[x265] [PATCH] primivites: rename luma_p2s to convert_p2s and move into PU
Steve Borho
steve at borho.org
Fri Apr 3 17:50:09 CEST 2015
On 04/03, rajesh at multicorewareinc.com wrote:
> # HG changeset patch
> # User Rajesh Paulraj<rajesh at multicorewareinc.com>
> # Date 1428065328 -19800
> # Fri Apr 03 18:18:48 2015 +0530
> # Node ID 24c96db729600c88a278c06cce4b8bf041d1357a
> # Parent 9a5fa67583feb6ffb7668f82632f7e93e5ec9415
> primivites: rename luma_p2s to convert_p2s and move into PU
this series is queued for testing
> diff -r 9a5fa67583fe -r 24c96db72960 source/common/ipfilter.cpp
> --- a/source/common/ipfilter.cpp Thu Apr 02 13:21:32 2015 -0500
> +++ b/source/common/ipfilter.cpp Fri Apr 03 18:18:48 2015 +0530
> @@ -34,27 +34,8 @@
> #endif
>
> namespace {
> -template<int dstStride, int width, int height>
> -void pixelToShort_c(const pixel* src, intptr_t srcStride, int16_t* dst)
> -{
> - int shift = IF_INTERNAL_PREC - X265_DEPTH;
> - int row, col;
> -
> - for (row = 0; row < height; row++)
> - {
> - for (col = 0; col < width; col++)
> - {
> - int16_t val = src[col] << shift;
> - dst[col] = val - (int16_t)IF_INTERNAL_OFFS;
> - }
> -
> - src += srcStride;
> - dst += dstStride;
> - }
> -}
> -
> -template<int dstStride>
> -void filterPixelToShort_c(const pixel* src, intptr_t srcStride, int16_t* dst, int width, int height)
> +template<int width, int height>
> +void filterPixelToShort_c(const pixel* src, intptr_t srcStride, int16_t* dst, intptr_t dstStride)
> {
> int shift = IF_INTERNAL_PREC - X265_DEPTH;
> int row, col;
> @@ -398,7 +379,7 @@
> p.chroma[X265_CSP_I420].pu[CHROMA_420_ ## W ## x ## H].filter_vps = interp_vert_ps_c<4, W, H>; \
> p.chroma[X265_CSP_I420].pu[CHROMA_420_ ## W ## x ## H].filter_vsp = interp_vert_sp_c<4, W, H>; \
> p.chroma[X265_CSP_I420].pu[CHROMA_420_ ## W ## x ## H].filter_vss = interp_vert_ss_c<4, W, H>; \
> - p.chroma[X265_CSP_I420].pu[CHROMA_420_ ## W ## x ## H].chroma_p2s = pixelToShort_c<MAX_CU_SIZE / 2, W, H>;
> + p.chroma[X265_CSP_I420].pu[CHROMA_420_ ## W ## x ## H].p2s = filterPixelToShort_c<W, H>;
>
> #define CHROMA_422(W, H) \
> p.chroma[X265_CSP_I422].pu[CHROMA_422_ ## W ## x ## H].filter_hpp = interp_horiz_pp_c<4, W, H>; \
> @@ -407,7 +388,7 @@
> p.chroma[X265_CSP_I422].pu[CHROMA_422_ ## W ## x ## H].filter_vps = interp_vert_ps_c<4, W, H>; \
> p.chroma[X265_CSP_I422].pu[CHROMA_422_ ## W ## x ## H].filter_vsp = interp_vert_sp_c<4, W, H>; \
> p.chroma[X265_CSP_I422].pu[CHROMA_422_ ## W ## x ## H].filter_vss = interp_vert_ss_c<4, W, H>; \
> - p.chroma[X265_CSP_I422].pu[CHROMA_422_ ## W ## x ## H].chroma_p2s = pixelToShort_c<MAX_CU_SIZE / 2, W, H>;
> + p.chroma[X265_CSP_I422].pu[CHROMA_422_ ## W ## x ## H].p2s = filterPixelToShort_c<W, H>;
>
> #define CHROMA_444(W, H) \
> p.chroma[X265_CSP_I444].pu[LUMA_ ## W ## x ## H].filter_hpp = interp_horiz_pp_c<4, W, H>; \
> @@ -416,7 +397,7 @@
> p.chroma[X265_CSP_I444].pu[LUMA_ ## W ## x ## H].filter_vps = interp_vert_ps_c<4, W, H>; \
> p.chroma[X265_CSP_I444].pu[LUMA_ ## W ## x ## H].filter_vsp = interp_vert_sp_c<4, W, H>; \
> p.chroma[X265_CSP_I444].pu[LUMA_ ## W ## x ## H].filter_vss = interp_vert_ss_c<4, W, H>; \
> - p.chroma[X265_CSP_I444].pu[LUMA_ ## W ## x ## H].chroma_p2s = pixelToShort_c<MAX_CU_SIZE, W, H>;
> + p.chroma[X265_CSP_I444].pu[LUMA_ ## W ## x ## H].p2s = filterPixelToShort_c<W, H>;
>
> #define LUMA(W, H) \
> p.pu[LUMA_ ## W ## x ## H].luma_hpp = interp_horiz_pp_c<8, W, H>; \
> @@ -426,7 +407,7 @@
> p.pu[LUMA_ ## W ## x ## H].luma_vsp = interp_vert_sp_c<8, W, H>; \
> p.pu[LUMA_ ## W ## x ## H].luma_vss = interp_vert_ss_c<8, W, H>; \
> p.pu[LUMA_ ## W ## x ## H].luma_hvpp = interp_hv_pp_c<8, W, H>; \
> - p.pu[LUMA_ ## W ## x ## H].filter_p2s = pixelToShort_c<MAX_CU_SIZE, W, H>
> + p.pu[LUMA_ ## W ## x ## H].convert_p2s = filterPixelToShort_c<W, H>;
>
> void setupFilterPrimitives_c(EncoderPrimitives& p)
> {
> @@ -530,11 +511,6 @@
> CHROMA_444(48, 64);
> CHROMA_444(64, 16);
> CHROMA_444(16, 64);
> - p.luma_p2s = filterPixelToShort_c<MAX_CU_SIZE>;
> -
> - p.chroma[X265_CSP_I444].p2s = filterPixelToShort_c<MAX_CU_SIZE>;
> - p.chroma[X265_CSP_I420].p2s = filterPixelToShort_c<MAX_CU_SIZE / 2>;
> - p.chroma[X265_CSP_I422].p2s = filterPixelToShort_c<MAX_CU_SIZE / 2>;
>
> p.extendRowBorder = extendCURowColBorder;
> }
> diff -r 9a5fa67583fe -r 24c96db72960 source/common/predict.cpp
> --- a/source/common/predict.cpp Thu Apr 02 13:21:32 2015 -0500
> +++ b/source/common/predict.cpp Fri Apr 03 18:18:48 2015 +0530
> @@ -273,7 +273,7 @@
> void Predict::predInterLumaShort(const PredictionUnit& pu, ShortYuv& dstSYuv, const PicYuv& refPic, const MV& mv) const
> {
> int16_t* dst = dstSYuv.getLumaAddr(pu.puAbsPartIdx);
> - int dstStride = dstSYuv.m_size;
> + intptr_t dstStride = dstSYuv.m_size;
>
> intptr_t srcStride = refPic.m_stride;
> intptr_t srcOffset = (mv.x >> 2) + (mv.y >> 2) * srcStride;
> @@ -288,7 +288,7 @@
> X265_CHECK(dstStride == MAX_CU_SIZE, "stride expected to be max cu size\n");
>
> if (!(yFrac | xFrac))
> - primitives.luma_p2s(src, srcStride, dst, pu.width, pu.height);
> + primitives.pu[partEnum].convert_p2s(src, srcStride, dst, dstStride);
> else if (!yFrac)
> primitives.pu[partEnum].luma_hps(src, srcStride, dst, dstStride, xFrac, 0);
> else if (!xFrac)
> @@ -375,14 +375,13 @@
> int partEnum = partitionFromSizes(pu.width, pu.height);
>
> uint32_t cxWidth = pu.width >> m_hChromaShift;
> - uint32_t cxHeight = pu.height >> m_vChromaShift;
>
> - X265_CHECK(((cxWidth | cxHeight) % 2) == 0, "chroma block size expected to be multiple of 2\n");
> + X265_CHECK(((cxWidth | (pu.height >> m_vChromaShift)) % 2) == 0, "chroma block size expected to be multiple of 2\n");
>
> if (!(yFrac | xFrac))
> {
> - primitives.chroma[m_csp].p2s(refCb, refStride, dstCb, cxWidth, cxHeight);
> - primitives.chroma[m_csp].p2s(refCr, refStride, dstCr, cxWidth, cxHeight);
> + primitives.chroma[m_csp].pu[partEnum].p2s(refCb, refStride, dstCb, dstStride);
> + primitives.chroma[m_csp].pu[partEnum].p2s(refCr, refStride, dstCr, dstStride);
> }
> else if (!yFrac)
> {
> diff -r 9a5fa67583fe -r 24c96db72960 source/common/primitives.cpp
> --- a/source/common/primitives.cpp Thu Apr 02 13:21:32 2015 -0500
> +++ b/source/common/primitives.cpp Fri Apr 03 18:18:48 2015 +0530
> @@ -90,7 +90,6 @@
>
> /* alias chroma 4:4:4 from luma primitives (all but chroma filters) */
>
> - p.chroma[X265_CSP_I444].p2s = p.luma_p2s;
> p.chroma[X265_CSP_I444].cu[BLOCK_4x4].sa8d = NULL;
>
> for (int i = 0; i < NUM_PU_SIZES; i++)
> @@ -98,7 +97,7 @@
> p.chroma[X265_CSP_I444].pu[i].copy_pp = p.pu[i].copy_pp;
> p.chroma[X265_CSP_I444].pu[i].addAvg = p.pu[i].addAvg;
> p.chroma[X265_CSP_I444].pu[i].satd = p.pu[i].satd;
> - p.chroma[X265_CSP_I444].pu[i].chroma_p2s = p.pu[i].filter_p2s;
> + p.chroma[X265_CSP_I444].pu[i].p2s = p.pu[i].convert_p2s;
> }
>
> for (int i = 0; i < NUM_CU_SIZES; i++)
> diff -r 9a5fa67583fe -r 24c96db72960 source/common/primitives.h
> --- a/source/common/primitives.h Thu Apr 02 13:21:32 2015 -0500
> +++ b/source/common/primitives.h Fri Apr 03 18:18:48 2015 +0530
> @@ -156,8 +156,7 @@
> typedef void (*filter_sp_t) (const int16_t* src, intptr_t srcStride, pixel* dst, intptr_t dstStride, int coeffIdx);
> typedef void (*filter_ss_t) (const int16_t* src, intptr_t srcStride, int16_t* dst, intptr_t dstStride, int coeffIdx);
> typedef void (*filter_hv_pp_t) (const pixel* src, intptr_t srcStride, pixel* dst, intptr_t dstStride, int idxX, int idxY);
> -typedef void (*filter_p2s_wxh_t)(const pixel* src, intptr_t srcStride, int16_t* dst, int width, int height);
> -typedef void (*filter_p2s_t)(const pixel* src, intptr_t srcStride, int16_t* dst);
> +typedef void (*filter_p2s_t)(const pixel* src, intptr_t srcStride, int16_t* dst, intptr_t dstStride);
>
> typedef void (*copy_pp_t)(pixel* dst, intptr_t dstStride, const pixel* src, intptr_t srcStride); // dst is aligned
> typedef void (*copy_sp_t)(pixel* dst, intptr_t dstStride, const int16_t* src, intptr_t srcStride);
> @@ -211,7 +210,7 @@
> addAvg_t addAvg; // bidir motion compensation, uses 16bit values
>
> copy_pp_t copy_pp;
> - filter_p2s_t filter_p2s;
> + filter_p2s_t convert_p2s;
> }
> pu[NUM_PU_SIZES];
>
> @@ -290,7 +289,6 @@
> weightp_sp_t weight_sp;
> weightp_pp_t weight_pp;
>
> - filter_p2s_wxh_t luma_p2s;
>
> findPosLast_t findPosLast;
>
> @@ -317,7 +315,7 @@
> filter_hps_t filter_hps;
> addAvg_t addAvg;
> copy_pp_t copy_pp;
> - filter_p2s_t chroma_p2s;
> + filter_p2s_t p2s;
>
> }
> pu[NUM_PU_SIZES];
> @@ -337,7 +335,6 @@
> }
> cu[NUM_CU_SIZES];
>
> - filter_p2s_wxh_t p2s; // takes width/height as arguments
> }
> chroma[X265_CSP_COUNT];
> };
> diff -r 9a5fa67583fe -r 24c96db72960 source/common/x86/asm-primitives.cpp
> --- a/source/common/x86/asm-primitives.cpp Thu Apr 02 13:21:32 2015 -0500
> +++ b/source/common/x86/asm-primitives.cpp Fri Apr 03 18:18:48 2015 +0530
> @@ -859,9 +859,6 @@
> PIXEL_AVG_W4(mmx2);
> LUMA_VAR(sse2);
>
> - p.luma_p2s = x265_luma_p2s_sse2;
> - p.chroma[X265_CSP_I420].p2s = x265_chroma_p2s_sse2;
> - p.chroma[X265_CSP_I422].p2s = x265_chroma_p2s_sse2;
>
> ALL_LUMA_TU(blockfill_s, blockfill_s, sse2);
> ALL_LUMA_TU_S(cpy1Dto2D_shr, cpy1Dto2D_shr_, sse2);
> @@ -1273,31 +1270,6 @@
> ASSIGN_SSE_PP(ssse3);
> p.cu[BLOCK_4x4].sse_pp = x265_pixel_ssd_4x4_ssse3;
> p.chroma[X265_CSP_I422].cu[BLOCK_422_4x8].sse_pp = x265_pixel_ssd_4x8_ssse3;
> - p.pu[LUMA_4x4].filter_p2s = x265_pixelToShort_4x4_ssse3;
> - p.pu[LUMA_4x8].filter_p2s = x265_pixelToShort_4x8_ssse3;
> - p.pu[LUMA_4x16].filter_p2s = x265_pixelToShort_4x16_ssse3;
> - p.pu[LUMA_8x4].filter_p2s = x265_pixelToShort_8x4_ssse3;
> - p.pu[LUMA_8x8].filter_p2s = x265_pixelToShort_8x8_ssse3;
> - p.pu[LUMA_8x16].filter_p2s = x265_pixelToShort_8x16_ssse3;
> - p.pu[LUMA_8x32].filter_p2s = x265_pixelToShort_8x32_ssse3;
> - p.pu[LUMA_16x4].filter_p2s = x265_pixelToShort_16x4_ssse3;
> - p.pu[LUMA_16x8].filter_p2s = x265_pixelToShort_16x8_ssse3;
> - p.pu[LUMA_16x12].filter_p2s = x265_pixelToShort_16x12_ssse3;
> - p.pu[LUMA_16x16].filter_p2s = x265_pixelToShort_16x16_ssse3;
> - p.pu[LUMA_16x32].filter_p2s = x265_pixelToShort_16x32_ssse3;
> - p.pu[LUMA_16x64].filter_p2s = x265_pixelToShort_16x64_ssse3;
> - p.pu[LUMA_32x8].filter_p2s = x265_pixelToShort_32x8_ssse3;
> - p.pu[LUMA_32x16].filter_p2s = x265_pixelToShort_32x16_ssse3;
> - p.pu[LUMA_32x24].filter_p2s = x265_pixelToShort_32x24_ssse3;
> - p.pu[LUMA_32x32].filter_p2s = x265_pixelToShort_32x32_ssse3;
> - p.pu[LUMA_32x64].filter_p2s = x265_pixelToShort_32x64_ssse3;
> - p.pu[LUMA_64x16].filter_p2s = x265_pixelToShort_64x16_ssse3;
> - p.pu[LUMA_64x32].filter_p2s = x265_pixelToShort_64x32_ssse3;
> - p.pu[LUMA_64x48].filter_p2s = x265_pixelToShort_64x48_ssse3;
> - p.pu[LUMA_64x64].filter_p2s = x265_pixelToShort_64x64_ssse3;
> -
> - p.chroma[X265_CSP_I420].p2s = x265_chroma_p2s_ssse3;
> - p.chroma[X265_CSP_I422].p2s = x265_chroma_p2s_ssse3;
>
> p.dst4x4 = x265_dst4_ssse3;
> p.cu[BLOCK_8x8].idct = x265_idct8_ssse3;
> diff -r 9a5fa67583fe -r 24c96db72960 source/common/x86/ipfilter8.asm
> --- a/source/common/x86/ipfilter8.asm Thu Apr 02 13:21:32 2015 -0500
> +++ b/source/common/x86/ipfilter8.asm Fri Apr 03 18:18:48 2015 +0530
> @@ -7740,320 +7740,6 @@
> FILTER_V4_W16n_H2 64, 48
> FILTER_V4_W16n_H2 48, 64
> FILTER_V4_W16n_H2 64, 16
> -;-----------------------------------------------------------------------------
> -; void pixelToShort(pixel *src, intptr_t srcStride, int16_t *dst, int width, int height)
> -;-----------------------------------------------------------------------------
> -%macro PIXEL_WH_4xN 2
> -INIT_XMM ssse3
> -cglobal pixelToShort_%1x%2, 3, 7, 6
> -
> - ; load width and height
> - mov r3d, %1
> - mov r4d, %2
> - ; load constant
> - mova m4, [pb_128]
> - mova m5, [tab_c_64_n64]
> -.loopH:
> - xor r5d, r5d
> -
> -.loopW:
> - mov r6, r0
> - movh m0, [r6]
> - punpcklbw m0, m4
> - pmaddubsw m0, m5
> -
> - movh m1, [r6 + r1]
> - punpcklbw m1, m4
> - pmaddubsw m1, m5
> -
> - movh m2, [r6 + r1 * 2]
> - punpcklbw m2, m4
> - pmaddubsw m2, m5
> -
> - lea r6, [r6 + r1 * 2]
> - movh m3, [r6 + r1]
> - punpcklbw m3, m4
> - pmaddubsw m3, m5
> -
> - add r5, 8
> - cmp r5, r3
> - jg .width4
> - movu [r2 + r5 * 2 + FENC_STRIDE * 0 - 16], m0
> - movu [r2 + r5 * 2 + FENC_STRIDE * 2 - 16], m1
> - movu [r2 + r5 * 2 + FENC_STRIDE * 4 - 16], m2
> - movu [r2 + r5 * 2 + FENC_STRIDE * 6 - 16], m3
> - je .nextH
> - jmp .loopW
> -
> -.width4:
> - movh [r2 + r5 * 2 + FENC_STRIDE * 0 - 16], m0
> - movh [r2 + r5 * 2 + FENC_STRIDE * 2 - 16], m1
> - movh [r2 + r5 * 2 + FENC_STRIDE * 4 - 16], m2
> - movh [r2 + r5 * 2 + FENC_STRIDE * 6 - 16], m3
> -
> -.nextH:
> - lea r0, [r0 + r1 * 4]
> - add r2, FENC_STRIDE * 8
> -
> - sub r4d, 4
> - jnz .loopH
> - RET
> -%endmacro
> -PIXEL_WH_4xN 4, 4
> -PIXEL_WH_4xN 4, 8
> -PIXEL_WH_4xN 4, 16
> -
> -;-----------------------------------------------------------------------------
> -; void pixelToShort(pixel *src, intptr_t srcStride, int16_t *dst, int width, int height)
> -;-----------------------------------------------------------------------------
> -%macro PIXEL_WH_8xN 2
> -INIT_XMM ssse3
> -cglobal pixelToShort_%1x%2, 3, 7, 6
> -
> - ; load width and height
> - mov r3d, %1
> - mov r4d, %2
> -
> - ; load constant
> - mova m4, [pb_128]
> - mova m5, [tab_c_64_n64]
> -
> -.loopH
> - xor r5d, r5d
> -.loopW
> - lea r6, [r0 + r5]
> -
> - movh m0, [r6]
> - punpcklbw m0, m4
> - pmaddubsw m0, m5
> -
> - movh m1, [r6 + r1]
> - punpcklbw m1, m4
> - pmaddubsw m1, m5
> -
> - movh m2, [r6 + r1 * 2]
> - punpcklbw m2, m4
> - pmaddubsw m2, m5
> -
> - lea r6, [r6 + r1 * 2]
> - movh m3, [r6 + r1]
> - punpcklbw m3, m4
> - pmaddubsw m3, m5
> -
> - add r5, 8
> - cmp r5, r3
> -
> - movu [r2 + FENC_STRIDE * 0], m0
> - movu [r2 + FENC_STRIDE * 2], m1
> - movu [r2 + FENC_STRIDE * 4], m2
> - movu [r2 + FENC_STRIDE * 6], m3
> -
> - je .nextH
> - jmp .loopW
> -
> -
> -.nextH:
> - lea r0, [r0 + r1 * 4]
> - add r2, FENC_STRIDE * 8
> -
> - sub r4d, 4
> - jnz .loopH
> - RET
> -%endmacro
> -PIXEL_WH_8xN 8, 8
> -PIXEL_WH_8xN 8, 4
> -PIXEL_WH_8xN 8, 16
> -PIXEL_WH_8xN 8, 32
> -
> -
> -;-----------------------------------------------------------------------------
> -; void pixelToShort(pixel *src, intptr_t srcStride, int16_t *dst, int width, int height)
> -;-----------------------------------------------------------------------------
> -%macro PIXEL_WH_16xN 2
> -INIT_XMM ssse3
> -cglobal pixelToShort_%1x%2, 3, 7, 6
> -
> - ; load width and height
> - mov r3d, %1
> - mov r4d, %2
> -
> - ; load constant
> - mova m4, [pb_128]
> - mova m5, [tab_c_64_n64]
> -
> -.loopH:
> - xor r5d, r5d
> -.loopW:
> - lea r6, [r0 + r5]
> -
> - movh m0, [r6]
> - punpcklbw m0, m4
> - pmaddubsw m0, m5
> -
> - movh m1, [r6 + r1]
> - punpcklbw m1, m4
> - pmaddubsw m1, m5
> -
> - movh m2, [r6 + r1 * 2]
> - punpcklbw m2, m4
> - pmaddubsw m2, m5
> -
> - lea r6, [r6 + r1 * 2]
> - movh m3, [r6 + r1]
> - punpcklbw m3, m4
> - pmaddubsw m3, m5
> -
> - add r5, 8
> - cmp r5, r3
> -
> - movu [r2 + r5 * 2 + FENC_STRIDE * 0 - 16], m0
> - movu [r2 + r5 * 2 + FENC_STRIDE * 2 - 16], m1
> - movu [r2 + r5 * 2 + FENC_STRIDE * 4 - 16], m2
> - movu [r2 + r5 * 2 + FENC_STRIDE * 6 - 16], m3
> - je .nextH
> - jmp .loopW
> -
> -
> -.nextH:
> - lea r0, [r0 + r1 * 4]
> - add r2, FENC_STRIDE * 8
> -
> - sub r4d, 4
> - jnz .loopH
> -
> - RET
> -%endmacro
> -PIXEL_WH_16xN 16, 16
> -PIXEL_WH_16xN 16, 8
> -PIXEL_WH_16xN 16, 4
> -PIXEL_WH_16xN 16, 12
> -PIXEL_WH_16xN 16, 32
> -PIXEL_WH_16xN 16, 64
> -
> -;-----------------------------------------------------------------------------
> -; void pixelToShort(pixel *src, intptr_t srcStride, int16_t *dst, int width, int height)
> -;-----------------------------------------------------------------------------
> -%macro PIXEL_WH_32xN 2
> -INIT_XMM ssse3
> -cglobal pixelToShort_%1x%2, 3, 7, 6
> -
> - ; load width and height
> - mov r3d, %1
> - mov r4d, %2
> -
> - ; load constant
> - mova m4, [pb_128]
> - mova m5, [tab_c_64_n64]
> -
> -.loopH:
> - xor r5d, r5d
> -.loopW:
> - lea r6, [r0 + r5]
> -
> - movh m0, [r6]
> - punpcklbw m0, m4
> - pmaddubsw m0, m5
> -
> - movh m1, [r6 + r1]
> - punpcklbw m1, m4
> - pmaddubsw m1, m5
> -
> - movh m2, [r6 + r1 * 2]
> - punpcklbw m2, m4
> - pmaddubsw m2, m5
> -
> - lea r6, [r6 + r1 * 2]
> - movh m3, [r6 + r1]
> - punpcklbw m3, m4
> - pmaddubsw m3, m5
> -
> - add r5, 8
> - cmp r5, r3
> -
> - movu [r2 + r5 * 2 + FENC_STRIDE * 0 - 16], m0
> - movu [r2 + r5 * 2 + FENC_STRIDE * 2 - 16], m1
> - movu [r2 + r5 * 2 + FENC_STRIDE * 4 - 16], m2
> - movu [r2 + r5 * 2 + FENC_STRIDE * 6 - 16], m3
> - je .nextH
> - jmp .loopW
> -
> -
> -.nextH:
> - lea r0, [r0 + r1 * 4]
> - add r2, FENC_STRIDE * 8
> -
> - sub r4d, 4
> - jnz .loopH
> -
> - RET
> -%endmacro
> -PIXEL_WH_32xN 32, 32
> -PIXEL_WH_32xN 32, 8
> -PIXEL_WH_32xN 32, 16
> -PIXEL_WH_32xN 32, 24
> -PIXEL_WH_32xN 32, 64
> -
> -;-----------------------------------------------------------------------------
> -; void pixelToShort(pixel *src, intptr_t srcStride, int16_t *dst, int width, int height)
> -;-----------------------------------------------------------------------------
> -%macro PIXEL_WH_64xN 2
> -INIT_XMM ssse3
> -cglobal pixelToShort_%1x%2, 3, 7, 6
> -
> - ; load width and height
> - mov r3d, %1
> - mov r4d, %2
> -
> - ; load constant
> - mova m4, [pb_128]
> - mova m5, [tab_c_64_n64]
> -
> -.loopH:
> - xor r5d, r5d
> -.loopW:
> - lea r6, [r0 + r5]
> -
> - movh m0, [r6]
> - punpcklbw m0, m4
> - pmaddubsw m0, m5
> -
> - movh m1, [r6 + r1]
> - punpcklbw m1, m4
> - pmaddubsw m1, m5
> -
> - movh m2, [r6 + r1 * 2]
> - punpcklbw m2, m4
> - pmaddubsw m2, m5
> -
> - lea r6, [r6 + r1 * 2]
> - movh m3, [r6 + r1]
> - punpcklbw m3, m4
> - pmaddubsw m3, m5
> -
> - add r5, 8
> - cmp r5, r3
> -
> - movu [r2 + r5 * 2 + FENC_STRIDE * 0 - 16], m0
> - movu [r2 + r5 * 2 + FENC_STRIDE * 2 - 16], m1
> - movu [r2 + r5 * 2 + FENC_STRIDE * 4 - 16], m2
> - movu [r2 + r5 * 2 + FENC_STRIDE * 6 - 16], m3
> - je .nextH
> - jmp .loopW
> -
> -
> -.nextH:
> - lea r0, [r0 + r1 * 4]
> - add r2, FENC_STRIDE * 8
> -
> - sub r4d, 4
> - jnz .loopH
> -
> - RET
> -%endmacro
> -PIXEL_WH_64xN 64, 64
> -PIXEL_WH_64xN 64, 16
> -PIXEL_WH_64xN 64, 32
> -PIXEL_WH_64xN 64, 48
>
> %macro PROCESS_LUMA_W4_4R 0
> movd m0, [r0]
> diff -r 9a5fa67583fe -r 24c96db72960 source/common/x86/ipfilter8.h
> --- a/source/common/x86/ipfilter8.h Thu Apr 02 13:21:32 2015 -0500
> +++ b/source/common/x86/ipfilter8.h Fri Apr 03 18:18:48 2015 +0530
> @@ -289,8 +289,6 @@
> SETUP_CHROMA_420_HORIZ_FUNC_DEF(64, 16, cpu); \
> SETUP_CHROMA_420_HORIZ_FUNC_DEF(16, 64, cpu)
>
> -void x265_chroma_p2s_sse2(const pixel* src, intptr_t srcStride, int16_t* dst, int width, int height);
> -void x265_luma_p2s_sse2(const pixel* src, intptr_t srcStride, int16_t* dst, int width, int height);
>
> CHROMA_420_VERT_FILTERS(_sse2);
> CHROMA_420_HORIZ_FILTERS(_sse4);
> @@ -594,7 +592,6 @@
> CHROMA_444_SP_FILTERS(_sse4);
> CHROMA_444_SS_FILTERS(_sse2);
>
> -void x265_chroma_p2s_ssse3(const pixel* src, intptr_t srcStride, int16_t* dst, int width, int height);
>
> #undef SETUP_CHROMA_FUNC_DEF
> #undef SETUP_CHROMA_SP_FUNC_DEF
> @@ -624,28 +621,6 @@
> LUMA_SP_FILTERS(_avx2);
> LUMA_SS_FILTERS(_avx2);
> void x265_interp_8tap_hv_pp_8x8_sse4(const pixel* src, intptr_t srcStride, pixel* dst, intptr_t dstStride, int idxX, int idxY);
> -void x265_pixelToShort_4x4_ssse3(const pixel* src, intptr_t srcStride, int16_t* dst);
> -void x265_pixelToShort_4x8_ssse3(const pixel* src, intptr_t srcStride, int16_t* dst);
> -void x265_pixelToShort_4x16_ssse3(const pixel* src, intptr_t srcStride, int16_t* dst);
> -void x265_pixelToShort_8x4_ssse3(const pixel* src, intptr_t srcStride, int16_t* dst);
> -void x265_pixelToShort_8x8_ssse3(const pixel* src, intptr_t srcStride, int16_t* dst);
> -void x265_pixelToShort_8x16_ssse3(const pixel* src, intptr_t srcStride, int16_t* dst);
> -void x265_pixelToShort_8x32_ssse3(const pixel* src, intptr_t srcStride, int16_t* dst);
> -void x265_pixelToShort_16x4_ssse3(const pixel* src, intptr_t srcStride, int16_t* dst);
> -void x265_pixelToShort_16x8_ssse3(const pixel* src, intptr_t srcStride, int16_t* dst);
> -void x265_pixelToShort_16x12_ssse3(const pixel* src, intptr_t srcStride, int16_t* dst);
> -void x265_pixelToShort_16x16_ssse3(const pixel* src, intptr_t srcStride, int16_t* dst);
> -void x265_pixelToShort_16x32_ssse3(const pixel* src, intptr_t srcStride, int16_t* dst);
> -void x265_pixelToShort_16x64_ssse3(const pixel* src, intptr_t srcStride, int16_t* dst);
> -void x265_pixelToShort_32x8_ssse3(const pixel* src, intptr_t srcStride, int16_t* dst);
> -void x265_pixelToShort_32x16_ssse3(const pixel* src, intptr_t srcStride, int16_t* dst);
> -void x265_pixelToShort_32x24_ssse3(const pixel* src, intptr_t srcStride, int16_t* dst);
> -void x265_pixelToShort_32x32_ssse3(const pixel* src, intptr_t srcStride, int16_t* dst);
> -void x265_pixelToShort_32x64_ssse3(const pixel* src, intptr_t srcStride, int16_t* dst);
> -void x265_pixelToShort_64x16_ssse3(const pixel* src, intptr_t srcStride, int16_t* dst);
> -void x265_pixelToShort_64x32_ssse3(const pixel* src, intptr_t srcStride, int16_t* dst);
> -void x265_pixelToShort_64x48_ssse3(const pixel* src, intptr_t srcStride, int16_t* dst);
> -void x265_pixelToShort_64x64_ssse3(const pixel* src, intptr_t srcStride, int16_t* dst);
> #undef LUMA_FILTERS
> #undef LUMA_SP_FILTERS
> #undef LUMA_SS_FILTERS
> diff -r 9a5fa67583fe -r 24c96db72960 source/test/ipfilterharness.cpp
> --- a/source/test/ipfilterharness.cpp Thu Apr 02 13:21:32 2015 -0500
> +++ b/source/test/ipfilterharness.cpp Fri Apr 03 18:18:48 2015 +0530
> @@ -61,55 +61,6 @@
> }
> }
>
> -bool IPFilterHarness::check_IPFilter_primitive(filter_p2s_wxh_t ref, filter_p2s_wxh_t opt, int isChroma, int csp)
> -{
> - intptr_t rand_srcStride;
> - int min_size = isChroma ? 2 : 4;
> - int max_size = isChroma ? (MAX_CU_SIZE >> 1) : MAX_CU_SIZE;
> -
> - if (isChroma && (csp == X265_CSP_I444))
> - {
> - min_size = 4;
> - max_size = MAX_CU_SIZE;
> - }
> -
> - for (int i = 0; i < ITERS; i++)
> - {
> - int index = i % TEST_CASES;
> - int rand_height = (int16_t)rand() % 100;
> - int rand_width = (int16_t)rand() % 100;
> -
> - rand_srcStride = rand_width + rand() % 100;
> - if (rand_srcStride < rand_width)
> - rand_srcStride = rand_width;
> -
> - rand_width &= ~(min_size - 1);
> - rand_width = x265_clip3(min_size, max_size, rand_width);
> -
> - rand_height &= ~(min_size - 1);
> - rand_height = x265_clip3(min_size, max_size, rand_height);
> -
> - ref(pixel_test_buff[index],
> - rand_srcStride,
> - IPF_C_output_s,
> - rand_width,
> - rand_height);
> -
> - checked(opt, pixel_test_buff[index],
> - rand_srcStride,
> - IPF_vec_output_s,
> - rand_width,
> - rand_height);
> -
> - if (memcmp(IPF_vec_output_s, IPF_C_output_s, TEST_BUF_SIZE * sizeof(int16_t)))
> - return false;
> -
> - reportfail();
> - }
> -
> - return true;
> -}
> -
> bool IPFilterHarness::check_IPFilterChroma_primitive(filter_pp_t ref, filter_pp_t opt)
> {
> intptr_t rand_srcStride, rand_dstStride;
> @@ -518,12 +469,13 @@
> {
> intptr_t rand_srcStride = rand() % 100;
> int index = i % TEST_CASES;
> + intptr_t dstStride = rand() % 100 + 64;
>
> - ref(pixel_test_buff[index] + i, rand_srcStride, IPF_C_output_s);
> + ref(pixel_test_buff[index] + i, rand_srcStride, IPF_C_output_s, dstStride);
>
> - checked(opt, pixel_test_buff[index] + i, rand_srcStride, IPF_vec_output_s);
> + checked(opt, pixel_test_buff[index] + i, rand_srcStride, IPF_vec_output_s, dstStride);
>
> - if (memcmp(IPF_vec_output_s, IPF_C_output_s, TEST_BUF_SIZE * sizeof(pixel)))
> + if (memcmp(IPF_vec_output_s, IPF_C_output_s, TEST_BUF_SIZE * sizeof(int16_t)))
> return false;
>
> reportfail();
> @@ -538,12 +490,13 @@
> {
> intptr_t rand_srcStride = rand() % 100;
> int index = i % TEST_CASES;
> + intptr_t dstStride = rand() % 100 + 64;
>
> - ref(pixel_test_buff[index] + i, rand_srcStride, IPF_C_output_s);
> + ref(pixel_test_buff[index] + i, rand_srcStride, IPF_C_output_s, dstStride);
>
> - checked(opt, pixel_test_buff[index] + i, rand_srcStride, IPF_vec_output_s);
> + checked(opt, pixel_test_buff[index] + i, rand_srcStride, IPF_vec_output_s, dstStride);
>
> - if (memcmp(IPF_vec_output_s, IPF_C_output_s, TEST_BUF_SIZE * sizeof(pixel)))
> + if (memcmp(IPF_vec_output_s, IPF_C_output_s, TEST_BUF_SIZE * sizeof(int16_t)))
> return false;
>
> reportfail();
> @@ -554,15 +507,6 @@
>
> bool IPFilterHarness::testCorrectness(const EncoderPrimitives& ref, const EncoderPrimitives& opt)
> {
> - if (opt.luma_p2s)
> - {
> - // last parameter does not matter in case of luma
> - if (!check_IPFilter_primitive(ref.luma_p2s, opt.luma_p2s, 0, 1))
> - {
> - printf("luma_p2s failed\n");
> - return false;
> - }
> - }
>
> for (int value = 0; value < NUM_PU_SIZES; value++)
> {
> @@ -622,11 +566,11 @@
> return false;
> }
> }
> - if (opt.pu[value].filter_p2s)
> + if (opt.pu[value].convert_p2s)
> {
> - if (!check_IPFilterLumaP2S_primitive(ref.pu[value].filter_p2s, opt.pu[value].filter_p2s))
> + if (!check_IPFilterLumaP2S_primitive(ref.pu[value].convert_p2s, opt.pu[value].convert_p2s))
> {
> - printf("filter_p2s[%s]", lumaPartStr[value]);
> + printf("convert_p2s[%s]", lumaPartStr[value]);
> return false;
> }
> }
> @@ -634,14 +578,6 @@
>
> for (int csp = X265_CSP_I420; csp < X265_CSP_COUNT; csp++)
> {
> - if (opt.chroma[csp].p2s)
> - {
> - if (!check_IPFilter_primitive(ref.chroma[csp].p2s, opt.chroma[csp].p2s, 1, csp))
> - {
> - printf("chroma_p2s[%s]", x265_source_csp_names[csp]);
> - return false;
> - }
> - }
> for (int value = 0; value < NUM_PU_SIZES; value++)
> {
> if (opt.chroma[csp].pu[value].filter_hpp)
> @@ -692,9 +628,9 @@
> return false;
> }
> }
> - if (opt.chroma[csp].pu[value].chroma_p2s)
> + if (opt.chroma[csp].pu[value].p2s)
> {
> - if (!check_IPFilterChromaP2S_primitive(ref.chroma[csp].pu[value].chroma_p2s, opt.chroma[csp].pu[value].chroma_p2s))
> + if (!check_IPFilterChromaP2S_primitive(ref.chroma[csp].pu[value].p2s, opt.chroma[csp].pu[value].p2s))
> {
> printf("chroma_p2s[%s]", chromaPartStr[csp][value]);
> return false;
> @@ -708,19 +644,10 @@
>
> void IPFilterHarness::measureSpeed(const EncoderPrimitives& ref, const EncoderPrimitives& opt)
> {
> - int height = 64;
> - int width = 64;
> int16_t srcStride = 96;
> int16_t dstStride = 96;
> int maxVerticalfilterHalfDistance = 3;
>
> - if (opt.luma_p2s)
> - {
> - printf("luma_p2s\t");
> - REPORT_SPEEDUP(opt.luma_p2s, ref.luma_p2s,
> - pixel_buff, srcStride, IPF_vec_output_s, width, height);
> - }
> -
> for (int value = 0; value < NUM_PU_SIZES; value++)
> {
> if (opt.pu[value].luma_hpp)
> @@ -777,23 +704,18 @@
> pixel_buff + 3 * srcStride, srcStride, IPF_vec_output_p, srcStride, 1, 3);
> }
>
> - if (opt.pu[value].filter_p2s)
> + if (opt.pu[value].convert_p2s)
> {
> - printf("filter_p2s [%s]\t", lumaPartStr[value]);
> - REPORT_SPEEDUP(opt.pu[value].filter_p2s, ref.pu[value].filter_p2s,
> - pixel_buff, srcStride, IPF_vec_output_s);
> + printf("convert_p2s[%s]\t", lumaPartStr[value]);
> + REPORT_SPEEDUP(opt.pu[value].convert_p2s, ref.pu[value].convert_p2s,
> + pixel_buff, srcStride,
> + IPF_vec_output_s, dstStride);
> }
> }
>
> for (int csp = X265_CSP_I420; csp < X265_CSP_COUNT; csp++)
> {
> printf("= Color Space %s =\n", x265_source_csp_names[csp]);
> - if (opt.chroma[csp].p2s)
> - {
> - printf("chroma_p2s\t");
> - REPORT_SPEEDUP(opt.chroma[csp].p2s, ref.chroma[csp].p2s,
> - pixel_buff, srcStride, IPF_vec_output_s, width, height);
> - }
> for (int value = 0; value < NUM_PU_SIZES; value++)
> {
> if (opt.chroma[csp].pu[value].filter_hpp)
> @@ -836,13 +758,11 @@
> short_buff + maxVerticalfilterHalfDistance * srcStride, srcStride,
> IPF_vec_output_s, dstStride, 1);
> }
> -
> - if (opt.chroma[csp].pu[value].chroma_p2s)
> + if (opt.chroma[csp].pu[value].p2s)
> {
> printf("chroma_p2s[%s]\t", chromaPartStr[csp][value]);
> - REPORT_SPEEDUP(opt.chroma[csp].pu[value].chroma_p2s, ref.chroma[csp].pu[value].chroma_p2s,
> - pixel_buff, srcStride,
> - IPF_vec_output_s);
> + REPORT_SPEEDUP(opt.chroma[csp].pu[value].p2s, ref.chroma[csp].pu[value].p2s,
> + pixel_buff, srcStride, IPF_vec_output_s, dstStride);
> }
> }
> }
> diff -r 9a5fa67583fe -r 24c96db72960 source/test/ipfilterharness.h
> --- a/source/test/ipfilterharness.h Thu Apr 02 13:21:32 2015 -0500
> +++ b/source/test/ipfilterharness.h Fri Apr 03 18:18:48 2015 +0530
> @@ -50,7 +50,6 @@
> pixel pixel_test_buff[TEST_CASES][TEST_BUF_SIZE];
> int16_t short_test_buff[TEST_CASES][TEST_BUF_SIZE];
>
> - bool check_IPFilter_primitive(filter_p2s_wxh_t ref, filter_p2s_wxh_t opt, int isChroma, int csp);
> bool check_IPFilterChroma_primitive(filter_pp_t ref, filter_pp_t opt);
> bool check_IPFilterChroma_ps_primitive(filter_ps_t ref, filter_ps_t opt);
> bool check_IPFilterChroma_hps_primitive(filter_hps_t ref, filter_hps_t opt);
> _______________________________________________
> x265-devel mailing list
> x265-devel at videolan.org
> https://mailman.videolan.org/listinfo/x265-devel
--
Steve Borho
More information about the x265-devel
mailing list