From: Simon Pilgrim <llvm-dev@redking.me.uk>
Date: Fri, 12 Oct 2018 10:26:59 +0000 (+0000)
Subject: [X86][AVX] Add examples of shuffles that can be reduced to a cross-lane shuffle follo... 
X-Git-Tag: android-x86-9.0-r1~11903
X-Git-Url: http://git.osdn.net/view?a=commitdiff_plain;h=0b50ad3f83305d5904d9f0ffa94424134a7b6d82;p=android-x86%2Fexternal-llvm.git

[X86][AVX] Add examples of shuffles that can be reduced to a cross-lane shuffle followed by a in-lane permute

Suitable for lowering by D53148

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@344332 91177308-0d34-0410-b5e6-96231b3b80d8
---

diff --git a/test/CodeGen/X86/vector-shuffle-256-v16.ll b/test/CodeGen/X86/vector-shuffle-256-v16.ll
index de587beadc1..90970f15fea 100644
--- a/test/CodeGen/X86/vector-shuffle-256-v16.ll
+++ b/test/CodeGen/X86/vector-shuffle-256-v16.ll
@@ -4052,6 +4052,32 @@ define <16 x i16> @shuffle_v16i16_05_06_07_16_17_18_19_28_13_14_15_24_25_26_27_2
   ret <16 x i16> %shuffle
 }
 
+define <16 x i16> @shuffle_v16i16_00_02_04_06_01_03_05_07_31_30_29_28_27_26_25_24(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: shuffle_v16i16_00_02_04_06_01_03_05_07_31_30_29_28_27_26_25_24:
+; AVX1:       # %bb.0:
+; AVX1-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,1,4,5,8,9,12,13,2,3,6,7,10,11,14,15]
+; AVX1-NEXT:    vextractf128 $1, %ymm1, %xmm1
+; AVX1-NEXT:    vpshufb {{.*#+}} xmm1 = xmm1[14,15,12,13,10,11,8,9,6,7,4,5,2,3,0,1]
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: shuffle_v16i16_00_02_04_06_01_03_05_07_31_30_29_28_27_26_25_24:
+; AVX2:       # %bb.0:
+; AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[0,1,4,5,8,9,12,13,2,3,6,7,10,11,14,15]
+; AVX2-NEXT:    vextracti128 $1, %ymm1, %xmm1
+; AVX2-NEXT:    vpshufb {{.*#+}} xmm1 = xmm1[14,15,12,13,10,11,8,9,6,7,4,5,2,3,0,1]
+; AVX2-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm0
+; AVX2-NEXT:    retq
+;
+; AVX512VL-LABEL: shuffle_v16i16_00_02_04_06_01_03_05_07_31_30_29_28_27_26_25_24:
+; AVX512VL:       # %bb.0:
+; AVX512VL-NEXT:    vmovdqa {{.*#+}} ymm2 = [0,2,4,6,1,3,5,7,31,30,29,28,27,26,25,24]
+; AVX512VL-NEXT:    vpermt2w %ymm1, %ymm2, %ymm0
+; AVX512VL-NEXT:    retq
+  %shuffle = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <i32 0, i32 2, i32 4, i32 6, i32 1, i32 3, i32 5, i32 7, i32 31, i32 30, i32 29, i32 28, i32 27, i32 26, i32 25, i32 24>
+  ret <16 x i16> %shuffle
+}
+
 define <16 x i16> @shuffle_v16i16_uu_06_uu_uu_17_18_19_uu_uu_14_uu_uu_25_26_27_uu(<16 x i16> %a, <16 x i16> %b) {
 ; AVX1-LABEL: shuffle_v16i16_uu_06_uu_uu_17_18_19_uu_uu_14_uu_uu_25_26_27_uu:
 ; AVX1:       # %bb.0:
diff --git a/test/CodeGen/X86/vector-shuffle-256-v32.ll b/test/CodeGen/X86/vector-shuffle-256-v32.ll
index 3e36b4a3b6a..5e9f30a727d 100644
--- a/test/CodeGen/X86/vector-shuffle-256-v32.ll
+++ b/test/CodeGen/X86/vector-shuffle-256-v32.ll
@@ -2495,6 +2495,37 @@ define <32 x i8> @shuffle_v32i8_00_32_01_33_02_34_03_35_04_36_05_37_06_38_07_39_
   ret <32 x i8> %shuffle
 }
 
+define <32 x i8> @shuffle_v32i8_15_14_13_12_11_10_09_08_07_06_05_04_03_02_01_00_32_34_36_38_40_42_44_46_33_35_37_39_41_43_45_47(<32 x i8> %a, <32 x i8> %b) {
+; AVX1-LABEL: shuffle_v32i8_15_14_13_12_11_10_09_08_07_06_05_04_03_02_01_00_32_34_36_38_40_42_44_46_33_35_37_39_41_43_45_47:
+; AVX1:       # %bb.0:
+; AVX1-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[15,14,13,12,11,10,9,8,7,6,5,4,3,2,1,0]
+; AVX1-NEXT:    vpshufb {{.*#+}} xmm1 = xmm1[0,2,4,6,8,10,12,14,1,3,5,7,9,11,13,15]
+; AVX1-NEXT:    vinsertf128 $1, %xmm1, %ymm0, %ymm0
+; AVX1-NEXT:    retq
+;
+; AVX2-LABEL: shuffle_v32i8_15_14_13_12_11_10_09_08_07_06_05_04_03_02_01_00_32_34_36_38_40_42_44_46_33_35_37_39_41_43_45_47:
+; AVX2:       # %bb.0:
+; AVX2-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[15,14,13,12,11,10,9,8,7,6,5,4,3,2,1,0]
+; AVX2-NEXT:    vpshufb {{.*#+}} xmm1 = xmm1[0,2,4,6,8,10,12,14,1,3,5,7,9,11,13,15]
+; AVX2-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm0
+; AVX2-NEXT:    retq
+;
+; AVX512VLBW-LABEL: shuffle_v32i8_15_14_13_12_11_10_09_08_07_06_05_04_03_02_01_00_32_34_36_38_40_42_44_46_33_35_37_39_41_43_45_47:
+; AVX512VLBW:       # %bb.0:
+; AVX512VLBW-NEXT:    vpshufb {{.*#+}} xmm0 = xmm0[15,14,13,12,11,10,9,8,7,6,5,4,3,2,1,0]
+; AVX512VLBW-NEXT:    vpshufb {{.*#+}} xmm1 = xmm1[0,2,4,6,8,10,12,14,1,3,5,7,9,11,13,15]
+; AVX512VLBW-NEXT:    vinserti128 $1, %xmm1, %ymm0, %ymm0
+; AVX512VLBW-NEXT:    retq
+;
+; AVX512VLVBMI-LABEL: shuffle_v32i8_15_14_13_12_11_10_09_08_07_06_05_04_03_02_01_00_32_34_36_38_40_42_44_46_33_35_37_39_41_43_45_47:
+; AVX512VLVBMI:       # %bb.0:
+; AVX512VLVBMI-NEXT:    vmovdqa {{.*#+}} ymm2 = [15,14,13,12,11,10,9,8,7,6,5,4,3,2,1,0,32,34,36,38,40,42,44,46,33,35,37,39,41,43,45,47]
+; AVX512VLVBMI-NEXT:    vpermt2b %ymm1, %ymm2, %ymm0
+; AVX512VLVBMI-NEXT:    retq
+  %shuffle = shufflevector <32 x i8> %a, <32 x i8> %b, <32 x i32> <i32 15, i32 14, i32 13, i32 12, i32 11, i32 10, i32 9, i32 8, i32 7, i32 6, i32 5, i32 4, i32 3, i32 2, i32 1, i32 0, i32 32, i32 34, i32 36, i32 38, i32 40, i32 42, i32 44, i32 46, i32 33, i32 35, i32 37, i32 39, i32 41, i32 43, i32 45, i32 47>
+  ret <32 x i8> %shuffle
+}
+
 define <32 x i8> @shuffle_v32i8_zz_zz_zz_zz_zz_zz_zz_zz_zz_zz_zz_zz_zz_zz_zz_32_zz_zz_zz_zz_zz_zz_zz_zz_zz_zz_zz_zz_zz_zz_zz_48(<32 x i8> %a) {
 ; AVX1-LABEL: shuffle_v32i8_zz_zz_zz_zz_zz_zz_zz_zz_zz_zz_zz_zz_zz_zz_zz_32_zz_zz_zz_zz_zz_zz_zz_zz_zz_zz_zz_zz_zz_zz_zz_48:
 ; AVX1:       # %bb.0: