1
0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2025-01-31 12:41:49 +01:00

[PowerPC] Recommit r340016 after fixing the reported issue

The internal benchmark failure reported by Google was due to a missing
check for the result type for the sign-extend and shift DAG. This commit
adds the check and re-commits the patch.

llvm-svn: 340734
This commit is contained in:
Nemanja Ivanovic 2018-08-27 11:20:27 +00:00
parent 6a6f9554e3
commit ee319d2073
5 changed files with 55 additions and 3 deletions

View File

@ -1351,6 +1351,7 @@ const char *PPCTargetLowering::getTargetNodeName(unsigned Opcode) const {
case PPCISD::QBFLT: return "PPCISD::QBFLT";
case PPCISD::QVLFSb: return "PPCISD::QVLFSb";
case PPCISD::BUILD_FP128: return "PPCISD::BUILD_FP128";
case PPCISD::EXTSWSLI: return "PPCISD::EXTSWSLI";
}
return nullptr;
}
@ -14131,7 +14132,30 @@ SDValue PPCTargetLowering::combineSHL(SDNode *N, DAGCombinerInfo &DCI) const {
if (auto Value = stripModuloOnShift(*this, N, DCI.DAG))
return Value;
return SDValue();
SDValue N0 = N->getOperand(0);
ConstantSDNode *CN1 = dyn_cast<ConstantSDNode>(N->getOperand(1));
if (!Subtarget.isISA3_0() ||
N0.getOpcode() != ISD::SIGN_EXTEND ||
N0.getOperand(0).getValueType() != MVT::i32 ||
CN1 == nullptr || N->getValueType(0) != MVT::i64)
return SDValue();
// We can't save an operation here if the value is already extended, and
// the existing shift is easier to combine.
SDValue ExtsSrc = N0.getOperand(0);
if (ExtsSrc.getOpcode() == ISD::TRUNCATE &&
ExtsSrc.getOperand(0).getOpcode() == ISD::AssertSext)
return SDValue();
SDLoc DL(N0);
SDValue ShiftBy = SDValue(CN1, 0);
// We want the shift amount to be i32 on the extswli, but the shift could
// have an i64.
if (ShiftBy.getValueType() == MVT::i64)
ShiftBy = DCI.DAG.getConstant(CN1->getZExtValue(), DL, MVT::i32);
return DCI.DAG.getNode(PPCISD::EXTSWSLI, DL, MVT::i64, N0->getOperand(0),
ShiftBy);
}
SDValue PPCTargetLowering::combineSRA(SDNode *N, DAGCombinerInfo &DCI) const {

View File

@ -149,6 +149,10 @@ namespace llvm {
/// For vector types, only the last n bits are used. See vsld.
SRL, SRA, SHL,
/// EXTSWSLI = The PPC extswsli instruction, which does an extend-sign
/// word and shift left immediate.
EXTSWSLI,
/// The combination of sra[wd]i and addze used to implemented signed
/// integer division by a power of 2. The first operand is the dividend,
/// and the second is the constant shift amount (representing the

View File

@ -717,9 +717,10 @@ defm SRADI : XSForm_1rc<31, 413, (outs g8rc:$rA), (ins g8rc:$rS, u6imm:$SH),
"sradi", "$rA, $rS, $SH", IIC_IntRotateDI,
[(set i64:$rA, (sra i64:$rS, (i32 imm:$SH)))]>, isPPC64;
defm EXTSWSLI : XSForm_1r<31, 445, (outs g8rc:$rA), (ins g8rc:$rS, u6imm:$SH),
defm EXTSWSLI : XSForm_1r<31, 445, (outs g8rc:$rA), (ins gprc:$rS, u6imm:$SH),
"extswsli", "$rA, $rS, $SH", IIC_IntRotateDI,
[]>, isPPC64;
[(set i64:$rA, (PPCextswsli i32:$rS, (i32 imm:$SH)))]>,
isPPC64, Requires<[IsISA3_0]>;
// For fast-isel:
let isCodeGenOnly = 1, Defs = [CARRY] in

View File

@ -114,6 +114,10 @@ def SDT_PPCqvlfsb : SDTypeProfile<1, 1, [
SDTCisVec<0>, SDTCisPtrTy<1>
]>;
def SDT_PPCextswsli : SDTypeProfile<1, 2, [ // extswsli
SDTCisInt<0>, SDTCisInt<1>, SDTCisOpSmallerThanOp<1, 0>, SDTCisInt<2>
]>;
//===----------------------------------------------------------------------===//
// PowerPC specific DAG Nodes.
//
@ -218,6 +222,8 @@ def PPCsrl : SDNode<"PPCISD::SRL" , SDTIntShiftOp>;
def PPCsra : SDNode<"PPCISD::SRA" , SDTIntShiftOp>;
def PPCshl : SDNode<"PPCISD::SHL" , SDTIntShiftOp>;
def PPCextswsli : SDNode<"PPCISD::EXTSWSLI" , SDT_PPCextswsli>;
// Move 2 i64 values into a VSX register
def PPCbuild_fp128: SDNode<"PPCISD::BUILD_FP128",
SDTypeProfile<1, 2,

View File

@ -0,0 +1,17 @@
; RUN: llc -verify-machineinstrs -mtriple=powerpc64le-unknown-linux-gnu \
; RUN: -mcpu=pwr9 -ppc-asm-full-reg-names < %s | FileCheck %s
@z = external local_unnamed_addr global i32*, align 8
; Function Attrs: norecurse nounwind readonly
define signext i32 @_Z2tcii(i32 signext %x, i32 signext %y) local_unnamed_addr #0 {
entry:
%0 = load i32*, i32** @z, align 8
%add = add nsw i32 %y, %x
%idxprom = sext i32 %add to i64
%arrayidx = getelementptr inbounds i32, i32* %0, i64 %idxprom
%1 = load i32, i32* %arrayidx, align 4
ret i32 %1
; CHECK-LABEL: @_Z2tcii
; CHECK: extswsli {{r[0-9]+}}, {{r[0-9]+}}, 2
}