- Sep 15, 2007
-
-
Chris Lattner authored
llvm-svn: 41984
-
Chris Lattner authored
GCC optimizes away things like ptr < NULL to false. To "fix" this, have the CBE emit casts of pointers to intptr_t when doing relational pointer comparisons. llvm-svn: 41983
-
Chris Lattner authored
llvm-svn: 41979
-
Dan Gohman authored
llvm-svn: 41975
-
Chris Lattner authored
llvm-svn: 41972
-
Owen Anderson authored
llvm-svn: 41971
-
Owen Anderson authored
llvm-svn: 41968
-
Dale Johannesen authored
double from some of the many places in the optimizers it appears, and do something reasonable with x86 long double. Make APInt::dump() public, remove newline, use it to dump ConstantSDNode's. Allow APFloats in FoldingSet. Expand X86 backend handling of long doubles (conversions to/from int, mostly). llvm-svn: 41967
-
- Sep 14, 2007
-
-
Evan Cheng authored
llvm-svn: 41962
-
Dan Gohman authored
ambiguity. llvm-svn: 41960
-
Dan Gohman authored
llvm-svn: 41959
-
Dan Gohman authored
isRegister, isImmediate, and isMachineBasicBlock, which are equivalent, and more popular. llvm-svn: 41958
-
Dan Gohman authored
for consistency with many other transforms. llvm-svn: 41957
-
Dan Gohman authored
are strict about such things. llvm-svn: 41956
-
Rafael Espindola authored
llvm-svn: 41953
-
Chris Lattner authored
miscompilation of 188.ammp. Reject select and bitcast in ValueIsOnlyUsedLocallyOrStoredToOneGlobal because RewriteHeapSROALoadUser can't handle it. llvm-svn: 41950
-
Chris Lattner authored
llvm-svn: 41949
-
Evan Cheng authored
llvm-svn: 41947
-
Tanya Lattner authored
llvm-svn: 41941
-
Bill Wendling authored
(http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20070910/053370.html). It's causing SPASS to fail. llvm-svn: 41938
-
- Sep 13, 2007
-
-
Chris Lattner authored
a limited form of PHI nodes. This finally fixes PR1639, speeding 179.art up from 7.84s to 3.13s on PPC. llvm-svn: 41933
-
Chris Lattner authored
along the way of PR1639 llvm-svn: 41930
-
Chris Lattner authored
llvm-svn: 41929
-
Chris Lattner authored
bitcasts and phis. This is a step to fixing PR1639. llvm-svn: 41928
-
Chris Lattner authored
nodes. This is the first step of the fix for PR1639. llvm-svn: 41927
-
Chris Lattner authored
llvm-svn: 41925
-
Chris Lattner authored
llvm-svn: 41924
-
Chris Lattner authored
llvm-svn: 41923
-
Evan Cheng authored
llvm-svn: 41901
-
Evan Cheng authored
llvm-svn: 41900
-
Evan Cheng authored
llvm-svn: 41899
-
Evan Cheng authored
llvm-svn: 41898
-
Evan Cheng authored
1. [(set GR32:$dst, (add GR32:$src1, GR32:$src2)), (modify EFLAGS)] This indicates the source pattern expects the instruction would produce 2 values. The first is the result of the addition. The second is an implicit definition in register EFLAGS. 2. def : Pat<(parallel (addc GR32:$src1, GR32:$src2), (modify EFLAGS)), ()> Similar to #1 except this is used for def : Pat patterns. llvm-svn: 41897
-
Evan Cheng authored
llvm-svn: 41893
-
- Sep 12, 2007
-
-
Chris Lattner authored
llvm-svn: 41881
-
Chris Lattner authored
using it for live-in values etc. llvm-svn: 41879
-
Dale Johannesen authored
llvm-svn: 41876
-
Dale Johannesen authored
Next round of x87 long double stuff. Getting close now, basically works. llvm-svn: 41875
-
Dale Johannesen authored
(could break hash table in ConstantFP) llvm-svn: 41874
-
Bill Wendling authored
llvm-svn: 41873
-