#include "llvm/Support/MathExtras.h"
#include "llvm/Support/raw_ostream.h"
#include "llvm/Target/TargetLowering.h"
-#include "llvm/Target/TargetMachine.h"
#include "llvm/Target/TargetOptions.h"
#include "llvm/Target/TargetRegisterInfo.h"
#include "llvm/Target/TargetSubtargetInfo.h"
SDValue visitFCEIL(SDNode *N);
SDValue visitFTRUNC(SDNode *N);
SDValue visitFFLOOR(SDNode *N);
+ SDValue visitFMINNUM(SDNode *N);
+ SDValue visitFMAXNUM(SDNode *N);
SDValue visitBRCOND(SDNode *N);
SDValue visitBR_CC(SDNode *N);
SDValue visitLOAD(SDNode *N);
SDValue BuildUDIV(SDNode *N);
SDValue BuildReciprocalEstimate(SDValue Op);
SDValue BuildRsqrtEstimate(SDValue Op);
+ SDValue BuildRsqrtNROneConst(SDValue Op, SDValue Est, unsigned Iterations);
+ SDValue BuildRsqrtNRTwoConst(SDValue Op, SDValue Est, unsigned Iterations);
SDValue MatchBSwapHWordLow(SDNode *N, SDValue N0, SDValue N1,
bool DemandHighBits = true);
SDValue MatchBSwapHWord(SDNode *N, SDValue N0, SDValue N1);
case ISD::FNEG: return visitFNEG(N);
case ISD::FABS: return visitFABS(N);
case ISD::FFLOOR: return visitFFLOOR(N);
+ case ISD::FMINNUM: return visitFMINNUM(N);
+ case ISD::FMAXNUM: return visitFMAXNUM(N);
case ISD::FCEIL: return visitFCEIL(N);
case ISD::FTRUNC: return visitFTRUNC(N);
case ISD::BRCOND: return visitBRCOND(N);
return DAG.getNode(ISD::SUB, DL, VT, N1, ZExt);
}
+ // add X, (sextinreg Y i1) -> sub X, (and Y 1)
+ if (N1.getOpcode() == ISD::SIGN_EXTEND_INREG) {
+ VTSDNode *TN = cast<VTSDNode>(N1.getOperand(1));
+ if (TN->getVT() == MVT::i1) {
+ SDLoc DL(N);
+ SDValue ZExt = DAG.getNode(ISD::AND, DL, VT, N1.getOperand(0),
+ DAG.getConstant(1, VT));
+ return DAG.getNode(ISD::SUB, DL, VT, N0, ZExt);
+ }
+ }
+
return SDValue();
}
VT);
}
+ // sub X, (sextinreg Y i1) -> add X, (and Y 1)
+ if (N1.getOpcode() == ISD::SIGN_EXTEND_INREG) {
+ VTSDNode *TN = cast<VTSDNode>(N1.getOperand(1));
+ if (TN->getVT() == MVT::i1) {
+ SDLoc DL(N);
+ SDValue ZExt = DAG.getNode(ISD::AND, DL, VT, N1.getOperand(0),
+ DAG.getConstant(1, VT));
+ return DAG.getNode(ISD::ADD, DL, VT, N0, ZExt);
+ }
+ }
+
return SDValue();
}
/// ((x & 0x0000ff00) >> 8) |
/// ((x & 0x00ff0000) << 8) |
/// ((x & 0xff000000) >> 8)
-static bool isBSwapHWordElement(SDValue N, SmallVectorImpl<SDNode *> &Parts) {
+static bool isBSwapHWordElement(SDValue N, MutableArrayRef<SDNode *> Parts) {
if (!N.getNode()->hasOneUse())
return false;
if (!TLI.isOperationLegal(ISD::BSWAP, VT))
return SDValue();
- SmallVector<SDNode*,4> Parts(4, (SDNode*)nullptr);
// Look for either
// (or (or (and), (and)), (or (and), (and)))
// (or (or (or (and), (and)), (and)), (and))
return SDValue();
SDValue N00 = N0.getOperand(0);
SDValue N01 = N0.getOperand(1);
+ SDNode *Parts[4] = {};
if (N1.getOpcode() == ISD::OR &&
N00.getNumOperands() == 2 && N01.getNumOperands() == 2) {
if (N0.getOpcode() == ISD::SETCC) {
if ((!LegalOperations &&
TLI.isOperationLegalOrCustom(ISD::SELECT_CC, VT)) ||
- TLI.isOperationLegal(ISD::SELECT_CC, VT))
+ TLI.isOperationLegal(ISD::SELECT_CC, VT))
return DAG.getNode(ISD::SELECT_CC, SDLoc(N), VT,
N0.getOperand(0), N0.getOperand(1),
N1, N2, N0.getOperand(2));
if (!LegalOperations || TLI.isOperationLegal(ISD::SETCC, SetCCVT)) {
SDLoc DL(N);
ISD::CondCode CC = cast<CondCodeSDNode>(N0.getOperand(2))->get();
- SDValue SetCC = DAG.getSetCC(DL,
- SetCCVT,
+ SDValue SetCC = DAG.getSetCC(DL, SetCCVT,
N0.getOperand(0), N0.getOperand(1), CC);
- EVT SelectVT = getSetCCResultType(VT);
- return DAG.getSelect(DL, VT,
- DAG.getSExtOrTrunc(SetCC, DL, SelectVT),
+ return DAG.getSelect(DL, VT, SetCC,
NegOne, DAG.getConstant(0, VT));
-
}
}
}
ConstantFPSDNode *N1CFP = dyn_cast<ConstantFPSDNode>(N1);
EVT VT = N->getValueType(0);
const TargetOptions &Options = DAG.getTarget().Options;
-
+
// fold vector ops
if (VT.isVector()) {
SDValue FoldedVOp = SimplifyVBinOp(N);
isNegatibleForFree(N1, LegalOperations, TLI, &Options) == 2)
return DAG.getNode(ISD::FSUB, SDLoc(N), VT, N0,
GetNegatedExpression(N1, DAG, LegalOperations));
-
+
// fold (fadd (fneg A), B) -> (fsub B, A)
if ((!LegalOperations || TLI.isOperationLegalOrCustom(ISD::FSUB, VT)) &&
isNegatibleForFree(N0, LegalOperations, TLI, &Options) == 2)
// No FP constant should be created after legalization as Instruction
// Selection pass has a hard time dealing with FP constants.
bool AllowNewConst = (Level < AfterLegalizeDAG);
-
+
// fold (fadd A, 0) -> A
if (N1CFP && N1CFP->getValueAPF().isZero())
return N0;
return DAG.getNode(ISD::FADD, SDLoc(N), VT, N0.getOperand(0),
DAG.getNode(ISD::FADD, SDLoc(N), VT,
N0.getOperand(1), N1));
-
+
// If allowed, fold (fadd (fneg x), x) -> 0.0
if (AllowNewConst && N0.getOpcode() == ISD::FNEG && N0.getOperand(0) == N1)
return DAG.getConstantFP(0.0, VT);
-
+
// If allowed, fold (fadd x, (fneg x)) -> 0.0
if (AllowNewConst && N1.getOpcode() == ISD::FNEG && N1.getOperand(0) == N0)
return DAG.getConstantFP(0.0, VT);
-
+
// We can fold chains of FADD's of the same value into multiplications.
// This transform is not safe in general because we are reducing the number
// of rounding steps.
if (N0.getOpcode() == ISD::FMUL) {
ConstantFPSDNode *CFP00 = dyn_cast<ConstantFPSDNode>(N0.getOperand(0));
ConstantFPSDNode *CFP01 = dyn_cast<ConstantFPSDNode>(N0.getOperand(1));
-
+
// (fadd (fmul x, c), x) -> (fmul x, c+1)
if (CFP01 && !CFP00 && N0.getOperand(0) == N1) {
SDValue NewCFP = DAG.getNode(ISD::FADD, SDLoc(N), VT,
DAG.getConstantFP(1.0, VT));
return DAG.getNode(ISD::FMUL, SDLoc(N), VT, N1, NewCFP);
}
-
+
// (fadd (fmul x, c), (fadd x, x)) -> (fmul x, c+2)
if (CFP01 && !CFP00 && N1.getOpcode() == ISD::FADD &&
N1.getOperand(0) == N1.getOperand(1) &&
N0.getOperand(0), NewCFP);
}
}
-
+
if (N1.getOpcode() == ISD::FMUL) {
ConstantFPSDNode *CFP10 = dyn_cast<ConstantFPSDNode>(N1.getOperand(0));
ConstantFPSDNode *CFP11 = dyn_cast<ConstantFPSDNode>(N1.getOperand(1));
-
+
// (fadd x, (fmul x, c)) -> (fmul x, c+1)
if (CFP11 && !CFP10 && N1.getOperand(0) == N0) {
SDValue NewCFP = DAG.getNode(ISD::FADD, SDLoc(N), VT,
return DAG.getNode(ISD::FMUL, SDLoc(N), VT,
N1, DAG.getConstantFP(3.0, VT));
}
-
+
if (N1.getOpcode() == ISD::FADD && AllowNewConst) {
ConstantFPSDNode *CFP10 = dyn_cast<ConstantFPSDNode>(N1.getOperand(0));
// (fadd x, (fadd x, x)) -> (fmul x, 3.0)
return DAG.getNode(ISD::FMUL, SDLoc(N), VT,
N0, DAG.getConstantFP(3.0, VT));
}
-
+
// (fadd (fadd x, x), (fadd x, x)) -> (fmul x, 4.0)
if (AllowNewConst &&
N0.getOpcode() == ISD::FADD && N1.getOpcode() == ISD::FADD &&
N0.getOperand(0), DAG.getConstantFP(4.0, VT));
}
} // enable-unsafe-fp-math
-
+
// FADD -> FMA combines:
if ((Options.AllowFPOpFusion == FPOpFusion::Fast || Options.UnsafeFPMath) &&
- DAG.getTarget()
- .getSubtargetImpl()
- ->getTargetLowering()
- ->isFMAFasterThanFMulAndFAdd(VT) &&
+ TLI.isFMAFasterThanFMulAndFAdd(VT) &&
(!LegalOperations || TLI.isOperationLegalOrCustom(ISD::FMA, VT))) {
// fold (fadd (fmul x, y), z) -> (fma x, y, z)
// FSUB -> FMA combines:
if ((Options.AllowFPOpFusion == FPOpFusion::Fast || Options.UnsafeFPMath) &&
- DAG.getTarget().getSubtargetImpl()
- ->getTargetLowering()
- ->isFMAFasterThanFMulAndFAdd(VT) &&
+ TLI.isFMAFasterThanFMulAndFAdd(VT) &&
(!LegalOperations || TLI.isOperationLegalOrCustom(ISD::FMA, VT))) {
// fold (fsub (fmul x, y), z) -> (fma x, y, (fneg z))
return DAG.getNode(ISD::FMUL, SDLoc(N), VT, N0,
DAG.getConstantFP(Recip, VT));
}
-
+
// If this FDIV is part of a reciprocal square root, it may be folded
// into a target-specific square root estimate instruction.
if (N1.getOpcode() == ISD::FSQRT) {
if (SDValue RV = BuildRsqrtEstimate(N1.getOperand(0))) {
- AddToWorklist(RV.getNode());
return DAG.getNode(ISD::FMUL, DL, VT, N0, RV);
}
} else if (N1.getOpcode() == ISD::FP_EXTEND &&
N1.getOperand(0).getOpcode() == ISD::FSQRT) {
if (SDValue RV = BuildRsqrtEstimate(N1.getOperand(0).getOperand(0))) {
- AddToWorklist(RV.getNode());
RV = DAG.getNode(ISD::FP_EXTEND, SDLoc(N1), VT, RV);
AddToWorklist(RV.getNode());
return DAG.getNode(ISD::FMUL, DL, VT, N0, RV);
} else if (N1.getOpcode() == ISD::FP_ROUND &&
N1.getOperand(0).getOpcode() == ISD::FSQRT) {
if (SDValue RV = BuildRsqrtEstimate(N1.getOperand(0).getOperand(0))) {
- AddToWorklist(RV.getNode());
RV = DAG.getNode(ISD::FP_ROUND, SDLoc(N1), VT, RV, N1.getOperand(1));
AddToWorklist(RV.getNode());
return DAG.getNode(ISD::FMUL, DL, VT, N0, RV);
}
+ } else if (N1.getOpcode() == ISD::FMUL) {
+ // Look through an FMUL. Even though this won't remove the FDIV directly,
+ // it's still worthwhile to get rid of the FSQRT if possible.
+ SDValue SqrtOp;
+ SDValue OtherOp;
+ if (N1.getOperand(0).getOpcode() == ISD::FSQRT) {
+ SqrtOp = N1.getOperand(0);
+ OtherOp = N1.getOperand(1);
+ } else if (N1.getOperand(1).getOpcode() == ISD::FSQRT) {
+ SqrtOp = N1.getOperand(1);
+ OtherOp = N1.getOperand(0);
+ }
+ if (SqrtOp.getNode()) {
+ // We found a FSQRT, so try to make this fold:
+ // x / (y * sqrt(z)) -> x * (rsqrt(z) / y)
+ if (SDValue RV = BuildRsqrtEstimate(SqrtOp.getOperand(0))) {
+ RV = DAG.getNode(ISD::FDIV, SDLoc(N1), VT, RV, OtherOp);
+ AddToWorklist(RV.getNode());
+ return DAG.getNode(ISD::FMUL, DL, VT, N0, RV);
+ }
+ }
}
-
+
// Fold into a reciprocal estimate and multiply instead of a real divide.
if (SDValue RV = BuildReciprocalEstimate(N1)) {
AddToWorklist(RV.getNode());
SDValue DAGCombiner::visitFSQRT(SDNode *N) {
if (DAG.getTarget().Options.UnsafeFPMath) {
- // Compute this as 1/(1/sqrt(X)): the reciprocal of the reciprocal sqrt.
+ // Compute this as X * (1/sqrt(X)) = X * (X ** -0.5)
if (SDValue RV = BuildRsqrtEstimate(N->getOperand(0))) {
+ EVT VT = RV.getValueType();
+ RV = DAG.getNode(ISD::FMUL, SDLoc(N), VT, N->getOperand(0), RV);
AddToWorklist(RV.getNode());
- RV = BuildReciprocalEstimate(RV);
- if (RV.getNode()) {
- // Unfortunately, RV is now NaN if the input was exactly 0.
- // Select out this case and force the answer to 0.
- EVT VT = RV.getValueType();
-
- SDValue Zero = DAG.getConstantFP(0.0, VT);
- SDValue ZeroCmp =
- DAG.getSetCC(SDLoc(N), TLI.getSetCCResultType(*DAG.getContext(), VT),
- N->getOperand(0), Zero, ISD::SETEQ);
- AddToWorklist(ZeroCmp.getNode());
- AddToWorklist(RV.getNode());
- RV = DAG.getNode(VT.isVector() ? ISD::VSELECT : ISD::SELECT,
- SDLoc(N), VT, ZeroCmp, Zero, RV);
- return RV;
- }
+ // Unfortunately, RV is now NaN if the input was exactly 0.
+ // Select out this case and force the answer to 0.
+ SDValue Zero = DAG.getConstantFP(0.0, VT);
+ SDValue ZeroCmp =
+ DAG.getSetCC(SDLoc(N), TLI.getSetCCResultType(*DAG.getContext(), VT),
+ N->getOperand(0), Zero, ISD::SETEQ);
+ AddToWorklist(ZeroCmp.getNode());
+ AddToWorklist(RV.getNode());
+
+ RV = DAG.getNode(VT.isVector() ? ISD::VSELECT : ISD::SELECT,
+ SDLoc(N), VT, ZeroCmp, Zero, RV);
+ return RV;
}
}
return SDValue();
return SDValue();
}
+SDValue DAGCombiner::visitFMINNUM(SDNode *N) {
+ SDValue N0 = N->getOperand(0);
+ SDValue N1 = N->getOperand(1);
+ const ConstantFPSDNode *N0CFP = dyn_cast<ConstantFPSDNode>(N0);
+ const ConstantFPSDNode *N1CFP = dyn_cast<ConstantFPSDNode>(N1);
+
+ if (N0CFP && N1CFP) {
+ const APFloat &C0 = N0CFP->getValueAPF();
+ const APFloat &C1 = N1CFP->getValueAPF();
+ return DAG.getConstantFP(minnum(C0, C1), N->getValueType(0));
+ }
+
+ if (N0CFP) {
+ EVT VT = N->getValueType(0);
+ // Canonicalize to constant on RHS.
+ return DAG.getNode(ISD::FMINNUM, SDLoc(N), VT, N1, N0);
+ }
+
+ return SDValue();
+}
+
+SDValue DAGCombiner::visitFMAXNUM(SDNode *N) {
+ SDValue N0 = N->getOperand(0);
+ SDValue N1 = N->getOperand(1);
+ const ConstantFPSDNode *N0CFP = dyn_cast<ConstantFPSDNode>(N0);
+ const ConstantFPSDNode *N1CFP = dyn_cast<ConstantFPSDNode>(N1);
+
+ if (N0CFP && N1CFP) {
+ const APFloat &C0 = N0CFP->getValueAPF();
+ const APFloat &C1 = N1CFP->getValueAPF();
+ return DAG.getConstantFP(maxnum(C0, C1), N->getValueType(0));
+ }
+
+ if (N0CFP) {
+ EVT VT = N->getValueType(0);
+ // Canonicalize to constant on RHS.
+ return DAG.getNode(ISD::FMAXNUM, SDLoc(N), VT, N1, N0);
+ }
+
+ return SDValue();
+}
+
SDValue DAGCombiner::visitFABS(SDNode *N) {
SDValue N0 = N->getOperand(0);
EVT VT = N->getValueType(0);
// fold (fabs c1) -> fabs(c1)
if (isa<ConstantFPSDNode>(N0))
return DAG.getNode(ISD::FABS, SDLoc(N), VT, N0);
-
+
// fold (fabs (fabs x)) -> (fabs x)
if (N0.getOpcode() == ISD::FABS)
return N->getOperand(0);
}
}
- bool UseAA = CombinerAA.getNumOccurrences() > 0 ? CombinerAA :
- TLI.getTargetMachine().getSubtarget<TargetSubtargetInfo>().useAA();
+ bool UseAA = CombinerAA.getNumOccurrences() > 0 ? CombinerAA
+ : DAG.getSubtarget().useAA();
#ifndef NDEBUG
if (CombinerAAOnlyFunc.getNumOccurrences() &&
CombinerAAOnlyFunc != DAG.getMachineFunction().getName())
// At this point, we know that we perform a cross-register-bank copy.
// Check if it is expensive.
- const TargetRegisterInfo *TRI =
- TLI.getTargetMachine().getSubtargetImpl()->getRegisterInfo();
+ const TargetRegisterInfo *TRI = DAG->getSubtarget().getRegisterInfo();
// Assume bitcasts are cheap, unless both register classes do not
// explicitly share a common sub class.
if (!TRI || TRI->getCommonSubClass(ArgRC, ResRC))
if (NewST.getNode())
return NewST;
- bool UseAA = CombinerAA.getNumOccurrences() > 0 ? CombinerAA :
- TLI.getTargetMachine().getSubtarget<TargetSubtargetInfo>().useAA();
+ bool UseAA = CombinerAA.getNumOccurrences() > 0 ? CombinerAA
+ : DAG.getSubtarget().useAA();
#ifndef NDEBUG
if (CombinerAAOnlyFunc.getNumOccurrences() &&
CombinerAAOnlyFunc != DAG.getMachineFunction().getName())
if (isUndefMask)
return DAG.getUNDEF(VT);
-
+
bool CommuteOperands = false;
if (N0.getOperand(1).getOpcode() != ISD::UNDEF) {
// To be valid, the combine shuffle mask should only reference elements
// The combined shuffle must map each index to itself.
IsIdentityMask = (unsigned)Mask[i] == i + BaseMaskIndex;
}
-
+
if (IsIdentityMask) {
if (CommuteOperands)
// optimize shuffle(shuffle(x, y), undef) -> y.
return OtherSV->getOperand(1);
-
+
// optimize shuffle(shuffle(x, undef), undef) -> x
// optimize shuffle(shuffle(x, y), undef) -> x
return OtherSV->getOperand(0);
// It is safe to replace the two loads if they have different alignments,
// but the new load must be the minimum (most restrictive) alignment of the
// inputs.
- bool isInvariant = LLD->getAlignment() & RLD->getAlignment();
+ bool isInvariant = LLD->isInvariant() & RLD->isInvariant();
unsigned Alignment = std::min(LLD->getAlignment(), RLD->getAlignment());
if (LLD->getExtensionType() == ISD::NON_EXTLOAD) {
Load = DAG.getLoad(TheSelect->getValueType(0),
return SDValue();
}
-SDValue DAGCombiner::BuildRsqrtEstimate(SDValue Op) {
- if (Level >= AfterLegalizeDAG)
- return SDValue();
+/// Newton iteration for a function: F(X) is X_{i+1} = X_i - F(X_i)/F'(X_i)
+/// For the reciprocal sqrt, we need to find the zero of the function:
+/// F(X) = 1/X^2 - A [which has a zero at X = 1/sqrt(A)]
+/// =>
+/// X_{i+1} = X_i (1.5 - A X_i^2 / 2)
+/// As a result, we precompute A/2 prior to the iteration loop.
+SDValue DAGCombiner::BuildRsqrtNROneConst(SDValue Arg, SDValue Est,
+ unsigned Iterations) {
+ EVT VT = Arg.getValueType();
+ SDLoc DL(Arg);
+ SDValue ThreeHalves = DAG.getConstantFP(1.5, VT);
- // Expose the DAG combiner to the target combiner implementations.
- TargetLowering::DAGCombinerInfo DCI(DAG, Level, false, this);
- unsigned Iterations = 0;
- if (SDValue Est = TLI.getRsqrtEstimate(Op, DCI, Iterations)) {
- if (Iterations) {
- // Newton iteration for a function: F(X) is X_{i+1} = X_i - F(X_i)/F'(X_i)
- // For the reciprocal sqrt, we need to find the zero of the function:
- // F(X) = 1/X^2 - A [which has a zero at X = 1/sqrt(A)]
- // =>
- // X_{i+1} = X_i (1.5 - A X_i^2 / 2)
- // As a result, we precompute A/2 prior to the iteration loop.
- EVT VT = Op.getValueType();
- SDLoc DL(Op);
- SDValue FPThreeHalves = DAG.getConstantFP(1.5, VT);
+ // We now need 0.5 * Arg which we can write as (1.5 * Arg - Arg) so that
+ // this entire sequence requires only one FP constant.
+ SDValue HalfArg = DAG.getNode(ISD::FMUL, DL, VT, ThreeHalves, Arg);
+ AddToWorklist(HalfArg.getNode());
- AddToWorklist(Est.getNode());
+ HalfArg = DAG.getNode(ISD::FSUB, DL, VT, HalfArg, Arg);
+ AddToWorklist(HalfArg.getNode());
- // We now need 0.5 * Arg which we can write as (1.5 * Arg - Arg) so that
- // this entire sequence requires only one FP constant.
- SDValue HalfArg = DAG.getNode(ISD::FMUL, DL, VT, FPThreeHalves, Op);
- AddToWorklist(HalfArg.getNode());
+ // Newton iterations: Est = Est * (1.5 - HalfArg * Est * Est)
+ for (unsigned i = 0; i < Iterations; ++i) {
+ SDValue NewEst = DAG.getNode(ISD::FMUL, DL, VT, Est, Est);
+ AddToWorklist(NewEst.getNode());
- HalfArg = DAG.getNode(ISD::FSUB, DL, VT, HalfArg, Op);
- AddToWorklist(HalfArg.getNode());
+ NewEst = DAG.getNode(ISD::FMUL, DL, VT, HalfArg, NewEst);
+ AddToWorklist(NewEst.getNode());
- // Newton iterations: Est = Est * (1.5 - HalfArg * Est * Est)
- for (unsigned i = 0; i < Iterations; ++i) {
- SDValue NewEst = DAG.getNode(ISD::FMUL, DL, VT, Est, Est);
- AddToWorklist(NewEst.getNode());
+ NewEst = DAG.getNode(ISD::FSUB, DL, VT, ThreeHalves, NewEst);
+ AddToWorklist(NewEst.getNode());
- NewEst = DAG.getNode(ISD::FMUL, DL, VT, HalfArg, NewEst);
- AddToWorklist(NewEst.getNode());
+ Est = DAG.getNode(ISD::FMUL, DL, VT, Est, NewEst);
+ AddToWorklist(Est.getNode());
+ }
+ return Est;
+}
- NewEst = DAG.getNode(ISD::FSUB, DL, VT, FPThreeHalves, NewEst);
- AddToWorklist(NewEst.getNode());
+/// Newton iteration for a function: F(X) is X_{i+1} = X_i - F(X_i)/F'(X_i)
+/// For the reciprocal sqrt, we need to find the zero of the function:
+/// F(X) = 1/X^2 - A [which has a zero at X = 1/sqrt(A)]
+/// =>
+/// X_{i+1} = (-0.5 * X_i) * (A * X_i * X_i + (-3.0))
+SDValue DAGCombiner::BuildRsqrtNRTwoConst(SDValue Arg, SDValue Est,
+ unsigned Iterations) {
+ EVT VT = Arg.getValueType();
+ SDLoc DL(Arg);
+ SDValue MinusThree = DAG.getConstantFP(-3.0, VT);
+ SDValue MinusHalf = DAG.getConstantFP(-0.5, VT);
- Est = DAG.getNode(ISD::FMUL, DL, VT, Est, NewEst);
- AddToWorklist(Est.getNode());
- }
+ // Newton iterations: Est = -0.5 * Est * (-3.0 + Arg * Est * Est)
+ for (unsigned i = 0; i < Iterations; ++i) {
+ SDValue HalfEst = DAG.getNode(ISD::FMUL, DL, VT, Est, MinusHalf);
+ AddToWorklist(HalfEst.getNode());
+
+ Est = DAG.getNode(ISD::FMUL, DL, VT, Est, Est);
+ AddToWorklist(Est.getNode());
+
+ Est = DAG.getNode(ISD::FMUL, DL, VT, Est, Arg);
+ AddToWorklist(Est.getNode());
+
+ Est = DAG.getNode(ISD::FADD, DL, VT, Est, MinusThree);
+ AddToWorklist(Est.getNode());
+
+ Est = DAG.getNode(ISD::FMUL, DL, VT, Est, HalfEst);
+ AddToWorklist(Est.getNode());
+ }
+ return Est;
+}
+
+SDValue DAGCombiner::BuildRsqrtEstimate(SDValue Op) {
+ if (Level >= AfterLegalizeDAG)
+ return SDValue();
+
+ // Expose the DAG combiner to the target combiner implementations.
+ TargetLowering::DAGCombinerInfo DCI(DAG, Level, false, this);
+ unsigned Iterations = 0;
+ bool UseOneConstNR = false;
+ if (SDValue Est = TLI.getRsqrtEstimate(Op, DCI, Iterations, UseOneConstNR)) {
+ AddToWorklist(Est.getNode());
+ if (Iterations) {
+ Est = UseOneConstNR ?
+ BuildRsqrtNROneConst(Op, Est, Iterations) :
+ BuildRsqrtNRTwoConst(Op, Est, Iterations);
}
return Est;
}
return false;
}
- bool UseAA = CombinerGlobalAA.getNumOccurrences() > 0 ? CombinerGlobalAA :
- TLI.getTargetMachine().getSubtarget<TargetSubtargetInfo>().useAA();
+ bool UseAA = CombinerGlobalAA.getNumOccurrences() > 0
+ ? CombinerGlobalAA
+ : DAG.getSubtarget().useAA();
#ifndef NDEBUG
if (CombinerAAOnlyFunc.getNumOccurrences() &&
CombinerAAOnlyFunc != DAG.getMachineFunction().getName())
projects / oota-llvm.git / blobdiff