llvmbot wrote:
<!--LLVM PR SUMMARY COMMENT--> @llvm/pr-subscribers-clang Author: Sam Tebbs (SamTebbs33) <details> <summary>Changes</summary> Adds the following SME2 builtins: svmin_single_(s8|s16|s32|s64)_x(2|4) svmin_single_(u8|u16|u32|u64)_x(2|4) svmin_single_(f16|f32|f64)_x(2|4) svmin_(s8|s16|s32|s64)_x(2|4) svmin_(u8|u16|u32|u64)_x(2|4) svmin_(f16|f32|f64)_x(2|4) svmax_single_(s8|s16|s32|s64)_x(2|4) svmax_single_(u8|u16|u32|u64)_x(2|4) svmax_single_(f16|f32|f64)_x(2|4) svmax_(s8|s16|s32|s64)_x(2|4) svmax_(u8|u16|u32|u64)_x(2|4) svmax_(f16|f32|f64)_x(2|4) See [https://github.com/ARM-software/acle/pull/217](https://github.com/ARM-software/acle/pull/217) --- Patch is 355.99 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/71688.diff 3 Files Affected: - (modified) clang/include/clang/Basic/arm_sme.td (+23) - (added) clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_max.c (+1562) - (added) clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_min.c (+1562) ``````````diff diff --git a/clang/include/clang/Basic/arm_sme.td b/clang/include/clang/Basic/arm_sme.td index b5655afdf419ecf..8c1cd8f26514b48 100644 --- a/clang/include/clang/Basic/arm_sme.td +++ b/clang/include/clang/Basic/arm_sme.td @@ -296,5 +296,28 @@ multiclass ZAAddSub<string n_suffix> { } } +//////////////////////////////////////////////////////////////////////////////// +// SME2 - MIN, MAX + +multiclass MinMaxIntr<string i, string zm, string mul, string t> { + def SVS # NAME : SInst<"sv" # i # "[" # zm # "_{d}_" # mul # "]", t, "csil", MergeNone, "aarch64_sve_s" # i # zm # "_" # mul, [IsStreaming], []>; + def SVU # NAME : SInst<"sv" # i # "[" # zm # "_{d}_" # mul # "]", t, "UcUsUiUl", MergeNone, "aarch64_sve_u" # i # zm # "_" # mul, [IsStreaming], []>; + def SVF # NAME : SInst<"sv" # i # "[" # zm # "_{d}_" # mul # "]", t, "hfd", MergeNone, "aarch64_sve_f" # i # zm # "_" # mul, [IsStreaming], []>; +} + +let TargetGuard = "sme2" in { + // SMAX / UMAX / FMAX + defm MAX_SINGLE_X2 : MinMaxIntr<"max", "_single", "x2", "22d">; + defm MAX_MULTI_X2 : MinMaxIntr<"max", "", "x2", "222">; + defm MAX_SINGLE_X4 : MinMaxIntr<"max", "_single", "x4", "44d">; + defm MAX_MULTI_X4 : MinMaxIntr<"max", "", "x4", "444">; + + // SMIN / UMIN / FMIN + defm MIN_SINGLE_X2 : MinMaxIntr<"min", "_single", "x2", "22d">; + defm MIN_MULTI_X2 : MinMaxIntr<"min", "", "x2", "222">; + defm MIN_SINGLE_X4 : MinMaxIntr<"min", "_single", "x4", "44d">; + defm MIN_MULTI_X4 : MinMaxIntr<"min", "", "x4", "444">; +} + defm SVADD : ZAAddSub<"add">; defm SVSUB : ZAAddSub<"sub">; diff --git a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_max.c b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_max.c new file mode 100644 index 000000000000000..40c40b6edd078fd --- /dev/null +++ b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_max.c @@ -0,0 +1,1562 @@ +// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py +// RUN: %clang_cc1 -D__ARM_FEATURE_SME -D__ARM_FEATURE_SME2 -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -p mem2reg,instcombine,tailcallelim | FileCheck %s +// RUN: %clang_cc1 -D__ARM_FEATURE_SME -D__ARM_FEATURE_SME2 -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -p mem2reg,instcombine,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK +// RUN: %clang_cc1 -D__ARM_FEATURE_SME -D__ARM_FEATURE_SME2 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -p mem2reg,instcombine,tailcallelim | FileCheck %s +// RUN: %clang_cc1 -D__ARM_FEATURE_SME -D__ARM_FEATURE_SME2 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -p mem2reg,instcombine,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK +// RUN: %clang_cc1 -D__ARM_FEATURE_SME -D__ARM_FEATURE_SME2 -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s +#include <arm_sme_draft_spec_subject_to_change.h> + +#ifdef SVE_OVERLOADED_FORMS +// A simple used,unused... macro, long enough to represent any SVE builtin. +#define SVE_ACLE_FUNC(A1,A2_UNUSED) A1 +#else +#define SVE_ACLE_FUNC(A1,A2) A1##A2 +#endif + +// Single, x2 + +// CHECK-LABEL: @test_svmax_single_s8_x2( +// CHECK-NEXT: entry: +// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.vector.extract.nxv16i8.nxv32i8(<vscale x 32 x i8> [[ZDN:%.*]], i64 0) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 16 x i8> @llvm.vector.extract.nxv16i8.nxv32i8(<vscale x 32 x i8> [[ZDN]], i64 16) +// CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.smax.single.x2.nxv16i8(<vscale x 16 x i8> [[TMP0]], <vscale x 16 x i8> [[TMP1]], <vscale x 16 x i8> [[ZM:%.*]]) +// CHECK-NEXT: [[TMP3:%.*]] = extractvalue { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP2]], 0 +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 32 x i8> @llvm.vector.insert.nxv32i8.nxv16i8(<vscale x 32 x i8> poison, <vscale x 16 x i8> [[TMP3]], i64 0) +// CHECK-NEXT: [[TMP5:%.*]] = extractvalue { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP2]], 1 +// CHECK-NEXT: [[TMP6:%.*]] = tail call <vscale x 32 x i8> @llvm.vector.insert.nxv32i8.nxv16i8(<vscale x 32 x i8> [[TMP4]], <vscale x 16 x i8> [[TMP5]], i64 16) +// CHECK-NEXT: ret <vscale x 32 x i8> [[TMP6]] +// +// CPP-CHECK-LABEL: @_Z23test_svmax_single_s8_x210svint8x2_tu10__SVInt8_t( +// CPP-CHECK-NEXT: entry: +// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.vector.extract.nxv16i8.nxv32i8(<vscale x 32 x i8> [[ZDN:%.*]], i64 0) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 16 x i8> @llvm.vector.extract.nxv16i8.nxv32i8(<vscale x 32 x i8> [[ZDN]], i64 16) +// CPP-CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.smax.single.x2.nxv16i8(<vscale x 16 x i8> [[TMP0]], <vscale x 16 x i8> [[TMP1]], <vscale x 16 x i8> [[ZM:%.*]]) +// CPP-CHECK-NEXT: [[TMP3:%.*]] = extractvalue { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP2]], 0 +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 32 x i8> @llvm.vector.insert.nxv32i8.nxv16i8(<vscale x 32 x i8> poison, <vscale x 16 x i8> [[TMP3]], i64 0) +// CPP-CHECK-NEXT: [[TMP5:%.*]] = extractvalue { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP2]], 1 +// CPP-CHECK-NEXT: [[TMP6:%.*]] = tail call <vscale x 32 x i8> @llvm.vector.insert.nxv32i8.nxv16i8(<vscale x 32 x i8> [[TMP4]], <vscale x 16 x i8> [[TMP5]], i64 16) +// CPP-CHECK-NEXT: ret <vscale x 32 x i8> [[TMP6]] +// +svint8x2_t test_svmax_single_s8_x2(svint8x2_t zdn, svint8_t zm) { + return SVE_ACLE_FUNC(svmax,_single_s8_x2)(zdn, zm); +} + +// CHECK-LABEL: @test_svmax_single_s16_x2( +// CHECK-NEXT: entry: +// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i16> @llvm.vector.extract.nxv8i16.nxv16i16(<vscale x 16 x i16> [[ZDN:%.*]], i64 0) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.vector.extract.nxv8i16.nxv16i16(<vscale x 16 x i16> [[ZDN]], i64 8) +// CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 8 x i16>, <vscale x 8 x i16> } @llvm.aarch64.sve.smax.single.x2.nxv8i16(<vscale x 8 x i16> [[TMP0]], <vscale x 8 x i16> [[TMP1]], <vscale x 8 x i16> [[ZM:%.*]]) +// CHECK-NEXT: [[TMP3:%.*]] = extractvalue { <vscale x 8 x i16>, <vscale x 8 x i16> } [[TMP2]], 0 +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 16 x i16> @llvm.vector.insert.nxv16i16.nxv8i16(<vscale x 16 x i16> poison, <vscale x 8 x i16> [[TMP3]], i64 0) +// CHECK-NEXT: [[TMP5:%.*]] = extractvalue { <vscale x 8 x i16>, <vscale x 8 x i16> } [[TMP2]], 1 +// CHECK-NEXT: [[TMP6:%.*]] = tail call <vscale x 16 x i16> @llvm.vector.insert.nxv16i16.nxv8i16(<vscale x 16 x i16> [[TMP4]], <vscale x 8 x i16> [[TMP5]], i64 8) +// CHECK-NEXT: ret <vscale x 16 x i16> [[TMP6]] +// +// CPP-CHECK-LABEL: @_Z24test_svmax_single_s16_x211svint16x2_tu11__SVInt16_t( +// CPP-CHECK-NEXT: entry: +// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i16> @llvm.vector.extract.nxv8i16.nxv16i16(<vscale x 16 x i16> [[ZDN:%.*]], i64 0) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.vector.extract.nxv8i16.nxv16i16(<vscale x 16 x i16> [[ZDN]], i64 8) +// CPP-CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 8 x i16>, <vscale x 8 x i16> } @llvm.aarch64.sve.smax.single.x2.nxv8i16(<vscale x 8 x i16> [[TMP0]], <vscale x 8 x i16> [[TMP1]], <vscale x 8 x i16> [[ZM:%.*]]) +// CPP-CHECK-NEXT: [[TMP3:%.*]] = extractvalue { <vscale x 8 x i16>, <vscale x 8 x i16> } [[TMP2]], 0 +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 16 x i16> @llvm.vector.insert.nxv16i16.nxv8i16(<vscale x 16 x i16> poison, <vscale x 8 x i16> [[TMP3]], i64 0) +// CPP-CHECK-NEXT: [[TMP5:%.*]] = extractvalue { <vscale x 8 x i16>, <vscale x 8 x i16> } [[TMP2]], 1 +// CPP-CHECK-NEXT: [[TMP6:%.*]] = tail call <vscale x 16 x i16> @llvm.vector.insert.nxv16i16.nxv8i16(<vscale x 16 x i16> [[TMP4]], <vscale x 8 x i16> [[TMP5]], i64 8) +// CPP-CHECK-NEXT: ret <vscale x 16 x i16> [[TMP6]] +// +svint16x2_t test_svmax_single_s16_x2(svint16x2_t zdn, svint16_t zm) { + return SVE_ACLE_FUNC(svmax,_single_s16_x2)(zdn, zm); +} + +// CHECK-LABEL: @test_svmax_single_s32_x2( +// CHECK-NEXT: entry: +// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i32> @llvm.vector.extract.nxv4i32.nxv8i32(<vscale x 8 x i32> [[ZDN:%.*]], i64 0) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i32> @llvm.vector.extract.nxv4i32.nxv8i32(<vscale x 8 x i32> [[ZDN]], i64 4) +// CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 4 x i32>, <vscale x 4 x i32> } @llvm.aarch64.sve.smax.single.x2.nxv4i32(<vscale x 4 x i32> [[TMP0]], <vscale x 4 x i32> [[TMP1]], <vscale x 4 x i32> [[ZM:%.*]]) +// CHECK-NEXT: [[TMP3:%.*]] = extractvalue { <vscale x 4 x i32>, <vscale x 4 x i32> } [[TMP2]], 0 +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x i32> @llvm.vector.insert.nxv8i32.nxv4i32(<vscale x 8 x i32> poison, <vscale x 4 x i32> [[TMP3]], i64 0) +// CHECK-NEXT: [[TMP5:%.*]] = extractvalue { <vscale x 4 x i32>, <vscale x 4 x i32> } [[TMP2]], 1 +// CHECK-NEXT: [[TMP6:%.*]] = tail call <vscale x 8 x i32> @llvm.vector.insert.nxv8i32.nxv4i32(<vscale x 8 x i32> [[TMP4]], <vscale x 4 x i32> [[TMP5]], i64 4) +// CHECK-NEXT: ret <vscale x 8 x i32> [[TMP6]] +// +// CPP-CHECK-LABEL: @_Z24test_svmax_single_s32_x211svint32x2_tu11__SVInt32_t( +// CPP-CHECK-NEXT: entry: +// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i32> @llvm.vector.extract.nxv4i32.nxv8i32(<vscale x 8 x i32> [[ZDN:%.*]], i64 0) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i32> @llvm.vector.extract.nxv4i32.nxv8i32(<vscale x 8 x i32> [[ZDN]], i64 4) +// CPP-CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 4 x i32>, <vscale x 4 x i32> } @llvm.aarch64.sve.smax.single.x2.nxv4i32(<vscale x 4 x i32> [[TMP0]], <vscale x 4 x i32> [[TMP1]], <vscale x 4 x i32> [[ZM:%.*]]) +// CPP-CHECK-NEXT: [[TMP3:%.*]] = extractvalue { <vscale x 4 x i32>, <vscale x 4 x i32> } [[TMP2]], 0 +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x i32> @llvm.vector.insert.nxv8i32.nxv4i32(<vscale x 8 x i32> poison, <vscale x 4 x i32> [[TMP3]], i64 0) +// CPP-CHECK-NEXT: [[TMP5:%.*]] = extractvalue { <vscale x 4 x i32>, <vscale x 4 x i32> } [[TMP2]], 1 +// CPP-CHECK-NEXT: [[TMP6:%.*]] = tail call <vscale x 8 x i32> @llvm.vector.insert.nxv8i32.nxv4i32(<vscale x 8 x i32> [[TMP4]], <vscale x 4 x i32> [[TMP5]], i64 4) +// CPP-CHECK-NEXT: ret <vscale x 8 x i32> [[TMP6]] +// +svint32x2_t test_svmax_single_s32_x2(svint32x2_t zdn, svint32_t zm) { + return SVE_ACLE_FUNC(svmax,_single_s32_x2)(zdn, zm); +} + +// CHECK-LABEL: @test_svmax_single_s64_x2( +// CHECK-NEXT: entry: +// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i64> @llvm.vector.extract.nxv2i64.nxv4i64(<vscale x 4 x i64> [[ZDN:%.*]], i64 0) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i64> @llvm.vector.extract.nxv2i64.nxv4i64(<vscale x 4 x i64> [[ZDN]], i64 2) +// CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 2 x i64>, <vscale x 2 x i64> } @llvm.aarch64.sve.smax.single.x2.nxv2i64(<vscale x 2 x i64> [[TMP0]], <vscale x 2 x i64> [[TMP1]], <vscale x 2 x i64> [[ZM:%.*]]) +// CHECK-NEXT: [[TMP3:%.*]] = extractvalue { <vscale x 2 x i64>, <vscale x 2 x i64> } [[TMP2]], 0 +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i64> @llvm.vector.insert.nxv4i64.nxv2i64(<vscale x 4 x i64> poison, <vscale x 2 x i64> [[TMP3]], i64 0) +// CHECK-NEXT: [[TMP5:%.*]] = extractvalue { <vscale x 2 x i64>, <vscale x 2 x i64> } [[TMP2]], 1 +// CHECK-NEXT: [[TMP6:%.*]] = tail call <vscale x 4 x i64> @llvm.vector.insert.nxv4i64.nxv2i64(<vscale x 4 x i64> [[TMP4]], <vscale x 2 x i64> [[TMP5]], i64 2) +// CHECK-NEXT: ret <vscale x 4 x i64> [[TMP6]] +// +// CPP-CHECK-LABEL: @_Z24test_svmax_single_s64_x211svint64x2_tu11__SVInt64_t( +// CPP-CHECK-NEXT: entry: +// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i64> @llvm.vector.extract.nxv2i64.nxv4i64(<vscale x 4 x i64> [[ZDN:%.*]], i64 0) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i64> @llvm.vector.extract.nxv2i64.nxv4i64(<vscale x 4 x i64> [[ZDN]], i64 2) +// CPP-CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 2 x i64>, <vscale x 2 x i64> } @llvm.aarch64.sve.smax.single.x2.nxv2i64(<vscale x 2 x i64> [[TMP0]], <vscale x 2 x i64> [[TMP1]], <vscale x 2 x i64> [[ZM:%.*]]) +// CPP-CHECK-NEXT: [[TMP3:%.*]] = extractvalue { <vscale x 2 x i64>, <vscale x 2 x i64> } [[TMP2]], 0 +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i64> @llvm.vector.insert.nxv4i64.nxv2i64(<vscale x 4 x i64> poison, <vscale x 2 x i64> [[TMP3]], i64 0) +// CPP-CHECK-NEXT: [[TMP5:%.*]] = extractvalue { <vscale x 2 x i64>, <vscale x 2 x i64> } [[TMP2]], 1 +// CPP-CHECK-NEXT: [[TMP6:%.*]] = tail call <vscale x 4 x i64> @llvm.vector.insert.nxv4i64.nxv2i64(<vscale x 4 x i64> [[TMP4]], <vscale x 2 x i64> [[TMP5]], i64 2) +// CPP-CHECK-NEXT: ret <vscale x 4 x i64> [[TMP6]] +// +svint64x2_t test_svmax_single_s64_x2(svint64x2_t zdn, svint64_t zm) { + return SVE_ACLE_FUNC(svmax,_single_s64_x2)(zdn, zm); +} + +// CHECK-LABEL: @test_svmax_single_u8_x2( +// CHECK-NEXT: entry: +// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.vector.extract.nxv16i8.nxv32i8(<vscale x 32 x i8> [[ZDN:%.*]], i64 0) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 16 x i8> @llvm.vector.extract.nxv16i8.nxv32i8(<vscale x 32 x i8> [[ZDN]], i64 16) +// CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.umax.single.x2.nxv16i8(<vscale x 16 x i8> [[TMP0]], <vscale x 16 x i8> [[TMP1]], <vscale x 16 x i8> [[ZM:%.*]]) +// CHECK-NEXT: [[TMP3:%.*]] = extractvalue { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP2]], 0 +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 32 x i8> @llvm.vector.insert.nxv32i8.nxv16i8(<vscale x 32 x i8> poison, <vscale x 16 x i8> [[TMP3]], i64 0) +// CHECK-NEXT: [[TMP5:%.*]] = extractvalue { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP2]], 1 +// CHECK-NEXT: [[TMP6:%.*]] = tail call <vscale x 32 x i8> @llvm.vector.insert.nxv32i8.nxv16i8(<vscale x 32 x i8> [[TMP4]], <vscale x 16 x i8> [[TMP5]], i64 16) +// CHECK-NEXT: ret <vscale x 32 x i8> [[TMP6]] +// +// CPP-CHECK-LABEL: @_Z23test_svmax_single_u8_x211svuint8x2_tu11__SVUint8_t( +// CPP-CHECK-NEXT: entry: +// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.vector.extract.nxv16i8.nxv32i8(<vscale x 32 x i8> [[ZDN:%.*]], i64 0) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 16 x i8> @llvm.vector.extract.nxv16i8.nxv32i8(<vscale x 32 x i8> [[ZDN]], i64 16) +// CPP-CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 x i8> } @llvm.aarch64.sve.umax.single.x2.nxv16i8(<vscale x 16 x i8> [[TMP0]], <vscale x 16 x i8> [[TMP1]], <vscale x 16 x i8> [[ZM:%.*]]) +// CPP-CHECK-NEXT: [[TMP3:%.*]] = extractvalue { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP2]], 0 +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 32 x i8> @llvm.vector.insert.nxv32i8.nxv16i8(<vscale x 32 x i8> poison, <vscale x 16 x i8> [[TMP3]], i64 0) +// CPP-CHECK-NEXT: [[TMP5:%.*]] = extractvalue { <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP2]], 1 +// CPP-CHECK-NEXT: [[TMP6:%.*]] = tail call <vscale x 32 x i8> @llvm.vector.insert.nxv32i8.nxv16i8(<vscale x 32 x i8> [[TMP4]], <vscale x 16 x i8> [[TMP5]], i64 16) +// CPP-CHECK-NEXT: ret <vscale x 32 x i8> [[TMP6]] +// +svuint8x2_t test_svmax_single_u8_x2(svuint8x2_t zdn, svuint8_t zm) { + return SVE_ACLE_FUNC(svmax,_single_u8_x2)(zdn, zm); +} + +// CHECK-LABEL: @test_svmax_single_u16_x2( +// CHECK-NEXT: entry: +// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i16> @llvm.vector.extract.nxv8i16.nxv16i16(<vscale x 16 x i16> [[ZDN:%.*]], i64 0) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.vector.extract.nxv8i16.nxv16i16(<vscale x 16 x i16> [[ZDN]], i64 8) +// CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 8 x i16>, <vscale x 8 x i16> } @llvm.aarch64.sve.umax.single.x2.nxv8i16(<vscale x 8 x i16> [[TMP0]], <vscale x 8 x i16> [[TMP1]], <vscale x 8 x i16> [[ZM:%.*]]) +// CHECK-NEXT: [[TMP3:%.*]] = extractvalue { <vscale x 8 x i16>, <vscale x 8 x i16> } [[TMP2]], 0 +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 16 x i16> @llvm.vector.insert.nxv16i16.nxv8i16(<vscale x 16 x i16> poison, <vscale x 8 x i16> [[TMP3]], i64 0) +// CHECK-NEXT: [[TMP5:%.*]] = extractvalue { <vscale x 8 x i16>, <vscale x 8 x i16> } [[TMP2]], 1 +// CHECK-NEXT: [[TMP6:%.*]] = tail call <vscale x 16 x i16> @llvm.vector.insert.nxv16i16.nxv8i16(<vscale x 16 x i16> [[TMP4]], <vscale x 8 x i16> [[TMP5]], i64 8) +// CHECK-NEXT: ret <vscale x 16 x i16> [[TMP6]] +// +// CPP-CHECK-LABEL: @_Z24test_svmax_single_u16_x212svuint16x2_tu12__SVUint16_t( +// CPP-CHECK-NEXT: entry: +// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i16> @llvm.vector.extract.nxv8i16.nxv16i16(<vscale x 16 x i16> [[ZDN:%.*]], i64 0) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.vector.extract.nxv8i16.nxv16i16(<vscale x 16 x i16> [[ZDN]], i64 8) +// CPP-CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 8 x i16>, <vscale x 8 x i16> } @llvm.aarch64.sve.umax.single.x2.nxv8i16(<vscale x 8 x i16> [[TMP0]], <vscale x 8 x i16> [[TMP1]], <vscale x 8 x i16> [[ZM:%.*]]) +// CPP-CHECK-NEXT: [[TMP3:%.*]] = extractvalue { <vscale x 8 x i16>, <vscale x 8 x i16> } [[TMP2]], 0 +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 16 x i16> @llvm.vector.insert.nxv16i16.nxv8i16(<vscale x 16 x i16> poison, <vscale x 8 x i16> [[TMP3]], i64 0) +// CPP-CHECK-NEXT: [[TMP5:%.*]] = extractvalue { <vscale x 8 x i16>, <vscale x 8 x i16> } [[TMP2]], 1 +// CPP-CHECK-NEXT: [[TMP6:%.*]] = tail call <vscale x 16 x i16> @llvm.vector.insert.nxv16i16.nxv8i16(<vscale x 16 x i16> [[TMP4]], <vscale x 8 x i16> [[TMP5]], i64 8) +// CPP-CHECK-NEXT: ret <vscale x 16 x i16> [[TMP6]] +// +svuint16x2_t test_svmax_single_u16_x2(svuint16x2_t zdn, svuint16_t zm) { + return SVE_ACLE_FUNC(svmax,_single_u16_x2)(zdn, zm); +} + +// CHECK-LABEL: @test_svmax_single_u32_x2( +// CHECK-NEXT: entry: +// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i32> @llvm.vector.extract.nxv4i32.nxv8i32(<vscale x 8 x i32> [[ZDN:%.*]], i64 0) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i32> @llvm.vector.extract.nxv4i32.nxv8i32(<vscale x 8 x i32> [[ZDN]], i64 4) +// CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 4 x i32>, <vscale x 4 x i32> } @llvm.aarch64.sve.umax.single.x2.nxv4i32(<vscale x 4 x i32> [[TMP0]], <vscale x 4 x i32> [[TMP1]], <vscale x 4 x i32> [[ZM:%.*]]) +// CHECK-NEXT: [[TMP3:%.*]] = extractvalue { <vscale x 4 x i32>, <vscale x 4 x i32> } [[TMP2]], 0 +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x i32> @llvm.vector.insert.nxv8i32.nxv4i32(<vscale x 8 x i32> poison, <vscale x 4 x i32> [[TMP3]], i64 0) +// CHECK-NEXT: [[TMP5:%.*]] = extractvalue { <vscale x 4 x i32>, <vscale x 4 x i32> } [[TMP2]], 1 +// CHECK-NEXT: [[TMP6:%.*]] = tail call <vscale x 8 x i32> @llvm.vector.insert.nxv8i32.nxv4i32(<vscale x 8 x i32> [[TMP4]], <vscale x 4 x i32> [[TMP5]], i64 4) +// CHECK-NEXT: ret <vscale x 8 x i32> [[TMP6]] +// +// CPP-CHECK-LABEL: @_Z24test_svmax_single_u32_x212svuint32x2_tu12__SVUint32_t( +// CPP-CHECK-NEXT: entry: +// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i32> @llvm.vector.extract.nxv4i32.nxv8i32(<vscale x 8 x i32> [[ZDN:%.*]], i64 0) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i32> @llvm.vector.extract.nxv4i32.nxv8i32(<vscale x 8 x i32> [[ZDN]], i64 4) +// CPP-CHECK-NEXT: [[TMP2:%.*]] = tail call { <vscale x 4 x i32>, <vscale x 4 x i32> } @llvm.aarch64.sve.umax.single.x2.nxv4i32(<vscale x 4 x i32> [[TMP0]], <vscale x 4 x i32> [[TMP1]], <vscale x 4 x i32> [[ZM:%.*]]) +// CPP-CHECK-NEXT: [[TMP3:%.*]] = extractvalue { <vscale x 4 x i32>, <vscale x 4 x i32> } [[TMP2]], 0 +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x i32> @llvm.ve... [truncated] `````````` </details> https://github.com/llvm/llvm-project/pull/71688 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits