diff options
Diffstat (limited to 'chromium/third_party/dav1d/libdav1d/src/x86/itx_avx2.asm')
| -rw-r--r-- | chromium/third_party/dav1d/libdav1d/src/x86/itx_avx2.asm | 84 |
1 files changed, 35 insertions, 49 deletions
diff --git a/chromium/third_party/dav1d/libdav1d/src/x86/itx_avx2.asm b/chromium/third_party/dav1d/libdav1d/src/x86/itx_avx2.asm index 092c842786d..a67f053a61b 100644 --- a/chromium/third_party/dav1d/libdav1d/src/x86/itx_avx2.asm +++ b/chromium/third_party/dav1d/libdav1d/src/x86/itx_avx2.asm @@ -126,7 +126,7 @@ pw_m2751_3035x8: dw -2751*8, 3035*8 SECTION .text -; Code size reduction trickery: Intead of using rip-relative loads with +; Code size reduction trickery: Instead of using rip-relative loads with ; mandatory 4-byte offsets everywhere, we can set up a base pointer with a ; single rip-relative lea and then address things relative from that with ; 1-byte offsets as long as data is within +-128 bytes of the base pointer. @@ -1194,13 +1194,9 @@ cglobal iidentity_4x16_internal_8bpc, 0, 5, 11, dst, stride, c, eob, tx2 %ifidn %1_%2, dct_dct movd xm1, [o(pw_2896x8)] pmulhrsw xm0, xm1, [cq] + mov [cq], eobd pmulhrsw xm0, xm1 - movd xm2, [o(pw_2048)] - pmulhrsw xm0, xm1 - pmulhrsw xm0, xm2 - vpbroadcastw m0, xm0 - mova m1, m0 - jmp m(iadst_8x4_internal_8bpc).end3 + jmp m(inv_txfm_add_dct_dct_8x8_8bpc).dconly2 %endif %endmacro @@ -1340,20 +1336,20 @@ cglobal iidentity_8x4_internal_8bpc, 0, 5, 7, dst, stride, c, eob, tx2 pmulhrsw xm0, xm1, [cq] movd xm2, [o(pw_16384)] mov [cq], eobd + or r3d, 8 +.dconly: pmulhrsw xm0, xm2 - psrlw xm2, 3 ; pw_2048 +.dconly2: + movd xm2, [pw_2048] pmulhrsw xm0, xm1 + lea r2, [strideq*3] pmulhrsw xm0, xm2 vpbroadcastw m0, xm0 -.end: - mov r2d, 2 -.end2: - lea r3, [strideq*3] -.loop: - WRITE_8X4 0, 0, 1, 2 +.dconly_loop: + WRITE_8X4 0, 0, 1, 2, strideq*1, strideq*2, r2 lea dstq, [dstq+strideq*4] - dec r2d - jg .loop + sub r3d, 4 + jg .dconly_loop RET %endif %endmacro @@ -1543,13 +1539,8 @@ cglobal iidentity_8x8_internal_8bpc, 0, 5, 7, dst, stride, c, eob, tx2 movd xm2, [o(pw_16384)] mov [cq], eobd pmulhrsw xm0, xm1 - pmulhrsw xm0, xm2 - psrlw xm2, 3 ; pw_2048 - pmulhrsw xm0, xm1 - pmulhrsw xm0, xm2 - vpbroadcastw m0, xm0 - mov r2d, 4 - jmp m(inv_txfm_add_dct_dct_8x8_8bpc).end2 + or r3d, 16 + jmp m(inv_txfm_add_dct_dct_8x8_8bpc).dconly %endif %endmacro @@ -1902,7 +1893,7 @@ cglobal iidentity_8x16_internal_8bpc, 0, 5, 13, dst, stride, c, eob, tx2 pmulhrsw xm0, xm1, [cq] movd xm2, [o(pw_16384)] mov [cq], eobd - mov r2d, 2 + or r3d, 4 .dconly: pmulhrsw xm0, xm2 movd xm2, [pw_2048] ; intentionally rip-relative @@ -1911,17 +1902,17 @@ cglobal iidentity_8x16_internal_8bpc, 0, 5, 13, dst, stride, c, eob, tx2 vpbroadcastw m0, xm0 pxor m3, m3 .dconly_loop: - mova xm1, [dstq] - vinserti128 m1, [dstq+strideq], 1 + mova xm1, [dstq+strideq*0] + vinserti128 m1, [dstq+strideq*1], 1 punpckhbw m2, m1, m3 punpcklbw m1, m3 paddw m2, m0 paddw m1, m0 packuswb m1, m2 - mova [dstq], xm1 - vextracti128 [dstq+strideq], m1, 1 + mova [dstq+strideq*0], xm1 + vextracti128 [dstq+strideq*1], m1, 1 lea dstq, [dstq+strideq*2] - dec r2d + sub r3d, 2 jg .dconly_loop RET %endif @@ -2162,7 +2153,7 @@ cglobal iidentity_16x4_internal_8bpc, 0, 5, 11, dst, stride, c, eob, tx2 movd xm2, [o(pw_16384)] mov [cq], eobd pmulhrsw xm0, xm1 - mov r2d, 4 + or r3d, 8 jmp m(inv_txfm_add_dct_dct_16x4_8bpc).dconly %endif %endmacro @@ -2473,7 +2464,7 @@ cglobal iidentity_16x8_internal_8bpc, 0, 5, 13, dst, stride, c, eob, tx2 pmulhrsw xm0, xm1, [cq] movd xm2, [o(pw_8192)] mov [cq], eobd - mov r2d, 8 + or r3d, 16 jmp m(inv_txfm_add_dct_dct_16x4_8bpc).dconly %endif %endmacro @@ -3120,13 +3111,8 @@ cglobal inv_txfm_add_dct_dct_8x32_8bpc, 4, 4, 0, dst, stride, c, eob pmulhrsw xm0, xm1, [cq] movd xm2, [o(pw_8192)] mov [cq], eobd - pmulhrsw xm0, xm2 - psrlw xm2, 2 ; pw_2048 - pmulhrsw xm0, xm1 - pmulhrsw xm0, xm2 - vpbroadcastw m0, xm0 - mov r2d, 8 - jmp m(inv_txfm_add_dct_dct_8x8_8bpc).end2 + or r3d, 32 + jmp m(inv_txfm_add_dct_dct_8x8_8bpc).dconly .full: REPX {pmulhrsw x, m9}, m12, m13, m14, m15 pmulhrsw m6, m9, [rsp+32*2] @@ -3290,7 +3276,7 @@ cglobal inv_txfm_add_dct_dct_32x8_8bpc, 4, 4, 0, dst, stride, c, eob pmulhrsw xm0, xm1, [cq] movd xm2, [o(pw_8192)] mov [cq], eobd - mov r2d, 8 + or r3d, 8 .dconly: pmulhrsw xm0, xm2 movd xm2, [pw_2048] ; intentionally rip-relative @@ -3307,7 +3293,7 @@ cglobal inv_txfm_add_dct_dct_32x8_8bpc, 4, 4, 0, dst, stride, c, eob packuswb m1, m2 mova [dstq], m1 add dstq, strideq - dec r2d + dec r3d jg .dconly_loop RET .normal: @@ -3672,7 +3658,7 @@ cglobal inv_txfm_add_dct_dct_16x32_8bpc, 4, 4, 0, dst, stride, c, eob movd xm2, [o(pw_16384)] mov [cq], eobd pmulhrsw xm0, xm1 - mov r2d, 16 + or r3d, 32 jmp m(inv_txfm_add_dct_dct_16x4_8bpc).dconly .full: mova [tmp1q-32*4], m1 @@ -3991,7 +3977,7 @@ cglobal inv_txfm_add_dct_dct_32x16_8bpc, 4, 4, 0, dst, stride, c, eob movd xm2, [o(pw_16384)] mov [cq], eobd pmulhrsw xm0, xm1 - mov r2d, 16 + or r3d, 16 jmp m(inv_txfm_add_dct_dct_32x8_8bpc).dconly .normal: PROLOGUE 0, 6, 16, 32*19, dst, stride, c, eob, tmp1, tmp2 @@ -4222,7 +4208,7 @@ cglobal inv_txfm_add_dct_dct_32x32_8bpc, 4, 4, 0, dst, stride, c, eob pmulhrsw xm0, xm1, [cq] movd xm2, [o(pw_8192)] mov [cq], eobd - mov r2d, 32 + or r3d, 32 jmp m(inv_txfm_add_dct_dct_32x8_8bpc).dconly .normal: PROLOGUE 0, 9, 16, 32*67, dst, stride, c, eob, tmp1, tmp2, \ @@ -4486,7 +4472,7 @@ cglobal inv_txfm_add_dct_dct_16x64_8bpc, 4, 4, 0, dst, stride, c, eob pmulhrsw xm0, xm1, [cq] movd xm2, [o(pw_8192)] mov [cq], eobd - mov r2d, 32 + or r3d, 64 jmp m(inv_txfm_add_dct_dct_16x4_8bpc).dconly .normal: PROLOGUE 0, 10, 16, 32*67, dst, stride, c, eob, tmp1, tmp2 @@ -4832,7 +4818,7 @@ cglobal inv_txfm_add_dct_dct_64x16_8bpc, 4, 4, 0, dst, stride, c, eob pmulhrsw xm0, xm1, [cq] movd xm2, [o(pw_8192)] mov [cq], eobd - mov r2d, 16 + or r3d, 16 .dconly: pmulhrsw xm0, xm2 movd xm2, [o(pw_2048)] @@ -4856,7 +4842,7 @@ cglobal inv_txfm_add_dct_dct_64x16_8bpc, 4, 4, 0, dst, stride, c, eob mova [dstq+32*0], m2 mova [dstq+32*1], m3 add dstq, strideq - dec r2d + dec r3d jg .dconly_loop RET .normal: @@ -4997,7 +4983,7 @@ cglobal inv_txfm_add_dct_dct_32x64_8bpc, 4, 4, 0, dst, stride, c, eob movd xm2, [o(pw_16384)] mov [cq], eobd pmulhrsw xm0, xm1 - mov r2d, 64 + or r3d, 64 jmp m(inv_txfm_add_dct_dct_32x8_8bpc).dconly .normal: PROLOGUE 0, 11, 16, 32*99, dst, stride, c, eob, tmp1, tmp2 @@ -5200,7 +5186,7 @@ cglobal inv_txfm_add_dct_dct_64x32_8bpc, 4, 4, 0, dst, stride, c, eob movd xm2, [o(pw_16384)] mov [cq], eobd pmulhrsw xm0, xm1 - mov r2d, 32 + or r3d, 32 jmp m(inv_txfm_add_dct_dct_64x16_8bpc).dconly .normal: PROLOGUE 0, 9, 16, 32*131, dst, stride, c, eob, tmp1, tmp2, \ @@ -5381,7 +5367,7 @@ cglobal inv_txfm_add_dct_dct_64x64_8bpc, 4, 4, 0, dst, stride, c, eob pmulhrsw xm0, xm1, [cq] movd xm2, [o(pw_8192)] mov [cq], eobd - mov r2d, 64 + or r3d, 64 jmp m(inv_txfm_add_dct_dct_64x16_8bpc).dconly .normal: PROLOGUE 0, 11, 16, 32*199, dst, stride, c, eob, tmp1, tmp2 |