[clang] [AArch64][SME2] Add builtins for SQDMULH (PR #75326)

[via cfe-commits]

llvmbot wrote:


<!--LLVM PR SUMMARY COMMENT-->

@llvm/pr-subscribers-clang

Author: Dinar Temirbulatov (dtemirbulatov)

<details>
<summary>Changes</summary>

Patch by: Kerry McLaughlin &lt;kerry.mclaughlin@<!-- -->arm.com&gt;

---

Patch is 66.51 KiB, truncated to 20.00 KiB below, full version: 
https://github.com/llvm/llvm-project/pull/75326.diff


2 Files Affected:

- (modified) clang/include/clang/Basic/arm_sve.td (+6) 
- (added) clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_sqdmulh.c (+584) 


``````````diff
diff --git a/clang/include/clang/Basic/arm_sve.td 
b/clang/include/clang/Basic/arm_sve.td
index aa9b105364a51a..7f46f20ba7a1a7 100644
--- a/clang/include/clang/Basic/arm_sve.td
+++ b/clang/include/clang/Basic/arm_sve.td
@@ -2168,6 +2168,12 @@ let TargetGuard = "sme2" in {
 
   def REINTERPRET_SVBOOL_TO_SVCOUNT : Inst<"svreinterpret[_c]", "}P", "Pc", 
MergeNone, "", [IsStreamingCompatible], []>;
   def REINTERPRET_SVCOUNT_TO_SVBOOL : Inst<"svreinterpret[_b]", "P}", "Pc", 
MergeNone, "", [IsStreamingCompatible], []>;
+
+  // SQDMULH
+  def SVSQDMULH_SINGLE_X2 : SInst<"svqdmulh[_single_{d}_x2]", "22d", "csil", 
MergeNone, "aarch64_sve_sqdmulh_single_vgx2", [IsStreaming], []>;
+  def SVSQDMULH_SINGLE_X4 : SInst<"svqdmulh[_single_{d}_x4]", "44d", "csil", 
MergeNone, "aarch64_sve_sqdmulh_single_vgx4", [IsStreaming], []>;
+  def SVSQDMULH_X2        : SInst<"svqdmulh[_{d}_x2]",        "222", "csil", 
MergeNone, "aarch64_sve_sqdmulh_vgx2",        [IsStreaming], []>;
+  def SVSQDMULH_X4        : SInst<"svqdmulh[_{d}_x4]",        "444", "csil", 
MergeNone, "aarch64_sve_sqdmulh_vgx4",        [IsStreaming], []>;
 }
 
 let TargetGuard = "sve2p1" in {
diff --git a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_sqdmulh.c 
b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_sqdmulh.c
new file mode 100644
index 00000000000000..6bbd23ccd32a52
--- /dev/null
+++ b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_sqdmulh.c
@@ -0,0 +1,584 @@
+// NOTE: Assertions have been autogenerated by utils/update_cc_test_checks.py
+// REQUIRES: aarch64-registered-target
+
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme2 -S 
-disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -p 
mem2reg,instcombine,tailcallelim | FileCheck %s
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme2 -S 
-disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -p 
mem2reg,instcombine,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu 
-target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | 
opt -S -p mem2reg,instcombine,tailcallelim | FileCheck %s
+// RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu 
-target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x 
c++ %s | opt -S -p mem2reg,instcombine,tailcallelim | FileCheck %s 
-check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme2 -S 
-disable-O0-optnone -Werror -Wall -o /dev/null %s
+#include <arm_sme_draft_spec_subject_to_change.h>
+
+#ifdef SVE_OVERLOADED_FORMS
+// A simple used,unused... macro, long enough to represent any SVE builtin.
+#define SVE_ACLE_FUNC(A1,A2_UNUSED,A3,A4_UNUSED,A5) A1##A3##A5
+#else
+#define SVE_ACLE_FUNC(A1,A2,A3,A4,A5) A1##A2##A3##A4##A5
+#endif
+
+// Single, x2
+
+// CHECK-LABEL: @test_svqdmulh_single_s8_x2(
+// CHECK-NEXT:  entry:
+// CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 16 x i8> 
@llvm.vector.extract.nxv16i8.nxv32i8(<vscale x 32 x i8> [[ZDN:%.*]], i64 0)
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 16 x i8> 
@llvm.vector.extract.nxv16i8.nxv32i8(<vscale x 32 x i8> [[ZDN]], i64 16)
+// CHECK-NEXT:    [[TMP2:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 
x i8> } @llvm.aarch64.sve.sqdmulh.single.vgx2.nxv16i8(<vscale x 16 x i8> 
[[TMP0]], <vscale x 16 x i8> [[TMP1]], <vscale x 16 x i8> [[ZM:%.*]])
+// CHECK-NEXT:    [[TMP3:%.*]] = extractvalue { <vscale x 16 x i8>, <vscale x 
16 x i8> } [[TMP2]], 0
+// CHECK-NEXT:    [[TMP4:%.*]] = tail call <vscale x 32 x i8> 
@llvm.vector.insert.nxv32i8.nxv16i8(<vscale x 32 x i8> poison, <vscale x 16 x 
i8> [[TMP3]], i64 0)
+// CHECK-NEXT:    [[TMP5:%.*]] = extractvalue { <vscale x 16 x i8>, <vscale x 
16 x i8> } [[TMP2]], 1
+// CHECK-NEXT:    [[TMP6:%.*]] = tail call <vscale x 32 x i8> 
@llvm.vector.insert.nxv32i8.nxv16i8(<vscale x 32 x i8> [[TMP4]], <vscale x 16 x 
i8> [[TMP5]], i64 16)
+// CHECK-NEXT:    ret <vscale x 32 x i8> [[TMP6]]
+//
+// CPP-CHECK-LABEL: @_Z26test_svqdmulh_single_s8_x210svint8x2_tu10__SVInt8_t(
+// CPP-CHECK-NEXT:  entry:
+// CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 16 x i8> 
@llvm.vector.extract.nxv16i8.nxv32i8(<vscale x 32 x i8> [[ZDN:%.*]], i64 0)
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 16 x i8> 
@llvm.vector.extract.nxv16i8.nxv32i8(<vscale x 32 x i8> [[ZDN]], i64 16)
+// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 
16 x i8> } @llvm.aarch64.sve.sqdmulh.single.vgx2.nxv16i8(<vscale x 16 x i8> 
[[TMP0]], <vscale x 16 x i8> [[TMP1]], <vscale x 16 x i8> [[ZM:%.*]])
+// CPP-CHECK-NEXT:    [[TMP3:%.*]] = extractvalue { <vscale x 16 x i8>, 
<vscale x 16 x i8> } [[TMP2]], 0
+// CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call <vscale x 32 x i8> 
@llvm.vector.insert.nxv32i8.nxv16i8(<vscale x 32 x i8> poison, <vscale x 16 x 
i8> [[TMP3]], i64 0)
+// CPP-CHECK-NEXT:    [[TMP5:%.*]] = extractvalue { <vscale x 16 x i8>, 
<vscale x 16 x i8> } [[TMP2]], 1
+// CPP-CHECK-NEXT:    [[TMP6:%.*]] = tail call <vscale x 32 x i8> 
@llvm.vector.insert.nxv32i8.nxv16i8(<vscale x 32 x i8> [[TMP4]], <vscale x 16 x 
i8> [[TMP5]], i64 16)
+// CPP-CHECK-NEXT:    ret <vscale x 32 x i8> [[TMP6]]
+//
+svint8x2_t test_svqdmulh_single_s8_x2(svint8x2_t zdn, svint8_t zm) 
__arm_streaming {
+  return SVE_ACLE_FUNC(svqdmulh,_single_s8_x2,,,)(zdn, zm);
+}
+
+// CHECK-LABEL: @test_svqdmulh_single_s16_x2(
+// CHECK-NEXT:  entry:
+// CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i16> 
@llvm.vector.extract.nxv8i16.nxv16i16(<vscale x 16 x i16> [[ZDN:%.*]], i64 0)
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x i16> 
@llvm.vector.extract.nxv8i16.nxv16i16(<vscale x 16 x i16> [[ZDN]], i64 8)
+// CHECK-NEXT:    [[TMP2:%.*]] = tail call { <vscale x 8 x i16>, <vscale x 8 x 
i16> } @llvm.aarch64.sve.sqdmulh.single.vgx2.nxv8i16(<vscale x 8 x i16> 
[[TMP0]], <vscale x 8 x i16> [[TMP1]], <vscale x 8 x i16> [[ZM:%.*]])
+// CHECK-NEXT:    [[TMP3:%.*]] = extractvalue { <vscale x 8 x i16>, <vscale x 
8 x i16> } [[TMP2]], 0
+// CHECK-NEXT:    [[TMP4:%.*]] = tail call <vscale x 16 x i16> 
@llvm.vector.insert.nxv16i16.nxv8i16(<vscale x 16 x i16> poison, <vscale x 8 x 
i16> [[TMP3]], i64 0)
+// CHECK-NEXT:    [[TMP5:%.*]] = extractvalue { <vscale x 8 x i16>, <vscale x 
8 x i16> } [[TMP2]], 1
+// CHECK-NEXT:    [[TMP6:%.*]] = tail call <vscale x 16 x i16> 
@llvm.vector.insert.nxv16i16.nxv8i16(<vscale x 16 x i16> [[TMP4]], <vscale x 8 
x i16> [[TMP5]], i64 8)
+// CHECK-NEXT:    ret <vscale x 16 x i16> [[TMP6]]
+//
+// CPP-CHECK-LABEL: 
@_Z27test_svqdmulh_single_s16_x211svint16x2_tu11__SVInt16_t(
+// CPP-CHECK-NEXT:  entry:
+// CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i16> 
@llvm.vector.extract.nxv8i16.nxv16i16(<vscale x 16 x i16> [[ZDN:%.*]], i64 0)
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x i16> 
@llvm.vector.extract.nxv8i16.nxv16i16(<vscale x 16 x i16> [[ZDN]], i64 8)
+// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call { <vscale x 8 x i16>, <vscale x 
8 x i16> } @llvm.aarch64.sve.sqdmulh.single.vgx2.nxv8i16(<vscale x 8 x i16> 
[[TMP0]], <vscale x 8 x i16> [[TMP1]], <vscale x 8 x i16> [[ZM:%.*]])
+// CPP-CHECK-NEXT:    [[TMP3:%.*]] = extractvalue { <vscale x 8 x i16>, 
<vscale x 8 x i16> } [[TMP2]], 0
+// CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call <vscale x 16 x i16> 
@llvm.vector.insert.nxv16i16.nxv8i16(<vscale x 16 x i16> poison, <vscale x 8 x 
i16> [[TMP3]], i64 0)
+// CPP-CHECK-NEXT:    [[TMP5:%.*]] = extractvalue { <vscale x 8 x i16>, 
<vscale x 8 x i16> } [[TMP2]], 1
+// CPP-CHECK-NEXT:    [[TMP6:%.*]] = tail call <vscale x 16 x i16> 
@llvm.vector.insert.nxv16i16.nxv8i16(<vscale x 16 x i16> [[TMP4]], <vscale x 8 
x i16> [[TMP5]], i64 8)
+// CPP-CHECK-NEXT:    ret <vscale x 16 x i16> [[TMP6]]
+//
+svint16x2_t test_svqdmulh_single_s16_x2(svint16x2_t zdn, svint16_t zm) 
__arm_streaming {
+  return SVE_ACLE_FUNC(svqdmulh,_single_s16_x2,,,)(zdn, zm);
+}
+
+// CHECK-LABEL: @test_svqdmulh_single_s32_x2(
+// CHECK-NEXT:  entry:
+// CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i32> 
@llvm.vector.extract.nxv4i32.nxv8i32(<vscale x 8 x i32> [[ZDN:%.*]], i64 0)
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x i32> 
@llvm.vector.extract.nxv4i32.nxv8i32(<vscale x 8 x i32> [[ZDN]], i64 4)
+// CHECK-NEXT:    [[TMP2:%.*]] = tail call { <vscale x 4 x i32>, <vscale x 4 x 
i32> } @llvm.aarch64.sve.sqdmulh.single.vgx2.nxv4i32(<vscale x 4 x i32> 
[[TMP0]], <vscale x 4 x i32> [[TMP1]], <vscale x 4 x i32> [[ZM:%.*]])
+// CHECK-NEXT:    [[TMP3:%.*]] = extractvalue { <vscale x 4 x i32>, <vscale x 
4 x i32> } [[TMP2]], 0
+// CHECK-NEXT:    [[TMP4:%.*]] = tail call <vscale x 8 x i32> 
@llvm.vector.insert.nxv8i32.nxv4i32(<vscale x 8 x i32> poison, <vscale x 4 x 
i32> [[TMP3]], i64 0)
+// CHECK-NEXT:    [[TMP5:%.*]] = extractvalue { <vscale x 4 x i32>, <vscale x 
4 x i32> } [[TMP2]], 1
+// CHECK-NEXT:    [[TMP6:%.*]] = tail call <vscale x 8 x i32> 
@llvm.vector.insert.nxv8i32.nxv4i32(<vscale x 8 x i32> [[TMP4]], <vscale x 4 x 
i32> [[TMP5]], i64 4)
+// CHECK-NEXT:    ret <vscale x 8 x i32> [[TMP6]]
+//
+// CPP-CHECK-LABEL: 
@_Z27test_svqdmulh_single_s32_x211svint32x2_tu11__SVInt32_t(
+// CPP-CHECK-NEXT:  entry:
+// CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 4 x i32> 
@llvm.vector.extract.nxv4i32.nxv8i32(<vscale x 8 x i32> [[ZDN:%.*]], i64 0)
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 4 x i32> 
@llvm.vector.extract.nxv4i32.nxv8i32(<vscale x 8 x i32> [[ZDN]], i64 4)
+// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call { <vscale x 4 x i32>, <vscale x 
4 x i32> } @llvm.aarch64.sve.sqdmulh.single.vgx2.nxv4i32(<vscale x 4 x i32> 
[[TMP0]], <vscale x 4 x i32> [[TMP1]], <vscale x 4 x i32> [[ZM:%.*]])
+// CPP-CHECK-NEXT:    [[TMP3:%.*]] = extractvalue { <vscale x 4 x i32>, 
<vscale x 4 x i32> } [[TMP2]], 0
+// CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call <vscale x 8 x i32> 
@llvm.vector.insert.nxv8i32.nxv4i32(<vscale x 8 x i32> poison, <vscale x 4 x 
i32> [[TMP3]], i64 0)
+// CPP-CHECK-NEXT:    [[TMP5:%.*]] = extractvalue { <vscale x 4 x i32>, 
<vscale x 4 x i32> } [[TMP2]], 1
+// CPP-CHECK-NEXT:    [[TMP6:%.*]] = tail call <vscale x 8 x i32> 
@llvm.vector.insert.nxv8i32.nxv4i32(<vscale x 8 x i32> [[TMP4]], <vscale x 4 x 
i32> [[TMP5]], i64 4)
+// CPP-CHECK-NEXT:    ret <vscale x 8 x i32> [[TMP6]]
+//
+svint32x2_t test_svqdmulh_single_s32_x2(svint32x2_t zdn, svint32_t zm) 
__arm_streaming {
+  return SVE_ACLE_FUNC(svqdmulh,_single_s32_x2,,,)(zdn, zm);
+}
+
+// CHECK-LABEL: @test_svqdmulh_single_s64_x2(
+// CHECK-NEXT:  entry:
+// CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i64> 
@llvm.vector.extract.nxv2i64.nxv4i64(<vscale x 4 x i64> [[ZDN:%.*]], i64 0)
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x i64> 
@llvm.vector.extract.nxv2i64.nxv4i64(<vscale x 4 x i64> [[ZDN]], i64 2)
+// CHECK-NEXT:    [[TMP2:%.*]] = tail call { <vscale x 2 x i64>, <vscale x 2 x 
i64> } @llvm.aarch64.sve.sqdmulh.single.vgx2.nxv2i64(<vscale x 2 x i64> 
[[TMP0]], <vscale x 2 x i64> [[TMP1]], <vscale x 2 x i64> [[ZM:%.*]])
+// CHECK-NEXT:    [[TMP3:%.*]] = extractvalue { <vscale x 2 x i64>, <vscale x 
2 x i64> } [[TMP2]], 0
+// CHECK-NEXT:    [[TMP4:%.*]] = tail call <vscale x 4 x i64> 
@llvm.vector.insert.nxv4i64.nxv2i64(<vscale x 4 x i64> poison, <vscale x 2 x 
i64> [[TMP3]], i64 0)
+// CHECK-NEXT:    [[TMP5:%.*]] = extractvalue { <vscale x 2 x i64>, <vscale x 
2 x i64> } [[TMP2]], 1
+// CHECK-NEXT:    [[TMP6:%.*]] = tail call <vscale x 4 x i64> 
@llvm.vector.insert.nxv4i64.nxv2i64(<vscale x 4 x i64> [[TMP4]], <vscale x 2 x 
i64> [[TMP5]], i64 2)
+// CHECK-NEXT:    ret <vscale x 4 x i64> [[TMP6]]
+//
+// CPP-CHECK-LABEL: 
@_Z27test_svqdmulh_single_s64_x211svint64x2_tu11__SVInt64_t(
+// CPP-CHECK-NEXT:  entry:
+// CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 2 x i64> 
@llvm.vector.extract.nxv2i64.nxv4i64(<vscale x 4 x i64> [[ZDN:%.*]], i64 0)
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 2 x i64> 
@llvm.vector.extract.nxv2i64.nxv4i64(<vscale x 4 x i64> [[ZDN]], i64 2)
+// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call { <vscale x 2 x i64>, <vscale x 
2 x i64> } @llvm.aarch64.sve.sqdmulh.single.vgx2.nxv2i64(<vscale x 2 x i64> 
[[TMP0]], <vscale x 2 x i64> [[TMP1]], <vscale x 2 x i64> [[ZM:%.*]])
+// CPP-CHECK-NEXT:    [[TMP3:%.*]] = extractvalue { <vscale x 2 x i64>, 
<vscale x 2 x i64> } [[TMP2]], 0
+// CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call <vscale x 4 x i64> 
@llvm.vector.insert.nxv4i64.nxv2i64(<vscale x 4 x i64> poison, <vscale x 2 x 
i64> [[TMP3]], i64 0)
+// CPP-CHECK-NEXT:    [[TMP5:%.*]] = extractvalue { <vscale x 2 x i64>, 
<vscale x 2 x i64> } [[TMP2]], 1
+// CPP-CHECK-NEXT:    [[TMP6:%.*]] = tail call <vscale x 4 x i64> 
@llvm.vector.insert.nxv4i64.nxv2i64(<vscale x 4 x i64> [[TMP4]], <vscale x 2 x 
i64> [[TMP5]], i64 2)
+// CPP-CHECK-NEXT:    ret <vscale x 4 x i64> [[TMP6]]
+//
+svint64x2_t test_svqdmulh_single_s64_x2(svint64x2_t zdn, svint64_t zm) 
__arm_streaming {
+  return SVE_ACLE_FUNC(svqdmulh,_single_s64_x2,,,)(zdn, zm);
+}
+
+// Single, x4
+
+// CHECK-LABEL: @test_svqdmulh_single_s8_x4(
+// CHECK-NEXT:  entry:
+// CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 16 x i8> 
@llvm.vector.extract.nxv16i8.nxv64i8(<vscale x 64 x i8> [[ZDN:%.*]], i64 0)
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 16 x i8> 
@llvm.vector.extract.nxv16i8.nxv64i8(<vscale x 64 x i8> [[ZDN]], i64 16)
+// CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 16 x i8> 
@llvm.vector.extract.nxv16i8.nxv64i8(<vscale x 64 x i8> [[ZDN]], i64 32)
+// CHECK-NEXT:    [[TMP3:%.*]] = tail call <vscale x 16 x i8> 
@llvm.vector.extract.nxv16i8.nxv64i8(<vscale x 64 x i8> [[ZDN]], i64 48)
+// CHECK-NEXT:    [[TMP4:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 16 
x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } 
@llvm.aarch64.sve.sqdmulh.single.vgx4.nxv16i8(<vscale x 16 x i8> [[TMP0]], 
<vscale x 16 x i8> [[TMP1]], <vscale x 16 x i8> [[TMP2]], <vscale x 16 x i8> 
[[TMP3]], <vscale x 16 x i8> [[ZM:%.*]])
+// CHECK-NEXT:    [[TMP5:%.*]] = extractvalue { <vscale x 16 x i8>, <vscale x 
16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP4]], 0
+// CHECK-NEXT:    [[TMP6:%.*]] = tail call <vscale x 64 x i8> 
@llvm.vector.insert.nxv64i8.nxv16i8(<vscale x 64 x i8> poison, <vscale x 16 x 
i8> [[TMP5]], i64 0)
+// CHECK-NEXT:    [[TMP7:%.*]] = extractvalue { <vscale x 16 x i8>, <vscale x 
16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP4]], 1
+// CHECK-NEXT:    [[TMP8:%.*]] = tail call <vscale x 64 x i8> 
@llvm.vector.insert.nxv64i8.nxv16i8(<vscale x 64 x i8> [[TMP6]], <vscale x 16 x 
i8> [[TMP7]], i64 16)
+// CHECK-NEXT:    [[TMP9:%.*]] = extractvalue { <vscale x 16 x i8>, <vscale x 
16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP4]], 2
+// CHECK-NEXT:    [[TMP10:%.*]] = tail call <vscale x 64 x i8> 
@llvm.vector.insert.nxv64i8.nxv16i8(<vscale x 64 x i8> [[TMP8]], <vscale x 16 x 
i8> [[TMP9]], i64 32)
+// CHECK-NEXT:    [[TMP11:%.*]] = extractvalue { <vscale x 16 x i8>, <vscale x 
16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP4]], 3
+// CHECK-NEXT:    [[TMP12:%.*]] = tail call <vscale x 64 x i8> 
@llvm.vector.insert.nxv64i8.nxv16i8(<vscale x 64 x i8> [[TMP10]], <vscale x 16 
x i8> [[TMP11]], i64 48)
+// CHECK-NEXT:    ret <vscale x 64 x i8> [[TMP12]]
+//
+// CPP-CHECK-LABEL: @_Z26test_svqdmulh_single_s8_x410svint8x4_tu10__SVInt8_t(
+// CPP-CHECK-NEXT:  entry:
+// CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 16 x i8> 
@llvm.vector.extract.nxv16i8.nxv64i8(<vscale x 64 x i8> [[ZDN:%.*]], i64 0)
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 16 x i8> 
@llvm.vector.extract.nxv16i8.nxv64i8(<vscale x 64 x i8> [[ZDN]], i64 16)
+// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 16 x i8> 
@llvm.vector.extract.nxv16i8.nxv64i8(<vscale x 64 x i8> [[ZDN]], i64 32)
+// CPP-CHECK-NEXT:    [[TMP3:%.*]] = tail call <vscale x 16 x i8> 
@llvm.vector.extract.nxv16i8.nxv64i8(<vscale x 64 x i8> [[ZDN]], i64 48)
+// CPP-CHECK-NEXT:    [[TMP4:%.*]] = tail call { <vscale x 16 x i8>, <vscale x 
16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } 
@llvm.aarch64.sve.sqdmulh.single.vgx4.nxv16i8(<vscale x 16 x i8> [[TMP0]], 
<vscale x 16 x i8> [[TMP1]], <vscale x 16 x i8> [[TMP2]], <vscale x 16 x i8> 
[[TMP3]], <vscale x 16 x i8> [[ZM:%.*]])
+// CPP-CHECK-NEXT:    [[TMP5:%.*]] = extractvalue { <vscale x 16 x i8>, 
<vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP4]], 0
+// CPP-CHECK-NEXT:    [[TMP6:%.*]] = tail call <vscale x 64 x i8> 
@llvm.vector.insert.nxv64i8.nxv16i8(<vscale x 64 x i8> poison, <vscale x 16 x 
i8> [[TMP5]], i64 0)
+// CPP-CHECK-NEXT:    [[TMP7:%.*]] = extractvalue { <vscale x 16 x i8>, 
<vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP4]], 1
+// CPP-CHECK-NEXT:    [[TMP8:%.*]] = tail call <vscale x 64 x i8> 
@llvm.vector.insert.nxv64i8.nxv16i8(<vscale x 64 x i8> [[TMP6]], <vscale x 16 x 
i8> [[TMP7]], i64 16)
+// CPP-CHECK-NEXT:    [[TMP9:%.*]] = extractvalue { <vscale x 16 x i8>, 
<vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP4]], 2
+// CPP-CHECK-NEXT:    [[TMP10:%.*]] = tail call <vscale x 64 x i8> 
@llvm.vector.insert.nxv64i8.nxv16i8(<vscale x 64 x i8> [[TMP8]], <vscale x 16 x 
i8> [[TMP9]], i64 32)
+// CPP-CHECK-NEXT:    [[TMP11:%.*]] = extractvalue { <vscale x 16 x i8>, 
<vscale x 16 x i8>, <vscale x 16 x i8>, <vscale x 16 x i8> } [[TMP4]], 3
+// CPP-CHECK-NEXT:    [[TMP12:%.*]] = tail call <vscale x 64 x i8> 
@llvm.vector.insert.nxv64i8.nxv16i8(<vscale x 64 x i8> [[TMP10]], <vscale x 16 
x i8> [[TMP11]], i64 48)
+// CPP-CHECK-NEXT:    ret <vscale x 64 x i8> [[TMP12]]
+//
+svint8x4_t test_svqdmulh_single_s8_x4(svint8x4_t zdn, svint8_t zm) 
__arm_streaming {
+  return SVE_ACLE_FUNC(svqdmulh,_single_s8_x4,,,)(zdn, zm);
+}
+
+// CHECK-LABEL: @test_svqdmulh_single_s16_x4(
+// CHECK-NEXT:  entry:
+// CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i16> 
@llvm.vector.extract.nxv8i16.nxv32i16(<vscale x 32 x i16> [[ZDN:%.*]], i64 0)
+// CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x i16> 
@llvm.vector.extract.nxv8i16.nxv32i16(<vscale x 32 x i16> [[ZDN]], i64 8)
+// CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 8 x i16> 
@llvm.vector.extract.nxv8i16.nxv32i16(<vscale x 32 x i16> [[ZDN]], i64 16)
+// CHECK-NEXT:    [[TMP3:%.*]] = tail call <vscale x 8 x i16> 
@llvm.vector.extract.nxv8i16.nxv32i16(<vscale x 32 x i16> [[ZDN]], i64 24)
+// CHECK-NEXT:    [[TMP4:%.*]] = tail call { <vscale x 8 x i16>, <vscale x 8 x 
i16>, <vscale x 8 x i16>, <vscale x 8 x i16> } 
@llvm.aarch64.sve.sqdmulh.single.vgx4.nxv8i16(<vscale x 8 x i16> [[TMP0]], 
<vscale x 8 x i16> [[TMP1]], <vscale x 8 x i16> [[TMP2]], <vscale x 8 x i16> 
[[TMP3]], <vscale x 8 x i16> [[ZM:%.*]])
+// CHECK-NEXT:    [[TMP5:%.*]] = extractvalue { <vscale x 8 x i16>, <vscale x 
8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16> } [[TMP4]], 0
+// CHECK-NEXT:    [[TMP6:%.*]] = tail call <vscale x 32 x i16> 
@llvm.vector.insert.nxv32i16.nxv8i16(<vscale x 32 x i16> poison, <vscale x 8 x 
i16> [[TMP5]], i64 0)
+// CHECK-NEXT:    [[TMP7:%.*]] = extractvalue { <vscale x 8 x i16>, <vscale x 
8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16> } [[TMP4]], 1
+// CHECK-NEXT:    [[TMP8:%.*]] = tail call <vscale x 32 x i16> 
@llvm.vector.insert.nxv32i16.nxv8i16(<vscale x 32 x i16> [[TMP6]], <vscale x 8 
x i16> [[TMP7]], i64 8)
+// CHECK-NEXT:    [[TMP9:%.*]] = extractvalue { <vscale x 8 x i16>, <vscale x 
8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16> } [[TMP4]], 2
+// CHECK-NEXT:    [[TMP10:%.*]] = tail call <vscale x 32 x i16> 
@llvm.vector.insert.nxv32i16.nxv8i16(<vscale x 32 x i16> [[TMP8]], <vscale x 8 
x i16> [[TMP9]], i64 16)
+// CHECK-NEXT:    [[TMP11:%.*]] = extractvalue { <vscale x 8 x i16>, <vscale x 
8 x i16>, <vscale x 8 x i16>, <vscale x 8 x i16> } [[TMP4]], 3
+// CHECK-NEXT:    [[TMP12:%.*]] = tail call <vscale x 32 x i16> 
@llvm.vector.insert.nxv32i16.nxv8i16(<vscale x 32 x i16> [[TMP10]], <vscale x 8 
x i16> [[TMP11]], i64 24)
+// CHECK-NEXT:    ret <vscale x 32 x i16> [[TMP12]]
+//
+// CPP-CHECK-LABEL: 
@_Z27test_svqdmulh_single_s16_x411svint16x4_tu11__SVInt16_t(
+// CPP-CHECK-NEXT:  entry:
+// CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 8 x i16> 
@llvm.vector.extract.nxv8i16.nxv32i16(<vscale x 32 x i16> [[ZDN:%.*]], i64 0)
+// CPP-CHECK-NEXT:    [[TMP1:%.*]] = tail call <vscale x 8 x i16> 
@llvm.vector.extract.nxv8i16.nxv32i16(<vscale x 32 x i16> [[ZDN]], i64 8)
+// CPP-CHECK-NEXT:    [[TMP2:%.*]] = tail call <vscale x 8 x i16> 
@llvm.vector.extract.nxv8i16.nxv32i16(<vscale x...
[truncated]

``````````

</details>


https://github.com/llvm/llvm-project/pull/75326
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits