2009-05-05 |
Dan Gohman | If a MachineBasicBlock has multiple ways of reaching... |
tree | commitdiff |
2009-05-05 |
Evan Cheng | Enable stack coloring with regs at -O3. |
tree | commitdiff |
2009-05-05 |
Chris Lattner | Do not require variable debug info nodes to have a... |
tree | commitdiff |
2009-05-05 |
Evan Cheng | Do not substitute if the new register isn't in the... |
tree | commitdiff |
2009-05-05 |
Evan Cheng | Move getInstrOperandRegClass from the scheduler to... |
tree | commitdiff |
2009-05-04 |
Evan Cheng | Do forward and backward substitution to eliminate loads... |
tree | commitdiff |
2009-05-04 |
Chris Lattner | Make DBG_STOPPOINT nodes, and therefore DBG_LABEL label... |
tree | commitdiff |
2009-05-04 |
Argyrios Kyrtzidis | Restore a comment. |
tree | commitdiff |
2009-05-04 |
Mike Stump | Restore minor deletion. |
tree | commitdiff |
2009-05-03 |
Evan Cheng | Typo. |
tree | commitdiff |
2009-05-03 |
Evan Cheng | In some rare cases, the register allocator can spill... |
tree | commitdiff |
2009-05-03 |
Anton Korobeynikov | Fix typo |
tree | commitdiff |
2009-05-03 |
Anton Korobeynikov | Properly handle sdiv / udiv / srem / urem libcalls |
tree | commitdiff |
2009-05-03 |
Anton Korobeynikov | Proper name 16 bit libcalls |
tree | commitdiff |
2009-05-03 |
Anton Korobeynikov | Add libcall expansion for 16 and 128 bit muls |
tree | commitdiff |
2009-05-03 |
Argyrios Kyrtzidis | -Move the DwarfWriter::ValidDebugInfo check to a static... |
tree | commitdiff |
2009-05-01 |
Bob Wilson | Allow CONCAT_VECTORS nodes to be legal or have custom... |
tree | commitdiff |
2009-05-01 |
Bill Wendling | Simplify more code and add timer stuff. |
tree | commitdiff |
2009-05-01 |
Bill Wendling | Simplify more code. |
tree | commitdiff |
2009-05-01 |
Bill Wendling | Simplify some code. |
tree | commitdiff |
2009-05-01 |
Bill Wendling | Fix whitespace. It was confusing me. |
tree | commitdiff |
2009-05-01 |
Evan Cheng | Code clean up. Bye bye PhysRegTracker. |
tree | commitdiff |
2009-04-30 |
Argyrios Kyrtzidis | Make DebugLoc independent of DwarfWriter. |
tree | commitdiff |
2009-04-30 |
Jakob Stoklund Olesen | Join cross class copies using getCommonSubClass() |
tree | commitdiff |
2009-04-30 |
Evan Cheng | Add a smarter heuristic to determine when to coalesce... |
tree | commitdiff |
2009-04-30 |
Jay Foad | Move helper functions for optimizing division by consta... |
tree | commitdiff |
2009-04-30 |
Chris Lattner | fix a regression handling indirect results: these need... |
tree | commitdiff |
2009-04-29 |
Bill Wendling | Instead of passing in an unsigned value for the optimiz... |
tree | commitdiff |
2009-04-29 |
Nate Begeman | Fix infinite recursion in the C++ code which handles... |
tree | commitdiff |
2009-04-29 |
Jakob Stoklund Olesen | MachineInstr::isRegTiedTo{Use,Def}Operand can safely... |
tree | commitdiff |
2009-04-29 |
Nate Begeman | Update comment, replace theoretically impossible check... |
tree | commitdiff |
2009-04-29 |
Evan Cheng | spillPhysRegAroundRegDefsUses() may have invalidated... |
tree | commitdiff |
2009-04-29 |
Nate Begeman | Implement review feedback for vector shuffle work. |
tree | commitdiff |
2009-04-29 |
Sanjiv Gupta | Add a public method called getAddressSpace() to the... |
tree | commitdiff |
2009-04-29 |
Chris Lattner | Disable the load-shrinking optimization from looking at |
tree | commitdiff |
2009-04-29 |
Evan Cheng | Determine allocation 'preference' with right register... |
tree | commitdiff |
2009-04-29 |
Bill Wendling | Second attempt: |
tree | commitdiff |
2009-04-28 |
Evan Cheng | Move getMatchingSuperReg() out of coalescer and into... |
tree | commitdiff |
2009-04-28 |
Jakob Stoklund Olesen | Don't coalesce a physical register with an incompatible... |
tree | commitdiff |
2009-04-28 |
Sanjiv Gupta | Initialized arrays can be in any address space. |
tree | commitdiff |
2009-04-28 |
Jakob Stoklund Olesen | Move getSubRegisterRegClass from ScheduleDagSDNodesEmit... |
tree | commitdiff |
2009-04-28 |
Evan Cheng | Fix PR4034. Bug in LiveInterval::join when it's compact... |
tree | commitdiff |
2009-04-28 |
Evan Cheng | Fix for PR4051. When 2address pass delete an instructio... |
tree | commitdiff |
2009-04-28 |
Bill Wendling | r70270 isn't ready yet. Back this out. Sorry for the... |
tree | commitdiff |
2009-04-28 |
Bill Wendling | Massive check in. This changes the "-fast" flag to... |
tree | commitdiff |
2009-04-27 |
Evan Cheng | Fix PR4076. Correctly create live interval of physical... |
tree | commitdiff |
2009-04-27 |
Owen Anderson | Don't skip the CopyMI when removing kill markers. |
tree | commitdiff |
2009-04-27 |
Duncan Sands | Now that PR2957 is resolved, remove a bunch of |
tree | commitdiff |
2009-04-27 |
Nate Begeman | 2nd attempt, fixing SSE4.1 issues and implementing... |
tree | commitdiff |
2009-04-27 |
Evan Cheng | Fix PR4056. It's possible a physical register def is... |
tree | commitdiff |
2009-04-27 |
Evan Cheng | Also delete last unused val#. |
tree | commitdiff |
2009-04-27 |
Dan Gohman | When transforming sext(trunc(load(x))) into sext(smalle... |
tree | commitdiff |
2009-04-25 |
Evan Cheng | Reuse unused val#'s to avoid running out of memory... |
tree | commitdiff |
2009-04-25 |
Dan Gohman | Refactor the code to grab the low and high parts of... |
tree | commitdiff |
2009-04-25 |
Dan Gohman | Add a top-level comment about DAGCombiner's role in... |
tree | commitdiff |
2009-04-25 |
Evan Cheng | Do not share a single unknown val# for all the live... |
tree | commitdiff |
2009-04-24 |
Dale Johannesen | Fix PR 4057, a crash doing float->char const folding. |
tree | commitdiff |
2009-04-24 |
Rafael Espindola | Revert 69952. Causes testsuite failures on linux x86-64. |
tree | commitdiff |
2009-04-24 |
Nate Begeman | PR2957 |
tree | commitdiff |
2009-04-24 |
Dan Gohman | Instead of requiring TLI.LowerCallTo to return an ISD... |
tree | commitdiff |
2009-04-23 |
Dan Gohman | Factor out a bit of code that appears in several places... |
tree | commitdiff |
2009-04-23 |
Dan Gohman | Handle Void types in ComputeValueVTs. This doesn't... |
tree | commitdiff |
2009-04-23 |
Evan Cheng | Update comments. |
tree | commitdiff |
2009-04-23 |
Evan Cheng | Fix an obvious type. |
tree | commitdiff |
2009-04-21 |
Evan Cheng | It has finally happened. Spiller is now using live... |
tree | commitdiff |
2009-04-21 |
Devang Patel | Fix Visual Studio 2008 build failure. |
tree | commitdiff |
2009-04-20 |
Dan Gohman | Make X86's copyRegToReg able to handle copies to and... |
tree | commitdiff |
2009-04-20 |
Dan Gohman | Simplify this code. getConstant knows how to make |
tree | commitdiff |
2009-04-20 |
Bob Wilson | Move duplicated AddLiveIn function from X86 and ARM... |
tree | commitdiff |
2009-04-20 |
Bob Wilson | Revise my previous change 68996 as suggested by Duncan. |
tree | commitdiff |
2009-04-20 |
Evan Cheng | - Remove an arbitrary spill weight tweak that should... |
tree | commitdiff |
2009-04-20 |
Evan Cheng | Added a linearscan register allocation optimization... |
tree | commitdiff |
2009-04-19 |
Duncan Sands | Now that BUILD_VECTOR operands are allowed to be |
tree | commitdiff |
2009-04-18 |
Chris Lattner | Fix PR3898, which manifests as failures on are an Xcore, |
tree | commitdiff |
2009-04-18 |
Duncan Sands | Don't try to make BUILD_VECTOR operands have the same |
tree | commitdiff |
2009-04-18 |
Evan Cheng | Add a new LiveInterval::overlaps(). It checks if the... |
tree | commitdiff |
2009-04-18 |
Dale Johannesen | Inline asm's were still introducing bogus dependencies; |
tree | commitdiff |
2009-04-17 |
Evan Cheng | Teach spiller to unfold instructions which modref spill... |
tree | commitdiff |
2009-04-16 |
Dan Gohman | In the list-burr's pseudo two-addr dependency heuristic... |
tree | commitdiff |
2009-04-16 |
Dan Gohman | Handle SUBREG_TO_REG instructions with the same heuristics |
tree | commitdiff |
2009-04-16 |
Devang Patel | Do not treat beginning of inlined scope as beginning... |
tree | commitdiff |
2009-04-16 |
Devang Patel | Record line number at the beginning of a func.start. |
tree | commitdiff |
2009-04-16 |
Devang Patel | In -fast mode do what FastISel does. |
tree | commitdiff |
2009-04-16 |
Devang Patel | If FastISel is run and it has known DebugLoc then use it. |
tree | commitdiff |
2009-04-16 |
Devang Patel | If location where the function was inlined is not know... |
tree | commitdiff |
2009-04-15 |
Devang Patel | s/RootDbgScope/FunctionDbgScope/g |
tree | commitdiff |
2009-04-15 |
Devang Patel | Add DISubprogram is not null check. |
tree | commitdiff |
2009-04-15 |
Dan Gohman | Generalize one of the SelectionDAG::ReplaceAllUsesWith... |
tree | commitdiff |
2009-04-15 |
Devang Patel | Check isInlinedSubroutine() before creating DW_TAG_inli... |
tree | commitdiff |
2009-04-15 |
Dan Gohman | Fix MachineInstr::getNumExplicitOperands to count |
tree | commitdiff |
2009-04-15 |
Dan Gohman | Move MachineRegisterInfo::setRegClass out of line. |
tree | commitdiff |
2009-04-15 |
Dan Gohman | Move MachineJumpTableInfo::ReplaceMBBInJumpTables out... |
tree | commitdiff |
2009-04-15 |
Dan Gohman | Give RemoveRegOperandFromRegInfo a comment and move the |
tree | commitdiff |
2009-04-15 |
Devang Patel | Construct and emit DW_TAG_inlined_subroutine DIEs for... |
tree | commitdiff |
2009-04-14 |
Dan Gohman | When the result of an EXTRACT_SUBREG, INSERT_SUBREG... |
tree | commitdiff |
2009-04-14 |
Dale Johannesen | Do not force asm's to be chained if they don't touch |
tree | commitdiff |
2009-04-14 |
Evan Cheng | Fix PR3934 part 2. findOnlyInterestingUse() was not... |
tree | commitdiff |
2009-04-13 |
Daniel Dunbar | Make these errors more noticable in build logs. |
tree | commitdiff |
2009-04-13 |
Bob Wilson | Change SelectionDAG type legalization to allow BUILD_VE... |
tree | commitdiff |
2009-04-13 |
Dan Gohman | Rename COPY_TO_SUBCLASS to COPY_TO_REGCLASS, and generalize |
tree | commitdiff |
next |