[llvm-commits] CVS: llvm/lib/Target/TargetLowering.cpp
Chris Lattner
lattner at cs.uiuc.edu
Sat May 6 16:40:42 PDT 2006
Changes in directory llvm/lib/Target:
TargetLowering.cpp updated: 1.58 -> 1.59
---
Log message:
Add some more sign propagation cases
---
Diffs of the changes: (+77 -10)
TargetLowering.cpp | 87 ++++++++++++++++++++++++++++++++++++++++++++++-------
1 files changed, 77 insertions(+), 10 deletions(-)
Index: llvm/lib/Target/TargetLowering.cpp
diff -u llvm/lib/Target/TargetLowering.cpp:1.58 llvm/lib/Target/TargetLowering.cpp:1.59
--- llvm/lib/Target/TargetLowering.cpp:1.58 Sat May 6 17:39:59 2006
+++ llvm/lib/Target/TargetLowering.cpp Sat May 6 18:40:29 2006
@@ -1068,16 +1068,6 @@
return Tmp - C->getValue();
}
break;
- case ISD::ADD:
- case ISD::SUB:
- // Add and sub can have at most one carry bit. Thus we know that the output
- // is, at worst, one more bit than the inputs.
- Tmp = ComputeNumSignBits(Op.getOperand(0), Depth+1);
- if (Tmp == 1) return 1; // Early out.
- Tmp2 = ComputeNumSignBits(Op.getOperand(1), Depth+1);
- if (Tmp2 == 1) return 1;
- return std::min(Tmp, Tmp2)-1;
-
case ISD::AND:
case ISD::OR:
case ISD::XOR: // NOT is handled here.
@@ -1098,6 +1088,83 @@
if (getSetCCResultContents() == ZeroOrNegativeOneSetCCResult)
return VTBits;
break;
+ case ISD::ROTL:
+ case ISD::ROTR:
+ if (ConstantSDNode *C = dyn_cast<ConstantSDNode>(Op.getOperand(1))) {
+ unsigned RotAmt = C->getValue() & (VTBits-1);
+
+ // Handle rotate right by N like a rotate left by 32-N.
+ if (Op.getOpcode() == ISD::ROTR)
+ RotAmt = (VTBits-RotAmt) & (VTBits-1);
+
+ // If we aren't rotating out all of the known-in sign bits, return the
+ // number that are left. This handles rotl(sext(x), 1) for example.
+ Tmp = ComputeNumSignBits(Op.getOperand(0), Depth+1);
+ if (Tmp > RotAmt+1) return Tmp-RotAmt;
+ }
+ break;
+ case ISD::ADD:
+ // Add can have at most one carry bit. Thus we know that the output
+ // is, at worst, one more bit than the inputs.
+ Tmp = ComputeNumSignBits(Op.getOperand(0), Depth+1);
+ if (Tmp == 1) return 1; // Early out.
+
+ // Special case decrementing a value (ADD X, -1):
+ if (ConstantSDNode *CRHS = dyn_cast<ConstantSDNode>(Op.getOperand(0)))
+ if (CRHS->isAllOnesValue()) {
+ uint64_t KnownZero, KnownOne;
+ uint64_t Mask = MVT::getIntVTBitMask(VT);
+ ComputeMaskedBits(Op.getOperand(0), Mask, KnownZero, KnownOne, Depth+1);
+
+ // If the input is known to be 0 or 1, the output is 0/-1, which is all
+ // sign bits set.
+ if ((KnownZero|1) == Mask)
+ return VTBits;
+
+ // If we are subtracting one from a positive number, there is no carry
+ // out of the result.
+ if (KnownZero & MVT::getIntVTSignBit(VT))
+ return Tmp;
+ }
+
+ Tmp2 = ComputeNumSignBits(Op.getOperand(1), Depth+1);
+ if (Tmp2 == 1) return 1;
+ return std::min(Tmp, Tmp2)-1;
+ break;
+
+ case ISD::SUB:
+ Tmp2 = ComputeNumSignBits(Op.getOperand(1), Depth+1);
+ if (Tmp2 == 1) return 1;
+
+ // Handle NEG.
+ if (ConstantSDNode *CLHS = dyn_cast<ConstantSDNode>(Op.getOperand(0)))
+ if (CLHS->getValue() == 0) {
+ uint64_t KnownZero, KnownOne;
+ uint64_t Mask = MVT::getIntVTBitMask(VT);
+ ComputeMaskedBits(Op.getOperand(1), Mask, KnownZero, KnownOne, Depth+1);
+ // If the input is known to be 0 or 1, the output is 0/-1, which is all
+ // sign bits set.
+ if ((KnownZero|1) == Mask)
+ return VTBits;
+
+ // If the input is known to be positive (the sign bit is known clear),
+ // the output of the NEG has the same number of sign bits as the input.
+ if (KnownZero & MVT::getIntVTSignBit(VT))
+ return Tmp2;
+
+ // Otherwise, we treat this like a SUB.
+ }
+
+ // Sub can have at most one carry bit. Thus we know that the output
+ // is, at worst, one more bit than the inputs.
+ Tmp = ComputeNumSignBits(Op.getOperand(0), Depth+1);
+ if (Tmp == 1) return 1; // Early out.
+ return std::min(Tmp, Tmp2)-1;
+ break;
+ case ISD::TRUNCATE:
+ // FIXME: it's tricky to do anything useful for this, but it is an important
+ // case for targets like X86.
+ break;
}
// Allow the target to implement this method for its nodes.
More information about the llvm-commits
mailing list