clang/test/CodeGen/AArch64/sme-intrinsics/acle_sme_add-i32.c

   1 // NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py UTC_ARGS: --version 3
   2 // REQUIRES: aarch64-registered-target
   3 // RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64 -target-feature +sme -O1 -Werror -emit-llvm -o - %s | FileCheck %s -check-prefixes=CHECK,CHECK-C
   4 // RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64 -target-feature +sme -O1 -Werror -emit-llvm -o - -x c++ %s | FileCheck %s -check-prefixes=CHECK,CHECK-CXX
   5 // RUN: %clang_cc1 -fclang-abi-compat=latest -DSME_OVERLOADED_FORMS -triple aarch64 -target-feature +sme -O1 -Werror -emit-llvm -o - %s | FileCheck %s -check-prefixes=CHECK,CHECK-C
   6 // RUN: %clang_cc1 -fclang-abi-compat=latest -DSME_OVERLOADED_FORMS -triple aarch64 -target-feature +sme -O1 -Werror -emit-llvm -o - -x c++ %s | FileCheck %s -check-prefixes=CHECK,CHECK-CXX
   7 // RUN: %clang_cc1 -fclang-abi-compat=latest -triple aarch64 -target-feature +sme -S -O1 -Werror -o /dev/null %s
   8
   9 #include <arm_sme.h>
  10
  11 #ifdef SME_OVERLOADED_FORMS
  12 #define SME_ACLE_FUNC(A1,A2_UNUSED,A3) A1##A3
  13 #else
  14 #define SME_ACLE_FUNC(A1,A2,A3) A1##A2##A3
  15 #endif
  16
  17 // CHECK-C-LABEL: define dso_local void @test_svaddha_za32_u32(
  18 // CHECK-C-SAME: <vscale x 16 x i1> [[PN:%.*]], <vscale x 16 x i1> [[PM:%.*]], <vscale x 4 x i32> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0:[0-9]+]] {
  19 // CHECK-C-NEXT:  entry:
  20 // CHECK-C-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PN]])
  21 // CHECK-C-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PM]])
  22 // CHECK-C-NEXT:    tail call void @llvm.aarch64.sme.addha.nxv4i32(i32 0, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i1> [[TMP1]], <vscale x 4 x i32> [[ZN]])
  23 // CHECK-C-NEXT:    ret void
  24 //
  25 // CHECK-CXX-LABEL: define dso_local void @_Z21test_svaddha_za32_u32u10__SVBool_tS_u12__SVUint32_t(
  26 // CHECK-CXX-SAME: <vscale x 16 x i1> [[PN:%.*]], <vscale x 16 x i1> [[PM:%.*]], <vscale x 4 x i32> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0:[0-9]+]] {
  27 // CHECK-CXX-NEXT:  entry:
  28 // CHECK-CXX-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PN]])
  29 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PM]])
  30 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.addha.nxv4i32(i32 0, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i1> [[TMP1]], <vscale x 4 x i32> [[ZN]])
  31 // CHECK-CXX-NEXT:    ret void
  32 //
  33 void test_svaddha_za32_u32(svbool_t pn, svbool_t pm, svuint32_t zn) __arm_streaming __arm_inout("za") {
  34   SME_ACLE_FUNC(svaddha_za32, _u32, _m)(0, pn, pm, zn);
  35 }
  36
  37 // CHECK-C-LABEL: define dso_local void @test_svaddha_za32_u32_1(
  38 // CHECK-C-SAME: <vscale x 16 x i1> [[PN:%.*]], <vscale x 16 x i1> [[PM:%.*]], <vscale x 4 x i32> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
  39 // CHECK-C-NEXT:  entry:
  40 // CHECK-C-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PN]])
  41 // CHECK-C-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PM]])
  42 // CHECK-C-NEXT:    tail call void @llvm.aarch64.sme.addha.nxv4i32(i32 3, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i1> [[TMP1]], <vscale x 4 x i32> [[ZN]])
  43 // CHECK-C-NEXT:    ret void
  44 //
  45 // CHECK-CXX-LABEL: define dso_local void @_Z23test_svaddha_za32_u32_1u10__SVBool_tS_u12__SVUint32_t(
  46 // CHECK-CXX-SAME: <vscale x 16 x i1> [[PN:%.*]], <vscale x 16 x i1> [[PM:%.*]], <vscale x 4 x i32> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
  47 // CHECK-CXX-NEXT:  entry:
  48 // CHECK-CXX-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PN]])
  49 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PM]])
  50 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.addha.nxv4i32(i32 3, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i1> [[TMP1]], <vscale x 4 x i32> [[ZN]])
  51 // CHECK-CXX-NEXT:    ret void
  52 //
  53 void test_svaddha_za32_u32_1(svbool_t pn, svbool_t pm, svuint32_t zn) __arm_streaming __arm_inout("za") {
  54   SME_ACLE_FUNC(svaddha_za32, _u32, _m)(3, pn, pm, zn);
  55 }
  56
  57 // CHECK-C-LABEL: define dso_local void @test_svaddha_za32_s32(
  58 // CHECK-C-SAME: <vscale x 16 x i1> [[PN:%.*]], <vscale x 16 x i1> [[PM:%.*]], <vscale x 4 x i32> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
  59 // CHECK-C-NEXT:  entry:
  60 // CHECK-C-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PN]])
  61 // CHECK-C-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PM]])
  62 // CHECK-C-NEXT:    tail call void @llvm.aarch64.sme.addha.nxv4i32(i32 0, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i1> [[TMP1]], <vscale x 4 x i32> [[ZN]])
  63 // CHECK-C-NEXT:    ret void
  64 //
  65 // CHECK-CXX-LABEL: define dso_local void @_Z21test_svaddha_za32_s32u10__SVBool_tS_u11__SVInt32_t(
  66 // CHECK-CXX-SAME: <vscale x 16 x i1> [[PN:%.*]], <vscale x 16 x i1> [[PM:%.*]], <vscale x 4 x i32> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
  67 // CHECK-CXX-NEXT:  entry:
  68 // CHECK-CXX-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PN]])
  69 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PM]])
  70 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.addha.nxv4i32(i32 0, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i1> [[TMP1]], <vscale x 4 x i32> [[ZN]])
  71 // CHECK-CXX-NEXT:    ret void
  72 //
  73 void test_svaddha_za32_s32(svbool_t pn, svbool_t pm, svint32_t zn) __arm_streaming __arm_inout("za") {
  74   SME_ACLE_FUNC(svaddha_za32, _s32, _m)(0, pn, pm, zn);
  75 }
  76
  77 // CHECK-C-LABEL: define dso_local void @test_svaddha_za32_s32_1(
  78 // CHECK-C-SAME: <vscale x 16 x i1> [[PN:%.*]], <vscale x 16 x i1> [[PM:%.*]], <vscale x 4 x i32> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
  79 // CHECK-C-NEXT:  entry:
  80 // CHECK-C-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PN]])
  81 // CHECK-C-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PM]])
  82 // CHECK-C-NEXT:    tail call void @llvm.aarch64.sme.addha.nxv4i32(i32 3, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i1> [[TMP1]], <vscale x 4 x i32> [[ZN]])
  83 // CHECK-C-NEXT:    ret void
  84 //
  85 // CHECK-CXX-LABEL: define dso_local void @_Z23test_svaddha_za32_s32_1u10__SVBool_tS_u11__SVInt32_t(
  86 // CHECK-CXX-SAME: <vscale x 16 x i1> [[PN:%.*]], <vscale x 16 x i1> [[PM:%.*]], <vscale x 4 x i32> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
  87 // CHECK-CXX-NEXT:  entry:
  88 // CHECK-CXX-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PN]])
  89 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PM]])
  90 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.addha.nxv4i32(i32 3, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i1> [[TMP1]], <vscale x 4 x i32> [[ZN]])
  91 // CHECK-CXX-NEXT:    ret void
  92 //
  93 void test_svaddha_za32_s32_1(svbool_t pn, svbool_t pm, svint32_t zn) __arm_streaming __arm_inout("za") {
  94   SME_ACLE_FUNC(svaddha_za32, _s32, _m)(3, pn, pm, zn);
  95 }
  96
  97 // CHECK-C-LABEL: define dso_local void @test_svaddva_za32_u32(
  98 // CHECK-C-SAME: <vscale x 16 x i1> [[PN:%.*]], <vscale x 16 x i1> [[PM:%.*]], <vscale x 4 x i32> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
  99 // CHECK-C-NEXT:  entry:
 100 // CHECK-C-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PN]])
 101 // CHECK-C-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PM]])
 102 // CHECK-C-NEXT:    tail call void @llvm.aarch64.sme.addva.nxv4i32(i32 0, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i1> [[TMP1]], <vscale x 4 x i32> [[ZN]])
 103 // CHECK-C-NEXT:    ret void
 104 //
 105 // CHECK-CXX-LABEL: define dso_local void @_Z21test_svaddva_za32_u32u10__SVBool_tS_u12__SVUint32_t(
 106 // CHECK-CXX-SAME: <vscale x 16 x i1> [[PN:%.*]], <vscale x 16 x i1> [[PM:%.*]], <vscale x 4 x i32> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
 107 // CHECK-CXX-NEXT:  entry:
 108 // CHECK-CXX-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PN]])
 109 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PM]])
 110 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.addva.nxv4i32(i32 0, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i1> [[TMP1]], <vscale x 4 x i32> [[ZN]])
 111 // CHECK-CXX-NEXT:    ret void
 112 //
 113 void test_svaddva_za32_u32(svbool_t pn, svbool_t pm, svuint32_t zn) __arm_streaming __arm_inout("za") {
 114   SME_ACLE_FUNC(svaddva_za32, _u32, _m)(0, pn, pm, zn);
 115 }
 116
 117 // CHECK-C-LABEL: define dso_local void @test_svaddva_za32_u32_1(
 118 // CHECK-C-SAME: <vscale x 16 x i1> [[PN:%.*]], <vscale x 16 x i1> [[PM:%.*]], <vscale x 4 x i32> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
 119 // CHECK-C-NEXT:  entry:
 120 // CHECK-C-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PN]])
 121 // CHECK-C-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PM]])
 122 // CHECK-C-NEXT:    tail call void @llvm.aarch64.sme.addva.nxv4i32(i32 3, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i1> [[TMP1]], <vscale x 4 x i32> [[ZN]])
 123 // CHECK-C-NEXT:    ret void
 124 //
 125 // CHECK-CXX-LABEL: define dso_local void @_Z23test_svaddva_za32_u32_1u10__SVBool_tS_u12__SVUint32_t(
 126 // CHECK-CXX-SAME: <vscale x 16 x i1> [[PN:%.*]], <vscale x 16 x i1> [[PM:%.*]], <vscale x 4 x i32> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
 127 // CHECK-CXX-NEXT:  entry:
 128 // CHECK-CXX-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PN]])
 129 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PM]])
 130 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.addva.nxv4i32(i32 3, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i1> [[TMP1]], <vscale x 4 x i32> [[ZN]])
 131 // CHECK-CXX-NEXT:    ret void
 132 //
 133 void test_svaddva_za32_u32_1(svbool_t pn, svbool_t pm, svuint32_t zn) __arm_streaming __arm_inout("za") {
 134   SME_ACLE_FUNC(svaddva_za32, _u32, _m)(3, pn, pm, zn);
 135 }
 136
 137 // CHECK-C-LABEL: define dso_local void @test_svaddva_za32_s32(
 138 // CHECK-C-SAME: <vscale x 16 x i1> [[PN:%.*]], <vscale x 16 x i1> [[PM:%.*]], <vscale x 4 x i32> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
 139 // CHECK-C-NEXT:  entry:
 140 // CHECK-C-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PN]])
 141 // CHECK-C-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PM]])
 142 // CHECK-C-NEXT:    tail call void @llvm.aarch64.sme.addva.nxv4i32(i32 0, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i1> [[TMP1]], <vscale x 4 x i32> [[ZN]])
 143 // CHECK-C-NEXT:    ret void
 144 //
 145 // CHECK-CXX-LABEL: define dso_local void @_Z21test_svaddva_za32_s32u10__SVBool_tS_u11__SVInt32_t(
 146 // CHECK-CXX-SAME: <vscale x 16 x i1> [[PN:%.*]], <vscale x 16 x i1> [[PM:%.*]], <vscale x 4 x i32> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
 147 // CHECK-CXX-NEXT:  entry:
 148 // CHECK-CXX-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PN]])
 149 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PM]])
 150 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.addva.nxv4i32(i32 0, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i1> [[TMP1]], <vscale x 4 x i32> [[ZN]])
 151 // CHECK-CXX-NEXT:    ret void
 152 //
 153 void test_svaddva_za32_s32(svbool_t pn, svbool_t pm, svint32_t zn) __arm_streaming __arm_inout("za") {
 154   SME_ACLE_FUNC(svaddva_za32, _s32, _m)(0, pn, pm, zn);
 155 }
 156
 157 // CHECK-C-LABEL: define dso_local void @test_svaddva_za32_s32_1(
 158 // CHECK-C-SAME: <vscale x 16 x i1> [[PN:%.*]], <vscale x 16 x i1> [[PM:%.*]], <vscale x 4 x i32> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
 159 // CHECK-C-NEXT:  entry:
 160 // CHECK-C-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PN]])
 161 // CHECK-C-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PM]])
 162 // CHECK-C-NEXT:    tail call void @llvm.aarch64.sme.addva.nxv4i32(i32 3, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i1> [[TMP1]], <vscale x 4 x i32> [[ZN]])
 163 // CHECK-C-NEXT:    ret void
 164 //
 165 // CHECK-CXX-LABEL: define dso_local void @_Z23test_svaddva_za32_s32_1u10__SVBool_tS_u11__SVInt32_t(
 166 // CHECK-CXX-SAME: <vscale x 16 x i1> [[PN:%.*]], <vscale x 16 x i1> [[PM:%.*]], <vscale x 4 x i32> [[ZN:%.*]]) local_unnamed_addr #[[ATTR0]] {
 167 // CHECK-CXX-NEXT:  entry:
 168 // CHECK-CXX-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PN]])
 169 // CHECK-CXX-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x i1> @llvm.aarch64.sve.convert.from.svbool.nxv4i1(<vscale x 16 x i1> [[PM]])
 170 // CHECK-CXX-NEXT:    tail call void @llvm.aarch64.sme.addva.nxv4i32(i32 3, <vscale x 4 x i1> [[TMP0]], <vscale x 4 x i1> [[TMP1]], <vscale x 4 x i32> [[ZN]])
 171 // CHECK-CXX-NEXT:    ret void
 172 //
 173 void test_svaddva_za32_s32_1(svbool_t pn, svbool_t pm, svint32_t zn) __arm_streaming __arm_inout("za") {
 174   SME_ACLE_FUNC(svaddva_za32, _s32, _m)(3, pn, pm, zn);
 175 }
 176 //// NOTE: These prefixes are unused and the list is autogenerated. Do not add tests below this line:
 177 // CHECK: {{.*}}