[llvm] [PhaseOrdering][X86] Copy backend horizontal min/max reduction tests to phaseordering (PR #194601)
Simon Pilgrim via llvm-commits
llvm-commits at lists.llvm.org
Tue Apr 28 04:47:10 PDT 2026
https://github.com/RKSimon created https://github.com/llvm/llvm-project/pull/194601
As discussed on #194473 - add middleend test coverage to ensure we're creating vXi8/vXi16 llvm.vector.reduce calls to ensure we can lower to PHMINPOS instructions
Also demonstrates that we're still not matching partial reduction patterns in vectorcombine
>From 7dff8471daeddc7844e2aeac211ac76fd064b634 Mon Sep 17 00:00:00 2001
From: Simon Pilgrim <llvm-dev at redking.me.uk>
Date: Tue, 28 Apr 2026 12:45:39 +0100
Subject: [PATCH] [PhaseOrdering][X86] Copy backend horizontal min/max
reduction tests to phaseordering
As discussed on #194473 - add middleend test coverage to ensure we're creating vXi8/vXi16 llvm.vector.reduce calls to ensure we can lower to PHMINPOS instructions
Also demonstrates that we're still not matching partial reduction patterns in vectorcombine
---
.../X86/horizontal-reduce-smax.ll | 419 ++++++++++++++++++
.../X86/horizontal-reduce-smin.ll | 419 ++++++++++++++++++
.../X86/horizontal-reduce-umax.ll | 419 ++++++++++++++++++
.../X86/horizontal-reduce-umin.ll | 419 ++++++++++++++++++
4 files changed, 1676 insertions(+)
create mode 100644 llvm/test/Transforms/PhaseOrdering/X86/horizontal-reduce-smax.ll
create mode 100644 llvm/test/Transforms/PhaseOrdering/X86/horizontal-reduce-smin.ll
create mode 100644 llvm/test/Transforms/PhaseOrdering/X86/horizontal-reduce-umax.ll
create mode 100644 llvm/test/Transforms/PhaseOrdering/X86/horizontal-reduce-umin.ll
diff --git a/llvm/test/Transforms/PhaseOrdering/X86/horizontal-reduce-smax.ll b/llvm/test/Transforms/PhaseOrdering/X86/horizontal-reduce-smax.ll
new file mode 100644
index 0000000000000..85186dba0891f
--- /dev/null
+++ b/llvm/test/Transforms/PhaseOrdering/X86/horizontal-reduce-smax.ll
@@ -0,0 +1,419 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
+; RUN: opt < %s -passes="default<O2>" -S -mtriple=i686-- -mcpu=x86-64 | FileCheck %s --check-prefixes=CHECK,SSE
+; RUN: opt < %s -passes="default<O2>" -S -mtriple=i686-- -mcpu=x86-64-v2 | FileCheck %s --check-prefixes=CHECK,SSE
+; RUN: opt < %s -passes="default<O2>" -S -mtriple=i686-- -mcpu=x86-64-v3 | FileCheck %s --check-prefixes=CHECK,AVX,AVX2
+; RUN: opt < %s -passes="default<O2>" -S -mtriple=i686-- -mcpu=x86-64-v4 | FileCheck %s --check-prefixes=CHECK,AVX,AVX512
+; RUN: opt < %s -passes="default<O2>" -S -mtriple=x86_64-- -mcpu=x86-64 | FileCheck %s --check-prefixes=CHECK,SSE
+; RUN: opt < %s -passes="default<O2>" -S -mtriple=x86_64-- -mcpu=x86-64-v2 | FileCheck %s --check-prefixes=CHECK,SSE
+; RUN: opt < %s -passes="default<O2>" -S -mtriple=x86_64-- -mcpu=x86-64-v3 | FileCheck %s --check-prefixes=CHECK,AVX,AVX2
+; RUN: opt < %s -passes="default<O2>" -S -mtriple=x86_64-- -mcpu=x86-64-v4 | FileCheck %s --check-prefixes=CHECK,AVX,AVX512
+
+;
+; 128-bit Vectors
+;
+
+define i64 @test_reduce_v2i64(<2 x i64> %a0) {
+; CHECK-LABEL: @test_reduce_v2i64(
+; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <2 x i64> [[A0:%.*]], <2 x i64> poison, <2 x i32> <i32 1, i32 poison>
+; CHECK-NEXT: [[TMP2:%.*]] = tail call <2 x i64> @llvm.smax.v2i64(<2 x i64> [[A0]], <2 x i64> [[TMP1]])
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <2 x i64> [[TMP2]], i64 0
+; CHECK-NEXT: ret i64 [[TMP4]]
+;
+ %1 = shufflevector <2 x i64> %a0, <2 x i64> undef, <2 x i32> <i32 1, i32 undef>
+ %2 = icmp sgt <2 x i64> %a0, %1
+ %3 = select <2 x i1> %2, <2 x i64> %a0, <2 x i64> %1
+ %4 = extractelement <2 x i64> %3, i32 0
+ ret i64 %4
+}
+
+define i32 @test_reduce_v4i32(<4 x i32> %a0) {
+; CHECK-LABEL: @test_reduce_v4i32(
+; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <4 x i32> [[A0:%.*]], <4 x i32> poison, <4 x i32> <i32 2, i32 3, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP2:%.*]] = tail call <4 x i32> @llvm.smax.v4i32(<4 x i32> [[A0]], <4 x i32> [[TMP1]])
+; CHECK-NEXT: [[TMP3:%.*]] = shufflevector <4 x i32> [[TMP2]], <4 x i32> poison, <4 x i32> <i32 1, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP4:%.*]] = tail call <4 x i32> @llvm.smax.v4i32(<4 x i32> [[TMP2]], <4 x i32> [[TMP3]])
+; CHECK-NEXT: [[TMP7:%.*]] = extractelement <4 x i32> [[TMP4]], i64 0
+; CHECK-NEXT: ret i32 [[TMP7]]
+;
+ %1 = shufflevector <4 x i32> %a0, <4 x i32> undef, <4 x i32> <i32 2, i32 3, i32 undef, i32 undef>
+ %2 = icmp sgt <4 x i32> %a0, %1
+ %3 = select <4 x i1> %2, <4 x i32> %a0, <4 x i32> %1
+ %4 = shufflevector <4 x i32> %3, <4 x i32> undef, <4 x i32> <i32 1, i32 undef, i32 undef, i32 undef>
+ %5 = icmp sgt <4 x i32> %3, %4
+ %6 = select <4 x i1> %5, <4 x i32> %3, <4 x i32> %4
+ %7 = extractelement <4 x i32> %6, i32 0
+ ret i32 %7
+}
+
+define i16 @test_reduce_v8i16(<8 x i16> %a0) {
+; CHECK-LABEL: @test_reduce_v8i16(
+; CHECK-NEXT: [[TMP10:%.*]] = tail call i16 @llvm.vector.reduce.smax.v8i16(<8 x i16> [[A0:%.*]])
+; CHECK-NEXT: ret i16 [[TMP10]]
+;
+ %1 = shufflevector <8 x i16> %a0, <8 x i16> undef, <8 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp sgt <8 x i16> %a0, %1
+ %3 = select <8 x i1> %2, <8 x i16> %a0, <8 x i16> %1
+ %4 = shufflevector <8 x i16> %3, <8 x i16> undef, <8 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp sgt <8 x i16> %3, %4
+ %6 = select <8 x i1> %5, <8 x i16> %3, <8 x i16> %4
+ %7 = shufflevector <8 x i16> %6, <8 x i16> undef, <8 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp sgt <8 x i16> %6, %7
+ %9 = select <8 x i1> %8, <8 x i16> %6, <8 x i16> %7
+ %10 = extractelement <8 x i16> %9, i32 0
+ ret i16 %10
+}
+
+define i8 @test_reduce_v16i8(<16 x i8> %a0) {
+; CHECK-LABEL: @test_reduce_v16i8(
+; CHECK-NEXT: [[TMP13:%.*]] = tail call i8 @llvm.vector.reduce.smax.v16i8(<16 x i8> [[A0:%.*]])
+; CHECK-NEXT: ret i8 [[TMP13]]
+;
+ %1 = shufflevector <16 x i8> %a0, <16 x i8> undef, <16 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp sgt <16 x i8> %a0, %1
+ %3 = select <16 x i1> %2, <16 x i8> %a0, <16 x i8> %1
+ %4 = shufflevector <16 x i8> %3, <16 x i8> undef, <16 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp sgt <16 x i8> %3, %4
+ %6 = select <16 x i1> %5, <16 x i8> %3, <16 x i8> %4
+ %7 = shufflevector <16 x i8> %6, <16 x i8> undef, <16 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp sgt <16 x i8> %6, %7
+ %9 = select <16 x i1> %8, <16 x i8> %6, <16 x i8> %7
+ %10 = shufflevector <16 x i8> %9, <16 x i8> undef, <16 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %11 = icmp sgt <16 x i8> %9, %10
+ %12 = select <16 x i1> %11, <16 x i8> %9, <16 x i8> %10
+ %13 = extractelement <16 x i8> %12, i32 0
+ ret i8 %13
+}
+
+;
+; 256-bit Vectors
+;
+
+define i64 @test_reduce_v4i64(<4 x i64> %a0) {
+; SSE-LABEL: @test_reduce_v4i64(
+; SSE-NEXT: [[TMP1:%.*]] = tail call i64 @llvm.vector.reduce.smax.v4i64(<4 x i64> [[A0:%.*]])
+; SSE-NEXT: ret i64 [[TMP1]]
+;
+; AVX-LABEL: @test_reduce_v4i64(
+; AVX-NEXT: [[TMP1:%.*]] = shufflevector <4 x i64> [[A0:%.*]], <4 x i64> poison, <4 x i32> <i32 2, i32 3, i32 poison, i32 poison>
+; AVX-NEXT: [[TMP2:%.*]] = tail call <4 x i64> @llvm.smax.v4i64(<4 x i64> [[A0]], <4 x i64> [[TMP1]])
+; AVX-NEXT: [[TMP3:%.*]] = shufflevector <4 x i64> [[TMP2]], <4 x i64> poison, <4 x i32> <i32 1, i32 poison, i32 poison, i32 poison>
+; AVX-NEXT: [[TMP4:%.*]] = tail call <4 x i64> @llvm.smax.v4i64(<4 x i64> [[TMP2]], <4 x i64> [[TMP3]])
+; AVX-NEXT: [[TMP5:%.*]] = extractelement <4 x i64> [[TMP4]], i64 0
+; AVX-NEXT: ret i64 [[TMP5]]
+;
+ %1 = shufflevector <4 x i64> %a0, <4 x i64> undef, <4 x i32> <i32 2, i32 3, i32 undef, i32 undef>
+ %2 = icmp sgt <4 x i64> %a0, %1
+ %3 = select <4 x i1> %2, <4 x i64> %a0, <4 x i64> %1
+ %4 = shufflevector <4 x i64> %3, <4 x i64> undef, <4 x i32> <i32 1, i32 undef, i32 undef, i32 undef>
+ %5 = icmp sgt <4 x i64> %3, %4
+ %6 = select <4 x i1> %5, <4 x i64> %3, <4 x i64> %4
+ %7 = extractelement <4 x i64> %6, i32 0
+ ret i64 %7
+}
+
+define i32 @test_reduce_v8i32(<8 x i32> %a0) {
+; SSE-LABEL: @test_reduce_v8i32(
+; SSE-NEXT: [[TMP1:%.*]] = tail call i32 @llvm.vector.reduce.smax.v8i32(<8 x i32> [[A0:%.*]])
+; SSE-NEXT: ret i32 [[TMP1]]
+;
+; AVX-LABEL: @test_reduce_v8i32(
+; AVX-NEXT: [[TMP1:%.*]] = shufflevector <8 x i32> [[A0:%.*]], <8 x i32> poison, <8 x i32> <i32 4, i32 5, i32 6, i32 7, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX-NEXT: [[TMP2:%.*]] = tail call <8 x i32> @llvm.smax.v8i32(<8 x i32> [[A0]], <8 x i32> [[TMP1]])
+; AVX-NEXT: [[TMP3:%.*]] = shufflevector <8 x i32> [[TMP2]], <8 x i32> poison, <8 x i32> <i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX-NEXT: [[TMP4:%.*]] = tail call <8 x i32> @llvm.smax.v8i32(<8 x i32> [[TMP2]], <8 x i32> [[TMP3]])
+; AVX-NEXT: [[TMP5:%.*]] = shufflevector <8 x i32> [[TMP4]], <8 x i32> poison, <8 x i32> <i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX-NEXT: [[TMP6:%.*]] = tail call <8 x i32> @llvm.smax.v8i32(<8 x i32> [[TMP4]], <8 x i32> [[TMP5]])
+; AVX-NEXT: [[TMP7:%.*]] = extractelement <8 x i32> [[TMP6]], i64 0
+; AVX-NEXT: ret i32 [[TMP7]]
+;
+ %1 = shufflevector <8 x i32> %a0, <8 x i32> undef, <8 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp sgt <8 x i32> %a0, %1
+ %3 = select <8 x i1> %2, <8 x i32> %a0, <8 x i32> %1
+ %4 = shufflevector <8 x i32> %3, <8 x i32> undef, <8 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp sgt <8 x i32> %3, %4
+ %6 = select <8 x i1> %5, <8 x i32> %3, <8 x i32> %4
+ %7 = shufflevector <8 x i32> %6, <8 x i32> undef, <8 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp sgt <8 x i32> %6, %7
+ %9 = select <8 x i1> %8, <8 x i32> %6, <8 x i32> %7
+ %10 = extractelement <8 x i32> %9, i32 0
+ ret i32 %10
+}
+
+define i16 @test_reduce_v16i16(<16 x i16> %a0) {
+; CHECK-LABEL: @test_reduce_v16i16(
+; CHECK-NEXT: [[TMP13:%.*]] = tail call i16 @llvm.vector.reduce.smax.v16i16(<16 x i16> [[A0:%.*]])
+; CHECK-NEXT: ret i16 [[TMP13]]
+;
+ %1 = shufflevector <16 x i16> %a0, <16 x i16> undef, <16 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp sgt <16 x i16> %a0, %1
+ %3 = select <16 x i1> %2, <16 x i16> %a0, <16 x i16> %1
+ %4 = shufflevector <16 x i16> %3, <16 x i16> undef, <16 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp sgt <16 x i16> %3, %4
+ %6 = select <16 x i1> %5, <16 x i16> %3, <16 x i16> %4
+ %7 = shufflevector <16 x i16> %6, <16 x i16> undef, <16 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp sgt <16 x i16> %6, %7
+ %9 = select <16 x i1> %8, <16 x i16> %6, <16 x i16> %7
+ %10 = shufflevector <16 x i16> %9, <16 x i16> undef, <16 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %11 = icmp sgt <16 x i16> %9, %10
+ %12 = select <16 x i1> %11, <16 x i16> %9, <16 x i16> %10
+ %13 = extractelement <16 x i16> %12, i32 0
+ ret i16 %13
+}
+
+define i8 @test_reduce_v32i8(<32 x i8> %a0) {
+; CHECK-LABEL: @test_reduce_v32i8(
+; CHECK-NEXT: [[TMP16:%.*]] = tail call i8 @llvm.vector.reduce.smax.v32i8(<32 x i8> [[A0:%.*]])
+; CHECK-NEXT: ret i8 [[TMP16]]
+;
+ %1 = shufflevector <32 x i8> %a0, <32 x i8> undef, <32 x i32> <i32 16, i32 17, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23, i32 24, i32 25, i32 26, i32 27, i32 28, i32 29, i32 30, i32 31, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp sgt <32 x i8> %a0, %1
+ %3 = select <32 x i1> %2, <32 x i8> %a0, <32 x i8> %1
+ %4 = shufflevector <32 x i8> %3, <32 x i8> undef, <32 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp sgt <32 x i8> %3, %4
+ %6 = select <32 x i1> %5, <32 x i8> %3, <32 x i8> %4
+ %7 = shufflevector <32 x i8> %6, <32 x i8> undef, <32 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp sgt <32 x i8> %6, %7
+ %9 = select <32 x i1> %8, <32 x i8> %6, <32 x i8> %7
+ %10 = shufflevector <32 x i8> %9, <32 x i8> undef, <32 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %11 = icmp sgt <32 x i8> %9, %10
+ %12 = select <32 x i1> %11, <32 x i8> %9, <32 x i8> %10
+ %13 = shufflevector <32 x i8> %12, <32 x i8> undef, <32 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %14 = icmp sgt <32 x i8> %12, %13
+ %15 = select <32 x i1> %14, <32 x i8> %12, <32 x i8> %13
+ %16 = extractelement <32 x i8> %15, i32 0
+ ret i8 %16
+}
+
+;
+; 512-bit Vectors
+;
+
+define i64 @test_reduce_v8i64(<8 x i64> %a0) {
+; SSE-LABEL: @test_reduce_v8i64(
+; SSE-NEXT: [[TMP1:%.*]] = tail call i64 @llvm.vector.reduce.smax.v8i64(<8 x i64> [[A0:%.*]])
+; SSE-NEXT: ret i64 [[TMP1]]
+;
+; AVX2-LABEL: @test_reduce_v8i64(
+; AVX2-NEXT: [[TMP1:%.*]] = tail call i64 @llvm.vector.reduce.smax.v8i64(<8 x i64> [[A0:%.*]])
+; AVX2-NEXT: ret i64 [[TMP1]]
+;
+; AVX512-LABEL: @test_reduce_v8i64(
+; AVX512-NEXT: [[TMP1:%.*]] = shufflevector <8 x i64> [[A0:%.*]], <8 x i64> poison, <8 x i32> <i32 4, i32 5, i32 6, i32 7, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX512-NEXT: [[TMP2:%.*]] = tail call <8 x i64> @llvm.smax.v8i64(<8 x i64> [[A0]], <8 x i64> [[TMP1]])
+; AVX512-NEXT: [[TMP3:%.*]] = shufflevector <8 x i64> [[TMP2]], <8 x i64> poison, <8 x i32> <i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX512-NEXT: [[TMP4:%.*]] = tail call <8 x i64> @llvm.smax.v8i64(<8 x i64> [[TMP2]], <8 x i64> [[TMP3]])
+; AVX512-NEXT: [[TMP5:%.*]] = shufflevector <8 x i64> [[TMP4]], <8 x i64> poison, <8 x i32> <i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX512-NEXT: [[TMP6:%.*]] = tail call <8 x i64> @llvm.smax.v8i64(<8 x i64> [[TMP4]], <8 x i64> [[TMP5]])
+; AVX512-NEXT: [[TMP7:%.*]] = extractelement <8 x i64> [[TMP6]], i64 0
+; AVX512-NEXT: ret i64 [[TMP7]]
+;
+ %1 = shufflevector <8 x i64> %a0, <8 x i64> undef, <8 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp sgt <8 x i64> %a0, %1
+ %3 = select <8 x i1> %2, <8 x i64> %a0, <8 x i64> %1
+ %4 = shufflevector <8 x i64> %3, <8 x i64> undef, <8 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp sgt <8 x i64> %3, %4
+ %6 = select <8 x i1> %5, <8 x i64> %3, <8 x i64> %4
+ %7 = shufflevector <8 x i64> %6, <8 x i64> undef, <8 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp sgt <8 x i64> %6, %7
+ %9 = select <8 x i1> %8, <8 x i64> %6, <8 x i64> %7
+ %10 = extractelement <8 x i64> %9, i32 0
+ ret i64 %10
+}
+
+define i32 @test_reduce_v16i32(<16 x i32> %a0) {
+; SSE-LABEL: @test_reduce_v16i32(
+; SSE-NEXT: [[TMP1:%.*]] = tail call i32 @llvm.vector.reduce.smax.v16i32(<16 x i32> [[A0:%.*]])
+; SSE-NEXT: ret i32 [[TMP1]]
+;
+; AVX2-LABEL: @test_reduce_v16i32(
+; AVX2-NEXT: [[TMP1:%.*]] = tail call i32 @llvm.vector.reduce.smax.v16i32(<16 x i32> [[A0:%.*]])
+; AVX2-NEXT: ret i32 [[TMP1]]
+;
+; AVX512-LABEL: @test_reduce_v16i32(
+; AVX512-NEXT: [[TMP1:%.*]] = shufflevector <16 x i32> [[A0:%.*]], <16 x i32> poison, <16 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX512-NEXT: [[TMP2:%.*]] = tail call <16 x i32> @llvm.smax.v16i32(<16 x i32> [[A0]], <16 x i32> [[TMP1]])
+; AVX512-NEXT: [[TMP3:%.*]] = shufflevector <16 x i32> [[TMP2]], <16 x i32> poison, <16 x i32> <i32 4, i32 5, i32 6, i32 7, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX512-NEXT: [[TMP4:%.*]] = tail call <16 x i32> @llvm.smax.v16i32(<16 x i32> [[TMP2]], <16 x i32> [[TMP3]])
+; AVX512-NEXT: [[TMP5:%.*]] = shufflevector <16 x i32> [[TMP4]], <16 x i32> poison, <16 x i32> <i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX512-NEXT: [[TMP6:%.*]] = tail call <16 x i32> @llvm.smax.v16i32(<16 x i32> [[TMP4]], <16 x i32> [[TMP5]])
+; AVX512-NEXT: [[TMP7:%.*]] = shufflevector <16 x i32> [[TMP6]], <16 x i32> poison, <16 x i32> <i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX512-NEXT: [[TMP8:%.*]] = tail call <16 x i32> @llvm.smax.v16i32(<16 x i32> [[TMP6]], <16 x i32> [[TMP7]])
+; AVX512-NEXT: [[TMP9:%.*]] = extractelement <16 x i32> [[TMP8]], i64 0
+; AVX512-NEXT: ret i32 [[TMP9]]
+;
+ %1 = shufflevector <16 x i32> %a0, <16 x i32> undef, <16 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp sgt <16 x i32> %a0, %1
+ %3 = select <16 x i1> %2, <16 x i32> %a0, <16 x i32> %1
+ %4 = shufflevector <16 x i32> %3, <16 x i32> undef, <16 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp sgt <16 x i32> %3, %4
+ %6 = select <16 x i1> %5, <16 x i32> %3, <16 x i32> %4
+ %7 = shufflevector <16 x i32> %6, <16 x i32> undef, <16 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp sgt <16 x i32> %6, %7
+ %9 = select <16 x i1> %8, <16 x i32> %6, <16 x i32> %7
+ %10 = shufflevector <16 x i32> %9, <16 x i32> undef, <16 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %11 = icmp sgt <16 x i32> %9, %10
+ %12 = select <16 x i1> %11, <16 x i32> %9, <16 x i32> %10
+ %13 = extractelement <16 x i32> %12, i32 0
+ ret i32 %13
+}
+
+define i16 @test_reduce_v32i16(<32 x i16> %a0) {
+; CHECK-LABEL: @test_reduce_v32i16(
+; CHECK-NEXT: [[TMP16:%.*]] = tail call i16 @llvm.vector.reduce.smax.v32i16(<32 x i16> [[A0:%.*]])
+; CHECK-NEXT: ret i16 [[TMP16]]
+;
+ %1 = shufflevector <32 x i16> %a0, <32 x i16> undef, <32 x i32> <i32 16, i32 17, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23, i32 24, i32 25, i32 26, i32 27, i32 28, i32 29, i32 30, i32 31, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp sgt <32 x i16> %a0, %1
+ %3 = select <32 x i1> %2, <32 x i16> %a0, <32 x i16> %1
+ %4 = shufflevector <32 x i16> %3, <32 x i16> undef, <32 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp sgt <32 x i16> %3, %4
+ %6 = select <32 x i1> %5, <32 x i16> %3, <32 x i16> %4
+ %7 = shufflevector <32 x i16> %6, <32 x i16> undef, <32 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp sgt <32 x i16> %6, %7
+ %9 = select <32 x i1> %8, <32 x i16> %6, <32 x i16> %7
+ %10 = shufflevector <32 x i16> %9, <32 x i16> undef, <32 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %11 = icmp sgt <32 x i16> %9, %10
+ %12 = select <32 x i1> %11, <32 x i16> %9, <32 x i16> %10
+ %13 = shufflevector <32 x i16> %12, <32 x i16> undef, <32 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %14 = icmp sgt <32 x i16> %12, %13
+ %15 = select <32 x i1> %14, <32 x i16> %12, <32 x i16> %13
+ %16 = extractelement <32 x i16> %15, i32 0
+ ret i16 %16
+}
+
+define i8 @test_reduce_v64i8(<64 x i8> %a0) {
+; CHECK-LABEL: @test_reduce_v64i8(
+; CHECK-NEXT: [[TMP19:%.*]] = tail call i8 @llvm.vector.reduce.smax.v64i8(<64 x i8> [[A0:%.*]])
+; CHECK-NEXT: ret i8 [[TMP19]]
+;
+ %1 = shufflevector <64 x i8> %a0, <64 x i8> undef, <64 x i32> <i32 32, i32 33, i32 34, i32 35, i32 36, i32 37, i32 38, i32 39, i32 40, i32 41, i32 42, i32 43, i32 44, i32 45, i32 46, i32 47, i32 48, i32 49, i32 50, i32 51, i32 52, i32 53, i32 54, i32 55, i32 56, i32 57, i32 58, i32 59, i32 60, i32 61, i32 62, i32 63, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp sgt <64 x i8> %a0, %1
+ %3 = select <64 x i1> %2, <64 x i8> %a0, <64 x i8> %1
+ %4 = shufflevector <64 x i8> %3, <64 x i8> undef, <64 x i32> <i32 16, i32 17, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23, i32 24, i32 25, i32 26, i32 27, i32 28, i32 29, i32 30, i32 31, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp sgt <64 x i8> %3, %4
+ %6 = select <64 x i1> %5, <64 x i8> %3, <64 x i8> %4
+ %7 = shufflevector <64 x i8> %6, <64 x i8> undef, <64 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp sgt <64 x i8> %6, %7
+ %9 = select <64 x i1> %8, <64 x i8> %6, <64 x i8> %7
+ %10 = shufflevector <64 x i8> %9, <64 x i8> undef, <64 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %11 = icmp sgt <64 x i8> %9, %10
+ %12 = select <64 x i1> %11, <64 x i8> %9, <64 x i8> %10
+ %13 = shufflevector <64 x i8> %12, <64 x i8> undef, <64 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %14 = icmp sgt <64 x i8> %12, %13
+ %15 = select <64 x i1> %14, <64 x i8> %12, <64 x i8> %13
+ %16 = shufflevector <64 x i8> %15, <64 x i8> undef, <64 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %17 = icmp sgt <64 x i8> %15, %16
+ %18 = select <64 x i1> %17, <64 x i8> %15, <64 x i8> %16
+ %19 = extractelement <64 x i8> %18, i32 0
+ ret i8 %19
+}
+
+;
+; Partial Vector Reductions
+;
+
+define i16 @test_reduce_v16i16_v8i16(<16 x i16> %a0) {
+; CHECK-LABEL: @test_reduce_v16i16_v8i16(
+; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <16 x i16> [[A0:%.*]], <16 x i16> poison, <16 x i32> <i32 4, i32 5, i32 6, i32 7, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP2:%.*]] = tail call <16 x i16> @llvm.smax.v16i16(<16 x i16> [[A0]], <16 x i16> [[TMP1]])
+; CHECK-NEXT: [[TMP3:%.*]] = shufflevector <16 x i16> [[TMP2]], <16 x i16> poison, <16 x i32> <i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP4:%.*]] = tail call <16 x i16> @llvm.smax.v16i16(<16 x i16> [[TMP2]], <16 x i16> [[TMP3]])
+; CHECK-NEXT: [[TMP5:%.*]] = shufflevector <16 x i16> [[TMP4]], <16 x i16> poison, <16 x i32> <i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP6:%.*]] = tail call <16 x i16> @llvm.smax.v16i16(<16 x i16> [[TMP4]], <16 x i16> [[TMP5]])
+; CHECK-NEXT: [[TMP10:%.*]] = extractelement <16 x i16> [[TMP6]], i64 0
+; CHECK-NEXT: ret i16 [[TMP10]]
+;
+ %1 = shufflevector <16 x i16> %a0, <16 x i16> undef, <16 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp sgt <16 x i16> %a0, %1
+ %3 = select <16 x i1> %2, <16 x i16> %a0, <16 x i16> %1
+ %4 = shufflevector <16 x i16> %3, <16 x i16> undef, <16 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp sgt <16 x i16> %3, %4
+ %6 = select <16 x i1> %5, <16 x i16> %3, <16 x i16> %4
+ %7 = shufflevector <16 x i16> %6, <16 x i16> undef, <16 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp sgt <16 x i16> %6, %7
+ %9 = select <16 x i1> %8, <16 x i16> %6, <16 x i16> %7
+ %10 = extractelement <16 x i16> %9, i32 0
+ ret i16 %10
+}
+
+define i16 @test_reduce_v32i16_v8i16(<32 x i16> %a0) {
+; CHECK-LABEL: @test_reduce_v32i16_v8i16(
+; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <32 x i16> [[A0:%.*]], <32 x i16> poison, <32 x i32> <i32 4, i32 5, i32 6, i32 7, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP2:%.*]] = tail call <32 x i16> @llvm.smax.v32i16(<32 x i16> [[A0]], <32 x i16> [[TMP1]])
+; CHECK-NEXT: [[TMP3:%.*]] = shufflevector <32 x i16> [[TMP2]], <32 x i16> poison, <32 x i32> <i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP4:%.*]] = tail call <32 x i16> @llvm.smax.v32i16(<32 x i16> [[TMP2]], <32 x i16> [[TMP3]])
+; CHECK-NEXT: [[TMP5:%.*]] = shufflevector <32 x i16> [[TMP4]], <32 x i16> poison, <32 x i32> <i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP6:%.*]] = tail call <32 x i16> @llvm.smax.v32i16(<32 x i16> [[TMP4]], <32 x i16> [[TMP5]])
+; CHECK-NEXT: [[TMP10:%.*]] = extractelement <32 x i16> [[TMP6]], i64 0
+; CHECK-NEXT: ret i16 [[TMP10]]
+;
+ %1 = shufflevector <32 x i16> %a0, <32 x i16> undef, <32 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp sgt <32 x i16> %a0, %1
+ %3 = select <32 x i1> %2, <32 x i16> %a0, <32 x i16> %1
+ %4 = shufflevector <32 x i16> %3, <32 x i16> undef, <32 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp sgt <32 x i16> %3, %4
+ %6 = select <32 x i1> %5, <32 x i16> %3, <32 x i16> %4
+ %7 = shufflevector <32 x i16> %6, <32 x i16> undef, <32 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp sgt <32 x i16> %6, %7
+ %9 = select <32 x i1> %8, <32 x i16> %6, <32 x i16> %7
+ %10 = extractelement <32 x i16> %9, i32 0
+ ret i16 %10
+}
+
+define i8 @test_reduce_v32i8_v16i8(<32 x i8> %a0) {
+; CHECK-LABEL: @test_reduce_v32i8_v16i8(
+; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <32 x i8> [[A0:%.*]], <32 x i8> poison, <32 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP2:%.*]] = tail call <32 x i8> @llvm.smax.v32i8(<32 x i8> [[A0]], <32 x i8> [[TMP1]])
+; CHECK-NEXT: [[TMP3:%.*]] = shufflevector <32 x i8> [[TMP2]], <32 x i8> poison, <32 x i32> <i32 4, i32 5, i32 6, i32 7, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP4:%.*]] = tail call <32 x i8> @llvm.smax.v32i8(<32 x i8> [[TMP2]], <32 x i8> [[TMP3]])
+; CHECK-NEXT: [[TMP5:%.*]] = shufflevector <32 x i8> [[TMP4]], <32 x i8> poison, <32 x i32> <i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP6:%.*]] = tail call <32 x i8> @llvm.smax.v32i8(<32 x i8> [[TMP4]], <32 x i8> [[TMP5]])
+; CHECK-NEXT: [[TMP7:%.*]] = shufflevector <32 x i8> [[TMP6]], <32 x i8> poison, <32 x i32> <i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP8:%.*]] = tail call <32 x i8> @llvm.smax.v32i8(<32 x i8> [[TMP6]], <32 x i8> [[TMP7]])
+; CHECK-NEXT: [[TMP13:%.*]] = extractelement <32 x i8> [[TMP8]], i64 0
+; CHECK-NEXT: ret i8 [[TMP13]]
+;
+ %1 = shufflevector <32 x i8> %a0, <32 x i8> undef, <32 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp sgt <32 x i8> %a0, %1
+ %3 = select <32 x i1> %2, <32 x i8> %a0, <32 x i8> %1
+ %4 = shufflevector <32 x i8> %3, <32 x i8> undef, <32 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp sgt <32 x i8> %3, %4
+ %6 = select <32 x i1> %5, <32 x i8> %3, <32 x i8> %4
+ %7 = shufflevector <32 x i8> %6, <32 x i8> undef, <32 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp sgt <32 x i8> %6, %7
+ %9 = select <32 x i1> %8, <32 x i8> %6, <32 x i8> %7
+ %10 = shufflevector <32 x i8> %9, <32 x i8> undef, <32 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %11 = icmp sgt <32 x i8> %9, %10
+ %12 = select <32 x i1> %11, <32 x i8> %9, <32 x i8> %10
+ %13 = extractelement <32 x i8> %12, i32 0
+ ret i8 %13
+}
+
+define i8 @test_reduce_v64i8_v16i8(<64 x i8> %a0) {
+; CHECK-LABEL: @test_reduce_v64i8_v16i8(
+; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <64 x i8> [[A0:%.*]], <64 x i8> poison, <64 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP2:%.*]] = tail call <64 x i8> @llvm.smax.v64i8(<64 x i8> [[A0]], <64 x i8> [[TMP1]])
+; CHECK-NEXT: [[TMP3:%.*]] = shufflevector <64 x i8> [[TMP2]], <64 x i8> poison, <64 x i32> <i32 4, i32 5, i32 6, i32 7, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP4:%.*]] = tail call <64 x i8> @llvm.smax.v64i8(<64 x i8> [[TMP2]], <64 x i8> [[TMP3]])
+; CHECK-NEXT: [[TMP5:%.*]] = shufflevector <64 x i8> [[TMP4]], <64 x i8> poison, <64 x i32> <i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP6:%.*]] = tail call <64 x i8> @llvm.smax.v64i8(<64 x i8> [[TMP4]], <64 x i8> [[TMP5]])
+; CHECK-NEXT: [[TMP7:%.*]] = shufflevector <64 x i8> [[TMP6]], <64 x i8> poison, <64 x i32> <i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP8:%.*]] = tail call <64 x i8> @llvm.smax.v64i8(<64 x i8> [[TMP6]], <64 x i8> [[TMP7]])
+; CHECK-NEXT: [[TMP13:%.*]] = extractelement <64 x i8> [[TMP8]], i64 0
+; CHECK-NEXT: ret i8 [[TMP13]]
+;
+ %1 = shufflevector <64 x i8> %a0, <64 x i8> undef, <64 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp sgt <64 x i8> %a0, %1
+ %3 = select <64 x i1> %2, <64 x i8> %a0, <64 x i8> %1
+ %4 = shufflevector <64 x i8> %3, <64 x i8> undef, <64 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp sgt <64 x i8> %3, %4
+ %6 = select <64 x i1> %5, <64 x i8> %3, <64 x i8> %4
+ %7 = shufflevector <64 x i8> %6, <64 x i8> undef, <64 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp sgt <64 x i8> %6, %7
+ %9 = select <64 x i1> %8, <64 x i8> %6, <64 x i8> %7
+ %10 = shufflevector <64 x i8> %9, <64 x i8> undef, <64 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %11 = icmp sgt <64 x i8> %9, %10
+ %12 = select <64 x i1> %11, <64 x i8> %9, <64 x i8> %10
+ %13 = extractelement <64 x i8> %12, i32 0
+ ret i8 %13
+}
diff --git a/llvm/test/Transforms/PhaseOrdering/X86/horizontal-reduce-smin.ll b/llvm/test/Transforms/PhaseOrdering/X86/horizontal-reduce-smin.ll
new file mode 100644
index 0000000000000..80c2929b5d5cf
--- /dev/null
+++ b/llvm/test/Transforms/PhaseOrdering/X86/horizontal-reduce-smin.ll
@@ -0,0 +1,419 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
+; RUN: opt < %s -passes="default<O2>" -S -mtriple=i686-- -mcpu=x86-64 | FileCheck %s --check-prefixes=CHECK,SSE
+; RUN: opt < %s -passes="default<O2>" -S -mtriple=i686-- -mcpu=x86-64-v2 | FileCheck %s --check-prefixes=CHECK,SSE
+; RUN: opt < %s -passes="default<O2>" -S -mtriple=i686-- -mcpu=x86-64-v3 | FileCheck %s --check-prefixes=CHECK,AVX,AVX2
+; RUN: opt < %s -passes="default<O2>" -S -mtriple=i686-- -mcpu=x86-64-v4 | FileCheck %s --check-prefixes=CHECK,AVX,AVX512
+; RUN: opt < %s -passes="default<O2>" -S -mtriple=x86_64-- -mcpu=x86-64 | FileCheck %s --check-prefixes=CHECK,SSE
+; RUN: opt < %s -passes="default<O2>" -S -mtriple=x86_64-- -mcpu=x86-64-v2 | FileCheck %s --check-prefixes=CHECK,SSE
+; RUN: opt < %s -passes="default<O2>" -S -mtriple=x86_64-- -mcpu=x86-64-v3 | FileCheck %s --check-prefixes=CHECK,AVX,AVX2
+; RUN: opt < %s -passes="default<O2>" -S -mtriple=x86_64-- -mcpu=x86-64-v4 | FileCheck %s --check-prefixes=CHECK,AVX,AVX512
+
+;
+; 128-bit Vectors
+;
+
+define i64 @test_reduce_v2i64(<2 x i64> %a0) {
+; CHECK-LABEL: @test_reduce_v2i64(
+; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <2 x i64> [[A0:%.*]], <2 x i64> poison, <2 x i32> <i32 1, i32 poison>
+; CHECK-NEXT: [[TMP2:%.*]] = tail call <2 x i64> @llvm.smin.v2i64(<2 x i64> [[A0]], <2 x i64> [[TMP1]])
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <2 x i64> [[TMP2]], i64 0
+; CHECK-NEXT: ret i64 [[TMP4]]
+;
+ %1 = shufflevector <2 x i64> %a0, <2 x i64> undef, <2 x i32> <i32 1, i32 undef>
+ %2 = icmp slt <2 x i64> %a0, %1
+ %3 = select <2 x i1> %2, <2 x i64> %a0, <2 x i64> %1
+ %4 = extractelement <2 x i64> %3, i32 0
+ ret i64 %4
+}
+
+define i32 @test_reduce_v4i32(<4 x i32> %a0) {
+; CHECK-LABEL: @test_reduce_v4i32(
+; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <4 x i32> [[A0:%.*]], <4 x i32> poison, <4 x i32> <i32 2, i32 3, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP2:%.*]] = tail call <4 x i32> @llvm.smin.v4i32(<4 x i32> [[A0]], <4 x i32> [[TMP1]])
+; CHECK-NEXT: [[TMP3:%.*]] = shufflevector <4 x i32> [[TMP2]], <4 x i32> poison, <4 x i32> <i32 1, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP4:%.*]] = tail call <4 x i32> @llvm.smin.v4i32(<4 x i32> [[TMP2]], <4 x i32> [[TMP3]])
+; CHECK-NEXT: [[TMP7:%.*]] = extractelement <4 x i32> [[TMP4]], i64 0
+; CHECK-NEXT: ret i32 [[TMP7]]
+;
+ %1 = shufflevector <4 x i32> %a0, <4 x i32> undef, <4 x i32> <i32 2, i32 3, i32 undef, i32 undef>
+ %2 = icmp slt <4 x i32> %a0, %1
+ %3 = select <4 x i1> %2, <4 x i32> %a0, <4 x i32> %1
+ %4 = shufflevector <4 x i32> %3, <4 x i32> undef, <4 x i32> <i32 1, i32 undef, i32 undef, i32 undef>
+ %5 = icmp slt <4 x i32> %3, %4
+ %6 = select <4 x i1> %5, <4 x i32> %3, <4 x i32> %4
+ %7 = extractelement <4 x i32> %6, i32 0
+ ret i32 %7
+}
+
+define i16 @test_reduce_v8i16(<8 x i16> %a0) {
+; CHECK-LABEL: @test_reduce_v8i16(
+; CHECK-NEXT: [[TMP10:%.*]] = tail call i16 @llvm.vector.reduce.smin.v8i16(<8 x i16> [[A0:%.*]])
+; CHECK-NEXT: ret i16 [[TMP10]]
+;
+ %1 = shufflevector <8 x i16> %a0, <8 x i16> undef, <8 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp slt <8 x i16> %a0, %1
+ %3 = select <8 x i1> %2, <8 x i16> %a0, <8 x i16> %1
+ %4 = shufflevector <8 x i16> %3, <8 x i16> undef, <8 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp slt <8 x i16> %3, %4
+ %6 = select <8 x i1> %5, <8 x i16> %3, <8 x i16> %4
+ %7 = shufflevector <8 x i16> %6, <8 x i16> undef, <8 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp slt <8 x i16> %6, %7
+ %9 = select <8 x i1> %8, <8 x i16> %6, <8 x i16> %7
+ %10 = extractelement <8 x i16> %9, i32 0
+ ret i16 %10
+}
+
+define i8 @test_reduce_v16i8(<16 x i8> %a0) {
+; CHECK-LABEL: @test_reduce_v16i8(
+; CHECK-NEXT: [[TMP13:%.*]] = tail call i8 @llvm.vector.reduce.smin.v16i8(<16 x i8> [[A0:%.*]])
+; CHECK-NEXT: ret i8 [[TMP13]]
+;
+ %1 = shufflevector <16 x i8> %a0, <16 x i8> undef, <16 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp slt <16 x i8> %a0, %1
+ %3 = select <16 x i1> %2, <16 x i8> %a0, <16 x i8> %1
+ %4 = shufflevector <16 x i8> %3, <16 x i8> undef, <16 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp slt <16 x i8> %3, %4
+ %6 = select <16 x i1> %5, <16 x i8> %3, <16 x i8> %4
+ %7 = shufflevector <16 x i8> %6, <16 x i8> undef, <16 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp slt <16 x i8> %6, %7
+ %9 = select <16 x i1> %8, <16 x i8> %6, <16 x i8> %7
+ %10 = shufflevector <16 x i8> %9, <16 x i8> undef, <16 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %11 = icmp slt <16 x i8> %9, %10
+ %12 = select <16 x i1> %11, <16 x i8> %9, <16 x i8> %10
+ %13 = extractelement <16 x i8> %12, i32 0
+ ret i8 %13
+}
+
+;
+; 256-bit Vectors
+;
+
+define i64 @test_reduce_v4i64(<4 x i64> %a0) {
+; SSE-LABEL: @test_reduce_v4i64(
+; SSE-NEXT: [[TMP1:%.*]] = tail call i64 @llvm.vector.reduce.smin.v4i64(<4 x i64> [[A0:%.*]])
+; SSE-NEXT: ret i64 [[TMP1]]
+;
+; AVX-LABEL: @test_reduce_v4i64(
+; AVX-NEXT: [[TMP1:%.*]] = shufflevector <4 x i64> [[A0:%.*]], <4 x i64> poison, <4 x i32> <i32 2, i32 3, i32 poison, i32 poison>
+; AVX-NEXT: [[TMP2:%.*]] = tail call <4 x i64> @llvm.smin.v4i64(<4 x i64> [[A0]], <4 x i64> [[TMP1]])
+; AVX-NEXT: [[TMP3:%.*]] = shufflevector <4 x i64> [[TMP2]], <4 x i64> poison, <4 x i32> <i32 1, i32 poison, i32 poison, i32 poison>
+; AVX-NEXT: [[TMP4:%.*]] = tail call <4 x i64> @llvm.smin.v4i64(<4 x i64> [[TMP2]], <4 x i64> [[TMP3]])
+; AVX-NEXT: [[TMP5:%.*]] = extractelement <4 x i64> [[TMP4]], i64 0
+; AVX-NEXT: ret i64 [[TMP5]]
+;
+ %1 = shufflevector <4 x i64> %a0, <4 x i64> undef, <4 x i32> <i32 2, i32 3, i32 undef, i32 undef>
+ %2 = icmp slt <4 x i64> %a0, %1
+ %3 = select <4 x i1> %2, <4 x i64> %a0, <4 x i64> %1
+ %4 = shufflevector <4 x i64> %3, <4 x i64> undef, <4 x i32> <i32 1, i32 undef, i32 undef, i32 undef>
+ %5 = icmp slt <4 x i64> %3, %4
+ %6 = select <4 x i1> %5, <4 x i64> %3, <4 x i64> %4
+ %7 = extractelement <4 x i64> %6, i32 0
+ ret i64 %7
+}
+
+define i32 @test_reduce_v8i32(<8 x i32> %a0) {
+; SSE-LABEL: @test_reduce_v8i32(
+; SSE-NEXT: [[TMP1:%.*]] = tail call i32 @llvm.vector.reduce.smin.v8i32(<8 x i32> [[A0:%.*]])
+; SSE-NEXT: ret i32 [[TMP1]]
+;
+; AVX-LABEL: @test_reduce_v8i32(
+; AVX-NEXT: [[TMP1:%.*]] = shufflevector <8 x i32> [[A0:%.*]], <8 x i32> poison, <8 x i32> <i32 4, i32 5, i32 6, i32 7, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX-NEXT: [[TMP2:%.*]] = tail call <8 x i32> @llvm.smin.v8i32(<8 x i32> [[A0]], <8 x i32> [[TMP1]])
+; AVX-NEXT: [[TMP3:%.*]] = shufflevector <8 x i32> [[TMP2]], <8 x i32> poison, <8 x i32> <i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX-NEXT: [[TMP4:%.*]] = tail call <8 x i32> @llvm.smin.v8i32(<8 x i32> [[TMP2]], <8 x i32> [[TMP3]])
+; AVX-NEXT: [[TMP5:%.*]] = shufflevector <8 x i32> [[TMP4]], <8 x i32> poison, <8 x i32> <i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX-NEXT: [[TMP6:%.*]] = tail call <8 x i32> @llvm.smin.v8i32(<8 x i32> [[TMP4]], <8 x i32> [[TMP5]])
+; AVX-NEXT: [[TMP7:%.*]] = extractelement <8 x i32> [[TMP6]], i64 0
+; AVX-NEXT: ret i32 [[TMP7]]
+;
+ %1 = shufflevector <8 x i32> %a0, <8 x i32> undef, <8 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp slt <8 x i32> %a0, %1
+ %3 = select <8 x i1> %2, <8 x i32> %a0, <8 x i32> %1
+ %4 = shufflevector <8 x i32> %3, <8 x i32> undef, <8 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp slt <8 x i32> %3, %4
+ %6 = select <8 x i1> %5, <8 x i32> %3, <8 x i32> %4
+ %7 = shufflevector <8 x i32> %6, <8 x i32> undef, <8 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp slt <8 x i32> %6, %7
+ %9 = select <8 x i1> %8, <8 x i32> %6, <8 x i32> %7
+ %10 = extractelement <8 x i32> %9, i32 0
+ ret i32 %10
+}
+
+define i16 @test_reduce_v16i16(<16 x i16> %a0) {
+; CHECK-LABEL: @test_reduce_v16i16(
+; CHECK-NEXT: [[TMP13:%.*]] = tail call i16 @llvm.vector.reduce.smin.v16i16(<16 x i16> [[A0:%.*]])
+; CHECK-NEXT: ret i16 [[TMP13]]
+;
+ %1 = shufflevector <16 x i16> %a0, <16 x i16> undef, <16 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp slt <16 x i16> %a0, %1
+ %3 = select <16 x i1> %2, <16 x i16> %a0, <16 x i16> %1
+ %4 = shufflevector <16 x i16> %3, <16 x i16> undef, <16 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp slt <16 x i16> %3, %4
+ %6 = select <16 x i1> %5, <16 x i16> %3, <16 x i16> %4
+ %7 = shufflevector <16 x i16> %6, <16 x i16> undef, <16 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp slt <16 x i16> %6, %7
+ %9 = select <16 x i1> %8, <16 x i16> %6, <16 x i16> %7
+ %10 = shufflevector <16 x i16> %9, <16 x i16> undef, <16 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %11 = icmp slt <16 x i16> %9, %10
+ %12 = select <16 x i1> %11, <16 x i16> %9, <16 x i16> %10
+ %13 = extractelement <16 x i16> %12, i32 0
+ ret i16 %13
+}
+
+define i8 @test_reduce_v32i8(<32 x i8> %a0) {
+; CHECK-LABEL: @test_reduce_v32i8(
+; CHECK-NEXT: [[TMP16:%.*]] = tail call i8 @llvm.vector.reduce.smin.v32i8(<32 x i8> [[A0:%.*]])
+; CHECK-NEXT: ret i8 [[TMP16]]
+;
+ %1 = shufflevector <32 x i8> %a0, <32 x i8> undef, <32 x i32> <i32 16, i32 17, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23, i32 24, i32 25, i32 26, i32 27, i32 28, i32 29, i32 30, i32 31, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp slt <32 x i8> %a0, %1
+ %3 = select <32 x i1> %2, <32 x i8> %a0, <32 x i8> %1
+ %4 = shufflevector <32 x i8> %3, <32 x i8> undef, <32 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp slt <32 x i8> %3, %4
+ %6 = select <32 x i1> %5, <32 x i8> %3, <32 x i8> %4
+ %7 = shufflevector <32 x i8> %6, <32 x i8> undef, <32 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp slt <32 x i8> %6, %7
+ %9 = select <32 x i1> %8, <32 x i8> %6, <32 x i8> %7
+ %10 = shufflevector <32 x i8> %9, <32 x i8> undef, <32 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %11 = icmp slt <32 x i8> %9, %10
+ %12 = select <32 x i1> %11, <32 x i8> %9, <32 x i8> %10
+ %13 = shufflevector <32 x i8> %12, <32 x i8> undef, <32 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %14 = icmp slt <32 x i8> %12, %13
+ %15 = select <32 x i1> %14, <32 x i8> %12, <32 x i8> %13
+ %16 = extractelement <32 x i8> %15, i32 0
+ ret i8 %16
+}
+
+;
+; 512-bit Vectors
+;
+
+define i64 @test_reduce_v8i64(<8 x i64> %a0) {
+; SSE-LABEL: @test_reduce_v8i64(
+; SSE-NEXT: [[TMP1:%.*]] = tail call i64 @llvm.vector.reduce.smin.v8i64(<8 x i64> [[A0:%.*]])
+; SSE-NEXT: ret i64 [[TMP1]]
+;
+; AVX2-LABEL: @test_reduce_v8i64(
+; AVX2-NEXT: [[TMP1:%.*]] = tail call i64 @llvm.vector.reduce.smin.v8i64(<8 x i64> [[A0:%.*]])
+; AVX2-NEXT: ret i64 [[TMP1]]
+;
+; AVX512-LABEL: @test_reduce_v8i64(
+; AVX512-NEXT: [[TMP1:%.*]] = shufflevector <8 x i64> [[A0:%.*]], <8 x i64> poison, <8 x i32> <i32 4, i32 5, i32 6, i32 7, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX512-NEXT: [[TMP2:%.*]] = tail call <8 x i64> @llvm.smin.v8i64(<8 x i64> [[A0]], <8 x i64> [[TMP1]])
+; AVX512-NEXT: [[TMP3:%.*]] = shufflevector <8 x i64> [[TMP2]], <8 x i64> poison, <8 x i32> <i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX512-NEXT: [[TMP4:%.*]] = tail call <8 x i64> @llvm.smin.v8i64(<8 x i64> [[TMP2]], <8 x i64> [[TMP3]])
+; AVX512-NEXT: [[TMP5:%.*]] = shufflevector <8 x i64> [[TMP4]], <8 x i64> poison, <8 x i32> <i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX512-NEXT: [[TMP6:%.*]] = tail call <8 x i64> @llvm.smin.v8i64(<8 x i64> [[TMP4]], <8 x i64> [[TMP5]])
+; AVX512-NEXT: [[TMP7:%.*]] = extractelement <8 x i64> [[TMP6]], i64 0
+; AVX512-NEXT: ret i64 [[TMP7]]
+;
+ %1 = shufflevector <8 x i64> %a0, <8 x i64> undef, <8 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp slt <8 x i64> %a0, %1
+ %3 = select <8 x i1> %2, <8 x i64> %a0, <8 x i64> %1
+ %4 = shufflevector <8 x i64> %3, <8 x i64> undef, <8 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp slt <8 x i64> %3, %4
+ %6 = select <8 x i1> %5, <8 x i64> %3, <8 x i64> %4
+ %7 = shufflevector <8 x i64> %6, <8 x i64> undef, <8 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp slt <8 x i64> %6, %7
+ %9 = select <8 x i1> %8, <8 x i64> %6, <8 x i64> %7
+ %10 = extractelement <8 x i64> %9, i32 0
+ ret i64 %10
+}
+
+define i32 @test_reduce_v16i32(<16 x i32> %a0) {
+; SSE-LABEL: @test_reduce_v16i32(
+; SSE-NEXT: [[TMP1:%.*]] = tail call i32 @llvm.vector.reduce.smin.v16i32(<16 x i32> [[A0:%.*]])
+; SSE-NEXT: ret i32 [[TMP1]]
+;
+; AVX2-LABEL: @test_reduce_v16i32(
+; AVX2-NEXT: [[TMP1:%.*]] = tail call i32 @llvm.vector.reduce.smin.v16i32(<16 x i32> [[A0:%.*]])
+; AVX2-NEXT: ret i32 [[TMP1]]
+;
+; AVX512-LABEL: @test_reduce_v16i32(
+; AVX512-NEXT: [[TMP1:%.*]] = shufflevector <16 x i32> [[A0:%.*]], <16 x i32> poison, <16 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX512-NEXT: [[TMP2:%.*]] = tail call <16 x i32> @llvm.smin.v16i32(<16 x i32> [[A0]], <16 x i32> [[TMP1]])
+; AVX512-NEXT: [[TMP3:%.*]] = shufflevector <16 x i32> [[TMP2]], <16 x i32> poison, <16 x i32> <i32 4, i32 5, i32 6, i32 7, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX512-NEXT: [[TMP4:%.*]] = tail call <16 x i32> @llvm.smin.v16i32(<16 x i32> [[TMP2]], <16 x i32> [[TMP3]])
+; AVX512-NEXT: [[TMP5:%.*]] = shufflevector <16 x i32> [[TMP4]], <16 x i32> poison, <16 x i32> <i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX512-NEXT: [[TMP6:%.*]] = tail call <16 x i32> @llvm.smin.v16i32(<16 x i32> [[TMP4]], <16 x i32> [[TMP5]])
+; AVX512-NEXT: [[TMP7:%.*]] = shufflevector <16 x i32> [[TMP6]], <16 x i32> poison, <16 x i32> <i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX512-NEXT: [[TMP8:%.*]] = tail call <16 x i32> @llvm.smin.v16i32(<16 x i32> [[TMP6]], <16 x i32> [[TMP7]])
+; AVX512-NEXT: [[TMP9:%.*]] = extractelement <16 x i32> [[TMP8]], i64 0
+; AVX512-NEXT: ret i32 [[TMP9]]
+;
+ %1 = shufflevector <16 x i32> %a0, <16 x i32> undef, <16 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp slt <16 x i32> %a0, %1
+ %3 = select <16 x i1> %2, <16 x i32> %a0, <16 x i32> %1
+ %4 = shufflevector <16 x i32> %3, <16 x i32> undef, <16 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp slt <16 x i32> %3, %4
+ %6 = select <16 x i1> %5, <16 x i32> %3, <16 x i32> %4
+ %7 = shufflevector <16 x i32> %6, <16 x i32> undef, <16 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp slt <16 x i32> %6, %7
+ %9 = select <16 x i1> %8, <16 x i32> %6, <16 x i32> %7
+ %10 = shufflevector <16 x i32> %9, <16 x i32> undef, <16 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %11 = icmp slt <16 x i32> %9, %10
+ %12 = select <16 x i1> %11, <16 x i32> %9, <16 x i32> %10
+ %13 = extractelement <16 x i32> %12, i32 0
+ ret i32 %13
+}
+
+define i16 @test_reduce_v32i16(<32 x i16> %a0) {
+; CHECK-LABEL: @test_reduce_v32i16(
+; CHECK-NEXT: [[TMP16:%.*]] = tail call i16 @llvm.vector.reduce.smin.v32i16(<32 x i16> [[A0:%.*]])
+; CHECK-NEXT: ret i16 [[TMP16]]
+;
+ %1 = shufflevector <32 x i16> %a0, <32 x i16> undef, <32 x i32> <i32 16, i32 17, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23, i32 24, i32 25, i32 26, i32 27, i32 28, i32 29, i32 30, i32 31, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp slt <32 x i16> %a0, %1
+ %3 = select <32 x i1> %2, <32 x i16> %a0, <32 x i16> %1
+ %4 = shufflevector <32 x i16> %3, <32 x i16> undef, <32 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp slt <32 x i16> %3, %4
+ %6 = select <32 x i1> %5, <32 x i16> %3, <32 x i16> %4
+ %7 = shufflevector <32 x i16> %6, <32 x i16> undef, <32 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp slt <32 x i16> %6, %7
+ %9 = select <32 x i1> %8, <32 x i16> %6, <32 x i16> %7
+ %10 = shufflevector <32 x i16> %9, <32 x i16> undef, <32 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %11 = icmp slt <32 x i16> %9, %10
+ %12 = select <32 x i1> %11, <32 x i16> %9, <32 x i16> %10
+ %13 = shufflevector <32 x i16> %12, <32 x i16> undef, <32 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %14 = icmp slt <32 x i16> %12, %13
+ %15 = select <32 x i1> %14, <32 x i16> %12, <32 x i16> %13
+ %16 = extractelement <32 x i16> %15, i32 0
+ ret i16 %16
+}
+
+define i8 @test_reduce_v64i8(<64 x i8> %a0) {
+; CHECK-LABEL: @test_reduce_v64i8(
+; CHECK-NEXT: [[TMP19:%.*]] = tail call i8 @llvm.vector.reduce.smin.v64i8(<64 x i8> [[A0:%.*]])
+; CHECK-NEXT: ret i8 [[TMP19]]
+;
+ %1 = shufflevector <64 x i8> %a0, <64 x i8> undef, <64 x i32> <i32 32, i32 33, i32 34, i32 35, i32 36, i32 37, i32 38, i32 39, i32 40, i32 41, i32 42, i32 43, i32 44, i32 45, i32 46, i32 47, i32 48, i32 49, i32 50, i32 51, i32 52, i32 53, i32 54, i32 55, i32 56, i32 57, i32 58, i32 59, i32 60, i32 61, i32 62, i32 63, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp slt <64 x i8> %a0, %1
+ %3 = select <64 x i1> %2, <64 x i8> %a0, <64 x i8> %1
+ %4 = shufflevector <64 x i8> %3, <64 x i8> undef, <64 x i32> <i32 16, i32 17, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23, i32 24, i32 25, i32 26, i32 27, i32 28, i32 29, i32 30, i32 31, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp slt <64 x i8> %3, %4
+ %6 = select <64 x i1> %5, <64 x i8> %3, <64 x i8> %4
+ %7 = shufflevector <64 x i8> %6, <64 x i8> undef, <64 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp slt <64 x i8> %6, %7
+ %9 = select <64 x i1> %8, <64 x i8> %6, <64 x i8> %7
+ %10 = shufflevector <64 x i8> %9, <64 x i8> undef, <64 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %11 = icmp slt <64 x i8> %9, %10
+ %12 = select <64 x i1> %11, <64 x i8> %9, <64 x i8> %10
+ %13 = shufflevector <64 x i8> %12, <64 x i8> undef, <64 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %14 = icmp slt <64 x i8> %12, %13
+ %15 = select <64 x i1> %14, <64 x i8> %12, <64 x i8> %13
+ %16 = shufflevector <64 x i8> %15, <64 x i8> undef, <64 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %17 = icmp slt <64 x i8> %15, %16
+ %18 = select <64 x i1> %17, <64 x i8> %15, <64 x i8> %16
+ %19 = extractelement <64 x i8> %18, i32 0
+ ret i8 %19
+}
+
+;
+; Partial Vector Reductions
+;
+
+define i16 @test_reduce_v16i16_v8i16(<16 x i16> %a0) {
+; CHECK-LABEL: @test_reduce_v16i16_v8i16(
+; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <16 x i16> [[A0:%.*]], <16 x i16> poison, <16 x i32> <i32 4, i32 5, i32 6, i32 7, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP2:%.*]] = tail call <16 x i16> @llvm.smin.v16i16(<16 x i16> [[A0]], <16 x i16> [[TMP1]])
+; CHECK-NEXT: [[TMP3:%.*]] = shufflevector <16 x i16> [[TMP2]], <16 x i16> poison, <16 x i32> <i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP4:%.*]] = tail call <16 x i16> @llvm.smin.v16i16(<16 x i16> [[TMP2]], <16 x i16> [[TMP3]])
+; CHECK-NEXT: [[TMP5:%.*]] = shufflevector <16 x i16> [[TMP4]], <16 x i16> poison, <16 x i32> <i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP6:%.*]] = tail call <16 x i16> @llvm.smin.v16i16(<16 x i16> [[TMP4]], <16 x i16> [[TMP5]])
+; CHECK-NEXT: [[TMP10:%.*]] = extractelement <16 x i16> [[TMP6]], i64 0
+; CHECK-NEXT: ret i16 [[TMP10]]
+;
+ %1 = shufflevector <16 x i16> %a0, <16 x i16> undef, <16 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp slt <16 x i16> %a0, %1
+ %3 = select <16 x i1> %2, <16 x i16> %a0, <16 x i16> %1
+ %4 = shufflevector <16 x i16> %3, <16 x i16> undef, <16 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp slt <16 x i16> %3, %4
+ %6 = select <16 x i1> %5, <16 x i16> %3, <16 x i16> %4
+ %7 = shufflevector <16 x i16> %6, <16 x i16> undef, <16 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp slt <16 x i16> %6, %7
+ %9 = select <16 x i1> %8, <16 x i16> %6, <16 x i16> %7
+ %10 = extractelement <16 x i16> %9, i32 0
+ ret i16 %10
+}
+
+define i16 @test_reduce_v32i16_v8i16(<32 x i16> %a0) {
+; CHECK-LABEL: @test_reduce_v32i16_v8i16(
+; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <32 x i16> [[A0:%.*]], <32 x i16> poison, <32 x i32> <i32 4, i32 5, i32 6, i32 7, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP2:%.*]] = tail call <32 x i16> @llvm.smin.v32i16(<32 x i16> [[A0]], <32 x i16> [[TMP1]])
+; CHECK-NEXT: [[TMP3:%.*]] = shufflevector <32 x i16> [[TMP2]], <32 x i16> poison, <32 x i32> <i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP4:%.*]] = tail call <32 x i16> @llvm.smin.v32i16(<32 x i16> [[TMP2]], <32 x i16> [[TMP3]])
+; CHECK-NEXT: [[TMP5:%.*]] = shufflevector <32 x i16> [[TMP4]], <32 x i16> poison, <32 x i32> <i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP6:%.*]] = tail call <32 x i16> @llvm.smin.v32i16(<32 x i16> [[TMP4]], <32 x i16> [[TMP5]])
+; CHECK-NEXT: [[TMP10:%.*]] = extractelement <32 x i16> [[TMP6]], i64 0
+; CHECK-NEXT: ret i16 [[TMP10]]
+;
+ %1 = shufflevector <32 x i16> %a0, <32 x i16> undef, <32 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp slt <32 x i16> %a0, %1
+ %3 = select <32 x i1> %2, <32 x i16> %a0, <32 x i16> %1
+ %4 = shufflevector <32 x i16> %3, <32 x i16> undef, <32 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp slt <32 x i16> %3, %4
+ %6 = select <32 x i1> %5, <32 x i16> %3, <32 x i16> %4
+ %7 = shufflevector <32 x i16> %6, <32 x i16> undef, <32 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp slt <32 x i16> %6, %7
+ %9 = select <32 x i1> %8, <32 x i16> %6, <32 x i16> %7
+ %10 = extractelement <32 x i16> %9, i32 0
+ ret i16 %10
+}
+
+define i8 @test_reduce_v32i8_v16i8(<32 x i8> %a0) {
+; CHECK-LABEL: @test_reduce_v32i8_v16i8(
+; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <32 x i8> [[A0:%.*]], <32 x i8> poison, <32 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP2:%.*]] = tail call <32 x i8> @llvm.smin.v32i8(<32 x i8> [[A0]], <32 x i8> [[TMP1]])
+; CHECK-NEXT: [[TMP3:%.*]] = shufflevector <32 x i8> [[TMP2]], <32 x i8> poison, <32 x i32> <i32 4, i32 5, i32 6, i32 7, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP4:%.*]] = tail call <32 x i8> @llvm.smin.v32i8(<32 x i8> [[TMP2]], <32 x i8> [[TMP3]])
+; CHECK-NEXT: [[TMP5:%.*]] = shufflevector <32 x i8> [[TMP4]], <32 x i8> poison, <32 x i32> <i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP6:%.*]] = tail call <32 x i8> @llvm.smin.v32i8(<32 x i8> [[TMP4]], <32 x i8> [[TMP5]])
+; CHECK-NEXT: [[TMP7:%.*]] = shufflevector <32 x i8> [[TMP6]], <32 x i8> poison, <32 x i32> <i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP8:%.*]] = tail call <32 x i8> @llvm.smin.v32i8(<32 x i8> [[TMP6]], <32 x i8> [[TMP7]])
+; CHECK-NEXT: [[TMP13:%.*]] = extractelement <32 x i8> [[TMP8]], i64 0
+; CHECK-NEXT: ret i8 [[TMP13]]
+;
+ %1 = shufflevector <32 x i8> %a0, <32 x i8> undef, <32 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp slt <32 x i8> %a0, %1
+ %3 = select <32 x i1> %2, <32 x i8> %a0, <32 x i8> %1
+ %4 = shufflevector <32 x i8> %3, <32 x i8> undef, <32 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp slt <32 x i8> %3, %4
+ %6 = select <32 x i1> %5, <32 x i8> %3, <32 x i8> %4
+ %7 = shufflevector <32 x i8> %6, <32 x i8> undef, <32 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp slt <32 x i8> %6, %7
+ %9 = select <32 x i1> %8, <32 x i8> %6, <32 x i8> %7
+ %10 = shufflevector <32 x i8> %9, <32 x i8> undef, <32 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %11 = icmp slt <32 x i8> %9, %10
+ %12 = select <32 x i1> %11, <32 x i8> %9, <32 x i8> %10
+ %13 = extractelement <32 x i8> %12, i32 0
+ ret i8 %13
+}
+
+define i8 @test_reduce_v64i8_v16i8(<64 x i8> %a0) {
+; CHECK-LABEL: @test_reduce_v64i8_v16i8(
+; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <64 x i8> [[A0:%.*]], <64 x i8> poison, <64 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP2:%.*]] = tail call <64 x i8> @llvm.smin.v64i8(<64 x i8> [[A0]], <64 x i8> [[TMP1]])
+; CHECK-NEXT: [[TMP3:%.*]] = shufflevector <64 x i8> [[TMP2]], <64 x i8> poison, <64 x i32> <i32 4, i32 5, i32 6, i32 7, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP4:%.*]] = tail call <64 x i8> @llvm.smin.v64i8(<64 x i8> [[TMP2]], <64 x i8> [[TMP3]])
+; CHECK-NEXT: [[TMP5:%.*]] = shufflevector <64 x i8> [[TMP4]], <64 x i8> poison, <64 x i32> <i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP6:%.*]] = tail call <64 x i8> @llvm.smin.v64i8(<64 x i8> [[TMP4]], <64 x i8> [[TMP5]])
+; CHECK-NEXT: [[TMP7:%.*]] = shufflevector <64 x i8> [[TMP6]], <64 x i8> poison, <64 x i32> <i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP8:%.*]] = tail call <64 x i8> @llvm.smin.v64i8(<64 x i8> [[TMP6]], <64 x i8> [[TMP7]])
+; CHECK-NEXT: [[TMP13:%.*]] = extractelement <64 x i8> [[TMP8]], i64 0
+; CHECK-NEXT: ret i8 [[TMP13]]
+;
+ %1 = shufflevector <64 x i8> %a0, <64 x i8> undef, <64 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp slt <64 x i8> %a0, %1
+ %3 = select <64 x i1> %2, <64 x i8> %a0, <64 x i8> %1
+ %4 = shufflevector <64 x i8> %3, <64 x i8> undef, <64 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp slt <64 x i8> %3, %4
+ %6 = select <64 x i1> %5, <64 x i8> %3, <64 x i8> %4
+ %7 = shufflevector <64 x i8> %6, <64 x i8> undef, <64 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp slt <64 x i8> %6, %7
+ %9 = select <64 x i1> %8, <64 x i8> %6, <64 x i8> %7
+ %10 = shufflevector <64 x i8> %9, <64 x i8> undef, <64 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %11 = icmp slt <64 x i8> %9, %10
+ %12 = select <64 x i1> %11, <64 x i8> %9, <64 x i8> %10
+ %13 = extractelement <64 x i8> %12, i32 0
+ ret i8 %13
+}
diff --git a/llvm/test/Transforms/PhaseOrdering/X86/horizontal-reduce-umax.ll b/llvm/test/Transforms/PhaseOrdering/X86/horizontal-reduce-umax.ll
new file mode 100644
index 0000000000000..dbb448c4b96e5
--- /dev/null
+++ b/llvm/test/Transforms/PhaseOrdering/X86/horizontal-reduce-umax.ll
@@ -0,0 +1,419 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
+; RUN: opt < %s -passes="default<O2>" -S -mtriple=i686-- -mcpu=x86-64 | FileCheck %s --check-prefixes=CHECK,SSE
+; RUN: opt < %s -passes="default<O2>" -S -mtriple=i686-- -mcpu=x86-64-v2 | FileCheck %s --check-prefixes=CHECK,SSE
+; RUN: opt < %s -passes="default<O2>" -S -mtriple=i686-- -mcpu=x86-64-v3 | FileCheck %s --check-prefixes=CHECK,AVX,AVX2
+; RUN: opt < %s -passes="default<O2>" -S -mtriple=i686-- -mcpu=x86-64-v4 | FileCheck %s --check-prefixes=CHECK,AVX,AVX512
+; RUN: opt < %s -passes="default<O2>" -S -mtriple=x86_64-- -mcpu=x86-64 | FileCheck %s --check-prefixes=CHECK,SSE
+; RUN: opt < %s -passes="default<O2>" -S -mtriple=x86_64-- -mcpu=x86-64-v2 | FileCheck %s --check-prefixes=CHECK,SSE
+; RUN: opt < %s -passes="default<O2>" -S -mtriple=x86_64-- -mcpu=x86-64-v3 | FileCheck %s --check-prefixes=CHECK,AVX,AVX2
+; RUN: opt < %s -passes="default<O2>" -S -mtriple=x86_64-- -mcpu=x86-64-v4 | FileCheck %s --check-prefixes=CHECK,AVX,AVX512
+
+;
+; 128-bit Vectors
+;
+
+define i64 @test_reduce_v2i64(<2 x i64> %a0) {
+; CHECK-LABEL: @test_reduce_v2i64(
+; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <2 x i64> [[A0:%.*]], <2 x i64> poison, <2 x i32> <i32 1, i32 poison>
+; CHECK-NEXT: [[TMP2:%.*]] = tail call <2 x i64> @llvm.umax.v2i64(<2 x i64> [[A0]], <2 x i64> [[TMP1]])
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <2 x i64> [[TMP2]], i64 0
+; CHECK-NEXT: ret i64 [[TMP4]]
+;
+ %1 = shufflevector <2 x i64> %a0, <2 x i64> undef, <2 x i32> <i32 1, i32 undef>
+ %2 = icmp ugt <2 x i64> %a0, %1
+ %3 = select <2 x i1> %2, <2 x i64> %a0, <2 x i64> %1
+ %4 = extractelement <2 x i64> %3, i32 0
+ ret i64 %4
+}
+
+define i32 @test_reduce_v4i32(<4 x i32> %a0) {
+; CHECK-LABEL: @test_reduce_v4i32(
+; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <4 x i32> [[A0:%.*]], <4 x i32> poison, <4 x i32> <i32 2, i32 3, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP2:%.*]] = tail call <4 x i32> @llvm.umax.v4i32(<4 x i32> [[A0]], <4 x i32> [[TMP1]])
+; CHECK-NEXT: [[TMP3:%.*]] = shufflevector <4 x i32> [[TMP2]], <4 x i32> poison, <4 x i32> <i32 1, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP4:%.*]] = tail call <4 x i32> @llvm.umax.v4i32(<4 x i32> [[TMP2]], <4 x i32> [[TMP3]])
+; CHECK-NEXT: [[TMP7:%.*]] = extractelement <4 x i32> [[TMP4]], i64 0
+; CHECK-NEXT: ret i32 [[TMP7]]
+;
+ %1 = shufflevector <4 x i32> %a0, <4 x i32> undef, <4 x i32> <i32 2, i32 3, i32 undef, i32 undef>
+ %2 = icmp ugt <4 x i32> %a0, %1
+ %3 = select <4 x i1> %2, <4 x i32> %a0, <4 x i32> %1
+ %4 = shufflevector <4 x i32> %3, <4 x i32> undef, <4 x i32> <i32 1, i32 undef, i32 undef, i32 undef>
+ %5 = icmp ugt <4 x i32> %3, %4
+ %6 = select <4 x i1> %5, <4 x i32> %3, <4 x i32> %4
+ %7 = extractelement <4 x i32> %6, i32 0
+ ret i32 %7
+}
+
+define i16 @test_reduce_v8i16(<8 x i16> %a0) {
+; CHECK-LABEL: @test_reduce_v8i16(
+; CHECK-NEXT: [[TMP10:%.*]] = tail call i16 @llvm.vector.reduce.umax.v8i16(<8 x i16> [[A0:%.*]])
+; CHECK-NEXT: ret i16 [[TMP10]]
+;
+ %1 = shufflevector <8 x i16> %a0, <8 x i16> undef, <8 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp ugt <8 x i16> %a0, %1
+ %3 = select <8 x i1> %2, <8 x i16> %a0, <8 x i16> %1
+ %4 = shufflevector <8 x i16> %3, <8 x i16> undef, <8 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp ugt <8 x i16> %3, %4
+ %6 = select <8 x i1> %5, <8 x i16> %3, <8 x i16> %4
+ %7 = shufflevector <8 x i16> %6, <8 x i16> undef, <8 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp ugt <8 x i16> %6, %7
+ %9 = select <8 x i1> %8, <8 x i16> %6, <8 x i16> %7
+ %10 = extractelement <8 x i16> %9, i32 0
+ ret i16 %10
+}
+
+define i8 @test_reduce_v16i8(<16 x i8> %a0) {
+; CHECK-LABEL: @test_reduce_v16i8(
+; CHECK-NEXT: [[TMP13:%.*]] = tail call i8 @llvm.vector.reduce.umax.v16i8(<16 x i8> [[A0:%.*]])
+; CHECK-NEXT: ret i8 [[TMP13]]
+;
+ %1 = shufflevector <16 x i8> %a0, <16 x i8> undef, <16 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp ugt <16 x i8> %a0, %1
+ %3 = select <16 x i1> %2, <16 x i8> %a0, <16 x i8> %1
+ %4 = shufflevector <16 x i8> %3, <16 x i8> undef, <16 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp ugt <16 x i8> %3, %4
+ %6 = select <16 x i1> %5, <16 x i8> %3, <16 x i8> %4
+ %7 = shufflevector <16 x i8> %6, <16 x i8> undef, <16 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp ugt <16 x i8> %6, %7
+ %9 = select <16 x i1> %8, <16 x i8> %6, <16 x i8> %7
+ %10 = shufflevector <16 x i8> %9, <16 x i8> undef, <16 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %11 = icmp ugt <16 x i8> %9, %10
+ %12 = select <16 x i1> %11, <16 x i8> %9, <16 x i8> %10
+ %13 = extractelement <16 x i8> %12, i32 0
+ ret i8 %13
+}
+
+;
+; 256-bit Vectors
+;
+
+define i64 @test_reduce_v4i64(<4 x i64> %a0) {
+; SSE-LABEL: @test_reduce_v4i64(
+; SSE-NEXT: [[TMP1:%.*]] = tail call i64 @llvm.vector.reduce.umax.v4i64(<4 x i64> [[A0:%.*]])
+; SSE-NEXT: ret i64 [[TMP1]]
+;
+; AVX-LABEL: @test_reduce_v4i64(
+; AVX-NEXT: [[TMP1:%.*]] = shufflevector <4 x i64> [[A0:%.*]], <4 x i64> poison, <4 x i32> <i32 2, i32 3, i32 poison, i32 poison>
+; AVX-NEXT: [[TMP2:%.*]] = tail call <4 x i64> @llvm.umax.v4i64(<4 x i64> [[A0]], <4 x i64> [[TMP1]])
+; AVX-NEXT: [[TMP3:%.*]] = shufflevector <4 x i64> [[TMP2]], <4 x i64> poison, <4 x i32> <i32 1, i32 poison, i32 poison, i32 poison>
+; AVX-NEXT: [[TMP4:%.*]] = tail call <4 x i64> @llvm.umax.v4i64(<4 x i64> [[TMP2]], <4 x i64> [[TMP3]])
+; AVX-NEXT: [[TMP5:%.*]] = extractelement <4 x i64> [[TMP4]], i64 0
+; AVX-NEXT: ret i64 [[TMP5]]
+;
+ %1 = shufflevector <4 x i64> %a0, <4 x i64> undef, <4 x i32> <i32 2, i32 3, i32 undef, i32 undef>
+ %2 = icmp ugt <4 x i64> %a0, %1
+ %3 = select <4 x i1> %2, <4 x i64> %a0, <4 x i64> %1
+ %4 = shufflevector <4 x i64> %3, <4 x i64> undef, <4 x i32> <i32 1, i32 undef, i32 undef, i32 undef>
+ %5 = icmp ugt <4 x i64> %3, %4
+ %6 = select <4 x i1> %5, <4 x i64> %3, <4 x i64> %4
+ %7 = extractelement <4 x i64> %6, i32 0
+ ret i64 %7
+}
+
+define i32 @test_reduce_v8i32(<8 x i32> %a0) {
+; SSE-LABEL: @test_reduce_v8i32(
+; SSE-NEXT: [[TMP1:%.*]] = tail call i32 @llvm.vector.reduce.umax.v8i32(<8 x i32> [[A0:%.*]])
+; SSE-NEXT: ret i32 [[TMP1]]
+;
+; AVX-LABEL: @test_reduce_v8i32(
+; AVX-NEXT: [[TMP1:%.*]] = shufflevector <8 x i32> [[A0:%.*]], <8 x i32> poison, <8 x i32> <i32 4, i32 5, i32 6, i32 7, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX-NEXT: [[TMP2:%.*]] = tail call <8 x i32> @llvm.umax.v8i32(<8 x i32> [[A0]], <8 x i32> [[TMP1]])
+; AVX-NEXT: [[TMP3:%.*]] = shufflevector <8 x i32> [[TMP2]], <8 x i32> poison, <8 x i32> <i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX-NEXT: [[TMP4:%.*]] = tail call <8 x i32> @llvm.umax.v8i32(<8 x i32> [[TMP2]], <8 x i32> [[TMP3]])
+; AVX-NEXT: [[TMP5:%.*]] = shufflevector <8 x i32> [[TMP4]], <8 x i32> poison, <8 x i32> <i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX-NEXT: [[TMP6:%.*]] = tail call <8 x i32> @llvm.umax.v8i32(<8 x i32> [[TMP4]], <8 x i32> [[TMP5]])
+; AVX-NEXT: [[TMP7:%.*]] = extractelement <8 x i32> [[TMP6]], i64 0
+; AVX-NEXT: ret i32 [[TMP7]]
+;
+ %1 = shufflevector <8 x i32> %a0, <8 x i32> undef, <8 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp ugt <8 x i32> %a0, %1
+ %3 = select <8 x i1> %2, <8 x i32> %a0, <8 x i32> %1
+ %4 = shufflevector <8 x i32> %3, <8 x i32> undef, <8 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp ugt <8 x i32> %3, %4
+ %6 = select <8 x i1> %5, <8 x i32> %3, <8 x i32> %4
+ %7 = shufflevector <8 x i32> %6, <8 x i32> undef, <8 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp ugt <8 x i32> %6, %7
+ %9 = select <8 x i1> %8, <8 x i32> %6, <8 x i32> %7
+ %10 = extractelement <8 x i32> %9, i32 0
+ ret i32 %10
+}
+
+define i16 @test_reduce_v16i16(<16 x i16> %a0) {
+; CHECK-LABEL: @test_reduce_v16i16(
+; CHECK-NEXT: [[TMP13:%.*]] = tail call i16 @llvm.vector.reduce.umax.v16i16(<16 x i16> [[A0:%.*]])
+; CHECK-NEXT: ret i16 [[TMP13]]
+;
+ %1 = shufflevector <16 x i16> %a0, <16 x i16> undef, <16 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp ugt <16 x i16> %a0, %1
+ %3 = select <16 x i1> %2, <16 x i16> %a0, <16 x i16> %1
+ %4 = shufflevector <16 x i16> %3, <16 x i16> undef, <16 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp ugt <16 x i16> %3, %4
+ %6 = select <16 x i1> %5, <16 x i16> %3, <16 x i16> %4
+ %7 = shufflevector <16 x i16> %6, <16 x i16> undef, <16 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp ugt <16 x i16> %6, %7
+ %9 = select <16 x i1> %8, <16 x i16> %6, <16 x i16> %7
+ %10 = shufflevector <16 x i16> %9, <16 x i16> undef, <16 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %11 = icmp ugt <16 x i16> %9, %10
+ %12 = select <16 x i1> %11, <16 x i16> %9, <16 x i16> %10
+ %13 = extractelement <16 x i16> %12, i32 0
+ ret i16 %13
+}
+
+define i8 @test_reduce_v32i8(<32 x i8> %a0) {
+; CHECK-LABEL: @test_reduce_v32i8(
+; CHECK-NEXT: [[TMP16:%.*]] = tail call i8 @llvm.vector.reduce.umax.v32i8(<32 x i8> [[A0:%.*]])
+; CHECK-NEXT: ret i8 [[TMP16]]
+;
+ %1 = shufflevector <32 x i8> %a0, <32 x i8> undef, <32 x i32> <i32 16, i32 17, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23, i32 24, i32 25, i32 26, i32 27, i32 28, i32 29, i32 30, i32 31, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp ugt <32 x i8> %a0, %1
+ %3 = select <32 x i1> %2, <32 x i8> %a0, <32 x i8> %1
+ %4 = shufflevector <32 x i8> %3, <32 x i8> undef, <32 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp ugt <32 x i8> %3, %4
+ %6 = select <32 x i1> %5, <32 x i8> %3, <32 x i8> %4
+ %7 = shufflevector <32 x i8> %6, <32 x i8> undef, <32 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp ugt <32 x i8> %6, %7
+ %9 = select <32 x i1> %8, <32 x i8> %6, <32 x i8> %7
+ %10 = shufflevector <32 x i8> %9, <32 x i8> undef, <32 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %11 = icmp ugt <32 x i8> %9, %10
+ %12 = select <32 x i1> %11, <32 x i8> %9, <32 x i8> %10
+ %13 = shufflevector <32 x i8> %12, <32 x i8> undef, <32 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %14 = icmp ugt <32 x i8> %12, %13
+ %15 = select <32 x i1> %14, <32 x i8> %12, <32 x i8> %13
+ %16 = extractelement <32 x i8> %15, i32 0
+ ret i8 %16
+}
+
+;
+; 512-bit Vectors
+;
+
+define i64 @test_reduce_v8i64(<8 x i64> %a0) {
+; SSE-LABEL: @test_reduce_v8i64(
+; SSE-NEXT: [[TMP1:%.*]] = tail call i64 @llvm.vector.reduce.umax.v8i64(<8 x i64> [[A0:%.*]])
+; SSE-NEXT: ret i64 [[TMP1]]
+;
+; AVX2-LABEL: @test_reduce_v8i64(
+; AVX2-NEXT: [[TMP1:%.*]] = tail call i64 @llvm.vector.reduce.umax.v8i64(<8 x i64> [[A0:%.*]])
+; AVX2-NEXT: ret i64 [[TMP1]]
+;
+; AVX512-LABEL: @test_reduce_v8i64(
+; AVX512-NEXT: [[TMP1:%.*]] = shufflevector <8 x i64> [[A0:%.*]], <8 x i64> poison, <8 x i32> <i32 4, i32 5, i32 6, i32 7, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX512-NEXT: [[TMP2:%.*]] = tail call <8 x i64> @llvm.umax.v8i64(<8 x i64> [[A0]], <8 x i64> [[TMP1]])
+; AVX512-NEXT: [[TMP3:%.*]] = shufflevector <8 x i64> [[TMP2]], <8 x i64> poison, <8 x i32> <i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX512-NEXT: [[TMP4:%.*]] = tail call <8 x i64> @llvm.umax.v8i64(<8 x i64> [[TMP2]], <8 x i64> [[TMP3]])
+; AVX512-NEXT: [[TMP5:%.*]] = shufflevector <8 x i64> [[TMP4]], <8 x i64> poison, <8 x i32> <i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX512-NEXT: [[TMP6:%.*]] = tail call <8 x i64> @llvm.umax.v8i64(<8 x i64> [[TMP4]], <8 x i64> [[TMP5]])
+; AVX512-NEXT: [[TMP7:%.*]] = extractelement <8 x i64> [[TMP6]], i64 0
+; AVX512-NEXT: ret i64 [[TMP7]]
+;
+ %1 = shufflevector <8 x i64> %a0, <8 x i64> undef, <8 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp ugt <8 x i64> %a0, %1
+ %3 = select <8 x i1> %2, <8 x i64> %a0, <8 x i64> %1
+ %4 = shufflevector <8 x i64> %3, <8 x i64> undef, <8 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp ugt <8 x i64> %3, %4
+ %6 = select <8 x i1> %5, <8 x i64> %3, <8 x i64> %4
+ %7 = shufflevector <8 x i64> %6, <8 x i64> undef, <8 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp ugt <8 x i64> %6, %7
+ %9 = select <8 x i1> %8, <8 x i64> %6, <8 x i64> %7
+ %10 = extractelement <8 x i64> %9, i32 0
+ ret i64 %10
+}
+
+define i32 @test_reduce_v16i32(<16 x i32> %a0) {
+; SSE-LABEL: @test_reduce_v16i32(
+; SSE-NEXT: [[TMP1:%.*]] = tail call i32 @llvm.vector.reduce.umax.v16i32(<16 x i32> [[A0:%.*]])
+; SSE-NEXT: ret i32 [[TMP1]]
+;
+; AVX2-LABEL: @test_reduce_v16i32(
+; AVX2-NEXT: [[TMP1:%.*]] = tail call i32 @llvm.vector.reduce.umax.v16i32(<16 x i32> [[A0:%.*]])
+; AVX2-NEXT: ret i32 [[TMP1]]
+;
+; AVX512-LABEL: @test_reduce_v16i32(
+; AVX512-NEXT: [[TMP1:%.*]] = shufflevector <16 x i32> [[A0:%.*]], <16 x i32> poison, <16 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX512-NEXT: [[TMP2:%.*]] = tail call <16 x i32> @llvm.umax.v16i32(<16 x i32> [[A0]], <16 x i32> [[TMP1]])
+; AVX512-NEXT: [[TMP3:%.*]] = shufflevector <16 x i32> [[TMP2]], <16 x i32> poison, <16 x i32> <i32 4, i32 5, i32 6, i32 7, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX512-NEXT: [[TMP4:%.*]] = tail call <16 x i32> @llvm.umax.v16i32(<16 x i32> [[TMP2]], <16 x i32> [[TMP3]])
+; AVX512-NEXT: [[TMP5:%.*]] = shufflevector <16 x i32> [[TMP4]], <16 x i32> poison, <16 x i32> <i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX512-NEXT: [[TMP6:%.*]] = tail call <16 x i32> @llvm.umax.v16i32(<16 x i32> [[TMP4]], <16 x i32> [[TMP5]])
+; AVX512-NEXT: [[TMP7:%.*]] = shufflevector <16 x i32> [[TMP6]], <16 x i32> poison, <16 x i32> <i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX512-NEXT: [[TMP8:%.*]] = tail call <16 x i32> @llvm.umax.v16i32(<16 x i32> [[TMP6]], <16 x i32> [[TMP7]])
+; AVX512-NEXT: [[TMP9:%.*]] = extractelement <16 x i32> [[TMP8]], i64 0
+; AVX512-NEXT: ret i32 [[TMP9]]
+;
+ %1 = shufflevector <16 x i32> %a0, <16 x i32> undef, <16 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp ugt <16 x i32> %a0, %1
+ %3 = select <16 x i1> %2, <16 x i32> %a0, <16 x i32> %1
+ %4 = shufflevector <16 x i32> %3, <16 x i32> undef, <16 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp ugt <16 x i32> %3, %4
+ %6 = select <16 x i1> %5, <16 x i32> %3, <16 x i32> %4
+ %7 = shufflevector <16 x i32> %6, <16 x i32> undef, <16 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp ugt <16 x i32> %6, %7
+ %9 = select <16 x i1> %8, <16 x i32> %6, <16 x i32> %7
+ %10 = shufflevector <16 x i32> %9, <16 x i32> undef, <16 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %11 = icmp ugt <16 x i32> %9, %10
+ %12 = select <16 x i1> %11, <16 x i32> %9, <16 x i32> %10
+ %13 = extractelement <16 x i32> %12, i32 0
+ ret i32 %13
+}
+
+define i16 @test_reduce_v32i16(<32 x i16> %a0) {
+; CHECK-LABEL: @test_reduce_v32i16(
+; CHECK-NEXT: [[TMP16:%.*]] = tail call i16 @llvm.vector.reduce.umax.v32i16(<32 x i16> [[A0:%.*]])
+; CHECK-NEXT: ret i16 [[TMP16]]
+;
+ %1 = shufflevector <32 x i16> %a0, <32 x i16> undef, <32 x i32> <i32 16, i32 17, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23, i32 24, i32 25, i32 26, i32 27, i32 28, i32 29, i32 30, i32 31, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp ugt <32 x i16> %a0, %1
+ %3 = select <32 x i1> %2, <32 x i16> %a0, <32 x i16> %1
+ %4 = shufflevector <32 x i16> %3, <32 x i16> undef, <32 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp ugt <32 x i16> %3, %4
+ %6 = select <32 x i1> %5, <32 x i16> %3, <32 x i16> %4
+ %7 = shufflevector <32 x i16> %6, <32 x i16> undef, <32 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp ugt <32 x i16> %6, %7
+ %9 = select <32 x i1> %8, <32 x i16> %6, <32 x i16> %7
+ %10 = shufflevector <32 x i16> %9, <32 x i16> undef, <32 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %11 = icmp ugt <32 x i16> %9, %10
+ %12 = select <32 x i1> %11, <32 x i16> %9, <32 x i16> %10
+ %13 = shufflevector <32 x i16> %12, <32 x i16> undef, <32 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %14 = icmp ugt <32 x i16> %12, %13
+ %15 = select <32 x i1> %14, <32 x i16> %12, <32 x i16> %13
+ %16 = extractelement <32 x i16> %15, i32 0
+ ret i16 %16
+}
+
+define i8 @test_reduce_v64i8(<64 x i8> %a0) {
+; CHECK-LABEL: @test_reduce_v64i8(
+; CHECK-NEXT: [[TMP19:%.*]] = tail call i8 @llvm.vector.reduce.umax.v64i8(<64 x i8> [[A0:%.*]])
+; CHECK-NEXT: ret i8 [[TMP19]]
+;
+ %1 = shufflevector <64 x i8> %a0, <64 x i8> undef, <64 x i32> <i32 32, i32 33, i32 34, i32 35, i32 36, i32 37, i32 38, i32 39, i32 40, i32 41, i32 42, i32 43, i32 44, i32 45, i32 46, i32 47, i32 48, i32 49, i32 50, i32 51, i32 52, i32 53, i32 54, i32 55, i32 56, i32 57, i32 58, i32 59, i32 60, i32 61, i32 62, i32 63, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp ugt <64 x i8> %a0, %1
+ %3 = select <64 x i1> %2, <64 x i8> %a0, <64 x i8> %1
+ %4 = shufflevector <64 x i8> %3, <64 x i8> undef, <64 x i32> <i32 16, i32 17, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23, i32 24, i32 25, i32 26, i32 27, i32 28, i32 29, i32 30, i32 31, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp ugt <64 x i8> %3, %4
+ %6 = select <64 x i1> %5, <64 x i8> %3, <64 x i8> %4
+ %7 = shufflevector <64 x i8> %6, <64 x i8> undef, <64 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp ugt <64 x i8> %6, %7
+ %9 = select <64 x i1> %8, <64 x i8> %6, <64 x i8> %7
+ %10 = shufflevector <64 x i8> %9, <64 x i8> undef, <64 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %11 = icmp ugt <64 x i8> %9, %10
+ %12 = select <64 x i1> %11, <64 x i8> %9, <64 x i8> %10
+ %13 = shufflevector <64 x i8> %12, <64 x i8> undef, <64 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %14 = icmp ugt <64 x i8> %12, %13
+ %15 = select <64 x i1> %14, <64 x i8> %12, <64 x i8> %13
+ %16 = shufflevector <64 x i8> %15, <64 x i8> undef, <64 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %17 = icmp ugt <64 x i8> %15, %16
+ %18 = select <64 x i1> %17, <64 x i8> %15, <64 x i8> %16
+ %19 = extractelement <64 x i8> %18, i32 0
+ ret i8 %19
+}
+
+;
+; Partial Vector Reductions
+;
+
+define i16 @test_reduce_v16i16_v8i16(<16 x i16> %a0) {
+; CHECK-LABEL: @test_reduce_v16i16_v8i16(
+; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <16 x i16> [[A0:%.*]], <16 x i16> poison, <16 x i32> <i32 4, i32 5, i32 6, i32 7, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP2:%.*]] = tail call <16 x i16> @llvm.umax.v16i16(<16 x i16> [[A0]], <16 x i16> [[TMP1]])
+; CHECK-NEXT: [[TMP3:%.*]] = shufflevector <16 x i16> [[TMP2]], <16 x i16> poison, <16 x i32> <i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP4:%.*]] = tail call <16 x i16> @llvm.umax.v16i16(<16 x i16> [[TMP2]], <16 x i16> [[TMP3]])
+; CHECK-NEXT: [[TMP5:%.*]] = shufflevector <16 x i16> [[TMP4]], <16 x i16> poison, <16 x i32> <i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP6:%.*]] = tail call <16 x i16> @llvm.umax.v16i16(<16 x i16> [[TMP4]], <16 x i16> [[TMP5]])
+; CHECK-NEXT: [[TMP10:%.*]] = extractelement <16 x i16> [[TMP6]], i64 0
+; CHECK-NEXT: ret i16 [[TMP10]]
+;
+ %1 = shufflevector <16 x i16> %a0, <16 x i16> undef, <16 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp ugt <16 x i16> %a0, %1
+ %3 = select <16 x i1> %2, <16 x i16> %a0, <16 x i16> %1
+ %4 = shufflevector <16 x i16> %3, <16 x i16> undef, <16 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp ugt <16 x i16> %3, %4
+ %6 = select <16 x i1> %5, <16 x i16> %3, <16 x i16> %4
+ %7 = shufflevector <16 x i16> %6, <16 x i16> undef, <16 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp ugt <16 x i16> %6, %7
+ %9 = select <16 x i1> %8, <16 x i16> %6, <16 x i16> %7
+ %10 = extractelement <16 x i16> %9, i32 0
+ ret i16 %10
+}
+
+define i16 @test_reduce_v32i16_v8i16(<32 x i16> %a0) {
+; CHECK-LABEL: @test_reduce_v32i16_v8i16(
+; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <32 x i16> [[A0:%.*]], <32 x i16> poison, <32 x i32> <i32 4, i32 5, i32 6, i32 7, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP2:%.*]] = tail call <32 x i16> @llvm.umax.v32i16(<32 x i16> [[A0]], <32 x i16> [[TMP1]])
+; CHECK-NEXT: [[TMP3:%.*]] = shufflevector <32 x i16> [[TMP2]], <32 x i16> poison, <32 x i32> <i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP4:%.*]] = tail call <32 x i16> @llvm.umax.v32i16(<32 x i16> [[TMP2]], <32 x i16> [[TMP3]])
+; CHECK-NEXT: [[TMP5:%.*]] = shufflevector <32 x i16> [[TMP4]], <32 x i16> poison, <32 x i32> <i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP6:%.*]] = tail call <32 x i16> @llvm.umax.v32i16(<32 x i16> [[TMP4]], <32 x i16> [[TMP5]])
+; CHECK-NEXT: [[TMP10:%.*]] = extractelement <32 x i16> [[TMP6]], i64 0
+; CHECK-NEXT: ret i16 [[TMP10]]
+;
+ %1 = shufflevector <32 x i16> %a0, <32 x i16> undef, <32 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp ugt <32 x i16> %a0, %1
+ %3 = select <32 x i1> %2, <32 x i16> %a0, <32 x i16> %1
+ %4 = shufflevector <32 x i16> %3, <32 x i16> undef, <32 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp ugt <32 x i16> %3, %4
+ %6 = select <32 x i1> %5, <32 x i16> %3, <32 x i16> %4
+ %7 = shufflevector <32 x i16> %6, <32 x i16> undef, <32 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp ugt <32 x i16> %6, %7
+ %9 = select <32 x i1> %8, <32 x i16> %6, <32 x i16> %7
+ %10 = extractelement <32 x i16> %9, i32 0
+ ret i16 %10
+}
+
+define i8 @test_reduce_v32i8_v16i8(<32 x i8> %a0) {
+; CHECK-LABEL: @test_reduce_v32i8_v16i8(
+; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <32 x i8> [[A0:%.*]], <32 x i8> poison, <32 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP2:%.*]] = tail call <32 x i8> @llvm.umax.v32i8(<32 x i8> [[A0]], <32 x i8> [[TMP1]])
+; CHECK-NEXT: [[TMP3:%.*]] = shufflevector <32 x i8> [[TMP2]], <32 x i8> poison, <32 x i32> <i32 4, i32 5, i32 6, i32 7, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP4:%.*]] = tail call <32 x i8> @llvm.umax.v32i8(<32 x i8> [[TMP2]], <32 x i8> [[TMP3]])
+; CHECK-NEXT: [[TMP5:%.*]] = shufflevector <32 x i8> [[TMP4]], <32 x i8> poison, <32 x i32> <i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP6:%.*]] = tail call <32 x i8> @llvm.umax.v32i8(<32 x i8> [[TMP4]], <32 x i8> [[TMP5]])
+; CHECK-NEXT: [[TMP7:%.*]] = shufflevector <32 x i8> [[TMP6]], <32 x i8> poison, <32 x i32> <i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP8:%.*]] = tail call <32 x i8> @llvm.umax.v32i8(<32 x i8> [[TMP6]], <32 x i8> [[TMP7]])
+; CHECK-NEXT: [[TMP13:%.*]] = extractelement <32 x i8> [[TMP8]], i64 0
+; CHECK-NEXT: ret i8 [[TMP13]]
+;
+ %1 = shufflevector <32 x i8> %a0, <32 x i8> undef, <32 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp ugt <32 x i8> %a0, %1
+ %3 = select <32 x i1> %2, <32 x i8> %a0, <32 x i8> %1
+ %4 = shufflevector <32 x i8> %3, <32 x i8> undef, <32 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp ugt <32 x i8> %3, %4
+ %6 = select <32 x i1> %5, <32 x i8> %3, <32 x i8> %4
+ %7 = shufflevector <32 x i8> %6, <32 x i8> undef, <32 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp ugt <32 x i8> %6, %7
+ %9 = select <32 x i1> %8, <32 x i8> %6, <32 x i8> %7
+ %10 = shufflevector <32 x i8> %9, <32 x i8> undef, <32 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %11 = icmp ugt <32 x i8> %9, %10
+ %12 = select <32 x i1> %11, <32 x i8> %9, <32 x i8> %10
+ %13 = extractelement <32 x i8> %12, i32 0
+ ret i8 %13
+}
+
+define i8 @test_reduce_v64i8_v16i8(<64 x i8> %a0) {
+; CHECK-LABEL: @test_reduce_v64i8_v16i8(
+; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <64 x i8> [[A0:%.*]], <64 x i8> poison, <64 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP2:%.*]] = tail call <64 x i8> @llvm.umax.v64i8(<64 x i8> [[A0]], <64 x i8> [[TMP1]])
+; CHECK-NEXT: [[TMP3:%.*]] = shufflevector <64 x i8> [[TMP2]], <64 x i8> poison, <64 x i32> <i32 4, i32 5, i32 6, i32 7, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP4:%.*]] = tail call <64 x i8> @llvm.umax.v64i8(<64 x i8> [[TMP2]], <64 x i8> [[TMP3]])
+; CHECK-NEXT: [[TMP5:%.*]] = shufflevector <64 x i8> [[TMP4]], <64 x i8> poison, <64 x i32> <i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP6:%.*]] = tail call <64 x i8> @llvm.umax.v64i8(<64 x i8> [[TMP4]], <64 x i8> [[TMP5]])
+; CHECK-NEXT: [[TMP7:%.*]] = shufflevector <64 x i8> [[TMP6]], <64 x i8> poison, <64 x i32> <i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP8:%.*]] = tail call <64 x i8> @llvm.umax.v64i8(<64 x i8> [[TMP6]], <64 x i8> [[TMP7]])
+; CHECK-NEXT: [[TMP13:%.*]] = extractelement <64 x i8> [[TMP8]], i64 0
+; CHECK-NEXT: ret i8 [[TMP13]]
+;
+ %1 = shufflevector <64 x i8> %a0, <64 x i8> undef, <64 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp ugt <64 x i8> %a0, %1
+ %3 = select <64 x i1> %2, <64 x i8> %a0, <64 x i8> %1
+ %4 = shufflevector <64 x i8> %3, <64 x i8> undef, <64 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp ugt <64 x i8> %3, %4
+ %6 = select <64 x i1> %5, <64 x i8> %3, <64 x i8> %4
+ %7 = shufflevector <64 x i8> %6, <64 x i8> undef, <64 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp ugt <64 x i8> %6, %7
+ %9 = select <64 x i1> %8, <64 x i8> %6, <64 x i8> %7
+ %10 = shufflevector <64 x i8> %9, <64 x i8> undef, <64 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %11 = icmp ugt <64 x i8> %9, %10
+ %12 = select <64 x i1> %11, <64 x i8> %9, <64 x i8> %10
+ %13 = extractelement <64 x i8> %12, i32 0
+ ret i8 %13
+}
diff --git a/llvm/test/Transforms/PhaseOrdering/X86/horizontal-reduce-umin.ll b/llvm/test/Transforms/PhaseOrdering/X86/horizontal-reduce-umin.ll
new file mode 100644
index 0000000000000..bd2366d49a951
--- /dev/null
+++ b/llvm/test/Transforms/PhaseOrdering/X86/horizontal-reduce-umin.ll
@@ -0,0 +1,419 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
+; RUN: opt < %s -passes="default<O2>" -S -mtriple=i686-- -mcpu=x86-64 | FileCheck %s --check-prefixes=CHECK,SSE
+; RUN: opt < %s -passes="default<O2>" -S -mtriple=i686-- -mcpu=x86-64-v2 | FileCheck %s --check-prefixes=CHECK,SSE
+; RUN: opt < %s -passes="default<O2>" -S -mtriple=i686-- -mcpu=x86-64-v3 | FileCheck %s --check-prefixes=CHECK,AVX,AVX2
+; RUN: opt < %s -passes="default<O2>" -S -mtriple=i686-- -mcpu=x86-64-v4 | FileCheck %s --check-prefixes=CHECK,AVX,AVX512
+; RUN: opt < %s -passes="default<O2>" -S -mtriple=x86_64-- -mcpu=x86-64 | FileCheck %s --check-prefixes=CHECK,SSE
+; RUN: opt < %s -passes="default<O2>" -S -mtriple=x86_64-- -mcpu=x86-64-v2 | FileCheck %s --check-prefixes=CHECK,SSE
+; RUN: opt < %s -passes="default<O2>" -S -mtriple=x86_64-- -mcpu=x86-64-v3 | FileCheck %s --check-prefixes=CHECK,AVX,AVX2
+; RUN: opt < %s -passes="default<O2>" -S -mtriple=x86_64-- -mcpu=x86-64-v4 | FileCheck %s --check-prefixes=CHECK,AVX,AVX512
+
+;
+; 128-bit Vectors
+;
+
+define i64 @test_reduce_v2i64(<2 x i64> %a0) {
+; CHECK-LABEL: @test_reduce_v2i64(
+; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <2 x i64> [[A0:%.*]], <2 x i64> poison, <2 x i32> <i32 1, i32 poison>
+; CHECK-NEXT: [[TMP2:%.*]] = tail call <2 x i64> @llvm.umin.v2i64(<2 x i64> [[A0]], <2 x i64> [[TMP1]])
+; CHECK-NEXT: [[TMP4:%.*]] = extractelement <2 x i64> [[TMP2]], i64 0
+; CHECK-NEXT: ret i64 [[TMP4]]
+;
+ %1 = shufflevector <2 x i64> %a0, <2 x i64> undef, <2 x i32> <i32 1, i32 undef>
+ %2 = icmp ult <2 x i64> %a0, %1
+ %3 = select <2 x i1> %2, <2 x i64> %a0, <2 x i64> %1
+ %4 = extractelement <2 x i64> %3, i32 0
+ ret i64 %4
+}
+
+define i32 @test_reduce_v4i32(<4 x i32> %a0) {
+; CHECK-LABEL: @test_reduce_v4i32(
+; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <4 x i32> [[A0:%.*]], <4 x i32> poison, <4 x i32> <i32 2, i32 3, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP2:%.*]] = tail call <4 x i32> @llvm.umin.v4i32(<4 x i32> [[A0]], <4 x i32> [[TMP1]])
+; CHECK-NEXT: [[TMP3:%.*]] = shufflevector <4 x i32> [[TMP2]], <4 x i32> poison, <4 x i32> <i32 1, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP4:%.*]] = tail call <4 x i32> @llvm.umin.v4i32(<4 x i32> [[TMP2]], <4 x i32> [[TMP3]])
+; CHECK-NEXT: [[TMP7:%.*]] = extractelement <4 x i32> [[TMP4]], i64 0
+; CHECK-NEXT: ret i32 [[TMP7]]
+;
+ %1 = shufflevector <4 x i32> %a0, <4 x i32> undef, <4 x i32> <i32 2, i32 3, i32 undef, i32 undef>
+ %2 = icmp ult <4 x i32> %a0, %1
+ %3 = select <4 x i1> %2, <4 x i32> %a0, <4 x i32> %1
+ %4 = shufflevector <4 x i32> %3, <4 x i32> undef, <4 x i32> <i32 1, i32 undef, i32 undef, i32 undef>
+ %5 = icmp ult <4 x i32> %3, %4
+ %6 = select <4 x i1> %5, <4 x i32> %3, <4 x i32> %4
+ %7 = extractelement <4 x i32> %6, i32 0
+ ret i32 %7
+}
+
+define i16 @test_reduce_v8i16(<8 x i16> %a0) {
+; CHECK-LABEL: @test_reduce_v8i16(
+; CHECK-NEXT: [[TMP10:%.*]] = tail call i16 @llvm.vector.reduce.umin.v8i16(<8 x i16> [[A0:%.*]])
+; CHECK-NEXT: ret i16 [[TMP10]]
+;
+ %1 = shufflevector <8 x i16> %a0, <8 x i16> undef, <8 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp ult <8 x i16> %a0, %1
+ %3 = select <8 x i1> %2, <8 x i16> %a0, <8 x i16> %1
+ %4 = shufflevector <8 x i16> %3, <8 x i16> undef, <8 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp ult <8 x i16> %3, %4
+ %6 = select <8 x i1> %5, <8 x i16> %3, <8 x i16> %4
+ %7 = shufflevector <8 x i16> %6, <8 x i16> undef, <8 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp ult <8 x i16> %6, %7
+ %9 = select <8 x i1> %8, <8 x i16> %6, <8 x i16> %7
+ %10 = extractelement <8 x i16> %9, i32 0
+ ret i16 %10
+}
+
+define i8 @test_reduce_v16i8(<16 x i8> %a0) {
+; CHECK-LABEL: @test_reduce_v16i8(
+; CHECK-NEXT: [[TMP13:%.*]] = tail call i8 @llvm.vector.reduce.umin.v16i8(<16 x i8> [[A0:%.*]])
+; CHECK-NEXT: ret i8 [[TMP13]]
+;
+ %1 = shufflevector <16 x i8> %a0, <16 x i8> undef, <16 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp ult <16 x i8> %a0, %1
+ %3 = select <16 x i1> %2, <16 x i8> %a0, <16 x i8> %1
+ %4 = shufflevector <16 x i8> %3, <16 x i8> undef, <16 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp ult <16 x i8> %3, %4
+ %6 = select <16 x i1> %5, <16 x i8> %3, <16 x i8> %4
+ %7 = shufflevector <16 x i8> %6, <16 x i8> undef, <16 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp ult <16 x i8> %6, %7
+ %9 = select <16 x i1> %8, <16 x i8> %6, <16 x i8> %7
+ %10 = shufflevector <16 x i8> %9, <16 x i8> undef, <16 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %11 = icmp ult <16 x i8> %9, %10
+ %12 = select <16 x i1> %11, <16 x i8> %9, <16 x i8> %10
+ %13 = extractelement <16 x i8> %12, i32 0
+ ret i8 %13
+}
+
+;
+; 256-bit Vectors
+;
+
+define i64 @test_reduce_v4i64(<4 x i64> %a0) {
+; SSE-LABEL: @test_reduce_v4i64(
+; SSE-NEXT: [[TMP1:%.*]] = tail call i64 @llvm.vector.reduce.umin.v4i64(<4 x i64> [[A0:%.*]])
+; SSE-NEXT: ret i64 [[TMP1]]
+;
+; AVX-LABEL: @test_reduce_v4i64(
+; AVX-NEXT: [[TMP1:%.*]] = shufflevector <4 x i64> [[A0:%.*]], <4 x i64> poison, <4 x i32> <i32 2, i32 3, i32 poison, i32 poison>
+; AVX-NEXT: [[TMP2:%.*]] = tail call <4 x i64> @llvm.umin.v4i64(<4 x i64> [[A0]], <4 x i64> [[TMP1]])
+; AVX-NEXT: [[TMP3:%.*]] = shufflevector <4 x i64> [[TMP2]], <4 x i64> poison, <4 x i32> <i32 1, i32 poison, i32 poison, i32 poison>
+; AVX-NEXT: [[TMP4:%.*]] = tail call <4 x i64> @llvm.umin.v4i64(<4 x i64> [[TMP2]], <4 x i64> [[TMP3]])
+; AVX-NEXT: [[TMP5:%.*]] = extractelement <4 x i64> [[TMP4]], i64 0
+; AVX-NEXT: ret i64 [[TMP5]]
+;
+ %1 = shufflevector <4 x i64> %a0, <4 x i64> undef, <4 x i32> <i32 2, i32 3, i32 undef, i32 undef>
+ %2 = icmp ult <4 x i64> %a0, %1
+ %3 = select <4 x i1> %2, <4 x i64> %a0, <4 x i64> %1
+ %4 = shufflevector <4 x i64> %3, <4 x i64> undef, <4 x i32> <i32 1, i32 undef, i32 undef, i32 undef>
+ %5 = icmp ult <4 x i64> %3, %4
+ %6 = select <4 x i1> %5, <4 x i64> %3, <4 x i64> %4
+ %7 = extractelement <4 x i64> %6, i32 0
+ ret i64 %7
+}
+
+define i32 @test_reduce_v8i32(<8 x i32> %a0) {
+; SSE-LABEL: @test_reduce_v8i32(
+; SSE-NEXT: [[TMP1:%.*]] = tail call i32 @llvm.vector.reduce.umin.v8i32(<8 x i32> [[A0:%.*]])
+; SSE-NEXT: ret i32 [[TMP1]]
+;
+; AVX-LABEL: @test_reduce_v8i32(
+; AVX-NEXT: [[TMP1:%.*]] = shufflevector <8 x i32> [[A0:%.*]], <8 x i32> poison, <8 x i32> <i32 4, i32 5, i32 6, i32 7, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX-NEXT: [[TMP2:%.*]] = tail call <8 x i32> @llvm.umin.v8i32(<8 x i32> [[A0]], <8 x i32> [[TMP1]])
+; AVX-NEXT: [[TMP3:%.*]] = shufflevector <8 x i32> [[TMP2]], <8 x i32> poison, <8 x i32> <i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX-NEXT: [[TMP4:%.*]] = tail call <8 x i32> @llvm.umin.v8i32(<8 x i32> [[TMP2]], <8 x i32> [[TMP3]])
+; AVX-NEXT: [[TMP5:%.*]] = shufflevector <8 x i32> [[TMP4]], <8 x i32> poison, <8 x i32> <i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX-NEXT: [[TMP6:%.*]] = tail call <8 x i32> @llvm.umin.v8i32(<8 x i32> [[TMP4]], <8 x i32> [[TMP5]])
+; AVX-NEXT: [[TMP7:%.*]] = extractelement <8 x i32> [[TMP6]], i64 0
+; AVX-NEXT: ret i32 [[TMP7]]
+;
+ %1 = shufflevector <8 x i32> %a0, <8 x i32> undef, <8 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp ult <8 x i32> %a0, %1
+ %3 = select <8 x i1> %2, <8 x i32> %a0, <8 x i32> %1
+ %4 = shufflevector <8 x i32> %3, <8 x i32> undef, <8 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp ult <8 x i32> %3, %4
+ %6 = select <8 x i1> %5, <8 x i32> %3, <8 x i32> %4
+ %7 = shufflevector <8 x i32> %6, <8 x i32> undef, <8 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp ult <8 x i32> %6, %7
+ %9 = select <8 x i1> %8, <8 x i32> %6, <8 x i32> %7
+ %10 = extractelement <8 x i32> %9, i32 0
+ ret i32 %10
+}
+
+define i16 @test_reduce_v16i16(<16 x i16> %a0) {
+; CHECK-LABEL: @test_reduce_v16i16(
+; CHECK-NEXT: [[TMP13:%.*]] = tail call i16 @llvm.vector.reduce.umin.v16i16(<16 x i16> [[A0:%.*]])
+; CHECK-NEXT: ret i16 [[TMP13]]
+;
+ %1 = shufflevector <16 x i16> %a0, <16 x i16> undef, <16 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp ult <16 x i16> %a0, %1
+ %3 = select <16 x i1> %2, <16 x i16> %a0, <16 x i16> %1
+ %4 = shufflevector <16 x i16> %3, <16 x i16> undef, <16 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp ult <16 x i16> %3, %4
+ %6 = select <16 x i1> %5, <16 x i16> %3, <16 x i16> %4
+ %7 = shufflevector <16 x i16> %6, <16 x i16> undef, <16 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp ult <16 x i16> %6, %7
+ %9 = select <16 x i1> %8, <16 x i16> %6, <16 x i16> %7
+ %10 = shufflevector <16 x i16> %9, <16 x i16> undef, <16 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %11 = icmp ult <16 x i16> %9, %10
+ %12 = select <16 x i1> %11, <16 x i16> %9, <16 x i16> %10
+ %13 = extractelement <16 x i16> %12, i32 0
+ ret i16 %13
+}
+
+define i8 @test_reduce_v32i8(<32 x i8> %a0) {
+; CHECK-LABEL: @test_reduce_v32i8(
+; CHECK-NEXT: [[TMP16:%.*]] = tail call i8 @llvm.vector.reduce.umin.v32i8(<32 x i8> [[A0:%.*]])
+; CHECK-NEXT: ret i8 [[TMP16]]
+;
+ %1 = shufflevector <32 x i8> %a0, <32 x i8> undef, <32 x i32> <i32 16, i32 17, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23, i32 24, i32 25, i32 26, i32 27, i32 28, i32 29, i32 30, i32 31, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp ult <32 x i8> %a0, %1
+ %3 = select <32 x i1> %2, <32 x i8> %a0, <32 x i8> %1
+ %4 = shufflevector <32 x i8> %3, <32 x i8> undef, <32 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp ult <32 x i8> %3, %4
+ %6 = select <32 x i1> %5, <32 x i8> %3, <32 x i8> %4
+ %7 = shufflevector <32 x i8> %6, <32 x i8> undef, <32 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp ult <32 x i8> %6, %7
+ %9 = select <32 x i1> %8, <32 x i8> %6, <32 x i8> %7
+ %10 = shufflevector <32 x i8> %9, <32 x i8> undef, <32 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %11 = icmp ult <32 x i8> %9, %10
+ %12 = select <32 x i1> %11, <32 x i8> %9, <32 x i8> %10
+ %13 = shufflevector <32 x i8> %12, <32 x i8> undef, <32 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %14 = icmp ult <32 x i8> %12, %13
+ %15 = select <32 x i1> %14, <32 x i8> %12, <32 x i8> %13
+ %16 = extractelement <32 x i8> %15, i32 0
+ ret i8 %16
+}
+
+;
+; 512-bit Vectors
+;
+
+define i64 @test_reduce_v8i64(<8 x i64> %a0) {
+; SSE-LABEL: @test_reduce_v8i64(
+; SSE-NEXT: [[TMP1:%.*]] = tail call i64 @llvm.vector.reduce.umin.v8i64(<8 x i64> [[A0:%.*]])
+; SSE-NEXT: ret i64 [[TMP1]]
+;
+; AVX2-LABEL: @test_reduce_v8i64(
+; AVX2-NEXT: [[TMP1:%.*]] = tail call i64 @llvm.vector.reduce.umin.v8i64(<8 x i64> [[A0:%.*]])
+; AVX2-NEXT: ret i64 [[TMP1]]
+;
+; AVX512-LABEL: @test_reduce_v8i64(
+; AVX512-NEXT: [[TMP1:%.*]] = shufflevector <8 x i64> [[A0:%.*]], <8 x i64> poison, <8 x i32> <i32 4, i32 5, i32 6, i32 7, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX512-NEXT: [[TMP2:%.*]] = tail call <8 x i64> @llvm.umin.v8i64(<8 x i64> [[A0]], <8 x i64> [[TMP1]])
+; AVX512-NEXT: [[TMP3:%.*]] = shufflevector <8 x i64> [[TMP2]], <8 x i64> poison, <8 x i32> <i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX512-NEXT: [[TMP4:%.*]] = tail call <8 x i64> @llvm.umin.v8i64(<8 x i64> [[TMP2]], <8 x i64> [[TMP3]])
+; AVX512-NEXT: [[TMP5:%.*]] = shufflevector <8 x i64> [[TMP4]], <8 x i64> poison, <8 x i32> <i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX512-NEXT: [[TMP6:%.*]] = tail call <8 x i64> @llvm.umin.v8i64(<8 x i64> [[TMP4]], <8 x i64> [[TMP5]])
+; AVX512-NEXT: [[TMP7:%.*]] = extractelement <8 x i64> [[TMP6]], i64 0
+; AVX512-NEXT: ret i64 [[TMP7]]
+;
+ %1 = shufflevector <8 x i64> %a0, <8 x i64> undef, <8 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp ult <8 x i64> %a0, %1
+ %3 = select <8 x i1> %2, <8 x i64> %a0, <8 x i64> %1
+ %4 = shufflevector <8 x i64> %3, <8 x i64> undef, <8 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp ult <8 x i64> %3, %4
+ %6 = select <8 x i1> %5, <8 x i64> %3, <8 x i64> %4
+ %7 = shufflevector <8 x i64> %6, <8 x i64> undef, <8 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp ult <8 x i64> %6, %7
+ %9 = select <8 x i1> %8, <8 x i64> %6, <8 x i64> %7
+ %10 = extractelement <8 x i64> %9, i32 0
+ ret i64 %10
+}
+
+define i32 @test_reduce_v16i32(<16 x i32> %a0) {
+; SSE-LABEL: @test_reduce_v16i32(
+; SSE-NEXT: [[TMP1:%.*]] = tail call i32 @llvm.vector.reduce.umin.v16i32(<16 x i32> [[A0:%.*]])
+; SSE-NEXT: ret i32 [[TMP1]]
+;
+; AVX2-LABEL: @test_reduce_v16i32(
+; AVX2-NEXT: [[TMP1:%.*]] = tail call i32 @llvm.vector.reduce.umin.v16i32(<16 x i32> [[A0:%.*]])
+; AVX2-NEXT: ret i32 [[TMP1]]
+;
+; AVX512-LABEL: @test_reduce_v16i32(
+; AVX512-NEXT: [[TMP1:%.*]] = shufflevector <16 x i32> [[A0:%.*]], <16 x i32> poison, <16 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX512-NEXT: [[TMP2:%.*]] = tail call <16 x i32> @llvm.umin.v16i32(<16 x i32> [[A0]], <16 x i32> [[TMP1]])
+; AVX512-NEXT: [[TMP3:%.*]] = shufflevector <16 x i32> [[TMP2]], <16 x i32> poison, <16 x i32> <i32 4, i32 5, i32 6, i32 7, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX512-NEXT: [[TMP4:%.*]] = tail call <16 x i32> @llvm.umin.v16i32(<16 x i32> [[TMP2]], <16 x i32> [[TMP3]])
+; AVX512-NEXT: [[TMP5:%.*]] = shufflevector <16 x i32> [[TMP4]], <16 x i32> poison, <16 x i32> <i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX512-NEXT: [[TMP6:%.*]] = tail call <16 x i32> @llvm.umin.v16i32(<16 x i32> [[TMP4]], <16 x i32> [[TMP5]])
+; AVX512-NEXT: [[TMP7:%.*]] = shufflevector <16 x i32> [[TMP6]], <16 x i32> poison, <16 x i32> <i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; AVX512-NEXT: [[TMP8:%.*]] = tail call <16 x i32> @llvm.umin.v16i32(<16 x i32> [[TMP6]], <16 x i32> [[TMP7]])
+; AVX512-NEXT: [[TMP9:%.*]] = extractelement <16 x i32> [[TMP8]], i64 0
+; AVX512-NEXT: ret i32 [[TMP9]]
+;
+ %1 = shufflevector <16 x i32> %a0, <16 x i32> undef, <16 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp ult <16 x i32> %a0, %1
+ %3 = select <16 x i1> %2, <16 x i32> %a0, <16 x i32> %1
+ %4 = shufflevector <16 x i32> %3, <16 x i32> undef, <16 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp ult <16 x i32> %3, %4
+ %6 = select <16 x i1> %5, <16 x i32> %3, <16 x i32> %4
+ %7 = shufflevector <16 x i32> %6, <16 x i32> undef, <16 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp ult <16 x i32> %6, %7
+ %9 = select <16 x i1> %8, <16 x i32> %6, <16 x i32> %7
+ %10 = shufflevector <16 x i32> %9, <16 x i32> undef, <16 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %11 = icmp ult <16 x i32> %9, %10
+ %12 = select <16 x i1> %11, <16 x i32> %9, <16 x i32> %10
+ %13 = extractelement <16 x i32> %12, i32 0
+ ret i32 %13
+}
+
+define i16 @test_reduce_v32i16(<32 x i16> %a0) {
+; CHECK-LABEL: @test_reduce_v32i16(
+; CHECK-NEXT: [[TMP16:%.*]] = tail call i16 @llvm.vector.reduce.umin.v32i16(<32 x i16> [[A0:%.*]])
+; CHECK-NEXT: ret i16 [[TMP16]]
+;
+ %1 = shufflevector <32 x i16> %a0, <32 x i16> undef, <32 x i32> <i32 16, i32 17, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23, i32 24, i32 25, i32 26, i32 27, i32 28, i32 29, i32 30, i32 31, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp ult <32 x i16> %a0, %1
+ %3 = select <32 x i1> %2, <32 x i16> %a0, <32 x i16> %1
+ %4 = shufflevector <32 x i16> %3, <32 x i16> undef, <32 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp ult <32 x i16> %3, %4
+ %6 = select <32 x i1> %5, <32 x i16> %3, <32 x i16> %4
+ %7 = shufflevector <32 x i16> %6, <32 x i16> undef, <32 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp ult <32 x i16> %6, %7
+ %9 = select <32 x i1> %8, <32 x i16> %6, <32 x i16> %7
+ %10 = shufflevector <32 x i16> %9, <32 x i16> undef, <32 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %11 = icmp ult <32 x i16> %9, %10
+ %12 = select <32 x i1> %11, <32 x i16> %9, <32 x i16> %10
+ %13 = shufflevector <32 x i16> %12, <32 x i16> undef, <32 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %14 = icmp ult <32 x i16> %12, %13
+ %15 = select <32 x i1> %14, <32 x i16> %12, <32 x i16> %13
+ %16 = extractelement <32 x i16> %15, i32 0
+ ret i16 %16
+}
+
+define i8 @test_reduce_v64i8(<64 x i8> %a0) {
+; CHECK-LABEL: @test_reduce_v64i8(
+; CHECK-NEXT: [[TMP19:%.*]] = tail call i8 @llvm.vector.reduce.umin.v64i8(<64 x i8> [[A0:%.*]])
+; CHECK-NEXT: ret i8 [[TMP19]]
+;
+ %1 = shufflevector <64 x i8> %a0, <64 x i8> undef, <64 x i32> <i32 32, i32 33, i32 34, i32 35, i32 36, i32 37, i32 38, i32 39, i32 40, i32 41, i32 42, i32 43, i32 44, i32 45, i32 46, i32 47, i32 48, i32 49, i32 50, i32 51, i32 52, i32 53, i32 54, i32 55, i32 56, i32 57, i32 58, i32 59, i32 60, i32 61, i32 62, i32 63, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp ult <64 x i8> %a0, %1
+ %3 = select <64 x i1> %2, <64 x i8> %a0, <64 x i8> %1
+ %4 = shufflevector <64 x i8> %3, <64 x i8> undef, <64 x i32> <i32 16, i32 17, i32 18, i32 19, i32 20, i32 21, i32 22, i32 23, i32 24, i32 25, i32 26, i32 27, i32 28, i32 29, i32 30, i32 31, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp ult <64 x i8> %3, %4
+ %6 = select <64 x i1> %5, <64 x i8> %3, <64 x i8> %4
+ %7 = shufflevector <64 x i8> %6, <64 x i8> undef, <64 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp ult <64 x i8> %6, %7
+ %9 = select <64 x i1> %8, <64 x i8> %6, <64 x i8> %7
+ %10 = shufflevector <64 x i8> %9, <64 x i8> undef, <64 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %11 = icmp ult <64 x i8> %9, %10
+ %12 = select <64 x i1> %11, <64 x i8> %9, <64 x i8> %10
+ %13 = shufflevector <64 x i8> %12, <64 x i8> undef, <64 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %14 = icmp ult <64 x i8> %12, %13
+ %15 = select <64 x i1> %14, <64 x i8> %12, <64 x i8> %13
+ %16 = shufflevector <64 x i8> %15, <64 x i8> undef, <64 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %17 = icmp ult <64 x i8> %15, %16
+ %18 = select <64 x i1> %17, <64 x i8> %15, <64 x i8> %16
+ %19 = extractelement <64 x i8> %18, i32 0
+ ret i8 %19
+}
+
+;
+; Partial Vector Reductions
+;
+
+define i16 @test_reduce_v16i16_v8i16(<16 x i16> %a0) {
+; CHECK-LABEL: @test_reduce_v16i16_v8i16(
+; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <16 x i16> [[A0:%.*]], <16 x i16> poison, <16 x i32> <i32 4, i32 5, i32 6, i32 7, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP2:%.*]] = tail call <16 x i16> @llvm.umin.v16i16(<16 x i16> [[A0]], <16 x i16> [[TMP1]])
+; CHECK-NEXT: [[TMP3:%.*]] = shufflevector <16 x i16> [[TMP2]], <16 x i16> poison, <16 x i32> <i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP4:%.*]] = tail call <16 x i16> @llvm.umin.v16i16(<16 x i16> [[TMP2]], <16 x i16> [[TMP3]])
+; CHECK-NEXT: [[TMP5:%.*]] = shufflevector <16 x i16> [[TMP4]], <16 x i16> poison, <16 x i32> <i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP6:%.*]] = tail call <16 x i16> @llvm.umin.v16i16(<16 x i16> [[TMP4]], <16 x i16> [[TMP5]])
+; CHECK-NEXT: [[TMP10:%.*]] = extractelement <16 x i16> [[TMP6]], i64 0
+; CHECK-NEXT: ret i16 [[TMP10]]
+;
+ %1 = shufflevector <16 x i16> %a0, <16 x i16> undef, <16 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp ult <16 x i16> %a0, %1
+ %3 = select <16 x i1> %2, <16 x i16> %a0, <16 x i16> %1
+ %4 = shufflevector <16 x i16> %3, <16 x i16> undef, <16 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp ult <16 x i16> %3, %4
+ %6 = select <16 x i1> %5, <16 x i16> %3, <16 x i16> %4
+ %7 = shufflevector <16 x i16> %6, <16 x i16> undef, <16 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp ult <16 x i16> %6, %7
+ %9 = select <16 x i1> %8, <16 x i16> %6, <16 x i16> %7
+ %10 = extractelement <16 x i16> %9, i32 0
+ ret i16 %10
+}
+
+define i16 @test_reduce_v32i16_v8i16(<32 x i16> %a0) {
+; CHECK-LABEL: @test_reduce_v32i16_v8i16(
+; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <32 x i16> [[A0:%.*]], <32 x i16> poison, <32 x i32> <i32 4, i32 5, i32 6, i32 7, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP2:%.*]] = tail call <32 x i16> @llvm.umin.v32i16(<32 x i16> [[A0]], <32 x i16> [[TMP1]])
+; CHECK-NEXT: [[TMP3:%.*]] = shufflevector <32 x i16> [[TMP2]], <32 x i16> poison, <32 x i32> <i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP4:%.*]] = tail call <32 x i16> @llvm.umin.v32i16(<32 x i16> [[TMP2]], <32 x i16> [[TMP3]])
+; CHECK-NEXT: [[TMP5:%.*]] = shufflevector <32 x i16> [[TMP4]], <32 x i16> poison, <32 x i32> <i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP6:%.*]] = tail call <32 x i16> @llvm.umin.v32i16(<32 x i16> [[TMP4]], <32 x i16> [[TMP5]])
+; CHECK-NEXT: [[TMP10:%.*]] = extractelement <32 x i16> [[TMP6]], i64 0
+; CHECK-NEXT: ret i16 [[TMP10]]
+;
+ %1 = shufflevector <32 x i16> %a0, <32 x i16> undef, <32 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp ult <32 x i16> %a0, %1
+ %3 = select <32 x i1> %2, <32 x i16> %a0, <32 x i16> %1
+ %4 = shufflevector <32 x i16> %3, <32 x i16> undef, <32 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp ult <32 x i16> %3, %4
+ %6 = select <32 x i1> %5, <32 x i16> %3, <32 x i16> %4
+ %7 = shufflevector <32 x i16> %6, <32 x i16> undef, <32 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp ult <32 x i16> %6, %7
+ %9 = select <32 x i1> %8, <32 x i16> %6, <32 x i16> %7
+ %10 = extractelement <32 x i16> %9, i32 0
+ ret i16 %10
+}
+
+define i8 @test_reduce_v32i8_v16i8(<32 x i8> %a0) {
+; CHECK-LABEL: @test_reduce_v32i8_v16i8(
+; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <32 x i8> [[A0:%.*]], <32 x i8> poison, <32 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP2:%.*]] = tail call <32 x i8> @llvm.umin.v32i8(<32 x i8> [[A0]], <32 x i8> [[TMP1]])
+; CHECK-NEXT: [[TMP3:%.*]] = shufflevector <32 x i8> [[TMP2]], <32 x i8> poison, <32 x i32> <i32 4, i32 5, i32 6, i32 7, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP4:%.*]] = tail call <32 x i8> @llvm.umin.v32i8(<32 x i8> [[TMP2]], <32 x i8> [[TMP3]])
+; CHECK-NEXT: [[TMP5:%.*]] = shufflevector <32 x i8> [[TMP4]], <32 x i8> poison, <32 x i32> <i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP6:%.*]] = tail call <32 x i8> @llvm.umin.v32i8(<32 x i8> [[TMP4]], <32 x i8> [[TMP5]])
+; CHECK-NEXT: [[TMP7:%.*]] = shufflevector <32 x i8> [[TMP6]], <32 x i8> poison, <32 x i32> <i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP8:%.*]] = tail call <32 x i8> @llvm.umin.v32i8(<32 x i8> [[TMP6]], <32 x i8> [[TMP7]])
+; CHECK-NEXT: [[TMP13:%.*]] = extractelement <32 x i8> [[TMP8]], i64 0
+; CHECK-NEXT: ret i8 [[TMP13]]
+;
+ %1 = shufflevector <32 x i8> %a0, <32 x i8> undef, <32 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp ult <32 x i8> %a0, %1
+ %3 = select <32 x i1> %2, <32 x i8> %a0, <32 x i8> %1
+ %4 = shufflevector <32 x i8> %3, <32 x i8> undef, <32 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp ult <32 x i8> %3, %4
+ %6 = select <32 x i1> %5, <32 x i8> %3, <32 x i8> %4
+ %7 = shufflevector <32 x i8> %6, <32 x i8> undef, <32 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp ult <32 x i8> %6, %7
+ %9 = select <32 x i1> %8, <32 x i8> %6, <32 x i8> %7
+ %10 = shufflevector <32 x i8> %9, <32 x i8> undef, <32 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %11 = icmp ult <32 x i8> %9, %10
+ %12 = select <32 x i1> %11, <32 x i8> %9, <32 x i8> %10
+ %13 = extractelement <32 x i8> %12, i32 0
+ ret i8 %13
+}
+
+define i8 @test_reduce_v64i8_v16i8(<64 x i8> %a0) {
+; CHECK-LABEL: @test_reduce_v64i8_v16i8(
+; CHECK-NEXT: [[TMP1:%.*]] = shufflevector <64 x i8> [[A0:%.*]], <64 x i8> poison, <64 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP2:%.*]] = tail call <64 x i8> @llvm.umin.v64i8(<64 x i8> [[A0]], <64 x i8> [[TMP1]])
+; CHECK-NEXT: [[TMP3:%.*]] = shufflevector <64 x i8> [[TMP2]], <64 x i8> poison, <64 x i32> <i32 4, i32 5, i32 6, i32 7, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP4:%.*]] = tail call <64 x i8> @llvm.umin.v64i8(<64 x i8> [[TMP2]], <64 x i8> [[TMP3]])
+; CHECK-NEXT: [[TMP5:%.*]] = shufflevector <64 x i8> [[TMP4]], <64 x i8> poison, <64 x i32> <i32 2, i32 3, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP6:%.*]] = tail call <64 x i8> @llvm.umin.v64i8(<64 x i8> [[TMP4]], <64 x i8> [[TMP5]])
+; CHECK-NEXT: [[TMP7:%.*]] = shufflevector <64 x i8> [[TMP6]], <64 x i8> poison, <64 x i32> <i32 1, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison, i32 poison>
+; CHECK-NEXT: [[TMP8:%.*]] = tail call <64 x i8> @llvm.umin.v64i8(<64 x i8> [[TMP6]], <64 x i8> [[TMP7]])
+; CHECK-NEXT: [[TMP13:%.*]] = extractelement <64 x i8> [[TMP8]], i64 0
+; CHECK-NEXT: ret i8 [[TMP13]]
+;
+ %1 = shufflevector <64 x i8> %a0, <64 x i8> undef, <64 x i32> <i32 8, i32 9, i32 10, i32 11, i32 12, i32 13, i32 14, i32 15, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %2 = icmp ult <64 x i8> %a0, %1
+ %3 = select <64 x i1> %2, <64 x i8> %a0, <64 x i8> %1
+ %4 = shufflevector <64 x i8> %3, <64 x i8> undef, <64 x i32> <i32 4, i32 5, i32 6, i32 7, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %5 = icmp ult <64 x i8> %3, %4
+ %6 = select <64 x i1> %5, <64 x i8> %3, <64 x i8> %4
+ %7 = shufflevector <64 x i8> %6, <64 x i8> undef, <64 x i32> <i32 2, i32 3, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %8 = icmp ult <64 x i8> %6, %7
+ %9 = select <64 x i1> %8, <64 x i8> %6, <64 x i8> %7
+ %10 = shufflevector <64 x i8> %9, <64 x i8> undef, <64 x i32> <i32 1, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
+ %11 = icmp ult <64 x i8> %9, %10
+ %12 = select <64 x i1> %11, <64 x i8> %9, <64 x i8> %10
+ %13 = extractelement <64 x i8> %12, i32 0
+ ret i8 %13
+}
More information about the llvm-commits
mailing list