| 2009-10-31 |
Dan Gohman | Rename forgetLoopBackedgeTakenCount to forgetLoop,... |
tree | commitdiff |
| 2009-10-31 |
Dan Gohman | Replace LoopUnrollPass.cpp's custom code-size estimatio... |
tree | commitdiff |
| 2009-10-31 |
Dan Gohman | Simplify this code. |
tree | commitdiff |
| 2009-10-31 |
Dan Gohman | Remove an unnecessary #include. |
tree | commitdiff |
| 2009-10-31 |
Dan Gohman | Update CMakeLists for recent renames. |
tree | commitdiff |
| 2009-10-31 |
Dan Gohman | Rename UnrollLoop.cpp to LoopUnroll.cpp, and LoopUnroll... |
tree | commitdiff |
| 2009-10-31 |
Dan Gohman | Remove CodeGenLICM. It's largely obsoleted by MachineLI... |
tree | commitdiff |
| 2009-10-31 |
Dan Gohman | Make ScalarEvolutionAliasAnalysis slightly more aggress... |
tree | commitdiff |
| 2009-10-31 |
Dan Gohman | Reapply r85634, with the bug fixed. |
tree | commitdiff |
| 2009-10-31 |
Dan Gohman | When discarding SrcValue information, discard all of... |
tree | commitdiff |
| 2009-10-31 |
Dan Gohman | Fix 80-column violation. |
tree | commitdiff |
| 2009-10-31 |
Eric Christopher | Fix warning with gcc-4.0 and signed/unsigned. |
tree | commitdiff |
| 2009-10-31 |
Evan Cheng | It's safe to remat t2LDRpci; Add PseudoSourceValue... |
tree | commitdiff |
| 2009-10-31 |
Evan Cheng | Revert 85634. It's breaking consumer-typeset (and others). |
tree | commitdiff |
| 2009-10-30 |
Dan Gohman | Add assertion checks here to turn silent miscompiles... |
tree | commitdiff |
| 2009-10-30 |
Dan Gohman | Don't mark registers dead here when processing nodes... |
tree | commitdiff |
| 2009-10-30 |
Dan Gohman | Add a comment about a missed opportunity. |
tree | commitdiff |
| 2009-10-30 |
Dan Gohman | Optimize around the fact that pred_iterator is slow... |
tree | commitdiff |
| 2009-10-30 |
Kevin Enderby | Updates to the ARM target assembler for llvm-mc per... |
tree | commitdiff |
| 2009-10-30 |
Devang Patel | If string field is empty then return NULL. |
tree | commitdiff |
| 2009-10-30 |
Chris Lattner | if basic blocks are destroyed while there are *just... |
tree | commitdiff |
| 2009-10-30 |
Dan Gohman | Teach SimplifyCFG how to eliminate duplicate PHI nodes... |
tree | commitdiff |
| 2009-10-30 |
Chris Lattner | make hasAddressTaken() constant time by storing a refco... |
tree | commitdiff |
| 2009-10-30 |
Bob Wilson | Add a note about Robert Muth's alternate jump table... |
tree | commitdiff |
| 2009-10-30 |
Dan Gohman | Sort the incoming values in PHI nodes to match the... |
tree | commitdiff |
| 2009-10-30 |
Dan Gohman | Fix MachineLICM to use the correct virtual register... |
tree | commitdiff |
| 2009-10-30 |
Chris Lattner | it isn't valid to take the address of the entry block. |
tree | commitdiff |
| 2009-10-30 |
Devang Patel | If a type is derived from a derived type then calculate... |
tree | commitdiff |
| 2009-10-30 |
Bob Wilson | Fix a comment. |
tree | commitdiff |
| 2009-10-30 |
Evan Cheng | Add option to createGVNPass to disable PRE. |
tree | commitdiff |
| 2009-10-30 |
Lang Hames | Stop the iterator in ValueLiveAt from potentially runni... |
tree | commitdiff |
| 2009-10-30 |
Rafael Espindola | This fixes functions like |
tree | commitdiff |
| 2009-10-30 |
Bob Wilson | Add ARM codegen for indirect branches. |
tree | commitdiff |
| 2009-10-30 |
Dan Gohman | Most stack straces don't need 3 digits worth of levels. |
tree | commitdiff |
| 2009-10-30 |
Dan Gohman | Don't delete blocks which have their address taken. |
tree | commitdiff |
| 2009-10-30 |
Dan Gohman | Mention if a block has its address taken in debug output. |
tree | commitdiff |
| 2009-10-30 |
Dan Gohman | Simplify this code and avoid an extra space character... |
tree | commitdiff |
| 2009-10-30 |
Dan Gohman | Add support for BlockAddress static initializers. |
tree | commitdiff |
| 2009-10-30 |
Dan Gohman | Add a FIXME comment. |
tree | commitdiff |
| 2009-10-30 |
Dan Gohman | Add some comments. |
tree | commitdiff |
| 2009-10-30 |
Dan Gohman | Initial x86 support for BlockAddresses. |
tree | commitdiff |
| 2009-10-30 |
Dan Gohman | Initial target-independent CodeGen support for BlockAdd... |
tree | commitdiff |
| 2009-10-30 |
Jim Grosbach | Dial back the realignment a bit. |
tree | commitdiff |
| 2009-10-29 |
David Goodwin | Between scheduling regions, correctly maintain anti... |
tree | commitdiff |
| 2009-10-29 |
Dan Gohman | Remove some unnecessary spaces in debug output. |
tree | commitdiff |
| 2009-10-29 |
Dan Gohman | Move some code from being emitted as boilerplate duplic... |
tree | commitdiff |
| 2009-10-29 |
David Goodwin | Fix a couple of bugs in aggressive anti-dep breaking. |
tree | commitdiff |
| 2009-10-29 |
Bob Wilson | Refactor complicated predicate into a separate function. |
tree | commitdiff |
| 2009-10-29 |
Devang Patel | First bitcase use may not lead to a dbg.declare intrins... |
tree | commitdiff |
| 2009-10-29 |
Dan Gohman | Rename usesCustomDAGSchedInserter to usesCustomInserter... |
tree | commitdiff |
| 2009-10-29 |
Dan Gohman | Refactor the code for unfolding a load into a separate... |
tree | commitdiff |
| 2009-10-29 |
Bill Wendling | Reapply r85338. |
tree | commitdiff |
| 2009-10-29 |
Benjamin Kramer | Fix MSVC build. |
tree | commitdiff |
| 2009-10-29 |
Nick Lewycky | Apply some cleanups. No functionality changes. |
tree | commitdiff |
| 2009-10-29 |
Chris Lattner | add sanity check for indbr. |
tree | commitdiff |
| 2009-10-29 |
Chris Lattner | just for the hell of it, allow globalopt to statically... |
tree | commitdiff |
| 2009-10-29 |
Chris Lattner | add interpreter support for indirect goto / blockaddres... |
tree | commitdiff |
| 2009-10-29 |
Zhongxing Xu | fix 80-col. |
tree | commitdiff |
| 2009-10-29 |
Zhongxing Xu | Explicitly convert to double to suppress Visual C+... |
tree | commitdiff |
| 2009-10-29 |
Jim Grosbach | To get more thorough testing from llc-beta nightly... |
tree | commitdiff |
| 2009-10-29 |
Jim Grosbach | When the function is doing dynamic stack realignment... |
tree | commitdiff |
| 2009-10-29 |
Chris Lattner | teach various passes about blockaddress. We no longer |
tree | commitdiff |
| 2009-10-29 |
Dale Johannesen | When there is a 2-instruction spill sequence, record |
tree | commitdiff |
| 2009-10-29 |
Bill Wendling | Don't put in these EH changes. |
tree | commitdiff |
| 2009-10-29 |
Chris Lattner | teach ValueMapper about BlockAddress', making bugpoint... |
tree | commitdiff |
| 2009-10-29 |
Chris Lattner | unindent massive blocks, no functionality change. |
tree | commitdiff |
| 2009-10-29 |
Bill Wendling | Reverting r85338 for now. It's causing a bootstrap... |
tree | commitdiff |
| 2009-10-29 |
Dan Gohman | Add a hasAddressTaken for BasicBlock. |
tree | commitdiff |
| 2009-10-28 |
Bob Wilson | Reimplement BranchFolding change to avoid tail merging... |
tree | commitdiff |
| 2009-10-28 |
Dale Johannesen | When we generate spill code, then decide we don't need |
tree | commitdiff |
| 2009-10-28 |
Eric Christopher | Make sure we return the right sized type here. |
tree | commitdiff |
| 2009-10-28 |
Bob Wilson | Revert r85346 change to control tail merging by CodeGen... |
tree | commitdiff |
| 2009-10-28 |
Victor Hernandez | Extend getMallocArraySize() to determine the array... |
tree | commitdiff |
| 2009-10-28 |
David Goodwin | Make AntiDepReg.h internal. |
tree | commitdiff |
| 2009-10-28 |
Bob Wilson | Add a Thumb BRIND pattern. Change the ARM BRIND assemb... |
tree | commitdiff |
| 2009-10-28 |
Evan Cheng | fconsts and fconstd are obviously re-materializable. |
tree | commitdiff |
| 2009-10-28 |
Jim Grosbach | Cleanup now that frame index scavenging via post-pass... |
tree | commitdiff |
| 2009-10-28 |
Devang Patel | llvm.dbg.global_variables do not exist anymore. |
tree | commitdiff |
| 2009-10-28 |
Chris Lattner | add a new 'SetCurrentDebugType' API (requested by Andre... |
tree | commitdiff |
| 2009-10-28 |
Dan Gohman | Don't call SDNode::isPredecessorOf when it isn't necess... |
tree | commitdiff |
| 2009-10-28 |
Dan Gohman | Simplify this code: if the unfolded load can't be hoist... |
tree | commitdiff |
| 2009-10-28 |
Edward O'Callaghan | No newline at end of file. |
tree | commitdiff |
| 2009-10-28 |
Benjamin Kramer | Update CMake file. |
tree | commitdiff |
| 2009-10-28 |
Owen Anderson | Treat lifetime begin/end markers as allocations/frees... |
tree | commitdiff |
| 2009-10-28 |
Nick Lewycky | Add ABCD, a generalized implementation of the Eliminati... |
tree | commitdiff |
| 2009-10-28 |
Evan Cheng | Give ARMISD::EH_SJLJ_LONGJMP and EH_SJLJ_SETJMP names. |
tree | commitdiff |
| 2009-10-28 |
Owen Anderson | Be more careful about invariance reasoning on "store... |
tree | commitdiff |
| 2009-10-28 |
Evan Cheng | X86 palignr intrinsics immediate field is in bits.... |
tree | commitdiff |
| 2009-10-28 |
Owen Anderson | Add trivial support for the invariance intrinsics to... |
tree | commitdiff |
| 2009-10-28 |
Chris Lattner | add bitcode reader support for blockaddress. We can... |
tree | commitdiff |
| 2009-10-28 |
Chris Lattner | bitcode writer support for blockaddress. |
tree | commitdiff |
| 2009-10-28 |
Chris Lattner | Previously, all operands to Constant were themselves... |
tree | commitdiff |
| 2009-10-28 |
Chris Lattner | 'static const void *X = &&y' can only be put in the |
tree | commitdiff |
| 2009-10-28 |
Dan Gohman | Rewrite SelectionDAG::isPredecessorOf to be iterative... |
tree | commitdiff |
| 2009-10-28 |
Chris Lattner | full asmparser support for blockaddress. We can now do: |
tree | commitdiff |
| 2009-10-28 |
Chris Lattner | asmprinter support for BlockAddress. |
tree | commitdiff |
| 2009-10-28 |
Chris Lattner | when we tear down a module, we need to be careful to |
tree | commitdiff |
| 2009-10-28 |
Dan Gohman | Teach MachineLICM to unfold loads from constant memory... |
tree | commitdiff |
| 2009-10-28 |
Evan Cheng | Use fconsts and fconstd to materialize small fp constants. |
tree | commitdiff |
| 2009-10-28 |
Evan Cheng | Add a second ValueType argument to isFPImmLegal. |
tree | commitdiff |
| next |
projects / oota-llvm.git / history