RPCS3/llvm-mirror
1
0
Fork 0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-11-23 11:13:28 +01:00
Code Issues Projects Releases Wiki Activity

reuse the result of some expensive computations in getSignExtendExpr() and getZeroExtendExpr()

Browse Source 
this gives a speedup of > 80 in a debug build in the test case of PR12825 (php_sha512_crypt_r)

llvm-svn: 156849
This commit is contained in:
Nuno Lopes 2012-05-15 20:20:14 +00:00
parent 032842545f
commit 40c48e0924
1 changed files with 20 additions and 18 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

38
lib/Analysis/ScalarEvolution.cpp
View File

@ -974,12 +974,15 @@ const SCEV *ScalarEvolution::getZeroExtendExpr(const SCEV *Op,
Type *WideTy = IntegerType::get(getContext(), BitWidth * 2);
// Check whether Start+Step*MaxBECount has no unsigned overflow.
const SCEV *ZMul = getMulExpr(CastedMaxBECount, Step);
const SCEV *Add = getAddExpr(Start, ZMul);
const SCEV *ZAdd = getZeroExtendExpr(getAddExpr(Start, ZMul), WideTy);
const SCEV *WideStart = getZeroExtendExpr(Start, WideTy);
const SCEV *WideMaxBECount =
getZeroExtendExpr(CastedMaxBECount, WideTy);
const SCEV *OperandExtendedAdd =
getAddExpr(getZeroExtendExpr(Start, WideTy),
getMulExpr(getZeroExtendExpr(CastedMaxBECount, WideTy),
getAddExpr(WideStart,
getMulExpr(WideMaxBECount,
getZeroExtendExpr(Step, WideTy)));
if (getZeroExtendExpr(Add, WideTy) == OperandExtendedAdd) {
if (ZAdd == OperandExtendedAdd) {
// Cache knowledge of AR NUW, which is propagated to this AddRec.
const_cast<SCEVAddRecExpr *>(AR)->setNoWrapFlags(SCEV::FlagNUW);
// Return the expression with the addrec on the outside.
@ -989,13 +992,11 @@ const SCEV *ScalarEvolution::getZeroExtendExpr(const SCEV *Op,
}
// Similar to above, only this time treat the step value as signed.
// This covers loops that count down.
const SCEV *SMul = getMulExpr(CastedMaxBECount, Step);
Add = getAddExpr(Start, SMul);
OperandExtendedAdd =
getAddExpr(getZeroExtendExpr(Start, WideTy),
getMulExpr(getZeroExtendExpr(CastedMaxBECount, WideTy),
getAddExpr(WideStart,
getMulExpr(WideMaxBECount,
getSignExtendExpr(Step, WideTy)));
if (getZeroExtendExpr(Add, WideTy) == OperandExtendedAdd) {
if (ZAdd == OperandExtendedAdd) {
// Cache knowledge of AR NW, which is propagated to this AddRec.
// Negative step causes unsigned wrap, but it still can't self-wrap.
const_cast<SCEVAddRecExpr *>(AR)->setNoWrapFlags(SCEV::FlagNW);
@ -1239,12 +1240,15 @@ const SCEV *ScalarEvolution::getSignExtendExpr(const SCEV *Op,
Type *WideTy = IntegerType::get(getContext(), BitWidth * 2);
// Check whether Start+Step*MaxBECount has no signed overflow.
const SCEV *SMul = getMulExpr(CastedMaxBECount, Step);
const SCEV *Add = getAddExpr(Start, SMul);
const SCEV *SAdd = getSignExtendExpr(getAddExpr(Start, SMul), WideTy);
const SCEV *WideStart = getSignExtendExpr(Start, WideTy);
const SCEV *WideMaxBECount =
getZeroExtendExpr(CastedMaxBECount, WideTy);
const SCEV *OperandExtendedAdd =
getAddExpr(getSignExtendExpr(Start, WideTy),
getMulExpr(getZeroExtendExpr(CastedMaxBECount, WideTy),
getAddExpr(WideStart,
getMulExpr(WideMaxBECount,
getSignExtendExpr(Step, WideTy)));
if (getSignExtendExpr(Add, WideTy) == OperandExtendedAdd) {
if (SAdd == OperandExtendedAdd) {
// Cache knowledge of AR NSW, which is propagated to this AddRec.
const_cast<SCEVAddRecExpr *>(AR)->setNoWrapFlags(SCEV::FlagNSW);
// Return the expression with the addrec on the outside.
@ -1254,13 +1258,11 @@ const SCEV *ScalarEvolution::getSignExtendExpr(const SCEV *Op,
}
// Similar to above, only this time treat the step value as unsigned.
// This covers loops that count up with an unsigned step.
const SCEV *UMul = getMulExpr(CastedMaxBECount, Step);
Add = getAddExpr(Start, UMul);
OperandExtendedAdd =
getAddExpr(getSignExtendExpr(Start, WideTy),
getMulExpr(getZeroExtendExpr(CastedMaxBECount, WideTy),
getAddExpr(WideStart,
getMulExpr(WideMaxBECount,
getZeroExtendExpr(Step, WideTy)));
if (getSignExtendExpr(Add, WideTy) == OperandExtendedAdd) {
if (SAdd == OperandExtendedAdd) {
// Cache knowledge of AR NSW, which is propagated to this AddRec.
const_cast<SCEVAddRecExpr *>(AR)->setNoWrapFlags(SCEV::FlagNSW);
// Return the expression with the addrec on the outside.