[x265] [PATCH] asm: new algorithm for intra_ang_16 modes 3 & 33, improved over 23%
rajesh at multicorewareinc.com
rajesh at multicorewareinc.com
Thu Aug 6 09:45:44 CEST 2015
# HG changeset patch
# User Rajesh Paulraj<rajesh at multicorewareinc.com>
# Date 1438846342 -19800
# Thu Aug 06 13:02:22 2015 +0530
# Node ID 4078c3fa7b2a362cdab1b1ea54e13a29ae0ef4f2
# Parent 377a996a8d74110f838ff2e3cef1c42781d6d730
asm: new algorithm for intra_ang_16 modes 3 & 33, improved over 23%
than previous avx2 code
diff -r 377a996a8d74 -r 4078c3fa7b2a source/common/x86/intrapred8.asm
--- a/source/common/x86/intrapred8.asm Wed Aug 05 15:09:14 2015 +0530
+++ b/source/common/x86/intrapred8.asm Thu Aug 06 13:02:22 2015 +0530
@@ -294,32 +294,6 @@
db 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16
ALIGN 32
-c_ang16_mode_33: db 6, 26, 6, 26, 6, 26, 6, 26, 6, 26, 6, 26, 6, 26, 6, 26, 6, 26, 6, 26, 6, 26, 6, 26, 6, 26, 6, 26, 6, 26, 6, 26
- db 12, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12, 20
- db 18, 14, 18, 14, 18, 14, 18, 14, 18, 14, 18, 14, 18, 14, 18, 14, 18, 14, 18, 14, 18, 14, 18, 14, 18, 14, 18, 14, 18, 14, 18, 14
- db 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 24, 8
- db 30, 2, 30, 2, 30, 2, 30, 2, 30, 2, 30, 2, 30, 2, 30, 2, 4, 28, 4, 28, 4, 28, 4, 28, 4, 28, 4, 28, 4, 28, 4, 28
- db 10, 22, 10, 22, 10, 22, 10, 22, 10, 22, 10, 22, 10, 22, 10, 22, 10, 22, 10, 22, 10, 22, 10, 22, 10, 22, 10, 22, 10, 22, 10, 22
- db 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16
- db 22, 10, 22, 10, 22, 10, 22, 10, 22, 10, 22, 10, 22, 10, 22, 10, 22, 10, 22, 10, 22, 10, 22, 10, 22, 10, 22, 10, 22, 10, 22, 10
- db 28, 4, 28, 4, 28, 4, 28, 4, 28, 4, 28, 4, 28, 4, 28, 4, 2, 30, 2, 30, 2, 30, 2, 30, 2, 30, 2, 30, 2, 30, 2, 30
- db 8, 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 24
- db 14, 18, 14, 18, 14, 18, 14, 18, 14, 18, 14, 18, 14, 18, 14, 18, 14, 18, 14, 18, 14, 18, 14, 18, 14, 18, 14, 18, 14, 18, 14, 18
- db 20, 12, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12
- db 26, 6, 26, 6, 26, 6, 26, 6, 26, 6, 26, 6, 26, 6, 26, 6, 26, 6, 26, 6, 26, 6, 26, 6, 26, 6, 26, 6, 26, 6, 26, 6
- db 32, 0, 32, 0, 32, 0, 32, 0, 32, 0, 32, 0, 32, 0, 32, 0, 32, 0, 32, 0, 32, 0, 32, 0, 32, 0, 32, 0, 32, 0, 32, 0
-
-ALIGN 32
-c_ang16_mode_3: db 6, 26, 6, 26, 6, 26, 6, 26, 6, 26, 6, 26, 6, 26, 6, 26, 22, 10, 22, 10, 22, 10, 22, 10, 22, 10, 22, 10, 22, 10, 22, 10
- db 12, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12, 20, 28, 4, 28, 4, 28, 4, 28, 4, 28, 4, 28, 4, 28, 4, 28, 4
- db 18, 14, 18, 14, 18, 14, 18, 14, 18, 14, 18, 14, 18, 14, 18, 14, 2, 30, 2, 30, 2, 30, 2, 30, 2, 30, 2, 30, 2, 30, 2, 30
- db 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 8, 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 24, 8, 24
- db 30, 2, 30, 2, 30, 2, 30, 2, 30, 2, 30, 2, 30, 2, 30, 2, 14, 18, 14, 18, 14, 18, 14, 18, 14, 18, 14, 18, 14, 18, 14, 18
- db 4, 28, 4, 28, 4, 28, 4, 28, 4, 28, 4, 28, 4, 28, 4, 28, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12
- db 10, 22, 10, 22, 10, 22, 10, 22, 10, 22, 10, 22, 10, 22, 10, 22, 26, 6, 26, 6, 26, 6, 26, 6, 26, 6, 26, 6, 26, 6, 26, 6
- db 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 16, 32, 0, 32, 0, 32, 0, 32, 0, 32, 0, 32, 0, 32, 0, 32, 0
-
-ALIGN 32
c_ang16_mode_24: db 5, 27, 5, 27, 5, 27, 5, 27, 5, 27, 5, 27, 5, 27, 5, 27, 10, 22, 10, 22, 10, 22, 10, 22, 10, 22, 10, 22, 10, 22, 10, 22
db 15, 17, 15, 17, 15, 17, 15, 17, 15, 17, 15, 17, 15, 17, 15, 17, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12, 20, 12
db 25, 7, 25, 7, 25, 7, 25, 7, 25, 7, 25, 7, 25, 7, 25, 7, 30, 2, 30, 2, 30, 2, 30, 2, 30, 2, 30, 2, 30, 2, 30, 2
@@ -13534,131 +13508,226 @@
INTRA_PRED_TRANS_STORE_16x16
RET
-
-INIT_YMM avx2
-cglobal intra_pred_ang16_3, 3, 6, 12
- mova m11, [pw_1024]
- lea r5, [intra_pred_shuff_0_8]
-
- movu xm9, [r2 + 1 + 32]
- pshufb xm9, [r5]
- movu xm10, [r2 + 9 + 32]
- pshufb xm10, [r5]
-
- movu xm7, [r2 + 8 + 32]
- pshufb xm7, [r5]
- vinserti128 m9, m9, xm7, 1
-
- movu xm8, [r2 + 16 + 32]
- pshufb xm8, [r5]
- vinserti128 m10, m10, xm8, 1
-
- lea r3, [3 * r1]
- lea r4, [c_ang16_mode_3]
-
- INTRA_PRED_ANG16_CAL_ROW m0, m1, 0
-
- movu xm9, [r2 + 2 + 32]
- pshufb xm9, [r5]
- movu xm10, [r2 + 10 + 32]
- pshufb xm10, [r5]
-
- movu xm7, [r2 + 9 + 32]
- pshufb xm7, [r5]
- vinserti128 m9, m9, xm7, 1
-
- movu xm8, [r2 + 17 + 32]
- pshufb xm8, [r5]
- vinserti128 m10, m10, xm8, 1
-
- INTRA_PRED_ANG16_CAL_ROW m1, m2, 1
-
- movu xm7, [r2 + 3 + 32]
- pshufb xm7, [r5]
- vinserti128 m9, m9, xm7, 0
-
- movu xm8, [r2 + 11 + 32]
- pshufb xm8, [r5]
- vinserti128 m10, m10, xm8, 0
-
- INTRA_PRED_ANG16_CAL_ROW m2, m3, 2
-
- movu xm9, [r2 + 4 + 32]
- pshufb xm9, [r5]
- movu xm10, [r2 + 12 + 32]
- pshufb xm10, [r5]
-
- movu xm7, [r2 + 10 + 32]
- pshufb xm7, [r5]
- vinserti128 m9, m9, xm7, 1
-
- movu xm8, [r2 + 18 + 32]
- pshufb xm8, [r5]
- vinserti128 m10, m10, xm8, 1
-
- INTRA_PRED_ANG16_CAL_ROW m3, m4, 3
-
- movu xm9, [r2 + 5 + 32]
- pshufb xm9, [r5]
- movu xm10, [r2 + 13 + 32]
- pshufb xm10, [r5]
-
- movu xm7, [r2 + 11 + 32]
- pshufb xm7, [r5]
- vinserti128 m9, m9, xm7, 1
-
- movu xm8, [r2 + 19 + 32]
- pshufb xm8, [r5]
- vinserti128 m10, m10, xm8, 1
-
- add r4, 4 * mmsize
-
- INTRA_PRED_ANG16_CAL_ROW m4, m5, 0
-
- movu xm7, [r2 + 12 + 32]
- pshufb xm7, [r5]
- vinserti128 m9, m9, xm7, 1
-
- movu xm8, [r2 + 20 + 32]
- pshufb xm8, [r5]
- vinserti128 m10, m10, xm8, 1
-
- INTRA_PRED_ANG16_CAL_ROW m5, m6, 1
-
- movu xm9, [r2 + 6 + 32]
- pshufb xm9, [r5]
- movu xm10, [r2 + 14 + 32]
- pshufb xm10, [r5]
-
- movu xm7, [r2 + 13 + 32]
- pshufb xm7, [r5]
- vinserti128 m9, m9, xm7, 1
-
- movu xm8, [r2 + 21 + 32]
- pshufb xm8, [r5]
- vinserti128 m10, m10, xm8, 1
-
- INTRA_PRED_ANG16_CAL_ROW m6, m7, 2
-
- movu xm9, [r2 + 7 + 32]
- pshufb xm9, [r5]
- movu xm10, [r2 + 15 + 32]
- pshufb xm10, [r5]
-
- movu xm7, [r2 + 14 + 32]
- pshufb xm7, [r5]
- vinserti128 m9, m9, xm7, 1
-
- movu xm8, [r2 + 22 + 32]
- pshufb xm8, [r5]
- vinserti128 m10, m10, xm8, 1
-
- INTRA_PRED_ANG16_CAL_ROW m7, m8, 3
-
- ; transpose and store
- INTRA_PRED_TRANS_STORE_16x16
- RET
-
+; transpose 8x32 to 16x16, used for intra_ang16x16 avx2 asm
+%if ARCH_X86_64 == 1
+INIT_YMM avx2
+%macro TRANSPOSE_STORE_8x32 12
+ jc .skip
+
+ punpcklbw m%9, m%1, m%2
+ punpckhbw m%1, m%2
+ punpcklbw m%10, m%3, m%4
+ punpckhbw m%3, m%4
+
+ punpcklwd m%11, m%9, m%10
+ punpckhwd m%9, m%10
+ punpcklwd m%10, m%1, m%3
+ punpckhwd m%1, m%3
+
+ punpckldq m%12, m%11, m%10
+ punpckhdq m%11, m%10
+ punpckldq m%10, m%9, m%1
+ punpckhdq m%9, m%1
+
+ punpcklbw m%1, m%5, m%6
+ punpckhbw m%5, m%6
+ punpcklbw m%2, m%7, m%8
+ punpckhbw m%7, m%8
+
+ punpcklwd m%3, m%1, m%2
+ punpckhwd m%1, m%2
+ punpcklwd m%4, m%5, m%7
+ punpckhwd m%5, m%7
+
+ punpckldq m%2, m%3, m%4
+ punpckhdq m%3, m%4
+ punpckldq m%4, m%1, m%5
+ punpckhdq m%1, m%5
+
+ punpckldq m%5, m%12, m%2
+ punpckhdq m%6, m%12, m%2
+ punpckldq m%7, m%10, m%4
+ punpckhdq m%8, m%10, m%4
+
+ punpckldq m%2, m%11, m%3
+ punpckhdq m%11, m%11, m%3
+ punpckldq m%4, m%9, m%1
+ punpckhdq m%9, m%9, m%1
+
+ movu [r0 + r1 * 0], xm%5
+ movu [r0 + r1 * 1], xm%6
+ movu [r0 + r1 * 2], xm%2
+ movu [r0 + r5 * 1], xm%11
+
+ lea r0, [r0 + r6]
+
+ movu [r0 + r1 * 0], xm%7
+ movu [r0 + r1 * 1], xm%8
+ movu [r0 + r1 * 2], xm%4
+ movu [r0 + r5 * 1], xm%9
+
+ lea r0, [r0 + r6]
+
+ vextracti128 [r0 + r1 * 0], m%5, 1
+ vextracti128 [r0 + r1 * 1], m%6, 1
+ vextracti128 [r0 + r1 * 2], m%2, 1
+ vextracti128 [r0 + r5 * 1], m%11, 1
+
+ lea r0, [r0 + r6]
+
+ vextracti128 [r0 + r1 * 0], m%7, 1
+ vextracti128 [r0 + r1 * 1], m%8, 1
+ vextracti128 [r0 + r1 * 2], m%4, 1
+ vextracti128 [r0 + r5 * 1], m%9, 1
+ jmp .end
+
+.skip:
+ vpermq m%1, m%1, q3120
+ vpermq m%2, m%2, q3120
+ vpermq m%3, m%3, q3120
+ vpermq m%4, m%4, q3120
+ vpermq m%5, m%5, q3120
+ vpermq m%6, m%6, q3120
+ vpermq m%7, m%7, q3120
+ vpermq m%8, m%8, q3120
+
+ movu [r0 + r1 * 0], xm%1
+ movu [r0 + r1 * 1], xm%2
+ movu [r0 + r1 * 2], xm%3
+ movu [r0 + r5 * 1], xm%4
+
+ lea r0, [r0 + r6]
+
+ movu [r0 + r1 * 0], xm%5
+ movu [r0 + r1 * 1], xm%6
+ movu [r0 + r1 * 2], xm%7
+ movu [r0 + r5 * 1], xm%8
+
+ lea r0, [r0 + r6]
+
+ vextracti128 [r0 + r1 * 0], m%1, 1
+ vextracti128 [r0 + r1 * 1], m%2, 1
+ vextracti128 [r0 + r1 * 2], m%3, 1
+ vextracti128 [r0 + r5 * 1], m%4, 1
+
+ lea r0, [r0 + r6]
+
+ vextracti128 [r0 + r1 * 0], m%5, 1
+ vextracti128 [r0 + r1 * 1], m%6, 1
+ vextracti128 [r0 + r1 * 2], m%7, 1
+ vextracti128 [r0 + r5 * 1], m%8, 1
+.end:
+%endmacro
+
+cglobal ang16_mode_3_33
+ ; rows 0 to 7
+ movu m0, [r2 + 1] ; [32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1]
+ movu m1, [r2 + 2] ; [33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2]
+
+ punpckhbw m2, m0, m1 ; [33 32 32 31 31 30 30 29 29 28 28 27 27 26 26 25 17 16 16 15 15 14 14 13 13 12 12 11 11 10 10 9]
+ punpcklbw m0, m1 ; [25 24 24 23 23 22 22 21 21 20 20 19 19 18 18 17 9 8 8 7 7 6 6 5 5 4 4 3 3 2 2 1]
+ vextracti128 xm1, m0, 1
+ vperm2i128 m0, m0, m2, 0x20 ; [17 16 16 15 15 14 14 13 13 12 12 11 11 10 10 9 9 8 8 7 7 6 6 5 5 4 4 3 3 2 2 1]
+ vperm2i128 m2, m2, m1, 0x20 ; [25 24 24 23 23 22 22 21 21 20 20 19 19 18 18 17 17 16 16 15 15 14 14 13 13 12 12 11 11 10 10 9]
+
+ pmaddubsw m4, m0, [r3 + 10 * 32] ; [26]
+ pmulhrsw m4, m7
+
+ palignr m5, m2, m0, 2
+ pmaddubsw m5, [r3 + 4 * 32] ; [20]
+ pmulhrsw m5, m7
+
+ palignr m6, m2, m0, 4
+ palignr m8, m2, m0, 6
+ pmaddubsw m6, [r3 - 2 * 32] ; [14]
+ pmulhrsw m6, m7
+ pmaddubsw m8, [r3 - 8 * 32] ; [8]
+ pmulhrsw m8, m7
+
+ palignr m10, m2, m0, 8
+ pmaddubsw m9, m10, [r3 - 14 * 32] ; [2]
+ pmulhrsw m9, m7
+ pmaddubsw m10, [r3 + 12 * 32] ; [28]
+ pmulhrsw m10, m7
+
+ palignr m11, m2, m0, 10
+ palignr m12, m2, m0, 12
+ pmaddubsw m11, [r3 + 6 * 32] ; [22]
+ pmulhrsw m11, m7
+ pmaddubsw m12, [r3] ; [16]
+ pmulhrsw m12, m7
+
+ ; rows 8 to 15
+ palignr m3, m2, m0, 14
+ palignr m1, m1, m2, 14
+ pmaddubsw m3, [r3 - 6 * 32] ; [10]
+ pmulhrsw m3, m7
+ packuswb m4, m3
+
+ pmaddubsw m3, m2, [r3 - 12 * 32] ; [4]
+ pmulhrsw m3, m7
+ packuswb m5, m3
+
+ pmaddubsw m3, m2, [r3 + 14 * 32] ; [30]
+ pmulhrsw m3, m7
+ packuswb m6, m3
+
+ movu xm0, [r2 + 25]
+ movu xm1, [r2 + 26]
+ punpcklbw m0, m1
+ mova m1, m2
+ vinserti128 m1, m1, xm0, 0
+ vpermq m1, m1, 01001110b
+
+ palignr m3, m1, m2, 2
+ pmaddubsw m3, [r3 + 8 * 32] ; [24]
+ pmulhrsw m3, m7
+ packuswb m8, m3
+
+ palignr m3, m1, m2, 4
+ pmaddubsw m3, [r3 + 2 * 32] ; [18]
+ pmulhrsw m3, m7
+ packuswb m9, m3
+
+ palignr m3, m1, m2, 6
+ pmaddubsw m3, [r3 - 4 * 32] ; [12]
+ pmulhrsw m3, m7
+ packuswb m10, m3
+
+ palignr m3, m1, m2, 8
+ pmaddubsw m3, [r3 - 10 * 32] ; [6]
+ pmulhrsw m3, m7
+ packuswb m11, m3
+
+ pmovzxbw m1, [r2 + 14]
+ packuswb m12, m1
+
+ TRANSPOSE_STORE_8x32 4, 5, 6, 8, 9, 10, 11, 12, 0, 1, 2, 3
+ ret
+
+INIT_YMM avx2
+cglobal intra_pred_ang16_3, 3, 7, 13
+ add r2, 32
+ lea r3, [ang_table_avx2 + 16 * 32]
+ lea r5, [r1 * 3] ; r5 -> 3 * stride
+ lea r6, [r1 * 4] ; r6 -> 4 * stride
+ mova m7, [pw_1024]
+ clc
+
+ call ang16_mode_3_33
+ RET
+
+INIT_YMM avx2
+cglobal intra_pred_ang16_33, 3, 7, 13
+ lea r3, [ang_table_avx2 + 16 * 32]
+ lea r5, [r1 * 3] ; r5 -> 3 * stride
+ lea r6, [r1 * 4] ; r6 -> 4 * stride
+ mova m7, [pw_1024]
+ stc
+
+ call ang16_mode_3_33
+ RET
+%endif ; ARCH_X86_64
INIT_YMM avx2
cglobal intra_pred_ang16_4, 3, 6, 12
@@ -14358,75 +14427,6 @@
RET
INIT_YMM avx2
-cglobal intra_pred_ang16_33, 3, 5, 6
- mova m0, [pw_1024]
- mova m5, [intra_pred_shuff_0_8]
- lea r3, [3 * r1]
- lea r4, [c_ang16_mode_33]
-
- INTRA_PRED_ANG16_MC2 1
- vperm2i128 m1, m1, m2, 00100000b
- pmaddubsw m3, m1, [r4 + 0 * mmsize]
- pmulhrsw m3, m0
-
- INTRA_PRED_ANG16_MC2 2
- INTRA_PRED_ANG16_MC4 r0, r0 + r1, 1
-
- INTRA_PRED_ANG16_MC2 3
- vperm2i128 m1, m1, m2, 00100000b
- pmaddubsw m3, m1, [r4 + 2 * mmsize]
- pmulhrsw m3, m0
-
- INTRA_PRED_ANG16_MC2 4
- INTRA_PRED_ANG16_MC4 r0 + 2 * r1, r0 + r3, 3
-
- lea r0, [r0 + 4 * r1]
- add r4, 4 * mmsize
-
- INTRA_PRED_ANG16_MC2 5
- INTRA_PRED_ANG16_MC0 r0, r0 + r1, 0
-
- INTRA_PRED_ANG16_MC2 6
- vperm2i128 m1, m1, m2, 00100000b
- pmaddubsw m3, m1, [r4 + 1 * mmsize]
- pmulhrsw m3, m0
-
- INTRA_PRED_ANG16_MC2 7
- INTRA_PRED_ANG16_MC4 r0 + 2 * r1, r0 + r3, 2
-
- INTRA_PRED_ANG16_MC2 8
- lea r0, [r0 + 4 * r1]
- INTRA_PRED_ANG16_MC3 r0, 3
-
- INTRA_PRED_ANG16_MC2 9
- add r4, 4 * mmsize
- INTRA_PRED_ANG16_MC0 r0 + r1, r0 + 2 * r1, 0
-
- INTRA_PRED_ANG16_MC2 10
- vperm2i128 m1, m1, m2, 00100000b
- pmaddubsw m3, m1, [r4 + 1 * mmsize]
- pmulhrsw m3, m0
-
- INTRA_PRED_ANG16_MC2 11
- INTRA_PRED_ANG16_MC4 r0 + r3, r0 + 4 * r1, 2
-
- lea r0, [r0 + 4 * r1]
-
- INTRA_PRED_ANG16_MC2 12
- vperm2i128 m1, m1, m2, 00100000b
- pmaddubsw m3, m1, [r4 + 3 * mmsize]
- pmulhrsw m3, m0
-
- INTRA_PRED_ANG16_MC2 13
- INTRA_PRED_ANG16_MC4 r0 + r1, r0 + 2 * r1, 4
-
- add r4, 4 * mmsize
-
- INTRA_PRED_ANG16_MC2 14
- INTRA_PRED_ANG16_MC3 r0 + r3, 1
- RET
-
-INIT_YMM avx2
cglobal intra_pred_ang16_24, 3, 5, 6
mova m0, [pw_1024]
mova m5, [intra_pred_shuff_0_8]
More information about the x265-devel
mailing list