- Aug 17, 2010
-
-
Evan Cheng authored
Move the decision logic whether it's a good idea to split a critical edge to clients. Also fixed an erroneous check. An edge is only a back edge when the from and to blocks are in the same loop. llvm-svn: 111256
-
Dan Gohman authored
loop, making the resulting loop significantly less ugly. Also, zap its trivial PHI nodes, since it's easy. llvm-svn: 111255
-
Jim Grosbach authored
llvm-svn: 111254
-
Benjamin Kramer authored
llvm-svn: 111252
-
Bob Wilson authored
printing "lsl #0". This fixes the remaining parts of pr7792. Make corresponding changes for encoding/decoding these instructions. llvm-svn: 111251
-
Evan Cheng authored
llvm-svn: 111250
-
Chris Lattner authored
and the formatting being thrown off. I admit to not knowing what is going on here. llvm-svn: 111249
-
Dan Gohman authored
what it does manually. llvm-svn: 111248
-
Benjamin Kramer authored
llvm-svn: 111247
-
Chris Lattner authored
llvm-svn: 111241
-
Nick Lewycky authored
against it. This affects Windows. llvm-svn: 111240
-
Benjamin Kramer authored
llvm-svn: 111232
-
Chris Lattner authored
cygwin when built with ENABLE_SHARED=1. Patch by NAKAMURA Takumi! llvm-svn: 111231
-
Benjamin Kramer authored
whole integer to the stream instead of doing byte-sized writes. llvm-svn: 111230
-
Benjamin Kramer authored
llvm-svn: 111228
-
Bob Wilson authored
llvm-svn: 111226
-
Evan Cheng authored
Add an option to disable codegen prepare critical edge splitting. In theory, PHI elimination is already doing all (most?) of the splitting needed. But machine-licm and machine-sink seem to miss some important optimizations when splitting is disabled. llvm-svn: 111224
-
Eric Christopher authored
llvm-svn: 111223
-
Eric Christopher authored
we're adding predicates and optional defs to the MachineInstrs. llvm-svn: 111222
-
Evan Cheng authored
PHI elimination should not break back edge. It can cause some significant code placement issues. rdar://8263994 good: LBB0_2: mov r2, r0 . . . mov r1, r2 bne LBB0_2 bad: LBB0_2: mov r2, r0 . . . @ BB#3: mov r1, r2 b LBB0_2 llvm-svn: 111221
-
Eric Christopher authored
and optional def operands are handled in the backend support. llvm-svn: 111220
-
Eric Christopher authored
llvm-svn: 111219
-
Benjamin Kramer authored
llvm-svn: 111214
-
Benjamin Kramer authored
llvm-svn: 111213
-
Bob Wilson authored
llvm-svn: 111212
-
Owen Anderson authored
llvm-svn: 111210
-
Bob Wilson authored
llvm-svn: 111208
-
Jim Grosbach authored
llvm-svn: 111206
-
Benjamin Kramer authored
llvm-svn: 111201
-
Dan Gohman authored
llvm-svn: 111200
-
Dan Gohman authored
a Pass abstraction, since that's the level it's actually used at. Rename Pass' dumpPassStructure to dumpPass. This eliminates an awkward use of getAsPass() to convert a PMDataManager* into a Pass* just to permit a dumpPassStructure call. llvm-svn: 111199
-
Jim Grosbach authored
mapping. Have the local block track its alignment requirement, and then apply that when the block itself is allocated. Previously, offsets could get adjusted in PEI to be different, relative to one another, than the block allocation thought they would be, which defeats the point of doing the allocation this way. Continuing rdar://8277890 llvm-svn: 111197
-
Bob Wilson authored
that the high halfword is zero. The shift need not be exactly 16 bits. llvm-svn: 111196
-
Bob Wilson authored
llvm-svn: 111195
-
Dan Gohman authored
make its constructor protected. llvm-svn: 111193
-
- Aug 16, 2010
-
-
Dan Gohman authored
llvm-svn: 111192
-
Dan Gohman authored
PMTopLevelManager's constructor take a PMDataManager *, which already provides the needed abstraction support. llvm-svn: 111189
-
Eli Friedman authored
llvm-svn: 111185
-
Eli Friedman authored
llvm-svn: 111183
-
Eli Friedman authored
llvm-svn: 111182
-