Add SME2 builtins for pfalse and ptrue (llvm#71953)

Extend pfalse and ptrue builtins with svcount_t return types to be
enabled for sve2p1 and sme2

See ARM-software/acle#217

Author: MDevereau

clang/include/clang/Basic/arm_sve.td

@@ -1925,8 +1925,6 @@ def SVBGRP_N : SInst<"svbgrp[_n_{d}]", "dda", "UcUsUiUl", MergeNone, "aarch64_sv
 
 let TargetGuard = "sve2p1" in {
 def SVFCLAMP   : SInst<"svclamp[_{d}]", "dddd", "hfd", MergeNone, "aarch64_sve_fclamp", [], []>;
-def SVPTRUE_COUNT  : SInst<"svptrue_{d}", "}v", "QcQsQiQl", MergeNone, "aarch64_sve_ptrue_{d}", [IsOverloadNone], []>;
-def SVPFALSE_COUNT_ALIAS : SInst<"svpfalse_c", "}v", "", MergeNone, "", [IsOverloadNone]>;
 
 def SVPEXT_SINGLE : SInst<"svpext_lane_{d}", "P}i", "QcQsQiQl", MergeNone, "aarch64_sve_pext", [], [ImmCheck<1, ImmCheck0_3>]>;
 def SVPEXT_X2     : SInst<"svpext_lane_{d}_x2", "2.P}i", "QcQsQiQl", MergeNone, "aarch64_sve_pext_x2", [], [ImmCheck<1, ImmCheck0_1>]>;
@@ -2045,6 +2043,12 @@ def SVCNTP_COUNT : SInst<"svcntp_{d}", "n}i", "QcQsQiQl", MergeNone, "aarch64_sv
 defm SVREVD : SInstZPZ<"svrevd", "csilUcUsUiUl", "aarch64_sve_revd">;
 }
 
+let TargetGuard = "sve2p1|sme2" in {
+  //FIXME: Replace IsStreamingCompatible with IsStreamingOrHasSVE2p1 when available
+  def SVPTRUE_COUNT  : SInst<"svptrue_{d}", "}v", "QcQsQiQl", MergeNone, "aarch64_sve_ptrue_{d}", [IsOverloadNone, IsStreamingCompatible], []>;
+
+  def SVPFALSE_COUNT_ALIAS : SInst<"svpfalse_c", "}v", "", MergeNone, "", [IsOverloadNone, IsStreamingCompatible]>;
+}
 
 let TargetGuard = "sve2p1,b16b16" in {
 defm SVMUL_BF  : SInstZPZZ<"svmul",  "b", "aarch64_sve_fmul",   "aarch64_sve_fmul_u">;

clang/lib/Sema/Sema.cpp

@@ -2084,8 +2084,8 @@ void Sema::checkTypeSupport(QualType Ty, SourceLocation Loc, ValueDecl *D) {
     if (Ty->isSVESizelessBuiltinType() && FD && FD->hasBody()) {
       llvm::StringMap<bool> CallerFeatureMap;
       Context.getFunctionFeatureMap(CallerFeatureMap, FD);
-      if (!Builtin::evaluateRequiredTargetFeatures(
-          "sve", CallerFeatureMap))
+      if (!Builtin::evaluateRequiredTargetFeatures("sve", CallerFeatureMap) &&
+          !Builtin::evaluateRequiredTargetFeatures("sme", CallerFeatureMap))
         Diag(D->getLocation(), diag::err_sve_vector_in_non_sve_target) << Ty;
     }
   };

clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_pfalse.c

@@ -1,18 +1,13 @@
 // 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 +sve2p1 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
-// RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s
-// RUN: %clang_cc1 -DSVE_OVERLOADED_FORMS -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -emit-llvm -o - -x c++ %s | opt -S -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
+// 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 -passes=mem2reg,tailcallelim | FileCheck %s -check-prefix=CPP-CHECK
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
-#include <arm_sve.h>
+// RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -disable-O0-optnone -Werror -Wall -o /dev/null %s
 
-#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
+#include <arm_sve.h>
 
 // CHECK-LABEL: @test_svpfalse_c(
 // CHECK-NEXT:  entry:
@@ -24,7 +19,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.convert.from.svbool.taarch64.svcountt(<vscale x 16 x i1> zeroinitializer)
 // CPP-CHECK-NEXT:    ret target("aarch64.svcount") [[TMP0]]
 //
-svcount_t test_svpfalse_c()
+svcount_t test_svpfalse_c(void) __arm_streaming_compatible
 {
-  return SVE_ACLE_FUNC(svpfalse_c,,,)();
+  return svpfalse_c();
 }

clang/test/CodeGen/aarch64-sve2p1-intrinsics/acle_sve2p1_ptrue.c

@@ -1,10 +1,18 @@
 // 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 +sve2p1 -S -O1 -Werror -emit-llvm -o - %s | FileCheck %s
+// RUN: %clang_cc1 -DTEST_SME2 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -O1 -Werror -emit-llvm -o - %s | FileCheck %s
 // RUN: %clang_cc1 -triple aarch64-none-linux-gnu -target-feature +sve2p1 -S -O1 -Werror -emit-llvm -o - -x c++ %s | FileCheck %s -check-prefix=CPP-CHECK
+// RUN: %clang_cc1 -DTEST_SME2 -triple aarch64-none-linux-gnu -target-feature +sme2 -S -O1 -Werror -emit-llvm -o - -x c++ %s | FileCheck %s -check-prefix=CPP-CHECK
 
 #include <arm_sve.h>
 
+#ifndef TEST_SME2
+#define ATTR
+#else
+#define ATTR __arm_streaming
+#endif
+
 // CHECK-LABEL: @test_svptrue_c8(
 // CHECK-NEXT:  entry:
 // CHECK-NEXT:    [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.ptrue.c8()
@@ -15,7 +23,7 @@
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.ptrue.c8()
 // CPP-CHECK-NEXT:    ret target("aarch64.svcount") [[TMP0]]
 //
-svcount_t test_svptrue_c8(void) {
+svcount_t test_svptrue_c8(void) ATTR {
   return svptrue_c8();
 }
 
@@ -29,7 +37,7 @@ svcount_t test_svptrue_c8(void) {
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.ptrue.c16()
 // CPP-CHECK-NEXT:    ret target("aarch64.svcount") [[TMP0]]
 //
-svcount_t test_svptrue_c16(void) {
+svcount_t test_svptrue_c16(void) ATTR {
   return svptrue_c16();
 }
 
@@ -43,7 +51,7 @@ svcount_t test_svptrue_c16(void) {
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.ptrue.c32()
 // CPP-CHECK-NEXT:    ret target("aarch64.svcount") [[TMP0]]
 //
-svcount_t test_svptrue_c32(void) {
+svcount_t test_svptrue_c32(void) ATTR {
   return svptrue_c32();
 }
 
@@ -57,6 +65,6 @@ svcount_t test_svptrue_c32(void) {
 // CPP-CHECK-NEXT:    [[TMP0:%.*]] = tail call target("aarch64.svcount") @llvm.aarch64.sve.ptrue.c64()
 // CPP-CHECK-NEXT:    ret target("aarch64.svcount") [[TMP0]]
 //
-svcount_t test_svptrue_c64(void) {
+svcount_t test_svptrue_c64(void) ATTR {
   return svptrue_c64();
 }