[llvm] [X86] Add v16i16 shuffle tests for "shift + blend" pack patterns (PR #208481)
Simon Pilgrim via llvm-commits
llvm-commits at lists.llvm.org
Thu Jul 9 08:10:12 PDT 2026
https://github.com/RKSimon created https://github.com/llvm/llvm-project/pull/208481
Shuffles that could be expanded by shifting one input and blending with the second inplace input
>From b505c36dc62a7ca95be6ea7338ca81681f945b5b Mon Sep 17 00:00:00 2001
From: Simon Pilgrim <llvm-dev at redking.me.uk>
Date: Thu, 9 Jul 2026 15:54:53 +0100
Subject: [PATCH] [X86] Add v16i16 shuffle tests for "shift + blend" pack
patterns
Shuffles that be expanded by shifting one input and blending with the second inplace input
---
.../CodeGen/X86/vector-shuffle-256-v16.ll | 100 ++++++++++++++++++
1 file changed, 100 insertions(+)
diff --git a/llvm/test/CodeGen/X86/vector-shuffle-256-v16.ll b/llvm/test/CodeGen/X86/vector-shuffle-256-v16.ll
index 4f07ff9f02c6f..32f995d488e1d 100644
--- a/llvm/test/CodeGen/X86/vector-shuffle-256-v16.ll
+++ b/llvm/test/CodeGen/X86/vector-shuffle-256-v16.ll
@@ -6975,6 +6975,106 @@ define <16 x i16> @shuffle_v16i16_9_9_9_9_9_9_9_9_u_u_u_u_u_u_u_u(<16 x i16> %a,
ret <16 x i16> %shuffle
}
+define <16 x i16> @shuffle_v16i16_00_16_02_18_04_20_06_22_08_24_10_26_12_28_14_30(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: shuffle_v16i16_00_16_02_18_04_20_06_22_08_24_10_26_12_28_14_30:
+; AVX1: # %bb.0:
+; AVX1-NEXT: vextractf128 $1, %ymm1, %xmm2
+; AVX1-NEXT: vmovdqa {{.*#+}} xmm3 = [0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15]
+; AVX1-NEXT: vpshufb %xmm3, %xmm2, %xmm2
+; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm4
+; AVX1-NEXT: vpshufb %xmm3, %xmm4, %xmm4
+; AVX1-NEXT: vpunpcklwd {{.*#+}} xmm2 = xmm4[0],xmm2[0],xmm4[1],xmm2[1],xmm4[2],xmm2[2],xmm4[3],xmm2[3]
+; AVX1-NEXT: vpshufb %xmm3, %xmm1, %xmm1
+; AVX1-NEXT: vpshufb %xmm3, %xmm0, %xmm0
+; AVX1-NEXT: vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX1-NEXT: vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; AVX1-NEXT: retq
+;
+; AVX2-LABEL: shuffle_v16i16_00_16_02_18_04_20_06_22_08_24_10_26_12_28_14_30:
+; AVX2: # %bb.0:
+; AVX2-NEXT: vmovdqa {{.*#+}} ymm2 = [0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15,16,17,20,21,24,25,28,29,24,25,28,29,28,29,30,31]
+; AVX2-NEXT: vpshufb %ymm2, %ymm1, %ymm1
+; AVX2-NEXT: vpshufb %ymm2, %ymm0, %ymm0
+; AVX2-NEXT: vpunpcklwd {{.*#+}} ymm0 = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[8],ymm1[8],ymm0[9],ymm1[9],ymm0[10],ymm1[10],ymm0[11],ymm1[11]
+; AVX2-NEXT: retq
+;
+; AVX512VL-LABEL: shuffle_v16i16_00_16_02_18_04_20_06_22_08_24_10_26_12_28_14_30:
+; AVX512VL: # %bb.0:
+; AVX512VL-NEXT: vpmovsxbw {{.*#+}} ymm2 = [0,16,2,18,4,20,6,22,8,24,10,26,12,28,14,30]
+; AVX512VL-NEXT: vpermt2w %ymm1, %ymm2, %ymm0
+; AVX512VL-NEXT: retq
+;
+; XOPAVX1-LABEL: shuffle_v16i16_00_16_02_18_04_20_06_22_08_24_10_26_12_28_14_30:
+; XOPAVX1: # %bb.0:
+; XOPAVX1-NEXT: vextractf128 $1, %ymm1, %xmm2
+; XOPAVX1-NEXT: vextractf128 $1, %ymm0, %xmm3
+; XOPAVX1-NEXT: vmovdqa {{.*#+}} xmm4 = [0,1,16,17,4,5,20,21,8,9,24,25,12,13,28,29]
+; XOPAVX1-NEXT: vpperm %xmm4, %xmm2, %xmm3, %xmm2
+; XOPAVX1-NEXT: vpperm %xmm4, %xmm1, %xmm0, %xmm0
+; XOPAVX1-NEXT: vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; XOPAVX1-NEXT: retq
+;
+; XOPAVX2-LABEL: shuffle_v16i16_00_16_02_18_04_20_06_22_08_24_10_26_12_28_14_30:
+; XOPAVX2: # %bb.0:
+; XOPAVX2-NEXT: vmovdqa {{.*#+}} ymm2 = [0,1,4,5,8,9,12,13,8,9,12,13,12,13,14,15,16,17,20,21,24,25,28,29,24,25,28,29,28,29,30,31]
+; XOPAVX2-NEXT: vpshufb %ymm2, %ymm1, %ymm1
+; XOPAVX2-NEXT: vpshufb %ymm2, %ymm0, %ymm0
+; XOPAVX2-NEXT: vpunpcklwd {{.*#+}} ymm0 = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[8],ymm1[8],ymm0[9],ymm1[9],ymm0[10],ymm1[10],ymm0[11],ymm1[11]
+; XOPAVX2-NEXT: retq
+ %r = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <i32 0, i32 16, i32 2, i32 18, i32 4, i32 20, i32 6, i32 22, i32 8, i32 24, i32 10, i32 26, i32 12, i32 28, i32 14, i32 30>
+ ret <16 x i16> %r
+}
+
+define <16 x i16> @shuffle_v16i16_01_17_03_19_05_21_07_23_09_25_11_27_13_29_15_31(<16 x i16> %a, <16 x i16> %b) {
+; AVX1-LABEL: shuffle_v16i16_01_17_03_19_05_21_07_23_09_25_11_27_13_29_15_31:
+; AVX1: # %bb.0:
+; AVX1-NEXT: vextractf128 $1, %ymm1, %xmm2
+; AVX1-NEXT: vmovdqa {{.*#+}} xmm3 = [2,3,6,7,10,11,14,15,14,15,10,11,12,13,14,15]
+; AVX1-NEXT: vpshufb %xmm3, %xmm2, %xmm2
+; AVX1-NEXT: vextractf128 $1, %ymm0, %xmm4
+; AVX1-NEXT: vpshufb %xmm3, %xmm4, %xmm4
+; AVX1-NEXT: vpunpcklwd {{.*#+}} xmm2 = xmm4[0],xmm2[0],xmm4[1],xmm2[1],xmm4[2],xmm2[2],xmm4[3],xmm2[3]
+; AVX1-NEXT: vpshufb %xmm3, %xmm1, %xmm1
+; AVX1-NEXT: vpshufb %xmm3, %xmm0, %xmm0
+; AVX1-NEXT: vpunpcklwd {{.*#+}} xmm0 = xmm0[0],xmm1[0],xmm0[1],xmm1[1],xmm0[2],xmm1[2],xmm0[3],xmm1[3]
+; AVX1-NEXT: vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; AVX1-NEXT: retq
+;
+; AVX2-LABEL: shuffle_v16i16_01_17_03_19_05_21_07_23_09_25_11_27_13_29_15_31:
+; AVX2: # %bb.0:
+; AVX2-NEXT: vmovdqa {{.*#+}} ymm2 = [2,3,6,7,10,11,14,15,14,15,10,11,12,13,14,15,18,19,22,23,26,27,30,31,30,31,26,27,28,29,30,31]
+; AVX2-NEXT: vpshufb %ymm2, %ymm1, %ymm1
+; AVX2-NEXT: vpshufb %ymm2, %ymm0, %ymm0
+; AVX2-NEXT: vpunpcklwd {{.*#+}} ymm0 = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[8],ymm1[8],ymm0[9],ymm1[9],ymm0[10],ymm1[10],ymm0[11],ymm1[11]
+; AVX2-NEXT: retq
+;
+; AVX512VL-LABEL: shuffle_v16i16_01_17_03_19_05_21_07_23_09_25_11_27_13_29_15_31:
+; AVX512VL: # %bb.0:
+; AVX512VL-NEXT: vpmovsxbw {{.*#+}} ymm2 = [1,17,3,19,5,21,7,23,9,25,11,27,13,29,15,31]
+; AVX512VL-NEXT: vpermt2w %ymm1, %ymm2, %ymm0
+; AVX512VL-NEXT: retq
+;
+; XOPAVX1-LABEL: shuffle_v16i16_01_17_03_19_05_21_07_23_09_25_11_27_13_29_15_31:
+; XOPAVX1: # %bb.0:
+; XOPAVX1-NEXT: vextractf128 $1, %ymm1, %xmm2
+; XOPAVX1-NEXT: vextractf128 $1, %ymm0, %xmm3
+; XOPAVX1-NEXT: vmovdqa {{.*#+}} xmm4 = [2,3,18,19,6,7,22,23,10,11,26,27,14,15,30,31]
+; XOPAVX1-NEXT: vpperm %xmm4, %xmm2, %xmm3, %xmm2
+; XOPAVX1-NEXT: vpperm %xmm4, %xmm1, %xmm0, %xmm0
+; XOPAVX1-NEXT: vinsertf128 $1, %xmm2, %ymm0, %ymm0
+; XOPAVX1-NEXT: retq
+;
+; XOPAVX2-LABEL: shuffle_v16i16_01_17_03_19_05_21_07_23_09_25_11_27_13_29_15_31:
+; XOPAVX2: # %bb.0:
+; XOPAVX2-NEXT: vmovdqa {{.*#+}} ymm2 = [2,3,6,7,10,11,14,15,14,15,10,11,12,13,14,15,18,19,22,23,26,27,30,31,30,31,26,27,28,29,30,31]
+; XOPAVX2-NEXT: vpshufb %ymm2, %ymm1, %ymm1
+; XOPAVX2-NEXT: vpshufb %ymm2, %ymm0, %ymm0
+; XOPAVX2-NEXT: vpunpcklwd {{.*#+}} ymm0 = ymm0[0],ymm1[0],ymm0[1],ymm1[1],ymm0[2],ymm1[2],ymm0[3],ymm1[3],ymm0[8],ymm1[8],ymm0[9],ymm1[9],ymm0[10],ymm1[10],ymm0[11],ymm1[11]
+; XOPAVX2-NEXT: retq
+ %r = shufflevector <16 x i16> %a, <16 x i16> %b, <16 x i32> <i32 1, i32 17, i32 3, i32 19, i32 5, i32 21, i32 7, i32 23, i32 9, i32 25, i32 11, i32 27, i32 13, i32 29, i32 15, i32 31>
+ ret <16 x i16> %r
+}
+
define <16 x i16> @shuffle_v16i16_02_18_03_19_00_16_01_17_10_26_11_27_08_24_09_25(<16 x i16> %a0, <16 x i16> %a1) {
; AVX1-LABEL: shuffle_v16i16_02_18_03_19_00_16_01_17_10_26_11_27_08_24_09_25:
; AVX1: # %bb.0:
More information about the llvm-commits
mailing list