1. Dan Villiom Podlaski Christiansen
  2. LLVM

Commits

Show all
Author Commit Message Date Builds
Tanya Lattner (Tanya Brethour)
Remove llvm-upgrade and update tests.
Tags
2 tags
Chris Lattner
dead pass
Chris Lattner
remove the LowerSelect pass. The last client was the old Sparc backend, which is long dead by now.
Chris Lattner
remove the lower packed pass. It can never work and even the parts that could work don't work fully. This fixes PR1705. Oh yeah, we don't have packed types anymore either ;-)
Owen Anderson
Refactor this method a bit, and correct a test that was completely wrong but happened to work out anyways. :-)
Chris Lattner
isa+cast -> dyncast.
Chris Lattner
simplify this code again, try 2 :)
Owen Anderson
Fix a comment.
Owen Anderson
In addition to arguments passed to it, memcpy (and all other calls) can ModRef pointers that alias their arguments as well. This fixes PR2057.
Owen Anderson
Major improvements to yesterday's return slot optimization. Remove some unneccessary constraints, and add some others that should have been in from the first place. Document the whole thing better.
Chris Lattner
Fix some minor issues folding undef, PR2052
Chris Lattner
fdiv/frem of undef can produce undef, because the undef operand can be a SNaN. We could be more aggressive and turn this into unreachable, but that is less nice, and not really worth it.
Chris Lattner
regenerate
Chris Lattner
Fix PR2060 by rejecting invalid types for integer constants.
Owen Anderson
Factor the profitability check for return slot optimization out into a static function. At some point in the future, this check will become smarter.
Owen Anderson
An sret parameter is required to be the first parameter, so there's no need to loop over all the parameters of the callee looking for it.
Nick Lewycky
Don't spew stats to stderr.
Owen Anderson
Cleanup some of my patches from yesterday. Refactor the check for which xform to apply to a memcpy into processInstruction. Also, fix a bug in the check due to missing braces.
Nick Lewycky
Fix up the run line for this new test. llc: for the -info-output-file option: requires a value!
Owen Anderson
Fix Transforms/GVN/memcpy.ll, which Chris broke in r47275 by reordering the branches. memcpy's are a kind of CallInst.
Evan Cheng
New test.
Evan Cheng
PR1909: Tail merging pass ran wild. It makes no sense to merge blocks in order to save a single instruction since a branch will be inserted for each BB.
Evan Cheng
Me not like duplicated comments.
Tanya Lattner (Tanya Brethour)
Remove llvm-upgrade and update tests.
Tanya Lattner (Tanya Brethour)
Remove llvm-upgrade and update tests.
Evan Cheng
- When DAG combiner is folding a bit convert into a BUILD_VECTOR, it should check if it's essentially a SCALAR_TO_VECTOR. Avoid turning (v8i16) <10, u, u, u> to <10, 0, u, u, u, u, u, u>. Instead, simply convert it to a SCALAR_TO_VECTOR of the proper type. - X86 now normalize SCALAR_TO_VECTOR to (BIT_CONVERT (v4i32 SCALAR_TO_VECTOR)). Get rid of X86ISD::S2VEC.
Nick Lewycky
Correctly fold divide-by-constant, even when faced with overflow.
Dan Gohman
Chris pointed out that it's not necessary to set i64 MUL to Expand on x86-32 since i64 itself is not a Legal type. And, update some comments.
Evan Cheng
- Remove the previous check which broke coalescer-commute3.ll - For now, conservatively ignore copy MI whose source is a physical register. Commuting its def MI can cause a physical register live interval to be live through a loop (since we know it's live coming into the def MI).
Chris Lattner
upgrade some tests.
  1. Prev
  2. Next