[Intrinsic] Rename flt.rounds intrinsic to get.rounding

[qiucf]

Summary

This is to address the inconsistency between FLT_ROUNDS_ and SET_ROUNDING SDAG node. Rename FLT_ROUNDS_ to GET_ROUNDING and add llvm.get.rounding intrinsic to replace flt.rounds.

[sepavloff]

Thank you for working on this!

Is there any reason why we should keep the old intrinsic?

[RKSimon]

In D139507#3978449, @sepavloff wrote:

Thank you for working on this!

Is there any reason why we should keep the old intrinsic?

I'm not clear on the general policy, but for x86 we tend to provide bc of old intrinsics/sigtaures via AutoUpgrade.cpp, supposedly only for a few releases but tbh we remove them very rarely....

[qiucf]

In D139507#3978449, @sepavloff wrote:

Thank you for working on this!

Is there any reason why we should keep the old intrinsic?

In case any user outside of clang references it (although I believe no), we can deprecate it and remove after a few releases.

[sepavloff]

LGTM.

In D139507#3980555, @qiucf wrote:

In D139507#3978449, @sepavloff wrote:

Thank you for working on this!

Is there any reason why we should keep the old intrinsic?

In case any user outside of clang references it (although I believe no), we can deprecate it and remove after a few releases.

Yes, this could be a good solution.

[nikic]

In D139507#3980555, @qiucf wrote:

In D139507#3978449, @sepavloff wrote:

Thank you for working on this!

Is there any reason why we should keep the old intrinsic?

In case any user outside of clang references it (although I believe no), we can deprecate it and remove after a few releases.

This is not how intrinsic changes work. You need to add AutoUpgrade support.

[qiucf]

Use AutoUpgrade to rename flt.rounds to get.rounding.

[nikic]

libcxxabi/test/test_demangle.pass.cpp

12917 ↗

(On Diff #482381)

Demangling tests should not be updated.

llvm/docs/LangRef.rst

24567

Can drop this note now.

llvm/lib/IR/AutoUpgrade.cpp

916

This needs a test in llvm/test/Bitcode. Create a bitcode file using flt.rounds prior to this change, and then test llvm-dis on it.

[nikic]

(per previous comment, mostly looks good though)

[nikic]

LGTM

clang/lib/CodeGen/CGBuiltin.cpp

       Builder.CreateFCmpUGE(Abs, ConstantFP::get(V->getContext(), Smallest),
                             "isnormal");
     V = Builder.CreateAnd(Eq, IsLessThanInf, "and");
     V = Builder.CreateAnd(V, IsNormal, "and");
     return RValue::get(Builder.CreateZExt(V, ConvertType(E->getType())));
   }
 
   case Builtin::BI__builtin_flt_rounds: {
-    Function *F = CGM.getIntrinsic(Intrinsic::flt_rounds);
+    Function *F = CGM.getIntrinsic(Intrinsic::get_rounding);
 
     llvm::Type *ResultType = ConvertType(E->getType());
     Value *Result = Builder.CreateCall(F);
     if (Result->getType() != ResultType)
       Result = Builder.CreateIntCast(Result, ResultType, /*isSigned*/true,
                                      "cast");
     return RValue::get(Result);
   }

clang/test/CodeGen/builtins-msp430.c

 // REQUIRES: msp430-registered-target
 // RUN: %clang_cc1 -triple msp430-unknown-unknown -emit-llvm %s -o - | FileCheck %s
 
 int test_builtin_flt_rounds() {
-  // CHECK:     [[V0:[%A-Za-z0-9.]+]] = call i32 @llvm.flt.rounds()
+  // CHECK:     [[V0:[%A-Za-z0-9.]+]] = call i32 @llvm.get.rounding()
   // CHECK-DAG: [[V1:[%A-Za-z0-9.]+]] = trunc i32 [[V0]] to i16
   // CHECK-DAG: ret i16 [[V1]]
   return __builtin_flt_rounds();
 }
 

clang/test/CodeGen/builtins.c

   // CHECK: fcmp oeq float
   // CHECK: call float @llvm.fabs.f32(float
   // CHECK: fcmp ult float {{.*}}, 0x7FF0000000000000
   // CHECK: fcmp uge float {{.*}}, 0x3810000000000000
   // CHECK: and i1
   // CHECK: and i1
 
   res = __builtin_flt_rounds();
-  // CHECK: call i32 @llvm.flt.rounds(
+  // CHECK: call i32 @llvm.get.rounding(
 }
 
 // CHECK-LABEL: define{{.*}} void @test_float_builtin_ops
 void test_float_builtin_ops(float F, double D, long double LD) {
   volatile float resf;
   volatile double resd;
   volatile long double resld;
   volatile long int resli;

llvm/docs/LangRef.rst

     %cond = call i1 @cond()
     br i1 %cond, label %loop, label %exit
 
   exit:
     ret void
   }
 
   !0 = !{!0} ; domain
   !1 = !{!1, !0} ; scope
   !2 = !{!1} ; scope list
 
 Multiple calls to `@llvm.experimental.noalias.scope.decl` for the same scope
 are possible, but one should never dominate another. Violations are pointed out
 by the verifier as they indicate a problem in either a transformation pass or
 the input.
 
 
 Floating Point Environment Manipulation intrinsics
 --------------------------------------------------
 
 These functions read or write floating point environment, such as rounding
 mode or state of floating point exceptions. Altering the floating point
 environment requires special care. See :ref:`Floating Point Environment <floatenv>`.
 
-'``llvm.flt.rounds``' Intrinsic
+'``llvm.get.rounding``' Intrinsic
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 
 Syntax:
 """""""
 
 ::
 
-      declare i32 @llvm.flt.rounds()
+      declare i32 @llvm.get.rounding()
 
 Overview:
 """""""""
 
-The '``llvm.flt.rounds``' intrinsic reads the current rounding mode.
+The '``llvm.get.rounding``' intrinsic reads the current rounding mode.
 
 Semantics:
 """"""""""
 
-The '``llvm.flt.rounds``' intrinsic returns the current rounding mode.
+The '``llvm.get.rounding``' intrinsic returns the current rounding mode.
 Encoding of the returned values is same as the result of ``FLT_ROUNDS``,
 specified by C standard:
 
 ::
 
     0  - toward zero
     1  - to nearest, ties to even
     2  - toward positive infinity
     3  - toward negative infinity
     4  - to nearest, ties away from zero
 
 Other values may be used to represent additional rounding modes, supported by a
 target. These values are target-specific.
 
-
 '``llvm.set.rounding``' Intrinsic
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
 
 Syntax:
 """""""
 
 ::
 
       declare void @llvm.set.rounding(i32 <val>)
 
 Overview:
 """""""""
 
 The '``llvm.set.rounding``' intrinsic sets current rounding mode.
 
 Arguments:
 """"""""""
 
 The argument is the required rounding mode. Encoding of rounding mode is
-the same as used by '``llvm.flt.rounds``'.
+the same as used by '``llvm.get.rounding``'.
 
 Semantics:
 """"""""""
 
 The '``llvm.set.rounding``' intrinsic sets the current rounding mode. It is
 similar to C library function 'fesetround', however this intrinsic does not
 return any value and uses platform-independent representation of IEEE rounding
 modes.

llvm/include/llvm/CodeGen/ISDOpcodes.h

   /// Returns current rounding mode:
   /// -1 Undefined
   ///  0 Round to 0
   ///  1 Round to nearest, ties to even
   ///  2 Round to +inf
   ///  3 Round to -inf
   ///  4 Round to nearest, ties to zero
   /// Result is rounding mode and chain. Input is a chain.
-  /// TODO: Rename this node to GET_ROUNDING.
-  FLT_ROUNDS_,
+  GET_ROUNDING,
 
   /// Set rounding mode.
   /// The first operand is a chain pointer. The second specifies the required
-  /// rounding mode, encoded in the same way as used in '``FLT_ROUNDS_``'.
+  /// rounding mode, encoded in the same way as used in '``GET_ROUNDING``'.
   SET_ROUNDING,
 
   /// X = FP_EXTEND(Y) - Extend a smaller FP type into a larger FP type.
   FP_EXTEND,
 
   /// BITCAST - This operator converts between integer, vector and FP
   /// values, as if the value was stored to memory with one type and loaded
   /// from the same address with the other type (or equivalently for vector

llvm/include/llvm/IR/Intrinsics.td

                                 ImmArg<ArgIndex<1>>, ImmArg<ArgIndex<2>>,
                                 ImmArg<ArgIndex<3>>]>,
                                ClangBuiltin<"__builtin_object_size">;
 
 //===--------------- Access to Floating Point Environment -----------------===//
 //
 
 let IntrProperties = [IntrInaccessibleMemOnly, IntrWillReturn] in {
-  def int_flt_rounds    : DefaultAttrsIntrinsic<[llvm_i32_ty], []>;
+  def int_get_rounding  : DefaultAttrsIntrinsic<[llvm_i32_ty], []>;
   def int_set_rounding  : DefaultAttrsIntrinsic<[], [llvm_i32_ty]>;
 }
 
 //===--------------- Floating Point Properties ----------------------------===//
 //
 
 def int_is_fpclass
     : DefaultAttrsIntrinsic<[LLVMScalarOrSameVectorWidth<0, llvm_i1_ty>],

llvm/lib/CodeGen/IntrinsicLowering.cpp

   case Intrinsic::roundeven: {
     ReplaceFPIntrinsicWithCall(CI, "roundevenf", "roundeven", "roundevenl");
     break;
   }
   case Intrinsic::copysign: {
     ReplaceFPIntrinsicWithCall(CI, "copysignf", "copysign", "copysignl");
     break;
   }
-  case Intrinsic::flt_rounds:
+  case Intrinsic::get_rounding:
      // Lower to "round to the nearest"
      if (!CI->getType()->isVoidTy())
        CI->replaceAllUsesWith(ConstantInt::get(CI->getType(), 1));
      break;
   case Intrinsic::invariant_start:
   case Intrinsic::lifetime_start:
     // Discard region information.
     CI->replaceAllUsesWith(UndefValue::get(CI->getType()));

llvm/lib/CodeGen/SelectionDAG/LegalizeDAG.cpp

   case ISD::CALLSEQ_START:
   case ISD::CALLSEQ_END:
     // FIXME: This shouldn't be necessary.  These nodes have special properties
     // dealing with the recursive nature of legalization.  Removing this
     // special case should be done as part of making LegalizeDAG non-recursive.
     SimpleFinishLegalizing = false;
     break;
   case ISD::EXTRACT_ELEMENT:
-  case ISD::FLT_ROUNDS_:
+  case ISD::GET_ROUNDING:
   case ISD::MERGE_VALUES:
   case ISD::EH_RETURN:
   case ISD::FRAME_TO_ARGS_OFFSET:
   case ISD::EH_DWARF_CFA:
   case ISD::EH_SJLJ_SETJMP:
   case ISD::EH_SJLJ_LONGJMP:
   case ISD::EH_SJLJ_SETUP_DISPATCH:
     // These operations lie about being legal: when they claim to be legal,

                                  CfaArg);
     SDValue FA = DAG.getNode(
         ISD::FRAMEADDR, dl, TLI.getPointerTy(DAG.getDataLayout()),
         DAG.getConstant(0, dl, TLI.getPointerTy(DAG.getDataLayout())));
     Results.push_back(DAG.getNode(ISD::ADD, dl, FA.getValueType(),
                                   FA, Offset));
     break;
   }
-  case ISD::FLT_ROUNDS_:
+  case ISD::GET_ROUNDING:
     Results.push_back(DAG.getConstant(1, dl, Node->getValueType(0)));
     Results.push_back(Node->getOperand(0));
     break;
   case ISD::EH_RETURN:
   case ISD::EH_LABEL:
   case ISD::PREFETCH:
   case ISD::VAEND:
   case ISD::EH_SJLJ_LONGJMP:

llvm/lib/CodeGen/SelectionDAG/LegalizeIntegerTypes.cpp

   case ISD::FP_TO_UINT_SAT:
                          Res = PromoteIntRes_FP_TO_XINT_SAT(N); break;
 
   case ISD::FP_TO_BF16:
   case ISD::FP_TO_FP16:
     Res = PromoteIntRes_FP_TO_FP16_BF16(N);
     break;
 
-  case ISD::FLT_ROUNDS_: Res = PromoteIntRes_FLT_ROUNDS(N); break;
+  case ISD::GET_ROUNDING: Res = PromoteIntRes_GET_ROUNDING(N); break;
 
   case ISD::AND:
   case ISD::OR:
   case ISD::XOR:
   case ISD::ADD:
   case ISD::SUB:
   case ISD::MUL:
   case ISD::VP_AND:

 
 SDValue DAGTypeLegalizer::PromoteIntRes_FP_TO_FP16_BF16(SDNode *N) {
   EVT NVT = TLI.getTypeToTransformTo(*DAG.getContext(), N->getValueType(0));
   SDLoc dl(N);
 
   return DAG.getNode(N->getOpcode(), dl, NVT, N->getOperand(0));
 }
 
-SDValue DAGTypeLegalizer::PromoteIntRes_FLT_ROUNDS(SDNode *N) {
+SDValue DAGTypeLegalizer::PromoteIntRes_GET_ROUNDING(SDNode *N) {
   EVT NVT = TLI.getTypeToTransformTo(*DAG.getContext(), N->getValueType(0));
   SDLoc dl(N);
 
   SDValue Res =
       DAG.getNode(N->getOpcode(), dl, {NVT, MVT::Other}, N->getOperand(0));
 
   // Legalize the chain result - switch anything that used the old chain to
   // use the new one.

   case ISD::PARITY:      ExpandIntRes_PARITY(N, Lo, Hi); break;
   case ISD::Constant:    ExpandIntRes_Constant(N, Lo, Hi); break;
   case ISD::ABS:         ExpandIntRes_ABS(N, Lo, Hi); break;
   case ISD::CTLZ_ZERO_UNDEF:
   case ISD::CTLZ:        ExpandIntRes_CTLZ(N, Lo, Hi); break;
   case ISD::CTPOP:       ExpandIntRes_CTPOP(N, Lo, Hi); break;
   case ISD::CTTZ_ZERO_UNDEF:
   case ISD::CTTZ:        ExpandIntRes_CTTZ(N, Lo, Hi); break;
-  case ISD::FLT_ROUNDS_: ExpandIntRes_FLT_ROUNDS(N, Lo, Hi); break;
+  case ISD::GET_ROUNDING:ExpandIntRes_GET_ROUNDING(N, Lo, Hi); break;
   case ISD::STRICT_FP_TO_SINT:
   case ISD::FP_TO_SINT:  ExpandIntRes_FP_TO_SINT(N, Lo, Hi); break;
   case ISD::STRICT_FP_TO_UINT:
   case ISD::FP_TO_UINT:  ExpandIntRes_FP_TO_UINT(N, Lo, Hi); break;
   case ISD::FP_TO_SINT_SAT:
   case ISD::FP_TO_UINT_SAT: ExpandIntRes_FP_TO_XINT_SAT(N, Lo, Hi); break;
   case ISD::STRICT_LLROUND:
   case ISD::STRICT_LLRINT:

 
   Lo = DAG.getSelect(dl, NVT, LoNotZero, LoLZ,
                      DAG.getNode(ISD::ADD, dl, NVT, HiLZ,
                                  DAG.getConstant(NVT.getSizeInBits(), dl,
                                                  NVT)));
   Hi = DAG.getConstant(0, dl, NVT);
 }
 
-void DAGTypeLegalizer::ExpandIntRes_FLT_ROUNDS(SDNode *N, SDValue &Lo,
+void DAGTypeLegalizer::ExpandIntRes_GET_ROUNDING(SDNode *N, SDValue &Lo,
                                                SDValue &Hi) {
   SDLoc dl(N);
   EVT NVT = TLI.getTypeToTransformTo(*DAG.getContext(), N->getValueType(0));
   unsigned NBitWidth = NVT.getSizeInBits();
 
-  Lo = DAG.getNode(ISD::FLT_ROUNDS_, dl, {NVT, MVT::Other}, N->getOperand(0));
+  Lo = DAG.getNode(ISD::GET_ROUNDING, dl, {NVT, MVT::Other}, N->getOperand(0));
   SDValue Chain = Lo.getValue(1);
-  // The high part is the sign of Lo, as -1 is a valid value for FLT_ROUNDS
+  // The high part is the sign of Lo, as -1 is a valid value for GET_ROUNDING
   Hi = DAG.getNode(ISD::SRA, dl, NVT, Lo,
                    DAG.getShiftAmountConstant(NBitWidth - 1, NVT, dl));
 
   // Legalize the chain result - switch anything that used the old chain to
   // use the new one.
   ReplaceValueWith(SDValue(N, 1), Chain);
 }
 

llvm/lib/CodeGen/SelectionDAG/LegalizeTypes.h

   SDValue PromoteIntRes_SADDSUBO_CARRY(SDNode *N, unsigned ResNo);
   SDValue PromoteIntRes_UNDEF(SDNode *N);
   SDValue PromoteIntRes_VAARG(SDNode *N);
   SDValue PromoteIntRes_VSCALE(SDNode *N);
   SDValue PromoteIntRes_XMULO(SDNode *N, unsigned ResNo);
   SDValue PromoteIntRes_ADDSUBSHLSAT(SDNode *N);
   SDValue PromoteIntRes_MULFIX(SDNode *N);
   SDValue PromoteIntRes_DIVFIX(SDNode *N);
-  SDValue PromoteIntRes_FLT_ROUNDS(SDNode *N);
+  SDValue PromoteIntRes_GET_ROUNDING(SDNode *N);
   SDValue PromoteIntRes_VECREDUCE(SDNode *N);
   SDValue PromoteIntRes_VP_REDUCE(SDNode *N);
   SDValue PromoteIntRes_ABS(SDNode *N);
   SDValue PromoteIntRes_Rotate(SDNode *N);
   SDValue PromoteIntRes_FunnelShift(SDNode *N);
   SDValue PromoteIntRes_IS_FPCLASS(SDNode *N);
 
   // Integer Operand Promotion.

   void ExpandIntRes_CTPOP             (SDNode *N, SDValue &Lo, SDValue &Hi);
   void ExpandIntRes_CTTZ              (SDNode *N, SDValue &Lo, SDValue &Hi);
   void ExpandIntRes_LOAD          (LoadSDNode *N, SDValue &Lo, SDValue &Hi);
   void ExpandIntRes_READCYCLECOUNTER  (SDNode *N, SDValue &Lo, SDValue &Hi);
   void ExpandIntRes_SIGN_EXTEND       (SDNode *N, SDValue &Lo, SDValue &Hi);
   void ExpandIntRes_SIGN_EXTEND_INREG (SDNode *N, SDValue &Lo, SDValue &Hi);
   void ExpandIntRes_TRUNCATE          (SDNode *N, SDValue &Lo, SDValue &Hi);
   void ExpandIntRes_ZERO_EXTEND       (SDNode *N, SDValue &Lo, SDValue &Hi);
-  void ExpandIntRes_FLT_ROUNDS        (SDNode *N, SDValue &Lo, SDValue &Hi);
+  void ExpandIntRes_GET_ROUNDING      (SDNode *N, SDValue &Lo, SDValue &Hi);
   void ExpandIntRes_FP_TO_SINT        (SDNode *N, SDValue &Lo, SDValue &Hi);
   void ExpandIntRes_FP_TO_UINT        (SDNode *N, SDValue &Lo, SDValue &Hi);
   void ExpandIntRes_FP_TO_XINT_SAT    (SDNode *N, SDValue &Lo, SDValue &Hi);
   void ExpandIntRes_LLROUND_LLRINT    (SDNode *N, SDValue &Lo, SDValue &Hi);
 
   void ExpandIntRes_Logical           (SDNode *N, SDValue &Lo, SDValue &Hi);
   void ExpandIntRes_ADDSUB            (SDNode *N, SDValue &Lo, SDValue &Hi);
   void ExpandIntRes_ADDSUBC           (SDNode *N, SDValue &Lo, SDValue &Hi);

llvm/lib/CodeGen/SelectionDAG/SelectionDAGBuilder.cpp

 
     FrameIndexSDNode *FI = cast<FrameIndexSDNode>(getValue(Alloca).getNode());
     GFI->addStackRoot(FI->getIndex(), TypeMap);
     return;
   }
   case Intrinsic::gcread:
   case Intrinsic::gcwrite:
     llvm_unreachable("GC failed to lower gcread/gcwrite intrinsics!");
-  case Intrinsic::flt_rounds:
-    Res = DAG.getNode(ISD::FLT_ROUNDS_, sdl, {MVT::i32, MVT::Other}, getRoot());
+  case Intrinsic::get_rounding:
+    Res = DAG.getNode(ISD::GET_ROUNDING, sdl, {MVT::i32, MVT::Other}, getRoot());
     setValue(&I, Res);
     DAG.setRoot(Res.getValue(1));
     return;
 
   case Intrinsic::expect:
     // Just replace __builtin_expect(exp, c) with EXP.
     setValue(&I, getValue(I.getArgOperand(0)));
     return;

llvm/lib/CodeGen/SelectionDAG/SelectionDAGDumper.cpp

   case ISD::GET_DYNAMIC_AREA_OFFSET:    return "get.dynamic.area.offset";
   case ISD::FREEZE:                     return "freeze";
   case ISD::PREALLOCATED_SETUP:
     return "call_setup";
   case ISD::PREALLOCATED_ARG:
     return "call_alloc";
 
   // Floating point environment manipulation
-  case ISD::FLT_ROUNDS_:                return "flt_rounds";
+  case ISD::GET_ROUNDING:               return "get_rounding";
   case ISD::SET_ROUNDING:               return "set_rounding";
 
   // Bit manipulation
   case ISD::ABS:                        return "abs";
   case ISD::BITREVERSE:                 return "bitreverse";
   case ISD::BSWAP:                      return "bswap";
   case ISD::CTPOP:                      return "ctpop";
   case ISD::CTTZ:                       return "cttz";

llvm/lib/IR/AutoUpgrade.cpp

         auto Args = F->getFunctionType()->params();
         Type *Tys[] = {Args[1]};
         NewFn = Intrinsic::getDeclaration(F->getParent(), ID, Tys);
         return true;
       }
     }
     break;
   }
+  case 'f':
+    if (Name.startswith("flt.rounds")) {
+      rename(F);
+      NewFn = Intrinsic::getDeclaration(F->getParent(), Intrinsic::get_rounding);
+      return true;
+    }
+    break;
   case 'i':
   case 'l': {
     bool IsLifetimeStart = Name.startswith("lifetime.start");
     if (IsLifetimeStart || Name.startswith("invariant.start")) {
       Intrinsic::ID ID = IsLifetimeStart ?
         Intrinsic::lifetime_start : Intrinsic::invariant_start;
       auto Args = F->getFunctionType()->params();
       Type* ObjectPtr[1] = {Args[1]};

llvm/lib/Target/AArch64/AArch64ISelLowering.h

   SDValue LowerDarwin_VASTART(SDValue Op, SelectionDAG &DAG) const;
   SDValue LowerWin64_VASTART(SDValue Op, SelectionDAG &DAG) const;
   SDValue LowerVASTART(SDValue Op, SelectionDAG &DAG) const;
   SDValue LowerVACOPY(SDValue Op, SelectionDAG &DAG) const;
   SDValue LowerVAARG(SDValue Op, SelectionDAG &DAG) const;
   SDValue LowerFRAMEADDR(SDValue Op, SelectionDAG &DAG) const;
   SDValue LowerSPONENTRY(SDValue Op, SelectionDAG &DAG) const;
   SDValue LowerRETURNADDR(SDValue Op, SelectionDAG &DAG) const;
-  SDValue LowerFLT_ROUNDS_(SDValue Op, SelectionDAG &DAG) const;
+  SDValue LowerGET_ROUNDING(SDValue Op, SelectionDAG &DAG) const;
   SDValue LowerSET_ROUNDING(SDValue Op, SelectionDAG &DAG) const;
   SDValue LowerINSERT_VECTOR_ELT(SDValue Op, SelectionDAG &DAG) const;
   SDValue LowerEXTRACT_VECTOR_ELT(SDValue Op, SelectionDAG &DAG) const;
   SDValue LowerSCALAR_TO_VECTOR(SDValue Op, SelectionDAG &DAG) const;
   SDValue LowerBUILD_VECTOR(SDValue Op, SelectionDAG &DAG) const;
   SDValue LowerVECTOR_SHUFFLE(SDValue Op, SelectionDAG &DAG) const;
   SDValue LowerSPLAT_VECTOR(SDValue Op, SelectionDAG &DAG) const;
   SDValue LowerDUPQLane(SDValue Op, SelectionDAG &DAG) const;

llvm/lib/Target/AArch64/AArch64ISelLowering.cpp

   }
 
   // Strict conversion to a larger type is legal
   for (auto VT : {MVT::f32, MVT::f64})
     setOperationAction(ISD::STRICT_FP_EXTEND, VT, Legal);
 
   setOperationAction(ISD::PREFETCH, MVT::Other, Custom);
 
-  setOperationAction(ISD::FLT_ROUNDS_, MVT::i32, Custom);
+  setOperationAction(ISD::GET_ROUNDING, MVT::i32, Custom);
   setOperationAction(ISD::SET_ROUNDING, MVT::Other, Custom);
 
   setOperationAction(ISD::ATOMIC_CMP_SWAP, MVT::i128, Custom);
   setOperationAction(ISD::ATOMIC_LOAD_SUB, MVT::i32, Custom);
   setOperationAction(ISD::ATOMIC_LOAD_SUB, MVT::i64, Custom);
   setOperationAction(ISD::ATOMIC_LOAD_AND, MVT::i32, Custom);
   setOperationAction(ISD::ATOMIC_LOAD_AND, MVT::i64, Custom);
 

     SDNode *N0 = N->getOperand(0).getNode();
     SDNode *N1 = N->getOperand(1).getNode();
     return N0->hasOneUse() && N1->hasOneUse() &&
       isZeroExtended(N0, DAG) && isZeroExtended(N1, DAG);
   }
   return false;
 }
 
-SDValue AArch64TargetLowering::LowerFLT_ROUNDS_(SDValue Op,
+SDValue AArch64TargetLowering::LowerGET_ROUNDING(SDValue Op,
                                                  SelectionDAG &DAG) const {
   // The rounding mode is in bits 23:22 of the FPSCR.
   // The ARM rounding mode value to FLT_ROUNDS mapping is 0->1, 1->2, 2->3, 3->0
   // The formula we use to implement this is (((FPSCR + 1 << 22) >> 22) & 3)
   // so that the shift + and get folded into a bitfield extract.
   SDLoc dl(Op);
 
   SDValue Chain = Op.getOperand(0);
   SDValue FPCR_64 = DAG.getNode(

   case ISD::STRICT_FP_TO_SINT:
   case ISD::STRICT_FP_TO_UINT:
     return LowerFP_TO_INT(Op, DAG);
   case ISD::FP_TO_SINT_SAT:
   case ISD::FP_TO_UINT_SAT:
     return LowerFP_TO_INT_SAT(Op, DAG);
   case ISD::FSINCOS:
     return LowerFSINCOS(Op, DAG);
-  case ISD::FLT_ROUNDS_:
-    return LowerFLT_ROUNDS_(Op, DAG);
+  case ISD::GET_ROUNDING:
+    return LowerGET_ROUNDING(Op, DAG);
   case ISD::SET_ROUNDING:
     return LowerSET_ROUNDING(Op, DAG);
   case ISD::MUL:
     return LowerMUL(Op, DAG);
   case ISD::MULHS:
     return LowerToPredicatedOp(Op, DAG, AArch64ISD::MULHS_PRED);
   case ISD::MULHU:
     return LowerToPredicatedOp(Op, DAG, AArch64ISD::MULHU_PRED);

llvm/lib/Target/ARM/ARMISelLowering.h

     SDValue LowerSELECT_CC(SDValue Op, SelectionDAG &DAG) const;
     SDValue LowerBRCOND(SDValue Op, SelectionDAG &DAG) const;
     SDValue LowerBR_CC(SDValue Op, SelectionDAG &DAG) const;
     SDValue LowerFCOPYSIGN(SDValue Op, SelectionDAG &DAG) const;
     SDValue LowerRETURNADDR(SDValue Op, SelectionDAG &DAG) const;
     SDValue LowerFRAMEADDR(SDValue Op, SelectionDAG &DAG) const;
     SDValue LowerShiftRightParts(SDValue Op, SelectionDAG &DAG) const;
     SDValue LowerShiftLeftParts(SDValue Op, SelectionDAG &DAG) const;
-    SDValue LowerFLT_ROUNDS_(SDValue Op, SelectionDAG &DAG) const;
+    SDValue LowerGET_ROUNDING(SDValue Op, SelectionDAG &DAG) const;
     SDValue LowerSET_ROUNDING(SDValue Op, SelectionDAG &DAG) const;
     SDValue LowerConstantFP(SDValue Op, SelectionDAG &DAG,
                             const ARMSubtarget *ST) const;
     SDValue LowerBUILD_VECTOR(SDValue Op, SelectionDAG &DAG,
                               const ARMSubtarget *ST) const;
     SDValue LowerINSERT_VECTOR_ELT(SDValue Op, SelectionDAG &DAG) const;
     SDValue LowerFSINCOS(SDValue Op, SelectionDAG &DAG) const;
     SDValue LowerDivRem(SDValue Op, SelectionDAG &DAG) const;

llvm/lib/Target/ARM/ARMISelLowering.cpp

   }
   setOperationAction(ISD::SIGN_EXTEND_INREG, MVT::i1, Expand);
 
   if (!Subtarget->useSoftFloat() && Subtarget->hasFPRegs() &&
       !Subtarget->isThumb1Only()) {
     // Turn f64->i64 into VMOVRRD, i64 -> f64 to VMOVDRR
     // iff target supports vfp2.
     setOperationAction(ISD::BITCAST, MVT::i64, Custom);
-    setOperationAction(ISD::FLT_ROUNDS_, MVT::i32, Custom);
+    setOperationAction(ISD::GET_ROUNDING, MVT::i32, Custom);
     setOperationAction(ISD::SET_ROUNDING, MVT::Other, Custom);
   }
 
   // We want to custom lower some of our intrinsics.
   setOperationAction(ISD::INTRINSIC_WO_CHAIN, MVT::Other, Custom);
   setOperationAction(ISD::EH_SJLJ_SETJMP, MVT::i32, Custom);
   setOperationAction(ISD::EH_SJLJ_LONGJMP, MVT::Other, Custom);
   setOperationAction(ISD::EH_SJLJ_SETUP_DISPATCH, MVT::Other, Custom);

   SDValue LoSmallShift = DAG.getNode(ISD::SHL, dl, VT, ShOpLo, ShAmt);
   SDValue Lo = DAG.getNode(ARMISD::CMOV, dl, VT, LoSmallShift,
                            DAG.getConstant(0, dl, VT), ARMcc, CCR, CmpLo);
 
   SDValue Ops[2] = { Lo, Hi };
   return DAG.getMergeValues(Ops, dl);
 }
 
-SDValue ARMTargetLowering::LowerFLT_ROUNDS_(SDValue Op,
+SDValue ARMTargetLowering::LowerGET_ROUNDING(SDValue Op,
                                              SelectionDAG &DAG) const {
   // The rounding mode is in bits 23:22 of the FPSCR.
   // The ARM rounding mode value to FLT_ROUNDS mapping is 0->1, 1->2, 2->3, 3->0
   // The formula we use to implement this is (((FPSCR + 1 << 22) >> 22) & 3)
   // so that the shift + and get folded into a bitfield extract.
   SDLoc dl(Op);
   SDValue Chain = Op.getOperand(0);
   SDValue Ops[] = {Chain,
                    DAG.getConstant(Intrinsic::arm_get_fpscr, dl, MVT::i32)};

   case ISD::VECTOR_SHUFFLE: return LowerVECTOR_SHUFFLE(Op, DAG, Subtarget);
   case ISD::EXTRACT_SUBVECTOR: return LowerEXTRACT_SUBVECTOR(Op, DAG, Subtarget);
   case ISD::INSERT_VECTOR_ELT: return LowerINSERT_VECTOR_ELT(Op, DAG);
   case ISD::EXTRACT_VECTOR_ELT: return LowerEXTRACT_VECTOR_ELT(Op, DAG, Subtarget);
   case ISD::CONCAT_VECTORS: return LowerCONCAT_VECTORS(Op, DAG, Subtarget);
   case ISD::TRUNCATE:      return LowerTruncate(Op.getNode(), DAG, Subtarget);
   case ISD::SIGN_EXTEND:
   case ISD::ZERO_EXTEND:   return LowerVectorExtend(Op.getNode(), DAG, Subtarget);
-  case ISD::FLT_ROUNDS_:   return LowerFLT_ROUNDS_(Op, DAG);
+  case ISD::GET_ROUNDING:  return LowerGET_ROUNDING(Op, DAG);
   case ISD::SET_ROUNDING:  return LowerSET_ROUNDING(Op, DAG);
   case ISD::MUL:           return LowerMUL(Op, DAG);
   case ISD::SDIV:
     if (Subtarget->isTargetWindows() && !Op.getValueType().isVector())
       return LowerDIV_Windows(Op, DAG, /* Signed */ true);
     return LowerSDIV(Op, DAG, Subtarget);
   case ISD::UDIV:
     if (Subtarget->isTargetWindows() && !Op.getValueType().isVector())

llvm/lib/Target/PowerPC/PPCISelLowering.h

     SDValue LowerEH_DWARF_CFA(SDValue Op, SelectionDAG &DAG) const;
     SDValue LowerLOAD(SDValue Op, SelectionDAG &DAG) const;
     SDValue LowerSTORE(SDValue Op, SelectionDAG &DAG) const;
     SDValue LowerTRUNCATE(SDValue Op, SelectionDAG &DAG) const;
     SDValue LowerSELECT_CC(SDValue Op, SelectionDAG &DAG) const;
     SDValue LowerFP_TO_INT(SDValue Op, SelectionDAG &DAG,
                            const SDLoc &dl) const;
     SDValue LowerINT_TO_FP(SDValue Op, SelectionDAG &DAG) const;
-    SDValue LowerFLT_ROUNDS_(SDValue Op, SelectionDAG &DAG) const;
+    SDValue LowerGET_ROUNDING(SDValue Op, SelectionDAG &DAG) const;
     SDValue LowerSHL_PARTS(SDValue Op, SelectionDAG &DAG) const;
     SDValue LowerSRL_PARTS(SDValue Op, SelectionDAG &DAG) const;
     SDValue LowerSRA_PARTS(SDValue Op, SelectionDAG &DAG) const;
     SDValue LowerFunnelShift(SDValue Op, SelectionDAG &DAG) const;
     SDValue LowerBUILD_VECTOR(SDValue Op, SelectionDAG &DAG) const;
     SDValue LowerVECTOR_SHUFFLE(SDValue Op, SelectionDAG &DAG) const;
     SDValue LowerVPERM(SDValue Op, SelectionDAG &DAG, ArrayRef<int> PermMask,
                        EVT VT, SDValue V1, SDValue V2) const;

llvm/lib/Target/PowerPC/PPCISelLowering.cpp

   } else {
     setOperationAction(ISD::FMA  , MVT::f64, Legal);
     setOperationAction(ISD::FMA  , MVT::f32, Legal);
   }
 
   if (Subtarget.hasSPE())
     setLoadExtAction(ISD::EXTLOAD, MVT::f64, MVT::f32, Expand);
 
-  setOperationAction(ISD::FLT_ROUNDS_, MVT::i32, Custom);
+  setOperationAction(ISD::GET_ROUNDING, MVT::i32, Custom);
 
   // If we're enabling GP optimizations, use hardware square root
   if (!Subtarget.hasFSQRT() &&
       !(TM.Options.UnsafeFPMath && Subtarget.hasFRSQRTE() &&
         Subtarget.hasFRE()))
     setOperationAction(ISD::FSQRT, MVT::f64, Expand);
 
   if (!Subtarget.hasFSQRT() &&

                        {Chain, FP, DAG.getIntPtrConstant(0, dl)}, Flags);
     else
       FP = DAG.getNode(ISD::FP_ROUND, dl, MVT::f32, FP,
                        DAG.getIntPtrConstant(0, dl, /*isTarget=*/true));
   }
   return FP;
 }
 
-SDValue PPCTargetLowering::LowerFLT_ROUNDS_(SDValue Op,
+SDValue PPCTargetLowering::LowerGET_ROUNDING(SDValue Op,
                                              SelectionDAG &DAG) const {
   SDLoc dl(Op);
   /*
    The rounding mode is in bits 30:31 of FPSR, and has the following
    settings:
      00 Round to nearest
      01 Round to 0
      10 Round to +inf
      11 Round to -inf
 
-  FLT_ROUNDS, on the other hand, expects the following:
+  GET_ROUNDING, on the other hand, expects the following:
     -1 Undefined
      0 Round to 0
      1 Round to nearest
      2 Round to +inf
      3 Round to -inf
 
   To perform the conversion, we do:
     ((FPSCR & 0x3) ^ ((~FPSCR & 0x3) >> 1))

   case ISD::STRICT_FP_TO_UINT:
   case ISD::STRICT_FP_TO_SINT:
   case ISD::FP_TO_UINT:
   case ISD::FP_TO_SINT:         return LowerFP_TO_INT(Op, DAG, SDLoc(Op));
   case ISD::STRICT_UINT_TO_FP:
   case ISD::STRICT_SINT_TO_FP:
   case ISD::UINT_TO_FP:
   case ISD::SINT_TO_FP:         return LowerINT_TO_FP(Op, DAG);
-  case ISD::FLT_ROUNDS_:        return LowerFLT_ROUNDS_(Op, DAG);
+  case ISD::GET_ROUNDING:       return LowerGET_ROUNDING(Op, DAG);
 
   // Lower 64-bit shifts.
   case ISD::SHL_PARTS:          return LowerSHL_PARTS(Op, DAG);
   case ISD::SRL_PARTS:          return LowerSRL_PARTS(Op, DAG);
   case ISD::SRA_PARTS:          return LowerSRA_PARTS(Op, DAG);
 
   case ISD::FSHL:               return LowerFunnelShift(Op, DAG);
   case ISD::FSHR:               return LowerFunnelShift(Op, DAG);

llvm/lib/Target/RISCV/RISCVISelLowering.cpp

   if (Subtarget.hasStdExtF()) {
     setOperationAction({ISD::FP_TO_UINT_SAT, ISD::FP_TO_SINT_SAT}, XLenVT,
                        Custom);
 
     setOperationAction({ISD::STRICT_FP_TO_UINT, ISD::STRICT_FP_TO_SINT,
                         ISD::STRICT_UINT_TO_FP, ISD::STRICT_SINT_TO_FP},
                        XLenVT, Legal);
 
-    setOperationAction(ISD::FLT_ROUNDS_, XLenVT, Custom);
+    setOperationAction(ISD::GET_ROUNDING, XLenVT, Custom);
     setOperationAction(ISD::SET_ROUNDING, MVT::Other, Custom);
   }
 
   setOperationAction({ISD::GlobalAddress, ISD::BlockAddress, ISD::ConstantPool,
                       ISD::JumpTable},
                      XLenVT, Custom);
 
   setOperationAction(ISD::GlobalTLSAddress, XLenVT, Custom);

   case ISD::FCOPYSIGN:
     return lowerFixedLengthVectorFCOPYSIGNToRVV(Op, DAG);
   case ISD::MGATHER:
   case ISD::VP_GATHER:
     return lowerMaskedGather(Op, DAG);
   case ISD::MSCATTER:
   case ISD::VP_SCATTER:
     return lowerMaskedScatter(Op, DAG);
-  case ISD::FLT_ROUNDS_:
+  case ISD::GET_ROUNDING:
     return lowerGET_ROUNDING(Op, DAG);
   case ISD::SET_ROUNDING:
     return lowerSET_ROUNDING(Op, DAG);
   case ISD::EH_DWARF_CFA:
     return lowerEH_DWARF_CFA(Op, DAG);
   case ISD::VP_SELECT:
     return lowerVPOp(Op, DAG, RISCVISD::VSELECT_VL);
   case ISD::VP_MERGE:

   case ISD::VP_REDUCE_XOR:
   case ISD::VP_REDUCE_SMAX:
   case ISD::VP_REDUCE_UMAX:
   case ISD::VP_REDUCE_SMIN:
   case ISD::VP_REDUCE_UMIN:
     if (SDValue V = lowerVPREDUCE(SDValue(N, 0), DAG))
       Results.push_back(V);
     break;
-  case ISD::FLT_ROUNDS_: {
+  case ISD::GET_ROUNDING: {
     SDVTList VTs = DAG.getVTList(Subtarget.getXLenVT(), MVT::Other);
-    SDValue Res = DAG.getNode(ISD::FLT_ROUNDS_, DL, VTs, N->getOperand(0));
+    SDValue Res = DAG.getNode(ISD::GET_ROUNDING, DL, VTs, N->getOperand(0));
     Results.push_back(Res.getValue(0));
     Results.push_back(Res.getValue(1));
     break;
   }
   }
 }
 
 // Try to fold (<bop> x, (reduction.<bop> vec, start))

llvm/lib/Target/X86/X86ISelLowering.h

     SDValue LowerADDROFRETURNADDR(SDValue Op, SelectionDAG &DAG) const;
     SDValue LowerFRAMEADDR(SDValue Op, SelectionDAG &DAG) const;
     SDValue LowerFRAME_TO_ARGS_OFFSET(SDValue Op, SelectionDAG &DAG) const;
     SDValue LowerEH_RETURN(SDValue Op, SelectionDAG &DAG) const;
     SDValue lowerEH_SJLJ_SETJMP(SDValue Op, SelectionDAG &DAG) const;
     SDValue lowerEH_SJLJ_LONGJMP(SDValue Op, SelectionDAG &DAG) const;
     SDValue lowerEH_SJLJ_SETUP_DISPATCH(SDValue Op, SelectionDAG &DAG) const;
     SDValue LowerINIT_TRAMPOLINE(SDValue Op, SelectionDAG &DAG) const;
-    SDValue LowerFLT_ROUNDS_(SDValue Op, SelectionDAG &DAG) const;
+    SDValue LowerGET_ROUNDING(SDValue Op, SelectionDAG &DAG) const;
     SDValue LowerSET_ROUNDING(SDValue Op, SelectionDAG &DAG) const;
     SDValue LowerWin64_i128OP(SDValue Op, SelectionDAG &DAG) const;
     SDValue LowerWin64_FP_TO_INT128(SDValue Op, SelectionDAG &DAG,
                                     SDValue &Chain) const;
     SDValue LowerWin64_INT128_TO_FP(SDValue Op, SelectionDAG &DAG) const;
     SDValue LowerGC_TRANSITION(SDValue Op, SelectionDAG &DAG) const;
     SDValue LowerINTRINSIC_WO_CHAIN(SDValue Op, SelectionDAG &DAG) const;
     SDValue lowerFaddFsub(SDValue Op, SelectionDAG &DAG) const;

llvm/lib/Target/X86/X86ISelLowering.cpp

   setOperationAction(ISD::SIGN_EXTEND_INREG, MVT::i1   , Expand);
 
   setOperationAction(ISD::FREM             , MVT::f32  , Expand);
   setOperationAction(ISD::FREM             , MVT::f64  , Expand);
   setOperationAction(ISD::FREM             , MVT::f80  , Expand);
   setOperationAction(ISD::FREM             , MVT::f128 , Expand);
 
   if (!Subtarget.useSoftFloat() && Subtarget.hasX87()) {
-    setOperationAction(ISD::FLT_ROUNDS_    , MVT::i32  , Custom);
+    setOperationAction(ISD::GET_ROUNDING   , MVT::i32  , Custom);
     setOperationAction(ISD::SET_ROUNDING   , MVT::Other, Custom);
   }
 
   // Promote the i8 variants and force them on up to i32 which has a shorter
   // encoding.
   setOperationPromotedToType(ISD::CTTZ           , MVT::i8   , MVT::i32);
   setOperationPromotedToType(ISD::CTTZ_ZERO_UNDEF, MVT::i8   , MVT::i32);
   // Promoted i16. tzcntw has a false dependency on Intel CPUs. For BSF, we emit

                        DAG.getConstant(6, dl, MVT::i32));
     OutChains[3] = DAG.getStore(Root, dl, Disp, Addr,
                                 MachinePointerInfo(TrmpAddr, 6), Align(1));
 
     return DAG.getNode(ISD::TokenFactor, dl, MVT::Other, OutChains);
   }
 }
 
-SDValue X86TargetLowering::LowerFLT_ROUNDS_(SDValue Op,
+SDValue X86TargetLowering::LowerGET_ROUNDING(SDValue Op,
                                              SelectionDAG &DAG) const {
   /*
    The rounding mode is in bits 11:10 of FPSR, and has the following
    settings:
      00 Round to nearest
      01 Round to -inf
      10 Round to +inf
      11 Round to 0
 
-  FLT_ROUNDS, on the other hand, expects the following:
+  GET_ROUNDING, on the other hand, expects the following:
     -1 Undefined
      0 Round to 0
      1 Round to nearest
      2 Round to +inf
      3 Round to -inf
 
   To perform the conversion, we use a packed lookup table of the four 2-bit
   values that we can index by FPSP[11:10]

   case ISD::DYNAMIC_STACKALLOC: return LowerDYNAMIC_STACKALLOC(Op, DAG);
   case ISD::EH_RETURN:          return LowerEH_RETURN(Op, DAG);
   case ISD::EH_SJLJ_SETJMP:     return lowerEH_SJLJ_SETJMP(Op, DAG);
   case ISD::EH_SJLJ_LONGJMP:    return lowerEH_SJLJ_LONGJMP(Op, DAG);
   case ISD::EH_SJLJ_SETUP_DISPATCH:
     return lowerEH_SJLJ_SETUP_DISPATCH(Op, DAG);
   case ISD::INIT_TRAMPOLINE:    return LowerINIT_TRAMPOLINE(Op, DAG);
   case ISD::ADJUST_TRAMPOLINE:  return LowerADJUST_TRAMPOLINE(Op, DAG);
-  case ISD::FLT_ROUNDS_:        return LowerFLT_ROUNDS_(Op, DAG);
+  case ISD::GET_ROUNDING:       return LowerGET_ROUNDING(Op, DAG);
   case ISD::SET_ROUNDING:       return LowerSET_ROUNDING(Op, DAG);
   case ISD::CTLZ:
   case ISD::CTLZ_ZERO_UNDEF:    return LowerCTLZ(Op, Subtarget, DAG);
   case ISD::CTTZ:
   case ISD::CTTZ_ZERO_UNDEF:    return LowerCTTZ(Op, Subtarget, DAG);
   case ISD::MUL:                return LowerMUL(Op, Subtarget, DAG);
   case ISD::MULHS:
   case ISD::MULHU:              return LowerMULH(Op, Subtarget, DAG);

llvm/test/Bitcode/auto_upgrade_flt_rounds.test

+RUN: llvm-dis %p/Inputs/auto_upgrade_flt_rounds.bc -o - | FileCheck %s
+
+CHECK-LABEL: define signext i32 @get_rounds()
+CHECK: tail call i32 @llvm.get.rounding()

llvm/test/CodeGen/AArch64/arm64-fpcr.ll

 declare i64 @llvm.aarch64.get.fpcr() #0
 
 define i32 @GetFltRounds() {
 ; CHECK-LABEL: GetFltRounds
 ; CHECK: mrs x8, FPCR
 ; CHECK: add w8, w8, #1024, lsl #12
 ; CHECK: ubfx w0, w8, #22, #2
 ; CHECK: ret
-  %1 = tail call i32 @llvm.flt.rounds()
+  %1 = tail call i32 @llvm.get.rounding()
   ret i32 %1
 }
 
-declare i32 @llvm.flt.rounds() #0
+declare i32 @llvm.get.rounding() #0

llvm/test/CodeGen/AArch64/strict-fp-opt.ll

   %sub = call float @llvm.experimental.constrained.fsub.f32(float %add1, float %add2, metadata !"round.dynamic", metadata !"fpexcept.ignore") #0
   ret float %sub
 }
 
 declare float @llvm.experimental.constrained.fadd.f32(float, float, metadata, metadata) #0
 declare float @llvm.experimental.constrained.fsub.f32(float, float, metadata, metadata) #0
 declare float @llvm.experimental.constrained.fmul.f32(float, float, metadata, metadata) #0
 declare float @llvm.experimental.constrained.fdiv.f32(float, float, metadata, metadata) #0
-declare i32 @llvm.flt.rounds()
+declare i32 @llvm.get.rounding()
 declare void @llvm.set.rounding(i32)
 
 attributes #0 = { "strictfp" }

llvm/test/CodeGen/ARM/2012-03-05-FPSCR-bug.ll

 
 ; <label>:2                                       ; preds = %1, %0
   br i1 undef, label %3, label %8
 
 ; <label>:3                                       ; preds = %2
   br i1 undef, label %4, label %7
 
 ; <label>:4                                       ; preds = %3
-  %5 = call i32 @llvm.flt.rounds()
+  %5 = call i32 @llvm.get.rounding()
   %6 = icmp eq i32 %5, 1
   br i1 %6, label %8, label %7
 
 ; <label>:7                                       ; preds = %4, %3
   unreachable
 
 ; <label>:8                                       ; preds = %4, %2
   br i1 undef, label %9, label %10
 
 ; <label>:9                                       ; preds = %8
   br label %10
 
 ; <label>:10                                      ; preds = %9, %8
   ret void
 }
 
-declare i32 @llvm.flt.rounds() nounwind
+declare i32 @llvm.get.rounding() nounwind

llvm/test/CodeGen/ARM/fpscr-intrinsics.ll

 ; RUN: llc < %s -O0 -mtriple=armv7-eabi -mcpu=cortex-a8 -mattr=+neon,+fp-armv8 | FileCheck %s
 ; RUN: llc < %s -O3 -mtriple=armv7-eabi -mcpu=cortex-a8 -mattr=+neon,+fp-armv8 | FileCheck %s
 
 @a = common global double 0.000000e+00, align 8
 
 ; Function Attrs: noinline nounwind uwtable
 define void @strtod() {
 entry:
   ; CHECK: vmrs r{{[0-9]+}}, fpscr
-  %0 = call i32 @llvm.flt.rounds()
+  %0 = call i32 @llvm.get.rounding()
   %tobool = icmp ne i32 %0, 0
   br i1 %tobool, label %if.then, label %if.end
 
 if.then:                                          ; preds = %entry
   store double 5.000000e-01, double* @a, align 8
   br label %if.end
 
 if.end:                                           ; preds = %if.then, %entry

 
 ; Function Attrs: nounwind readonly
 declare i32 @llvm.arm.get.fpscr()
 
 ; Function Attrs: nounwind writeonly
 declare void @llvm.arm.set.fpscr(i32)
 
 ; Function Attrs: nounwind
-declare i32 @llvm.flt.rounds()
+declare i32 @llvm.get.rounding()

llvm/test/CodeGen/ARM/no-fpscr-liveness.ll

 ; CHECK: vcmp
 ; ...
 ; CHECK: add sp, #8
 ; CHECK: bx lr
 define i32 @eggs(double* nocapture readnone %arg) {
 bb:
   %tmp = alloca %struct.wibble, align 4
   %tmp1 = bitcast %struct.wibble* %tmp to i8*
-  %tmp2 = tail call i32 @llvm.flt.rounds()
+  %tmp2 = tail call i32 @llvm.get.rounding()
   %tmp3 = ptrtoint %struct.wibble* %tmp to i32
   %tmp4 = sitofp i32 %tmp3 to double
   %tmp5 = fmul double %tmp4, 0x0123456789ABCDEF
   %tmp6 = fptosi double %tmp5 to i32
   %tmp7 = fcmp une double %tmp5, 0.000000e+00
   %tmp8 = sitofp i32 %tmp6 to double
   %tmp9 = fcmp une double %tmp5, %tmp8
   %tmp10 = and i1 %tmp7, %tmp9
   %tmp11 = sext i1 %tmp10 to i32
   %tmp12 = add nsw i32 %tmp11, %tmp6
   store i32 %tmp12, i32* @global, align 4
   %tmp13 = icmp ne i32 %tmp12, 0
   %tmp14 = icmp ne i32 %tmp2, 0
   %tmp15 = and i1 %tmp14, %tmp13
   br i1 %tmp15, label %bb16, label %bb18
 
 bb16:                                             ; preds = %bb
   %tmp17 = load i32, i32* @global.1, align 4
   br label %bb18
 
 bb18:                                             ; preds = %bb16, %bb
   ret i32 undef
 }
 
-declare i32 @llvm.flt.rounds()
+declare i32 @llvm.get.rounding()
 declare i32 @zot(...)

llvm/test/CodeGen/MSP430/flt_rounds.ll

 ; RUN: llc -verify-machineinstrs < %s -march=msp430
 
 define i16 @foo() {
 entry:
-  %0 = call i32 @llvm.flt.rounds()
+  %0 = call i32 @llvm.get.rounding()
   %1 = trunc i32 %0 to i16
   ret i16 %1
 }
 
-declare i32 @llvm.flt.rounds() nounwind 
+declare i32 @llvm.get.rounding() nounwind

llvm/test/CodeGen/PowerPC/frounds.ll

 ; DM-NEXT:    xor 3, 3, 4
 ; DM-NEXT:    stw 3, -8(1)
 ; DM-NEXT:    stw 3, -4(1)
 ; DM-NEXT:    blr
 entry:
   %retval = alloca i32    ; <ptr> [#uses=2]
   %tmp = alloca i32   ; <ptr> [#uses=2]
   %"alloca point" = bitcast i32 0 to i32    ; <i32> [#uses=0]
-  %tmp1 = call i32 @llvm.flt.rounds( )    ; <i32> [#uses=1]
+  %tmp1 = call i32 @llvm.get.rounding( )    ; <i32> [#uses=1]
   store i32 %tmp1, ptr %tmp, align 4
   %tmp2 = load i32, ptr %tmp, align 4    ; <i32> [#uses=1]
   store i32 %tmp2, ptr %retval, align 4
   br label %return
 
 return:   ; preds = %entry
   %retval3 = load i32, ptr %retval    ; <i32> [#uses=1]
   ret i32 %retval3
 }
 
-declare i32 @llvm.flt.rounds() nounwind
+declare i32 @llvm.get.rounding() nounwind

llvm/test/CodeGen/RISCV/flt-rounds.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc -mtriple=riscv32 -verify-machineinstrs < %s \
 ; RUN:   | FileCheck -check-prefix=RV32I %s
 ; RUN: llc -mtriple=riscv64 -verify-machineinstrs < %s \
 ; RUN:   | FileCheck -check-prefix=RV64I %s
 
-declare i32 @llvm.flt.rounds()
+declare i32 @llvm.get.rounding()
 
 define i32 @test_flt_rounds() nounwind {
 ; RV32I-LABEL: test_flt_rounds:
 ; RV32I:       # %bb.0:
 ; RV32I-NEXT:    li a0, 1
 ; RV32I-NEXT:    ret
 ;
 ; RV64I-LABEL: test_flt_rounds:
 ; RV64I:       # %bb.0:
 ; RV64I-NEXT:    li a0, 1
 ; RV64I-NEXT:    ret
-  %1 = call i32 @llvm.flt.rounds()
+  %1 = call i32 @llvm.get.rounding()
   ret i32 %1
 }

llvm/test/CodeGen/RISCV/fpenv.ll

 ; RV64IF:       # %bb.0:
 ; RV64IF-NEXT:    frrm a0
 ; RV64IF-NEXT:    slliw a0, a0, 2
 ; RV64IF-NEXT:    lui a1, 66
 ; RV64IF-NEXT:    addiw a1, a1, 769
 ; RV64IF-NEXT:    srl a0, a1, a0
 ; RV64IF-NEXT:    andi a0, a0, 7
 ; RV64IF-NEXT:    ret
-  %rm = call i32 @llvm.flt.rounds()
+  %rm = call i32 @llvm.get.rounding()
   ret i32 %rm
 }
 
 define void @func_02(i32 %rm) {
 ; RV32IF-LABEL: func_02:
 ; RV32IF:       # %bb.0:
 ; RV32IF-NEXT:    slli a0, a0, 2
 ; RV32IF-NEXT:    lui a1, 66

 ; RV64IF:       # %bb.0:
 ; RV64IF-NEXT:    fsrmi 4
 ; RV64IF-NEXT:    ret
   call void @llvm.set.rounding(i32 4)
   ret void
 }
 
 declare void @llvm.set.rounding(i32)
-declare i32 @llvm.flt.rounds()
+declare i32 @llvm.get.rounding()

llvm/test/CodeGen/X86/flt-rounds.ll

 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
 ; RUN: llc -mtriple=i686-unknown-linux-gnu -mattr=-sse -verify-machineinstrs < %s | FileCheck %s --check-prefix=X86
 ; RUN: llc -mtriple=i686-unknown-linux-gnu -mattr=-sse2 -verify-machineinstrs < %s | FileCheck %s --check-prefix=X86
 ; RUN: llc -mtriple=x86_64-unknown-linux-gnu -verify-machineinstrs < %s | FileCheck %s --check-prefix=X64
 
-declare i32 @llvm.flt.rounds()
+declare i32 @llvm.get.rounding()
 
 define i32 @test_flt_rounds() nounwind {
 ; X86-LABEL: test_flt_rounds:
 ; X86:       # %bb.0:
 ; X86-NEXT:    subl $2, %esp
 ; X86-NEXT:    fnstcw (%esp)
 ; X86-NEXT:    movzwl (%esp), %ecx
 ; X86-NEXT:    shrl $9, %ecx

 ; X64-NEXT:    movzwl -{{[0-9]+}}(%rsp), %ecx
 ; X64-NEXT:    shrl $9, %ecx
 ; X64-NEXT:    andb $6, %cl
 ; X64-NEXT:    movl $45, %eax
 ; X64-NEXT:    # kill: def $cl killed $cl killed $ecx
 ; X64-NEXT:    shrl %cl, %eax
 ; X64-NEXT:    andl $3, %eax
 ; X64-NEXT:    retq
-  %1 = call i32 @llvm.flt.rounds()
+  %1 = call i32 @llvm.get.rounding()
   ret i32 %1
 }
 
 ; Make sure we preserve order with fesetround.
 define i32 @multiple_flt_rounds() nounwind {
 ; X86-LABEL: multiple_flt_rounds:
 ; X86:       # %bb.0: # %entry
 ; X86-NEXT:    pushl %ebx

 ; X64-NEXT:    setne %al
 ; X64-NEXT:    addq $16, %rsp
 ; X64-NEXT:    popq %rbx
 ; X64-NEXT:    popq %r14
 ; X64-NEXT:    popq %rbp
 ; X64-NEXT:    retq
 entry:
   %call = tail call i32 @fesetround(i32 1024)
-  %0 = tail call i32 @llvm.flt.rounds()
+  %0 = tail call i32 @llvm.get.rounding()
   %cmp = icmp ne i32 %0, 3
   %spec.select = zext i1 %cmp to i32
   %call1 = tail call i32 @fesetround(i32 0)
-  %1 = tail call i32 @llvm.flt.rounds()
+  %1 = tail call i32 @llvm.get.rounding()
   %cmp2 = icmp eq i32 %1, 1
   %inc4 = select i1 %cmp, i32 2, i32 1
   %errs.1 = select i1 %cmp2, i32 %spec.select, i32 %inc4
   %call6 = tail call i32 @fesetround(i32 3072)
-  %2 = tail call i32 @llvm.flt.rounds()
+  %2 = tail call i32 @llvm.get.rounding()
   %cmp7 = icmp ne i32 %2, 0
   %inc9 = zext i1 %cmp7 to i32
   %spec.select22 = add nuw nsw i32 %errs.1, %inc9
   %call11 = tail call i32 @fesetround(i32 2048)
-  %3 = tail call i32 @llvm.flt.rounds()
+  %3 = tail call i32 @llvm.get.rounding()
   %cmp12 = icmp ne i32 %3, 2
   %inc14.neg = sext i1 %cmp12 to i32
   %cmp16 = icmp ne i32 %spec.select22, %inc14.neg
   %cond = zext i1 %cmp16 to i32
   ret i32 %cond
 }
 
 ; Function Attrs: nounwind
 declare dso_local i32 @fesetround(i32) local_unnamed_addr #1