- Mar 23, 2009
-
-
Evan Cheng authored
Model inline asm constraint which ties an input to an output register as machine operand TIED_TO constraint. This eliminated the need to pre-allocate registers for these. This also allows register allocator can eliminate the unneeded copies. llvm-svn: 67512
-
Evan Cheng authored
Do not fold away subreg_to_reg if the source register has a sub-register index. That means the source register is taking a sub-register of a larger register. e.g. On x86 %RAX<def> = ... %RAX<def> = SUBREG_TO_REG 0, %EAX:3<kill>, 3 The first def is defining RAX, not EAX so the top bits were not zero-extended. llvm-svn: 67511
-
Chris Lattner authored
llvm-svn: 67510
-
Chris Lattner authored
were when we came around, not to their default handler. This should fix PR3848 llvm-svn: 67509
-
Chris Lattner authored
llvm-svn: 67508
-
Dan Gohman authored
llvm-svn: 67507
-
Chris Lattner authored
llvm-svn: 67505
-
Chris Lattner authored
Do not recommend llvm::OStream anymore. Use raw_ostream or MemoryBuffer. llvm-svn: 67504
-
Dan Gohman authored
i8 return values. llvm-svn: 67502
-
Evan Cheng authored
llvm-svn: 67500
-
- Mar 22, 2009
-
-
Chris Lattner authored
llvm-svn: 67473
-
Duncan Sands authored
work. llvm-svn: 67471
-
Bill Wendling authored
llvm-svn: 67469
-
Bill Wendling authored
llvm-svn: 67468
-
Chris Lattner authored
llvm-svn: 67458
-
- Mar 21, 2009
-
-
Duncan Sands authored
llvm-svn: 67454
-
Evan Cheng authored
llvm-svn: 67451
-
Chris Lattner authored
unneeded bitcast is requested. This is common for frontends who just unconditionally cast even if the target is often the right type already. THis prevents going into getFoldedCast which switches on the opcode and does a bunch of other stuff before doing the same opzn. llvm-svn: 67435
-
Chris Lattner authored
tidy some df iteration stuff, patch by John Mosby! llvm-svn: 67428
-
Bruno Cardoso Lopes authored
Handle odd registers allocation in FGR32. llvm-svn: 67422
-
Bob Wilson authored
llvm-svn: 67416
-
- Mar 20, 2009
-
-
Bob Wilson authored
llvm-svn: 67412
-
Chris Lattner authored
shifts together. This fixes PR3851. llvm-svn: 67411
-
Duncan Sands authored
linkage: the value may be replaced with something different at link time. (Frontends that want to allow values to be loaded out of weak constants can give their constants weak_odr linkage). llvm-svn: 67407
-
Dan Gohman authored
llvm-svn: 67400
-
Evan Cheng authored
For inline asm output operand that matches an input. Encode the input operand index in the high bits. llvm-svn: 67387
-
Sanjiv Gupta authored
llvm-svn: 67373
-
Sanjiv Gupta authored
llvm-svn: 67372
-
Sanjiv Gupta authored
Removed unncessary code. No functionality change. llvm-svn: 67371
-
Sanjiv Gupta authored
llvm-svn: 67370
-
Nick Lewycky authored
llvm-svn: 67368
-
Chris Lattner authored
llvm-svn: 67364
-
Mon P Wang authored
for those architectures that support the instruction. llvm-svn: 67363
-
Dan Gohman authored
(return or unreachable) is a kill. llvm-svn: 67357
-
Evan Cheng authored
llvm-svn: 67349
-
Sebastian Redl authored
- Make type declarations match the struct/class keyword of the definition. - Move AddSignalHandler into the namespace where it belongs. - Correctly call functions from template base. - Some other small changes. With this patch, LLVM and Clang should build properly and with far less noise under VS2008. llvm-svn: 67347
-
- Mar 19, 2009
-
-
Evan Cheng authored
llvm-svn: 67335
-
Evan Cheng authored
More makefile changes to allow dejagnu tests to pass when system tools default to a different target from the llvm configuration (e.g. 64-bit gcc and 32-bit llvm). llvm-svn: 67334
-
Misha Brukman authored
llvm-svn: 67326
-
Dan Gohman authored
llvm-svn: 67319
-