2011-09-20 |
Andrew Trick | ARM isel bug fix for adds/subs operands. |
tree | commitdiff |
2011-09-20 |
Andrew Trick | whitespace |
tree | commitdiff |
2011-09-18 |
Nadav Rotem | white space cleanups |
tree | commitdiff |
2011-09-16 |
Benjamin Kramer | Namespacify. |
tree | commitdiff |
2011-09-16 |
Jakob Stoklund Olesen | Spill mode: Hoist back-copies locally. |
tree | commitdiff |
2011-09-16 |
Jakob Stoklund Olesen | Disable local spill hoisting for non-killing copies. |
tree | commitdiff |
2011-09-15 |
Eli Friedman | Some legalization fixes for atomic load and store. |
tree | commitdiff |
2011-09-15 |
Jakob Stoklund Olesen | Add an option to disable spill hoisting. |
tree | commitdiff |
2011-09-15 |
Jakob Stoklund Olesen | VirtRegMap is counting spill slots, not register spills. |
tree | commitdiff |
2011-09-15 |
Jakob Stoklund Olesen | Count correctly when a COPY turns into a spill or reload. |
tree | commitdiff |
2011-09-15 |
Jakob Stoklund Olesen | Count inserted spills and reloads more accurately. |
tree | commitdiff |
2011-09-15 |
Jakob Stoklund Olesen | Trace through sibling PHIs in bulk. |
tree | commitdiff |
2011-09-15 |
Jakob Stoklund Olesen | Speed up LiveIntervals::shrinkToUse with some caching. |
tree | commitdiff |
2011-09-15 |
Jakob Stoklund Olesen | Revert r139782, "RemoveCopyByCommutingDef doesn't need... |
tree | commitdiff |
2011-09-15 |
Jakob Stoklund Olesen | Stop verifying hasPHIKill() flags. |
tree | commitdiff |
2011-09-15 |
Jakob Stoklund Olesen | RemoveCopyByCommutingDef doesn't need hasPHIKill(). |
tree | commitdiff |
2011-09-15 |
Jakob Stoklund Olesen | It is safe to remat a value killed by phis. |
tree | commitdiff |
2011-09-15 |
Jakob Stoklund Olesen | Leave hasPHIKill flags alone in LiveInterval::RenumberV... |
tree | commitdiff |
2011-09-15 |
Andrew Trick | [regcoalescing] bug fix for RegistersDefinedFromSameValue. |
tree | commitdiff |
2011-09-14 |
Devang Patel | Add support to emit debug info for C++0x nullptr type. |
tree | commitdiff |
2011-09-14 |
Jakob Stoklund Olesen | Ignore the cloning of unknown registers. |
tree | commitdiff |
2011-09-14 |
Jakob Stoklund Olesen | Hoist back-copies to the least busy dominator. |
tree | commitdiff |
2011-09-14 |
Nadav Rotem | Add integer promotion support for vselect |
tree | commitdiff |
2011-09-13 |
Jakob Stoklund Olesen | Distinguish complex mapped values from forced recomputa... |
tree | commitdiff |
2011-09-13 |
Jakob Stoklund Olesen | Implement -split-spill-mode=size. |
tree | commitdiff |
2011-09-13 |
Eli Friedman | Fix check for unaligned load/store so it doesn't catch... |
tree | commitdiff |
2011-09-13 |
Eli Friedman | Error out on CodeGen of unaligned load/store. Fix... |
tree | commitdiff |
2011-09-13 |
Nadav Rotem | Fix the assertion which checks the size of the input... |
tree | commitdiff |
2011-09-13 |
Nadav Rotem | Add vselect target support for targets that do not... |
tree | commitdiff |
2011-09-13 |
Devang Patel | Use a cache to maintain list of machine basic blocks... |
tree | commitdiff |
2011-09-13 |
Jakob Stoklund Olesen | Add SplitEditor::markOverlappedComplement(). |
tree | commitdiff |
2011-09-13 |
Jakob Stoklund Olesen | Eliminate the extendRange() wrapper. |
tree | commitdiff |
2011-09-13 |
Jakob Stoklund Olesen | Switch extendInBlock() to take a kill slot instead... |
tree | commitdiff |
2011-09-13 |
Jakob Stoklund Olesen | Use a separate LiveRangeCalc for the complement in... |
tree | commitdiff |
2011-09-13 |
NAKAMURA Takumi | Unbreak msvc. |
tree | commitdiff |
2011-09-13 |
Jakob Stoklund Olesen | Extract live range calculations from SplitKit. |
tree | commitdiff |
2011-09-12 |
Bill Wendling | Introduce a bit of a hack. |
tree | commitdiff |
2011-09-12 |
Jakob Stoklund Olesen | Remove the -compact-regions flag. |
tree | commitdiff |
2011-09-12 |
Jakob Stoklund Olesen | Add an interface for SplitKit complement spill modes. |
tree | commitdiff |
2011-09-12 |
Jakob Stoklund Olesen | Update comments to reflect some (not so) recent changes. |
tree | commitdiff |
2011-09-10 |
Richard Trieu | Fix asserts in CodeGen from: |
tree | commitdiff |
2011-09-09 |
Chris Lattner | tidy up a bit |
tree | commitdiff |
2011-09-09 |
Eli Friedman | Make the SelectionDAG verify that all the operands... |
tree | commitdiff |
2011-09-09 |
Jakob Stoklund Olesen | Reapply r139247: Cache intermediate results during... |
tree | commitdiff |
2011-09-08 |
Devang Patel | Directly point debug info to the stack slot of the... |
tree | commitdiff |
2011-09-07 |
Jakob Stoklund Olesen | Revert r139247 "Cache intermediate results during trace... |
tree | commitdiff |
2011-09-07 |
Jakob Stoklund Olesen | Cache intermediate results during traceSiblingValue. |
tree | commitdiff |
2011-09-07 |
James Molloy | Refactor instprinter and mcdisassembler to take a Subta... |
tree | commitdiff |
2011-09-07 |
Eli Friedman | Relax the MemOperands on atomics a bit. Fixes -verify... |
tree | commitdiff |
2011-09-07 |
Devang Patel | While sinking machine instructions, sink matching DBG_V... |
tree | commitdiff |
2011-09-06 |
Duncan Sands | Add codegen support for vector select (in the IR this... |
tree | commitdiff |
2011-09-06 |
Duncan Sands | Split the init.trampoline intrinsic, which currently... |
tree | commitdiff |
2011-09-03 |
Owen Anderson | Fix a truly heinous bug in DAGCombine related to Assert... |
tree | commitdiff |
2011-09-02 |
Jakob Stoklund Olesen | Simplify by using isFullCopy(). |
tree | commitdiff |
2011-09-02 |
Duncan Sands | Darwin wants ctors/dtors to be ordered the other way... |
tree | commitdiff |
2011-09-01 |
Dan Gohman | Revert r131152, r129796, r129761. This code is currentl... |
tree | commitdiff |
2011-09-01 |
Benjamin Kramer | Don't drop alignment info on local common symbols. |
tree | commitdiff |
2011-09-01 |
Jakob Stoklund Olesen | Permit remat of partial register defs when it is safe. |
tree | commitdiff |
2011-09-01 |
Jakob Stoklund Olesen | Revert r138794, "Do not try to rematerialize a value... |
tree | commitdiff |
2011-09-01 |
Jakob Stoklund Olesen | Prevent remat of partial register redefinitions. |
tree | commitdiff |
2011-09-01 |
Evan Cheng | Teach MachineLICM reg pressure tracking code to deal... |
tree | commitdiff |
2011-09-01 |
Andrew Trick | PreRA scheduler should avoid cloning compares. |
tree | commitdiff |
2011-08-31 |
David Greene | Fix Size Typing |
tree | commitdiff |
2011-08-31 |
Eli Friedman | Misc cleanup; addresses Duncan's comments on r138877. |
tree | commitdiff |
2011-08-31 |
Eli Friedman | Fill in type legalization for MERGE_VALUES in all the... |
tree | commitdiff |
2011-08-31 |
Eli Friedman | Generic expansion for atomic load/store into cmpxchg... |
tree | commitdiff |
2011-08-31 |
David Greene | Compress Repeated Byte Output |
tree | commitdiff |
2011-08-31 |
Rafael Espindola | Spelling and grammar fixes to problems found by Duncan. |
tree | commitdiff |
2011-08-30 |
Rafael Espindola | Emit segmented-stack specific code into function prolog... |
tree | commitdiff |
2011-08-30 |
Evan Cheng | Follow up to r138791. |
tree | commitdiff |
2011-08-30 |
Bob Wilson | Do not try to rematerialize a value from a partial... |
tree | commitdiff |
2011-08-29 |
Jim Grosbach | Thumb2 parsing and encoding for IT blocks. |
tree | commitdiff |
2011-08-28 |
Duncan Sands | Fix PR5329: pay attention to constructor/destructor... |
tree | commitdiff |
2011-08-27 |
Bill Wendling | These splits should be done whether they are critical... |
tree | commitdiff |
2011-08-26 |
Bill Wendling | Update the dominator tree with the correct dominator... |
tree | commitdiff |
2011-08-26 |
Bill Wendling | Split the landing pad block only if it's a critical... |
tree | commitdiff |
2011-08-26 |
Eli Friedman | Atomic load/store on ARM/Thumb. |
tree | commitdiff |
2011-08-25 |
Bill Wendling | Look at only the terminators of the basic block. Also... |
tree | commitdiff |
2011-08-24 |
Eli Friedman | Basic x86 code generation for atomic load and store... |
tree | commitdiff |
2011-08-24 |
Evan Cheng | Move TargetRegistry and TargetSelect from Target to... |
tree | commitdiff |
2011-08-24 |
Jim Grosbach | Tidy up. Trailing whitespace. |
tree | commitdiff |
2011-08-24 |
Bill Wendling | Add the sentinal "no handle" value to the ResumeInst. |
tree | commitdiff |
2011-08-23 |
Bill Wendling | Don't replace *all* uses with the new stuff. |
tree | commitdiff |
2011-08-23 |
Bill Wendling | Look at the end of the entry block for an invoke. |
tree | commitdiff |
2011-08-23 |
Bill Wendling | A landingpad instruction is neither folded nor dead. |
tree | commitdiff |
2011-08-23 |
Evan Cheng | Fix 80 col violations. |
tree | commitdiff |
2011-08-22 |
Bill Wendling | Split the landing pad's edge. Then for all uses of... |
tree | commitdiff |
2011-08-22 |
Evan Cheng | Follow up to Jim's r138278. This fixes commuteInstructi... |
tree | commitdiff |
2011-08-22 |
Bill Wendling | Some whitespace fixes and #include reordering. |
tree | commitdiff |
2011-08-22 |
Nick Lewycky | Be less redundant. |
tree | commitdiff |
2011-08-19 |
Devang Patel | Do not use named md nodes to track variables that are... |
tree | commitdiff |
2011-08-19 |
Benjamin Kramer | Roll back the rest of r126557. It's a hack that will... |
tree | commitdiff |
2011-08-19 |
Nick Lewycky | Eli points out that this is what report_fatal_error... |
tree | commitdiff |
2011-08-19 |
Nick Lewycky | This is not actually unreachable, so don't use llvm_unr... |
tree | commitdiff |
2011-08-19 |
Jakob Stoklund Olesen | Add llc flags to disable machine DCE and CSE. |
tree | commitdiff |
2011-08-19 |
Benjamin Kramer | Make a bunch of symbols private. |
tree | commitdiff |
2011-08-19 |
Jakob Stoklund Olesen | Don't treat a partial <def,undef> operand as a read. |
tree | commitdiff |
2011-08-18 |
Renato Golin | add the comments of each declaration follow it, making... |
tree | commitdiff |
2011-08-18 |
Devang Patel | Eliminate unnecessary forwarding function. |
tree | commitdiff |
2011-08-18 |
Devang Patel | Add new DIE into the map asap. |
tree | commitdiff |
next |