SemaARM.cpp

Go to the documentation of this file.

//===------ SemaARM.cpp ---------- ARM target-specific routines -----------===//
//
// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
// See https://llvm.org/LICENSE.txt for license information.
// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
//
//===----------------------------------------------------------------------===//
//
//  This file implements semantic analysis functions specific to ARM.
//
//===----------------------------------------------------------------------===//
 
#include "clang/Sema/SemaARM.h"
#include "clang/Basic/DiagnosticSema.h"
#include "clang/Basic/TargetBuiltins.h"
#include "clang/Basic/TargetInfo.h"
#include "clang/Sema/Initialization.h"
#include "clang/Sema/ParsedAttr.h"
#include "clang/Sema/Sema.h"
 
namespace clang {
 
SemaARM::SemaARM(Sema &S) : SemaBase(S) {}
 
/// BuiltinARMMemoryTaggingCall - Handle calls of memory tagging extensions
bool SemaARM::BuiltinARMMemoryTaggingCall(unsigned BuiltinID,
                                          CallExpr *TheCall) {
  ASTContext &Context = getASTContext();
 
  if (BuiltinID == AArch64::BI__builtin_arm_irg) {
    if (SemaRef.checkArgCount(TheCall, 2))
      return true;
    Expr *Arg0 = TheCall->getArg(0);
    Expr *Arg1 = TheCall->getArg(1);
 
    ExprResult FirstArg = SemaRef.DefaultFunctionArrayLvalueConversion(Arg0);
    if (FirstArg.isInvalid())
      return true;
    QualType FirstArgType = FirstArg.get()->getType();
    if (!FirstArgType->isAnyPointerType())
      return Diag(TheCall->getBeginLoc(), diag::err_memtag_arg_must_be_pointer)
             << "first" << FirstArgType << Arg0->getSourceRange();
    TheCall->setArg(0, FirstArg.get());
 
    ExprResult SecArg = SemaRef.DefaultLvalueConversion(Arg1);
    if (SecArg.isInvalid())
      return true;
    QualType SecArgType = SecArg.get()->getType();
    if (!SecArgType->isIntegerType())
      return Diag(TheCall->getBeginLoc(), diag::err_memtag_arg_must_be_integer)
             << "second" << SecArgType << Arg1->getSourceRange();
 
    // Derive the return type from the pointer argument.
    TheCall->setType(FirstArgType);
    return false;
  }
 
  if (BuiltinID == AArch64::BI__builtin_arm_addg) {
    if (SemaRef.checkArgCount(TheCall, 2))
      return true;
 
    Expr *Arg0 = TheCall->getArg(0);
    ExprResult FirstArg = SemaRef.DefaultFunctionArrayLvalueConversion(Arg0);
    if (FirstArg.isInvalid())
      return true;
    QualType FirstArgType = FirstArg.get()->getType();
    if (!FirstArgType->isAnyPointerType())
      return Diag(TheCall->getBeginLoc(), diag::err_memtag_arg_must_be_pointer)
             << "first" << FirstArgType << Arg0->getSourceRange();
    TheCall->setArg(0, FirstArg.get());
 
    // Derive the return type from the pointer argument.
    TheCall->setType(FirstArgType);
 
    // Second arg must be an constant in range [0,15]
    return SemaRef.BuiltinConstantArgRange(TheCall, 1, 0, 15);
  }
 
  if (BuiltinID == AArch64::BI__builtin_arm_gmi) {
    if (SemaRef.checkArgCount(TheCall, 2))
      return true;
    Expr *Arg0 = TheCall->getArg(0);
    Expr *Arg1 = TheCall->getArg(1);
 
    ExprResult FirstArg = SemaRef.DefaultFunctionArrayLvalueConversion(Arg0);
    if (FirstArg.isInvalid())
      return true;
    QualType FirstArgType = FirstArg.get()->getType();
    if (!FirstArgType->isAnyPointerType())
      return Diag(TheCall->getBeginLoc(), diag::err_memtag_arg_must_be_pointer)
             << "first" << FirstArgType << Arg0->getSourceRange();
 
    QualType SecArgType = Arg1->getType();
    if (!SecArgType->isIntegerType())
      return Diag(TheCall->getBeginLoc(), diag::err_memtag_arg_must_be_integer)
             << "second" << SecArgType << Arg1->getSourceRange();
    TheCall->setType(Context.IntTy);
    return false;
  }
 
  if (BuiltinID == AArch64::BI__builtin_arm_ldg ||
      BuiltinID == AArch64::BI__builtin_arm_stg) {
    if (SemaRef.checkArgCount(TheCall, 1))
      return true;
    Expr *Arg0 = TheCall->getArg(0);
    ExprResult FirstArg = SemaRef.DefaultFunctionArrayLvalueConversion(Arg0);
    if (FirstArg.isInvalid())
      return true;
 
    QualType FirstArgType = FirstArg.get()->getType();
    if (!FirstArgType->isAnyPointerType())
      return Diag(TheCall->getBeginLoc(), diag::err_memtag_arg_must_be_pointer)
             << "first" << FirstArgType << Arg0->getSourceRange();
    TheCall->setArg(0, FirstArg.get());
 
    // Derive the return type from the pointer argument.
    if (BuiltinID == AArch64::BI__builtin_arm_ldg)
      TheCall->setType(FirstArgType);
    return false;
  }
 
  if (BuiltinID == AArch64::BI__builtin_arm_subp) {
    Expr *ArgA = TheCall->getArg(0);
    Expr *ArgB = TheCall->getArg(1);
 
    ExprResult ArgExprA = SemaRef.DefaultFunctionArrayLvalueConversion(ArgA);
    ExprResult ArgExprB = SemaRef.DefaultFunctionArrayLvalueConversion(ArgB);
 
    if (ArgExprA.isInvalid() || ArgExprB.isInvalid())
      return true;
 
    QualType ArgTypeA = ArgExprA.get()->getType();
    QualType ArgTypeB = ArgExprB.get()->getType();
 
    auto isNull = [&](Expr *E) -> bool {
      return E->isNullPointerConstant(Context,
                                      Expr::NPC_ValueDependentIsNotNull);
    };
 
    // argument should be either a pointer or null
    if (!ArgTypeA->isAnyPointerType() && !isNull(ArgA))
      return Diag(TheCall->getBeginLoc(), diag::err_memtag_arg_null_or_pointer)
             << "first" << ArgTypeA << ArgA->getSourceRange();
 
    if (!ArgTypeB->isAnyPointerType() && !isNull(ArgB))
      return Diag(TheCall->getBeginLoc(), diag::err_memtag_arg_null_or_pointer)
             << "second" << ArgTypeB << ArgB->getSourceRange();
 
    // Ensure Pointee types are compatible
    if (ArgTypeA->isAnyPointerType() && !isNull(ArgA) &&
        ArgTypeB->isAnyPointerType() && !isNull(ArgB)) {
      QualType pointeeA = ArgTypeA->getPointeeType();
      QualType pointeeB = ArgTypeB->getPointeeType();
      if (!Context.typesAreCompatible(
              Context.getCanonicalType(pointeeA).getUnqualifiedType(),
              Context.getCanonicalType(pointeeB).getUnqualifiedType())) {
        return Diag(TheCall->getBeginLoc(),
                    diag::err_typecheck_sub_ptr_compatible)
               << ArgTypeA << ArgTypeB << ArgA->getSourceRange()
               << ArgB->getSourceRange();
      }
    }
 
    // at least one argument should be pointer type
    if (!ArgTypeA->isAnyPointerType() && !ArgTypeB->isAnyPointerType())
      return Diag(TheCall->getBeginLoc(), diag::err_memtag_any2arg_pointer)
             << ArgTypeA << ArgTypeB << ArgA->getSourceRange();
 
    if (isNull(ArgA)) // adopt type of the other pointer
      ArgExprA =
          SemaRef.ImpCastExprToType(ArgExprA.get(), ArgTypeB, CK_NullToPointer);
 
    if (isNull(ArgB))
      ArgExprB =
          SemaRef.ImpCastExprToType(ArgExprB.get(), ArgTypeA, CK_NullToPointer);
 
    TheCall->setArg(0, ArgExprA.get());
    TheCall->setArg(1, ArgExprB.get());
    TheCall->setType(Context.LongLongTy);
    return false;
  }
  assert(false && "Unhandled ARM MTE intrinsic");
  return true;
}
 
/// BuiltinARMSpecialReg - Handle a check if argument ArgNum of CallExpr
/// TheCall is an ARM/AArch64 special register string literal.
bool SemaARM::BuiltinARMSpecialReg(unsigned BuiltinID, CallExpr *TheCall,
                                   int ArgNum, unsigned ExpectedFieldNum,
                                   bool AllowName) {
  bool IsARMBuiltin = BuiltinID == ARM::BI__builtin_arm_rsr64 ||
                      BuiltinID == ARM::BI__builtin_arm_wsr64 ||
                      BuiltinID == ARM::BI__builtin_arm_rsr ||
                      BuiltinID == ARM::BI__builtin_arm_rsrp ||
                      BuiltinID == ARM::BI__builtin_arm_wsr ||
                      BuiltinID == ARM::BI__builtin_arm_wsrp;
  bool IsAArch64Builtin = BuiltinID == AArch64::BI__builtin_arm_rsr64 ||
                          BuiltinID == AArch64::BI__builtin_arm_wsr64 ||
                          BuiltinID == AArch64::BI__builtin_arm_rsr128 ||
                          BuiltinID == AArch64::BI__builtin_arm_wsr128 ||
                          BuiltinID == AArch64::BI__builtin_arm_rsr ||
                          BuiltinID == AArch64::BI__builtin_arm_rsrp ||
                          BuiltinID == AArch64::BI__builtin_arm_wsr ||
                          BuiltinID == AArch64::BI__builtin_arm_wsrp;
  assert((IsARMBuiltin || IsAArch64Builtin) && "Unexpected ARM builtin.");
 
  // We can't check the value of a dependent argument.
  Expr *Arg = TheCall->getArg(ArgNum);
  if (Arg->isTypeDependent() || Arg->isValueDependent())
    return false;
 
  // Check if the argument is a string literal.
  if (!isa<StringLiteral>(Arg->IgnoreParenImpCasts()))
    return Diag(TheCall->getBeginLoc(), diag::err_expr_not_string_literal)
           << Arg->getSourceRange();
 
  // Check the type of special register given.
  StringRef Reg = cast<StringLiteral>(Arg->IgnoreParenImpCasts())->getString();
  SmallVector<StringRef, 6> Fields;
  Reg.split(Fields, ":");
 
  if (Fields.size() != ExpectedFieldNum && !(AllowName && Fields.size() == 1))
    return Diag(TheCall->getBeginLoc(), diag::err_arm_invalid_specialreg)
           << Arg->getSourceRange();
 
  // If the string is the name of a register then we cannot check that it is
  // valid here but if the string is of one the forms described in ACLE then we
  // can check that the supplied fields are integers and within the valid
  // ranges.
  if (Fields.size() > 1) {
    bool FiveFields = Fields.size() == 5;
 
    bool ValidString = true;
    if (IsARMBuiltin) {
      ValidString &= Fields[0].starts_with_insensitive("cp") ||
                     Fields[0].starts_with_insensitive("p");
      if (ValidString)
        Fields[0] = Fields[0].drop_front(
            Fields[0].starts_with_insensitive("cp") ? 2 : 1);
 
      ValidString &= Fields[2].starts_with_insensitive("c");
      if (ValidString)
        Fields[2] = Fields[2].drop_front(1);
 
      if (FiveFields) {
        ValidString &= Fields[3].starts_with_insensitive("c");
        if (ValidString)
          Fields[3] = Fields[3].drop_front(1);
      }
    }
 
    SmallVector<int, 5> Ranges;
    if (FiveFields)
      Ranges.append({IsAArch64Builtin ? 1 : 15, 7, 15, 15, 7});
    else
      Ranges.append({15, 7, 15});
 
    for (unsigned i = 0; i < Fields.size(); ++i) {
      int IntField;
      ValidString &= !Fields[i].getAsInteger(10, IntField);
      ValidString &= (IntField >= 0 && IntField <= Ranges[i]);
    }
 
    if (!ValidString)
      return Diag(TheCall->getBeginLoc(), diag::err_arm_invalid_specialreg)
             << Arg->getSourceRange();
  } else if (IsAArch64Builtin && Fields.size() == 1) {
    // This code validates writes to PSTATE registers.
 
    // Not a write.
    if (TheCall->getNumArgs() != 2)
      return false;
 
    // The 128-bit system register accesses do not touch PSTATE.
    if (BuiltinID == AArch64::BI__builtin_arm_rsr128 ||
        BuiltinID == AArch64::BI__builtin_arm_wsr128)
      return false;
 
    // These are the named PSTATE accesses using "MSR (immediate)" instructions,
    // along with the upper limit on the immediates allowed.
    auto MaxLimit = llvm::StringSwitch<std::optional<unsigned>>(Reg)
                        .CaseLower("spsel", 15)
                        .CaseLower("daifclr", 15)
                        .CaseLower("daifset", 15)
                        .CaseLower("pan", 15)
                        .CaseLower("uao", 15)
                        .CaseLower("dit", 15)
                        .CaseLower("ssbs", 15)
                        .CaseLower("tco", 15)
                        .CaseLower("allint", 1)
                        .CaseLower("pm", 1)
                        .Default(std::nullopt);
 
    // If this is not a named PSTATE, just continue without validating, as this
    // will be lowered to an "MSR (register)" instruction directly
    if (!MaxLimit)
      return false;
 
    // Here we only allow constants in the range for that pstate, as required by
    // the ACLE.
    //
    // While clang also accepts the names of system registers in its ACLE
    // intrinsics, we prevent this with the PSTATE names used in MSR (immediate)
    // as the value written via a register is different to the value used as an
    // immediate to have the same effect. e.g., for the instruction `msr tco,
    // x0`, it is bit 25 of register x0 that is written into PSTATE.TCO, but
    // with `msr tco, #imm`, it is bit 0 of xN that is written into PSTATE.TCO.
    //
    // If a programmer wants to codegen the MSR (register) form of `msr tco,
    // xN`, they can still do so by specifying the register using five
    // colon-separated numbers in a string.
    return SemaRef.BuiltinConstantArgRange(TheCall, 1, 0, *MaxLimit);
  }
 
  return false;
}
 
/// getNeonEltType - Return the QualType corresponding to the elements of
/// the vector type specified by the NeonTypeFlags.  This is used to check
/// the pointer arguments for Neon load/store intrinsics.
static QualType getNeonEltType(NeonTypeFlags Flags, ASTContext &Context,
                               bool IsPolyUnsigned, bool IsInt64Long) {
  switch (Flags.getEltType()) {
  case NeonTypeFlags::Int8:
    return Flags.isUnsigned() ? Context.UnsignedCharTy : Context.SignedCharTy;
  case NeonTypeFlags::Int16:
    return Flags.isUnsigned() ? Context.UnsignedShortTy : Context.ShortTy;
  case NeonTypeFlags::Int32:
    return Flags.isUnsigned() ? Context.UnsignedIntTy : Context.IntTy;
  case NeonTypeFlags::Int64:
    if (IsInt64Long)
      return Flags.isUnsigned() ? Context.UnsignedLongTy : Context.LongTy;
    else
      return Flags.isUnsigned() ? Context.UnsignedLongLongTy
                                : Context.LongLongTy;
  case NeonTypeFlags::Poly8:
    return IsPolyUnsigned ? Context.UnsignedCharTy : Context.SignedCharTy;
  case NeonTypeFlags::Poly16:
    return IsPolyUnsigned ? Context.UnsignedShortTy : Context.ShortTy;
  case NeonTypeFlags::Poly64:
    if (IsInt64Long)
      return Context.UnsignedLongTy;
    else
      return Context.UnsignedLongLongTy;
  case NeonTypeFlags::Poly128:
    break;
  case NeonTypeFlags::Float16:
    return Context.HalfTy;
  case NeonTypeFlags::Float32:
    return Context.FloatTy;
  case NeonTypeFlags::Float64:
    return Context.DoubleTy;
  case NeonTypeFlags::BFloat16:
    return Context.BFloat16Ty;
  }
  llvm_unreachable("Invalid NeonTypeFlag!");
}
 
enum ArmSMEState : unsigned {
  ArmNoState = 0,
 
  ArmInZA = 0b01,
  ArmOutZA = 0b10,
  ArmInOutZA = 0b11,
  ArmZAMask = 0b11,
 
  ArmInZT0 = 0b01 << 2,
  ArmOutZT0 = 0b10 << 2,
  ArmInOutZT0 = 0b11 << 2,
  ArmZT0Mask = 0b11 << 2
};
 
bool SemaARM::CheckImmediateArg(CallExpr *TheCall, unsigned CheckTy,
                                unsigned ArgIdx, unsigned EltBitWidth,
                                unsigned ContainerBitWidth) {
  // Function that checks whether the operand (ArgIdx) is an immediate
  // that is one of a given set of values.
  auto CheckImmediateInSet = [&](std::initializer_list<int64_t> Set,
                                 int ErrDiag) -> bool {
    // We can't check the value of a dependent argument.
    Expr *Arg = TheCall->getArg(ArgIdx);
    if (Arg->isTypeDependent() || Arg->isValueDependent())
      return false;
 
    // Check constant-ness first.
    llvm::APSInt Imm;
    if (SemaRef.BuiltinConstantArg(TheCall, ArgIdx, Imm))
      return true;
 
    if (std::find(Set.begin(), Set.end(), Imm.getSExtValue()) == Set.end())
      return Diag(TheCall->getBeginLoc(), ErrDiag) << Arg->getSourceRange();
    return false;
  };
 
  switch ((ImmCheckType)CheckTy) {
  case ImmCheckType::ImmCheck0_31:
    if (SemaRef.BuiltinConstantArgRange(TheCall, ArgIdx, 0, 31))
      return true;
    break;
  case ImmCheckType::ImmCheck0_13:
    if (SemaRef.BuiltinConstantArgRange(TheCall, ArgIdx, 0, 13))
      return true;
    break;
  case ImmCheckType::ImmCheck0_63:
    if (SemaRef.BuiltinConstantArgRange(TheCall, ArgIdx, 0, 63))
      return true;
    break;
  case ImmCheckType::ImmCheck1_16:
    if (SemaRef.BuiltinConstantArgRange(TheCall, ArgIdx, 1, 16))
      return true;
    break;
  case ImmCheckType::ImmCheck0_7:
    if (SemaRef.BuiltinConstantArgRange(TheCall, ArgIdx, 0, 7))
      return true;
    break;
  case ImmCheckType::ImmCheck1_1:
    if (SemaRef.BuiltinConstantArgRange(TheCall, ArgIdx, 1, 1))
      return true;
    break;
  case ImmCheckType::ImmCheck1_3:
    if (SemaRef.BuiltinConstantArgRange(TheCall, ArgIdx, 1, 3))
      return true;
    break;
  case ImmCheckType::ImmCheck1_7:
    if (SemaRef.BuiltinConstantArgRange(TheCall, ArgIdx, 1, 7))
      return true;
    break;
  case ImmCheckType::ImmCheckExtract:
    if (SemaRef.BuiltinConstantArgRange(TheCall, ArgIdx, 0,
                                        (2048 / EltBitWidth) - 1))
      return true;
    break;
  case ImmCheckType::ImmCheckCvt:
  case ImmCheckType::ImmCheckShiftRight:
    if (SemaRef.BuiltinConstantArgRange(TheCall, ArgIdx, 1, EltBitWidth))
      return true;
    break;
  case ImmCheckType::ImmCheckShiftRightNarrow:
    if (SemaRef.BuiltinConstantArgRange(TheCall, ArgIdx, 1, EltBitWidth / 2))
      return true;
    break;
  case ImmCheckType::ImmCheckShiftLeft:
    if (SemaRef.BuiltinConstantArgRange(TheCall, ArgIdx, 0, EltBitWidth - 1))
      return true;
    break;
  case ImmCheckType::ImmCheckLaneIndex:
    if (SemaRef.BuiltinConstantArgRange(TheCall, ArgIdx, 0,
                                        (ContainerBitWidth / EltBitWidth) - 1))
      return true;
    break;
  case ImmCheckType::ImmCheckLaneIndexCompRotate:
    if (SemaRef.BuiltinConstantArgRange(
            TheCall, ArgIdx, 0, (ContainerBitWidth / (2 * EltBitWidth)) - 1))
      return true;
    break;
  case ImmCheckType::ImmCheckLaneIndexDot:
    if (SemaRef.BuiltinConstantArgRange(
            TheCall, ArgIdx, 0, (ContainerBitWidth / (4 * EltBitWidth)) - 1))
      return true;
    break;
  case ImmCheckType::ImmCheckComplexRot90_270:
    if (CheckImmediateInSet({90, 270}, diag::err_rotation_argument_to_cadd))
      return true;
    break;
  case ImmCheckType::ImmCheckComplexRotAll90:
    if (CheckImmediateInSet({0, 90, 180, 270},
                            diag::err_rotation_argument_to_cmla))
      return true;
    break;
  case ImmCheckType::ImmCheck0_1:
    if (SemaRef.BuiltinConstantArgRange(TheCall, ArgIdx, 0, 1))
      return true;
    break;
  case ImmCheckType::ImmCheck0_2:
    if (SemaRef.BuiltinConstantArgRange(TheCall, ArgIdx, 0, 2))
      return true;
    break;
  case ImmCheckType::ImmCheck0_3:
    if (SemaRef.BuiltinConstantArgRange(TheCall, ArgIdx, 0, 3))
      return true;
    break;
  case ImmCheckType::ImmCheck0_0:
    if (SemaRef.BuiltinConstantArgRange(TheCall, ArgIdx, 0, 0))
      return true;
    break;
  case ImmCheckType::ImmCheck0_15:
    if (SemaRef.BuiltinConstantArgRange(TheCall, ArgIdx, 0, 15))
      return true;
    break;
  case ImmCheckType::ImmCheck0_255:
    if (SemaRef.BuiltinConstantArgRange(TheCall, ArgIdx, 0, 255))
      return true;
    break;
  case ImmCheckType::ImmCheck1_32:
    if (SemaRef.BuiltinConstantArgRange(TheCall, ArgIdx, 1, 32))
      return true;
    break;
  case ImmCheckType::ImmCheck1_64:
    if (SemaRef.BuiltinConstantArgRange(TheCall, ArgIdx, 1, 64))
      return true;
    break;
  case ImmCheckType::ImmCheck2_4_Mul2:
    if (SemaRef.BuiltinConstantArgRange(TheCall, ArgIdx, 2, 4) ||
        SemaRef.BuiltinConstantArgMultiple(TheCall, ArgIdx, 2))
      return true;
    break;
  }
  return false;
}
 
bool SemaARM::PerformNeonImmChecks(
    CallExpr *TheCall,
    SmallVectorImpl<std::tuple<int, int, int, int>> &ImmChecks,
    int OverloadType) {
  bool HasError = false;
 
  for (const auto &I : ImmChecks) {
    auto [ArgIdx, CheckTy, ElementBitWidth, VecBitWidth] = I;
 
    if (OverloadType >= 0)
      ElementBitWidth = NeonTypeFlags(OverloadType).getEltSizeInBits();
 
    HasError |= CheckImmediateArg(TheCall, CheckTy, ArgIdx, ElementBitWidth,
                                  VecBitWidth);
  }
 
  return HasError;
}
 
bool SemaARM::PerformSVEImmChecks(
    CallExpr *TheCall, SmallVectorImpl<std::tuple<int, int, int>> &ImmChecks) {
  bool HasError = false;
 
  for (const auto &I : ImmChecks) {
    auto [ArgIdx, CheckTy, ElementBitWidth] = I;
    HasError |=
        CheckImmediateArg(TheCall, CheckTy, ArgIdx, ElementBitWidth, 128);
  }
 
  return HasError;
}
 
SemaARM::ArmStreamingType getArmStreamingFnType(const FunctionDecl *FD) {
  if (FD->hasAttr<ArmLocallyStreamingAttr>())
    return SemaARM::ArmStreaming;
  if (const Type *Ty = FD->getType().getTypePtrOrNull()) {
    if (const auto *FPT = Ty->getAs<FunctionProtoType>()) {
      if (FPT->getAArch64SMEAttributes() &
          FunctionType::SME_PStateSMEnabledMask)
        return SemaARM::ArmStreaming;
      if (FPT->getAArch64SMEAttributes() &
          FunctionType::SME_PStateSMCompatibleMask)
        return SemaARM::ArmStreamingCompatible;
    }
  }
  return SemaARM::ArmNonStreaming;
}
 
static bool checkArmStreamingBuiltin(Sema &S, CallExpr *TheCall,
                                     const FunctionDecl *FD,
                                     SemaARM::ArmStreamingType BuiltinType,
                                     unsigned BuiltinID) {
  SemaARM::ArmStreamingType FnType = getArmStreamingFnType(FD);
 
  // Check if the intrinsic is available in the right mode, i.e.
  // * When compiling for SME only, the caller must be in streaming mode.
  // * When compiling for SVE only, the caller must be in non-streaming mode.
  // * When compiling for both SVE and SME, the caller can be in either mode.
  if (BuiltinType == SemaARM::VerifyRuntimeMode) {
    llvm::StringMap<bool> CallerFeatureMapWithoutSVE;
    S.Context.getFunctionFeatureMap(CallerFeatureMapWithoutSVE, FD);
    CallerFeatureMapWithoutSVE["sve"] = false;
 
    // Avoid emitting diagnostics for a function that can never compile.
    if (FnType == SemaARM::ArmStreaming && !CallerFeatureMapWithoutSVE["sme"])
      return false;
 
    llvm::StringMap<bool> CallerFeatureMapWithoutSME;
    S.Context.getFunctionFeatureMap(CallerFeatureMapWithoutSME, FD);
    CallerFeatureMapWithoutSME["sme"] = false;
 
    // We know the builtin requires either some combination of SVE flags, or
    // some combination of SME flags, but we need to figure out which part
    // of the required features is satisfied by the target features.
    //
    // For a builtin with target guard 'sve2p1|sme2', if we compile with
    // '+sve2p1,+sme', then we know that it satisfies the 'sve2p1' part if we
    // evaluate the features for '+sve2p1,+sme,+nosme'.
    //
    // Similarly, if we compile with '+sve2,+sme2', then we know it satisfies
    // the 'sme2' part if we evaluate the features for '+sve2,+sme2,+nosve'.
    StringRef BuiltinTargetGuards(
        S.Context.BuiltinInfo.getRequiredFeatures(BuiltinID));
    bool SatisfiesSVE = Builtin::evaluateRequiredTargetFeatures(
        BuiltinTargetGuards, CallerFeatureMapWithoutSME);
    bool SatisfiesSME = Builtin::evaluateRequiredTargetFeatures(
        BuiltinTargetGuards, CallerFeatureMapWithoutSVE);
 
    if ((SatisfiesSVE && SatisfiesSME) ||
        (SatisfiesSVE && FnType == SemaARM::ArmStreamingCompatible))
      return false;
    else if (SatisfiesSVE)
      BuiltinType = SemaARM::ArmNonStreaming;
    else if (SatisfiesSME)
      BuiltinType = SemaARM::ArmStreaming;
    else
      // This should be diagnosed by CodeGen
      return false;
  }
 
  if (FnType != SemaARM::ArmNonStreaming &&
      BuiltinType == SemaARM::ArmNonStreaming)
    S.Diag(TheCall->getBeginLoc(), diag::err_attribute_arm_sm_incompat_builtin)
        << TheCall->getSourceRange() << "non-streaming";
  else if (FnType != SemaARM::ArmStreaming &&
           BuiltinType == SemaARM::ArmStreaming)
    S.Diag(TheCall->getBeginLoc(), diag::err_attribute_arm_sm_incompat_builtin)
        << TheCall->getSourceRange() << "streaming";
  else
    return false;
 
  return true;
}
 
static bool hasArmZAState(const FunctionDecl *FD) {
  const auto *T = FD->getType()->getAs<FunctionProtoType>();
  return (T && FunctionType::getArmZAState(T->getAArch64SMEAttributes()) !=
                   FunctionType::ARM_None) ||
         (FD->hasAttr<ArmNewAttr>() && FD->getAttr<ArmNewAttr>()->isNewZA());
}
 
static bool hasArmZT0State(const FunctionDecl *FD) {
  const auto *T = FD->getType()->getAs<FunctionProtoType>();
  return (T && FunctionType::getArmZT0State(T->getAArch64SMEAttributes()) !=
                   FunctionType::ARM_None) ||
         (FD->hasAttr<ArmNewAttr>() && FD->getAttr<ArmNewAttr>()->isNewZT0());
}
 
static ArmSMEState getSMEState(unsigned BuiltinID) {
  switch (BuiltinID) {
  default:
    return ArmNoState;
#define GET_SME_BUILTIN_GET_STATE
#include "clang/Basic/arm_sme_builtins_za_state.inc"
#undef GET_SME_BUILTIN_GET_STATE
  }
}
 
bool SemaARM::CheckSMEBuiltinFunctionCall(unsigned BuiltinID,
                                          CallExpr *TheCall) {
  if (const FunctionDecl *FD = SemaRef.getCurFunctionDecl()) {
    std::optional<ArmStreamingType> BuiltinType;
 
    switch (BuiltinID) {
#define GET_SME_STREAMING_ATTRS
#include "clang/Basic/arm_sme_streaming_attrs.inc"
#undef GET_SME_STREAMING_ATTRS
    }
 
    if (BuiltinType &&
        checkArmStreamingBuiltin(SemaRef, TheCall, FD, *BuiltinType, BuiltinID))
      return true;
 
    if ((getSMEState(BuiltinID) & ArmZAMask) && !hasArmZAState(FD))
      Diag(TheCall->getBeginLoc(),
           diag::warn_attribute_arm_za_builtin_no_za_state)
          << TheCall->getSourceRange();
 
    if ((getSMEState(BuiltinID) & ArmZT0Mask) && !hasArmZT0State(FD))
      Diag(TheCall->getBeginLoc(),
           diag::warn_attribute_arm_zt0_builtin_no_zt0_state)
          << TheCall->getSourceRange();
  }
 
  // Range check SME intrinsics that take immediate values.
  SmallVector<std::tuple<int, int, int>, 3> ImmChecks;
 
  switch (BuiltinID) {
  default:
    return false;
#define GET_SME_IMMEDIATE_CHECK
#include "clang/Basic/arm_sme_sema_rangechecks.inc"
#undef GET_SME_IMMEDIATE_CHECK
  }
 
  return PerformSVEImmChecks(TheCall, ImmChecks);
}
 
bool SemaARM::CheckSVEBuiltinFunctionCall(unsigned BuiltinID,
                                          CallExpr *TheCall) {
  if (const FunctionDecl *FD = SemaRef.getCurFunctionDecl()) {
    std::optional<ArmStreamingType> BuiltinType;
 
    switch (BuiltinID) {
#define GET_SVE_STREAMING_ATTRS
#include "clang/Basic/arm_sve_streaming_attrs.inc"
#undef GET_SVE_STREAMING_ATTRS
    }
    if (BuiltinType &&
        checkArmStreamingBuiltin(SemaRef, TheCall, FD, *BuiltinType, BuiltinID))
      return true;
  }
  // Range check SVE intrinsics that take immediate values.
  SmallVector<std::tuple<int, int, int>, 3> ImmChecks;
 
  switch (BuiltinID) {
  default:
    return false;
#define GET_SVE_IMMEDIATE_CHECK
#include "clang/Basic/arm_sve_sema_rangechecks.inc"
#undef GET_SVE_IMMEDIATE_CHECK
  }
 
  return PerformSVEImmChecks(TheCall, ImmChecks);
}
 
bool SemaARM::CheckNeonBuiltinFunctionCall(const TargetInfo &TI,
                                           unsigned BuiltinID,
                                           CallExpr *TheCall) {
  if (const FunctionDecl *FD = SemaRef.getCurFunctionDecl()) {
 
    switch (BuiltinID) {
    default:
      break;
#define GET_NEON_BUILTINS
#define TARGET_BUILTIN(id, ...) case NEON::BI##id:
#define BUILTIN(id, ...) case NEON::BI##id:
#include "clang/Basic/arm_neon.inc"
      if (checkArmStreamingBuiltin(