Commits

Show all
Author Commit Message Labels Comments Date
Ted Kremenek
Add sanity check in Clang TableGen backend to check if 'Component' is a string.
Tags
2 tags
Dale Johannesen
Fix internal representation of fp80 to be the same as a normal i80 {low64, high16} rather than its own {high64, low16}. A depressing number of places know about this; I think I got them all. Bitcode readers and writers convert back to the old form to avoid breaking compatibility.
jdm
README.txt: test commit w/blank line appended
Dan Gohman
When unfolding a load during scheduling, the new operator node has a data dependency on the load node, so it really needs a data-dependence edge to the load node, even if the load previously existed. And add a few comments.
Evan Cheng
Update test for pr3864.
Evan Cheng
Fix PR3391 and PR3864. Reg allocator infinite looping.
Dan Gohman
Don't set SUnit::hasPhysRegDefs to true unless the defs are actually have uses, which reflects the way it's used.
Dan Gohman
Fix canClobberPhysRegDefs to check all SDNodes grouped together in an SUnit, instead of just the first one. This fix is needed by some upcoming scheduler changes.
Tags
checker-0.179
Dan Gohman
Add a new bit to SUnit to record whether a node has implicit physreg defs, regardless of whether they are actually used.
Dan Gohman
Now that errs() is properly non-buffered, there's no need to explicitly flush it.
Dan Gohman
Clarify a comment.
Dan Gohman
LoopVR is not CFGOnly.
Dan Gohman
Enhance LiveValues to work on PHI operands.
Dan Gohman
Make getOperandNumForIncomingValue and getOperandNumForIncomingBlock static member functions, and add getIncomingValueNumForOperand and getIncomingBlockNumForOperand, which are the respective inverses.
Dan Gohman
Correct some comments. Operand numbers start at 0.
Tags
checker/checker-0.179
Evan Cheng
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.
Evan Cheng
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.
Chris Lattner
fix a bug Alexei Svitkine pointed out.
Chris Lattner
When we restore signal handlers, restore them back to what they were when we came around, not to their default handler. This should fix PR3848
Chris Lattner
factorize signal registration, part of PR3848.
Dan Gohman
Fix a grammaro in a comment that Bill noticed.
Chris Lattner
tt'ify.
Chris Lattner
VC++ 6.0 is not future work :) Do not recommend llvm::OStream anymore. Use raw_ostream or MemoryBuffer.
Dan Gohman
Add comments explaining why there's only one register for i8 return values.
Evan Cheng
Missed this as part of r67451.
Chris Lattner
Fix PR3860 by correcting a predicate.
Duncan Sands
Add some explanations of how apint loads and stores work.
Bill Wendling
Really should pass -dylib to the linker...
Bill Wendling
A dylib should be built as a dylib and not a bundle.
Chris Lattner
add method to access a template argument.
Tags
checker-0.178
  1. Prev
  2. Next