diff options
Diffstat (limited to 'clang/test/CodeGen')
22 files changed, 388 insertions, 394 deletions
diff --git a/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_ld1.c b/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_ld1.c index 19e5243..1ed59c6c 100644 --- a/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_ld1.c +++ b/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_ld1.c @@ -24,12 +24,12 @@ // CHECK-LABEL: @test_svld1_s8( // CHECK-NEXT: entry: -// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr [[BASE:%.*]], i32 1, <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr align 1 [[BASE:%.*]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> zeroinitializer) // CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]] // // CPP-CHECK-LABEL: @_Z13test_svld1_s8u10__SVBool_tPKa( // CPP-CHECK-NEXT: entry: -// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr [[BASE:%.*]], i32 1, <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr align 1 [[BASE:%.*]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> zeroinitializer) // CPP-CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]] // svint8_t test_svld1_s8(svbool_t pg, const int8_t *base) MODE_ATTR @@ -40,13 +40,13 @@ svint8_t test_svld1_s8(svbool_t pg, const int8_t *base) MODE_ATTR // CHECK-LABEL: @test_svld1_s16( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0(ptr [[BASE:%.*]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0(ptr align 1 [[BASE:%.*]], <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> zeroinitializer) // CHECK-NEXT: ret <vscale x 8 x i16> [[TMP1]] // // CPP-CHECK-LABEL: @_Z14test_svld1_s16u10__SVBool_tPKs( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0(ptr [[BASE:%.*]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0(ptr align 1 [[BASE:%.*]], <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> zeroinitializer) // CPP-CHECK-NEXT: ret <vscale x 8 x i16> [[TMP1]] // svint16_t test_svld1_s16(svbool_t pg, const int16_t *base) MODE_ATTR @@ -57,13 +57,13 @@ svint16_t test_svld1_s16(svbool_t pg, const int16_t *base) MODE_ATTR // CHECK-LABEL: @test_svld1_s32( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0(ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0(ptr align 1 [[BASE:%.*]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> zeroinitializer) // CHECK-NEXT: ret <vscale x 4 x i32> [[TMP1]] // // CPP-CHECK-LABEL: @_Z14test_svld1_s32u10__SVBool_tPKi( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0(ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0(ptr align 1 [[BASE:%.*]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> zeroinitializer) // CPP-CHECK-NEXT: ret <vscale x 4 x i32> [[TMP1]] // svint32_t test_svld1_s32(svbool_t pg, const int32_t *base) MODE_ATTR @@ -74,13 +74,13 @@ svint32_t test_svld1_s32(svbool_t pg, const int32_t *base) MODE_ATTR // CHECK-LABEL: @test_svld1_s64( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i64> @llvm.masked.load.nxv2i64.p0(ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i64> @llvm.masked.load.nxv2i64.p0(ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> zeroinitializer) // CHECK-NEXT: ret <vscale x 2 x i64> [[TMP1]] // // CPP-CHECK-LABEL: @_Z14test_svld1_s64u10__SVBool_tPKl( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i64> @llvm.masked.load.nxv2i64.p0(ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i64> @llvm.masked.load.nxv2i64.p0(ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> zeroinitializer) // CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP1]] // svint64_t test_svld1_s64(svbool_t pg, const int64_t *base) MODE_ATTR @@ -90,12 +90,12 @@ svint64_t test_svld1_s64(svbool_t pg, const int64_t *base) MODE_ATTR // CHECK-LABEL: @test_svld1_u8( // CHECK-NEXT: entry: -// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr [[BASE:%.*]], i32 1, <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr align 1 [[BASE:%.*]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> zeroinitializer) // CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]] // // CPP-CHECK-LABEL: @_Z13test_svld1_u8u10__SVBool_tPKh( // CPP-CHECK-NEXT: entry: -// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr [[BASE:%.*]], i32 1, <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr align 1 [[BASE:%.*]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> zeroinitializer) // CPP-CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]] // svuint8_t test_svld1_u8(svbool_t pg, const uint8_t *base) MODE_ATTR @@ -106,13 +106,13 @@ svuint8_t test_svld1_u8(svbool_t pg, const uint8_t *base) MODE_ATTR // CHECK-LABEL: @test_svld1_u16( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0(ptr [[BASE:%.*]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0(ptr align 1 [[BASE:%.*]], <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> zeroinitializer) // CHECK-NEXT: ret <vscale x 8 x i16> [[TMP1]] // // CPP-CHECK-LABEL: @_Z14test_svld1_u16u10__SVBool_tPKt( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0(ptr [[BASE:%.*]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0(ptr align 1 [[BASE:%.*]], <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> zeroinitializer) // CPP-CHECK-NEXT: ret <vscale x 8 x i16> [[TMP1]] // svuint16_t test_svld1_u16(svbool_t pg, const uint16_t *base) MODE_ATTR @@ -123,13 +123,13 @@ svuint16_t test_svld1_u16(svbool_t pg, const uint16_t *base) MODE_ATTR // CHECK-LABEL: @test_svld1_u32( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0(ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0(ptr align 1 [[BASE:%.*]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> zeroinitializer) // CHECK-NEXT: ret <vscale x 4 x i32> [[TMP1]] // // CPP-CHECK-LABEL: @_Z14test_svld1_u32u10__SVBool_tPKj( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0(ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0(ptr align 1 [[BASE:%.*]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> zeroinitializer) // CPP-CHECK-NEXT: ret <vscale x 4 x i32> [[TMP1]] // svuint32_t test_svld1_u32(svbool_t pg, const uint32_t *base) MODE_ATTR @@ -140,13 +140,13 @@ svuint32_t test_svld1_u32(svbool_t pg, const uint32_t *base) MODE_ATTR // CHECK-LABEL: @test_svld1_u64( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i64> @llvm.masked.load.nxv2i64.p0(ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i64> @llvm.masked.load.nxv2i64.p0(ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> zeroinitializer) // CHECK-NEXT: ret <vscale x 2 x i64> [[TMP1]] // // CPP-CHECK-LABEL: @_Z14test_svld1_u64u10__SVBool_tPKm( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i64> @llvm.masked.load.nxv2i64.p0(ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i64> @llvm.masked.load.nxv2i64.p0(ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> zeroinitializer) // CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP1]] // svuint64_t test_svld1_u64(svbool_t pg, const uint64_t *base) MODE_ATTR @@ -157,13 +157,13 @@ svuint64_t test_svld1_u64(svbool_t pg, const uint64_t *base) MODE_ATTR // CHECK-LABEL: @test_svld1_f16( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.masked.load.nxv8f16.p0(ptr [[BASE:%.*]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.masked.load.nxv8f16.p0(ptr align 1 [[BASE:%.*]], <vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> zeroinitializer) // CHECK-NEXT: ret <vscale x 8 x half> [[TMP1]] // // CPP-CHECK-LABEL: @_Z14test_svld1_f16u10__SVBool_tPKDh( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.masked.load.nxv8f16.p0(ptr [[BASE:%.*]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x half> @llvm.masked.load.nxv8f16.p0(ptr align 1 [[BASE:%.*]], <vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> zeroinitializer) // CPP-CHECK-NEXT: ret <vscale x 8 x half> [[TMP1]] // svfloat16_t test_svld1_f16(svbool_t pg, const float16_t *base) MODE_ATTR @@ -174,13 +174,13 @@ svfloat16_t test_svld1_f16(svbool_t pg, const float16_t *base) MODE_ATTR // CHECK-LABEL: @test_svld1_f32( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr align 1 [[BASE:%.*]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> zeroinitializer) // CHECK-NEXT: ret <vscale x 4 x float> [[TMP1]] // // CPP-CHECK-LABEL: @_Z14test_svld1_f32u10__SVBool_tPKf( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr align 1 [[BASE:%.*]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> zeroinitializer) // CPP-CHECK-NEXT: ret <vscale x 4 x float> [[TMP1]] // svfloat32_t test_svld1_f32(svbool_t pg, const float32_t *base) MODE_ATTR @@ -191,13 +191,13 @@ svfloat32_t test_svld1_f32(svbool_t pg, const float32_t *base) MODE_ATTR // CHECK-LABEL: @test_svld1_f64( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.masked.load.nxv2f64.p0(ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.masked.load.nxv2f64.p0(ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> zeroinitializer) // CHECK-NEXT: ret <vscale x 2 x double> [[TMP1]] // // CPP-CHECK-LABEL: @_Z14test_svld1_f64u10__SVBool_tPKd( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.masked.load.nxv2f64.p0(ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x double> @llvm.masked.load.nxv2f64.p0(ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> zeroinitializer) // CPP-CHECK-NEXT: ret <vscale x 2 x double> [[TMP1]] // svfloat64_t test_svld1_f64(svbool_t pg, const float64_t *base) MODE_ATTR @@ -207,12 +207,12 @@ svfloat64_t test_svld1_f64(svbool_t pg, const float64_t *base) MODE_ATTR // CHECK-LABEL: @test_svld1_mf8( // CHECK-NEXT: entry: -// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr [[BASE:%.*]], i32 1, <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr align 1 [[BASE:%.*]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> zeroinitializer) // CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]] // // CPP-CHECK-LABEL: @_Z14test_svld1_mf8u10__SVBool_tPKu6__mfp8( // CPP-CHECK-NEXT: entry: -// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr [[BASE:%.*]], i32 1, <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr align 1 [[BASE:%.*]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> zeroinitializer) // CPP-CHECK-NEXT: ret <vscale x 16 x i8> [[TMP0]] // svmfloat8_t test_svld1_mf8(svbool_t pg, const mfloat8_t *base) MODE_ATTR @@ -226,7 +226,7 @@ svmfloat8_t test_svld1_mf8(svbool_t pg, const mfloat8_t *base) MODE_ATTR // CHECK-NEXT: [[TMP1:%.*]] = shl nuw nsw i64 [[TMP0]], 4 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP1]] // CHECK-NEXT: [[TMP2:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP3:%.*]] = tail call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr [[TMP2]], i32 1, <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP3:%.*]] = tail call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr align 1 [[TMP2]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> zeroinitializer) // CHECK-NEXT: ret <vscale x 16 x i8> [[TMP3]] // // CPP-CHECK-LABEL: @_Z18test_svld1_vnum_s8u10__SVBool_tPKal( @@ -235,7 +235,7 @@ svmfloat8_t test_svld1_mf8(svbool_t pg, const mfloat8_t *base) MODE_ATTR // CPP-CHECK-NEXT: [[TMP1:%.*]] = shl nuw nsw i64 [[TMP0]], 4 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP1]] // CPP-CHECK-NEXT: [[TMP2:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP3:%.*]] = tail call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr [[TMP2]], i32 1, <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP3:%.*]] = tail call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr align 1 [[TMP2]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> zeroinitializer) // CPP-CHECK-NEXT: ret <vscale x 16 x i8> [[TMP3]] // svint8_t test_svld1_vnum_s8(svbool_t pg, const int8_t *base, int64_t vnum) MODE_ATTR @@ -250,7 +250,7 @@ svint8_t test_svld1_vnum_s8(svbool_t pg, const int8_t *base, int64_t vnum) MODE_ // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0(ptr [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0(ptr align 1 [[TMP3]], <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> zeroinitializer) // CHECK-NEXT: ret <vscale x 8 x i16> [[TMP4]] // // CPP-CHECK-LABEL: @_Z19test_svld1_vnum_s16u10__SVBool_tPKsl( @@ -260,7 +260,7 @@ svint8_t test_svld1_vnum_s8(svbool_t pg, const int8_t *base, int64_t vnum) MODE_ // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0(ptr [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0(ptr align 1 [[TMP3]], <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> zeroinitializer) // CPP-CHECK-NEXT: ret <vscale x 8 x i16> [[TMP4]] // svint16_t test_svld1_vnum_s16(svbool_t pg, const int16_t *base, int64_t vnum) MODE_ATTR @@ -275,7 +275,7 @@ svint16_t test_svld1_vnum_s16(svbool_t pg, const int16_t *base, int64_t vnum) MO // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0(ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0(ptr align 1 [[TMP3]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> zeroinitializer) // CHECK-NEXT: ret <vscale x 4 x i32> [[TMP4]] // // CPP-CHECK-LABEL: @_Z19test_svld1_vnum_s32u10__SVBool_tPKil( @@ -285,7 +285,7 @@ svint16_t test_svld1_vnum_s16(svbool_t pg, const int16_t *base, int64_t vnum) MO // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0(ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0(ptr align 1 [[TMP3]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> zeroinitializer) // CPP-CHECK-NEXT: ret <vscale x 4 x i32> [[TMP4]] // svint32_t test_svld1_vnum_s32(svbool_t pg, const int32_t *base, int64_t vnum) MODE_ATTR @@ -300,7 +300,7 @@ svint32_t test_svld1_vnum_s32(svbool_t pg, const int32_t *base, int64_t vnum) MO // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i64> @llvm.masked.load.nxv2i64.p0(ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i64> @llvm.masked.load.nxv2i64.p0(ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> zeroinitializer) // CHECK-NEXT: ret <vscale x 2 x i64> [[TMP4]] // // CPP-CHECK-LABEL: @_Z19test_svld1_vnum_s64u10__SVBool_tPKll( @@ -310,7 +310,7 @@ svint32_t test_svld1_vnum_s32(svbool_t pg, const int32_t *base, int64_t vnum) MO // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i64> @llvm.masked.load.nxv2i64.p0(ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i64> @llvm.masked.load.nxv2i64.p0(ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> zeroinitializer) // CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP4]] // svint64_t test_svld1_vnum_s64(svbool_t pg, const int64_t *base, int64_t vnum) MODE_ATTR @@ -324,7 +324,7 @@ svint64_t test_svld1_vnum_s64(svbool_t pg, const int64_t *base, int64_t vnum) MO // CHECK-NEXT: [[TMP1:%.*]] = shl nuw nsw i64 [[TMP0]], 4 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP1]] // CHECK-NEXT: [[TMP2:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP3:%.*]] = tail call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr [[TMP2]], i32 1, <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP3:%.*]] = tail call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr align 1 [[TMP2]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> zeroinitializer) // CHECK-NEXT: ret <vscale x 16 x i8> [[TMP3]] // // CPP-CHECK-LABEL: @_Z18test_svld1_vnum_u8u10__SVBool_tPKhl( @@ -333,7 +333,7 @@ svint64_t test_svld1_vnum_s64(svbool_t pg, const int64_t *base, int64_t vnum) MO // CPP-CHECK-NEXT: [[TMP1:%.*]] = shl nuw nsw i64 [[TMP0]], 4 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP1]] // CPP-CHECK-NEXT: [[TMP2:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP3:%.*]] = tail call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr [[TMP2]], i32 1, <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP3:%.*]] = tail call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr align 1 [[TMP2]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> zeroinitializer) // CPP-CHECK-NEXT: ret <vscale x 16 x i8> [[TMP3]] // svuint8_t test_svld1_vnum_u8(svbool_t pg, const uint8_t *base, int64_t vnum) MODE_ATTR @@ -348,7 +348,7 @@ svuint8_t test_svld1_vnum_u8(svbool_t pg, const uint8_t *base, int64_t vnum) MOD // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0(ptr [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0(ptr align 1 [[TMP3]], <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> zeroinitializer) // CHECK-NEXT: ret <vscale x 8 x i16> [[TMP4]] // // CPP-CHECK-LABEL: @_Z19test_svld1_vnum_u16u10__SVBool_tPKtl( @@ -358,7 +358,7 @@ svuint8_t test_svld1_vnum_u8(svbool_t pg, const uint8_t *base, int64_t vnum) MOD // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0(ptr [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x i16> @llvm.masked.load.nxv8i16.p0(ptr align 1 [[TMP3]], <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i16> zeroinitializer) // CPP-CHECK-NEXT: ret <vscale x 8 x i16> [[TMP4]] // svuint16_t test_svld1_vnum_u16(svbool_t pg, const uint16_t *base, int64_t vnum) MODE_ATTR @@ -373,7 +373,7 @@ svuint16_t test_svld1_vnum_u16(svbool_t pg, const uint16_t *base, int64_t vnum) // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0(ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0(ptr align 1 [[TMP3]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> zeroinitializer) // CHECK-NEXT: ret <vscale x 4 x i32> [[TMP4]] // // CPP-CHECK-LABEL: @_Z19test_svld1_vnum_u32u10__SVBool_tPKjl( @@ -383,7 +383,7 @@ svuint16_t test_svld1_vnum_u16(svbool_t pg, const uint16_t *base, int64_t vnum) // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0(ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i32> @llvm.masked.load.nxv4i32.p0(ptr align 1 [[TMP3]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i32> zeroinitializer) // CPP-CHECK-NEXT: ret <vscale x 4 x i32> [[TMP4]] // svuint32_t test_svld1_vnum_u32(svbool_t pg, const uint32_t *base, int64_t vnum) MODE_ATTR @@ -398,7 +398,7 @@ svuint32_t test_svld1_vnum_u32(svbool_t pg, const uint32_t *base, int64_t vnum) // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i64> @llvm.masked.load.nxv2i64.p0(ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i64> @llvm.masked.load.nxv2i64.p0(ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> zeroinitializer) // CHECK-NEXT: ret <vscale x 2 x i64> [[TMP4]] // // CPP-CHECK-LABEL: @_Z19test_svld1_vnum_u64u10__SVBool_tPKml( @@ -408,7 +408,7 @@ svuint32_t test_svld1_vnum_u32(svbool_t pg, const uint32_t *base, int64_t vnum) // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i64> @llvm.masked.load.nxv2i64.p0(ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i64> @llvm.masked.load.nxv2i64.p0(ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i64> zeroinitializer) // CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP4]] // svuint64_t test_svld1_vnum_u64(svbool_t pg, const uint64_t *base, int64_t vnum) MODE_ATTR @@ -423,7 +423,7 @@ svuint64_t test_svld1_vnum_u64(svbool_t pg, const uint64_t *base, int64_t vnum) // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x half> @llvm.masked.load.nxv8f16.p0(ptr [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x half> @llvm.masked.load.nxv8f16.p0(ptr align 1 [[TMP3]], <vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> zeroinitializer) // CHECK-NEXT: ret <vscale x 8 x half> [[TMP4]] // // CPP-CHECK-LABEL: @_Z19test_svld1_vnum_f16u10__SVBool_tPKDhl( @@ -433,7 +433,7 @@ svuint64_t test_svld1_vnum_u64(svbool_t pg, const uint64_t *base, int64_t vnum) // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x half> @llvm.masked.load.nxv8f16.p0(ptr [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x half> @llvm.masked.load.nxv8f16.p0(ptr align 1 [[TMP3]], <vscale x 8 x i1> [[TMP0]], <vscale x 8 x half> zeroinitializer) // CPP-CHECK-NEXT: ret <vscale x 8 x half> [[TMP4]] // svfloat16_t test_svld1_vnum_f16(svbool_t pg, const float16_t *base, int64_t vnum) MODE_ATTR @@ -448,7 +448,7 @@ svfloat16_t test_svld1_vnum_f16(svbool_t pg, const float16_t *base, int64_t vnum // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr align 1 [[TMP3]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> zeroinitializer) // CHECK-NEXT: ret <vscale x 4 x float> [[TMP4]] // // CPP-CHECK-LABEL: @_Z19test_svld1_vnum_f32u10__SVBool_tPKfl( @@ -458,7 +458,7 @@ svfloat16_t test_svld1_vnum_f16(svbool_t pg, const float16_t *base, int64_t vnum // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x float> @llvm.masked.load.nxv4f32.p0(ptr align 1 [[TMP3]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x float> zeroinitializer) // CPP-CHECK-NEXT: ret <vscale x 4 x float> [[TMP4]] // svfloat32_t test_svld1_vnum_f32(svbool_t pg, const float32_t *base, int64_t vnum) MODE_ATTR @@ -473,7 +473,7 @@ svfloat32_t test_svld1_vnum_f32(svbool_t pg, const float32_t *base, int64_t vnum // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x double> @llvm.masked.load.nxv2f64.p0(ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x double> @llvm.masked.load.nxv2f64.p0(ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> zeroinitializer) // CHECK-NEXT: ret <vscale x 2 x double> [[TMP4]] // // CPP-CHECK-LABEL: @_Z19test_svld1_vnum_f64u10__SVBool_tPKdl( @@ -483,7 +483,7 @@ svfloat32_t test_svld1_vnum_f32(svbool_t pg, const float32_t *base, int64_t vnum // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x double> @llvm.masked.load.nxv2f64.p0(ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x double> @llvm.masked.load.nxv2f64.p0(ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x double> zeroinitializer) // CPP-CHECK-NEXT: ret <vscale x 2 x double> [[TMP4]] // svfloat64_t test_svld1_vnum_f64(svbool_t pg, const float64_t *base, int64_t vnum) MODE_ATTR @@ -497,7 +497,7 @@ svfloat64_t test_svld1_vnum_f64(svbool_t pg, const float64_t *base, int64_t vnum // CHECK-NEXT: [[TMP1:%.*]] = shl nuw nsw i64 [[TMP0]], 4 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP1]] // CHECK-NEXT: [[TMP2:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP3:%.*]] = tail call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr [[TMP2]], i32 1, <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP3:%.*]] = tail call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr align 1 [[TMP2]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> zeroinitializer) // CHECK-NEXT: ret <vscale x 16 x i8> [[TMP3]] // // CPP-CHECK-LABEL: @_Z19test_svld1_vnum_mf8u10__SVBool_tPKu6__mfp8l( @@ -506,7 +506,7 @@ svfloat64_t test_svld1_vnum_f64(svbool_t pg, const float64_t *base, int64_t vnum // CPP-CHECK-NEXT: [[TMP1:%.*]] = shl nuw nsw i64 [[TMP0]], 4 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP1]] // CPP-CHECK-NEXT: [[TMP2:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP3:%.*]] = tail call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr [[TMP2]], i32 1, <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP3:%.*]] = tail call <vscale x 16 x i8> @llvm.masked.load.nxv16i8.p0(ptr align 1 [[TMP2]], <vscale x 16 x i1> [[PG:%.*]], <vscale x 16 x i8> zeroinitializer) // CPP-CHECK-NEXT: ret <vscale x 16 x i8> [[TMP3]] // svmfloat8_t test_svld1_vnum_mf8(svbool_t pg, const mfloat8_t *base, int64_t vnum) MODE_ATTR @@ -1205,13 +1205,13 @@ svfloat64_t test_svld1_gather_u64base_index_f64(svbool_t pg, svuint64_t bases, i // CHECK-LABEL: @test_svld1_bf16( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x bfloat> @llvm.masked.load.nxv8bf16.p0(ptr [[BASE:%.*]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x bfloat> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x bfloat> @llvm.masked.load.nxv8bf16.p0(ptr align 1 [[BASE:%.*]], <vscale x 8 x i1> [[TMP0]], <vscale x 8 x bfloat> zeroinitializer) // CHECK-NEXT: ret <vscale x 8 x bfloat> [[TMP1]] // // CPP-CHECK-LABEL: @_Z15test_svld1_bf16u10__SVBool_tPKu6__bf16( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x bfloat> @llvm.masked.load.nxv8bf16.p0(ptr [[BASE:%.*]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x bfloat> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x bfloat> @llvm.masked.load.nxv8bf16.p0(ptr align 1 [[BASE:%.*]], <vscale x 8 x i1> [[TMP0]], <vscale x 8 x bfloat> zeroinitializer) // CPP-CHECK-NEXT: ret <vscale x 8 x bfloat> [[TMP1]] // svbfloat16_t test_svld1_bf16(svbool_t pg, const bfloat16_t *base) MODE_ATTR @@ -1226,7 +1226,7 @@ svbfloat16_t test_svld1_bf16(svbool_t pg, const bfloat16_t *base) MODE_ATTR // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x bfloat> @llvm.masked.load.nxv8bf16.p0(ptr [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x bfloat> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x bfloat> @llvm.masked.load.nxv8bf16.p0(ptr align 1 [[TMP3]], <vscale x 8 x i1> [[TMP0]], <vscale x 8 x bfloat> zeroinitializer) // CHECK-NEXT: ret <vscale x 8 x bfloat> [[TMP4]] // // CPP-CHECK-LABEL: @_Z20test_svld1_vnum_bf16u10__SVBool_tPKu6__bf16l( @@ -1236,7 +1236,7 @@ svbfloat16_t test_svld1_bf16(svbool_t pg, const bfloat16_t *base) MODE_ATTR // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x bfloat> @llvm.masked.load.nxv8bf16.p0(ptr [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x bfloat> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x bfloat> @llvm.masked.load.nxv8bf16.p0(ptr align 1 [[TMP3]], <vscale x 8 x i1> [[TMP0]], <vscale x 8 x bfloat> zeroinitializer) // CPP-CHECK-NEXT: ret <vscale x 8 x bfloat> [[TMP4]] // svbfloat16_t test_svld1_vnum_bf16(svbool_t pg, const bfloat16_t *base, int64_t vnum) MODE_ATTR diff --git a/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_ld1sb.c b/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_ld1sb.c index 2757f28..eb40da2 100644 --- a/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_ld1sb.c +++ b/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_ld1sb.c @@ -25,14 +25,14 @@ // CHECK-LABEL: @test_svld1sb_s16( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr [[BASE:%.*]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr align 1 [[BASE:%.*]], <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i8> zeroinitializer) // CHECK-NEXT: [[TMP2:%.*]] = sext <vscale x 8 x i8> [[TMP1]] to <vscale x 8 x i16> // CHECK-NEXT: ret <vscale x 8 x i16> [[TMP2]] // // CPP-CHECK-LABEL: @_Z16test_svld1sb_s16u10__SVBool_tPKa( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr [[BASE:%.*]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i8> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr align 1 [[BASE:%.*]], <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i8> zeroinitializer) // CPP-CHECK-NEXT: [[TMP2:%.*]] = sext <vscale x 8 x i8> [[TMP1]] to <vscale x 8 x i16> // CPP-CHECK-NEXT: ret <vscale x 8 x i16> [[TMP2]] // @@ -44,14 +44,14 @@ svint16_t test_svld1sb_s16(svbool_t pg, const int8_t *base) MODE_ATTR // CHECK-LABEL: @test_svld1sb_s32( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr align 1 [[BASE:%.*]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i8> zeroinitializer) // CHECK-NEXT: [[TMP2:%.*]] = sext <vscale x 4 x i8> [[TMP1]] to <vscale x 4 x i32> // CHECK-NEXT: ret <vscale x 4 x i32> [[TMP2]] // // CPP-CHECK-LABEL: @_Z16test_svld1sb_s32u10__SVBool_tPKa( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i8> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr align 1 [[BASE:%.*]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i8> zeroinitializer) // CPP-CHECK-NEXT: [[TMP2:%.*]] = sext <vscale x 4 x i8> [[TMP1]] to <vscale x 4 x i32> // CPP-CHECK-NEXT: ret <vscale x 4 x i32> [[TMP2]] // @@ -63,14 +63,14 @@ svint32_t test_svld1sb_s32(svbool_t pg, const int8_t *base) MODE_ATTR // CHECK-LABEL: @test_svld1sb_s64( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i8> zeroinitializer) // CHECK-NEXT: [[TMP2:%.*]] = sext <vscale x 2 x i8> [[TMP1]] to <vscale x 2 x i64> // CHECK-NEXT: ret <vscale x 2 x i64> [[TMP2]] // // CPP-CHECK-LABEL: @_Z16test_svld1sb_s64u10__SVBool_tPKa( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i8> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i8> zeroinitializer) // CPP-CHECK-NEXT: [[TMP2:%.*]] = sext <vscale x 2 x i8> [[TMP1]] to <vscale x 2 x i64> // CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP2]] // @@ -82,14 +82,14 @@ svint64_t test_svld1sb_s64(svbool_t pg, const int8_t *base) MODE_ATTR // CHECK-LABEL: @test_svld1sb_u16( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr [[BASE:%.*]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr align 1 [[BASE:%.*]], <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i8> zeroinitializer) // CHECK-NEXT: [[TMP2:%.*]] = sext <vscale x 8 x i8> [[TMP1]] to <vscale x 8 x i16> // CHECK-NEXT: ret <vscale x 8 x i16> [[TMP2]] // // CPP-CHECK-LABEL: @_Z16test_svld1sb_u16u10__SVBool_tPKa( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr [[BASE:%.*]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i8> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr align 1 [[BASE:%.*]], <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i8> zeroinitializer) // CPP-CHECK-NEXT: [[TMP2:%.*]] = sext <vscale x 8 x i8> [[TMP1]] to <vscale x 8 x i16> // CPP-CHECK-NEXT: ret <vscale x 8 x i16> [[TMP2]] // @@ -101,14 +101,14 @@ svuint16_t test_svld1sb_u16(svbool_t pg, const int8_t *base) MODE_ATTR // CHECK-LABEL: @test_svld1sb_u32( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr align 1 [[BASE:%.*]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i8> zeroinitializer) // CHECK-NEXT: [[TMP2:%.*]] = sext <vscale x 4 x i8> [[TMP1]] to <vscale x 4 x i32> // CHECK-NEXT: ret <vscale x 4 x i32> [[TMP2]] // // CPP-CHECK-LABEL: @_Z16test_svld1sb_u32u10__SVBool_tPKa( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i8> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr align 1 [[BASE:%.*]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i8> zeroinitializer) // CPP-CHECK-NEXT: [[TMP2:%.*]] = sext <vscale x 4 x i8> [[TMP1]] to <vscale x 4 x i32> // CPP-CHECK-NEXT: ret <vscale x 4 x i32> [[TMP2]] // @@ -120,14 +120,14 @@ svuint32_t test_svld1sb_u32(svbool_t pg, const int8_t *base) MODE_ATTR // CHECK-LABEL: @test_svld1sb_u64( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i8> zeroinitializer) // CHECK-NEXT: [[TMP2:%.*]] = sext <vscale x 2 x i8> [[TMP1]] to <vscale x 2 x i64> // CHECK-NEXT: ret <vscale x 2 x i64> [[TMP2]] // // CPP-CHECK-LABEL: @_Z16test_svld1sb_u64u10__SVBool_tPKa( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i8> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i8> zeroinitializer) // CPP-CHECK-NEXT: [[TMP2:%.*]] = sext <vscale x 2 x i8> [[TMP1]] to <vscale x 2 x i64> // CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP2]] // @@ -143,7 +143,7 @@ svuint64_t test_svld1sb_u64(svbool_t pg, const int8_t *base) MODE_ATTR // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 3 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr align 1 [[TMP3]], <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i8> zeroinitializer) // CHECK-NEXT: [[TMP5:%.*]] = sext <vscale x 8 x i8> [[TMP4]] to <vscale x 8 x i16> // CHECK-NEXT: ret <vscale x 8 x i16> [[TMP5]] // @@ -154,7 +154,7 @@ svuint64_t test_svld1sb_u64(svbool_t pg, const int8_t *base) MODE_ATTR // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 3 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i8> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr align 1 [[TMP3]], <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i8> zeroinitializer) // CPP-CHECK-NEXT: [[TMP5:%.*]] = sext <vscale x 8 x i8> [[TMP4]] to <vscale x 8 x i16> // CPP-CHECK-NEXT: ret <vscale x 8 x i16> [[TMP5]] // @@ -170,7 +170,7 @@ svint16_t test_svld1sb_vnum_s16(svbool_t pg, const int8_t *base, int64_t vnum) M // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 2 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr align 1 [[TMP3]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i8> zeroinitializer) // CHECK-NEXT: [[TMP5:%.*]] = sext <vscale x 4 x i8> [[TMP4]] to <vscale x 4 x i32> // CHECK-NEXT: ret <vscale x 4 x i32> [[TMP5]] // @@ -181,7 +181,7 @@ svint16_t test_svld1sb_vnum_s16(svbool_t pg, const int8_t *base, int64_t vnum) M // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 2 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i8> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr align 1 [[TMP3]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i8> zeroinitializer) // CPP-CHECK-NEXT: [[TMP5:%.*]] = sext <vscale x 4 x i8> [[TMP4]] to <vscale x 4 x i32> // CPP-CHECK-NEXT: ret <vscale x 4 x i32> [[TMP5]] // @@ -197,7 +197,7 @@ svint32_t test_svld1sb_vnum_s32(svbool_t pg, const int8_t *base, int64_t vnum) M // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 1 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i8> zeroinitializer) // CHECK-NEXT: [[TMP5:%.*]] = sext <vscale x 2 x i8> [[TMP4]] to <vscale x 2 x i64> // CHECK-NEXT: ret <vscale x 2 x i64> [[TMP5]] // @@ -208,7 +208,7 @@ svint32_t test_svld1sb_vnum_s32(svbool_t pg, const int8_t *base, int64_t vnum) M // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 1 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i8> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i8> zeroinitializer) // CPP-CHECK-NEXT: [[TMP5:%.*]] = sext <vscale x 2 x i8> [[TMP4]] to <vscale x 2 x i64> // CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP5]] // @@ -224,7 +224,7 @@ svint64_t test_svld1sb_vnum_s64(svbool_t pg, const int8_t *base, int64_t vnum) M // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 3 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr align 1 [[TMP3]], <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i8> zeroinitializer) // CHECK-NEXT: [[TMP5:%.*]] = sext <vscale x 8 x i8> [[TMP4]] to <vscale x 8 x i16> // CHECK-NEXT: ret <vscale x 8 x i16> [[TMP5]] // @@ -235,7 +235,7 @@ svint64_t test_svld1sb_vnum_s64(svbool_t pg, const int8_t *base, int64_t vnum) M // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 3 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i8> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr align 1 [[TMP3]], <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i8> zeroinitializer) // CPP-CHECK-NEXT: [[TMP5:%.*]] = sext <vscale x 8 x i8> [[TMP4]] to <vscale x 8 x i16> // CPP-CHECK-NEXT: ret <vscale x 8 x i16> [[TMP5]] // @@ -251,7 +251,7 @@ svuint16_t test_svld1sb_vnum_u16(svbool_t pg, const int8_t *base, int64_t vnum) // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 2 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr align 1 [[TMP3]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i8> zeroinitializer) // CHECK-NEXT: [[TMP5:%.*]] = sext <vscale x 4 x i8> [[TMP4]] to <vscale x 4 x i32> // CHECK-NEXT: ret <vscale x 4 x i32> [[TMP5]] // @@ -262,7 +262,7 @@ svuint16_t test_svld1sb_vnum_u16(svbool_t pg, const int8_t *base, int64_t vnum) // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 2 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i8> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr align 1 [[TMP3]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i8> zeroinitializer) // CPP-CHECK-NEXT: [[TMP5:%.*]] = sext <vscale x 4 x i8> [[TMP4]] to <vscale x 4 x i32> // CPP-CHECK-NEXT: ret <vscale x 4 x i32> [[TMP5]] // @@ -278,7 +278,7 @@ svuint32_t test_svld1sb_vnum_u32(svbool_t pg, const int8_t *base, int64_t vnum) // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 1 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i8> zeroinitializer) // CHECK-NEXT: [[TMP5:%.*]] = sext <vscale x 2 x i8> [[TMP4]] to <vscale x 2 x i64> // CHECK-NEXT: ret <vscale x 2 x i64> [[TMP5]] // @@ -289,7 +289,7 @@ svuint32_t test_svld1sb_vnum_u32(svbool_t pg, const int8_t *base, int64_t vnum) // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 1 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i8> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i8> zeroinitializer) // CPP-CHECK-NEXT: [[TMP5:%.*]] = sext <vscale x 2 x i8> [[TMP4]] to <vscale x 2 x i64> // CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP5]] // diff --git a/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_ld1sh.c b/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_ld1sh.c index dbc762f..e1cbb53 100644 --- a/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_ld1sh.c +++ b/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_ld1sh.c @@ -25,14 +25,14 @@ // CHECK-LABEL: @test_svld1sh_s32( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i16> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr align 1 [[BASE:%.*]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i16> zeroinitializer) // CHECK-NEXT: [[TMP2:%.*]] = sext <vscale x 4 x i16> [[TMP1]] to <vscale x 4 x i32> // CHECK-NEXT: ret <vscale x 4 x i32> [[TMP2]] // // CPP-CHECK-LABEL: @_Z16test_svld1sh_s32u10__SVBool_tPKs( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i16> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr align 1 [[BASE:%.*]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i16> zeroinitializer) // CPP-CHECK-NEXT: [[TMP2:%.*]] = sext <vscale x 4 x i16> [[TMP1]] to <vscale x 4 x i32> // CPP-CHECK-NEXT: ret <vscale x 4 x i32> [[TMP2]] // @@ -44,14 +44,14 @@ svint32_t test_svld1sh_s32(svbool_t pg, const int16_t *base) MODE_ATTR // CHECK-LABEL: @test_svld1sh_s64( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i16> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i16> zeroinitializer) // CHECK-NEXT: [[TMP2:%.*]] = sext <vscale x 2 x i16> [[TMP1]] to <vscale x 2 x i64> // CHECK-NEXT: ret <vscale x 2 x i64> [[TMP2]] // // CPP-CHECK-LABEL: @_Z16test_svld1sh_s64u10__SVBool_tPKs( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i16> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i16> zeroinitializer) // CPP-CHECK-NEXT: [[TMP2:%.*]] = sext <vscale x 2 x i16> [[TMP1]] to <vscale x 2 x i64> // CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP2]] // @@ -63,14 +63,14 @@ svint64_t test_svld1sh_s64(svbool_t pg, const int16_t *base) MODE_ATTR // CHECK-LABEL: @test_svld1sh_u32( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i16> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr align 1 [[BASE:%.*]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i16> zeroinitializer) // CHECK-NEXT: [[TMP2:%.*]] = sext <vscale x 4 x i16> [[TMP1]] to <vscale x 4 x i32> // CHECK-NEXT: ret <vscale x 4 x i32> [[TMP2]] // // CPP-CHECK-LABEL: @_Z16test_svld1sh_u32u10__SVBool_tPKs( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i16> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr align 1 [[BASE:%.*]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i16> zeroinitializer) // CPP-CHECK-NEXT: [[TMP2:%.*]] = sext <vscale x 4 x i16> [[TMP1]] to <vscale x 4 x i32> // CPP-CHECK-NEXT: ret <vscale x 4 x i32> [[TMP2]] // @@ -82,14 +82,14 @@ svuint32_t test_svld1sh_u32(svbool_t pg, const int16_t *base) MODE_ATTR // CHECK-LABEL: @test_svld1sh_u64( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i16> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i16> zeroinitializer) // CHECK-NEXT: [[TMP2:%.*]] = sext <vscale x 2 x i16> [[TMP1]] to <vscale x 2 x i64> // CHECK-NEXT: ret <vscale x 2 x i64> [[TMP2]] // // CPP-CHECK-LABEL: @_Z16test_svld1sh_u64u10__SVBool_tPKs( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i16> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i16> zeroinitializer) // CPP-CHECK-NEXT: [[TMP2:%.*]] = sext <vscale x 2 x i16> [[TMP1]] to <vscale x 2 x i64> // CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP2]] // @@ -105,7 +105,7 @@ svuint64_t test_svld1sh_u64(svbool_t pg, const int16_t *base) MODE_ATTR // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 3 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i16> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr align 1 [[TMP3]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i16> zeroinitializer) // CHECK-NEXT: [[TMP5:%.*]] = sext <vscale x 4 x i16> [[TMP4]] to <vscale x 4 x i32> // CHECK-NEXT: ret <vscale x 4 x i32> [[TMP5]] // @@ -116,7 +116,7 @@ svuint64_t test_svld1sh_u64(svbool_t pg, const int16_t *base) MODE_ATTR // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 3 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i16> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr align 1 [[TMP3]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i16> zeroinitializer) // CPP-CHECK-NEXT: [[TMP5:%.*]] = sext <vscale x 4 x i16> [[TMP4]] to <vscale x 4 x i32> // CPP-CHECK-NEXT: ret <vscale x 4 x i32> [[TMP5]] // @@ -132,7 +132,7 @@ svint32_t test_svld1sh_vnum_s32(svbool_t pg, const int16_t *base, int64_t vnum) // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 2 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i16> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i16> zeroinitializer) // CHECK-NEXT: [[TMP5:%.*]] = sext <vscale x 2 x i16> [[TMP4]] to <vscale x 2 x i64> // CHECK-NEXT: ret <vscale x 2 x i64> [[TMP5]] // @@ -143,7 +143,7 @@ svint32_t test_svld1sh_vnum_s32(svbool_t pg, const int16_t *base, int64_t vnum) // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 2 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i16> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i16> zeroinitializer) // CPP-CHECK-NEXT: [[TMP5:%.*]] = sext <vscale x 2 x i16> [[TMP4]] to <vscale x 2 x i64> // CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP5]] // @@ -159,7 +159,7 @@ svint64_t test_svld1sh_vnum_s64(svbool_t pg, const int16_t *base, int64_t vnum) // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 3 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i16> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr align 1 [[TMP3]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i16> zeroinitializer) // CHECK-NEXT: [[TMP5:%.*]] = sext <vscale x 4 x i16> [[TMP4]] to <vscale x 4 x i32> // CHECK-NEXT: ret <vscale x 4 x i32> [[TMP5]] // @@ -170,7 +170,7 @@ svint64_t test_svld1sh_vnum_s64(svbool_t pg, const int16_t *base, int64_t vnum) // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 3 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i16> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr align 1 [[TMP3]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i16> zeroinitializer) // CPP-CHECK-NEXT: [[TMP5:%.*]] = sext <vscale x 4 x i16> [[TMP4]] to <vscale x 4 x i32> // CPP-CHECK-NEXT: ret <vscale x 4 x i32> [[TMP5]] // @@ -186,7 +186,7 @@ svuint32_t test_svld1sh_vnum_u32(svbool_t pg, const int16_t *base, int64_t vnum) // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 2 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i16> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i16> zeroinitializer) // CHECK-NEXT: [[TMP5:%.*]] = sext <vscale x 2 x i16> [[TMP4]] to <vscale x 2 x i64> // CHECK-NEXT: ret <vscale x 2 x i64> [[TMP5]] // @@ -197,7 +197,7 @@ svuint32_t test_svld1sh_vnum_u32(svbool_t pg, const int16_t *base, int64_t vnum) // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 2 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i16> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i16> zeroinitializer) // CPP-CHECK-NEXT: [[TMP5:%.*]] = sext <vscale x 2 x i16> [[TMP4]] to <vscale x 2 x i64> // CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP5]] // diff --git a/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_ld1sw.c b/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_ld1sw.c index 575d214..14ee095 100644 --- a/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_ld1sw.c +++ b/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_ld1sw.c @@ -25,14 +25,14 @@ // CHECK-LABEL: @test_svld1sw_s64( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i32> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i32> zeroinitializer) // CHECK-NEXT: [[TMP2:%.*]] = sext <vscale x 2 x i32> [[TMP1]] to <vscale x 2 x i64> // CHECK-NEXT: ret <vscale x 2 x i64> [[TMP2]] // // CPP-CHECK-LABEL: @_Z16test_svld1sw_s64u10__SVBool_tPKi( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i32> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i32> zeroinitializer) // CPP-CHECK-NEXT: [[TMP2:%.*]] = sext <vscale x 2 x i32> [[TMP1]] to <vscale x 2 x i64> // CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP2]] // @@ -44,14 +44,14 @@ svint64_t test_svld1sw_s64(svbool_t pg, const int32_t *base) MODE_ATTR // CHECK-LABEL: @test_svld1sw_u64( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i32> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i32> zeroinitializer) // CHECK-NEXT: [[TMP2:%.*]] = sext <vscale x 2 x i32> [[TMP1]] to <vscale x 2 x i64> // CHECK-NEXT: ret <vscale x 2 x i64> [[TMP2]] // // CPP-CHECK-LABEL: @_Z16test_svld1sw_u64u10__SVBool_tPKi( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i32> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i32> zeroinitializer) // CPP-CHECK-NEXT: [[TMP2:%.*]] = sext <vscale x 2 x i32> [[TMP1]] to <vscale x 2 x i64> // CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP2]] // @@ -67,7 +67,7 @@ svuint64_t test_svld1sw_u64(svbool_t pg, const int32_t *base) MODE_ATTR // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 3 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i32> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i32> zeroinitializer) // CHECK-NEXT: [[TMP5:%.*]] = sext <vscale x 2 x i32> [[TMP4]] to <vscale x 2 x i64> // CHECK-NEXT: ret <vscale x 2 x i64> [[TMP5]] // @@ -78,7 +78,7 @@ svuint64_t test_svld1sw_u64(svbool_t pg, const int32_t *base) MODE_ATTR // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 3 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i32> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i32> zeroinitializer) // CPP-CHECK-NEXT: [[TMP5:%.*]] = sext <vscale x 2 x i32> [[TMP4]] to <vscale x 2 x i64> // CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP5]] // @@ -94,7 +94,7 @@ svint64_t test_svld1sw_vnum_s64(svbool_t pg, const int32_t *base, int64_t vnum) // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 3 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i32> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i32> zeroinitializer) // CHECK-NEXT: [[TMP5:%.*]] = sext <vscale x 2 x i32> [[TMP4]] to <vscale x 2 x i64> // CHECK-NEXT: ret <vscale x 2 x i64> [[TMP5]] // @@ -105,7 +105,7 @@ svint64_t test_svld1sw_vnum_s64(svbool_t pg, const int32_t *base, int64_t vnum) // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 3 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i32> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i32> zeroinitializer) // CPP-CHECK-NEXT: [[TMP5:%.*]] = sext <vscale x 2 x i32> [[TMP4]] to <vscale x 2 x i64> // CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP5]] // diff --git a/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_ld1ub.c b/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_ld1ub.c index 07e8815..3e0d2827 100644 --- a/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_ld1ub.c +++ b/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_ld1ub.c @@ -25,14 +25,14 @@ // CHECK-LABEL: @test_svld1ub_s16( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr [[BASE:%.*]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr align 1 [[BASE:%.*]], <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i8> zeroinitializer) // CHECK-NEXT: [[TMP2:%.*]] = zext <vscale x 8 x i8> [[TMP1]] to <vscale x 8 x i16> // CHECK-NEXT: ret <vscale x 8 x i16> [[TMP2]] // // CPP-CHECK-LABEL: @_Z16test_svld1ub_s16u10__SVBool_tPKh( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr [[BASE:%.*]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i8> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr align 1 [[BASE:%.*]], <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i8> zeroinitializer) // CPP-CHECK-NEXT: [[TMP2:%.*]] = zext <vscale x 8 x i8> [[TMP1]] to <vscale x 8 x i16> // CPP-CHECK-NEXT: ret <vscale x 8 x i16> [[TMP2]] // @@ -44,14 +44,14 @@ svint16_t test_svld1ub_s16(svbool_t pg, const uint8_t *base) MODE_ATTR // CHECK-LABEL: @test_svld1ub_s32( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr align 1 [[BASE:%.*]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i8> zeroinitializer) // CHECK-NEXT: [[TMP2:%.*]] = zext <vscale x 4 x i8> [[TMP1]] to <vscale x 4 x i32> // CHECK-NEXT: ret <vscale x 4 x i32> [[TMP2]] // // CPP-CHECK-LABEL: @_Z16test_svld1ub_s32u10__SVBool_tPKh( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i8> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr align 1 [[BASE:%.*]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i8> zeroinitializer) // CPP-CHECK-NEXT: [[TMP2:%.*]] = zext <vscale x 4 x i8> [[TMP1]] to <vscale x 4 x i32> // CPP-CHECK-NEXT: ret <vscale x 4 x i32> [[TMP2]] // @@ -63,14 +63,14 @@ svint32_t test_svld1ub_s32(svbool_t pg, const uint8_t *base) MODE_ATTR // CHECK-LABEL: @test_svld1ub_s64( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i8> zeroinitializer) // CHECK-NEXT: [[TMP2:%.*]] = zext <vscale x 2 x i8> [[TMP1]] to <vscale x 2 x i64> // CHECK-NEXT: ret <vscale x 2 x i64> [[TMP2]] // // CPP-CHECK-LABEL: @_Z16test_svld1ub_s64u10__SVBool_tPKh( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i8> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i8> zeroinitializer) // CPP-CHECK-NEXT: [[TMP2:%.*]] = zext <vscale x 2 x i8> [[TMP1]] to <vscale x 2 x i64> // CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP2]] // @@ -82,14 +82,14 @@ svint64_t test_svld1ub_s64(svbool_t pg, const uint8_t *base) MODE_ATTR // CHECK-LABEL: @test_svld1ub_u16( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr [[BASE:%.*]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr align 1 [[BASE:%.*]], <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i8> zeroinitializer) // CHECK-NEXT: [[TMP2:%.*]] = zext <vscale x 8 x i8> [[TMP1]] to <vscale x 8 x i16> // CHECK-NEXT: ret <vscale x 8 x i16> [[TMP2]] // // CPP-CHECK-LABEL: @_Z16test_svld1ub_u16u10__SVBool_tPKh( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr [[BASE:%.*]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i8> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr align 1 [[BASE:%.*]], <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i8> zeroinitializer) // CPP-CHECK-NEXT: [[TMP2:%.*]] = zext <vscale x 8 x i8> [[TMP1]] to <vscale x 8 x i16> // CPP-CHECK-NEXT: ret <vscale x 8 x i16> [[TMP2]] // @@ -101,14 +101,14 @@ svuint16_t test_svld1ub_u16(svbool_t pg, const uint8_t *base) MODE_ATTR // CHECK-LABEL: @test_svld1ub_u32( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr align 1 [[BASE:%.*]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i8> zeroinitializer) // CHECK-NEXT: [[TMP2:%.*]] = zext <vscale x 4 x i8> [[TMP1]] to <vscale x 4 x i32> // CHECK-NEXT: ret <vscale x 4 x i32> [[TMP2]] // // CPP-CHECK-LABEL: @_Z16test_svld1ub_u32u10__SVBool_tPKh( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i8> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr align 1 [[BASE:%.*]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i8> zeroinitializer) // CPP-CHECK-NEXT: [[TMP2:%.*]] = zext <vscale x 4 x i8> [[TMP1]] to <vscale x 4 x i32> // CPP-CHECK-NEXT: ret <vscale x 4 x i32> [[TMP2]] // @@ -120,14 +120,14 @@ svuint32_t test_svld1ub_u32(svbool_t pg, const uint8_t *base) MODE_ATTR // CHECK-LABEL: @test_svld1ub_u64( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i8> zeroinitializer) // CHECK-NEXT: [[TMP2:%.*]] = zext <vscale x 2 x i8> [[TMP1]] to <vscale x 2 x i64> // CHECK-NEXT: ret <vscale x 2 x i64> [[TMP2]] // // CPP-CHECK-LABEL: @_Z16test_svld1ub_u64u10__SVBool_tPKh( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i8> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i8> zeroinitializer) // CPP-CHECK-NEXT: [[TMP2:%.*]] = zext <vscale x 2 x i8> [[TMP1]] to <vscale x 2 x i64> // CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP2]] // @@ -143,7 +143,7 @@ svuint64_t test_svld1ub_u64(svbool_t pg, const uint8_t *base) MODE_ATTR // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 3 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr align 1 [[TMP3]], <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i8> zeroinitializer) // CHECK-NEXT: [[TMP5:%.*]] = zext <vscale x 8 x i8> [[TMP4]] to <vscale x 8 x i16> // CHECK-NEXT: ret <vscale x 8 x i16> [[TMP5]] // @@ -154,7 +154,7 @@ svuint64_t test_svld1ub_u64(svbool_t pg, const uint8_t *base) MODE_ATTR // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 3 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i8> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr align 1 [[TMP3]], <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i8> zeroinitializer) // CPP-CHECK-NEXT: [[TMP5:%.*]] = zext <vscale x 8 x i8> [[TMP4]] to <vscale x 8 x i16> // CPP-CHECK-NEXT: ret <vscale x 8 x i16> [[TMP5]] // @@ -170,7 +170,7 @@ svint16_t test_svld1ub_vnum_s16(svbool_t pg, const uint8_t *base, int64_t vnum) // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 2 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr align 1 [[TMP3]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i8> zeroinitializer) // CHECK-NEXT: [[TMP5:%.*]] = zext <vscale x 4 x i8> [[TMP4]] to <vscale x 4 x i32> // CHECK-NEXT: ret <vscale x 4 x i32> [[TMP5]] // @@ -181,7 +181,7 @@ svint16_t test_svld1ub_vnum_s16(svbool_t pg, const uint8_t *base, int64_t vnum) // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 2 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i8> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr align 1 [[TMP3]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i8> zeroinitializer) // CPP-CHECK-NEXT: [[TMP5:%.*]] = zext <vscale x 4 x i8> [[TMP4]] to <vscale x 4 x i32> // CPP-CHECK-NEXT: ret <vscale x 4 x i32> [[TMP5]] // @@ -197,7 +197,7 @@ svint32_t test_svld1ub_vnum_s32(svbool_t pg, const uint8_t *base, int64_t vnum) // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 1 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i8> zeroinitializer) // CHECK-NEXT: [[TMP5:%.*]] = zext <vscale x 2 x i8> [[TMP4]] to <vscale x 2 x i64> // CHECK-NEXT: ret <vscale x 2 x i64> [[TMP5]] // @@ -208,7 +208,7 @@ svint32_t test_svld1ub_vnum_s32(svbool_t pg, const uint8_t *base, int64_t vnum) // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 1 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i8> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i8> zeroinitializer) // CPP-CHECK-NEXT: [[TMP5:%.*]] = zext <vscale x 2 x i8> [[TMP4]] to <vscale x 2 x i64> // CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP5]] // @@ -224,7 +224,7 @@ svint64_t test_svld1ub_vnum_s64(svbool_t pg, const uint8_t *base, int64_t vnum) // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 3 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr align 1 [[TMP3]], <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i8> zeroinitializer) // CHECK-NEXT: [[TMP5:%.*]] = zext <vscale x 8 x i8> [[TMP4]] to <vscale x 8 x i16> // CHECK-NEXT: ret <vscale x 8 x i16> [[TMP5]] // @@ -235,7 +235,7 @@ svint64_t test_svld1ub_vnum_s64(svbool_t pg, const uint8_t *base, int64_t vnum) // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 3 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i8> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 8 x i8> @llvm.masked.load.nxv8i8.p0(ptr align 1 [[TMP3]], <vscale x 8 x i1> [[TMP0]], <vscale x 8 x i8> zeroinitializer) // CPP-CHECK-NEXT: [[TMP5:%.*]] = zext <vscale x 8 x i8> [[TMP4]] to <vscale x 8 x i16> // CPP-CHECK-NEXT: ret <vscale x 8 x i16> [[TMP5]] // @@ -251,7 +251,7 @@ svuint16_t test_svld1ub_vnum_u16(svbool_t pg, const uint8_t *base, int64_t vnum) // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 2 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr align 1 [[TMP3]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i8> zeroinitializer) // CHECK-NEXT: [[TMP5:%.*]] = zext <vscale x 4 x i8> [[TMP4]] to <vscale x 4 x i32> // CHECK-NEXT: ret <vscale x 4 x i32> [[TMP5]] // @@ -262,7 +262,7 @@ svuint16_t test_svld1ub_vnum_u16(svbool_t pg, const uint8_t *base, int64_t vnum) // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 2 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i8> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i8> @llvm.masked.load.nxv4i8.p0(ptr align 1 [[TMP3]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i8> zeroinitializer) // CPP-CHECK-NEXT: [[TMP5:%.*]] = zext <vscale x 4 x i8> [[TMP4]] to <vscale x 4 x i32> // CPP-CHECK-NEXT: ret <vscale x 4 x i32> [[TMP5]] // @@ -278,7 +278,7 @@ svuint32_t test_svld1ub_vnum_u32(svbool_t pg, const uint8_t *base, int64_t vnum) // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 1 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i8> zeroinitializer) // CHECK-NEXT: [[TMP5:%.*]] = zext <vscale x 2 x i8> [[TMP4]] to <vscale x 2 x i64> // CHECK-NEXT: ret <vscale x 2 x i64> [[TMP5]] // @@ -289,7 +289,7 @@ svuint32_t test_svld1ub_vnum_u32(svbool_t pg, const uint8_t *base, int64_t vnum) // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 1 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i8> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i8> @llvm.masked.load.nxv2i8.p0(ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i8> zeroinitializer) // CPP-CHECK-NEXT: [[TMP5:%.*]] = zext <vscale x 2 x i8> [[TMP4]] to <vscale x 2 x i64> // CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP5]] // diff --git a/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_ld1uh.c b/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_ld1uh.c index 6d91c1e..18dfc082 100644 --- a/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_ld1uh.c +++ b/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_ld1uh.c @@ -25,14 +25,14 @@ // CHECK-LABEL: @test_svld1uh_s32( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i16> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr align 1 [[BASE:%.*]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i16> zeroinitializer) // CHECK-NEXT: [[TMP2:%.*]] = zext <vscale x 4 x i16> [[TMP1]] to <vscale x 4 x i32> // CHECK-NEXT: ret <vscale x 4 x i32> [[TMP2]] // // CPP-CHECK-LABEL: @_Z16test_svld1uh_s32u10__SVBool_tPKt( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i16> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr align 1 [[BASE:%.*]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i16> zeroinitializer) // CPP-CHECK-NEXT: [[TMP2:%.*]] = zext <vscale x 4 x i16> [[TMP1]] to <vscale x 4 x i32> // CPP-CHECK-NEXT: ret <vscale x 4 x i32> [[TMP2]] // @@ -44,14 +44,14 @@ svint32_t test_svld1uh_s32(svbool_t pg, const uint16_t *base) MODE_ATTR // CHECK-LABEL: @test_svld1uh_s64( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i16> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i16> zeroinitializer) // CHECK-NEXT: [[TMP2:%.*]] = zext <vscale x 2 x i16> [[TMP1]] to <vscale x 2 x i64> // CHECK-NEXT: ret <vscale x 2 x i64> [[TMP2]] // // CPP-CHECK-LABEL: @_Z16test_svld1uh_s64u10__SVBool_tPKt( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i16> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i16> zeroinitializer) // CPP-CHECK-NEXT: [[TMP2:%.*]] = zext <vscale x 2 x i16> [[TMP1]] to <vscale x 2 x i64> // CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP2]] // @@ -63,14 +63,14 @@ svint64_t test_svld1uh_s64(svbool_t pg, const uint16_t *base) MODE_ATTR // CHECK-LABEL: @test_svld1uh_u32( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i16> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr align 1 [[BASE:%.*]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i16> zeroinitializer) // CHECK-NEXT: [[TMP2:%.*]] = zext <vscale x 4 x i16> [[TMP1]] to <vscale x 4 x i32> // CHECK-NEXT: ret <vscale x 4 x i32> [[TMP2]] // // CPP-CHECK-LABEL: @_Z16test_svld1uh_u32u10__SVBool_tPKt( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i16> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr align 1 [[BASE:%.*]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i16> zeroinitializer) // CPP-CHECK-NEXT: [[TMP2:%.*]] = zext <vscale x 4 x i16> [[TMP1]] to <vscale x 4 x i32> // CPP-CHECK-NEXT: ret <vscale x 4 x i32> [[TMP2]] // @@ -82,14 +82,14 @@ svuint32_t test_svld1uh_u32(svbool_t pg, const uint16_t *base) MODE_ATTR // CHECK-LABEL: @test_svld1uh_u64( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i16> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i16> zeroinitializer) // CHECK-NEXT: [[TMP2:%.*]] = zext <vscale x 2 x i16> [[TMP1]] to <vscale x 2 x i64> // CHECK-NEXT: ret <vscale x 2 x i64> [[TMP2]] // // CPP-CHECK-LABEL: @_Z16test_svld1uh_u64u10__SVBool_tPKt( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i16> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i16> zeroinitializer) // CPP-CHECK-NEXT: [[TMP2:%.*]] = zext <vscale x 2 x i16> [[TMP1]] to <vscale x 2 x i64> // CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP2]] // @@ -105,7 +105,7 @@ svuint64_t test_svld1uh_u64(svbool_t pg, const uint16_t *base) MODE_ATTR // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 3 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i16> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr align 1 [[TMP3]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i16> zeroinitializer) // CHECK-NEXT: [[TMP5:%.*]] = zext <vscale x 4 x i16> [[TMP4]] to <vscale x 4 x i32> // CHECK-NEXT: ret <vscale x 4 x i32> [[TMP5]] // @@ -116,7 +116,7 @@ svuint64_t test_svld1uh_u64(svbool_t pg, const uint16_t *base) MODE_ATTR // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 3 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i16> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr align 1 [[TMP3]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i16> zeroinitializer) // CPP-CHECK-NEXT: [[TMP5:%.*]] = zext <vscale x 4 x i16> [[TMP4]] to <vscale x 4 x i32> // CPP-CHECK-NEXT: ret <vscale x 4 x i32> [[TMP5]] // @@ -132,7 +132,7 @@ svint32_t test_svld1uh_vnum_s32(svbool_t pg, const uint16_t *base, int64_t vnum) // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 2 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i16> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i16> zeroinitializer) // CHECK-NEXT: [[TMP5:%.*]] = zext <vscale x 2 x i16> [[TMP4]] to <vscale x 2 x i64> // CHECK-NEXT: ret <vscale x 2 x i64> [[TMP5]] // @@ -143,7 +143,7 @@ svint32_t test_svld1uh_vnum_s32(svbool_t pg, const uint16_t *base, int64_t vnum) // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 2 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i16> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i16> zeroinitializer) // CPP-CHECK-NEXT: [[TMP5:%.*]] = zext <vscale x 2 x i16> [[TMP4]] to <vscale x 2 x i64> // CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP5]] // @@ -159,7 +159,7 @@ svint64_t test_svld1uh_vnum_s64(svbool_t pg, const uint16_t *base, int64_t vnum) // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 3 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i16> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr align 1 [[TMP3]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i16> zeroinitializer) // CHECK-NEXT: [[TMP5:%.*]] = zext <vscale x 4 x i16> [[TMP4]] to <vscale x 4 x i32> // CHECK-NEXT: ret <vscale x 4 x i32> [[TMP5]] // @@ -170,7 +170,7 @@ svint64_t test_svld1uh_vnum_s64(svbool_t pg, const uint16_t *base, int64_t vnum) // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 3 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i16> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 4 x i16> @llvm.masked.load.nxv4i16.p0(ptr align 1 [[TMP3]], <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i16> zeroinitializer) // CPP-CHECK-NEXT: [[TMP5:%.*]] = zext <vscale x 4 x i16> [[TMP4]] to <vscale x 4 x i32> // CPP-CHECK-NEXT: ret <vscale x 4 x i32> [[TMP5]] // @@ -186,7 +186,7 @@ svuint32_t test_svld1uh_vnum_u32(svbool_t pg, const uint16_t *base, int64_t vnum // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 2 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i16> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i16> zeroinitializer) // CHECK-NEXT: [[TMP5:%.*]] = zext <vscale x 2 x i16> [[TMP4]] to <vscale x 2 x i64> // CHECK-NEXT: ret <vscale x 2 x i64> [[TMP5]] // @@ -197,7 +197,7 @@ svuint32_t test_svld1uh_vnum_u32(svbool_t pg, const uint16_t *base, int64_t vnum // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 2 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i16> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i16> @llvm.masked.load.nxv2i16.p0(ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i16> zeroinitializer) // CPP-CHECK-NEXT: [[TMP5:%.*]] = zext <vscale x 2 x i16> [[TMP4]] to <vscale x 2 x i64> // CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP5]] // diff --git a/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_ld1uw.c b/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_ld1uw.c index 7be2398..62637ff 100644 --- a/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_ld1uw.c +++ b/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_ld1uw.c @@ -25,14 +25,14 @@ // CHECK-LABEL: @test_svld1uw_s64( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i32> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i32> zeroinitializer) // CHECK-NEXT: [[TMP2:%.*]] = zext <vscale x 2 x i32> [[TMP1]] to <vscale x 2 x i64> // CHECK-NEXT: ret <vscale x 2 x i64> [[TMP2]] // // CPP-CHECK-LABEL: @_Z16test_svld1uw_s64u10__SVBool_tPKj( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i32> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i32> zeroinitializer) // CPP-CHECK-NEXT: [[TMP2:%.*]] = zext <vscale x 2 x i32> [[TMP1]] to <vscale x 2 x i64> // CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP2]] // @@ -44,14 +44,14 @@ svint64_t test_svld1uw_s64(svbool_t pg, const uint32_t *base) MODE_ATTR // CHECK-LABEL: @test_svld1uw_u64( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i32> zeroinitializer) +// CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i32> zeroinitializer) // CHECK-NEXT: [[TMP2:%.*]] = zext <vscale x 2 x i32> [[TMP1]] to <vscale x 2 x i64> // CHECK-NEXT: ret <vscale x 2 x i64> [[TMP2]] // // CPP-CHECK-LABEL: @_Z16test_svld1uw_u64u10__SVBool_tPKj( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i32> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP1:%.*]] = tail call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i32> zeroinitializer) // CPP-CHECK-NEXT: [[TMP2:%.*]] = zext <vscale x 2 x i32> [[TMP1]] to <vscale x 2 x i64> // CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP2]] // @@ -67,7 +67,7 @@ svuint64_t test_svld1uw_u64(svbool_t pg, const uint32_t *base) MODE_ATTR // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 3 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i32> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i32> zeroinitializer) // CHECK-NEXT: [[TMP5:%.*]] = zext <vscale x 2 x i32> [[TMP4]] to <vscale x 2 x i64> // CHECK-NEXT: ret <vscale x 2 x i64> [[TMP5]] // @@ -78,7 +78,7 @@ svuint64_t test_svld1uw_u64(svbool_t pg, const uint32_t *base) MODE_ATTR // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 3 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i32> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i32> zeroinitializer) // CPP-CHECK-NEXT: [[TMP5:%.*]] = zext <vscale x 2 x i32> [[TMP4]] to <vscale x 2 x i64> // CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP5]] // @@ -94,7 +94,7 @@ svint64_t test_svld1uw_vnum_s64(svbool_t pg, const uint32_t *base, int64_t vnum) // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 3 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i32> zeroinitializer) +// CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i32> zeroinitializer) // CHECK-NEXT: [[TMP5:%.*]] = zext <vscale x 2 x i32> [[TMP4]] to <vscale x 2 x i64> // CHECK-NEXT: ret <vscale x 2 x i64> [[TMP5]] // @@ -105,7 +105,7 @@ svint64_t test_svld1uw_vnum_s64(svbool_t pg, const uint32_t *base, int64_t vnum) // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 3 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i32> zeroinitializer) +// CPP-CHECK-NEXT: [[TMP4:%.*]] = tail call <vscale x 2 x i32> @llvm.masked.load.nxv2i32.p0(ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]], <vscale x 2 x i32> zeroinitializer) // CPP-CHECK-NEXT: [[TMP5:%.*]] = zext <vscale x 2 x i32> [[TMP4]] to <vscale x 2 x i64> // CPP-CHECK-NEXT: ret <vscale x 2 x i64> [[TMP5]] // diff --git a/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_st1.c b/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_st1.c index 56f8c32..4d0005e 100644 --- a/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_st1.c +++ b/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_st1.c @@ -24,12 +24,12 @@ // CHECK-LABEL: @test_svst1_s8( // CHECK-NEXT: entry: -// CHECK-NEXT: tail call void @llvm.masked.store.nxv16i8.p0(<vscale x 16 x i8> [[DATA:%.*]], ptr [[BASE:%.*]], i32 1, <vscale x 16 x i1> [[PG:%.*]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv16i8.p0(<vscale x 16 x i8> [[DATA:%.*]], ptr align 1 [[BASE:%.*]], <vscale x 16 x i1> [[PG:%.*]]) // CHECK-NEXT: ret void // // CPP-CHECK-LABEL: @_Z13test_svst1_s8u10__SVBool_tPau10__SVInt8_t( // CPP-CHECK-NEXT: entry: -// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv16i8.p0(<vscale x 16 x i8> [[DATA:%.*]], ptr [[BASE:%.*]], i32 1, <vscale x 16 x i1> [[PG:%.*]]) +// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv16i8.p0(<vscale x 16 x i8> [[DATA:%.*]], ptr align 1 [[BASE:%.*]], <vscale x 16 x i1> [[PG:%.*]]) // CPP-CHECK-NEXT: ret void // void test_svst1_s8(svbool_t pg, int8_t *base, svint8_t data) MODE_ATTR @@ -40,13 +40,13 @@ void test_svst1_s8(svbool_t pg, int8_t *base, svint8_t data) MODE_ATTR // CHECK-LABEL: @test_svst1_s16( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: tail call void @llvm.masked.store.nxv8i16.p0(<vscale x 8 x i16> [[DATA:%.*]], ptr [[BASE:%.*]], i32 1, <vscale x 8 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv8i16.p0(<vscale x 8 x i16> [[DATA:%.*]], ptr align 1 [[BASE:%.*]], <vscale x 8 x i1> [[TMP0]]) // CHECK-NEXT: ret void // // CPP-CHECK-LABEL: @_Z14test_svst1_s16u10__SVBool_tPsu11__SVInt16_t( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv8i16.p0(<vscale x 8 x i16> [[DATA:%.*]], ptr [[BASE:%.*]], i32 1, <vscale x 8 x i1> [[TMP0]]) +// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv8i16.p0(<vscale x 8 x i16> [[DATA:%.*]], ptr align 1 [[BASE:%.*]], <vscale x 8 x i1> [[TMP0]]) // CPP-CHECK-NEXT: ret void // void test_svst1_s16(svbool_t pg, int16_t *base, svint16_t data) MODE_ATTR @@ -57,13 +57,13 @@ void test_svst1_s16(svbool_t pg, int16_t *base, svint16_t data) MODE_ATTR // CHECK-LABEL: @test_svst1_s32( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: tail call void @llvm.masked.store.nxv4i32.p0(<vscale x 4 x i32> [[DATA:%.*]], ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv4i32.p0(<vscale x 4 x i32> [[DATA:%.*]], ptr align 1 [[BASE:%.*]], <vscale x 4 x i1> [[TMP0]]) // CHECK-NEXT: ret void // // CPP-CHECK-LABEL: @_Z14test_svst1_s32u10__SVBool_tPiu11__SVInt32_t( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv4i32.p0(<vscale x 4 x i32> [[DATA:%.*]], ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]]) +// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv4i32.p0(<vscale x 4 x i32> [[DATA:%.*]], ptr align 1 [[BASE:%.*]], <vscale x 4 x i1> [[TMP0]]) // CPP-CHECK-NEXT: ret void // void test_svst1_s32(svbool_t pg, int32_t *base, svint32_t data) MODE_ATTR @@ -74,13 +74,13 @@ void test_svst1_s32(svbool_t pg, int32_t *base, svint32_t data) MODE_ATTR // CHECK-LABEL: @test_svst1_s64( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: tail call void @llvm.masked.store.nxv2i64.p0(<vscale x 2 x i64> [[DATA:%.*]], ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv2i64.p0(<vscale x 2 x i64> [[DATA:%.*]], ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]]) // CHECK-NEXT: ret void // // CPP-CHECK-LABEL: @_Z14test_svst1_s64u10__SVBool_tPlu11__SVInt64_t( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv2i64.p0(<vscale x 2 x i64> [[DATA:%.*]], ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]]) +// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv2i64.p0(<vscale x 2 x i64> [[DATA:%.*]], ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]]) // CPP-CHECK-NEXT: ret void // void test_svst1_s64(svbool_t pg, int64_t *base, svint64_t data) MODE_ATTR @@ -90,12 +90,12 @@ void test_svst1_s64(svbool_t pg, int64_t *base, svint64_t data) MODE_ATTR // CHECK-LABEL: @test_svst1_u8( // CHECK-NEXT: entry: -// CHECK-NEXT: tail call void @llvm.masked.store.nxv16i8.p0(<vscale x 16 x i8> [[DATA:%.*]], ptr [[BASE:%.*]], i32 1, <vscale x 16 x i1> [[PG:%.*]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv16i8.p0(<vscale x 16 x i8> [[DATA:%.*]], ptr align 1 [[BASE:%.*]], <vscale x 16 x i1> [[PG:%.*]]) // CHECK-NEXT: ret void // // CPP-CHECK-LABEL: @_Z13test_svst1_u8u10__SVBool_tPhu11__SVUint8_t( // CPP-CHECK-NEXT: entry: -// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv16i8.p0(<vscale x 16 x i8> [[DATA:%.*]], ptr [[BASE:%.*]], i32 1, <vscale x 16 x i1> [[PG:%.*]]) +// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv16i8.p0(<vscale x 16 x i8> [[DATA:%.*]], ptr align 1 [[BASE:%.*]], <vscale x 16 x i1> [[PG:%.*]]) // CPP-CHECK-NEXT: ret void // void test_svst1_u8(svbool_t pg, uint8_t *base, svuint8_t data) MODE_ATTR @@ -106,13 +106,13 @@ void test_svst1_u8(svbool_t pg, uint8_t *base, svuint8_t data) MODE_ATTR // CHECK-LABEL: @test_svst1_u16( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: tail call void @llvm.masked.store.nxv8i16.p0(<vscale x 8 x i16> [[DATA:%.*]], ptr [[BASE:%.*]], i32 1, <vscale x 8 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv8i16.p0(<vscale x 8 x i16> [[DATA:%.*]], ptr align 1 [[BASE:%.*]], <vscale x 8 x i1> [[TMP0]]) // CHECK-NEXT: ret void // // CPP-CHECK-LABEL: @_Z14test_svst1_u16u10__SVBool_tPtu12__SVUint16_t( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv8i16.p0(<vscale x 8 x i16> [[DATA:%.*]], ptr [[BASE:%.*]], i32 1, <vscale x 8 x i1> [[TMP0]]) +// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv8i16.p0(<vscale x 8 x i16> [[DATA:%.*]], ptr align 1 [[BASE:%.*]], <vscale x 8 x i1> [[TMP0]]) // CPP-CHECK-NEXT: ret void // void test_svst1_u16(svbool_t pg, uint16_t *base, svuint16_t data) MODE_ATTR @@ -123,13 +123,13 @@ void test_svst1_u16(svbool_t pg, uint16_t *base, svuint16_t data) MODE_ATTR // CHECK-LABEL: @test_svst1_u32( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: tail call void @llvm.masked.store.nxv4i32.p0(<vscale x 4 x i32> [[DATA:%.*]], ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv4i32.p0(<vscale x 4 x i32> [[DATA:%.*]], ptr align 1 [[BASE:%.*]], <vscale x 4 x i1> [[TMP0]]) // CHECK-NEXT: ret void // // CPP-CHECK-LABEL: @_Z14test_svst1_u32u10__SVBool_tPju12__SVUint32_t( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv4i32.p0(<vscale x 4 x i32> [[DATA:%.*]], ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]]) +// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv4i32.p0(<vscale x 4 x i32> [[DATA:%.*]], ptr align 1 [[BASE:%.*]], <vscale x 4 x i1> [[TMP0]]) // CPP-CHECK-NEXT: ret void // void test_svst1_u32(svbool_t pg, uint32_t *base, svuint32_t data) MODE_ATTR @@ -140,13 +140,13 @@ void test_svst1_u32(svbool_t pg, uint32_t *base, svuint32_t data) MODE_ATTR // CHECK-LABEL: @test_svst1_u64( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: tail call void @llvm.masked.store.nxv2i64.p0(<vscale x 2 x i64> [[DATA:%.*]], ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv2i64.p0(<vscale x 2 x i64> [[DATA:%.*]], ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]]) // CHECK-NEXT: ret void // // CPP-CHECK-LABEL: @_Z14test_svst1_u64u10__SVBool_tPmu12__SVUint64_t( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv2i64.p0(<vscale x 2 x i64> [[DATA:%.*]], ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]]) +// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv2i64.p0(<vscale x 2 x i64> [[DATA:%.*]], ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]]) // CPP-CHECK-NEXT: ret void // void test_svst1_u64(svbool_t pg, uint64_t *base, svuint64_t data) MODE_ATTR @@ -157,13 +157,13 @@ void test_svst1_u64(svbool_t pg, uint64_t *base, svuint64_t data) MODE_ATTR // CHECK-LABEL: @test_svst1_f16( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: tail call void @llvm.masked.store.nxv8f16.p0(<vscale x 8 x half> [[DATA:%.*]], ptr [[BASE:%.*]], i32 1, <vscale x 8 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv8f16.p0(<vscale x 8 x half> [[DATA:%.*]], ptr align 1 [[BASE:%.*]], <vscale x 8 x i1> [[TMP0]]) // CHECK-NEXT: ret void // // CPP-CHECK-LABEL: @_Z14test_svst1_f16u10__SVBool_tPDhu13__SVFloat16_t( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv8f16.p0(<vscale x 8 x half> [[DATA:%.*]], ptr [[BASE:%.*]], i32 1, <vscale x 8 x i1> [[TMP0]]) +// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv8f16.p0(<vscale x 8 x half> [[DATA:%.*]], ptr align 1 [[BASE:%.*]], <vscale x 8 x i1> [[TMP0]]) // CPP-CHECK-NEXT: ret void // void test_svst1_f16(svbool_t pg, float16_t *base, svfloat16_t data) MODE_ATTR @@ -174,13 +174,13 @@ void test_svst1_f16(svbool_t pg, float16_t *base, svfloat16_t data) MODE_ATTR // CHECK-LABEL: @test_svst1_f32( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: tail call void @llvm.masked.store.nxv4f32.p0(<vscale x 4 x float> [[DATA:%.*]], ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv4f32.p0(<vscale x 4 x float> [[DATA:%.*]], ptr align 1 [[BASE:%.*]], <vscale x 4 x i1> [[TMP0]]) // CHECK-NEXT: ret void // // CPP-CHECK-LABEL: @_Z14test_svst1_f32u10__SVBool_tPfu13__SVFloat32_t( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv4f32.p0(<vscale x 4 x float> [[DATA:%.*]], ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]]) +// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv4f32.p0(<vscale x 4 x float> [[DATA:%.*]], ptr align 1 [[BASE:%.*]], <vscale x 4 x i1> [[TMP0]]) // CPP-CHECK-NEXT: ret void // void test_svst1_f32(svbool_t pg, float32_t *base, svfloat32_t data) MODE_ATTR @@ -191,13 +191,13 @@ void test_svst1_f32(svbool_t pg, float32_t *base, svfloat32_t data) MODE_ATTR // CHECK-LABEL: @test_svst1_f64( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: tail call void @llvm.masked.store.nxv2f64.p0(<vscale x 2 x double> [[DATA:%.*]], ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv2f64.p0(<vscale x 2 x double> [[DATA:%.*]], ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]]) // CHECK-NEXT: ret void // // CPP-CHECK-LABEL: @_Z14test_svst1_f64u10__SVBool_tPdu13__SVFloat64_t( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv2f64.p0(<vscale x 2 x double> [[DATA:%.*]], ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]]) +// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv2f64.p0(<vscale x 2 x double> [[DATA:%.*]], ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]]) // CPP-CHECK-NEXT: ret void // void test_svst1_f64(svbool_t pg, float64_t *base, svfloat64_t data) MODE_ATTR @@ -207,12 +207,12 @@ void test_svst1_f64(svbool_t pg, float64_t *base, svfloat64_t data) MODE_ATTR // CHECK-LABEL: @test_svst1_mf8( // CHECK-NEXT: entry: -// CHECK-NEXT: tail call void @llvm.masked.store.nxv16i8.p0(<vscale x 16 x i8> [[DATA:%.*]], ptr [[BASE:%.*]], i32 1, <vscale x 16 x i1> [[PG:%.*]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv16i8.p0(<vscale x 16 x i8> [[DATA:%.*]], ptr align 1 [[BASE:%.*]], <vscale x 16 x i1> [[PG:%.*]]) // CHECK-NEXT: ret void // // CPP-CHECK-LABEL: @_Z14test_svst1_mf8u10__SVBool_tPu6__mfp8u13__SVMfloat8_t( // CPP-CHECK-NEXT: entry: -// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv16i8.p0(<vscale x 16 x i8> [[DATA:%.*]], ptr [[BASE:%.*]], i32 1, <vscale x 16 x i1> [[PG:%.*]]) +// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv16i8.p0(<vscale x 16 x i8> [[DATA:%.*]], ptr align 1 [[BASE:%.*]], <vscale x 16 x i1> [[PG:%.*]]) // CPP-CHECK-NEXT: ret void // void test_svst1_mf8(svbool_t pg, mfloat8_t *base, svmfloat8_t data) MODE_ATTR @@ -226,7 +226,7 @@ void test_svst1_mf8(svbool_t pg, mfloat8_t *base, svmfloat8_t data) MODE_ATTR // CHECK-NEXT: [[TMP1:%.*]] = shl nuw nsw i64 [[TMP0]], 4 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP1]] // CHECK-NEXT: [[TMP2:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: tail call void @llvm.masked.store.nxv16i8.p0(<vscale x 16 x i8> [[DATA:%.*]], ptr [[TMP2]], i32 1, <vscale x 16 x i1> [[PG:%.*]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv16i8.p0(<vscale x 16 x i8> [[DATA:%.*]], ptr align 1 [[TMP2]], <vscale x 16 x i1> [[PG:%.*]]) // CHECK-NEXT: ret void // // CPP-CHECK-LABEL: @_Z18test_svst1_vnum_s8u10__SVBool_tPalu10__SVInt8_t( @@ -235,7 +235,7 @@ void test_svst1_mf8(svbool_t pg, mfloat8_t *base, svmfloat8_t data) MODE_ATTR // CPP-CHECK-NEXT: [[TMP1:%.*]] = shl nuw nsw i64 [[TMP0]], 4 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP1]] // CPP-CHECK-NEXT: [[TMP2:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv16i8.p0(<vscale x 16 x i8> [[DATA:%.*]], ptr [[TMP2]], i32 1, <vscale x 16 x i1> [[PG:%.*]]) +// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv16i8.p0(<vscale x 16 x i8> [[DATA:%.*]], ptr align 1 [[TMP2]], <vscale x 16 x i1> [[PG:%.*]]) // CPP-CHECK-NEXT: ret void // void test_svst1_vnum_s8(svbool_t pg, int8_t *base, int64_t vnum, svint8_t data) MODE_ATTR @@ -250,7 +250,7 @@ void test_svst1_vnum_s8(svbool_t pg, int8_t *base, int64_t vnum, svint8_t data) // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: tail call void @llvm.masked.store.nxv8i16.p0(<vscale x 8 x i16> [[DATA:%.*]], ptr [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv8i16.p0(<vscale x 8 x i16> [[DATA:%.*]], ptr align 1 [[TMP3]], <vscale x 8 x i1> [[TMP0]]) // CHECK-NEXT: ret void // // CPP-CHECK-LABEL: @_Z19test_svst1_vnum_s16u10__SVBool_tPslu11__SVInt16_t( @@ -260,7 +260,7 @@ void test_svst1_vnum_s8(svbool_t pg, int8_t *base, int64_t vnum, svint8_t data) // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv8i16.p0(<vscale x 8 x i16> [[DATA:%.*]], ptr [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]]) +// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv8i16.p0(<vscale x 8 x i16> [[DATA:%.*]], ptr align 1 [[TMP3]], <vscale x 8 x i1> [[TMP0]]) // CPP-CHECK-NEXT: ret void // void test_svst1_vnum_s16(svbool_t pg, int16_t *base, int64_t vnum, svint16_t data) MODE_ATTR @@ -275,7 +275,7 @@ void test_svst1_vnum_s16(svbool_t pg, int16_t *base, int64_t vnum, svint16_t dat // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: tail call void @llvm.masked.store.nxv4i32.p0(<vscale x 4 x i32> [[DATA:%.*]], ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv4i32.p0(<vscale x 4 x i32> [[DATA:%.*]], ptr align 1 [[TMP3]], <vscale x 4 x i1> [[TMP0]]) // CHECK-NEXT: ret void // // CPP-CHECK-LABEL: @_Z19test_svst1_vnum_s32u10__SVBool_tPilu11__SVInt32_t( @@ -285,7 +285,7 @@ void test_svst1_vnum_s16(svbool_t pg, int16_t *base, int64_t vnum, svint16_t dat // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv4i32.p0(<vscale x 4 x i32> [[DATA:%.*]], ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]]) +// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv4i32.p0(<vscale x 4 x i32> [[DATA:%.*]], ptr align 1 [[TMP3]], <vscale x 4 x i1> [[TMP0]]) // CPP-CHECK-NEXT: ret void // void test_svst1_vnum_s32(svbool_t pg, int32_t *base, int64_t vnum, svint32_t data) MODE_ATTR @@ -300,7 +300,7 @@ void test_svst1_vnum_s32(svbool_t pg, int32_t *base, int64_t vnum, svint32_t dat // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: tail call void @llvm.masked.store.nxv2i64.p0(<vscale x 2 x i64> [[DATA:%.*]], ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv2i64.p0(<vscale x 2 x i64> [[DATA:%.*]], ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]]) // CHECK-NEXT: ret void // // CPP-CHECK-LABEL: @_Z19test_svst1_vnum_s64u10__SVBool_tPllu11__SVInt64_t( @@ -310,7 +310,7 @@ void test_svst1_vnum_s32(svbool_t pg, int32_t *base, int64_t vnum, svint32_t dat // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv2i64.p0(<vscale x 2 x i64> [[DATA:%.*]], ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]]) +// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv2i64.p0(<vscale x 2 x i64> [[DATA:%.*]], ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]]) // CPP-CHECK-NEXT: ret void // void test_svst1_vnum_s64(svbool_t pg, int64_t *base, int64_t vnum, svint64_t data) MODE_ATTR @@ -324,7 +324,7 @@ void test_svst1_vnum_s64(svbool_t pg, int64_t *base, int64_t vnum, svint64_t dat // CHECK-NEXT: [[TMP1:%.*]] = shl nuw nsw i64 [[TMP0]], 4 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP1]] // CHECK-NEXT: [[TMP2:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: tail call void @llvm.masked.store.nxv16i8.p0(<vscale x 16 x i8> [[DATA:%.*]], ptr [[TMP2]], i32 1, <vscale x 16 x i1> [[PG:%.*]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv16i8.p0(<vscale x 16 x i8> [[DATA:%.*]], ptr align 1 [[TMP2]], <vscale x 16 x i1> [[PG:%.*]]) // CHECK-NEXT: ret void // // CPP-CHECK-LABEL: @_Z18test_svst1_vnum_u8u10__SVBool_tPhlu11__SVUint8_t( @@ -333,7 +333,7 @@ void test_svst1_vnum_s64(svbool_t pg, int64_t *base, int64_t vnum, svint64_t dat // CPP-CHECK-NEXT: [[TMP1:%.*]] = shl nuw nsw i64 [[TMP0]], 4 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP1]] // CPP-CHECK-NEXT: [[TMP2:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv16i8.p0(<vscale x 16 x i8> [[DATA:%.*]], ptr [[TMP2]], i32 1, <vscale x 16 x i1> [[PG:%.*]]) +// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv16i8.p0(<vscale x 16 x i8> [[DATA:%.*]], ptr align 1 [[TMP2]], <vscale x 16 x i1> [[PG:%.*]]) // CPP-CHECK-NEXT: ret void // void test_svst1_vnum_u8(svbool_t pg, uint8_t *base, int64_t vnum, svuint8_t data) MODE_ATTR @@ -348,7 +348,7 @@ void test_svst1_vnum_u8(svbool_t pg, uint8_t *base, int64_t vnum, svuint8_t data // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: tail call void @llvm.masked.store.nxv8i16.p0(<vscale x 8 x i16> [[DATA:%.*]], ptr [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv8i16.p0(<vscale x 8 x i16> [[DATA:%.*]], ptr align 1 [[TMP3]], <vscale x 8 x i1> [[TMP0]]) // CHECK-NEXT: ret void // // CPP-CHECK-LABEL: @_Z19test_svst1_vnum_u16u10__SVBool_tPtlu12__SVUint16_t( @@ -358,7 +358,7 @@ void test_svst1_vnum_u8(svbool_t pg, uint8_t *base, int64_t vnum, svuint8_t data // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv8i16.p0(<vscale x 8 x i16> [[DATA:%.*]], ptr [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]]) +// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv8i16.p0(<vscale x 8 x i16> [[DATA:%.*]], ptr align 1 [[TMP3]], <vscale x 8 x i1> [[TMP0]]) // CPP-CHECK-NEXT: ret void // void test_svst1_vnum_u16(svbool_t pg, uint16_t *base, int64_t vnum, svuint16_t data) MODE_ATTR @@ -373,7 +373,7 @@ void test_svst1_vnum_u16(svbool_t pg, uint16_t *base, int64_t vnum, svuint16_t d // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: tail call void @llvm.masked.store.nxv4i32.p0(<vscale x 4 x i32> [[DATA:%.*]], ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv4i32.p0(<vscale x 4 x i32> [[DATA:%.*]], ptr align 1 [[TMP3]], <vscale x 4 x i1> [[TMP0]]) // CHECK-NEXT: ret void // // CPP-CHECK-LABEL: @_Z19test_svst1_vnum_u32u10__SVBool_tPjlu12__SVUint32_t( @@ -383,7 +383,7 @@ void test_svst1_vnum_u16(svbool_t pg, uint16_t *base, int64_t vnum, svuint16_t d // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv4i32.p0(<vscale x 4 x i32> [[DATA:%.*]], ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]]) +// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv4i32.p0(<vscale x 4 x i32> [[DATA:%.*]], ptr align 1 [[TMP3]], <vscale x 4 x i1> [[TMP0]]) // CPP-CHECK-NEXT: ret void // void test_svst1_vnum_u32(svbool_t pg, uint32_t *base, int64_t vnum, svuint32_t data) MODE_ATTR @@ -398,7 +398,7 @@ void test_svst1_vnum_u32(svbool_t pg, uint32_t *base, int64_t vnum, svuint32_t d // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: tail call void @llvm.masked.store.nxv2i64.p0(<vscale x 2 x i64> [[DATA:%.*]], ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv2i64.p0(<vscale x 2 x i64> [[DATA:%.*]], ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]]) // CHECK-NEXT: ret void // // CPP-CHECK-LABEL: @_Z19test_svst1_vnum_u64u10__SVBool_tPmlu12__SVUint64_t( @@ -408,7 +408,7 @@ void test_svst1_vnum_u32(svbool_t pg, uint32_t *base, int64_t vnum, svuint32_t d // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv2i64.p0(<vscale x 2 x i64> [[DATA:%.*]], ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]]) +// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv2i64.p0(<vscale x 2 x i64> [[DATA:%.*]], ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]]) // CPP-CHECK-NEXT: ret void // void test_svst1_vnum_u64(svbool_t pg, uint64_t *base, int64_t vnum, svuint64_t data) MODE_ATTR @@ -423,7 +423,7 @@ void test_svst1_vnum_u64(svbool_t pg, uint64_t *base, int64_t vnum, svuint64_t d // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: tail call void @llvm.masked.store.nxv8f16.p0(<vscale x 8 x half> [[DATA:%.*]], ptr [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv8f16.p0(<vscale x 8 x half> [[DATA:%.*]], ptr align 1 [[TMP3]], <vscale x 8 x i1> [[TMP0]]) // CHECK-NEXT: ret void // // CPP-CHECK-LABEL: @_Z19test_svst1_vnum_f16u10__SVBool_tPDhlu13__SVFloat16_t( @@ -433,7 +433,7 @@ void test_svst1_vnum_u64(svbool_t pg, uint64_t *base, int64_t vnum, svuint64_t d // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv8f16.p0(<vscale x 8 x half> [[DATA:%.*]], ptr [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]]) +// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv8f16.p0(<vscale x 8 x half> [[DATA:%.*]], ptr align 1 [[TMP3]], <vscale x 8 x i1> [[TMP0]]) // CPP-CHECK-NEXT: ret void // void test_svst1_vnum_f16(svbool_t pg, float16_t *base, int64_t vnum, svfloat16_t data) MODE_ATTR @@ -448,7 +448,7 @@ void test_svst1_vnum_f16(svbool_t pg, float16_t *base, int64_t vnum, svfloat16_t // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: tail call void @llvm.masked.store.nxv4f32.p0(<vscale x 4 x float> [[DATA:%.*]], ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv4f32.p0(<vscale x 4 x float> [[DATA:%.*]], ptr align 1 [[TMP3]], <vscale x 4 x i1> [[TMP0]]) // CHECK-NEXT: ret void // // CPP-CHECK-LABEL: @_Z19test_svst1_vnum_f32u10__SVBool_tPflu13__SVFloat32_t( @@ -458,7 +458,7 @@ void test_svst1_vnum_f16(svbool_t pg, float16_t *base, int64_t vnum, svfloat16_t // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv4f32.p0(<vscale x 4 x float> [[DATA:%.*]], ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]]) +// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv4f32.p0(<vscale x 4 x float> [[DATA:%.*]], ptr align 1 [[TMP3]], <vscale x 4 x i1> [[TMP0]]) // CPP-CHECK-NEXT: ret void // void test_svst1_vnum_f32(svbool_t pg, float32_t *base, int64_t vnum, svfloat32_t data) MODE_ATTR @@ -473,7 +473,7 @@ void test_svst1_vnum_f32(svbool_t pg, float32_t *base, int64_t vnum, svfloat32_t // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: tail call void @llvm.masked.store.nxv2f64.p0(<vscale x 2 x double> [[DATA:%.*]], ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv2f64.p0(<vscale x 2 x double> [[DATA:%.*]], ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]]) // CHECK-NEXT: ret void // // CPP-CHECK-LABEL: @_Z19test_svst1_vnum_f64u10__SVBool_tPdlu13__SVFloat64_t( @@ -483,7 +483,7 @@ void test_svst1_vnum_f32(svbool_t pg, float32_t *base, int64_t vnum, svfloat32_t // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv2f64.p0(<vscale x 2 x double> [[DATA:%.*]], ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]]) +// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv2f64.p0(<vscale x 2 x double> [[DATA:%.*]], ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]]) // CPP-CHECK-NEXT: ret void // void test_svst1_vnum_f64(svbool_t pg, float64_t *base, int64_t vnum, svfloat64_t data) MODE_ATTR @@ -497,7 +497,7 @@ void test_svst1_vnum_f64(svbool_t pg, float64_t *base, int64_t vnum, svfloat64_t // CHECK-NEXT: [[TMP1:%.*]] = shl nuw nsw i64 [[TMP0]], 4 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP1]] // CHECK-NEXT: [[TMP2:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: tail call void @llvm.masked.store.nxv16i8.p0(<vscale x 16 x i8> [[DATA:%.*]], ptr [[TMP2]], i32 1, <vscale x 16 x i1> [[PG:%.*]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv16i8.p0(<vscale x 16 x i8> [[DATA:%.*]], ptr align 1 [[TMP2]], <vscale x 16 x i1> [[PG:%.*]]) // CHECK-NEXT: ret void // // CPP-CHECK-LABEL: @_Z19test_svst1_vnum_mf8u10__SVBool_tPu6__mfp8lu13__SVMfloat8_t( @@ -506,7 +506,7 @@ void test_svst1_vnum_f64(svbool_t pg, float64_t *base, int64_t vnum, svfloat64_t // CPP-CHECK-NEXT: [[TMP1:%.*]] = shl nuw nsw i64 [[TMP0]], 4 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP1]] // CPP-CHECK-NEXT: [[TMP2:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv16i8.p0(<vscale x 16 x i8> [[DATA:%.*]], ptr [[TMP2]], i32 1, <vscale x 16 x i1> [[PG:%.*]]) +// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv16i8.p0(<vscale x 16 x i8> [[DATA:%.*]], ptr align 1 [[TMP2]], <vscale x 16 x i1> [[PG:%.*]]) // CPP-CHECK-NEXT: ret void // void test_svst1_vnum_mf8(svbool_t pg, mfloat8_t *base, int64_t vnum, svmfloat8_t data) MODE_ATTR @@ -1247,13 +1247,13 @@ void test_svst1_scatter_u64base_index_f64(svbool_t pg, svuint64_t bases, int64_t // CHECK-LABEL: @test_svst1_bf16( // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]]) -// CHECK-NEXT: tail call void @llvm.masked.store.nxv8bf16.p0(<vscale x 8 x bfloat> [[DATA:%.*]], ptr [[BASE:%.*]], i32 1, <vscale x 8 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv8bf16.p0(<vscale x 8 x bfloat> [[DATA:%.*]], ptr align 1 [[BASE:%.*]], <vscale x 8 x i1> [[TMP0]]) // CHECK-NEXT: ret void // // CPP-CHECK-LABEL: @_Z15test_svst1_bf16u10__SVBool_tPu6__bf16u14__SVBfloat16_t( // CPP-CHECK-NEXT: entry: // CPP-CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]]) -// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv8bf16.p0(<vscale x 8 x bfloat> [[DATA:%.*]], ptr [[BASE:%.*]], i32 1, <vscale x 8 x i1> [[TMP0]]) +// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv8bf16.p0(<vscale x 8 x bfloat> [[DATA:%.*]], ptr align 1 [[BASE:%.*]], <vscale x 8 x i1> [[TMP0]]) // CPP-CHECK-NEXT: ret void // void test_svst1_bf16(svbool_t pg, bfloat16_t *base, svbfloat16_t data) MODE_ATTR @@ -1268,7 +1268,7 @@ void test_svst1_bf16(svbool_t pg, bfloat16_t *base, svbfloat16_t data) MODE_ATTR // CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CHECK-NEXT: tail call void @llvm.masked.store.nxv8bf16.p0(<vscale x 8 x bfloat> [[DATA:%.*]], ptr [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv8bf16.p0(<vscale x 8 x bfloat> [[DATA:%.*]], ptr align 1 [[TMP3]], <vscale x 8 x i1> [[TMP0]]) // CHECK-NEXT: ret void // // CPP-CHECK-LABEL: @_Z20test_svst1_vnum_bf16u10__SVBool_tPu6__bf16lu14__SVBfloat16_t( @@ -1278,7 +1278,7 @@ void test_svst1_bf16(svbool_t pg, bfloat16_t *base, svbfloat16_t data) MODE_ATTR // CPP-CHECK-NEXT: [[TMP2:%.*]] = shl nuw nsw i64 [[TMP1]], 4 // CPP-CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CPP-CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] -// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv8bf16.p0(<vscale x 8 x bfloat> [[DATA:%.*]], ptr [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]]) +// CPP-CHECK-NEXT: tail call void @llvm.masked.store.nxv8bf16.p0(<vscale x 8 x bfloat> [[DATA:%.*]], ptr align 1 [[TMP3]], <vscale x 8 x i1> [[TMP0]]) // CPP-CHECK-NEXT: ret void // void test_svst1_vnum_bf16(svbool_t pg, bfloat16_t *base, int64_t vnum, svbfloat16_t data) MODE_ATTR diff --git a/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_st1b.c b/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_st1b.c index c908bc2..3ac49e2 100644 --- a/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_st1b.c +++ b/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_st1b.c @@ -24,7 +24,7 @@ // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]]) // CHECK-NEXT: [[TMP1:%.*]] = trunc <vscale x 8 x i16> [[DATA:%.*]] to <vscale x 8 x i8> -// CHECK-NEXT: tail call void @llvm.masked.store.nxv8i8.p0(<vscale x 8 x i8> [[TMP1]], ptr [[BASE:%.*]], i32 1, <vscale x 8 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv8i8.p0(<vscale x 8 x i8> [[TMP1]], ptr align 1 [[BASE:%.*]], <vscale x 8 x i1> [[TMP0]]) // CHECK-NEXT: ret void // void test_svst1b_s16(svbool_t pg, int8_t *base, svint16_t data) MODE_ATTR @@ -36,7 +36,7 @@ void test_svst1b_s16(svbool_t pg, int8_t *base, svint16_t data) MODE_ATTR // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]]) // CHECK-NEXT: [[TMP1:%.*]] = trunc <vscale x 4 x i32> [[DATA:%.*]] to <vscale x 4 x i8> -// CHECK-NEXT: tail call void @llvm.masked.store.nxv4i8.p0(<vscale x 4 x i8> [[TMP1]], ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv4i8.p0(<vscale x 4 x i8> [[TMP1]], ptr align 1 [[BASE:%.*]], <vscale x 4 x i1> [[TMP0]]) // CHECK-NEXT: ret void // void test_svst1b_s32(svbool_t pg, int8_t *base, svint32_t data) MODE_ATTR @@ -48,7 +48,7 @@ void test_svst1b_s32(svbool_t pg, int8_t *base, svint32_t data) MODE_ATTR // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) // CHECK-NEXT: [[TMP1:%.*]] = trunc <vscale x 2 x i64> [[DATA:%.*]] to <vscale x 2 x i8> -// CHECK-NEXT: tail call void @llvm.masked.store.nxv2i8.p0(<vscale x 2 x i8> [[TMP1]], ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv2i8.p0(<vscale x 2 x i8> [[TMP1]], ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]]) // CHECK-NEXT: ret void // void test_svst1b_s64(svbool_t pg, int8_t *base, svint64_t data) MODE_ATTR @@ -60,7 +60,7 @@ void test_svst1b_s64(svbool_t pg, int8_t *base, svint64_t data) MODE_ATTR // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 8 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv8i1(<vscale x 16 x i1> [[PG:%.*]]) // CHECK-NEXT: [[TMP1:%.*]] = trunc <vscale x 8 x i16> [[DATA:%.*]] to <vscale x 8 x i8> -// CHECK-NEXT: tail call void @llvm.masked.store.nxv8i8.p0(<vscale x 8 x i8> [[TMP1]], ptr [[BASE:%.*]], i32 1, <vscale x 8 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv8i8.p0(<vscale x 8 x i8> [[TMP1]], ptr align 1 [[BASE:%.*]], <vscale x 8 x i1> [[TMP0]]) // CHECK-NEXT: ret void // void test_svst1b_u16(svbool_t pg, uint8_t *base, svuint16_t data) MODE_ATTR @@ -72,7 +72,7 @@ void test_svst1b_u16(svbool_t pg, uint8_t *base, svuint16_t data) MODE_ATTR // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]]) // CHECK-NEXT: [[TMP1:%.*]] = trunc <vscale x 4 x i32> [[DATA:%.*]] to <vscale x 4 x i8> -// CHECK-NEXT: tail call void @llvm.masked.store.nxv4i8.p0(<vscale x 4 x i8> [[TMP1]], ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv4i8.p0(<vscale x 4 x i8> [[TMP1]], ptr align 1 [[BASE:%.*]], <vscale x 4 x i1> [[TMP0]]) // CHECK-NEXT: ret void // void test_svst1b_u32(svbool_t pg, uint8_t *base, svuint32_t data) MODE_ATTR @@ -84,7 +84,7 @@ void test_svst1b_u32(svbool_t pg, uint8_t *base, svuint32_t data) MODE_ATTR // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) // CHECK-NEXT: [[TMP1:%.*]] = trunc <vscale x 2 x i64> [[DATA:%.*]] to <vscale x 2 x i8> -// CHECK-NEXT: tail call void @llvm.masked.store.nxv2i8.p0(<vscale x 2 x i8> [[TMP1]], ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv2i8.p0(<vscale x 2 x i8> [[TMP1]], ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]]) // CHECK-NEXT: ret void // void test_svst1b_u64(svbool_t pg, uint8_t *base, svuint64_t data) MODE_ATTR @@ -100,7 +100,7 @@ void test_svst1b_u64(svbool_t pg, uint8_t *base, svuint64_t data) MODE_ATTR // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] // CHECK-NEXT: [[TMP4:%.*]] = trunc <vscale x 8 x i16> [[DATA:%.*]] to <vscale x 8 x i8> -// CHECK-NEXT: tail call void @llvm.masked.store.nxv8i8.p0(<vscale x 8 x i8> [[TMP4]], ptr [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv8i8.p0(<vscale x 8 x i8> [[TMP4]], ptr align 1 [[TMP3]], <vscale x 8 x i1> [[TMP0]]) // CHECK-NEXT: ret void // void test_svst1b_vnum_s16(svbool_t pg, int8_t *base, int64_t vnum, svint16_t data) MODE_ATTR @@ -116,7 +116,7 @@ void test_svst1b_vnum_s16(svbool_t pg, int8_t *base, int64_t vnum, svint16_t dat // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] // CHECK-NEXT: [[TMP4:%.*]] = trunc <vscale x 4 x i32> [[DATA:%.*]] to <vscale x 4 x i8> -// CHECK-NEXT: tail call void @llvm.masked.store.nxv4i8.p0(<vscale x 4 x i8> [[TMP4]], ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv4i8.p0(<vscale x 4 x i8> [[TMP4]], ptr align 1 [[TMP3]], <vscale x 4 x i1> [[TMP0]]) // CHECK-NEXT: ret void // void test_svst1b_vnum_s32(svbool_t pg, int8_t *base, int64_t vnum, svint32_t data) MODE_ATTR @@ -132,7 +132,7 @@ void test_svst1b_vnum_s32(svbool_t pg, int8_t *base, int64_t vnum, svint32_t dat // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] // CHECK-NEXT: [[TMP4:%.*]] = trunc <vscale x 2 x i64> [[DATA:%.*]] to <vscale x 2 x i8> -// CHECK-NEXT: tail call void @llvm.masked.store.nxv2i8.p0(<vscale x 2 x i8> [[TMP4]], ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv2i8.p0(<vscale x 2 x i8> [[TMP4]], ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]]) // CHECK-NEXT: ret void // void test_svst1b_vnum_s64(svbool_t pg, int8_t *base, int64_t vnum, svint64_t data) MODE_ATTR @@ -148,7 +148,7 @@ void test_svst1b_vnum_s64(svbool_t pg, int8_t *base, int64_t vnum, svint64_t dat // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] // CHECK-NEXT: [[TMP4:%.*]] = trunc <vscale x 8 x i16> [[DATA:%.*]] to <vscale x 8 x i8> -// CHECK-NEXT: tail call void @llvm.masked.store.nxv8i8.p0(<vscale x 8 x i8> [[TMP4]], ptr [[TMP3]], i32 1, <vscale x 8 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv8i8.p0(<vscale x 8 x i8> [[TMP4]], ptr align 1 [[TMP3]], <vscale x 8 x i1> [[TMP0]]) // CHECK-NEXT: ret void // void test_svst1b_vnum_u16(svbool_t pg, uint8_t *base, int64_t vnum, svuint16_t data) MODE_ATTR @@ -164,7 +164,7 @@ void test_svst1b_vnum_u16(svbool_t pg, uint8_t *base, int64_t vnum, svuint16_t d // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] // CHECK-NEXT: [[TMP4:%.*]] = trunc <vscale x 4 x i32> [[DATA:%.*]] to <vscale x 4 x i8> -// CHECK-NEXT: tail call void @llvm.masked.store.nxv4i8.p0(<vscale x 4 x i8> [[TMP4]], ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv4i8.p0(<vscale x 4 x i8> [[TMP4]], ptr align 1 [[TMP3]], <vscale x 4 x i1> [[TMP0]]) // CHECK-NEXT: ret void // void test_svst1b_vnum_u32(svbool_t pg, uint8_t *base, int64_t vnum, svuint32_t data) MODE_ATTR @@ -180,7 +180,7 @@ void test_svst1b_vnum_u32(svbool_t pg, uint8_t *base, int64_t vnum, svuint32_t d // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] // CHECK-NEXT: [[TMP4:%.*]] = trunc <vscale x 2 x i64> [[DATA:%.*]] to <vscale x 2 x i8> -// CHECK-NEXT: tail call void @llvm.masked.store.nxv2i8.p0(<vscale x 2 x i8> [[TMP4]], ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv2i8.p0(<vscale x 2 x i8> [[TMP4]], ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]]) // CHECK-NEXT: ret void // void test_svst1b_vnum_u64(svbool_t pg, uint8_t *base, int64_t vnum, svuint64_t data) MODE_ATTR diff --git a/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_st1h.c b/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_st1h.c index 959b658..1e3e0b2 100644 --- a/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_st1h.c +++ b/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_st1h.c @@ -24,7 +24,7 @@ // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]]) // CHECK-NEXT: [[TMP1:%.*]] = trunc <vscale x 4 x i32> [[DATA:%.*]] to <vscale x 4 x i16> -// CHECK-NEXT: tail call void @llvm.masked.store.nxv4i16.p0(<vscale x 4 x i16> [[TMP1]], ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv4i16.p0(<vscale x 4 x i16> [[TMP1]], ptr align 1 [[BASE:%.*]], <vscale x 4 x i1> [[TMP0]]) // CHECK-NEXT: ret void // void test_svst1h_s32(svbool_t pg, int16_t *base, svint32_t data) MODE_ATTR @@ -36,7 +36,7 @@ void test_svst1h_s32(svbool_t pg, int16_t *base, svint32_t data) MODE_ATTR // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) // CHECK-NEXT: [[TMP1:%.*]] = trunc <vscale x 2 x i64> [[DATA:%.*]] to <vscale x 2 x i16> -// CHECK-NEXT: tail call void @llvm.masked.store.nxv2i16.p0(<vscale x 2 x i16> [[TMP1]], ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv2i16.p0(<vscale x 2 x i16> [[TMP1]], ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]]) // CHECK-NEXT: ret void // void test_svst1h_s64(svbool_t pg, int16_t *base, svint64_t data) MODE_ATTR @@ -48,7 +48,7 @@ void test_svst1h_s64(svbool_t pg, int16_t *base, svint64_t data) MODE_ATTR // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PG:%.*]]) // CHECK-NEXT: [[TMP1:%.*]] = trunc <vscale x 4 x i32> [[DATA:%.*]] to <vscale x 4 x i16> -// CHECK-NEXT: tail call void @llvm.masked.store.nxv4i16.p0(<vscale x 4 x i16> [[TMP1]], ptr [[BASE:%.*]], i32 1, <vscale x 4 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv4i16.p0(<vscale x 4 x i16> [[TMP1]], ptr align 1 [[BASE:%.*]], <vscale x 4 x i1> [[TMP0]]) // CHECK-NEXT: ret void // void test_svst1h_u32(svbool_t pg, uint16_t *base, svuint32_t data) MODE_ATTR @@ -60,7 +60,7 @@ void test_svst1h_u32(svbool_t pg, uint16_t *base, svuint32_t data) MODE_ATTR // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) // CHECK-NEXT: [[TMP1:%.*]] = trunc <vscale x 2 x i64> [[DATA:%.*]] to <vscale x 2 x i16> -// CHECK-NEXT: tail call void @llvm.masked.store.nxv2i16.p0(<vscale x 2 x i16> [[TMP1]], ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv2i16.p0(<vscale x 2 x i16> [[TMP1]], ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]]) // CHECK-NEXT: ret void // void test_svst1h_u64(svbool_t pg, uint16_t *base, svuint64_t data) MODE_ATTR @@ -76,7 +76,7 @@ void test_svst1h_u64(svbool_t pg, uint16_t *base, svuint64_t data) MODE_ATTR // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] // CHECK-NEXT: [[TMP4:%.*]] = trunc <vscale x 4 x i32> [[DATA:%.*]] to <vscale x 4 x i16> -// CHECK-NEXT: tail call void @llvm.masked.store.nxv4i16.p0(<vscale x 4 x i16> [[TMP4]], ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv4i16.p0(<vscale x 4 x i16> [[TMP4]], ptr align 1 [[TMP3]], <vscale x 4 x i1> [[TMP0]]) // CHECK-NEXT: ret void // void test_svst1h_vnum_s32(svbool_t pg, int16_t *base, int64_t vnum, svint32_t data) MODE_ATTR @@ -92,7 +92,7 @@ void test_svst1h_vnum_s32(svbool_t pg, int16_t *base, int64_t vnum, svint32_t da // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] // CHECK-NEXT: [[TMP4:%.*]] = trunc <vscale x 2 x i64> [[DATA:%.*]] to <vscale x 2 x i16> -// CHECK-NEXT: tail call void @llvm.masked.store.nxv2i16.p0(<vscale x 2 x i16> [[TMP4]], ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv2i16.p0(<vscale x 2 x i16> [[TMP4]], ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]]) // CHECK-NEXT: ret void // void test_svst1h_vnum_s64(svbool_t pg, int16_t *base, int64_t vnum, svint64_t data) MODE_ATTR @@ -108,7 +108,7 @@ void test_svst1h_vnum_s64(svbool_t pg, int16_t *base, int64_t vnum, svint64_t da // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] // CHECK-NEXT: [[TMP4:%.*]] = trunc <vscale x 4 x i32> [[DATA:%.*]] to <vscale x 4 x i16> -// CHECK-NEXT: tail call void @llvm.masked.store.nxv4i16.p0(<vscale x 4 x i16> [[TMP4]], ptr [[TMP3]], i32 1, <vscale x 4 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv4i16.p0(<vscale x 4 x i16> [[TMP4]], ptr align 1 [[TMP3]], <vscale x 4 x i1> [[TMP0]]) // CHECK-NEXT: ret void // void test_svst1h_vnum_u32(svbool_t pg, uint16_t *base, int64_t vnum, svuint32_t data) MODE_ATTR @@ -124,7 +124,7 @@ void test_svst1h_vnum_u32(svbool_t pg, uint16_t *base, int64_t vnum, svuint32_t // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] // CHECK-NEXT: [[TMP4:%.*]] = trunc <vscale x 2 x i64> [[DATA:%.*]] to <vscale x 2 x i16> -// CHECK-NEXT: tail call void @llvm.masked.store.nxv2i16.p0(<vscale x 2 x i16> [[TMP4]], ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv2i16.p0(<vscale x 2 x i16> [[TMP4]], ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]]) // CHECK-NEXT: ret void // void test_svst1h_vnum_u64(svbool_t pg, uint16_t *base, int64_t vnum, svuint64_t data) MODE_ATTR diff --git a/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_st1w.c b/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_st1w.c index 3d9e45bd..1a12412 100644 --- a/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_st1w.c +++ b/clang/test/CodeGen/AArch64/sve-intrinsics/acle_sve_st1w.c @@ -24,7 +24,7 @@ // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) // CHECK-NEXT: [[TMP1:%.*]] = trunc <vscale x 2 x i64> [[DATA:%.*]] to <vscale x 2 x i32> -// CHECK-NEXT: tail call void @llvm.masked.store.nxv2i32.p0(<vscale x 2 x i32> [[TMP1]], ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv2i32.p0(<vscale x 2 x i32> [[TMP1]], ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]]) // CHECK-NEXT: ret void // void test_svst1w_s64(svbool_t pg, int32_t *base, svint64_t data) MODE_ATTR @@ -36,7 +36,7 @@ void test_svst1w_s64(svbool_t pg, int32_t *base, svint64_t data) MODE_ATTR // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = tail call <vscale x 2 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv2i1(<vscale x 16 x i1> [[PG:%.*]]) // CHECK-NEXT: [[TMP1:%.*]] = trunc <vscale x 2 x i64> [[DATA:%.*]] to <vscale x 2 x i32> -// CHECK-NEXT: tail call void @llvm.masked.store.nxv2i32.p0(<vscale x 2 x i32> [[TMP1]], ptr [[BASE:%.*]], i32 1, <vscale x 2 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv2i32.p0(<vscale x 2 x i32> [[TMP1]], ptr align 1 [[BASE:%.*]], <vscale x 2 x i1> [[TMP0]]) // CHECK-NEXT: ret void // void test_svst1w_u64(svbool_t pg, uint32_t *base, svuint64_t data) MODE_ATTR @@ -52,7 +52,7 @@ void test_svst1w_u64(svbool_t pg, uint32_t *base, svuint64_t data) MODE_ATTR // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] // CHECK-NEXT: [[TMP4:%.*]] = trunc <vscale x 2 x i64> [[DATA:%.*]] to <vscale x 2 x i32> -// CHECK-NEXT: tail call void @llvm.masked.store.nxv2i32.p0(<vscale x 2 x i32> [[TMP4]], ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv2i32.p0(<vscale x 2 x i32> [[TMP4]], ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]]) // CHECK-NEXT: ret void // void test_svst1w_vnum_s64(svbool_t pg, int32_t *base, int64_t vnum, svint64_t data) MODE_ATTR @@ -68,7 +68,7 @@ void test_svst1w_vnum_s64(svbool_t pg, int32_t *base, int64_t vnum, svint64_t da // CHECK-NEXT: [[DOTIDX:%.*]] = mul i64 [[VNUM:%.*]], [[TMP2]] // CHECK-NEXT: [[TMP3:%.*]] = getelementptr i8, ptr [[BASE:%.*]], i64 [[DOTIDX]] // CHECK-NEXT: [[TMP4:%.*]] = trunc <vscale x 2 x i64> [[DATA:%.*]] to <vscale x 2 x i32> -// CHECK-NEXT: tail call void @llvm.masked.store.nxv2i32.p0(<vscale x 2 x i32> [[TMP4]], ptr [[TMP3]], i32 1, <vscale x 2 x i1> [[TMP0]]) +// CHECK-NEXT: tail call void @llvm.masked.store.nxv2i32.p0(<vscale x 2 x i32> [[TMP4]], ptr align 1 [[TMP3]], <vscale x 2 x i1> [[TMP0]]) // CHECK-NEXT: ret void // void test_svst1w_vnum_u64(svbool_t pg, uint32_t *base, int64_t vnum, svuint64_t data) MODE_ATTR diff --git a/clang/test/CodeGen/X86/avx10_2bf16-builtins.c b/clang/test/CodeGen/X86/avx10_2bf16-builtins.c index c7fea07..f8a4c51 100644 --- a/clang/test/CodeGen/X86/avx10_2bf16-builtins.c +++ b/clang/test/CodeGen/X86/avx10_2bf16-builtins.c @@ -274,7 +274,7 @@ __m256bh test_mm256_loadu_pbh(void *p) { __m128bh test_mm_load_sbh(void const *A) { // CHECK-LABEL: test_mm_load_sbh - // CHECK: %{{.*}} = call <8 x bfloat> @llvm.masked.load.v8bf16.p0(ptr %{{.*}}, i32 1, <8 x i1> bitcast (<1 x i8> splat (i8 1) to <8 x i1>), <8 x bfloat> %{{.*}}) + // CHECK: %{{.*}} = call <8 x bfloat> @llvm.masked.load.v8bf16.p0(ptr align 1 %{{.*}}, <8 x i1> bitcast (<1 x i8> splat (i8 1) to <8 x i1>), <8 x bfloat> %{{.*}}) return _mm_load_sbh(A); } @@ -305,7 +305,7 @@ void test_mm_store_sbh(void *A, __m128bh B) { void test_mm_mask_store_sbh(void *__P, __mmask8 __U, __m128bh __A) { // CHECK-LABEL: @test_mm_mask_store_sbh - // CHECK: call void @llvm.masked.store.v8bf16.p0(<8 x bfloat> %{{.*}}, ptr %{{.*}}, i32 1, <8 x i1> %{{.*}}) + // CHECK: call void @llvm.masked.store.v8bf16.p0(<8 x bfloat> %{{.*}}, ptr align 1 %{{.*}}, <8 x i1> %{{.*}}) _mm_mask_store_sbh(__P, __U, __A); } @@ -323,13 +323,13 @@ void test_mm_store_pbh(void *p, __m128bh a) { __m128bh test_mm_mask_load_sbh(__m128bh __A, __mmask8 __U, const void *__W) { // CHECK-LABEL: @test_mm_mask_load_sbh - // CHECK: %{{.*}} = call <8 x bfloat> @llvm.masked.load.v8bf16.p0(ptr %{{.*}}, i32 1, <8 x i1> %{{.*}}, <8 x bfloat> %{{.*}}) + // CHECK: %{{.*}} = call <8 x bfloat> @llvm.masked.load.v8bf16.p0(ptr align 1 %{{.*}}, <8 x i1> %{{.*}}, <8 x bfloat> %{{.*}}) return _mm_mask_load_sbh(__A, __U, __W); } __m128bh test_mm_maskz_load_sbh(__mmask8 __U, const void *__W) { // CHECK-LABEL: @test_mm_maskz_load_sbh - // CHECK: %{{.*}} = call <8 x bfloat> @llvm.masked.load.v8bf16.p0(ptr %{{.*}}, i32 1, <8 x i1> %{{.*}}, <8 x bfloat> %{{.*}}) + // CHECK: %{{.*}} = call <8 x bfloat> @llvm.masked.load.v8bf16.p0(ptr align 1 %{{.*}}, <8 x i1> %{{.*}}, <8 x bfloat> %{{.*}}) return _mm_maskz_load_sbh(__U, __W); } diff --git a/clang/test/CodeGen/X86/avx512bw-builtins.c b/clang/test/CodeGen/X86/avx512bw-builtins.c index 55bf482..d07e40a 100644 --- a/clang/test/CodeGen/X86/avx512bw-builtins.c +++ b/clang/test/CodeGen/X86/avx512bw-builtins.c @@ -2414,13 +2414,13 @@ __m512i test_mm512_loadu_epi16 (void *__P) __m512i test_mm512_mask_loadu_epi16(__m512i __W, __mmask32 __U, void const *__P) { // CHECK-LABEL: test_mm512_mask_loadu_epi16 - // CHECK: @llvm.masked.load.v32i16.p0(ptr %{{.*}}, i32 1, <32 x i1> %{{.*}}, <32 x i16> %{{.*}}) + // CHECK: @llvm.masked.load.v32i16.p0(ptr align 1 %{{.*}}, <32 x i1> %{{.*}}, <32 x i16> %{{.*}}) return _mm512_mask_loadu_epi16(__W, __U, __P); } __m512i test_mm512_maskz_loadu_epi16(__mmask32 __U, void const *__P) { // CHECK-LABEL: test_mm512_maskz_loadu_epi16 - // CHECK: @llvm.masked.load.v32i16.p0(ptr %{{.*}}, i32 1, <32 x i1> %{{.*}}, <32 x i16> %{{.*}}) + // CHECK: @llvm.masked.load.v32i16.p0(ptr align 1 %{{.*}}, <32 x i1> %{{.*}}, <32 x i16> %{{.*}}) return _mm512_maskz_loadu_epi16(__U, __P); } @@ -2433,13 +2433,13 @@ __m512i test_mm512_loadu_epi8 (void *__P) __m512i test_mm512_mask_loadu_epi8(__m512i __W, __mmask64 __U, void const *__P) { // CHECK-LABEL: test_mm512_mask_loadu_epi8 - // CHECK: @llvm.masked.load.v64i8.p0(ptr %{{.*}}, i32 1, <64 x i1> %{{.*}}, <64 x i8> %{{.*}}) + // CHECK: @llvm.masked.load.v64i8.p0(ptr align 1 %{{.*}}, <64 x i1> %{{.*}}, <64 x i8> %{{.*}}) return _mm512_mask_loadu_epi8(__W, __U, __P); } __m512i test_mm512_maskz_loadu_epi8(__mmask64 __U, void const *__P) { // CHECK-LABEL: test_mm512_maskz_loadu_epi8 - // CHECK: @llvm.masked.load.v64i8.p0(ptr %{{.*}}, i32 1, <64 x i1> %{{.*}}, <64 x i8> %{{.*}}) + // CHECK: @llvm.masked.load.v64i8.p0(ptr align 1 %{{.*}}, <64 x i1> %{{.*}}, <64 x i8> %{{.*}}) return _mm512_maskz_loadu_epi8(__U, __P); } @@ -2451,7 +2451,7 @@ void test_mm512_storeu_epi16(void *__P, __m512i __A) { void test_mm512_mask_storeu_epi16(void *__P, __mmask32 __U, __m512i __A) { // CHECK-LABEL: test_mm512_mask_storeu_epi16 - // CHECK: @llvm.masked.store.v32i16.p0(<32 x i16> %{{.*}}, ptr %{{.*}}, i32 1, <32 x i1> %{{.*}}) + // CHECK: @llvm.masked.store.v32i16.p0(<32 x i16> %{{.*}}, ptr align 1 %{{.*}}, <32 x i1> %{{.*}}) return _mm512_mask_storeu_epi16(__P, __U, __A); } @@ -2470,7 +2470,7 @@ void test_mm512_storeu_epi8(void *__P, __m512i __A) { void test_mm512_mask_storeu_epi8(void *__P, __mmask64 __U, __m512i __A) { // CHECK-LABEL: test_mm512_mask_storeu_epi8 - // CHECK: @llvm.masked.store.v64i8.p0(<64 x i8> %{{.*}}, ptr %{{.*}}, i32 1, <64 x i1> %{{.*}}) + // CHECK: @llvm.masked.store.v64i8.p0(<64 x i8> %{{.*}}, ptr align 1 %{{.*}}, <64 x i1> %{{.*}}) return _mm512_mask_storeu_epi8(__P, __U, __A); } __mmask64 test_mm512_mask_test_epi8_mask(__mmask64 __U, __m512i __A, __m512i __B) { diff --git a/clang/test/CodeGen/X86/avx512f-builtins.c b/clang/test/CodeGen/X86/avx512f-builtins.c index 3deaf8e..6073fdf 100644 --- a/clang/test/CodeGen/X86/avx512f-builtins.c +++ b/clang/test/CodeGen/X86/avx512f-builtins.c @@ -212,7 +212,7 @@ void test_mm512_storeu_pd(void *p, __m512d a) void test_mm512_mask_store_ps(void *p, __m512 a, __mmask16 m) { // CHECK-LABEL: test_mm512_mask_store_ps - // CHECK: @llvm.masked.store.v16f32.p0(<16 x float> %{{.*}}, ptr %{{.*}}, i32 64, <16 x i1> %{{.*}}) + // CHECK: @llvm.masked.store.v16f32.p0(<16 x float> %{{.*}}, ptr align 64 %{{.*}}, <16 x i1> %{{.*}}) _mm512_mask_store_ps(p, m, a); } @@ -260,7 +260,7 @@ void test_mm512_store_pd(void *p, __m512d a) void test_mm512_mask_store_pd(void *p, __m512d a, __mmask8 m) { // CHECK-LABEL: test_mm512_mask_store_pd - // CHECK: @llvm.masked.store.v8f64.p0(<8 x double> %{{.*}}, ptr %{{.*}}, i32 64, <8 x i1> %{{.*}}) + // CHECK: @llvm.masked.store.v8f64.p0(<8 x double> %{{.*}}, ptr align 64 %{{.*}}, <8 x i1> %{{.*}}) _mm512_mask_store_pd(p, m, a); } @@ -272,7 +272,7 @@ void test_mm512_storeu_epi32(void *__P, __m512i __A) { void test_mm512_mask_storeu_epi32(void *__P, __mmask16 __U, __m512i __A) { // CHECK-LABEL: test_mm512_mask_storeu_epi32 - // CHECK: @llvm.masked.store.v16i32.p0(<16 x i32> %{{.*}}, ptr %{{.*}}, i32 1, <16 x i1> %{{.*}}) + // CHECK: @llvm.masked.store.v16i32.p0(<16 x i32> %{{.*}}, ptr align 1 %{{.*}}, <16 x i1> %{{.*}}) return _mm512_mask_storeu_epi32(__P, __U, __A); } @@ -284,7 +284,7 @@ void test_mm512_storeu_epi64(void *__P, __m512i __A) { void test_mm512_mask_storeu_epi64(void *__P, __mmask8 __U, __m512i __A) { // CHECK-LABEL: test_mm512_mask_storeu_epi64 - // CHECK: @llvm.masked.store.v8i64.p0(<8 x i64> %{{.*}}, ptr %{{.*}}, i32 1, <8 x i1> %{{.*}}) + // CHECK: @llvm.masked.store.v8i64.p0(<8 x i64> %{{.*}}, ptr align 1 %{{.*}}, <8 x i1> %{{.*}}) return _mm512_mask_storeu_epi64(__P, __U, __A); } @@ -305,14 +305,14 @@ __m512i test_mm512_loadu_epi32 (void *__P) __m512i test_mm512_mask_loadu_epi32 (__m512i __W, __mmask16 __U, void *__P) { // CHECK-LABEL: test_mm512_mask_loadu_epi32 - // CHECK: @llvm.masked.load.v16i32.p0(ptr %{{.*}}, i32 1, <16 x i1> %{{.*}}, <16 x i32> %{{.*}}) + // CHECK: @llvm.masked.load.v16i32.p0(ptr align 1 %{{.*}}, <16 x i1> %{{.*}}, <16 x i32> %{{.*}}) return _mm512_mask_loadu_epi32 (__W,__U, __P); } __m512i test_mm512_maskz_loadu_epi32 (__mmask16 __U, void *__P) { // CHECK-LABEL: test_mm512_maskz_loadu_epi32 - // CHECK: @llvm.masked.load.v16i32.p0(ptr %{{.*}}, i32 1, <16 x i1> %{{.*}}, <16 x i32> %{{.*}}) + // CHECK: @llvm.masked.load.v16i32.p0(ptr align 1 %{{.*}}, <16 x i1> %{{.*}}, <16 x i32> %{{.*}}) return _mm512_maskz_loadu_epi32 (__U, __P); } @@ -326,14 +326,14 @@ __m512i test_mm512_loadu_epi64 (void *__P) __m512i test_mm512_mask_loadu_epi64 (__m512i __W, __mmask8 __U, void *__P) { // CHECK-LABEL: test_mm512_mask_loadu_epi64 - // CHECK: @llvm.masked.load.v8i64.p0(ptr %{{.*}}, i32 1, <8 x i1> %{{.*}}, <8 x i64> %{{.*}}) + // CHECK: @llvm.masked.load.v8i64.p0(ptr align 1 %{{.*}}, <8 x i1> %{{.*}}, <8 x i64> %{{.*}}) return _mm512_mask_loadu_epi64 (__W,__U, __P); } __m512i test_mm512_maskz_loadu_epi64 (__mmask16 __U, void *__P) { // CHECK-LABEL: test_mm512_maskz_loadu_epi64 - // CHECK: @llvm.masked.load.v8i64.p0(ptr %{{.*}}, i32 1, <8 x i1> %{{.*}}, <8 x i64> %{{.*}}) + // CHECK: @llvm.masked.load.v8i64.p0(ptr align 1 %{{.*}}, <8 x i1> %{{.*}}, <8 x i64> %{{.*}}) return _mm512_maskz_loadu_epi64 (__U, __P); } @@ -347,7 +347,7 @@ __m512 test_mm512_loadu_ps(void *p) __m512 test_mm512_mask_loadu_ps (__m512 __W, __mmask16 __U, void *__P) { // CHECK-LABEL: test_mm512_mask_loadu_ps - // CHECK: @llvm.masked.load.v16f32.p0(ptr %{{.*}}, i32 1, <16 x i1> %{{.*}}, <16 x float> %{{.*}}) + // CHECK: @llvm.masked.load.v16f32.p0(ptr align 1 %{{.*}}, <16 x i1> %{{.*}}, <16 x float> %{{.*}}) return _mm512_mask_loadu_ps (__W,__U, __P); } @@ -361,7 +361,7 @@ __m512d test_mm512_loadu_pd(void *p) __m512d test_mm512_mask_loadu_pd (__m512d __W, __mmask8 __U, void *__P) { // CHECK-LABEL: test_mm512_mask_loadu_pd - // CHECK: @llvm.masked.load.v8f64.p0(ptr %{{.*}}, i32 1, <8 x i1> %{{.*}}, <8 x double> %{{.*}}) + // CHECK: @llvm.masked.load.v8f64.p0(ptr align 1 %{{.*}}, <8 x i1> %{{.*}}, <8 x double> %{{.*}}) return _mm512_mask_loadu_pd (__W,__U, __P); } @@ -399,14 +399,14 @@ __m512 test_mm512_load_ps(void *p) __m512 test_mm512_mask_load_ps (__m512 __W, __mmask16 __U, void *__P) { // CHECK-LABEL: test_mm512_mask_load_ps - // CHECK: @llvm.masked.load.v16f32.p0(ptr %{{.*}}, i32 64, <16 x i1> %{{.*}}, <16 x float> %{{.*}}) + // CHECK: @llvm.masked.load.v16f32.p0(ptr align 64 %{{.*}}, <16 x i1> %{{.*}}, <16 x float> %{{.*}}) return _mm512_mask_load_ps (__W,__U, __P); } __m512 test_mm512_maskz_load_ps(__mmask16 __U, void *__P) { // CHECK-LABEL: test_mm512_maskz_load_ps - // CHECK: @llvm.masked.load.v16f32.p0(ptr %{{.*}}, i32 64, <16 x i1> %{{.*}}, <16 x float> %{{.*}}) + // CHECK: @llvm.masked.load.v16f32.p0(ptr align 64 %{{.*}}, <16 x i1> %{{.*}}, <16 x float> %{{.*}}) return _mm512_maskz_load_ps(__U, __P); } @@ -420,14 +420,14 @@ __m512d test_mm512_load_pd(void *p) __m512d test_mm512_mask_load_pd (__m512d __W, __mmask8 __U, void *__P) { // CHECK-LABEL: test_mm512_mask_load_pd - // CHECK: @llvm.masked.load.v8f64.p0(ptr %{{.*}}, i32 64, <8 x i1> %{{.*}}, <8 x double> %{{.*}}) + // CHECK: @llvm.masked.load.v8f64.p0(ptr align 64 %{{.*}}, <8 x i1> %{{.*}}, <8 x double> %{{.*}}) return _mm512_mask_load_pd (__W,__U, __P); } __m512d test_mm512_maskz_load_pd(__mmask8 __U, void *__P) { // CHECK-LABEL: test_mm512_maskz_load_pd - // CHECK: @llvm.masked.load.v8f64.p0(ptr %{{.*}}, i32 64, <8 x i1> %{{.*}}, <8 x double> %{{.*}}) + // CHECK: @llvm.masked.load.v8f64.p0(ptr align 64 %{{.*}}, <8 x i1> %{{.*}}, <8 x double> %{{.*}}) return _mm512_maskz_load_pd(__U, __P); } @@ -4560,13 +4560,13 @@ __m512i test_mm512_maskz_srli_epi64_2(__mmask8 __U, __m512i __A, unsigned int __ __m512i test_mm512_mask_load_epi32(__m512i __W, __mmask16 __U, void const *__P) { // CHECK-LABEL: test_mm512_mask_load_epi32 - // CHECK: @llvm.masked.load.v16i32.p0(ptr %{{.*}}, i32 64, <16 x i1> %{{.*}}, <16 x i32> %{{.*}}) + // CHECK: @llvm.masked.load.v16i32.p0(ptr align 64 %{{.*}}, <16 x i1> %{{.*}}, <16 x i32> %{{.*}}) return _mm512_mask_load_epi32(__W, __U, __P); } __m512i test_mm512_maskz_load_epi32(__mmask16 __U, void const *__P) { // CHECK-LABEL: test_mm512_maskz_load_epi32 - // CHECK: @llvm.masked.load.v16i32.p0(ptr %{{.*}}, i32 64, <16 x i1> %{{.*}}, <16 x i32> %{{.*}}) + // CHECK: @llvm.masked.load.v16i32.p0(ptr align 64 %{{.*}}, <16 x i1> %{{.*}}, <16 x i32> %{{.*}}) return _mm512_maskz_load_epi32(__U, __P); } @@ -4596,25 +4596,25 @@ __m512i test_mm512_maskz_mov_epi64(__mmask8 __U, __m512i __A) { __m512i test_mm512_mask_load_epi64(__m512i __W, __mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm512_mask_load_epi64 - // CHECK: @llvm.masked.load.v8i64.p0(ptr %{{.*}}, i32 64, <8 x i1> %{{.*}}, <8 x i64> %{{.*}}) + // CHECK: @llvm.masked.load.v8i64.p0(ptr align 64 %{{.*}}, <8 x i1> %{{.*}}, <8 x i64> %{{.*}}) return _mm512_mask_load_epi64(__W, __U, __P); } __m512i test_mm512_maskz_load_epi64(__mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm512_maskz_load_epi64 - // CHECK: @llvm.masked.load.v8i64.p0(ptr %{{.*}}, i32 64, <8 x i1> %{{.*}}, <8 x i64> %{{.*}}) + // CHECK: @llvm.masked.load.v8i64.p0(ptr align 64 %{{.*}}, <8 x i1> %{{.*}}, <8 x i64> %{{.*}}) return _mm512_maskz_load_epi64(__U, __P); } void test_mm512_mask_store_epi32(void *__P, __mmask16 __U, __m512i __A) { // CHECK-LABEL: test_mm512_mask_store_epi32 - // CHECK: @llvm.masked.store.v16i32.p0(<16 x i32> %{{.*}}, ptr %{{.*}}, i32 64, <16 x i1> %{{.*}}) + // CHECK: @llvm.masked.store.v16i32.p0(<16 x i32> %{{.*}}, ptr align 64 %{{.*}}, <16 x i1> %{{.*}}) return _mm512_mask_store_epi32(__P, __U, __A); } void test_mm512_mask_store_epi64(void *__P, __mmask8 __U, __m512i __A) { // CHECK-LABEL: test_mm512_mask_store_epi64 - // CHECK: @llvm.masked.store.v8i64.p0(<8 x i64> %{{.*}}, ptr %{{.*}}, i32 64, <8 x i1> %{{.*}}) + // CHECK: @llvm.masked.store.v8i64.p0(<8 x i64> %{{.*}}, ptr align 64 %{{.*}}, <8 x i1> %{{.*}}) return _mm512_mask_store_epi64(__P, __U, __A); } @@ -11572,42 +11572,42 @@ __m128d test_mm_maskz_move_sd (__mmask8 __U, __m128d __A, __m128d __B) void test_mm_mask_store_ss(float * __P, __mmask8 __U, __m128 __A) { // CHECK-LABEL: test_mm_mask_store_ss - // CHECK: call void @llvm.masked.store.v4f32.p0(<4 x float> %{{.*}}, ptr %{{.*}}, i32 1, <4 x i1> %{{.*}}) + // CHECK: call void @llvm.masked.store.v4f32.p0(<4 x float> %{{.*}}, ptr align 1 %{{.*}}, <4 x i1> %{{.*}}) _mm_mask_store_ss(__P, __U, __A); } void test_mm_mask_store_sd(double * __P, __mmask8 __U, __m128d __A) { // CHECK-LABEL: test_mm_mask_store_sd - // CHECK: call void @llvm.masked.store.v2f64.p0(<2 x double> %{{.*}}, ptr %{{.*}}, i32 1, <2 x i1> %{{.*}}) + // CHECK: call void @llvm.masked.store.v2f64.p0(<2 x double> %{{.*}}, ptr align 1 %{{.*}}, <2 x i1> %{{.*}}) _mm_mask_store_sd(__P, __U, __A); } __m128 test_mm_mask_load_ss(__m128 __A, __mmask8 __U, const float* __W) { // CHECK-LABEL: test_mm_mask_load_ss - // CHECK: call {{.*}}<4 x float> @llvm.masked.load.v4f32.p0(ptr %{{.*}}, i32 1, <4 x i1> %{{.*}}, <4 x float> %{{.*}}) + // CHECK: call {{.*}}<4 x float> @llvm.masked.load.v4f32.p0(ptr align 1 %{{.*}}, <4 x i1> %{{.*}}, <4 x float> %{{.*}}) return _mm_mask_load_ss(__A, __U, __W); } __m128 test_mm_maskz_load_ss (__mmask8 __U, const float * __W) { // CHECK-LABEL: test_mm_maskz_load_ss - // CHECK: call {{.*}}<4 x float> @llvm.masked.load.v4f32.p0(ptr %{{.*}}, i32 1, <4 x i1> %{{.*}}, <4 x float> %{{.*}}) + // CHECK: call {{.*}}<4 x float> @llvm.masked.load.v4f32.p0(ptr align 1 %{{.*}}, <4 x i1> %{{.*}}, <4 x float> %{{.*}}) return _mm_maskz_load_ss (__U, __W); } __m128d test_mm_mask_load_sd (__m128d __A, __mmask8 __U, const double * __W) { // CHECK-LABEL: test_mm_mask_load_sd - // CHECK: call {{.*}}<2 x double> @llvm.masked.load.v2f64.p0(ptr %{{.*}}, i32 1, <2 x i1> %{{.*}}, <2 x double> %{{.*}}) + // CHECK: call {{.*}}<2 x double> @llvm.masked.load.v2f64.p0(ptr align 1 %{{.*}}, <2 x i1> %{{.*}}, <2 x double> %{{.*}}) return _mm_mask_load_sd (__A, __U, __W); } __m128d test_mm_maskz_load_sd (__mmask8 __U, const double * __W) { // CHECK-LABEL: test_mm_maskz_load_sd - // CHECK: call {{.*}}<2 x double> @llvm.masked.load.v2f64.p0(ptr %{{.*}}, i32 1, <2 x i1> %{{.*}}, <2 x double> %{{.*}}) + // CHECK: call {{.*}}<2 x double> @llvm.masked.load.v2f64.p0(ptr align 1 %{{.*}}, <2 x i1> %{{.*}}, <2 x double> %{{.*}}) return _mm_maskz_load_sd (__U, __W); } diff --git a/clang/test/CodeGen/X86/avx512fp16-builtins.c b/clang/test/CodeGen/X86/avx512fp16-builtins.c index 2befff0..f0a0a3b 100644 --- a/clang/test/CodeGen/X86/avx512fp16-builtins.c +++ b/clang/test/CodeGen/X86/avx512fp16-builtins.c @@ -1505,13 +1505,13 @@ __m128h test_mm_load_sh(void const *A) { __m128h test_mm_mask_load_sh(__m128h __A, __mmask8 __U, const void *__W) { // CHECK-LABEL: test_mm_mask_load_sh - // CHECK: @llvm.masked.load.v8f16.p0(ptr %{{.*}}, i32 1, <8 x i1> %{{.*}}, <8 x half> %{{.*}}) + // CHECK: @llvm.masked.load.v8f16.p0(ptr align 1 %{{.*}}, <8 x i1> %{{.*}}, <8 x half> %{{.*}}) return _mm_mask_load_sh(__A, __U, __W); } __m128h test_mm_maskz_load_sh(__mmask8 __U, const void *__W) { // CHECK-LABEL: test_mm_maskz_load_sh - // CHECK: @llvm.masked.load.v8f16.p0(ptr %{{.*}}, i32 1, <8 x i1> %{{.*}}, <8 x half> %{{.*}}) + // CHECK: @llvm.masked.load.v8f16.p0(ptr align 1 %{{.*}}, <8 x i1> %{{.*}}, <8 x half> %{{.*}}) return _mm_maskz_load_sh(__U, __W); } @@ -1560,7 +1560,7 @@ void test_mm_store_sh(void *A, __m128h B) { void test_mm_mask_store_sh(void *__P, __mmask8 __U, __m128h __A) { // CHECK-LABEL: test_mm_mask_store_sh - // CHECK: call void @llvm.masked.store.v8f16.p0(<8 x half> %{{.*}}, ptr %{{.*}}, i32 1, <8 x i1> %{{.*}}) + // CHECK: call void @llvm.masked.store.v8f16.p0(<8 x half> %{{.*}}, ptr align 1 %{{.*}}, <8 x i1> %{{.*}}) _mm_mask_store_sh(__P, __U, __A); } diff --git a/clang/test/CodeGen/X86/avx512vl-builtins.c b/clang/test/CodeGen/X86/avx512vl-builtins.c index 9b6bfea9..9819dbd 100644 --- a/clang/test/CodeGen/X86/avx512vl-builtins.c +++ b/clang/test/CodeGen/X86/avx512vl-builtins.c @@ -7002,7 +7002,7 @@ void test_mm_store_epi32(void *__P, __m128i __A) { void test_mm_mask_store_epi32(void *__P, __mmask8 __U, __m128i __A) { // CHECK-LABEL: test_mm_mask_store_epi32 - // CHECK: @llvm.masked.store.v4i32.p0(<4 x i32> %{{.*}}, ptr %{{.}}, i32 16, <4 x i1> %{{.*}}) + // CHECK: @llvm.masked.store.v4i32.p0(<4 x i32> %{{.*}}, ptr align 16 %{{.}}, <4 x i1> %{{.*}}) return _mm_mask_store_epi32(__P, __U, __A); } @@ -7014,7 +7014,7 @@ void test_mm256_store_epi32(void *__P, __m256i __A) { void test_mm256_mask_store_epi32(void *__P, __mmask8 __U, __m256i __A) { // CHECK-LABEL: test_mm256_mask_store_epi32 - // CHECK: @llvm.masked.store.v8i32.p0(<8 x i32> %{{.*}}, ptr %{{.}}, i32 32, <8 x i1> %{{.*}}) + // CHECK: @llvm.masked.store.v8i32.p0(<8 x i32> %{{.*}}, ptr align 32 %{{.}}, <8 x i1> %{{.*}}) return _mm256_mask_store_epi32(__P, __U, __A); } @@ -7074,13 +7074,13 @@ __m128i test_mm_load_epi32(void const *__P) { __m128i test_mm_mask_load_epi32(__m128i __W, __mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm_mask_load_epi32 - // CHECK: @llvm.masked.load.v4i32.p0(ptr %{{.*}}, i32 16, <4 x i1> %{{.*}}, <4 x i32> %{{.*}}) + // CHECK: @llvm.masked.load.v4i32.p0(ptr align 16 %{{.*}}, <4 x i1> %{{.*}}, <4 x i32> %{{.*}}) return _mm_mask_load_epi32(__W, __U, __P); } __m128i test_mm_maskz_load_epi32(__mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm_maskz_load_epi32 - // CHECK: @llvm.masked.load.v4i32.p0(ptr %{{.*}}, i32 16, <4 x i1> %{{.*}}, <4 x i32> %{{.*}}) + // CHECK: @llvm.masked.load.v4i32.p0(ptr align 16 %{{.*}}, <4 x i1> %{{.*}}, <4 x i32> %{{.*}}) return _mm_maskz_load_epi32(__U, __P); } @@ -7092,13 +7092,13 @@ __m256i test_mm256_load_epi32(void const *__P) { __m256i test_mm256_mask_load_epi32(__m256i __W, __mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm256_mask_load_epi32 - // CHECK: @llvm.masked.load.v8i32.p0(ptr %{{.*}}, i32 32, <8 x i1> %{{.*}}, <8 x i32> %{{.*}}) + // CHECK: @llvm.masked.load.v8i32.p0(ptr align 32 %{{.*}}, <8 x i1> %{{.*}}, <8 x i32> %{{.*}}) return _mm256_mask_load_epi32(__W, __U, __P); } __m256i test_mm256_maskz_load_epi32(__mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm256_maskz_load_epi32 - // CHECK: @llvm.masked.load.v8i32.p0(ptr %{{.*}}, i32 32, <8 x i1> %{{.*}}, <8 x i32> %{{.*}}) + // CHECK: @llvm.masked.load.v8i32.p0(ptr align 32 %{{.*}}, <8 x i1> %{{.*}}, <8 x i32> %{{.*}}) return _mm256_maskz_load_epi32(__U, __P); } @@ -7110,13 +7110,13 @@ __m128i test_mm_load_epi64(void const *__P) { __m128i test_mm_mask_load_epi64(__m128i __W, __mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm_mask_load_epi64 - // CHECK: @llvm.masked.load.v2i64.p0(ptr %{{.*}}, i32 16, <2 x i1> %{{.*}}, <2 x i64> %{{.*}}) + // CHECK: @llvm.masked.load.v2i64.p0(ptr align 16 %{{.*}}, <2 x i1> %{{.*}}, <2 x i64> %{{.*}}) return _mm_mask_load_epi64(__W, __U, __P); } __m128i test_mm_maskz_load_epi64(__mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm_maskz_load_epi64 - // CHECK: @llvm.masked.load.v2i64.p0(ptr %{{.*}}, i32 16, <2 x i1> %{{.*}}, <2 x i64> %{{.*}}) + // CHECK: @llvm.masked.load.v2i64.p0(ptr align 16 %{{.*}}, <2 x i1> %{{.*}}, <2 x i64> %{{.*}}) return _mm_maskz_load_epi64(__U, __P); } @@ -7128,13 +7128,13 @@ __m256i test_mm256_load_epi64(void const *__P) { __m256i test_mm256_mask_load_epi64(__m256i __W, __mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm256_mask_load_epi64 - // CHECK: @llvm.masked.load.v4i64.p0(ptr %{{.*}}, i32 32, <4 x i1> %{{.*}}, <4 x i64> %{{.*}}) + // CHECK: @llvm.masked.load.v4i64.p0(ptr align 32 %{{.*}}, <4 x i1> %{{.*}}, <4 x i64> %{{.*}}) return _mm256_mask_load_epi64(__W, __U, __P); } __m256i test_mm256_maskz_load_epi64(__mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm256_maskz_load_epi64 - // CHECK: @llvm.masked.load.v4i64.p0(ptr %{{.*}}, i32 32, <4 x i1> %{{.*}}, <4 x i64> %{{.*}}) + // CHECK: @llvm.masked.load.v4i64.p0(ptr align 32 %{{.*}}, <4 x i1> %{{.*}}, <4 x i64> %{{.*}}) return _mm256_maskz_load_epi64(__U, __P); } @@ -7146,7 +7146,7 @@ void test_mm_store_epi64(void *__P, __m128i __A) { void test_mm_mask_store_epi64(void *__P, __mmask8 __U, __m128i __A) { // CHECK-LABEL: test_mm_mask_store_epi64 - // CHECK: @llvm.masked.store.v2i64.p0(<2 x i64> %{{.*}}, ptr %{{.*}}, i32 16, <2 x i1> %{{.*}}) + // CHECK: @llvm.masked.store.v2i64.p0(<2 x i64> %{{.*}}, ptr align 16 %{{.*}}, <2 x i1> %{{.*}}) return _mm_mask_store_epi64(__P, __U, __A); } @@ -7158,7 +7158,7 @@ void test_mm256_store_epi64(void *__P, __m256i __A) { void test_mm256_mask_store_epi64(void *__P, __mmask8 __U, __m256i __A) { // CHECK-LABEL: test_mm256_mask_store_epi64 - // CHECK: @llvm.masked.store.v4i64.p0(<4 x i64> %{{.*}}, ptr %{{.*}}, i32 32, <4 x i1> %{{.*}}) + // CHECK: @llvm.masked.store.v4i64.p0(<4 x i64> %{{.*}}, ptr align 32 %{{.*}}, <4 x i1> %{{.*}}) return _mm256_mask_store_epi64(__P, __U, __A); } @@ -7370,49 +7370,49 @@ __m256 test_mm256_maskz_fixupimm_ps(__mmask8 __U, __m256 __A, __m256 __B, __m256 __m128d test_mm_mask_load_pd(__m128d __W, __mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm_mask_load_pd - // CHECK: @llvm.masked.load.v2f64.p0(ptr %{{.*}}, i32 16, <2 x i1> %{{.*}}, <2 x double> %{{.*}}) + // CHECK: @llvm.masked.load.v2f64.p0(ptr align 16 %{{.*}}, <2 x i1> %{{.*}}, <2 x double> %{{.*}}) return _mm_mask_load_pd(__W, __U, __P); } __m128d test_mm_maskz_load_pd(__mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm_maskz_load_pd - // CHECK: @llvm.masked.load.v2f64.p0(ptr %{{.*}}, i32 16, <2 x i1> %{{.*}}, <2 x double> %{{.*}}) + // CHECK: @llvm.masked.load.v2f64.p0(ptr align 16 %{{.*}}, <2 x i1> %{{.*}}, <2 x double> %{{.*}}) return _mm_maskz_load_pd(__U, __P); } __m256d test_mm256_mask_load_pd(__m256d __W, __mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm256_mask_load_pd - // CHECK: @llvm.masked.load.v4f64.p0(ptr %{{.*}}, i32 32, <4 x i1> %{{.*}}, <4 x double> %{{.*}}) + // CHECK: @llvm.masked.load.v4f64.p0(ptr align 32 %{{.*}}, <4 x i1> %{{.*}}, <4 x double> %{{.*}}) return _mm256_mask_load_pd(__W, __U, __P); } __m256d test_mm256_maskz_load_pd(__mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm256_maskz_load_pd - // CHECK: @llvm.masked.load.v4f64.p0(ptr %{{.*}}, i32 32, <4 x i1> %{{.*}}, <4 x double> %{{.*}}) + // CHECK: @llvm.masked.load.v4f64.p0(ptr align 32 %{{.*}}, <4 x i1> %{{.*}}, <4 x double> %{{.*}}) return _mm256_maskz_load_pd(__U, __P); } __m128 test_mm_mask_load_ps(__m128 __W, __mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm_mask_load_ps - // CHECK: @llvm.masked.load.v4f32.p0(ptr %{{.*}}, i32 16, <4 x i1> %{{.*}}, <4 x float> %{{.*}}) + // CHECK: @llvm.masked.load.v4f32.p0(ptr align 16 %{{.*}}, <4 x i1> %{{.*}}, <4 x float> %{{.*}}) return _mm_mask_load_ps(__W, __U, __P); } __m128 test_mm_maskz_load_ps(__mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm_maskz_load_ps - // CHECK: @llvm.masked.load.v4f32.p0(ptr %{{.*}}, i32 16, <4 x i1> %{{.*}}, <4 x float> %{{.*}}) + // CHECK: @llvm.masked.load.v4f32.p0(ptr align 16 %{{.*}}, <4 x i1> %{{.*}}, <4 x float> %{{.*}}) return _mm_maskz_load_ps(__U, __P); } __m256 test_mm256_mask_load_ps(__m256 __W, __mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm256_mask_load_ps - // CHECK: @llvm.masked.load.v8f32.p0(ptr %{{.*}}, i32 32, <8 x i1> %{{.*}}, <8 x float> %{{.*}}) + // CHECK: @llvm.masked.load.v8f32.p0(ptr align 32 %{{.*}}, <8 x i1> %{{.*}}, <8 x float> %{{.*}}) return _mm256_mask_load_ps(__W, __U, __P); } __m256 test_mm256_maskz_load_ps(__mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm256_maskz_load_ps - // CHECK: @llvm.masked.load.v8f32.p0(ptr %{{.*}}, i32 32, <8 x i1> %{{.*}}, <8 x float> %{{.*}}) + // CHECK: @llvm.masked.load.v8f32.p0(ptr align 32 %{{.*}}, <8 x i1> %{{.*}}, <8 x float> %{{.*}}) return _mm256_maskz_load_ps(__U, __P); } @@ -7424,13 +7424,13 @@ __m128i test_mm_loadu_epi64(void const *__P) { __m128i test_mm_mask_loadu_epi64(__m128i __W, __mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm_mask_loadu_epi64 - // CHECK: @llvm.masked.load.v2i64.p0(ptr %{{.*}}, i32 1, <2 x i1> %{{.*}}, <2 x i64> %{{.*}}) + // CHECK: @llvm.masked.load.v2i64.p0(ptr align 1 %{{.*}}, <2 x i1> %{{.*}}, <2 x i64> %{{.*}}) return _mm_mask_loadu_epi64(__W, __U, __P); } __m128i test_mm_maskz_loadu_epi64(__mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm_maskz_loadu_epi64 - // CHECK: @llvm.masked.load.v2i64.p0(ptr %{{.*}}, i32 1, <2 x i1> %{{.*}}, <2 x i64> %{{.*}}) + // CHECK: @llvm.masked.load.v2i64.p0(ptr align 1 %{{.*}}, <2 x i1> %{{.*}}, <2 x i64> %{{.*}}) return _mm_maskz_loadu_epi64(__U, __P); } @@ -7442,13 +7442,13 @@ __m256i test_mm256_loadu_epi64(void const *__P) { __m256i test_mm256_mask_loadu_epi64(__m256i __W, __mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm256_mask_loadu_epi64 - // CHECK: @llvm.masked.load.v4i64.p0(ptr %{{.*}}, i32 1, <4 x i1> %{{.*}}, <4 x i64> %{{.*}}) + // CHECK: @llvm.masked.load.v4i64.p0(ptr align 1 %{{.*}}, <4 x i1> %{{.*}}, <4 x i64> %{{.*}}) return _mm256_mask_loadu_epi64(__W, __U, __P); } __m256i test_mm256_maskz_loadu_epi64(__mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm256_maskz_loadu_epi64 - // CHECK: @llvm.masked.load.v4i64.p0(ptr %{{.*}}, i32 1, <4 x i1> %{{.*}}, <4 x i64> %{{.*}}) + // CHECK: @llvm.masked.load.v4i64.p0(ptr align 1 %{{.*}}, <4 x i1> %{{.*}}, <4 x i64> %{{.*}}) return _mm256_maskz_loadu_epi64(__U, __P); } @@ -7460,13 +7460,13 @@ __m128i test_mm_loadu_epi32(void const *__P) { __m128i test_mm_mask_loadu_epi32(__m128i __W, __mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm_mask_loadu_epi32 - // CHECK: @llvm.masked.load.v4i32.p0(ptr %{{.*}}, i32 1, <4 x i1> %{{.*}}, <4 x i32> %{{.*}}) + // CHECK: @llvm.masked.load.v4i32.p0(ptr align 1 %{{.*}}, <4 x i1> %{{.*}}, <4 x i32> %{{.*}}) return _mm_mask_loadu_epi32(__W, __U, __P); } __m128i test_mm_maskz_loadu_epi32(__mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm_maskz_loadu_epi32 - // CHECK: @llvm.masked.load.v4i32.p0(ptr %{{.*}}, i32 1, <4 x i1> %{{.*}}, <4 x i32> %{{.*}}) + // CHECK: @llvm.masked.load.v4i32.p0(ptr align 1 %{{.*}}, <4 x i1> %{{.*}}, <4 x i32> %{{.*}}) return _mm_maskz_loadu_epi32(__U, __P); } @@ -7478,85 +7478,85 @@ __m256i test_mm256_loadu_epi32(void const *__P) { __m256i test_mm256_mask_loadu_epi32(__m256i __W, __mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm256_mask_loadu_epi32 - // CHECK: @llvm.masked.load.v8i32.p0(ptr %{{.*}}, i32 1, <8 x i1> %{{.*}}, <8 x i32> %{{.*}}) + // CHECK: @llvm.masked.load.v8i32.p0(ptr align 1 %{{.*}}, <8 x i1> %{{.*}}, <8 x i32> %{{.*}}) return _mm256_mask_loadu_epi32(__W, __U, __P); } __m256i test_mm256_maskz_loadu_epi32(__mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm256_maskz_loadu_epi32 - // CHECK: @llvm.masked.load.v8i32.p0(ptr %{{.*}}, i32 1, <8 x i1> %{{.*}}, <8 x i32> %{{.*}}) + // CHECK: @llvm.masked.load.v8i32.p0(ptr align 1 %{{.*}}, <8 x i1> %{{.*}}, <8 x i32> %{{.*}}) return _mm256_maskz_loadu_epi32(__U, __P); } __m128d test_mm_mask_loadu_pd(__m128d __W, __mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm_mask_loadu_pd - // CHECK: @llvm.masked.load.v2f64.p0(ptr %{{.*}}, i32 1, <2 x i1> %{{.*}}, <2 x double> %{{.*}}) + // CHECK: @llvm.masked.load.v2f64.p0(ptr align 1 %{{.*}}, <2 x i1> %{{.*}}, <2 x double> %{{.*}}) return _mm_mask_loadu_pd(__W, __U, __P); } __m128d test_mm_maskz_loadu_pd(__mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm_maskz_loadu_pd - // CHECK: @llvm.masked.load.v2f64.p0(ptr %{{.*}}, i32 1, <2 x i1> %{{.*}}, <2 x double> %{{.*}}) + // CHECK: @llvm.masked.load.v2f64.p0(ptr align 1 %{{.*}}, <2 x i1> %{{.*}}, <2 x double> %{{.*}}) return _mm_maskz_loadu_pd(__U, __P); } __m256d test_mm256_mask_loadu_pd(__m256d __W, __mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm256_mask_loadu_pd - // CHECK: @llvm.masked.load.v4f64.p0(ptr %{{.*}}, i32 1, <4 x i1> %{{.*}}, <4 x double> %{{.*}}) + // CHECK: @llvm.masked.load.v4f64.p0(ptr align 1 %{{.*}}, <4 x i1> %{{.*}}, <4 x double> %{{.*}}) return _mm256_mask_loadu_pd(__W, __U, __P); } __m256d test_mm256_maskz_loadu_pd(__mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm256_maskz_loadu_pd - // CHECK: @llvm.masked.load.v4f64.p0(ptr %{{.*}}, i32 1, <4 x i1> %{{.*}}, <4 x double> %{{.*}}) + // CHECK: @llvm.masked.load.v4f64.p0(ptr align 1 %{{.*}}, <4 x i1> %{{.*}}, <4 x double> %{{.*}}) return _mm256_maskz_loadu_pd(__U, __P); } __m128 test_mm_mask_loadu_ps(__m128 __W, __mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm_mask_loadu_ps - // CHECK: @llvm.masked.load.v4f32.p0(ptr %{{.*}}, i32 1, <4 x i1> %{{.*}}, <4 x float> %{{.*}}) + // CHECK: @llvm.masked.load.v4f32.p0(ptr align 1 %{{.*}}, <4 x i1> %{{.*}}, <4 x float> %{{.*}}) return _mm_mask_loadu_ps(__W, __U, __P); } __m128 test_mm_maskz_loadu_ps(__mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm_maskz_loadu_ps - // CHECK: @llvm.masked.load.v4f32.p0(ptr %{{.*}}, i32 1, <4 x i1> %{{.*}}, <4 x float> %{{.*}}) + // CHECK: @llvm.masked.load.v4f32.p0(ptr align 1 %{{.*}}, <4 x i1> %{{.*}}, <4 x float> %{{.*}}) return _mm_maskz_loadu_ps(__U, __P); } __m256 test_mm256_mask_loadu_ps(__m256 __W, __mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm256_mask_loadu_ps - // CHECK: @llvm.masked.load.v8f32.p0(ptr %{{.*}}, i32 1, <8 x i1> %{{.*}}, <8 x float> %{{.*}}) + // CHECK: @llvm.masked.load.v8f32.p0(ptr align 1 %{{.*}}, <8 x i1> %{{.*}}, <8 x float> %{{.*}}) return _mm256_mask_loadu_ps(__W, __U, __P); } __m256 test_mm256_maskz_loadu_ps(__mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm256_maskz_loadu_ps - // CHECK: @llvm.masked.load.v8f32.p0(ptr %{{.*}}, i32 1, <8 x i1> %{{.*}}, <8 x float> %{{.*}}) + // CHECK: @llvm.masked.load.v8f32.p0(ptr align 1 %{{.*}}, <8 x i1> %{{.*}}, <8 x float> %{{.*}}) return _mm256_maskz_loadu_ps(__U, __P); } void test_mm_mask_store_pd(void *__P, __mmask8 __U, __m128d __A) { // CHECK-LABEL: test_mm_mask_store_pd - // CHECK: @llvm.masked.store.v2f64.p0(<2 x double> %{{.*}}, ptr %{{.*}}, i32 16, <2 x i1> %{{.*}}) + // CHECK: @llvm.masked.store.v2f64.p0(<2 x double> %{{.*}}, ptr align 16 %{{.*}}, <2 x i1> %{{.*}}) return _mm_mask_store_pd(__P, __U, __A); } void test_mm256_mask_store_pd(void *__P, __mmask8 __U, __m256d __A) { // CHECK-LABEL: test_mm256_mask_store_pd - // CHECK: @llvm.masked.store.v4f64.p0(<4 x double> %{{.*}}, ptr %{{.*}}, i32 32, <4 x i1> %{{.*}}) + // CHECK: @llvm.masked.store.v4f64.p0(<4 x double> %{{.*}}, ptr align 32 %{{.*}}, <4 x i1> %{{.*}}) return _mm256_mask_store_pd(__P, __U, __A); } void test_mm_mask_store_ps(void *__P, __mmask8 __U, __m128 __A) { // CHECK-LABEL: test_mm_mask_store_ps - // CHECK: @llvm.masked.store.v4f32.p0(<4 x float> %{{.*}}, ptr %{{.*}}, i32 16, <4 x i1> %{{.*}}) + // CHECK: @llvm.masked.store.v4f32.p0(<4 x float> %{{.*}}, ptr align 16 %{{.*}}, <4 x i1> %{{.*}}) return _mm_mask_store_ps(__P, __U, __A); } void test_mm256_mask_store_ps(void *__P, __mmask8 __U, __m256 __A) { // CHECK-LABEL: test_mm256_mask_store_ps - // CHECK: @llvm.masked.store.v8f32.p0(<8 x float> %{{.*}}, ptr %{{.*}}, i32 32, <8 x i1> %{{.*}}) + // CHECK: @llvm.masked.store.v8f32.p0(<8 x float> %{{.*}}, ptr align 32 %{{.*}}, <8 x i1> %{{.*}}) return _mm256_mask_store_ps(__P, __U, __A); } @@ -7568,7 +7568,7 @@ void test_mm_storeu_epi64(void *__p, __m128i __a) { void test_mm_mask_storeu_epi64(void *__P, __mmask8 __U, __m128i __A) { // CHECK-LABEL: test_mm_mask_storeu_epi64 - // CHECK: @llvm.masked.store.v2i64.p0(<2 x i64> %{{.*}}, ptr %{{.*}}, i32 1, <2 x i1> %{{.*}}) + // CHECK: @llvm.masked.store.v2i64.p0(<2 x i64> %{{.*}}, ptr align 1 %{{.*}}, <2 x i1> %{{.*}}) return _mm_mask_storeu_epi64(__P, __U, __A); } @@ -7580,7 +7580,7 @@ void test_mm256_storeu_epi64(void *__P, __m256i __A) { void test_mm256_mask_storeu_epi64(void *__P, __mmask8 __U, __m256i __A) { // CHECK-LABEL: test_mm256_mask_storeu_epi64 - // CHECK: @llvm.masked.store.v4i64.p0(<4 x i64> %{{.*}}, ptr %{{.*}}, i32 1, <4 x i1> %{{.*}}) + // CHECK: @llvm.masked.store.v4i64.p0(<4 x i64> %{{.*}}, ptr align 1 %{{.*}}, <4 x i1> %{{.*}}) return _mm256_mask_storeu_epi64(__P, __U, __A); } @@ -7592,7 +7592,7 @@ void test_mm_storeu_epi32(void *__P, __m128i __A) { void test_mm_mask_storeu_epi32(void *__P, __mmask8 __U, __m128i __A) { // CHECK-LABEL: test_mm_mask_storeu_epi32 - // CHECK: @llvm.masked.store.v4i32.p0(<4 x i32> %{{.*}}, ptr %{{.*}}, i32 1, <4 x i1> %{{.*}}) + // CHECK: @llvm.masked.store.v4i32.p0(<4 x i32> %{{.*}}, ptr align 1 %{{.*}}, <4 x i1> %{{.*}}) return _mm_mask_storeu_epi32(__P, __U, __A); } @@ -7604,31 +7604,31 @@ void test_mm256_storeu_epi32(void *__P, __m256i __A) { void test_mm256_mask_storeu_epi32(void *__P, __mmask8 __U, __m256i __A) { // CHECK-LABEL: test_mm256_mask_storeu_epi32 - // CHECK: @llvm.masked.store.v8i32.p0(<8 x i32> %{{.*}}, ptr %{{.*}}, i32 1, <8 x i1> %{{.*}}) + // CHECK: @llvm.masked.store.v8i32.p0(<8 x i32> %{{.*}}, ptr align 1 %{{.*}}, <8 x i1> %{{.*}}) return _mm256_mask_storeu_epi32(__P, __U, __A); } void test_mm_mask_storeu_pd(void *__P, __mmask8 __U, __m128d __A) { // CHECK-LABEL: test_mm_mask_storeu_pd - // CHECK: @llvm.masked.store.v2f64.p0(<2 x double> %{{.*}}, ptr %{{.*}}, i32 1, <2 x i1> %{{.*}}) + // CHECK: @llvm.masked.store.v2f64.p0(<2 x double> %{{.*}}, ptr align 1 %{{.*}}, <2 x i1> %{{.*}}) return _mm_mask_storeu_pd(__P, __U, __A); } void test_mm256_mask_storeu_pd(void *__P, __mmask8 __U, __m256d __A) { // CHECK-LABEL: test_mm256_mask_storeu_pd - // CHECK: @llvm.masked.store.v4f64.p0(<4 x double> %{{.*}}, ptr %{{.*}}, i32 1, <4 x i1> %{{.*}}) + // CHECK: @llvm.masked.store.v4f64.p0(<4 x double> %{{.*}}, ptr align 1 %{{.*}}, <4 x i1> %{{.*}}) return _mm256_mask_storeu_pd(__P, __U, __A); } void test_mm_mask_storeu_ps(void *__P, __mmask8 __U, __m128 __A) { // CHECK-LABEL: test_mm_mask_storeu_ps - // CHECK: @llvm.masked.store.v4f32.p0(<4 x float> %{{.*}}, ptr %{{.*}}, i32 1, <4 x i1> %{{.*}}) + // CHECK: @llvm.masked.store.v4f32.p0(<4 x float> %{{.*}}, ptr align 1 %{{.*}}, <4 x i1> %{{.*}}) return _mm_mask_storeu_ps(__P, __U, __A); } void test_mm256_mask_storeu_ps(void *__P, __mmask8 __U, __m256 __A) { // CHECK-LABEL: test_mm256_mask_storeu_ps - // CHECK: @llvm.masked.store.v8f32.p0(<8 x float> %{{.*}}, ptr %{{.*}}, i32 1, <8 x i1> %{{.*}}) + // CHECK: @llvm.masked.store.v8f32.p0(<8 x float> %{{.*}}, ptr align 1 %{{.*}}, <8 x i1> %{{.*}}) return _mm256_mask_storeu_ps(__P, __U, __A); } diff --git a/clang/test/CodeGen/X86/avx512vlbw-builtins.c b/clang/test/CodeGen/X86/avx512vlbw-builtins.c index 95e4d40..116d86f 100644 --- a/clang/test/CodeGen/X86/avx512vlbw-builtins.c +++ b/clang/test/CodeGen/X86/avx512vlbw-builtins.c @@ -2776,13 +2776,13 @@ __m128i test_mm_loadu_epi16(void const *__P) { __m128i test_mm_mask_loadu_epi16(__m128i __W, __mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm_mask_loadu_epi16 - // CHECK: @llvm.masked.load.v8i16.p0(ptr %{{.*}}, i32 1, <8 x i1> %{{.*}}, <8 x i16> %{{.*}}) + // CHECK: @llvm.masked.load.v8i16.p0(ptr align 1 %{{.*}}, <8 x i1> %{{.*}}, <8 x i16> %{{.*}}) return _mm_mask_loadu_epi16(__W, __U, __P); } __m128i test_mm_maskz_loadu_epi16(__mmask8 __U, void const *__P) { // CHECK-LABEL: test_mm_maskz_loadu_epi16 - // CHECK: @llvm.masked.load.v8i16.p0(ptr %{{.*}}, i32 1, <8 x i1> %{{.*}}, <8 x i16> %{{.*}}) + // CHECK: @llvm.masked.load.v8i16.p0(ptr align 1 %{{.*}}, <8 x i1> %{{.*}}, <8 x i16> %{{.*}}) return _mm_maskz_loadu_epi16(__U, __P); } @@ -2794,13 +2794,13 @@ __m256i test_mm256_loadu_epi16(void const *__P) { __m256i test_mm256_mask_loadu_epi16(__m256i __W, __mmask16 __U, void const *__P) { // CHECK-LABEL: test_mm256_mask_loadu_epi16 - // CHECK: @llvm.masked.load.v16i16.p0(ptr %{{.*}}, i32 1, <16 x i1> %{{.*}}, <16 x i16> %{{.*}}) + // CHECK: @llvm.masked.load.v16i16.p0(ptr align 1 %{{.*}}, <16 x i1> %{{.*}}, <16 x i16> %{{.*}}) return _mm256_mask_loadu_epi16(__W, __U, __P); } __m256i test_mm256_maskz_loadu_epi16(__mmask16 __U, void const *__P) { // CHECK-LABEL: test_mm256_maskz_loadu_epi16 - // CHECK: @llvm.masked.load.v16i16.p0(ptr %{{.*}}, i32 1, <16 x i1> %{{.*}}, <16 x i16> %{{.*}}) + // CHECK: @llvm.masked.load.v16i16.p0(ptr align 1 %{{.*}}, <16 x i1> %{{.*}}, <16 x i16> %{{.*}}) return _mm256_maskz_loadu_epi16(__U, __P); } @@ -2812,13 +2812,13 @@ __m128i test_mm_loadu_epi8(void const *__P) { __m128i test_mm_mask_loadu_epi8(__m128i __W, __mmask16 __U, void const *__P) { // CHECK-LABEL: test_mm_mask_loadu_epi8 - // CHECK: @llvm.masked.load.v16i8.p0(ptr %{{.*}}, i32 1, <16 x i1> %{{.*}}, <16 x i8> %{{.*}}) + // CHECK: @llvm.masked.load.v16i8.p0(ptr align 1 %{{.*}}, <16 x i1> %{{.*}}, <16 x i8> %{{.*}}) return _mm_mask_loadu_epi8(__W, __U, __P); } __m128i test_mm_maskz_loadu_epi8(__mmask16 __U, void const *__P) { // CHECK-LABEL: test_mm_maskz_loadu_epi8 - // CHECK: @llvm.masked.load.v16i8.p0(ptr %{{.*}}, i32 1, <16 x i1> %{{.*}}, <16 x i8> %{{.*}}) + // CHECK: @llvm.masked.load.v16i8.p0(ptr align 1 %{{.*}}, <16 x i1> %{{.*}}, <16 x i8> %{{.*}}) return _mm_maskz_loadu_epi8(__U, __P); } @@ -2830,13 +2830,13 @@ __m256i test_mm256_loadu_epi8(void const *__P) { __m256i test_mm256_mask_loadu_epi8(__m256i __W, __mmask32 __U, void const *__P) { // CHECK-LABEL: test_mm256_mask_loadu_epi8 - // CHECK: @llvm.masked.load.v32i8.p0(ptr %{{.*}}, i32 1, <32 x i1> %{{.*}}, <32 x i8> %{{.*}}) + // CHECK: @llvm.masked.load.v32i8.p0(ptr align 1 %{{.*}}, <32 x i1> %{{.*}}, <32 x i8> %{{.*}}) return _mm256_mask_loadu_epi8(__W, __U, __P); } __m256i test_mm256_maskz_loadu_epi8(__mmask32 __U, void const *__P) { // CHECK-LABEL: test_mm256_maskz_loadu_epi8 - // CHECK: @llvm.masked.load.v32i8.p0(ptr %{{.*}}, i32 1, <32 x i1> %{{.*}}, <32 x i8> %{{.*}}) + // CHECK: @llvm.masked.load.v32i8.p0(ptr align 1 %{{.*}}, <32 x i1> %{{.*}}, <32 x i8> %{{.*}}) return _mm256_maskz_loadu_epi8(__U, __P); } @@ -2848,7 +2848,7 @@ void test_mm_storeu_epi16(void *__p, __m128i __a) { void test_mm_mask_storeu_epi16(void *__P, __mmask8 __U, __m128i __A) { // CHECK-LABEL: test_mm_mask_storeu_epi16 - // CHECK: @llvm.masked.store.v8i16.p0(<8 x i16> %{{.*}}, ptr %{{.*}}, i32 1, <8 x i1> %{{.*}}) + // CHECK: @llvm.masked.store.v8i16.p0(<8 x i16> %{{.*}}, ptr align 1 %{{.*}}, <8 x i1> %{{.*}}) return _mm_mask_storeu_epi16(__P, __U, __A); } @@ -2860,7 +2860,7 @@ void test_mm256_storeu_epi16(void *__P, __m256i __A) { void test_mm256_mask_storeu_epi16(void *__P, __mmask16 __U, __m256i __A) { // CHECK-LABEL: test_mm256_mask_storeu_epi16 - // CHECK: @llvm.masked.store.v16i16.p0(<16 x i16> %{{.*}}, ptr %{{.*}}, i32 1, <16 x i1> %{{.*}}) + // CHECK: @llvm.masked.store.v16i16.p0(<16 x i16> %{{.*}}, ptr align 1 %{{.*}}, <16 x i1> %{{.*}}) return _mm256_mask_storeu_epi16(__P, __U, __A); } @@ -2872,7 +2872,7 @@ void test_mm_storeu_epi8(void *__p, __m128i __a) { void test_mm_mask_storeu_epi8(void *__P, __mmask16 __U, __m128i __A) { // CHECK-LABEL: test_mm_mask_storeu_epi8 - // CHECK: @llvm.masked.store.v16i8.p0(<16 x i8> %{{.*}}, ptr %{{.*}}, i32 1, <16 x i1> %{{.*}}) + // CHECK: @llvm.masked.store.v16i8.p0(<16 x i8> %{{.*}}, ptr align 1 %{{.*}}, <16 x i1> %{{.*}}) return _mm_mask_storeu_epi8(__P, __U, __A); } @@ -2884,7 +2884,7 @@ void test_mm256_storeu_epi8(void *__P, __m256i __A) { void test_mm256_mask_storeu_epi8(void *__P, __mmask32 __U, __m256i __A) { // CHECK-LABEL: test_mm256_mask_storeu_epi8 - // CHECK: @llvm.masked.store.v32i8.p0(<32 x i8> %{{.*}}, ptr %{{.*}}, i32 1, <32 x i1> %{{.*}}) + // CHECK: @llvm.masked.store.v32i8.p0(<32 x i8> %{{.*}}, ptr align 1 %{{.*}}, <32 x i1> %{{.*}}) return _mm256_mask_storeu_epi8(__P, __U, __A); } __mmask16 test_mm_test_epi8_mask(__m128i __A, __m128i __B) { diff --git a/clang/test/CodeGen/arm-mve-intrinsics/cplusplus.cpp b/clang/test/CodeGen/arm-mve-intrinsics/cplusplus.cpp index 2971961..4e3d0ce 100644 --- a/clang/test/CodeGen/arm-mve-intrinsics/cplusplus.cpp +++ b/clang/test/CodeGen/arm-mve-intrinsics/cplusplus.cpp @@ -114,7 +114,7 @@ uint16x8_t test_vld1q_u16(const uint16_t *base) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP0]]) -// CHECK-NEXT: call void @llvm.masked.store.v4i32.p0(<4 x i32> [[VALUE:%.*]], ptr [[BASE:%.*]], i32 4, <4 x i1> [[TMP1]]) +// CHECK-NEXT: call void @llvm.masked.store.v4i32.p0(<4 x i32> [[VALUE:%.*]], ptr align 4 [[BASE:%.*]], <4 x i1> [[TMP1]]) // CHECK-NEXT: ret void // void test_vst1q_p_s32(int32_t *base, int32x4_t value, mve_pred16_t p) diff --git a/clang/test/CodeGen/arm-mve-intrinsics/dup.c b/clang/test/CodeGen/arm-mve-intrinsics/dup.c index c2c7a9c..e7113fd 100644 --- a/clang/test/CodeGen/arm-mve-intrinsics/dup.c +++ b/clang/test/CodeGen/arm-mve-intrinsics/dup.c @@ -244,8 +244,7 @@ uint32x4_t test_vdupq_m_n_u32(uint32x4_t inactive, uint32_t a, mve_pred16_t p) // CHECK-NEXT: [[TMP1:%.*]] = call <8 x i1> @llvm.arm.mve.pred.i2v.v8i1(i32 [[TMP0]]) // CHECK-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <8 x half> poison, half [[A:%.*]], i64 0 // CHECK-NEXT: [[DOTSPLAT:%.*]] = shufflevector <8 x half> [[DOTSPLATINSERT]], <8 x half> poison, <8 x i32> zeroinitializer -// CHECK-NEXT: [[TMP2:%.*]] = select <8 x i1> [[TMP1]], <8 x half> [[DOTSPLAT]], <8 x half> undef -// CHECK-NEXT: ret <8 x half> [[TMP2]] +// CHECK-NEXT: ret <8 x half> [[DOTSPLAT]] // float16x8_t test_vdupq_x_n_f16(float16_t a, mve_pred16_t p) { @@ -258,8 +257,7 @@ float16x8_t test_vdupq_x_n_f16(float16_t a, mve_pred16_t p) // CHECK-NEXT: [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP0]]) // CHECK-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <4 x float> poison, float [[A:%.*]], i64 0 // CHECK-NEXT: [[DOTSPLAT:%.*]] = shufflevector <4 x float> [[DOTSPLATINSERT]], <4 x float> poison, <4 x i32> zeroinitializer -// CHECK-NEXT: [[TMP2:%.*]] = select <4 x i1> [[TMP1]], <4 x float> [[DOTSPLAT]], <4 x float> undef -// CHECK-NEXT: ret <4 x float> [[TMP2]] +// CHECK-NEXT: ret <4 x float> [[DOTSPLAT]] // float32x4_t test_vdupq_x_n_f32(float32_t a, mve_pred16_t p) { @@ -272,8 +270,7 @@ float32x4_t test_vdupq_x_n_f32(float32_t a, mve_pred16_t p) // CHECK-NEXT: [[TMP1:%.*]] = call <16 x i1> @llvm.arm.mve.pred.i2v.v16i1(i32 [[TMP0]]) // CHECK-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <16 x i8> poison, i8 [[A:%.*]], i64 0 // CHECK-NEXT: [[DOTSPLAT:%.*]] = shufflevector <16 x i8> [[DOTSPLATINSERT]], <16 x i8> poison, <16 x i32> zeroinitializer -// CHECK-NEXT: [[TMP2:%.*]] = select <16 x i1> [[TMP1]], <16 x i8> [[DOTSPLAT]], <16 x i8> undef -// CHECK-NEXT: ret <16 x i8> [[TMP2]] +// CHECK-NEXT: ret <16 x i8> [[DOTSPLAT]] // int8x16_t test_vdupq_x_n_s8(int8_t a, mve_pred16_t p) { @@ -286,8 +283,7 @@ int8x16_t test_vdupq_x_n_s8(int8_t a, mve_pred16_t p) // CHECK-NEXT: [[TMP1:%.*]] = call <8 x i1> @llvm.arm.mve.pred.i2v.v8i1(i32 [[TMP0]]) // CHECK-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <8 x i16> poison, i16 [[A:%.*]], i64 0 // CHECK-NEXT: [[DOTSPLAT:%.*]] = shufflevector <8 x i16> [[DOTSPLATINSERT]], <8 x i16> poison, <8 x i32> zeroinitializer -// CHECK-NEXT: [[TMP2:%.*]] = select <8 x i1> [[TMP1]], <8 x i16> [[DOTSPLAT]], <8 x i16> undef -// CHECK-NEXT: ret <8 x i16> [[TMP2]] +// CHECK-NEXT: ret <8 x i16> [[DOTSPLAT]] // int16x8_t test_vdupq_x_n_s16(int16_t a, mve_pred16_t p) { @@ -300,8 +296,7 @@ int16x8_t test_vdupq_x_n_s16(int16_t a, mve_pred16_t p) // CHECK-NEXT: [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP0]]) // CHECK-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <4 x i32> poison, i32 [[A:%.*]], i64 0 // CHECK-NEXT: [[DOTSPLAT:%.*]] = shufflevector <4 x i32> [[DOTSPLATINSERT]], <4 x i32> poison, <4 x i32> zeroinitializer -// CHECK-NEXT: [[TMP2:%.*]] = select <4 x i1> [[TMP1]], <4 x i32> [[DOTSPLAT]], <4 x i32> undef -// CHECK-NEXT: ret <4 x i32> [[TMP2]] +// CHECK-NEXT: ret <4 x i32> [[DOTSPLAT]] // int32x4_t test_vdupq_x_n_s32(int32_t a, mve_pred16_t p) { @@ -314,8 +309,7 @@ int32x4_t test_vdupq_x_n_s32(int32_t a, mve_pred16_t p) // CHECK-NEXT: [[TMP1:%.*]] = call <16 x i1> @llvm.arm.mve.pred.i2v.v16i1(i32 [[TMP0]]) // CHECK-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <16 x i8> poison, i8 [[A:%.*]], i64 0 // CHECK-NEXT: [[DOTSPLAT:%.*]] = shufflevector <16 x i8> [[DOTSPLATINSERT]], <16 x i8> poison, <16 x i32> zeroinitializer -// CHECK-NEXT: [[TMP2:%.*]] = select <16 x i1> [[TMP1]], <16 x i8> [[DOTSPLAT]], <16 x i8> undef -// CHECK-NEXT: ret <16 x i8> [[TMP2]] +// CHECK-NEXT: ret <16 x i8> [[DOTSPLAT]] // uint8x16_t test_vdupq_x_n_u8(uint8_t a, mve_pred16_t p) { @@ -328,8 +322,7 @@ uint8x16_t test_vdupq_x_n_u8(uint8_t a, mve_pred16_t p) // CHECK-NEXT: [[TMP1:%.*]] = call <8 x i1> @llvm.arm.mve.pred.i2v.v8i1(i32 [[TMP0]]) // CHECK-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <8 x i16> poison, i16 [[A:%.*]], i64 0 // CHECK-NEXT: [[DOTSPLAT:%.*]] = shufflevector <8 x i16> [[DOTSPLATINSERT]], <8 x i16> poison, <8 x i32> zeroinitializer -// CHECK-NEXT: [[TMP2:%.*]] = select <8 x i1> [[TMP1]], <8 x i16> [[DOTSPLAT]], <8 x i16> undef -// CHECK-NEXT: ret <8 x i16> [[TMP2]] +// CHECK-NEXT: ret <8 x i16> [[DOTSPLAT]] // uint16x8_t test_vdupq_x_n_u16(uint16_t a, mve_pred16_t p) { @@ -342,8 +335,7 @@ uint16x8_t test_vdupq_x_n_u16(uint16_t a, mve_pred16_t p) // CHECK-NEXT: [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP0]]) // CHECK-NEXT: [[DOTSPLATINSERT:%.*]] = insertelement <4 x i32> poison, i32 [[A:%.*]], i64 0 // CHECK-NEXT: [[DOTSPLAT:%.*]] = shufflevector <4 x i32> [[DOTSPLATINSERT]], <4 x i32> poison, <4 x i32> zeroinitializer -// CHECK-NEXT: [[TMP2:%.*]] = select <4 x i1> [[TMP1]], <4 x i32> [[DOTSPLAT]], <4 x i32> undef -// CHECK-NEXT: ret <4 x i32> [[TMP2]] +// CHECK-NEXT: ret <4 x i32> [[DOTSPLAT]] // uint32x4_t test_vdupq_x_n_u32(uint32_t a, mve_pred16_t p) { diff --git a/clang/test/CodeGen/arm-mve-intrinsics/load-store.c b/clang/test/CodeGen/arm-mve-intrinsics/load-store.c index 50f7011..ede2e95 100644 --- a/clang/test/CodeGen/arm-mve-intrinsics/load-store.c +++ b/clang/test/CodeGen/arm-mve-intrinsics/load-store.c @@ -122,7 +122,7 @@ uint32x4_t test_vld1q_u32(const uint32_t *base) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <8 x i1> @llvm.arm.mve.pred.i2v.v8i1(i32 [[TMP0]]) -// CHECK-NEXT: [[TMP2:%.*]] = call <8 x half> @llvm.masked.load.v8f16.p0(ptr [[BASE:%.*]], i32 2, <8 x i1> [[TMP1]], <8 x half> zeroinitializer) +// CHECK-NEXT: [[TMP2:%.*]] = call <8 x half> @llvm.masked.load.v8f16.p0(ptr align 2 [[BASE:%.*]], <8 x i1> [[TMP1]], <8 x half> zeroinitializer) // CHECK-NEXT: ret <8 x half> [[TMP2]] // float16x8_t test_vld1q_z_f16(const float16_t *base, mve_pred16_t p) @@ -138,7 +138,7 @@ float16x8_t test_vld1q_z_f16(const float16_t *base, mve_pred16_t p) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP0]]) -// CHECK-NEXT: [[TMP2:%.*]] = call <4 x float> @llvm.masked.load.v4f32.p0(ptr [[BASE:%.*]], i32 4, <4 x i1> [[TMP1]], <4 x float> zeroinitializer) +// CHECK-NEXT: [[TMP2:%.*]] = call <4 x float> @llvm.masked.load.v4f32.p0(ptr align 4 [[BASE:%.*]], <4 x i1> [[TMP1]], <4 x float> zeroinitializer) // CHECK-NEXT: ret <4 x float> [[TMP2]] // float32x4_t test_vld1q_z_f32(const float32_t *base, mve_pred16_t p) @@ -154,7 +154,7 @@ float32x4_t test_vld1q_z_f32(const float32_t *base, mve_pred16_t p) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <16 x i1> @llvm.arm.mve.pred.i2v.v16i1(i32 [[TMP0]]) -// CHECK-NEXT: [[TMP2:%.*]] = call <16 x i8> @llvm.masked.load.v16i8.p0(ptr [[BASE:%.*]], i32 1, <16 x i1> [[TMP1]], <16 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP2:%.*]] = call <16 x i8> @llvm.masked.load.v16i8.p0(ptr align 1 [[BASE:%.*]], <16 x i1> [[TMP1]], <16 x i8> zeroinitializer) // CHECK-NEXT: ret <16 x i8> [[TMP2]] // int8x16_t test_vld1q_z_s8(const int8_t *base, mve_pred16_t p) @@ -170,7 +170,7 @@ int8x16_t test_vld1q_z_s8(const int8_t *base, mve_pred16_t p) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <8 x i1> @llvm.arm.mve.pred.i2v.v8i1(i32 [[TMP0]]) -// CHECK-NEXT: [[TMP2:%.*]] = call <8 x i16> @llvm.masked.load.v8i16.p0(ptr [[BASE:%.*]], i32 2, <8 x i1> [[TMP1]], <8 x i16> zeroinitializer) +// CHECK-NEXT: [[TMP2:%.*]] = call <8 x i16> @llvm.masked.load.v8i16.p0(ptr align 2 [[BASE:%.*]], <8 x i1> [[TMP1]], <8 x i16> zeroinitializer) // CHECK-NEXT: ret <8 x i16> [[TMP2]] // int16x8_t test_vld1q_z_s16(const int16_t *base, mve_pred16_t p) @@ -186,7 +186,7 @@ int16x8_t test_vld1q_z_s16(const int16_t *base, mve_pred16_t p) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP0]]) -// CHECK-NEXT: [[TMP2:%.*]] = call <4 x i32> @llvm.masked.load.v4i32.p0(ptr [[BASE:%.*]], i32 4, <4 x i1> [[TMP1]], <4 x i32> zeroinitializer) +// CHECK-NEXT: [[TMP2:%.*]] = call <4 x i32> @llvm.masked.load.v4i32.p0(ptr align 4 [[BASE:%.*]], <4 x i1> [[TMP1]], <4 x i32> zeroinitializer) // CHECK-NEXT: ret <4 x i32> [[TMP2]] // int32x4_t test_vld1q_z_s32(const int32_t *base, mve_pred16_t p) @@ -202,7 +202,7 @@ int32x4_t test_vld1q_z_s32(const int32_t *base, mve_pred16_t p) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <16 x i1> @llvm.arm.mve.pred.i2v.v16i1(i32 [[TMP0]]) -// CHECK-NEXT: [[TMP2:%.*]] = call <16 x i8> @llvm.masked.load.v16i8.p0(ptr [[BASE:%.*]], i32 1, <16 x i1> [[TMP1]], <16 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP2:%.*]] = call <16 x i8> @llvm.masked.load.v16i8.p0(ptr align 1 [[BASE:%.*]], <16 x i1> [[TMP1]], <16 x i8> zeroinitializer) // CHECK-NEXT: ret <16 x i8> [[TMP2]] // uint8x16_t test_vld1q_z_u8(const uint8_t *base, mve_pred16_t p) @@ -218,7 +218,7 @@ uint8x16_t test_vld1q_z_u8(const uint8_t *base, mve_pred16_t p) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <8 x i1> @llvm.arm.mve.pred.i2v.v8i1(i32 [[TMP0]]) -// CHECK-NEXT: [[TMP2:%.*]] = call <8 x i16> @llvm.masked.load.v8i16.p0(ptr [[BASE:%.*]], i32 2, <8 x i1> [[TMP1]], <8 x i16> zeroinitializer) +// CHECK-NEXT: [[TMP2:%.*]] = call <8 x i16> @llvm.masked.load.v8i16.p0(ptr align 2 [[BASE:%.*]], <8 x i1> [[TMP1]], <8 x i16> zeroinitializer) // CHECK-NEXT: ret <8 x i16> [[TMP2]] // uint16x8_t test_vld1q_z_u16(const uint16_t *base, mve_pred16_t p) @@ -234,7 +234,7 @@ uint16x8_t test_vld1q_z_u16(const uint16_t *base, mve_pred16_t p) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP0]]) -// CHECK-NEXT: [[TMP2:%.*]] = call <4 x i32> @llvm.masked.load.v4i32.p0(ptr [[BASE:%.*]], i32 4, <4 x i1> [[TMP1]], <4 x i32> zeroinitializer) +// CHECK-NEXT: [[TMP2:%.*]] = call <4 x i32> @llvm.masked.load.v4i32.p0(ptr align 4 [[BASE:%.*]], <4 x i1> [[TMP1]], <4 x i32> zeroinitializer) // CHECK-NEXT: ret <4 x i32> [[TMP2]] // uint32x4_t test_vld1q_z_u32(const uint32_t *base, mve_pred16_t p) @@ -314,7 +314,7 @@ uint32x4_t test_vldrbq_u32(const uint8_t *base) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <16 x i1> @llvm.arm.mve.pred.i2v.v16i1(i32 [[TMP0]]) -// CHECK-NEXT: [[TMP2:%.*]] = call <16 x i8> @llvm.masked.load.v16i8.p0(ptr [[BASE:%.*]], i32 1, <16 x i1> [[TMP1]], <16 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP2:%.*]] = call <16 x i8> @llvm.masked.load.v16i8.p0(ptr align 1 [[BASE:%.*]], <16 x i1> [[TMP1]], <16 x i8> zeroinitializer) // CHECK-NEXT: ret <16 x i8> [[TMP2]] // int8x16_t test_vldrbq_z_s8(const int8_t *base, mve_pred16_t p) @@ -326,7 +326,7 @@ int8x16_t test_vldrbq_z_s8(const int8_t *base, mve_pred16_t p) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <8 x i1> @llvm.arm.mve.pred.i2v.v8i1(i32 [[TMP0]]) -// CHECK-NEXT: [[TMP2:%.*]] = call <8 x i8> @llvm.masked.load.v8i8.p0(ptr [[BASE:%.*]], i32 1, <8 x i1> [[TMP1]], <8 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP2:%.*]] = call <8 x i8> @llvm.masked.load.v8i8.p0(ptr align 1 [[BASE:%.*]], <8 x i1> [[TMP1]], <8 x i8> zeroinitializer) // CHECK-NEXT: [[TMP3:%.*]] = sext <8 x i8> [[TMP2]] to <8 x i16> // CHECK-NEXT: ret <8 x i16> [[TMP3]] // @@ -339,7 +339,7 @@ int16x8_t test_vldrbq_z_s16(const int8_t *base, mve_pred16_t p) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP0]]) -// CHECK-NEXT: [[TMP2:%.*]] = call <4 x i8> @llvm.masked.load.v4i8.p0(ptr [[BASE:%.*]], i32 1, <4 x i1> [[TMP1]], <4 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP2:%.*]] = call <4 x i8> @llvm.masked.load.v4i8.p0(ptr align 1 [[BASE:%.*]], <4 x i1> [[TMP1]], <4 x i8> zeroinitializer) // CHECK-NEXT: [[TMP3:%.*]] = sext <4 x i8> [[TMP2]] to <4 x i32> // CHECK-NEXT: ret <4 x i32> [[TMP3]] // @@ -352,7 +352,7 @@ int32x4_t test_vldrbq_z_s32(const int8_t *base, mve_pred16_t p) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <16 x i1> @llvm.arm.mve.pred.i2v.v16i1(i32 [[TMP0]]) -// CHECK-NEXT: [[TMP2:%.*]] = call <16 x i8> @llvm.masked.load.v16i8.p0(ptr [[BASE:%.*]], i32 1, <16 x i1> [[TMP1]], <16 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP2:%.*]] = call <16 x i8> @llvm.masked.load.v16i8.p0(ptr align 1 [[BASE:%.*]], <16 x i1> [[TMP1]], <16 x i8> zeroinitializer) // CHECK-NEXT: ret <16 x i8> [[TMP2]] // uint8x16_t test_vldrbq_z_u8(const uint8_t *base, mve_pred16_t p) @@ -364,7 +364,7 @@ uint8x16_t test_vldrbq_z_u8(const uint8_t *base, mve_pred16_t p) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <8 x i1> @llvm.arm.mve.pred.i2v.v8i1(i32 [[TMP0]]) -// CHECK-NEXT: [[TMP2:%.*]] = call <8 x i8> @llvm.masked.load.v8i8.p0(ptr [[BASE:%.*]], i32 1, <8 x i1> [[TMP1]], <8 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP2:%.*]] = call <8 x i8> @llvm.masked.load.v8i8.p0(ptr align 1 [[BASE:%.*]], <8 x i1> [[TMP1]], <8 x i8> zeroinitializer) // CHECK-NEXT: [[TMP3:%.*]] = zext <8 x i8> [[TMP2]] to <8 x i16> // CHECK-NEXT: ret <8 x i16> [[TMP3]] // @@ -377,7 +377,7 @@ uint16x8_t test_vldrbq_z_u16(const uint8_t *base, mve_pred16_t p) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP0]]) -// CHECK-NEXT: [[TMP2:%.*]] = call <4 x i8> @llvm.masked.load.v4i8.p0(ptr [[BASE:%.*]], i32 1, <4 x i1> [[TMP1]], <4 x i8> zeroinitializer) +// CHECK-NEXT: [[TMP2:%.*]] = call <4 x i8> @llvm.masked.load.v4i8.p0(ptr align 1 [[BASE:%.*]], <4 x i1> [[TMP1]], <4 x i8> zeroinitializer) // CHECK-NEXT: [[TMP3:%.*]] = zext <4 x i8> [[TMP2]] to <4 x i32> // CHECK-NEXT: ret <4 x i32> [[TMP3]] // @@ -442,7 +442,7 @@ uint32x4_t test_vldrhq_u32(const uint16_t *base) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <8 x i1> @llvm.arm.mve.pred.i2v.v8i1(i32 [[TMP0]]) -// CHECK-NEXT: [[TMP2:%.*]] = call <8 x half> @llvm.masked.load.v8f16.p0(ptr [[BASE:%.*]], i32 2, <8 x i1> [[TMP1]], <8 x half> zeroinitializer) +// CHECK-NEXT: [[TMP2:%.*]] = call <8 x half> @llvm.masked.load.v8f16.p0(ptr align 2 [[BASE:%.*]], <8 x i1> [[TMP1]], <8 x half> zeroinitializer) // CHECK-NEXT: ret <8 x half> [[TMP2]] // float16x8_t test_vldrhq_z_f16(const float16_t *base, mve_pred16_t p) @@ -454,7 +454,7 @@ float16x8_t test_vldrhq_z_f16(const float16_t *base, mve_pred16_t p) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <8 x i1> @llvm.arm.mve.pred.i2v.v8i1(i32 [[TMP0]]) -// CHECK-NEXT: [[TMP2:%.*]] = call <8 x i16> @llvm.masked.load.v8i16.p0(ptr [[BASE:%.*]], i32 2, <8 x i1> [[TMP1]], <8 x i16> zeroinitializer) +// CHECK-NEXT: [[TMP2:%.*]] = call <8 x i16> @llvm.masked.load.v8i16.p0(ptr align 2 [[BASE:%.*]], <8 x i1> [[TMP1]], <8 x i16> zeroinitializer) // CHECK-NEXT: ret <8 x i16> [[TMP2]] // int16x8_t test_vldrhq_z_s16(const int16_t *base, mve_pred16_t p) @@ -466,7 +466,7 @@ int16x8_t test_vldrhq_z_s16(const int16_t *base, mve_pred16_t p) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP0]]) -// CHECK-NEXT: [[TMP2:%.*]] = call <4 x i16> @llvm.masked.load.v4i16.p0(ptr [[BASE:%.*]], i32 2, <4 x i1> [[TMP1]], <4 x i16> zeroinitializer) +// CHECK-NEXT: [[TMP2:%.*]] = call <4 x i16> @llvm.masked.load.v4i16.p0(ptr align 2 [[BASE:%.*]], <4 x i1> [[TMP1]], <4 x i16> zeroinitializer) // CHECK-NEXT: [[TMP3:%.*]] = sext <4 x i16> [[TMP2]] to <4 x i32> // CHECK-NEXT: ret <4 x i32> [[TMP3]] // @@ -479,7 +479,7 @@ int32x4_t test_vldrhq_z_s32(const int16_t *base, mve_pred16_t p) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <8 x i1> @llvm.arm.mve.pred.i2v.v8i1(i32 [[TMP0]]) -// CHECK-NEXT: [[TMP2:%.*]] = call <8 x i16> @llvm.masked.load.v8i16.p0(ptr [[BASE:%.*]], i32 2, <8 x i1> [[TMP1]], <8 x i16> zeroinitializer) +// CHECK-NEXT: [[TMP2:%.*]] = call <8 x i16> @llvm.masked.load.v8i16.p0(ptr align 2 [[BASE:%.*]], <8 x i1> [[TMP1]], <8 x i16> zeroinitializer) // CHECK-NEXT: ret <8 x i16> [[TMP2]] // uint16x8_t test_vldrhq_z_u16(const uint16_t *base, mve_pred16_t p) @@ -491,7 +491,7 @@ uint16x8_t test_vldrhq_z_u16(const uint16_t *base, mve_pred16_t p) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP0]]) -// CHECK-NEXT: [[TMP2:%.*]] = call <4 x i16> @llvm.masked.load.v4i16.p0(ptr [[BASE:%.*]], i32 2, <4 x i1> [[TMP1]], <4 x i16> zeroinitializer) +// CHECK-NEXT: [[TMP2:%.*]] = call <4 x i16> @llvm.masked.load.v4i16.p0(ptr align 2 [[BASE:%.*]], <4 x i1> [[TMP1]], <4 x i16> zeroinitializer) // CHECK-NEXT: [[TMP3:%.*]] = zext <4 x i16> [[TMP2]] to <4 x i32> // CHECK-NEXT: ret <4 x i32> [[TMP3]] // @@ -534,7 +534,7 @@ uint32x4_t test_vldrwq_u32(const uint32_t *base) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP0]]) -// CHECK-NEXT: [[TMP2:%.*]] = call <4 x float> @llvm.masked.load.v4f32.p0(ptr [[BASE:%.*]], i32 4, <4 x i1> [[TMP1]], <4 x float> zeroinitializer) +// CHECK-NEXT: [[TMP2:%.*]] = call <4 x float> @llvm.masked.load.v4f32.p0(ptr align 4 [[BASE:%.*]], <4 x i1> [[TMP1]], <4 x float> zeroinitializer) // CHECK-NEXT: ret <4 x float> [[TMP2]] // float32x4_t test_vldrwq_z_f32(const float32_t *base, mve_pred16_t p) @@ -546,7 +546,7 @@ float32x4_t test_vldrwq_z_f32(const float32_t *base, mve_pred16_t p) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP0]]) -// CHECK-NEXT: [[TMP2:%.*]] = call <4 x i32> @llvm.masked.load.v4i32.p0(ptr [[BASE:%.*]], i32 4, <4 x i1> [[TMP1]], <4 x i32> zeroinitializer) +// CHECK-NEXT: [[TMP2:%.*]] = call <4 x i32> @llvm.masked.load.v4i32.p0(ptr align 4 [[BASE:%.*]], <4 x i1> [[TMP1]], <4 x i32> zeroinitializer) // CHECK-NEXT: ret <4 x i32> [[TMP2]] // int32x4_t test_vldrwq_z_s32(const int32_t *base, mve_pred16_t p) @@ -558,7 +558,7 @@ int32x4_t test_vldrwq_z_s32(const int32_t *base, mve_pred16_t p) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP0]]) -// CHECK-NEXT: [[TMP2:%.*]] = call <4 x i32> @llvm.masked.load.v4i32.p0(ptr [[BASE:%.*]], i32 4, <4 x i1> [[TMP1]], <4 x i32> zeroinitializer) +// CHECK-NEXT: [[TMP2:%.*]] = call <4 x i32> @llvm.masked.load.v4i32.p0(ptr align 4 [[BASE:%.*]], <4 x i1> [[TMP1]], <4 x i32> zeroinitializer) // CHECK-NEXT: ret <4 x i32> [[TMP2]] // uint32x4_t test_vldrwq_z_u32(const uint32_t *base, mve_pred16_t p) @@ -682,7 +682,7 @@ void test_vst1q_u32(uint32_t *base, uint32x4_t value) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <8 x i1> @llvm.arm.mve.pred.i2v.v8i1(i32 [[TMP0]]) -// CHECK-NEXT: call void @llvm.masked.store.v8f16.p0(<8 x half> [[VALUE:%.*]], ptr [[BASE:%.*]], i32 2, <8 x i1> [[TMP1]]) +// CHECK-NEXT: call void @llvm.masked.store.v8f16.p0(<8 x half> [[VALUE:%.*]], ptr align 2 [[BASE:%.*]], <8 x i1> [[TMP1]]) // CHECK-NEXT: ret void // void test_vst1q_p_f16(float16_t *base, float16x8_t value, mve_pred16_t p) @@ -698,7 +698,7 @@ void test_vst1q_p_f16(float16_t *base, float16x8_t value, mve_pred16_t p) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP0]]) -// CHECK-NEXT: call void @llvm.masked.store.v4f32.p0(<4 x float> [[VALUE:%.*]], ptr [[BASE:%.*]], i32 4, <4 x i1> [[TMP1]]) +// CHECK-NEXT: call void @llvm.masked.store.v4f32.p0(<4 x float> [[VALUE:%.*]], ptr align 4 [[BASE:%.*]], <4 x i1> [[TMP1]]) // CHECK-NEXT: ret void // void test_vst1q_p_f32(float32_t *base, float32x4_t value, mve_pred16_t p) @@ -714,7 +714,7 @@ void test_vst1q_p_f32(float32_t *base, float32x4_t value, mve_pred16_t p) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <16 x i1> @llvm.arm.mve.pred.i2v.v16i1(i32 [[TMP0]]) -// CHECK-NEXT: call void @llvm.masked.store.v16i8.p0(<16 x i8> [[VALUE:%.*]], ptr [[BASE:%.*]], i32 1, <16 x i1> [[TMP1]]) +// CHECK-NEXT: call void @llvm.masked.store.v16i8.p0(<16 x i8> [[VALUE:%.*]], ptr align 1 [[BASE:%.*]], <16 x i1> [[TMP1]]) // CHECK-NEXT: ret void // void test_vst1q_p_s8(int8_t *base, int8x16_t value, mve_pred16_t p) @@ -730,7 +730,7 @@ void test_vst1q_p_s8(int8_t *base, int8x16_t value, mve_pred16_t p) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <8 x i1> @llvm.arm.mve.pred.i2v.v8i1(i32 [[TMP0]]) -// CHECK-NEXT: call void @llvm.masked.store.v8i16.p0(<8 x i16> [[VALUE:%.*]], ptr [[BASE:%.*]], i32 2, <8 x i1> [[TMP1]]) +// CHECK-NEXT: call void @llvm.masked.store.v8i16.p0(<8 x i16> [[VALUE:%.*]], ptr align 2 [[BASE:%.*]], <8 x i1> [[TMP1]]) // CHECK-NEXT: ret void // void test_vst1q_p_s16(int16_t *base, int16x8_t value, mve_pred16_t p) @@ -746,7 +746,7 @@ void test_vst1q_p_s16(int16_t *base, int16x8_t value, mve_pred16_t p) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP0]]) -// CHECK-NEXT: call void @llvm.masked.store.v4i32.p0(<4 x i32> [[VALUE:%.*]], ptr [[BASE:%.*]], i32 4, <4 x i1> [[TMP1]]) +// CHECK-NEXT: call void @llvm.masked.store.v4i32.p0(<4 x i32> [[VALUE:%.*]], ptr align 4 [[BASE:%.*]], <4 x i1> [[TMP1]]) // CHECK-NEXT: ret void // void test_vst1q_p_s32(int32_t *base, int32x4_t value, mve_pred16_t p) @@ -762,7 +762,7 @@ void test_vst1q_p_s32(int32_t *base, int32x4_t value, mve_pred16_t p) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <16 x i1> @llvm.arm.mve.pred.i2v.v16i1(i32 [[TMP0]]) -// CHECK-NEXT: call void @llvm.masked.store.v16i8.p0(<16 x i8> [[VALUE:%.*]], ptr [[BASE:%.*]], i32 1, <16 x i1> [[TMP1]]) +// CHECK-NEXT: call void @llvm.masked.store.v16i8.p0(<16 x i8> [[VALUE:%.*]], ptr align 1 [[BASE:%.*]], <16 x i1> [[TMP1]]) // CHECK-NEXT: ret void // void test_vst1q_p_u8(uint8_t *base, uint8x16_t value, mve_pred16_t p) @@ -778,7 +778,7 @@ void test_vst1q_p_u8(uint8_t *base, uint8x16_t value, mve_pred16_t p) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <8 x i1> @llvm.arm.mve.pred.i2v.v8i1(i32 [[TMP0]]) -// CHECK-NEXT: call void @llvm.masked.store.v8i16.p0(<8 x i16> [[VALUE:%.*]], ptr [[BASE:%.*]], i32 2, <8 x i1> [[TMP1]]) +// CHECK-NEXT: call void @llvm.masked.store.v8i16.p0(<8 x i16> [[VALUE:%.*]], ptr align 2 [[BASE:%.*]], <8 x i1> [[TMP1]]) // CHECK-NEXT: ret void // void test_vst1q_p_u16(uint16_t *base, uint16x8_t value, mve_pred16_t p) @@ -794,7 +794,7 @@ void test_vst1q_p_u16(uint16_t *base, uint16x8_t value, mve_pred16_t p) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP0]]) -// CHECK-NEXT: call void @llvm.masked.store.v4i32.p0(<4 x i32> [[VALUE:%.*]], ptr [[BASE:%.*]], i32 4, <4 x i1> [[TMP1]]) +// CHECK-NEXT: call void @llvm.masked.store.v4i32.p0(<4 x i32> [[VALUE:%.*]], ptr align 4 [[BASE:%.*]], <4 x i1> [[TMP1]]) // CHECK-NEXT: ret void // void test_vst1q_p_u32(uint32_t *base, uint32x4_t value, mve_pred16_t p) @@ -898,7 +898,7 @@ void test_vstrbq_u32(uint8_t *base, uint32x4_t value) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <16 x i1> @llvm.arm.mve.pred.i2v.v16i1(i32 [[TMP0]]) -// CHECK-NEXT: call void @llvm.masked.store.v16i8.p0(<16 x i8> [[VALUE:%.*]], ptr [[BASE:%.*]], i32 1, <16 x i1> [[TMP1]]) +// CHECK-NEXT: call void @llvm.masked.store.v16i8.p0(<16 x i8> [[VALUE:%.*]], ptr align 1 [[BASE:%.*]], <16 x i1> [[TMP1]]) // CHECK-NEXT: ret void // void test_vstrbq_p_s8(int8_t *base, int8x16_t value, mve_pred16_t p) @@ -915,7 +915,7 @@ void test_vstrbq_p_s8(int8_t *base, int8x16_t value, mve_pred16_t p) // CHECK-NEXT: [[TMP0:%.*]] = trunc <8 x i16> [[VALUE:%.*]] to <8 x i8> // CHECK-NEXT: [[TMP1:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP2:%.*]] = call <8 x i1> @llvm.arm.mve.pred.i2v.v8i1(i32 [[TMP1]]) -// CHECK-NEXT: call void @llvm.masked.store.v8i8.p0(<8 x i8> [[TMP0]], ptr [[BASE:%.*]], i32 1, <8 x i1> [[TMP2]]) +// CHECK-NEXT: call void @llvm.masked.store.v8i8.p0(<8 x i8> [[TMP0]], ptr align 1 [[BASE:%.*]], <8 x i1> [[TMP2]]) // CHECK-NEXT: ret void // void test_vstrbq_p_s16(int8_t *base, int16x8_t value, mve_pred16_t p) @@ -932,7 +932,7 @@ void test_vstrbq_p_s16(int8_t *base, int16x8_t value, mve_pred16_t p) // CHECK-NEXT: [[TMP0:%.*]] = trunc <4 x i32> [[VALUE:%.*]] to <4 x i8> // CHECK-NEXT: [[TMP1:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP2:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP1]]) -// CHECK-NEXT: call void @llvm.masked.store.v4i8.p0(<4 x i8> [[TMP0]], ptr [[BASE:%.*]], i32 1, <4 x i1> [[TMP2]]) +// CHECK-NEXT: call void @llvm.masked.store.v4i8.p0(<4 x i8> [[TMP0]], ptr align 1 [[BASE:%.*]], <4 x i1> [[TMP2]]) // CHECK-NEXT: ret void // void test_vstrbq_p_s32(int8_t *base, int32x4_t value, mve_pred16_t p) @@ -948,7 +948,7 @@ void test_vstrbq_p_s32(int8_t *base, int32x4_t value, mve_pred16_t p) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <16 x i1> @llvm.arm.mve.pred.i2v.v16i1(i32 [[TMP0]]) -// CHECK-NEXT: call void @llvm.masked.store.v16i8.p0(<16 x i8> [[VALUE:%.*]], ptr [[BASE:%.*]], i32 1, <16 x i1> [[TMP1]]) +// CHECK-NEXT: call void @llvm.masked.store.v16i8.p0(<16 x i8> [[VALUE:%.*]], ptr align 1 [[BASE:%.*]], <16 x i1> [[TMP1]]) // CHECK-NEXT: ret void // void test_vstrbq_p_u8(uint8_t *base, uint8x16_t value, mve_pred16_t p) @@ -965,7 +965,7 @@ void test_vstrbq_p_u8(uint8_t *base, uint8x16_t value, mve_pred16_t p) // CHECK-NEXT: [[TMP0:%.*]] = trunc <8 x i16> [[VALUE:%.*]] to <8 x i8> // CHECK-NEXT: [[TMP1:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP2:%.*]] = call <8 x i1> @llvm.arm.mve.pred.i2v.v8i1(i32 [[TMP1]]) -// CHECK-NEXT: call void @llvm.masked.store.v8i8.p0(<8 x i8> [[TMP0]], ptr [[BASE:%.*]], i32 1, <8 x i1> [[TMP2]]) +// CHECK-NEXT: call void @llvm.masked.store.v8i8.p0(<8 x i8> [[TMP0]], ptr align 1 [[BASE:%.*]], <8 x i1> [[TMP2]]) // CHECK-NEXT: ret void // void test_vstrbq_p_u16(uint8_t *base, uint16x8_t value, mve_pred16_t p) @@ -982,7 +982,7 @@ void test_vstrbq_p_u16(uint8_t *base, uint16x8_t value, mve_pred16_t p) // CHECK-NEXT: [[TMP0:%.*]] = trunc <4 x i32> [[VALUE:%.*]] to <4 x i8> // CHECK-NEXT: [[TMP1:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP2:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP1]]) -// CHECK-NEXT: call void @llvm.masked.store.v4i8.p0(<4 x i8> [[TMP0]], ptr [[BASE:%.*]], i32 1, <4 x i1> [[TMP2]]) +// CHECK-NEXT: call void @llvm.masked.store.v4i8.p0(<4 x i8> [[TMP0]], ptr align 1 [[BASE:%.*]], <4 x i1> [[TMP2]]) // CHECK-NEXT: ret void // void test_vstrbq_p_u32(uint8_t *base, uint32x4_t value, mve_pred16_t p) @@ -1070,7 +1070,7 @@ void test_vstrhq_u32(uint16_t *base, uint32x4_t value) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <8 x i1> @llvm.arm.mve.pred.i2v.v8i1(i32 [[TMP0]]) -// CHECK-NEXT: call void @llvm.masked.store.v8f16.p0(<8 x half> [[VALUE:%.*]], ptr [[BASE:%.*]], i32 2, <8 x i1> [[TMP1]]) +// CHECK-NEXT: call void @llvm.masked.store.v8f16.p0(<8 x half> [[VALUE:%.*]], ptr align 2 [[BASE:%.*]], <8 x i1> [[TMP1]]) // CHECK-NEXT: ret void // void test_vstrhq_p_f16(float16_t *base, float16x8_t value, mve_pred16_t p) @@ -1086,7 +1086,7 @@ void test_vstrhq_p_f16(float16_t *base, float16x8_t value, mve_pred16_t p) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <8 x i1> @llvm.arm.mve.pred.i2v.v8i1(i32 [[TMP0]]) -// CHECK-NEXT: call void @llvm.masked.store.v8i16.p0(<8 x i16> [[VALUE:%.*]], ptr [[BASE:%.*]], i32 2, <8 x i1> [[TMP1]]) +// CHECK-NEXT: call void @llvm.masked.store.v8i16.p0(<8 x i16> [[VALUE:%.*]], ptr align 2 [[BASE:%.*]], <8 x i1> [[TMP1]]) // CHECK-NEXT: ret void // void test_vstrhq_p_s16(int16_t *base, int16x8_t value, mve_pred16_t p) @@ -1103,7 +1103,7 @@ void test_vstrhq_p_s16(int16_t *base, int16x8_t value, mve_pred16_t p) // CHECK-NEXT: [[TMP0:%.*]] = trunc <4 x i32> [[VALUE:%.*]] to <4 x i16> // CHECK-NEXT: [[TMP1:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP2:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP1]]) -// CHECK-NEXT: call void @llvm.masked.store.v4i16.p0(<4 x i16> [[TMP0]], ptr [[BASE:%.*]], i32 2, <4 x i1> [[TMP2]]) +// CHECK-NEXT: call void @llvm.masked.store.v4i16.p0(<4 x i16> [[TMP0]], ptr align 2 [[BASE:%.*]], <4 x i1> [[TMP2]]) // CHECK-NEXT: ret void // void test_vstrhq_p_s32(int16_t *base, int32x4_t value, mve_pred16_t p) @@ -1119,7 +1119,7 @@ void test_vstrhq_p_s32(int16_t *base, int32x4_t value, mve_pred16_t p) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <8 x i1> @llvm.arm.mve.pred.i2v.v8i1(i32 [[TMP0]]) -// CHECK-NEXT: call void @llvm.masked.store.v8i16.p0(<8 x i16> [[VALUE:%.*]], ptr [[BASE:%.*]], i32 2, <8 x i1> [[TMP1]]) +// CHECK-NEXT: call void @llvm.masked.store.v8i16.p0(<8 x i16> [[VALUE:%.*]], ptr align 2 [[BASE:%.*]], <8 x i1> [[TMP1]]) // CHECK-NEXT: ret void // void test_vstrhq_p_u16(uint16_t *base, uint16x8_t value, mve_pred16_t p) @@ -1136,7 +1136,7 @@ void test_vstrhq_p_u16(uint16_t *base, uint16x8_t value, mve_pred16_t p) // CHECK-NEXT: [[TMP0:%.*]] = trunc <4 x i32> [[VALUE:%.*]] to <4 x i16> // CHECK-NEXT: [[TMP1:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP2:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP1]]) -// CHECK-NEXT: call void @llvm.masked.store.v4i16.p0(<4 x i16> [[TMP0]], ptr [[BASE:%.*]], i32 2, <4 x i1> [[TMP2]]) +// CHECK-NEXT: call void @llvm.masked.store.v4i16.p0(<4 x i16> [[TMP0]], ptr align 2 [[BASE:%.*]], <4 x i1> [[TMP2]]) // CHECK-NEXT: ret void // void test_vstrhq_p_u32(uint16_t *base, uint32x4_t value, mve_pred16_t p) @@ -1194,7 +1194,7 @@ void test_vstrwq_u32(uint32_t *base, uint32x4_t value) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP0]]) -// CHECK-NEXT: call void @llvm.masked.store.v4f32.p0(<4 x float> [[VALUE:%.*]], ptr [[BASE:%.*]], i32 4, <4 x i1> [[TMP1]]) +// CHECK-NEXT: call void @llvm.masked.store.v4f32.p0(<4 x float> [[VALUE:%.*]], ptr align 4 [[BASE:%.*]], <4 x i1> [[TMP1]]) // CHECK-NEXT: ret void // void test_vstrwq_p_f32(float32_t *base, float32x4_t value, mve_pred16_t p) @@ -1210,7 +1210,7 @@ void test_vstrwq_p_f32(float32_t *base, float32x4_t value, mve_pred16_t p) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP0]]) -// CHECK-NEXT: call void @llvm.masked.store.v4i32.p0(<4 x i32> [[VALUE:%.*]], ptr [[BASE:%.*]], i32 4, <4 x i1> [[TMP1]]) +// CHECK-NEXT: call void @llvm.masked.store.v4i32.p0(<4 x i32> [[VALUE:%.*]], ptr align 4 [[BASE:%.*]], <4 x i1> [[TMP1]]) // CHECK-NEXT: ret void // void test_vstrwq_p_s32(int32_t *base, int32x4_t value, mve_pred16_t p) @@ -1226,7 +1226,7 @@ void test_vstrwq_p_s32(int32_t *base, int32x4_t value, mve_pred16_t p) // CHECK-NEXT: entry: // CHECK-NEXT: [[TMP0:%.*]] = zext i16 [[P:%.*]] to i32 // CHECK-NEXT: [[TMP1:%.*]] = call <4 x i1> @llvm.arm.mve.pred.i2v.v4i1(i32 [[TMP0]]) -// CHECK-NEXT: call void @llvm.masked.store.v4i32.p0(<4 x i32> [[VALUE:%.*]], ptr [[BASE:%.*]], i32 4, <4 x i1> [[TMP1]]) +// CHECK-NEXT: call void @llvm.masked.store.v4i32.p0(<4 x i32> [[VALUE:%.*]], ptr align 4 [[BASE:%.*]], <4 x i1> [[TMP1]]) // CHECK-NEXT: ret void // void test_vstrwq_p_u32(uint32_t *base, uint32x4_t value, mve_pred16_t p) diff --git a/clang/test/CodeGen/builtin-masked.c b/clang/test/CodeGen/builtin-masked.c index e2b5e09..28b94b7 100644 --- a/clang/test/CodeGen/builtin-masked.c +++ b/clang/test/CodeGen/builtin-masked.c @@ -19,7 +19,7 @@ typedef _Bool v8b __attribute__((ext_vector_type(8))); // CHECK-NEXT: [[LOAD_BITS2:%.*]] = load i8, ptr [[M_ADDR]], align 1 // CHECK-NEXT: [[TMP1:%.*]] = bitcast i8 [[LOAD_BITS2]] to <8 x i1> // CHECK-NEXT: [[TMP2:%.*]] = load ptr, ptr [[P_ADDR]], align 8 -// CHECK-NEXT: [[MASKED_LOAD:%.*]] = call <8 x i32> @llvm.masked.load.v8i32.p0(ptr [[TMP2]], i32 4, <8 x i1> [[TMP1]], <8 x i32> poison) +// CHECK-NEXT: [[MASKED_LOAD:%.*]] = call <8 x i32> @llvm.masked.load.v8i32.p0(ptr align 4 [[TMP2]], <8 x i1> [[TMP1]], <8 x i32> poison) // CHECK-NEXT: ret <8 x i32> [[MASKED_LOAD]] // v8i test_load(v8b m, int *p) { @@ -45,7 +45,7 @@ v8i test_load(v8b m, int *p) { // CHECK-NEXT: [[TMP2:%.*]] = bitcast i8 [[LOAD_BITS2]] to <8 x i1> // CHECK-NEXT: [[TMP3:%.*]] = load ptr, ptr [[P_ADDR]], align 8 // CHECK-NEXT: [[TMP4:%.*]] = load <8 x i32>, ptr [[T_ADDR]], align 32 -// CHECK-NEXT: [[MASKED_LOAD:%.*]] = call <8 x i32> @llvm.masked.load.v8i32.p0(ptr [[TMP3]], i32 4, <8 x i1> [[TMP2]], <8 x i32> [[TMP4]]) +// CHECK-NEXT: [[MASKED_LOAD:%.*]] = call <8 x i32> @llvm.masked.load.v8i32.p0(ptr align 4 [[TMP3]], <8 x i1> [[TMP2]], <8 x i32> [[TMP4]]) // CHECK-NEXT: ret <8 x i32> [[MASKED_LOAD]] // v8i test_load_passthru(v8b m, int *p, v8i t) { @@ -97,7 +97,7 @@ v8i test_load_expand(v8b m, int *p, v8i t) { // CHECK-NEXT: [[TMP2:%.*]] = bitcast i8 [[LOAD_BITS2]] to <8 x i1> // CHECK-NEXT: [[TMP3:%.*]] = load <8 x i32>, ptr [[V_ADDR]], align 32 // CHECK-NEXT: [[TMP4:%.*]] = load ptr, ptr [[P_ADDR]], align 8 -// CHECK-NEXT: call void @llvm.masked.store.v8i32.p0(<8 x i32> [[TMP3]], ptr [[TMP4]], i32 4, <8 x i1> [[TMP2]]) +// CHECK-NEXT: call void @llvm.masked.store.v8i32.p0(<8 x i32> [[TMP3]], ptr align 4 [[TMP4]], <8 x i1> [[TMP2]]) // CHECK-NEXT: ret void // void test_store(v8b m, v8i v, int *p) { @@ -150,7 +150,7 @@ void test_compress_store(v8b m, v8i v, int *p) { // CHECK-NEXT: [[TMP3:%.*]] = load <8 x i32>, ptr [[IDX_ADDR]], align 32 // CHECK-NEXT: [[TMP4:%.*]] = load ptr, ptr [[PTR_ADDR]], align 8 // CHECK-NEXT: [[TMP5:%.*]] = getelementptr i32, ptr [[TMP4]], <8 x i32> [[TMP3]] -// CHECK-NEXT: [[MASKED_GATHER:%.*]] = call <8 x i32> @llvm.masked.gather.v8i32.v8p0(<8 x ptr> [[TMP5]], i32 4, <8 x i1> [[TMP2]], <8 x i32> poison) +// CHECK-NEXT: [[MASKED_GATHER:%.*]] = call <8 x i32> @llvm.masked.gather.v8i32.v8p0(<8 x ptr> align 4 [[TMP5]], <8 x i1> [[TMP2]], <8 x i32> poison) // CHECK-NEXT: ret <8 x i32> [[MASKED_GATHER]] // v8i test_gather(v8b mask, v8i idx, int *ptr) { @@ -181,7 +181,7 @@ v8i test_gather(v8b mask, v8i idx, int *ptr) { // CHECK-NEXT: [[TMP5:%.*]] = load <8 x i32>, ptr [[IDX_ADDR]], align 32 // CHECK-NEXT: [[TMP6:%.*]] = load ptr, ptr [[PTR_ADDR]], align 8 // CHECK-NEXT: [[TMP7:%.*]] = getelementptr i32, ptr [[TMP6]], <8 x i32> [[TMP4]] -// CHECK-NEXT: call void @llvm.masked.scatter.v8i32.v8p0(<8 x i32> [[TMP5]], <8 x ptr> [[TMP7]], i32 4, <8 x i1> [[TMP3]]) +// CHECK-NEXT: call void @llvm.masked.scatter.v8i32.v8p0(<8 x i32> [[TMP5]], <8 x ptr> align 4 [[TMP7]], <8 x i1> [[TMP3]]) // CHECK-NEXT: ret void // void test_scatter(v8b mask, v8i val, v8i idx, int *ptr) { @@ -203,7 +203,7 @@ void test_scatter(v8b mask, v8i val, v8i idx, int *ptr) { // CHECK-NEXT: [[LOAD_BITS2:%.*]] = load i8, ptr [[MASK_ADDR]], align 1 // CHECK-NEXT: [[TMP1:%.*]] = bitcast i8 [[LOAD_BITS2]] to <8 x i1> // CHECK-NEXT: [[TMP2:%.*]] = load ptr addrspace(42), ptr [[PTR_ADDR]], align 8 -// CHECK-NEXT: [[MASKED_LOAD:%.*]] = call <8 x i32> @llvm.masked.load.v8i32.p42(ptr addrspace(42) [[TMP2]], i32 4, <8 x i1> [[TMP1]], <8 x i32> poison) +// CHECK-NEXT: [[MASKED_LOAD:%.*]] = call <8 x i32> @llvm.masked.load.v8i32.p42(ptr addrspace(42) align 4 [[TMP2]], <8 x i1> [[TMP1]], <8 x i32> poison) // CHECK-NEXT: ret <8 x i32> [[MASKED_LOAD]] // v8i test_load_as(v8b mask, int __attribute__((address_space(42))) * ptr) { @@ -229,7 +229,7 @@ v8i test_load_as(v8b mask, int __attribute__((address_space(42))) * ptr) { // CHECK-NEXT: [[TMP2:%.*]] = bitcast i8 [[LOAD_BITS2]] to <8 x i1> // CHECK-NEXT: [[TMP3:%.*]] = load <8 x i32>, ptr [[V_ADDR]], align 32 // CHECK-NEXT: [[TMP4:%.*]] = load ptr addrspace(42), ptr [[P_ADDR]], align 8 -// CHECK-NEXT: call void @llvm.masked.store.v8i32.p42(<8 x i32> [[TMP3]], ptr addrspace(42) [[TMP4]], i32 4, <8 x i1> [[TMP2]]) +// CHECK-NEXT: call void @llvm.masked.store.v8i32.p42(<8 x i32> [[TMP3]], ptr addrspace(42) align 4 [[TMP4]], <8 x i1> [[TMP2]]) // CHECK-NEXT: ret void // void test_store_as(v8b m, v8i v, int __attribute__((address_space(42))) *p) { @@ -256,7 +256,7 @@ void test_store_as(v8b m, v8i v, int __attribute__((address_space(42))) *p) { // CHECK-NEXT: [[TMP3:%.*]] = load <8 x i32>, ptr [[IDX_ADDR]], align 32 // CHECK-NEXT: [[TMP4:%.*]] = load ptr addrspace(42), ptr [[PTR_ADDR]], align 8 // CHECK-NEXT: [[TMP5:%.*]] = getelementptr i32, ptr addrspace(42) [[TMP4]], <8 x i32> [[TMP3]] -// CHECK-NEXT: [[MASKED_GATHER:%.*]] = call <8 x i32> @llvm.masked.gather.v8i32.v8p42(<8 x ptr addrspace(42)> [[TMP5]], i32 4, <8 x i1> [[TMP2]], <8 x i32> poison) +// CHECK-NEXT: [[MASKED_GATHER:%.*]] = call <8 x i32> @llvm.masked.gather.v8i32.v8p42(<8 x ptr addrspace(42)> align 4 [[TMP5]], <8 x i1> [[TMP2]], <8 x i32> poison) // CHECK-NEXT: ret <8 x i32> [[MASKED_GATHER]] // v8i test_gather_as(v8b mask, v8i idx, int __attribute__((address_space(42))) *ptr) { @@ -287,7 +287,7 @@ v8i test_gather_as(v8b mask, v8i idx, int __attribute__((address_space(42))) *pt // CHECK-NEXT: [[TMP5:%.*]] = load <8 x i32>, ptr [[IDX_ADDR]], align 32 // CHECK-NEXT: [[TMP6:%.*]] = load ptr addrspace(42), ptr [[PTR_ADDR]], align 8 // CHECK-NEXT: [[TMP7:%.*]] = getelementptr i32, ptr addrspace(42) [[TMP6]], <8 x i32> [[TMP4]] -// CHECK-NEXT: call void @llvm.masked.scatter.v8i32.v8p42(<8 x i32> [[TMP5]], <8 x ptr addrspace(42)> [[TMP7]], i32 4, <8 x i1> [[TMP3]]) +// CHECK-NEXT: call void @llvm.masked.scatter.v8i32.v8p42(<8 x i32> [[TMP5]], <8 x ptr addrspace(42)> align 4 [[TMP7]], <8 x i1> [[TMP3]]) // CHECK-NEXT: ret void // void test_scatter_as(v8b mask, v8i val, v8i idx, int __attribute__((address_space(42))) *ptr) { diff --git a/clang/test/CodeGen/c11atomics-ios.c b/clang/test/CodeGen/c11atomics-ios.c index 8310270..34843ef 100644 --- a/clang/test/CodeGen/c11atomics-ios.c +++ b/clang/test/CodeGen/c11atomics-ios.c @@ -235,6 +235,7 @@ _Bool test_promoted_cmpxchg(_Atomic(PS) *addr, PS *desired, PS *new) { // CHECK: [[ATOMIC_DESIRED:%.*]] = alloca { %struct.PS, [2 x i8] }, align 8 // CHECK: [[ATOMIC_NEW:%.*]] = alloca { %struct.PS, [2 x i8] }, align 8 // CHECK: [[RES_ADDR:%.*]] = alloca i8, align 1 + // CHECK: [[OLD_TMP:%.*]] = alloca i64, align 8 // CHECK: store ptr %addr, ptr [[ADDR_ARG]], align 4 // CHECK: store ptr %desired, ptr [[DESIRED_ARG]], align 4 // CHECK: store ptr %new, ptr [[NEW_ARG]], align 4 @@ -251,7 +252,8 @@ _Bool test_promoted_cmpxchg(_Atomic(PS) *addr, PS *desired, PS *new) { // CHECK: [[RES_BOOL:%.*]] = extractvalue { i64, i1 } [[RES]], 1 // CHECK: br i1 [[RES_BOOL]], label {{%.*}}, label {{%.*}} - // CHECK: store i64 [[RES_VAL64]], ptr [[ATOMIC_DESIRED]], align 8 + // CHECK: store i64 [[RES_VAL64]], ptr [[OLD_TMP]], align 8 + // CHECK: call void @llvm.memcpy.p0.p0.i64(ptr align 2 [[DESIRED_ARG:%.*]], ptr align 8 [[OLD_TMP]], i64 6, i1 false) // CHECK: br label {{%.*}} // CHECK: [[RES_BOOL8:%.*]] = zext i1 [[RES_BOOL]] to i8 |