Author: evancheng
Date: Wed Dec 12 18:43:27 2007
New Revision: 44970
URL: http://llvm.org/viewvc/llvm-project?rev=44970&view=rev
Log:
Fold some and + shift in x86 addressing mode.
Added:
llvm/trunk/test/CodeGen/X86/fold-and-shift.ll
Modified:
llvm/trunk/lib/Target/X86/X86ISelDAGToDAG.cpp
llvm/trunk/lib/Target/X86/X86InstrInfo.td
Modified: llvm/trunk/lib/Target/X86/X86ISelDAGToDAG.cpp
URL:
http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/X86/X86ISelDAGToDAG.cpp?rev=44970&r1=44969&r2=44970&view=diff
==============================================================================
--- llvm/trunk/lib/Target/X86/X86ISelDAGToDAG.cpp (original)
+++ llvm/trunk/lib/Target/X86/X86ISelDAGToDAG.cpp Wed Dec 12 18:43:27 2007
@@ -582,7 +582,7 @@
}
int id = N.Val->getNodeId();
- bool Available = isSelected(id);
+ bool AlreadySelected = isSelected(id); // Already selected, not yet replaced.
switch (N.getOpcode()) {
default: break;
@@ -605,7 +605,7 @@
// If value is available in a register both base and index components have
// been picked, we can't fit the result available in the register in the
// addressing mode. Duplicate GlobalAddress or ConstantPool as
displacement.
- if (!Available || (AM.Base.Reg.Val && AM.IndexReg.Val)) {
+ if (!AlreadySelected || (AM.Base.Reg.Val && AM.IndexReg.Val)) {
bool isStatic = TM.getRelocationModel() == Reloc::Static;
SDOperand N0 = N.getOperand(0);
// Mac OS X X86-64 lower 4G address is not available.
@@ -653,7 +653,7 @@
break;
case ISD::SHL:
- if (Available || AM.IndexReg.Val != 0 || AM.Scale != 1)
+ if (AlreadySelected || AM.IndexReg.Val != 0 || AM.Scale != 1)
break;
if (ConstantSDNode *CN = dyn_cast<ConstantSDNode>(N.Val->getOperand(1))) {
@@ -690,7 +690,7 @@
// FALL THROUGH
case ISD::MUL:
// X*[3,5,9] -> X+X*[2,4,8]
- if (!Available &&
+ if (!AlreadySelected &&
AM.BaseType == X86ISelAddressMode::RegBase &&
AM.Base.Reg.Val == 0 &&
AM.IndexReg.Val == 0) {
@@ -725,7 +725,7 @@
break;
case ISD::ADD:
- if (!Available) {
+ if (!AlreadySelected) {
X86ISelAddressMode Backup = AM;
if (!MatchAddress(N.Val->getOperand(0), AM, false, Depth+1) &&
!MatchAddress(N.Val->getOperand(1), AM, false, Depth+1))
@@ -740,7 +740,7 @@
case ISD::OR:
// Handle "X | C" as "X + C" iff X is known to have C bits clear.
- if (Available) break;
+ if (AlreadySelected) break;
if (ConstantSDNode *CN = dyn_cast<ConstantSDNode>(N.getOperand(1))) {
X86ISelAddressMode Backup = AM;
@@ -758,6 +758,44 @@
AM = Backup;
}
break;
+
+ case ISD::AND: {
+ // Handle "(x << C1) & C2" as "(X & (C2>>C1)) << C1" if safe and if this
+ // allows us to fold the shift into this addressing mode.
+ if (AlreadySelected) break;
+ SDOperand Shift = N.getOperand(0);
+ if (Shift.getOpcode() != ISD::SHL) break;
+
+ // Scale must not be used already.
+ if (AM.IndexReg.Val != 0 || AM.Scale != 1) break;
+
+ ConstantSDNode *C2 = dyn_cast<ConstantSDNode>(N.getOperand(1));
+ ConstantSDNode *C1 = dyn_cast<ConstantSDNode>(Shift.getOperand(1));
+ if (!C1 || !C2) break;
+
+ // Not likely to be profitable if either the AND or SHIFT node has more
+ // than one use (unless all uses are for address computation). Besides,
+ // isel mechanism requires their node ids to be reused.
+ if (!N.hasOneUse() || !Shift.hasOneUse())
+ break;
+
+ // Verify that the shift amount is something we can fold.
+ unsigned ShiftCst = C1->getValue();
+ if (ShiftCst != 1 && ShiftCst != 2 && ShiftCst != 3)
+ break;
+
+ // Get the new AND mask, this folds to a constant.
+ SDOperand NewANDMask = CurDAG->getNode(ISD::SRL, N.getValueType(),
+ SDOperand(C2, 0), SDOperand(C1, 0));
+ SDOperand NewAND = CurDAG->getNode(ISD::AND, N.getValueType(),
+ Shift.getOperand(0), NewANDMask);
+ NewANDMask.Val->setNodeId(Shift.Val->getNodeId());
+ NewAND.Val->setNodeId(N.Val->getNodeId());
+
+ AM.Scale = 1 << ShiftCst;
+ AM.IndexReg = NewAND;
+ return false;
+ }
}
return MatchAddressBase(N, AM, isRoot, Depth);
Modified: llvm/trunk/lib/Target/X86/X86InstrInfo.td
URL:
http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/X86/X86InstrInfo.td?rev=44970&r1=44969&r2=44970&view=diff
==============================================================================
--- llvm/trunk/lib/Target/X86/X86InstrInfo.td (original)
+++ llvm/trunk/lib/Target/X86/X86InstrInfo.td Wed Dec 12 18:43:27 2007
@@ -2613,6 +2613,10 @@
def : Pat<(i32 (anyext (loadi8 addr:$src))), (MOVZX32rm8 addr:$src)>;
def : Pat<(i32 (anyext (loadi16 addr:$src))), (MOVZX32rm16 addr:$src)>;
+// (and (i32 load), 255) -> (zextload i8)
+def : Pat<(i32 (and (loadi32 addr:$src), (i32 255))), (MOVZX32rm8 addr:$src)>;
+def : Pat<(i32 (and (loadi32 addr:$src), (i32 65535))),(MOVZX32rm16
addr:$src)>;
+
//===----------------------------------------------------------------------===//
// Some peepholes
//===----------------------------------------------------------------------===//
Added: llvm/trunk/test/CodeGen/X86/fold-and-shift.ll
URL:
http://llvm.org/viewvc/llvm-project/llvm/trunk/test/CodeGen/X86/fold-and-shift.ll?rev=44970&view=auto
==============================================================================
--- llvm/trunk/test/CodeGen/X86/fold-and-shift.ll (added)
+++ llvm/trunk/test/CodeGen/X86/fold-and-shift.ll Wed Dec 12 18:43:27 2007
@@ -0,0 +1,21 @@
+; RUN: llvm-as < %s | llc -march=x86 | not grep and
+
+define i32 @t1(i8* %X, i32 %i) {
+entry:
+ %tmp2 = shl i32 %i, 2 ; <i32> [#uses=1]
+ %tmp4 = and i32 %tmp2, 1020 ; <i32> [#uses=1]
+ %tmp7 = getelementptr i8* %X, i32 %tmp4 ; <i8*> [#uses=1]
+ %tmp78 = bitcast i8* %tmp7 to i32* ; <i32*> [#uses=1]
+ %tmp9 = load i32* %tmp78, align 4 ; <i32> [#uses=1]
+ ret i32 %tmp9
+}
+
+define i32 @t2(i16* %X, i32 %i) {
+entry:
+ %tmp2 = shl i32 %i, 1 ; <i32> [#uses=1]
+ %tmp4 = and i32 %tmp2, 131070 ; <i32> [#uses=1]
+ %tmp7 = getelementptr i16* %X, i32 %tmp4 ; <i16*>
[#uses=1]
+ %tmp78 = bitcast i16* %tmp7 to i32* ; <i32*> [#uses=1]
+ %tmp9 = load i32* %tmp78, align 4 ; <i32> [#uses=1]
+ ret i32 %tmp9
+}
_______________________________________________
llvm-commits mailing list
llvm-commits@cs.uiuc.edu
http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
