Q1)/2) <= flim + .else + @ calculate hev and normal_limit: + vabd.u8 q12, q2, q3 @ abs(P1-P0) + vabd.u8 q13, q5, q4 @ abs(Q1-Q0) + vabd.u8 q10, q0, q1 @ abs(P3-P2) + vabd.u8 q11, q1, q2 @ abs(P2-P1) + vcle.u8 q8, q12, q15 @ abs(P1-P0) <= flim_I + vcle.u8 q9, q13, q15 @ abs(Q1-Q0) <= flim_I + vcle.u8 q10, q10, q15 @ abs(P3-P2) <= flim_I + vcle.u8 q11, q11, q15 @ abs(P2-P1) <= flim_I + vand q8, q8, q9 + vabd.u8 q9, q7, q6 @ abs(Q3-Q2) + vand q8, q8, q11 + vabd.u8 q11, q6, q5 @ abs(Q2-Q1) + vand q8, q8, q10 + vcle.u8 q10, q9, q15 @ abs(Q3-Q2) <= flim_I + vcle.u8 q11, q11, q15 @ abs(Q2-Q1) <= flim_I + vabd.u8 q9, q3, q4 @ abs(P0-Q0) + vabd.u8 q15, q2, q5 @ abs(P1-Q1) + vand q8, q8, q10 + vqadd.u8 q9, q9, q9 @ abs(P0-Q0) * 2 + vand q8, q8, q11 + vshr.u8 q10, q15, #1 @ abs(P1-Q1) / 2 + vdup.8 q15, r12 @ hev_thresh + vqadd.u8 q11, q9, q10 @ (abs(P0-Q0)*2) + (abs(P1-Q1)/2) + vcgt.u8 q12, q12, q15 @ abs(P1-P0) > hev_thresh + vcle.u8 q11, q11, q14 @ (abs(P0-Q0)*2) + (abs(P1-Q1)/2) <= flim_E + vcgt.u8 q14, q13, q15 @ abs(Q1-Q0) > hev_thresh + vand q8, q8, q11 + vmov.i8 q13, #0x80 + vorr q9, q12, q14 + .endif + + @ at this point: + @ q8: normal_limit + @ q9: hev + + @ convert to signed value: + veor q3, q3, q13 @ PS0 = P0 ^ 0x80 + veor q4, q4, q13 @ QS0 = Q0 ^ 0x80 + + vmov.i16 q12, #3 + vsubl.s8 q10, d8, d6 @ QS0 - PS0 + vsubl.s8 q11, d9, d7 @ (widened to 16bit) + veor q2, q2, q13 @ PS1 = P1 ^ 0x80 + veor q5, q5, q13 @ QS1 = Q1 ^ 0x80 + vmul.i16 q10, q10, q12 @ w = 3 * (QS0 - PS0) + vmul.i16 q11, q11, q12 + + vqsub.s8 q12, q2, q5 @ clamp(PS1-QS1) + vmov.i8 q14, #4 + vmov.i8 q15, #3 + .if \inner + vand q12, q12, q9 @ if(hev) w += clamp(PS1-QS1) + .endif + vaddw.s8 q10, q10, d24 @ w += clamp(PS1-QS1) + vaddw.s8 q11, q11, d25 + vqmovn.s16 d20, q10 @ narrow result back into q10 + vqmovn.s16 d21, q11 + .if !\inner && !\simple + veor q1, q1, q13 @ PS2 = P2 ^ 0x80 + veor q6, q6, q13 @ QS2 = Q2 ^ 0x80 + .endif + vand q10, q10, q8 @ w &= normal_limit + + @ registers used at this point.. + @ q0 -> P3 (don't corrupt) + @ q1-q6 -> PS2-QS2 + @ q7 -> Q3 (don't corrupt) + @ q9 -> hev + @ q10 -> w + @ q13 -> #0x80 + @ q14 -> #4 + @ q15 -> #3 + @ q8, q11, q12 -> unused + + @ filter_common: is4tap==1 + @ c1 = clamp(w + 4) >> 3; + @ c2 = clamp(w + 3) >> 3; + @ Q0 = s2u(QS0 - c1); + @ P0 = s2u(PS0 + c2); + + .if \simple + vqadd.s8 q11, q10, q14 @ c1 = clamp((w&hev)+4) + vqadd.s8 q12, q10, q15 @ c2 = clamp((w&hev)+3) + vshr.s8 q11, q11, #3 @ c1 >>= 3 + vshr.s8 q12, q12, #3 @ c2 >>= 3 + vqsub.s8 q4, q4, q11 @ QS0 = clamp(QS0-c1) + vqadd.s8 q3, q3, q12 @ PS0 = clamp(PS0+c2) + veor q4, q4, q13 @ Q0 = QS0 ^ 0x80 + veor q3, q3, q13 @ P0 = PS0 ^ 0x80 + veor q5, q5, q13 @ Q1 = QS1 ^ 0x80 + veor q2, q2, q13 @ P1 = PS1 ^ 0x80 + .elseif \inner + @ the !is4tap case of filter_common, only used for inner blocks + @ c3 = ((c1&~hev) + 1) >> 1; + @ Q1 = s2u(QS1 - c3); + @ P1 = s2u(PS1 + c3); + vqadd.s8 q11, q10, q14 @ c1 = clamp((w&hev)+4) + vqadd.s8 q12, q10, q15 @ c2 = clamp((w&hev)+3) + vshr.s8 q11, q11, #3 @ c1 >>= 3 + vshr.s8 q12, q12, #3 @ c2 >>= 3 + vqsub.s8 q4, q4, q11 @ QS0 = clamp(QS0-c1) + vqadd.s8 q3, q3, q12 @ PS0 = clamp(PS0+c2) + vbic q11, q11, q9 @ c1 & ~hev + veor q4, q4, q13 @ Q0 = QS0 ^ 0x80 + vrshr.s8 q11, q11, #1 @ c3 >>= 1 + veor q3, q3, q13 @ P0 = PS0 ^ 0x80 + vqsub.s8 q5, q5, q11 @ QS1 = clamp(QS1-c3) + vqadd.s8 q2, q2, q11 @ PS1 = clamp(PS1+c3) + veor q5, q5, q13 @ Q1 = QS1 ^ 0x80 + veor q2, q2, q13 @ P1 = PS1 ^ 0x80 + .else + vand q12, q10, q9 @ w & hev + vqadd.s8 q11, q12, q14 @ c1 = clamp((w&hev)+4) + vqadd.s8 q12, q12, q15 @ c2 = clamp((w&hev)+3) + vshr.s8 q11, q11, #3 @ c1 >>= 3 + vshr.s8 q12, q12, #3 @ c2 >>= 3 + vbic q10, q10, q9 @ w &= ~hev + vqsub.s8 q4, q4, q11 @ QS0 = clamp(QS0-c1) + vqadd.s8 q3, q3, q12 @ PS0 = clamp(PS0+c2) + + @ filter_mbedge: + @ a = clamp((27*w + 63) >> 7); + @ Q0 = s2u(QS0 - a); + @ P0 = s2u(PS0 + a); + @ a = clamp((18*w + 63) >> 7); + @ Q1 = s2u(QS1 - a); + @ P1 = s2u(PS1 + a); + @ a = clamp((9*w + 63) >> 7); + @ Q2 = s2u(QS2 - a); + @ P2 = s2u(PS2 + a); + vmov.i16 q9, #63 + vshll.s8 q14, d20, #3 + vshll.s8 q15, d21, #3 + vaddw.s8 q14, q14, d20 + vaddw.s8 q15, q15, d21 + vadd.s16 q8, q9, q14 + vadd.s16 q9, q9, q15 @ 9*w + 63 + vadd.s16 q11, q8, q14 + vadd.s16 q12, q9, q15 @ 18*w + 63 + vadd.s16 q14, q11, q14 + vadd.s16 q15, q12, q15 @ 27*w + 63 + vqshrn.s16 d16, q8, #7 + vqshrn.s16 d17, q9, #7 @ clamp(( 9*w + 63)>>7) + vqshrn.s16 d22, q11, #7 + vqshrn.s16 d23, q12, #7 @ clamp((18*w + 63)>>7) + vqshrn.s16 d28, q14, #7 + vqshrn.s16 d29, q15, #7 @ clamp((27*w + 63)>>7) + vqadd.s8 q1, q1, q8 @ PS2 = clamp(PS2+a) + vqsub.s8 q6, q6, q8 @ QS2 = clamp(QS2-a) + vqadd.s8 q2, q2, q11 @ PS1 = clamp(PS1+a) + vqsub.s8 q5, q5, q11 @ QS1 = clamp(QS1-a) + vqadd.s8 q3, q3, q14 @ PS0 = clamp(PS0+a) + vqsub.s8 q4, q4, q14 @ QS0 = clamp(QS0-a) + veor q3, q3, q13 @ P0 = PS0 ^ 0x80 + veor q4, q4, q13 @ Q0 = QS0 ^ 0x80 + veor q2, q2, q13 @ P1 = PS1 ^ 0x80 + veor q5, q5, q13 @ Q1 = QS1 ^ 0x80 + veor q1, q1, q13 @ P2 = PS2 ^ 0x80 + veor q6, q6, q13 @ Q2 = QS2 ^ 0x80 + .endif +.endm + +.macro transpose8x16matrix + vtrn.32 q0, q4 + vtrn.32 q1, q5 + vtrn.32 q2, q6 + vtrn.32 q3, q7 + + vtrn.16 q0, q2 + vtrn.16 q1, q3 + vtrn.16 q4, q6 + vtrn.16 q5, q7 + + vtrn.8 q0, q1 + vtrn.8 q2, q3 + vtrn.8 q4, q5 + vtrn.8 q6, q7 +.endm + +.macro vp8_v_loop_filter16 name, inner=0, simple=0 +function ff_vp8_v_loop_filter16\name\()_neon, export=1 + vpush {q4-q7} + sub r0, r0, r1, lsl #1+!\simple + + @ Load pixels: + .if !\simple + ldr r12, [sp, #64] @ hev_thresh + vld1.8 {q0}, [r0,:128], r1 @ P3 + vld1.8 {q1}, [r0,:128], r1 @ P2 + .endif + vld1.8 {q2}, [r0,:128], r1 @ P1 + vld1.8 {q3}, [r0,:128], r1 @ P0 + vld1.8 {q4}, [r0,:128], r1 @ Q0 + vld1.8 {q5}, [r0,:128], r1 @ Q1 + .if !\simple + vld1.8 {q6}, [r0,:128], r1 @ Q2 + vld1.8 {q7}, [r0,:128] @ Q3 + vdup.8 q15, r3 @ flim_I + .endif + vdup.8 q14, r2 @ flim_E + + vp8_loop_filter inner=\inner, simple=\simple + + @ back up to P2: dst -= stride * 6 + sub r0, r0, r1, lsl #2 + .if !\simple + sub r0, r0, r1, lsl #1 + + @ Store pixels: + vst1.8 {q1}, [r0,:128], r1 @ P2 + .endif + vst1.8 {q2}, [r0,:128], r1 @ P1 + vst1.8 {q3}, [r0,:128], r1 @ P0 + vst1.8 {q4}, [r0,:128], r1 @ Q0 + vst1.8 {q5}, [r0,:128], r1 @ Q1 + .if !\simple + vst1.8 {q6}, [r0,:128] @ Q2 + .endif + + vpop {q4-q7} + bx lr +endfunc +.endm + +vp8_v_loop_filter16 +vp8_v_loop_filter16 _inner, inner=1 +vp8_v_loop_filter16 _simple, simple=1 + +.macro vp8_v_loop_filter8uv name, inner=0 +function ff_vp8_v_loop_filter8uv\name\()_neon, export=1 + vpush {q4-q7} + sub r0, r0, r2, lsl #2 + sub r1, r1, r2, lsl #2 + ldr r12, [sp, #64] @ flim_I + + @ Load pixels: + vld1.8 {d0}, [r0,:64], r2 @ P3 + vld1.8 {d1}, [r1,:64], r2 @ P3 + vld1.8 {d2}, [r0,:64], r2 @ P2 + vld1.8 {d3}, [r1,:64], r2 @ P2 + vld1.8 {d4}, [r0,:64], r2 @ P1 + vld1.8 {d5}, [r1,:64], r2 @ P1 + vld1.8 {d6}, [r0,:64], r2 @ P0 + vld1.8 {d7}, [r1,:64], r2 @ P0 + vld1.8 {d8}, [r0,:64], r2 @ Q0 + vld1.8 {d9}, [r1,:64], r2 @ Q0 + vld1.8 {d10}, [r0,:64], r2 @ Q1 + vld1.8 {d11}, [r1,:64], r2 @ Q1 + vld1.8 {d12}, [r0,:64], r2 @ Q2 + vld1.8 {d13}, [r1,:64], r2 @ Q2 + vld1.8 {d14}, [r0,:64] @ Q3 + vld1.8 {d15}, [r1,:64] @ Q3 + + vdup.8 q14, r3 @ flim_E + vdup.8 q15, r12 @ flim_I + ldr r12, [sp, #68] @ hev_thresh + + vp8_loop_filter inner=\inner + + @ back up to P2: u,v -= stride * 6 + sub r0, r0, r2, lsl #2 + sub r1, r1, r2, lsl #2 + sub r0, r0, r2, lsl #1 + sub r1, r1, r2, lsl #1 + + @ Store pixels: + vst1.8 {d2}, [r0,:64], r2 @ P2 + vst1.8 {d3}, [r1,:64], r2 @ P2 + vst1.8 {d4}, [r0,:64], r2 @ P1 + vst1.8 {d5}, [r1,:64], r2 @ P1 + vst1.8 {d6}, [r0,:64], r2 @ P0 + vst1.8 {d7}, [r1,:64], r2 @ P0 + vst1.8 {d8}, [r0,:64], r2 @ Q0 + vst1.8 {d9}, [r1,:64], r2 @ Q0 + vst1.8 {d10}, [r0,:64], r2 @ Q1 + vst1.8 {d11}, [r1,:64], r2 @ Q1 + vst1.8 {d12}, [r0,:64] @ Q2 + vst1.8 {d13}, [r1,:64] @ Q2 + + vpop {q4-q7} + bx lr +endfunc +.endm + +vp8_v_loop_filter8uv +vp8_v_loop_filter8uv _inner, inner=1 + +.macro vp8_h_loop_filter16 name, inner=0, simple=0 +function ff_vp8_h_loop_filter16\name\()_neon, export=1 + vpush {q4-q7} + sub r0, r0, #4 + .if !\simple + ldr r12, [sp, #64] @ hev_thresh + .endif + + @ Load pixels: + vld1.8 {d0}, [r0], r1 @ load first 8-line src data + vld1.8 {d2}, [r0], r1 + vld1.8 {d4}, [r0], r1 + vld1.8 {d6}, [r0], r1 + vld1.8 {d8}, [r0], r1 + vld1.8 {d10}, [r0], r1 + vld1.8 {d12}, [r0], r1 + vld1.8 {d14}, [r0], r1 + vld1.8 {d1}, [r0], r1 @ load second 8-line src data + vld1.8 {d3}, [r0], r1 + vld1.8 {d5}, [r0], r1 + vld1.8 {d7}, [r0], r1 + vld1.8 {d9}, [r0], r1 + vld1.8 {d11}, [r0], r1 + vld1.8 {d13}, [r0], r1 + vld1.8 {d15}, [r0], r1 + + transpose8x16matrix + + vdup.8 q14, r2 @ flim_E + .if !\simple + vdup.8 q15, r3 @ flim_I + .endif + + vp8_loop_filter inner=\inner, simple=\simple + + sub r0, r0, r1, lsl #4 @ backup 16 rows + + transpose8x16matrix + + @ Store pixels: + vst1.8 {d0}, [r0], r1 + vst1.8 {d2}, [r0], r1 + vst1.8 {d4}, [r0], r1 + vst1.8 {d6}, [r0], r1 + vst1.8 {d8}, [r0], r1 + vst1.8 {d10}, [r0], r1 + vst1.8 {d12}, [r0], r1 + vst1.8 {d14}, [r0], r1 + vst1.8 {d1}, [r0], r1 + vst1.8 {d3}, [r0], r1 + vst1.8 {d5}, [r0], r1 + vst1.8 {d7}, [r0], r1 + vst1.8 {d9}, [r0], r1 + vst1.8 {d11}, [r0], r1 + vst1.8 {d13}, [r0], r1 + vst1.8 {d15}, [r0] + + vpop {q4-q7} + bx lr +endfunc +.endm + +vp8_h_loop_filter16 +vp8_h_loop_filter16 _inner, inner=1 +vp8_h_loop_filter16 _simple, simple=1 + +.macro vp8_h_loop_filter8uv name, inner=0 +function ff_vp8_h_loop_filter8uv\name\()_neon, export=1 + vpush {q4-q7} + sub r0, r0, #4 + sub r1, r1, #4 + ldr r12, [sp, #64] @ flim_I + + @ Load pixels: + vld1.8 {d0}, [r0], r2 @ load u + vld1.8 {d1}, [r1], r2 @ load v + vld1.8 {d2}, [r0], r2 + vld1.8 {d3}, [r1], r2 + vld1.8 {d4}, [r0], r2 + vld1.8 {d5}, [r1], r2 + vld1.8 {d6}, [r0], r2 + vld1.8 {d7}, [r1], r2 + vld1.8 {d8}, [r0], r2 + vld1.8 {d9}, [r1], r2 + vld1.8 {d10}, [r0], r2 + vld1.8 {d11}, [r1], r2 + vld1.8 {d12}, [r0], r2 + vld1.8 {d13}, [r1], r2 + vld1.8 {d14}, [r0], r2 + vld1.8 {d15}, [r1], r2 + + transpose8x16matrix + + vdup.8 q14, r3 @ flim_E + vdup.8 q15, r12 @ flim_I + ldr r12, [sp, #68] @ hev_thresh + + vp8_loop_filter inner=\inner + + sub r0, r0, r2, lsl #3 @ backup u 8 rows + sub r1, r1, r2, lsl #3 @ backup v 8 rows + + transpose8x16matrix + + @ Store pixels: + vst1.8 {d0}, [r0], r2 + vst1.8 {d1}, [r1], r2 + vst1.8 {d2}, [r0], r2 + vst1.8 {d3}, [r1], r2 + vst1.8 {d4}, [r0], r2 + vst1.8 {d5}, [r1], r2 + vst1.8 {d6}, [r0], r2 + vst1.8 {d7}, [r1], r2 + vst1.8 {d8}, [r0], r2 + vst1.8 {d9}, [r1], r2 + vst1.8 {d10}, [r0], r2 + vst1.8 {d11}, [r1], r2 + vst1.8 {d12}, [r0], r2 + vst1.8 {d13}, [r1], r2 + vst1.8 {d14}, [r0] + vst1.8 {d15}, [r1] + + vpop {q4-q7} + bx lr +endfunc +.endm + +vp8_h_loop_filter8uv +vp8_h_loop_filter8uv _inner, inner=1 + +function ff_put_vp8_pixels16_neon, export=1 + ldr r12, [sp, #0] @ h +1: + subs r12, r12, #4 + vld1.8 {q0}, [r2], r3 + vld1.8 {q1}, [r2], r3 + vld1.8 {q2}, [r2], r3 + vld1.8 {q3}, [r2], r3 + vst1.8 {q0}, [r0,:128], r1 + vst1.8 {q1}, [r0,:128], r1 + vst1.8 {q2}, [r0,:128], r1 + vst1.8 {q3}, [r0,:128], r1 + bgt 1b + bx lr +endfunc + +function ff_put_vp8_pixels8_neon, export=1 + ldr r12, [sp, #0] @ h +1: + subs r12, r12, #4 + vld1.8 {d0}, [r2], r3 + vld1.8 {d1}, [r2], r3 + vld1.8 {d2}, [r2], r3 + vld1.8 {d3}, [r2], r3 + vst1.8 {d0}, [r0,:64], r1 + vst1.8 {d1}, [r0,:64], r1 + vst1.8 {d2}, [r0,:64], r1 + vst1.8 {d3}, [r0,:64], r1 + bgt 1b + bx lr +endfunc + +function ff_put_vp8_pixels4_neon, export=1 + ldr r12, [sp, #0] @ h + push {r4-r6,lr} +1: + subs r12, r12, #4 + ldr r4, [r2], r3 + ldr r5, [r2], r3 + ldr r6, [r2], r3 + ldr lr, [r2], r3 + str r4, [r0], r1 + str r5, [r0], r1 + str r6, [r0], r1 + str lr, [r0], r1 + bgt 1b + pop {r4-r6,pc} +endfunc + +/* 4/6-tap 8th-pel MC */ + +.macro vp8_epel8_h6 d, a, b + vext.8 d27, \a, \b, #1 + vmovl.u8 q8, \a + vext.8 d28, \a, \b, #2 + vmovl.u8 q9, d27 + vext.8 d29, \a, \b, #3 + vmovl.u8 q10, d28 + vext.8 d30, \a, \b, #4 + vmovl.u8 q11, d29 + vext.8 d31, \a, \b, #5 + vmovl.u8 q12, d30 + vmul.u16 q10, q10, d0[2] + vmovl.u8 q13, d31 + vmul.u16 q11, q11, d0[3] + vmls.u16 q10, q9, d0[1] + vmls.u16 q11, q12, d1[0] + vmla.u16 q10, q8, d0[0] + vmla.u16 q11, q13, d1[1] + vqadd.s16 q11, q10, q11 + vqrshrun.s16 \d, q11, #7 +.endm + +.macro vp8_epel16_h6 d0, d1, s0, s1, s2, q0, q1 + vext.8 q14, \q0, \q1, #3 + vext.8 q15, \q0, \q1, #4 + vmovl.u8 q11, d28 + vmovl.u8 q14, d29 + vext.8 q3, \q0, \q1, #2 + vmovl.u8 q12, d30 + vmovl.u8 q15, d31 + vext.8 q8, \q0, \q1, #1 + vmovl.u8 q10, d6 + vmovl.u8 q3, d7 + vext.8 q2, \q0, \q1, #5 + vmovl.u8 q13, d4 + vmovl.u8 q2, d5 + vmovl.u8 q9, d16 + vmovl.u8 q8, d17 + vmul.u16 q11, q11, d0[3] + vmul.u16 q10, q10, d0[2] + vmul.u16 q3, q3, d0[2] + vmul.u16 q14, q14, d0[3] + vmls.u16 q11, q12, d1[0] + vmovl.u8 q12, \s0 + vmovl.u8 q1, \s1 + vmls.u16 q10, q9, d0[1] + vmls.u16 q3, q8, d0[1] + vmls.u16 q14, q15, d1[0] + vmla.u16 q10, q12, d0[0] + vmla.u16 q11, q13, d1[1] + vmla.u16 q3, q1, d0[0] + vmla.u16 q14, q2, d1[1] + vqadd.s16 q11, q10, q11 + vqadd.s16 q14, q3, q14 + vqrshrun.s16 \d0, q11, #7 + vqrshrun.s16 \d1, q14, #7 +.endm + +.macro vp8_epel8_v6 d0, s0, s1, s2, s3, s4, s5 + vmovl.u8 q10, \s2 + vmovl.u8 q11, \s3 + vmovl.u8 q9, \s1 + vmovl.u8 q12, \s4 + vmovl.u8 q8, \s0 + vmovl.u8 q13, \s5 + vmul.u16 q10, q10, d0[2] + vmul.u16 q11, q11, d0[3] + vmls.u16 q10, q9, d0[1] + vmls.u16 q11, q12, d1[0] + vmla.u16 q10, q8, d0[0] + vmla.u16 q11, q13, d1[1] + vqadd.s16 q11, q10, q11 + vqrshrun.s16 \d0, q11, #7 +.endm + +.macro vp8_epel8_v6_y2 d0, d1, s0, s1, s2, s3, s4, s5, s6 + vmovl.u8 q10, \s0 + vmovl.u8 q11, \s3 + vmovl.u8 q14, \s6 + vmovl.u8 q9, \s1 + vmovl.u8 q12, \s4 + vmovl.u8 q8, \s2 + vmovl.u8 q13, \s5 + vmul.u16 q10, q10, d0[0] + vmul.u16 q15, q11, d0[3] + vmul.u16 q11, q11, d0[2] + vmul.u16 q14, q14, d1[1] + vmls.u16 q10, q9, d0[1] + vmls.u16 q15, q12, d1[0] + vmls.u16 q11, q8, d0[1] + vmls.u16 q14, q13, d1[0] + vmla.u16 q10, q8, d0[2] + vmla.u16 q15, q13, d1[1] + vmla.u16 q11, q9, d0[0] + vmla.u16 q14, q12, d0[3] + vqadd.s16 q15, q10, q15 + vqadd.s16 q14, q11, q14 + vqrshrun.s16 \d0, q15, #7 + vqrshrun.s16 \d1, q14, #7 +.endm + +.macro vp8_epel8_h4 d, a, b + vext.8 d28, \a, \b, #1 + vmovl.u8 q9, \a + vext.8 d29, \a, \b, #2 + vmovl.u8 q10, d28 + vext.8 d30, \a, \b, #3 + vmovl.u8 q11, d29 + vmovl.u8 q12, d30 + vmul.u16 q10, q10, d0[2] + vmul.u16 q11, q11, d0[3] + vmls.u16 q10, q9, d0[1] + vmls.u16 q11, q12, d1[0] + vqadd.s16 q11, q10, q11 + vqrshrun.s16 \d, q11, #7 +.endm + +.macro vp8_epel8_v4_y2 d0, d1, s0, s1, s2, s3, s4 + vmovl.u8 q9, \s0 + vmovl.u8 q10, \s1 + vmovl.u8 q11, \s2 + vmovl.u8 q12, \s3 + vmovl.u8 q13, \s4 + vmul.u16 q8, q10, d0[2] + vmul.u16 q14, q11, d0[3] + vmul.u16 q11, q11, d0[2] + vmul.u16 q15, q12, d0[3] + vmls.u16 q8, q9, d0[1] + vmls.u16 q14, q12, d1[0] + vmls.u16 q11, q10, d0[1] + vmls.u16 q15, q13, d1[0] + vqadd.s16 q8, q8, q14 + vqadd.s16 q11, q11, q15 + vqrshrun.s16 \d0, q8, #7 + vqrshrun.s16 \d1, q11, #7 +.endm + +function ff_put_vp8_epel16_v6_neon, export=1 + sub r2, r2, r3, lsl #1 + push {r4,lr} + vpush {d8-d15} + + ldr r4, [sp, #80] @ my + movrel lr, subpel_filters-16 + ldr r12, [sp, #72] @ h + add r4, lr, r4, lsl #4 + vld1.16 {q0}, [r4,:128] +1: + vld1.8 {d2-d3}, [r2], r3 + vld1.8 {d4-d5}, [r2], r3 + vld1.8 {d6-d7}, [r2], r3 + vld1.8 {d8-d9}, [r2], r3 + vld1.8 {d10-d11},[r2], r3 + vld1.8 {d12-d13},[r2], r3 + vld1.8 {d14-d15},[r2] + sub r2, r2, r3, lsl #2 + + vp8_epel8_v6_y2 d2, d4, d2, d4, d6, d8, d10, d12, d14 + vp8_epel8_v6_y2 d3, d5, d3, d5, d7, d9, d11, d13, d15 + + vst1.8 {d2-d3}, [r0,:128], r1 + vst1.8 {d4-d5}, [r0,:128], r1 + subs r12, r12, #2 + bne 1b + + vpop {d8-d15} + pop {r4,pc} +endfunc + +function ff_put_vp8_epel16_h6_neon, export=1 + sub r2, r2, #2 + push {r4,lr} + + ldr r4, [sp, #12] @ mx + movrel lr, subpel_filters-16 + ldr r12, [sp, #8] @ h + add r4, lr, r4, lsl #4 + vld1.16 {q0}, [r4,:128] +1: + vld1.8 {d2-d4}, [r2], r3 + + vp8_epel16_h6 d2, d3, d2, d3, d4, q1, q2 + + vst1.8 {d2-d3}, [r0,:128], r1 + subs r12, r12, #1 + bne 1b + + pop {r4,pc} +endfunc + +function ff_put_vp8_epel16_h6v6_neon, export=1 + sub r2, r2, r3, lsl #1 + sub r2, r2, #2 + push {r4,lr} + vpush {d8-d9} + + @ first pass (horizontal): + ldr r4, [sp, #28] @ mx + movrel lr, subpel_filters-16 + ldr r12, [sp, #24] @ h + add r4, lr, r4, lsl #4 + sub sp, sp, #336+16 + vld1.16 {q0}, [r4,:128] + add lr, sp, #15 + add r12, r12, #5 + bic lr, lr, #15 +1: + vld1.8 {d2,d3,d4}, [r2], r3 + + vp8_epel16_h6 d2, d3, d2, d3, d4, q1, q2 + + vst1.8 {d2-d3}, [lr,:128]! + subs r12, r12, #1 + bne 1b + + @ second pass (vertical): + ldr r4, [sp, #336+16+32] @ my + movrel lr, subpel_filters-16 + ldr r12, [sp, #336+16+24] @ h + add r4, lr, r4, lsl #4 + add lr, sp, #15 + vld1.16 {q0}, [r4,:128] + bic lr, lr, #15 +2: + vld1.8 {d2-d5}, [lr,:128]! + vld1.8 {d6-d9}, [lr,:128]! + vld1.8 {d28-d31},[lr,:128] + sub lr, lr, #48 + + vp8_epel8_v6 d2, d2, d4, d6, d8, d28, d30 + vp8_epel8_v6 d3, d3, d5, d7, d9, d29, d31 + + vst1.8 {d2-d3}, [r0,:128], r1 + subs r12, r12, #1 + bne 2b + + add sp, sp, #336+16 + vpop {d8-d9} + pop {r4,pc} +endfunc + +function ff_put_vp8_epel8_v6_neon, export=1 + sub r2, r2, r3, lsl #1 + push {r4,lr} + + ldr r4, [sp, #16] @ my + movrel lr, subpel_filters-16 + ldr r12, [sp, #8] @ h + add r4, lr, r4, lsl #4 + vld1.16 {q0}, [r4,:128] +1: + vld1.8 {d2}, [r2], r3 + vld1.8 {d3}, [r2], r3 + vld1.8 {d4}, [r2], r3 + vld1.8 {d5}, [r2], r3 + vld1.8 {d6}, [r2], r3 + vld1.8 {d7}, [r2], r3 + vld1.8 {d28}, [r2] + + sub r2, r2, r3, lsl #2 + + vp8_epel8_v6_y2 d2, d3, d2, d3, d4, d5, d6, d7, d28 + + vst1.8 {d2}, [r0,:64], r1 + vst1.8 {d3}, [r0,:64], r1 + subs r12, r12, #2 + bne 1b + + pop {r4,pc} +endfunc + +function ff_put_vp8_epel8_h6_neon, export=1 + sub r2, r2, #2 + push {r4,lr} + + ldr r4, [sp, #12] @ mx + movrel lr, subpel_filters-16 + ldr r12, [sp, #8] @ h + add r4, lr, r4, lsl #4 + vld1.16 {q0}, [r4,:128] +1: + vld1.8 {d2,d3}, [r2], r3 + + vp8_epel8_h6 d2, d2, d3 + + vst1.8 {d2}, [r0,:64], r1 + subs r12, r12, #1 + bne 1b + + pop {r4,pc} +endfunc + +function ff_put_vp8_epel8_h6v6_neon, export=1 + sub r2, r2, r3, lsl #1 + sub r2, r2, #2 + push {r4,lr} + + @ first pass (horizontal): + ldr r4, [sp, #12] @ mx + movrel lr, subpel_filters-16 + ldr r12, [sp, #8] @ h + add r4, lr, r4, lsl #4 + sub sp, sp, #168+16 + vld1.16 {q0}, [r4,:128] + add lr, sp, #15 + add r12, r12, #5 + bic lr, lr, #15 +1: + vld1.8 {d2,d3}, [r2], r3 + + vp8_epel8_h6 d2, d2, d3 + + vst1.8 {d2}, [lr,:64]! + subs r12, r12, #1 + bne 1b + + @ second pass (vertical): + ldr r4, [sp, #168+16+16] @ my + movrel lr, subpel_filters-16 + ldr r12, [sp, #168+16+8] @ h + add r4, lr, r4, lsl #4 + add lr, sp, #15 + vld1.16 {q0}, [r4,:128] + bic lr, lr, #15 +2: + vld1.8 {d2-d5}, [lr,:128]! + vld1.8 {d6-d7}, [lr,:128]! + vld1.8 {d30}, [lr,:64] + sub lr, lr, #32 + + vp8_epel8_v6_y2 d2, d3, d2, d3, d4, d5, d6, d7, d30 + + vst1.8 {d2}, [r0,:64], r1 + vst1.8 {d3}, [r0,:64], r1 + subs r12, r12, #2 + bne 2b + + add sp, sp, #168+16 + pop {r4,pc} +endfunc + +function ff_put_vp8_epel8_v4_neon, export=1 + sub r2, r2, r3 + push {r4,lr} + + ldr r4, [sp, #16] @ my + movrel lr, subpel_filters-16 + ldr r12, [sp, #8] @ h + add r4, lr, r4, lsl #4 + vld1.16 {q0}, [r4,:128] +1: + vld1.8 {d2}, [r2], r3 + vld1.8 {d3}, [r2], r3 + vld1.8 {d4}, [r2], r3 + vld1.8 {d5}, [r2], r3 + vld1.8 {d6}, [r2] + sub r2, r2, r3, lsl #1 + + vp8_epel8_v4_y2 d2, d3, d2, d3, d4, d5, d6 + + vst1.8 {d2}, [r0,:64], r1 + vst1.8 {d3}, [r0,:64], r1 + subs r12, r12, #2 + bne 1b + + pop {r4,pc} +endfunc + +function ff_put_vp8_epel8_h4_neon, export=1 + sub r2, r2, #1 + push {r4,lr} + + ldr r4, [sp, #12] @ mx + movrel lr, subpel_filters-16 + ldr r12, [sp, #8] @ h + add r4, lr, r4, lsl #4 + vld1.16 {q0}, [r4,:128] +1: + vld1.8 {d2,d3}, [r2], r3 + + vp8_epel8_h4 d2, d2, d3 + + vst1.8 {d2}, [r0,:64], r1 + subs r12, r12, #1 + bne 1b + + pop {r4,pc} +endfunc + +function ff_put_vp8_epel8_h4v4_neon, export=1 + sub r2, r2, r3 + sub r2, r2, #1 + push {r4,lr} + + @ first pass (horizontal): + ldr r4, [sp, #12] @ mx + movrel lr, subpel_filters-16 + ldr r12, [sp, #8] @ h + add r4, lr, r4, lsl #4 + sub sp, sp, #168+16 + vld1.16 {q0}, [r4,:128] + add lr, sp, #15 + add r12, r12, #3 + bic lr, lr, #15 +1: + vld1.8 {d2,d3}, [r2], r3 + + vp8_epel8_h4 d2, d2, d3 + + vst1.8 {d2}, [lr,:64]! + subs r12, r12, #1 + bne 1b + + @ second pass (vertical): + ldr r4, [sp, #168+16+16] @ my + movrel lr, subpel_filters-16 + ldr r12, [sp, #168+16+8] @ h + add r4, lr, r4, lsl #4 + add lr, sp, #15 + vld1.16 {q0}, [r4,:128] + bic lr, lr, #15 +2: + vld1.8 {d2-d5}, [lr,:128]! + vld1.8 {d6}, [lr,:64] + sub lr, lr, #16 + + vp8_epel8_v4_y2 d2, d3, d2, d3, d4, d5, d6 + + vst1.8 {d2}, [r0,:64], r1 + vst1.8 {d3}, [r0,:64], r1 + subs r12, r12, #2 + bne 2b + + add sp, sp, #168+16 + pop {r4,pc} +endfunc + +function ff_put_vp8_epel8_h6v4_neon, export=1 + sub r2, r2, r3 + sub r2, r2, #2 + push {r4,lr} + + @ first pass (horizontal): + ldr r4, [sp, #12] @ mx + movrel lr, subpel_filters-16 + ldr r12, [sp, #8] @ h + add r4, lr, r4, lsl #4 + sub sp, sp, #168+16 + vld1.16 {q0}, [r4,:128] + add lr, sp, #15 + add r12, r12, #3 + bic lr, lr, #15 +1: + vld1.8 {d2,d3}, [r2], r3 + + vp8_epel8_h6 d2, d2, d3 + + vst1.8 {d2}, [lr,:64]! + subs r12, r12, #1 + bne 1b + + @ second pass (vertical): + ldr r4, [sp, #168+16+16] @ my + movrel lr, subpel_filters-16 + ldr r12, [sp, #168+16+8] @ h + add r4, lr, r4, lsl #4 + add lr, sp, #15 + vld1.16 {q0}, [r4,:128] + bic lr, lr, #15 +2: + vld1.8 {d2-d5}, [lr,:128]! + vld1.8 {d6}, [lr,:64] + sub lr, lr, #16 + + vp8_epel8_v4_y2 d2, d3, d2, d3, d4, d5, d6 + + vst1.8 {d2}, [r0,:64], r1 + vst1.8 {d3}, [r0,:64], r1 + subs r12, r12, #2 + bne 2b + + add sp, sp, #168+16 + pop {r4,pc} +endfunc + +function ff_put_vp8_epel8_h4v6_neon, export=1 + sub r2, r2, r3, lsl #1 + sub r2, r2, #1 + push {r4,lr} + + @ first pass (horizontal): + ldr r4, [sp, #12] @ mx + movrel lr, subpel_filters-16 + ldr r12, [sp, #8] @ h + add r4, lr, r4, lsl #4 + sub sp, sp, #168+16 + vld1.16 {q0}, [r4,:128] + add lr, sp, #15 + add r12, r12, #5 + bic lr, lr, #15 +1: + vld1.8 {d2,d3}, [r2], r3 + + vp8_epel8_h4 d2, d2, d3 + + vst1.8 {d2}, [lr,:64]! + subs r12, r12, #1 + bne 1b + + @ second pass (vertical): + ldr r4, [sp, #168+16+16] @ my + movrel lr, subpel_filters-16 + ldr r12, [sp, #168+16+8] @ h + add r4, lr, r4, lsl #4 + add lr, sp, #15 + vld1.16 {q0}, [r4,:128] + bic lr, lr, #15 +2: + vld1.8 {d2-d5}, [lr,:128]! + vld1.8 {d6-d7}, [lr,:128]! + vld1.8 {d30}, [lr,:64] + sub lr, lr, #32 + + vp8_epel8_v6_y2 d2, d3, d2, d3, d4, d5, d6, d7, d30 + + vst1.8 {d2}, [r0,:64], r1 + vst1.8 {d3}, [r0,:64], r1 + subs r12, r12, #2 + bne 2b + + add sp, sp, #168+16 + pop {r4,pc} +endfunc + +function ff_put_vp8_epel4_v6_neon, export=1 + sub r2, r2, r3, lsl #1 + push {r4,lr} + + ldr r4, [sp, #16] @ my + movrel lr, subpel_filters-16 + ldr r12, [sp, #8] @ h + add r4, lr, r4, lsl #4 + vld1.16 {q0}, [r4,:128] +1: + vld1.32 {d2[]}, [r2], r3 + vld1.32 {d3[]}, [r2], r3 + vld1.32 {d4[]}, [r2], r3 + vld1.32 {d5[]}, [r2], r3 + vld1.32 {d6[]}, [r2], r3 + vld1.32 {d7[]}, [r2], r3 + vld1.32 {d28[]}, [r2] + sub r2, r2, r3, lsl #2 + vld1.32 {d2[1]}, [r2], r3 + vld1.32 {d3[1]}, [r2], r3 + vld1.32 {d4[1]}, [r2], r3 + vld1.32 {d5[1]}, [r2], r3 + vld1.32 {d6[1]}, [r2], r3 + vld1.32 {d7[1]}, [r2], r3 + vld1.32 {d28[1]}, [r2] + sub r2, r2, r3, lsl #2 + + vp8_epel8_v6_y2 d2, d3, d2, d3, d4, d5, d6, d7, d28 + + vst1.32 {d2[0]}, [r0,:32], r1 + vst1.32 {d3[0]}, [r0,:32], r1 + vst1.32 {d2[1]}, [r0,:32], r1 + vst1.32 {d3[1]}, [r0,:32], r1 + subs r12, r12, #4 + bne 1b + + pop {r4,pc} +endfunc + +function ff_put_vp8_epel4_h6_neon, export=1 + sub r2, r2, #2 + push {r4,lr} + + ldr r4, [sp, #12] @ mx + movrel lr, subpel_filters-16 + ldr r12, [sp, #8] @ h + add r4, lr, r4, lsl #4 + vld1.16 {q0}, [r4,:128] +1: + vld1.8 {q1}, [r2], r3 + vp8_epel8_h6 d2, d2, d3 + vst1.32 {d2[0]}, [r0,:32], r1 + subs r12, r12, #1 + bne 1b + + pop {r4,pc} +endfunc + +function ff_put_vp8_epel4_h6v6_neon, export=1 + sub r2, r2, r3, lsl #1 + sub r2, r2, #2 + push {r4,lr} + + ldr r4, [sp, #12] @ mx + movrel lr, subpel_filters-16 + ldr r12, [sp, #8] @ h + add r4, lr, r4, lsl #4 + sub sp, sp, #52+16 + vld1.16 {q0}, [r4,:128] + add lr, sp, #15 + add r12, r12, #5 + bic lr, lr, #15 +1: + vld1.8 {q1}, [r2], r3 + vp8_epel8_h6 d2, d2, d3 + vst1.32 {d2[0]}, [lr,:32]! + subs r12, r12, #1 + bne 1b + + ldr r4, [sp, #52+16+16] @ my + movrel lr, subpel_filters-16 + ldr r12, [sp, #52+16+8] @ h + add r4, lr, r4, lsl #4 + add lr, sp, #15 + vld1.16 {q0}, [r4,:128] + bic lr, lr, #15 +2: + vld1.8 {d2-d3}, [lr,:128]! + vld1.8 {d6}, [lr,:64]! + vld1.32 {d28[]}, [lr,:32] + sub lr, lr, #16 + vld1.8 {d4-d5}, [lr]! + vld1.8 {d7}, [lr,:64]! + vld1.32 {d28[1]}, [lr,:32] + sub lr, lr, #16 + vtrn.32 q1, q2 + vtrn.32 d6, d7 + vp8_epel8_v6_y2 d2, d3, d2, d4, d3, d5, d6, d7, d28 + vst1.32 {d2[0]}, [r0,:32], r1 + vst1.32 {d3[0]}, [r0,:32], r1 + vst1.32 {d2[1]}, [r0,:32], r1 + vst1.32 {d3[1]}, [r0,:32], r1 + subs r12, r12, #4 + bne 2b + + add sp, sp, #52+16 + pop {r4,pc} +endfunc + +function ff_put_vp8_epel4_h4v6_neon, export=1 + sub r2, r2, r3, lsl #1 + sub r2, r2, #1 + push {r4,lr} + + ldr r4, [sp, #12] @ mx + movrel lr, subpel_filters-16 + ldr r12, [sp, #8] @ h + add r4, lr, r4, lsl #4 + sub sp, sp, #52+16 + vld1.16 {q0}, [r4,:128] + add lr, sp, #15 + add r12, r12, #5 + bic lr, lr, #15 +1: + vld1.8 {d2}, [r2], r3 + vp8_epel8_h4 d2, d2, d2 + vst1.32 {d2[0]}, [lr,:32]! + subs r12, r12, #1 + bne 1b + + ldr r4, [sp, #52+16+16] @ my + movrel lr, subpel_filters-16 + ldr r12, [sp, #52+16+8] @ h + add r4, lr, r4, lsl #4 + add lr, sp, #15 + vld1.16 {q0}, [r4,:128] + bic lr, lr, #15 +2: + vld1.8 {d2-d3}, [lr,:128]! + vld1.8 {d6}, [lr,:64]! + vld1.32 {d28[]}, [lr,:32] + sub lr, lr, #16 + vld1.8 {d4-d5}, [lr]! + vld1.8 {d7}, [lr,:64]! + vld1.32 {d28[1]}, [lr,:32] + sub lr, lr, #16 + vtrn.32 q1, q2 + vtrn.32 d6, d7 + vp8_epel8_v6_y2 d2, d3, d2, d4, d3, d5, d6, d7, d28 + vst1.32 {d2[0]}, [r0,:32], r1 + vst1.32 {d3[0]}, [r0,:32], r1 + vst1.32 {d2[1]}, [r0,:32], r1 + vst1.32 {d3[1]}, [r0,:32], r1 + subs r12, r12, #4 + bne 2b + + add sp, sp, #52+16 + pop {r4,pc} +endfunc + +function ff_put_vp8_epel4_h6v4_neon, export=1 + sub r2, r2, r3 + sub r2, r2, #2 + push {r4,lr} + + ldr r4, [sp, #12] @ mx + movrel lr, subpel_filters-16 + ldr r12, [sp, #8] @ h + add r4, lr, r4, lsl #4 + sub sp, sp, #44+16 + vld1.16 {q0}, [r4,:128] + add lr, sp, #15 + add r12, r12, #3 + bic lr, lr, #15 +1: + vld1.8 {q1}, [r2], r3 + vp8_epel8_h6 d2, d2, d3 + vst1.32 {d2[0]}, [lr,:32]! + subs r12, r12, #1 + bne 1b + + ldr r4, [sp, #44+16+16] @ my + movrel lr, subpel_filters-16 + ldr r12, [sp, #44+16+8] @ h + add r4, lr, r4, lsl #4 + add lr, sp, #15 + vld1.16 {q0}, [r4,:128] + bic lr, lr, #15 +2: + vld1.8 {d2-d3}, [lr,:128]! + vld1.32 {d6[]}, [lr,:32] + sub lr, lr, #8 + vld1.8 {d4-d5}, [lr]! + vld1.32 {d6[1]}, [lr,:32] + sub lr, lr, #8 + vtrn.32 q1, q2 + vp8_epel8_v4_y2 d2, d3, d2, d4, d3, d5, d6 + vst1.32 {d2[0]}, [r0,:32], r1 + vst1.32 {d3[0]}, [r0,:32], r1 + vst1.32 {d2[1]}, [r0,:32], r1 + vst1.32 {d3[1]}, [r0,:32], r1 + subs r12, r12, #4 + bne 2b + + add sp, sp, #44+16 + pop {r4,pc} +endfunc + +function ff_put_vp8_epel4_h4_neon, export=1 + sub r2, r2, #1 + push {r4,lr} + + ldr r4, [sp, #12] @ mx + movrel lr, subpel_filters-16 + ldr r12, [sp, #8] @ h + add r4, lr, r4, lsl #4 + vld1.16 {q0}, [r4,:128] +1: + vld1.8 {d2}, [r2], r3 + vp8_epel8_h4 d2, d2, d2 + vst1.32 {d2[0]}, [r0,:32], r1 + subs r12, r12, #1 + bne 1b + + pop {r4,pc} +endfunc + +function ff_put_vp8_epel4_v4_neon, export=1 + sub r2, r2, r3 + push {r4,lr} + + ldr r4, [sp, #16] @ my + movrel lr, subpel_filters-16 + ldr r12, [sp, #8] @ h + add r4, lr, r4, lsl #4 + vld1.16 {q0}, [r4,:128] +1: + vld1.32 {d2[]}, [r2], r3 + vld1.32 {d3[]}, [r2], r3 + vld1.32 {d4[]}, [r2], r3 + vld1.32 {d5[]}, [r2], r3 + vld1.32 {d6[]}, [r2] + sub r2, r2, r3, lsl #1 + vld1.32 {d2[1]}, [r2], r3 + vld1.32 {d3[1]}, [r2], r3 + vld1.32 {d4[1]}, [r2], r3 + vld1.32 {d5[1]}, [r2], r3 + vld1.32 {d6[1]}, [r2] + sub r2, r2, r3, lsl #1 + + vp8_epel8_v4_y2 d2, d3, d2, d3, d4, d5, d6 + + vst1.32 {d2[0]}, [r0,:32], r1 + vst1.32 {d3[0]}, [r0,:32], r1 + vst1.32 {d2[1]}, [r0,:32], r1 + vst1.32 {d3[1]}, [r0,:32], r1 + subs r12, r12, #4 + bne 1b + + pop {r4,pc} +endfunc + +function ff_put_vp8_epel4_h4v4_neon, export=1 + sub r2, r2, r3 + sub r2, r2, #1 + push {r4,lr} + + ldr r4, [sp, #12] @ mx + movrel lr, subpel_filters-16 + ldr r12, [sp, #8] @ h + add r4, lr, r4, lsl #4 + sub sp, sp, #44+16 + vld1.16 {q0}, [r4,:128] + add lr, sp, #15 + add r12, r12, #3 + bic lr, lr, #15 +1: + vld1.8 {d2}, [r2], r3 + vp8_epel8_h4 d2, d2, d3 + vst1.32 {d2[0]}, [lr,:32]! + subs r12, r12, #1 + bne 1b + + ldr r4, [sp, #44+16+16] @ my + movrel lr, subpel_filters-16 + ldr r12, [sp, #44+16+8] @ h + add r4, lr, r4, lsl #4 + add lr, sp, #15 + vld1.16 {q0}, [r4,:128] + bic lr, lr, #15 +2: + vld1.8 {d2-d3}, [lr,:128]! + vld1.32 {d6[]}, [lr,:32] + sub lr, lr, #8 + vld1.8 {d4-d5}, [lr]! + vld1.32 {d6[1]}, [lr,:32] + sub lr, lr, #8 + vtrn.32 q1, q2 + vp8_epel8_v4_y2 d2, d3, d2, d4, d3, d5, d6 + vst1.32 {d2[0]}, [r0,:32], r1 + vst1.32 {d3[0]}, [r0,:32], r1 + vst1.32 {d2[1]}, [r0,:32], r1 + vst1.32 {d3[1]}, [r0,:32], r1 + subs r12, r12, #4 + bne 2b + + add sp, sp, #44+16 + pop {r4,pc} +endfunc + +@ note: worst case sum of all 6-tap filter values * 255 is 0x7f80 so 16 bit +@ arithmatic can be used to apply filters +const subpel_filters, align=4 + .short 0, 6, 123, 12, 1, 0, 0, 0 + .short 2, 11, 108, 36, 8, 1, 0, 0 + .short 0, 9, 93, 50, 6, 0, 0, 0 + .short 3, 16, 77, 77, 16, 3, 0, 0 + .short 0, 6, 50, 93, 9, 0, 0, 0 + .short 1, 8, 36, 108, 11, 2, 0, 0 + .short 0, 1, 12, 123, 6, 0, 0, 0 +endconst + +/* Bilinear MC */ + +function ff_put_vp8_bilin16_h_neon, export=1 + ldr r3, [sp, #4] @ mx + rsb r12, r3, #8 + vdup.8 d0, r3 + vdup.8 d1, r12 + ldr r12, [sp] @ h +1: + subs r12, r12, #2 + vld1.8 {d2-d4}, [r2], r1 + vext.8 q2, q1, q2, #1 + vmull.u8 q8, d2, d1 + vmlal.u8 q8, d4, d0 + vld1.8 {d18-d20},[r2], r1 + vmull.u8 q3, d3, d1 + vmlal.u8 q3, d5, d0 + vext.8 q10, q9, q10, #1 + vmull.u8 q11, d18, d1 + vmlal.u8 q11, d20, d0 + vmull.u8 q12, d19, d1 + vmlal.u8 q12, d21, d0 + vrshrn.u16 d4, q8, #3 + vrshrn.u16 d5, q3, #3 + vrshrn.u16 d6, q11, #3 + vrshrn.u16 d7, q12, #3 + vst1.8 {q2}, [r0,:128], r1 + vst1.8 {q3}, [r0,:128], r1 + bgt 1b + + bx lr +endfunc + +function ff_put_vp8_bilin16_v_neon, export=1 + ldr r3, [sp, #8] @ my + rsb r12, r3, #8 + vdup.8 d0, r3 + vdup.8 d1, r12 + ldr r12, [sp] @ h + vld1.8 {q1}, [r2], r1 +1: + subs r12, r12, #2 + vld1.8 {q2}, [r2], r1 + vmull.u8 q3, d2, d1 + vmlal.u8 q3, d4, d0 + vmull.u8 q8, d3, d1 + vmlal.u8 q8, d5, d0 + vld1.8 {q1}, [r2], r1 + vmull.u8 q9, d4, d1 + vmlal.u8 q9, d2, d0 + vmull.u8 q10, d5, d1 + vmlal.u8 q10, d3, d0 + vrshrn.u16 d4, q3, #3 + vrshrn.u16 d5, q8, #3 + vrshrn.u16 d6, q9, #3 + vrshrn.u16 d7, q10, #3 + vst1.8 {q2}, [r0,:128], r1 + vst1.8 {q3}, [r0,:128], r1 + bgt 1b + + bx lr +endfunc + +function ff_put_vp8_bilin16_hv_neon, export=1 + ldr r3, [sp, #4] @ mx + rsb r12, r3, #8 + vdup.8 d0, r3 + vdup.8 d1, r12 + ldr r3, [sp, #8] @ my + rsb r12, r3, #8 + vdup.8 d2, r3 + vdup.8 d3, r12 + ldr r12, [sp] @ h + + vld1.8 {d4-d6}, [r2], r1 + vext.8 q3, q2, q3, #1 + vmull.u8 q8, d4, d1 + vmlal.u8 q8, d6, d0 + vmull.u8 q9, d5, d1 + vmlal.u8 q9, d7, d0 + vrshrn.u16 d4, q8, #3 + vrshrn.u16 d5, q9, #3 +1: + subs r12, r12, #2 + vld1.8 {d18-d20},[r2], r1 + vext.8 q10, q9, q10, #1 + vmull.u8 q11, d18, d1 + vmlal.u8 q11, d20, d0 + vld1.8 {d26-d28},[r2], r1 + vmull.u8 q12, d19, d1 + vmlal.u8 q12, d21, d0 + vext.8 q14, q13, q14, #1 + vmull.u8 q8, d26, d1 + vmlal.u8 q8, d28, d0 + vmull.u8 q9, d27, d1 + vmlal.u8 q9, d29, d0 + vrshrn.u16 d6, q11, #3 + vrshrn.u16 d7, q12, #3 + vmull.u8 q12, d4, d3 + vmlal.u8 q12, d6, d2 + vmull.u8 q15, d5, d3 + vmlal.u8 q15, d7, d2 + vrshrn.u16 d4, q8, #3 + vrshrn.u16 d5, q9, #3 + vmull.u8 q10, d6, d3 + vmlal.u8 q10, d4, d2 + vmull.u8 q11, d7, d3 + vmlal.u8 q11, d5, d2 + vrshrn.u16 d24, q12, #3 + vrshrn.u16 d25, q15, #3 + vst1.8 {q12}, [r0,:128], r1 + vrshrn.u16 d20, q10, #3 + vrshrn.u16 d21, q11, #3 + vst1.8 {q10}, [r0,:128], r1 + bgt 1b + + bx lr +endfunc + +function ff_put_vp8_bilin8_h_neon, export=1 + ldr r3, [sp, #4] @ mx + rsb r12, r3, #8 + vdup.8 d0, r3 + vdup.8 d1, r12 + ldr r12, [sp] @ h +1: + subs r12, r12, #2 + vld1.8 {q1}, [r2], r1 + vext.8 d3, d2, d3, #1 + vmull.u8 q2, d2, d1 + vmlal.u8 q2, d3, d0 + vld1.8 {q3}, [r2], r1 + vext.8 d7, d6, d7, #1 + vmull.u8 q8, d6, d1 + vmlal.u8 q8, d7, d0 + vrshrn.u16 d4, q2, #3 + vrshrn.u16 d16, q8, #3 + vst1.8 {d4}, [r0,:64], r1 + vst1.8 {d16}, [r0,:64], r1 + bgt 1b + + bx lr +endfunc + +function ff_put_vp8_bilin8_v_neon, export=1 + ldr r3, [sp, #8] @ my + rsb r12, r3, #8 + vdup.8 d0, r3 + vdup.8 d1, r12 + ldr r12, [sp] @ h + vld1.8 {d2}, [r2], r1 +1: + subs r12, r12, #2 + vld1.8 {d3}, [r2], r1 + vmull.u8 q2, d2, d1 + vmlal.u8 q2, d3, d0 + vld1.8 {d2}, [r2], r1 + vmull.u8 q3, d3, d1 + vmlal.u8 q3, d2, d0 + vrshrn.u16 d4, q2, #3 + vrshrn.u16 d6, q3, #3 + vst1.8 {d4}, [r0,:64], r1 + vst1.8 {d6}, [r0,:64], r1 + bgt 1b + + bx lr +endfunc + +function ff_put_vp8_bilin8_hv_neon, export=1 + ldr r3, [sp, #4] @ mx + rsb r12, r3, #8 + vdup.8 d0, r3 + vdup.8 d1, r12 + ldr r3, [sp, #8] @ my + rsb r12, r3, #8 + vdup.8 d2, r3 + vdup.8 d3, r12 + ldr r12, [sp] @ h + + vld1.8 {q2}, [r2], r1 + vext.8 d5, d4, d5, #1 + vmull.u8 q9, d4, d1 + vmlal.u8 q9, d5, d0 + vrshrn.u16 d22, q9, #3 +1: + subs r12, r12, #2 + vld1.8 {q3}, [r2], r1 + vext.8 d7, d6, d7, #1 + vmull.u8 q8, d6, d1 + vmlal.u8 q8, d7, d0 + vld1.8 {q2}, [r2], r1 + vext.8 d5, d4, d5, #1 + vmull.u8 q9, d4, d1 + vmlal.u8 q9, d5, d0 + vrshrn.u16 d16, q8, #3 + vmull.u8 q10, d22, d3 + vmlal.u8 q10, d16, d2 + vrshrn.u16 d22, q9, #3 + vmull.u8 q12, d16, d3 + vmlal.u8 q12, d22, d2 + vrshrn.u16 d20, q10, #3 + vst1.8 {d20}, [r0,:64], r1 + vrshrn.u16 d23, q12, #3 + vst1.8 {d23}, [r0,:64], r1 + bgt 1b + + bx lr +endfunc + +function ff_put_vp8_bilin4_h_neon, export=1 + ldr r3, [sp, #4] @ mx + rsb r12, r3, #8 + vdup.8 d0, r3 + vdup.8 d1, r12 + ldr r12, [sp] @ h +1: + subs r12, r12, #2 + vld1.8 {d2}, [r2], r1 + vext.8 d3, d2, d3, #1 + vld1.8 {d6}, [r2], r1 + vext.8 d7, d6, d7, #1 + vtrn.32 q1, q3 + vmull.u8 q2, d2, d1 + vmlal.u8 q2, d3, d0 + vrshrn.u16 d4, q2, #3 + vst1.32 {d4[0]}, [r0,:32], r1 + vst1.32 {d4[1]}, [r0,:32], r1 + bgt 1b + + bx lr +endfunc + +function ff_put_vp8_bilin4_v_neon, export=1 + ldr r3, [sp, #8] @ my + rsb r12, r3, #8 + vdup.8 d0, r3 + vdup.8 d1, r12 + ldr r12, [sp] @ h + vld1.32 {d2[]}, [r2], r1 +1: + vld1.32 {d3[]}, [r2] + vld1.32 {d2[1]}, [r2], r1 + vld1.32 {d3[1]}, [r2], r1 + vmull.u8 q2, d2, d1 + vmlal.u8 q2, d3, d0 + vtrn.32 d3, d2 + vrshrn.u16 d4, q2, #3 + vst1.32 {d4[0]}, [r0,:32], r1 + vst1.32 {d4[1]}, [r0,:32], r1 + subs r12, r12, #2 + bgt 1b + + bx lr +endfunc + +function ff_put_vp8_bilin4_hv_neon, export=1 + ldr r3, [sp, #4] @ mx + rsb r12, r3, #8 + vdup.8 d0, r3 + vdup.8 d1, r12 + ldr r3, [sp, #8] @ my + rsb r12, r3, #8 + vdup.8 d2, r3 + vdup.8 d3, r12 + ldr r12, [sp] @ h + + vld1.8 {d4}, [r2], r1 + vext.8 d5, d4, d4, #1 + vmull.u8 q9, d4, d1 + vmlal.u8 q9, d5, d0 + vrshrn.u16 d22, q9, #3 +1: + subs r12, r12, #2 + vld1.8 {d6}, [r2], r1 + vext.8 d7, d6, d6, #1 + vld1.8 {d4}, [r2], r1 + vext.8 d5, d4, d4, #1 + vtrn.32 q3, q2 + vmull.u8 q8, d6, d1 + vmlal.u8 q8, d7, d0 + vrshrn.u16 d16, q8, #3 + vmull.u8 q10, d16, d2 + vtrn.32 d22, d16 + vmlal.u8 q10, d22, d3 + vrev64.32 d22, d16 + vrshrn.u16 d20, q10, #3 + vst1.32 {d20[0]}, [r0,:32], r1 + vst1.32 {d20[1]}, [r0,:32], r1 + bgt 1b + + bx lr +endfunc diff --git a/libavcodec/vp8dsp.c b/libavcodec/vp8dsp.c index 30699cecbe..4391a7ebc7 100644 --- a/libavcodec/vp8dsp.c +++ b/libavcodec/vp8dsp.c @@ -523,4 +523,6 @@ av_cold void ff_vp8dsp_init(VP8DSPContext *dsp) ff_vp8dsp_init_x86(dsp); if (HAVE_ALTIVEC) ff_vp8dsp_init_altivec(dsp); + if (ARCH_ARM) + ff_vp8dsp_init_arm(dsp); } diff --git a/libavcodec/vp8dsp.h b/libavcodec/vp8dsp.h index ee5c7ec0fb..e8757ab7d0 100644 --- a/libavcodec/vp8dsp.h +++ b/libavcodec/vp8dsp.h @@ -79,5 +79,6 @@ void ff_put_vp8_pixels4_c(uint8_t *dst, uint8_t *src, int stride, int h, int x, void ff_vp8dsp_init(VP8DSPContext *c); void ff_vp8dsp_init_x86(VP8DSPContext *c); void ff_vp8dsp_init_altivec(VP8DSPContext *c); +void ff_vp8dsp_init_arm(VP8DSPContext *c); #endif /* AVCODEC_VP8DSP_H */