- Mar 17, 2007
-
-
Evan Cheng authored
llvm-svn: 35133
-
Evan Cheng authored
in that BB. llvm-svn: 35132
-
Devang Patel authored
llvm-svn: 35129
-
- Mar 16, 2007
-
-
Lauro Ramos Venancio authored
Fix MultiSource/Applications/aha test. llvm-svn: 35128
-
Evan Cheng authored
llvm-svn: 35127
-
Evan Cheng authored
llvm-svn: 35126
-
Bill Wendling authored
llvm-svn: 35125
-
Evan Cheng authored
computation used as GEP indexes and if the expression can be folded into target addressing mode of GEP load / store use types. llvm-svn: 35123
-
Evan Cheng authored
llvm-svn: 35122
-
Evan Cheng authored
folded into target addressing mode for the given type. llvm-svn: 35121
-
Nick Lewycky authored
Add new micro-optimizations. Add icmp predicate snuggling. Given %x ULT 4, "icmp ugt %x, 2" becomes "icmp eq %x, 3". This doesn't apply in any non-trivial cases yet due to missing support for NE values in ValueRanges. llvm-svn: 35119
-
- Mar 15, 2007
-
-
Bill Wendling authored
llvm-svn: 35118
-
Evan Cheng authored
llvm-svn: 35117
-
- Mar 14, 2007
-
-
Evan Cheng authored
it as a late BURR scheduling tie-breaker. Intuitively, it's good to push down instructions whose results are liveout so their long live ranges won't conflict with other values which are needed inside the BB. Further prioritize liveout instructions by the number of operands which are calculated within the BB. llvm-svn: 35109
-
Evan Cheng authored
llvm-svn: 35108
-
Evan Cheng authored
llvm-svn: 35107
-
Evan Cheng authored
llvm-svn: 35105
-
Evan Cheng authored
llvm-svn: 35104
-
Evan Cheng authored
llvm-svn: 35097
-
Evan Cheng authored
llvm-svn: 35096
-
Evan Cheng authored
llvm-svn: 35095
-
Zhou Sheng authored
llvm-svn: 35094
-
Zhou Sheng authored
llvm-svn: 35093
-
Evan Cheng authored
e.g. t1 = op t2, c1 t3 = op t4, c2 and the following instructions are both ready. t2 = op c3 t4 = op c4 Then schedule t2 = op first. i.e. t4 = op c4 t2 = op c3 t1 = op t2, c1 t3 = op t4, c2 This creates more short live intervals which work better with the register allocator. llvm-svn: 35089
-
- Mar 13, 2007
-
-
Evan Cheng authored
llvm-svn: 35088
-
Evan Cheng authored
llvm-svn: 35087
-
Evan Cheng authored
llvm-svn: 35086
-
Nicolas Geoffray authored
llvm-svn: 35083
-
Chris Lattner authored
Fix PR1244 and Transforms/InstCombine/2007-03-13-CompareMerge.ll llvm-svn: 35082
-
Zhou Sheng authored
"APInt::getAllOnesValue(ShiftAmt).zextOrCopy(BitWidth)", to handle ShiftAmt == BitWidth situation, use zextOrCopy() instead of zext(). llvm-svn: 35080
-
Zhou Sheng authored
1. Ensure VTy, KnownOne and KnownZero have same bitwidth. 2. Make code more efficient. llvm-svn: 35078
-
Evan Cheng authored
llvm-svn: 35077
-
Evan Cheng authored
More flexible TargetLowering LSR hooks for testing whether an immediate is a legal target address immediate or scale. llvm-svn: 35076
-
Evan Cheng authored
llvm-svn: 35075
-
Evan Cheng authored
More flexible TargetLowering LSR hooks for testing whether an immediate is a legal target address immediate or scale. llvm-svn: 35074
-
Evan Cheng authored
a legal target address immediate or scale. llvm-svn: 35073
-
Evan Cheng authored
llvm-svn: 35072
-
- Mar 12, 2007
-
-
Evan Cheng authored
llvm-svn: 35070
-
Reid Spencer authored
llvm-svn: 35068
-
Jeff Cohen authored
llvm-svn: 35067
-