2010-09-02 |
Chris Lattner | reapply 112894:
|
commit | commitdiff | tree |
2010-09-02 |
Chris Lattner | revert patch, need to update clang tests too
|
commit | commitdiff | tree |
2010-09-02 |
Chris Lattner | Remove #uses comments from functions: they we're padded...
|
commit | commitdiff | tree |
2010-09-02 |
Chris Lattner | more test cleanup
|
commit | commitdiff | tree |
2010-09-02 |
Chris Lattner | remove some noise from tests.
|
commit | commitdiff | tree |
2010-09-02 |
Chris Lattner | we are past the point where these tests are useful.
|
commit | commitdiff | tree |
2010-09-02 |
Chris Lattner | fix more AST updating bugs, correcting miscompilation...
|
commit | commitdiff | tree |
2010-09-02 |
Chris Lattner | updates to make it clear that this applies to clang...
|
commit | commitdiff | tree |
2010-09-01 |
Chris Lattner | deepen my MMX/SRoA hack to avoid hurting non-x86 codegen.
|
commit | commitdiff | tree |
2010-09-01 |
Chris Lattner | update the tutorial to use CreateFAdd to create fp...
|
commit | commitdiff | tree |
2010-09-01 |
Chris Lattner | have the makefiles check the llvm-config error code...
|
commit | commitdiff | tree |
2010-09-01 |
Chris Lattner | dead code patrol
|
commit | commitdiff | tree |
2010-09-01 |
Chris Lattner | zap dead code.
|
commit | commitdiff | tree |
2010-09-01 |
Chris Lattner | temporarily revert r112664, it is causing a decoding...
|
commit | commitdiff | tree |
2010-09-01 |
Chris Lattner | zap dead code.
|
commit | commitdiff | tree |
2010-09-01 |
Chris Lattner | remove dead code.
|
commit | commitdiff | tree |
2010-09-01 |
Chris Lattner | add a gross hack to work around a problem that Argiris...
|
commit | commitdiff | tree |
2010-09-01 |
Chris Lattner | filecheckize
|
commit | commitdiff | tree |
2010-08-31 |
Chris Lattner | licm is wasting time hoisting constant foldable operations,
|
commit | commitdiff | tree |
2010-08-31 |
Chris Lattner | tidy up
|
commit | commitdiff | tree |
2010-08-31 |
Chris Lattner | merge two tests.
|
commit | commitdiff | tree |
2010-08-31 |
Chris Lattner | merge two tests and convert to filecheck.
|
commit | commitdiff | tree |
2010-08-30 |
Chris Lattner | nuke dead ivar which was supposed to be committed with...
|
commit | commitdiff | tree |
2010-08-30 |
Chris Lattner | two changes:
|
commit | commitdiff | tree |
2010-08-29 |
Chris Lattner | revert 112457, it looks like it broke selfhost.
|
commit | commitdiff | tree |
2010-08-29 |
Chris Lattner | rewrite DwarfEHPrepare to use SSAUpdater to promote...
|
commit | commitdiff | tree |
2010-08-29 |
Chris Lattner | inline function into its only caller.
|
commit | commitdiff | tree |
2010-08-29 |
Chris Lattner | two changes: 1) make AliasSet hold the list of call...
|
commit | commitdiff | tree |
2010-08-29 |
Chris Lattner | LICM does get dead instructions input to it. Instead...
|
commit | commitdiff | tree |
2010-08-29 |
Chris Lattner | use moveBefore instead of remove+insert, it avoids...
|
commit | commitdiff | tree |
2010-08-29 |
Chris Lattner | revert 112448 for now.
|
commit | commitdiff | tree |
2010-08-29 |
Chris Lattner | optimize LICM::hoist to use moveBefore. Correct its...
|
commit | commitdiff | tree |
2010-08-29 |
Chris Lattner | fix some bugs (found by inspection) where LICM would...
|
commit | commitdiff | tree |
2010-08-29 |
Chris Lattner | rework the ownership of subloop alias information:...
|
commit | commitdiff | tree |
2010-08-29 |
Chris Lattner | apparently unswitch had the same "Feature". Stop its
|
commit | commitdiff | tree |
2010-08-29 |
Chris Lattner | now that loop passes don't use DomFrontier, there is...
|
commit | commitdiff | tree |
2010-08-29 |
Chris Lattner | Stop explicitly scheduling domfrontier before the loop...
|
commit | commitdiff | tree |
2010-08-29 |
Chris Lattner | licm preserves the cfg, it doesn't have to explicitly...
|
commit | commitdiff | tree |
2010-08-29 |
Chris Lattner | now that it doesn't use the PromoteMemToReg function...
|
commit | commitdiff | tree |
2010-08-29 |
Chris Lattner | completely rewrite the memory promotion algorithm in...
|
commit | commitdiff | tree |
2010-08-29 |
Chris Lattner | use getUniqueExitBlocks instead of a manual set.
|
commit | commitdiff | tree |
2010-08-29 |
Chris Lattner | reimplement LICM::sink to use SSAUpdater instead of...
|
commit | commitdiff | tree |
2010-08-29 |
Chris Lattner | implement SSAUpdater::RewriteUseAfterInsertions, a...
|
commit | commitdiff | tree |
2010-08-29 |
Chris Lattner | remove dead proto
|
commit | commitdiff | tree |
2010-08-29 |
Chris Lattner | reduce indentation in LICM::sink by using early exits...
|
commit | commitdiff | tree |
2010-08-29 |
Chris Lattner | modernize this pass a bit: use efficient set/map and...
|
commit | commitdiff | tree |
2010-08-29 |
Chris Lattner | when merging two alias sets, the result set is volatile...
|
commit | commitdiff | tree |
2010-08-29 |
Chris Lattner | more cleanup
|
commit | commitdiff | tree |
2010-08-29 |
Chris Lattner | clean this up
|
commit | commitdiff | tree |
2010-08-29 |
Chris Lattner | merge a bunch of shuffle tests into sse2.ll
|
commit | commitdiff | tree |
2010-08-29 |
Chris Lattner | add a bunch more common shuffles to the instprinter.
|
commit | commitdiff | tree |
2010-08-29 |
Chris Lattner | add some nounwind's
|
commit | commitdiff | tree |
2010-08-28 |
Chris Lattner | I have manually decoded the imm field of an insertps...
|
commit | commitdiff | tree |
2010-08-28 |
Chris Lattner | fixme accomplished
|
commit | commitdiff | tree |
2010-08-28 |
Chris Lattner | tidy up
|
commit | commitdiff | tree |
2010-08-28 |
Chris Lattner | fix the buildvector->insertp[sd] logic to not always...
|
commit | commitdiff | tree |
2010-08-28 |
Chris Lattner | fix the BuildVector -> unpcklps logic to not do pointless...
|
commit | commitdiff | tree |
2010-08-28 |
Chris Lattner | improve comments in the unpcklps generating logic,...
|
commit | commitdiff | tree |
2010-08-28 |
Chris Lattner | remove the MSIL backend. It isn't maintained, is buggy...
|
commit | commitdiff | tree |
2010-08-28 |
Chris Lattner | remove unions from LLVM IR. They are severely buggy...
|
commit | commitdiff | tree |
2010-08-28 |
Chris Lattner | remove the ABCD and SSI passes. They don't have any...
|
commit | commitdiff | tree |
2010-08-28 |
Chris Lattner | remove dead proto
|
commit | commitdiff | tree |
2010-08-28 |
Chris Lattner | more dead thing zapping.
|
commit | commitdiff | tree |
2010-08-28 |
Chris Lattner | zap dead method
|
commit | commitdiff | tree |
2010-08-28 |
Chris Lattner | for completeness, allow undef also.
|
commit | commitdiff | tree |
2010-08-28 |
Chris Lattner | squish dead code.
|
commit | commitdiff | tree |
2010-08-28 |
Chris Lattner | zap dead code
|
commit | commitdiff | tree |
2010-08-28 |
Chris Lattner | handle the constant case of vector insertion. For...
|
commit | commitdiff | tree |
2010-08-28 |
Chris Lattner | optimize bitcasts from large integers to vector into...
|
commit | commitdiff | tree |
2010-08-27 |
Chris Lattner | tidy up test.
|
commit | commitdiff | tree |
2010-08-27 |
Chris Lattner | no really, fix the test.
|
commit | commitdiff | tree |
2010-08-27 |
Chris Lattner | fix this test. It's not clear what it's really testing.
|
commit | commitdiff | tree |
2010-08-27 |
Chris Lattner | Enhance the shift propagator to handle the case when...
|
commit | commitdiff | tree |
2010-08-27 |
Chris Lattner | Implement a pretty general logical shift propagation
|
commit | commitdiff | tree |
2010-08-27 |
Chris Lattner | remove some special shift cases that have been subsumed...
|
commit | commitdiff | tree |
2010-08-27 |
Chris Lattner | merge and filecheckize test
|
commit | commitdiff | tree |
2010-08-27 |
Chris Lattner | merge two tests
|
commit | commitdiff | tree |
2010-08-27 |
Chris Lattner | teach the truncation optimization that an entire chain of
|
commit | commitdiff | tree |
2010-08-27 |
Chris Lattner | get this test passing on linux builders.
|
commit | commitdiff | tree |
2010-08-27 |
Chris Lattner | Add an instcombine to clean up a common pattern produced
|
commit | commitdiff | tree |
2010-08-26 |
Chris Lattner | filecheckize
|
commit | commitdiff | tree |
2010-08-26 |
Chris Lattner | rename test.
|
commit | commitdiff | tree |
2010-08-26 |
Chris Lattner | optimize "integer extraction out of the middle of a...
|
commit | commitdiff | tree |
2010-08-26 |
Chris Lattner | optimize bitcast(trunc(bitcast(x))) where the result...
|
commit | commitdiff | tree |
2010-08-26 |
Chris Lattner | filecheckize
|
commit | commitdiff | tree |
2010-08-26 |
Chris Lattner | rename test
|
commit | commitdiff | tree |
2010-08-26 |
Chris Lattner | add m_BitCast for matching a bitcast.
|
commit | commitdiff | tree |
2010-08-26 |
Chris Lattner | Add a hackaround for PR7993 which is causing failures...
|
commit | commitdiff | tree |
2010-08-26 |
Chris Lattner | I think enough general codegen bugs are fixed to allow...
|
commit | commitdiff | tree |
2010-08-26 |
Chris Lattner | implement SplitVecOp_CONCAT_VECTORS, fixing the included...
|
commit | commitdiff | tree |
2010-08-26 |
Chris Lattner | Make sure this forces the x86 targets
|
commit | commitdiff | tree |
2010-08-26 |
Chris Lattner | fix sse1 only codegen in x86-64 mode, which is something we
|
commit | commitdiff | tree |
2010-08-26 |
Chris Lattner | zap dead code.
|
commit | commitdiff | tree |
2010-08-26 |
Chris Lattner | remove dead proto
|
commit | commitdiff | tree |
2010-08-26 |
Chris Lattner | zap dead code.
|
commit | commitdiff | tree |
2010-08-25 |
Chris Lattner | temporarily disable this, which started failing on...
|
commit | commitdiff | tree |
2010-08-25 |
Chris Lattner | we should pattern match the SSE complex arithmetic...
|
commit | commitdiff | tree |
2010-08-25 |
Chris Lattner | add a specialization for the MVT form of getTypeAction...
|
commit | commitdiff | tree |
2010-08-25 |
Chris Lattner | remove some llvmcontext arguments that are now dead...
|
commit | commitdiff | tree |
2010-08-25 |
Chris Lattner | Change handling of illegal vector types to widen when...
|
commit | commitdiff | tree |