# HG changeset patch # User Vignesh Vijayakumar<vign...@multicorewareinc.com>> # Date 1522976950 25200 # Thu Apr 05 18:09:10 2018 -0700 # Node ID ab41c6957bc2f359e5df82f9936c3fd00a5d2ea5 # Parent 71f7869fac602953ef5e14c344f10adc374d7bfa x86: AVX512 optimise intermediate register load in chroma_vsp, chroma_vss, chroma_vps
diff -r 71f7869fac60 -r ab41c6957bc2 source/common/x86/ipfilter16.asm
--- a/source/common/x86/ipfilter16.asm Wed Nov 15 14:35:17 2017 +0530
+++ b/source/common/x86/ipfilter16.asm Thu Apr 05 18:09:10 2018 -0700
@@ -7527,10 +7527,10 @@
vinserti32x4 m4, [r8 + 4 * r1], 2
vinserti32x4 m4, [r9 + 4 * r1], 3
punpcklwd m6, m5, m4
- pmaddwd m6, [r5 + mmsize]
+ pmaddwd m6, m9
paddd m2, m6
punpckhwd m5, m4
- pmaddwd m5, [r5 + mmsize]
+ pmaddwd m5, m9
paddd m3, m5
paddd m0, m7
@@ -7561,7 +7561,7 @@
;-----------------------------------------------------------------------------------------------------------------
%if ARCH_X86_64
INIT_ZMM avx512
-cglobal interp_4tap_vert_ps_8x8, 5, 11, 8
+cglobal interp_4tap_vert_ps_8x8, 5, 11, 10
add r1d, r1d
add r3d, r3d
sub r0, r1
@@ -7576,13 +7576,15 @@
vbroadcasti32x4 m7, [INTERP_OFFSET_PS]
lea r10, [3 * r1]
lea r7, [3 * r3]
+ mova m8, [r5]
+ mova m9, [r5 + mmsize]
PROCESS_CHROMA_VERT_PS_8x8_AVX512
RET
%endif
%macro FILTER_VER_PS_CHROMA_8xN_AVX512 1
INIT_ZMM avx512
-cglobal interp_4tap_vert_ps_8x%1, 5, 11, 8
+cglobal interp_4tap_vert_ps_8x%1, 5, 11, 10
add r1d, r1d
add r3d, r3d
sub r0, r1
@@ -7597,6 +7599,8 @@
vbroadcasti32x4 m7, [INTERP_OFFSET_PS]
lea r10, [3 * r1]
lea r7, [3 * r3]
+ mova m8, [r5]
+ mova m9, [r5 + mmsize]
%rep %1/8 - 1
PROCESS_CHROMA_VERT_PS_8x8_AVX512
lea r0, [r8 + 4 * r1]
@@ -7619,33 +7623,33 @@
movu ym3, [r0 + r1]
vinserti32x8 m3, [r6 + r1], 1
punpcklwd m0, m1, m3
- pmaddwd m0, [r5]
+ pmaddwd m0, m8
punpckhwd m1, m3
- pmaddwd m1, [r5]
+ pmaddwd m1, m8
movu ym4, [r0 + 2 * r1]
vinserti32x8 m4, [r6 + 2 * r1], 1
punpcklwd m2, m3, m4
- pmaddwd m2, [r5]
+ pmaddwd m2, m8
punpckhwd m3, m4
- pmaddwd m3, [r5]
+ pmaddwd m3, m8
movu ym5, [r0 + r8]
vinserti32x8 m5, [r6 + r8], 1
punpcklwd m6, m4, m5
- pmaddwd m6, [r5 + mmsize]
+ pmaddwd m6, m9
paddd m0, m6
punpckhwd m4, m5
- pmaddwd m4, [r5 + mmsize]
+ pmaddwd m4, m9
paddd m1, m4
movu ym4, [r0 + 4 * r1]
vinserti32x8 m4, [r6 + 4 * r1], 1
punpcklwd m6, m5, m4
- pmaddwd m6, [r5 + mmsize]
+ pmaddwd m6, m9
paddd m2, m6
punpckhwd m5, m4
- pmaddwd m5, [r5 + mmsize]
+ pmaddwd m5, m9
paddd m3, m5
paddd m0, m7
@@ -7671,7 +7675,7 @@
;-----------------------------------------------------------------------------------------------------------------
%if ARCH_X86_64
INIT_ZMM avx512
-cglobal interp_4tap_vert_ps_16x4, 5, 9, 8
+cglobal interp_4tap_vert_ps_16x4, 5, 9, 10
add r1d, r1d
add r3d, r3d
sub r0, r1
@@ -7686,13 +7690,15 @@
vbroadcasti32x4 m7, [INTERP_OFFSET_PS]
lea r7, [3 * r3]
lea r8, [3 * r1]
+ mova m8, [r5]
+ mova m9, [r5 + mmsize]
PROCESS_CHROMA_VERT_PS_16x4_AVX512
RET
%endif
%macro FILTER_VER_PS_CHROMA_16xN_AVX512 1
INIT_ZMM avx512
-cglobal interp_4tap_vert_ps_16x%1, 5, 9, 8
+cglobal interp_4tap_vert_ps_16x%1, 5, 9, 10
add r1d, r1d
add r3d, r3d
sub r0, r1
@@ -7707,6 +7713,8 @@
vbroadcasti32x4 m7, [INTERP_OFFSET_PS]
lea r7, [3 * r3]
lea r8, [3 * r1]
+ mova m8, [r5]
+ mova m9, [r5 + mmsize]
%rep %1/4 - 1
PROCESS_CHROMA_VERT_PS_16x4_AVX512
lea r0, [r0 + 4 * r1]
@@ -7741,12 +7749,12 @@
punpcklwd m0, m1, m3
punpcklwd m9, m10, m12
- pmaddwd m0, [r5]
- pmaddwd m9, [r5]
+ pmaddwd m0, m16
+ pmaddwd m9, m16
punpckhwd m1, m3
punpckhwd m10, m12
- pmaddwd m1, [r5]
- pmaddwd m10, [r5]
+ pmaddwd m1, m16
+ pmaddwd m10, m16
movu ym4, [r0 + 2 * r1]
movu ym13, [r8 + 2 * r1]
@@ -7754,12 +7762,12 @@
vinserti32x8 m13, [r9 + 2 * r1], 1
punpcklwd m2, m3, m4
punpcklwd m11, m12, m13
- pmaddwd m2, [r5]
- pmaddwd m11, [r5]
+ pmaddwd m2, m16
+ pmaddwd m11, m16
punpckhwd m3, m4
punpckhwd m12, m13
- pmaddwd m3, [r5]
- pmaddwd m12, [r5]
+ pmaddwd m3, m16
+ pmaddwd m12, m16
movu ym5, [r0 + r10]
vinserti32x8 m5, [r6 + r10], 1
@@ -7767,14 +7775,14 @@
vinserti32x8 m14, [r9 + r10], 1
punpcklwd m6, m4, m5
punpcklwd m15, m13, m14
- pmaddwd m6, [r5 + mmsize]
- pmaddwd m15, [r5 + mmsize]
+ pmaddwd m6, m17
+ pmaddwd m15, m17
paddd m0, m6
paddd m9, m15
punpckhwd m4, m5
punpckhwd m13, m14
- pmaddwd m4, [r5 + mmsize]
- pmaddwd m13, [r5 + mmsize]
+ pmaddwd m4, m17
+ pmaddwd m13, m17
paddd m1, m4
paddd m10, m13
@@ -7784,14 +7792,14 @@
vinserti32x8 m13, [r9 + 4 * r1], 1
punpcklwd m6, m5, m4
punpcklwd m15, m14, m13
- pmaddwd m6, [r5 + mmsize]
- pmaddwd m15, [r5 + mmsize]
+ pmaddwd m6, m17
+ pmaddwd m15, m17
paddd m2, m6
paddd m11, m15
punpckhwd m5, m4
punpckhwd m14, m13
- pmaddwd m5, [r5 + mmsize]
- pmaddwd m14, [r5 + mmsize]
+ pmaddwd m5, m17
+ pmaddwd m14, m17
paddd m3, m5
paddd m12, m14
@@ -7836,28 +7844,28 @@
vinserti32x4 m3, [r8 + r1 + mmsize/2], 2
vinserti32x4 m3, [r9 + r1 + mmsize/2], 3
punpcklwd m0, m1,
m3
- pmaddwd m0, [r5]
+ pmaddwd m0, m16
punpckhwd m1, m3
- pmaddwd m1, [r5]
+ pmaddwd m1, m16
movu xm4, [r0 + 2 * r1 + mmsize/2]
vinserti32x4 m4, [r6 + 2 * r1 + mmsize/2], 1
vinserti32x4 m4, [r8 + 2 * r1 + mmsize/2], 2
vinserti32x4 m4, [r9 + 2 * r1 + mmsize/2], 3
punpcklwd m2, m3,
m4
- pmaddwd m2, [r5]
+ pmaddwd m2, m16
punpckhwd m3, m4
- pmaddwd m3, [r5]
+ pmaddwd m3, m16
movu xm5, [r0 + r10 + mmsize/2]
vinserti32x4 m5, [r6 + r10 + mmsize/2], 1
vinserti32x4 m5, [r8 + r10 + mmsize/2], 2
vinserti32x4 m5, [r9 + r10 + mmsize/2], 3
punpcklwd m6, m4,
m5
- pmaddwd m6, [r5 + mmsize]
+ pmaddwd m6, m17
paddd m0, m6
punpckhwd m4, m5
- pmaddwd m4, [r5 + mmsize]
+ pmaddwd m4, m17
paddd m1, m4
movu xm4, [r0 + 4 * r1 + mmsize/2]
@@ -7865,10 +7873,10 @@
vinserti32x4 m4, [r8 + 4 * r1 + mmsize/2], 2
vinserti32x4 m4, [r9 + 4 * r1 + mmsize/2], 3
punpcklwd m6, m5,
m4
- pmaddwd m6, [r5 + mmsize]
+ pmaddwd m6, m17
paddd m2, m6
punpckhwd m5, m4
- pmaddwd m5, [r5 + mmsize]
+ pmaddwd m5, m17
paddd m3, m5
paddd m0, m7
@@ -7896,7 +7904,7 @@
%macro FILTER_VER_PS_CHROMA_24xN_AVX512 1
INIT_ZMM avx512
-cglobal interp_4tap_vert_ps_24x%1, 5, 12, 16
+cglobal interp_4tap_vert_ps_24x%1, 5, 12, 18
add r1d, r1d
add r3d, r3d
sub r0, r1
@@ -7911,6 +7919,8 @@
vbroadcasti32x4 m7, [INTERP_OFFSET_PS]
lea r10, [3 * r1]
lea r7, [3 * r3]
+ mova m16, [r5]
+ mova m17, [r5 + mmsize]
%rep %1/8 - 1
PROCESS_CHROMA_VERT_PS_24x8_AVX512
lea r0, [r8 + 4 * r1]
@@ -7929,31 +7939,31 @@
movu m1, [r0]
movu m3, [r0 + r1]
punpcklwd m0, m1, m3
- pmaddwd m0, [r5]
+ pmaddwd m0, m9
punpckhwd m1, m3
- pmaddwd m1, [r5]
+ pmaddwd m1, m9
movu m4, [r0 + 2 * r1]
punpcklwd m2, m3, m4
- pmaddwd m2, [r5]
+ pmaddwd m2, m9
punpckhwd m3, m4
- pmaddwd m3, [r5]
+ pmaddwd m3, m9
lea r0, [r0 + 2 * r1]
movu m5, [r0 + r1]
punpcklwd m6, m4, m5
- pmaddwd m6, [r5 + mmsize]
+ pmaddwd m6, m10
paddd m0, m6
punpckhwd m4, m5
- pmaddwd m4, [r5 + mmsize]
+ pmaddwd m4, m10
paddd m1, m4
movu m4, [r0 + 2 * r1]
punpcklwd m6, m5, m4
- pmaddwd m6, [r5 + mmsize]
+ pmaddwd m6, m10
paddd m2, m6
punpckhwd m5, m4
- pmaddwd m5, [r5 + mmsize]
+ pmaddwd m5, m10
paddd m3, m5
paddd m0, m7
@@ -7976,7 +7986,7 @@
;-----------------------------------------------------------------------------------------------------------------
%macro FILTER_VER_PS_CHROMA_32xN_AVX512 1
INIT_ZMM avx512
-cglobal interp_4tap_vert_ps_32x%1, 5, 7, 9
+cglobal interp_4tap_vert_ps_32x%1, 5, 7, 11
add r1d, r1d
add r3d, r3d
sub r0, r1
@@ -7989,7 +7999,8 @@
lea r5, [tab_ChromaCoeffV_avx512 + r4]
%endif
vbroadcasti32x4 m7, [INTERP_OFFSET_PS]
-
+ mova m9, [r5]
+ mova m10, [r5 + mmsize]
%rep %1/2 - 1
PROCESS_CHROMA_VERT_PS_32x2_AVX512
lea r2, [r2 + 2 * r3]
@@ -8015,36 +8026,36 @@
movu m12, [r6 + r1]
punpcklwd m0, m1, m3
punpcklwd m9, m10, m12
- pmaddwd m0, [r5]
- pmaddwd m9, [r5]
+ pmaddwd m0, m16
+ pmaddwd m9, m16
punpckhwd m1, m3
punpckhwd m10, m12
- pmaddwd m1, [r5]
- pmaddwd m10, [r5]
+ pmaddwd m1, m16
+ pmaddwd m10, m16
movu m4, [r0 + 2 * r1]
movu m13, [r6 + 2 * r1]
punpcklwd m2, m3, m4
punpcklwd m11, m12, m13
- pmaddwd m2, [r5]
- pmaddwd m11, [r5]
+ pmaddwd m2, m16
+ pmaddwd m11, m16
punpckhwd m3, m4
punpckhwd m12, m13
- pmaddwd m3, [r5]
- pmaddwd m12, [r5]
+ pmaddwd m3, m16
+ pmaddwd m12, m16
movu m5, [r0 + r7]
movu m14, [r6 + r7]
punpcklwd m6, m4, m5
punpcklwd m15, m13, m14
- pmaddwd m6, [r5 + mmsize]
- pmaddwd m15, [r5 + mmsize]
+ pmaddwd m6, m17
+ pmaddwd m15, m17
paddd m0, m6
paddd m9, m15
punpckhwd m4, m5
punpckhwd m13, m14
- pmaddwd m4, [r5 + mmsize]
- pmaddwd m13, [r5 + mmsize]
+ pmaddwd m4, m17
+ pmaddwd m13, m17
paddd m1, m4
paddd m10, m13
@@ -8052,14 +8063,14 @@
movu m13, [r6 + 4 * r1]
punpcklwd m6, m5, m4
punpcklwd m15, m14, m13
- pmaddwd m6, [r5 + mmsize]
- pmaddwd m15, [r5 + mmsize]
+ pmaddwd m6, m17
+ pmaddwd m15, m17
paddd m2, m6
paddd m11, m15
punpckhwd m5, m4
punpckhwd m14, m13
- pmaddwd m5, [r5 + mmsize]
- pmaddwd m14, [r5 + mmsize]
+ pmaddwd m5, m17
+ pmaddwd m14, m17
paddd m3, m5
paddd m12, m14
@@ -8095,33 +8106,33 @@
movu ym3, [r0 + r1 + mmsize]
vinserti32x8 m3, [r6 + r1 + mmsize], 1
punpcklwd m0, m1, m3
- pmaddwd m0, [r5]
+ pmaddwd m0, m16
punpckhwd m1, m3
- pmaddwd m1, [r5]
+ pmaddwd m1, m16
movu ym4, [r0 + 2 * r1 + mmsize]
vinserti32x8 m4, [r6 + 2 * r1 + mmsize], 1
punpcklwd m2, m3, m4
- pmaddwd m2, [r5]
+ pmaddwd m2, m16
punpckhwd m3, m4
- pmaddwd m3, [r5]
+ pmaddwd m3, m16
movu ym5, [r0 + r7 + mmsize]
vinserti32x8 m5, [r6 + r7 + mmsize], 1
punpcklwd m6, m4, m5
- pmaddwd m6, [r5 + mmsize]
+ pmaddwd m6, m17
paddd m0, m6
punpckhwd m4, m5
- pmaddwd m4, [r5 + mmsize]
+ pmaddwd m4, m17
paddd m1, m4
movu ym4, [r0 + 4 * r1 + mmsize]
vinserti32x8 m4, [r6 + 4 * r1 + mmsize], 1
punpcklwd m6, m5, m4
- pmaddwd m6, [r5 + mmsize]
+ pmaddwd m6, m17
paddd m2, m6
punpckhwd m5, m4
- pmaddwd m5, [r5 + mmsize]
+ pmaddwd m5, m17
paddd m3, m5
paddd m0, m7
@@ -8144,7 +8155,7 @@
%if ARCH_X86_64
INIT_ZMM avx512
-cglobal interp_4tap_vert_ps_48x64, 5, 9, 16
+cglobal interp_4tap_vert_ps_48x64, 5, 9, 18
add r1d, r1d
add r3d, r3d
sub r0, r1
@@ -8158,7 +8169,8 @@
lea r7, [3 * r1]
lea r8, [3 * r3]
vbroadcasti32x4 m7, [INTERP_OFFSET_PS]
-
+ mova m16, [r5]
+ mova m17, [r5 + mmsize]
%rep 15
PROCESS_CHROMA_VERT_PS_48x4_AVX512
lea r0, [r0 + 4 * r1]
@@ -8172,60 +8184,60 @@
movu m1, [r0]
movu m3, [r0 + r1]
punpcklwd m0, m1, m3
- pmaddwd m0, [r5]
+ pmaddwd m0, m15
punpckhwd m1, m3
- pmaddwd m1, [r5]
+ pmaddwd m1, m15
movu m9, [r0 + mmsize]
movu m11, [r0 + r1 + mmsize]
punpcklwd m8, m9, m11
- pmaddwd m8, [r5]
+ pmaddwd m8, m15
punpckhwd m9, m11
- pmaddwd m9, [r5]
+ pmaddwd m9, m15
movu m4, [r0 + 2 * r1]
punpcklwd m2, m3, m4
- pmaddwd m2, [r5]
+ pmaddwd m2, m15
punpckhwd m3, m4
- pmaddwd m3, [r5]
+ pmaddwd m3, m15
movu m12, [r0 + 2 * r1 + mmsize]
punpcklwd m10, m11, m12
- pmaddwd m10, [r5]
+ pmaddwd m10, m15
punpckhwd m11, m12
- pmaddwd m11, [r5]
+ pmaddwd m11, m15
lea r0, [r0 + 2 * r1]
movu m5, [r0 + r1]
punpcklwd m6, m4, m5
- pmaddwd m6, [r5 + 1 * mmsize]
+ pmaddwd m6, m16
paddd m0, m6
punpckhwd m4, m5
- pmaddwd m4, [r5 + 1 * mmsize]
+ pmaddwd m4, m16
paddd m1, m4
movu m13, [r0 + r1 + mmsize]
punpcklwd m14, m12, m13
- pmaddwd m14, [r5 + 1 * mmsize]
+ pmaddwd m14, m16
paddd m8, m14
punpckhwd m12, m13
- pmaddwd m12, [r5 + 1 * mmsize]
+ pmaddwd m12, m16
paddd m9, m12
movu m4, [r0 + 2 * r1]
punpcklwd m6, m5, m4
- pmaddwd m6, [r5 + 1 * mmsize]
+ pmaddwd m6, m16
paddd m2, m6
punpckhwd m5, m4
- pmaddwd m5, [r5 + 1 * mmsize]
+ pmaddwd m5, m16
paddd m3, m5
movu m12, [r0 + 2 * r1 + mmsize]
punpcklwd m14, m13, m12
- pmaddwd m14, [r5 + 1 * mmsize]
+ pmaddwd m14, m16
paddd m10, m14
punpckhwd m13, m12
- pmaddwd m13, [r5 + 1 * mmsize]
+ pmaddwd m13, m16
paddd m11, m13
paddd m0, m7
@@ -8261,7 +8273,7 @@
;-----------------------------------------------------------------------------------------------------------------
%macro FILTER_VER_PS_CHROMA_64xN_AVX512 1
INIT_ZMM avx512
-cglobal interp_4tap_vert_ps_64x%1, 5, 7, 15
+cglobal interp_4tap_vert_ps_64x%1, 5, 7, 17
add r1d, r1d
add r3d, r3d
sub r0, r1
@@ -8274,6 +8286,8 @@
lea r5, [tab_ChromaCoeffV_avx512 + r4]
%endif
vbroadcasti32x4 m7, [INTERP_OFFSET_PS]
+ mova m15, [r5]
+ mova m16, [r5 + mmsize]
%rep %1/2 - 1
PROCESS_CHROMA_VERT_PS_64x2_AVX512
@@ -8308,28 +8322,28 @@
vinserti32x4 m3, [r8 + r1], 2
vinserti32x4 m3, [r9 + r1], 3
punpcklwd m0, m1, m3
- pmaddwd m0, [r5]
+ pmaddwd m0, m8
punpckhwd m1, m3
- pmaddwd m1, [r5]
+ pmaddwd m1, m8
movu xm4, [r0 + 2 * r1]
vinserti32x4 m4, [r6 + 2 * r1], 1
vinserti32x4 m4, [r8 + 2 * r1], 2
vinserti32x4 m4, [r9 + 2 * r1], 3
punpcklwd m2, m3, m4
- pmaddwd m2, [r5]
+ pmaddwd m2, m8
punpckhwd m3, m4
- pmaddwd m3, [r5]
+ pmaddwd m3, m8
movu xm5, [r0 + r10]
vinserti32x4 m5, [r6 + r10], 1
vinserti32x4 m5, [r8 + r10], 2
vinserti32x4 m5, [r9 + r10], 3
punpcklwd m6, m4, m5
- pmaddwd m6, [r5 + mmsize]
+ pmaddwd m6, m9
paddd m0, m6
punpckhwd m4, m5
- pmaddwd m4, [r5 + mmsize]
+ pmaddwd m4, m9
paddd m1, m4
movu xm4, [r0 + 4 * r1]
@@ -8337,10 +8351,10 @@
vinserti32x4 m4, [r8 + 4 * r1], 2
vinserti32x4 m4, [r9 + 4 * r1], 3
punpcklwd m6, m5, m4
- pmaddwd m6, [r5 + mmsize]
+ pmaddwd m6, m9
paddd m2, m6
punpckhwd m5, m4
- pmaddwd m5, [r5 + mmsize]
+ pmaddwd m5, m9
paddd m3, m5
%ifidn %1,sp
_______________________________________________
x265-devel mailing list
x265-devel@videolan.org
https://mailman.videolan.org/listinfo/x265-devel
