[AArch64][SME2] Add builtins to cast svbool from/to svcount.

[dtemirbulatov]

Add builtin: 'svreinterpret_b' to cast from svcount_t to svbool_t.
Add builtin: 'svreinterpret_c' to cast from svbool_t to svcount_t.

Patch by: Hassnaa Hamdi [email protected]

[llvmbot]

@llvm/pr-subscribers-clang-codegen

@llvm/pr-subscribers-clang

Author: Dinar Temirbulatov (dtemirbulatov)

Changes

Add builtin: 'svreinterpret_b' to cast from svcount_t to svbool_t.
Add builtin: 'svreinterpret_c' to cast from svbool_t to svcount_t.

Patch by: Hassnaa Hamdi <[email protected]>

---

Full diff: https://github.com/llvm/llvm-project/pull/74720.diff

3 Files Affected:

• (modified) clang/include/clang/Basic/arm_sve.td (+3)
• (modified) clang/lib/CodeGen/CGBuiltin.cpp (+16)
• (added) clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_reinterpret_svcount_svbool.c (+47)

diff --git a/clang/include/clang/Basic/arm_sve.td b/clang/include/clang/Basic/arm_sve.td
index 3f69a3df9e616..ec68489effefa 100644
--- a/clang/include/clang/Basic/arm_sve.td
+++ b/clang/include/clang/Basic/arm_sve.td
@@ -2130,6 +2130,9 @@ let TargetGuard = "sme2" in {
   def SVURSHL_X2 : SInst<"svrshl[_{d}_x2]", "222", "UcUsUiUl", MergeNone, "aarch64_sve_urshl_x2", [IsStreaming], []>;
   def SVSRSHL_X4 : SInst<"svrshl[_{d}_x4]", "444", "csil",     MergeNone, "aarch64_sve_srshl_x4", [IsStreaming], []>;
   def SVURSHL_X4 : SInst<"svrshl[_{d}_x4]", "444", "UcUsUiUl", MergeNone, "aarch64_sve_urshl_x4", [IsStreaming], []>;
+
+  def REINTERPRET_SVBOOL_TO_SVCOUNT : Inst<"svreinterpret[_c]", "}P", "Pc", MergeNone, "", [IsStreamingCompatible], [], MemEltTyDefault>;
+  def REINTERPRET_SVCOUNT_TO_SVBOOL : Inst<"svreinterpret[_b]", "P}", "Pc", MergeNone, "", [IsStreamingCompatible], [], MemEltTyDefault>;
 }
 
 let TargetGuard = "sve2p1" in {
diff --git a/clang/lib/CodeGen/CGBuiltin.cpp b/clang/lib/CodeGen/CGBuiltin.cpp
index 65d9862621061..b1d8edcd99123 100644
--- a/clang/lib/CodeGen/CGBuiltin.cpp
+++ b/clang/lib/CodeGen/CGBuiltin.cpp
@@ -10214,6 +10214,22 @@ Value *CodeGenFunction::EmitAArch64SVEBuiltinExpr(unsigned BuiltinID,
   switch (BuiltinID) {
   default:
     return nullptr;
+
+  case SVE::BI__builtin_sve_svreinterpret_b: {
+    auto SVCountTy =
+        llvm::TargetExtType::get(getLLVMContext(), "aarch64.svcount");
+    Function *CastFromSVCountF =
+        CGM.getIntrinsic(Intrinsic::aarch64_sve_convert_to_svbool, SVCountTy);
+    return Builder.CreateCall(CastFromSVCountF, Ops[0]);
+  }
+  case SVE::BI__builtin_sve_svreinterpret_c: {
+    auto SVCountTy =
+        llvm::TargetExtType::get(getLLVMContext(), "aarch64.svcount");
+    Function *CastToSVCountF =
+        CGM.getIntrinsic(Intrinsic::aarch64_sve_convert_from_svbool, SVCountTy);
+    return Builder.CreateCall(CastToSVCountF, Ops[0]);
+  }
+
   case SVE::BI__builtin_sve_svpsel_lane_b8:
   case SVE::BI__builtin_sve_svpsel_lane_b16:
   case SVE::BI__builtin_sve_svpsel_lane_b32:
diff --git a/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_reinterpret_svcount_svbool.c b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_reinterpret_svcount_svbool.c
new file mode 100644
index 0000000000000..82b30e8bbe9b9
--- /dev/null
+++ b/clang/test/CodeGen/aarch64-sme2-intrinsics/acle_sme2_reinterpret_svcount_svbool.c
@@ -0,0 +1,47 @@
+// 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 +sve -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -p mem2reg,instcombine,tailcallelim | FileCheck %s
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -p mem2reg,instcombine,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -p mem2reg,instcombine,tailcallelim | FileCheck %s
+// RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -p mem2reg,instcombine,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
+
+#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) A1##A3
+#else
+#define SVE_ACLE_FUNC(A1,A2,A3,A4) A1##A2##A3##A4
+#endif
+
+// CHECK-LABEL: @test_svreinterpret_svbool_svcnt(
+// CHECK-NEXT:  entry:
+// CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 16 x i1> @llvm.aarch64.sve.convert.to.svbool.taarch64.svcountt(target("aarch64.svcount") [[CNT:%.*]])
+// CHECK-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z31test_svreinterpret_svbool_svcntu11__SVCount_t(
+// CPP-CHECK-NEXT:  entry:
+// CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call <vscale x 16 x i1> @llvm.aarch64.sve.convert.to.svbool.taarch64.svcountt(target("aarch64.svcount") [[CNT:%.*]])
+// CPP-CHECK-NEXT:    ret <vscale x 16 x i1> [[TMP0]]
+//
+svbool_t test_svreinterpret_svbool_svcnt(svcount_t cnt) __arm_streaming_compatible
+{
+  return SVE_ACLE_FUNC(svreinterpret,_b,,)(cnt);
+}
+
+// CHECK-LABEL: @test_svreinterpret_svcnt_svbool(
+// CHECK-NEXT:  entry:
+// CHECK-NEXT:    [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.convert.from.svbool.taarch64.svcountt(<vscale x 16 x i1> [[PG:%.*]])
+// CHECK-NEXT:    ret target("aarch64.svcount") [[TMP0]]
+//
+// CPP-CHECK-LABEL: @_Z31test_svreinterpret_svcnt_svboolu10__SVBool_t(
+// CPP-CHECK-NEXT:  entry:
+// CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.convert.from.svbool.taarch64.svcountt(<vscale x 16 x i1> [[PG:%.*]])
+// CPP-CHECK-NEXT:    ret target("aarch64.svcount") [[TMP0]]
+//
+svcount_t test_svreinterpret_svcnt_svbool(svbool_t pg) __arm_streaming_compatible
+{
+  return SVE_ACLE_FUNC(svreinterpret,_c,,)(pg);
+}

[github-actions]

✅ With the latest revision this PR passed the C/C++ code formatter.

[SamTebbs33]

LGTM

[kmclaughlin-arm]

Thank you @dtemirbulatov, LGTM!