2010-06-08 |
Bruno Cardoso Lopes | Reapply r105521, this time appending "LLU" to 64 bit |
tree | commitdiff |
2010-06-08 |
Eric Christopher | Ensure that mov and not lea are used to stick the addre... |
tree | commitdiff |
2010-06-08 |
Jim Grosbach | fix typo |
tree | commitdiff |
2010-06-08 |
Daniel Dunbar | Use const_iterator where appropriate. |
tree | commitdiff |
2010-06-08 |
Daniel Dunbar | DeltaAlgorithm: Tweak split to split by first/second... |
tree | commitdiff |
2010-06-08 |
Daniel Dunbar | ADT: Add DAGDeltaAlgorithm, which is a DAG minimization... |
tree | commitdiff |
2010-06-08 |
Benjamin Kramer | Use realloc instead of malloc+memcpy when growing a... |
tree | commitdiff |
2010-06-08 |
Kalle Raiskila | Flag SPU's function call sequence together. |
tree | commitdiff |
2010-06-08 |
Bob Wilson | Fix up a comment. |
tree | commitdiff |
2010-06-07 |
Bob Wilson | Further changes for Neon vector shuffles: |
tree | commitdiff |
2010-06-07 |
Bob Wilson | Fix a mistake in my previous change r105437: don't... |
tree | commitdiff |
2010-06-07 |
Dan Gohman | Add some basic debug output. |
tree | commitdiff |
2010-06-07 |
Jim Grosbach | Handle dbg_value instructions (i.e., skip them) when... |
tree | commitdiff |
2010-06-07 |
Jim Grosbach | Cleanup. Process the dbg_values separately |
tree | commitdiff |
2010-06-07 |
Dan Gohman | Make bugpoint dead-argument-hacking actually work,... |
tree | commitdiff |
2010-06-07 |
Dan Gohman | The FoldingSet hash data includes pointer values, so... |
tree | commitdiff |
2010-06-07 |
Dan Gohman | Optimize this code somewhat by taking advantage of... |
tree | commitdiff |
2010-06-07 |
Bill Wendling | Another place where the code wanted to access the argum... |
tree | commitdiff |
2010-06-07 |
Dan Gohman | Micro-optimize this, to speed up this hotspot in debug... |
tree | commitdiff |
2010-06-07 |
Dan Gohman | Micro-optimize this. |
tree | commitdiff |
2010-06-07 |
Jim Grosbach | Move exit check where it really belongs. |
tree | commitdiff |
2010-06-07 |
Dan Gohman | Optimize ScalarEvolution's SCEVComplexityCompare predic... |
tree | commitdiff |
2010-06-07 |
Bill Wendling | Create new accessors to get arguments for call/invoke... |
tree | commitdiff |
2010-06-05 |
Kenneth Uildriks | Partial specialization was not checking the callsite... |
tree | commitdiff |
2010-06-05 |
Duncan Sands | This bug is also present in MSVC10. Requested by Elroo... |
tree | commitdiff |
2010-06-05 |
Chris Lattner | revert r105521, which is breaking the buildbots with... |
tree | commitdiff |
2010-06-05 |
Bruno Cardoso Lopes | Initial AVX support for some instructions. No patterns... |
tree | commitdiff |
2010-06-05 |
Dale Johannesen | Improvements to tail call code. No functional effect |
tree | commitdiff |
2010-06-05 |
Dan Gohman | getFoldedOffsetOf no longer does anything special with... |
tree | commitdiff |
2010-06-05 |
Stuart Hastings | Revert 105492 & 105493 due to a testcase regression... |
tree | commitdiff |
2010-06-05 |
Dan Gohman | LSR needs to remember inserted instructions even in... |
tree | commitdiff |
2010-06-05 |
Dale Johannesen | Fix some liveout handling related to tail calls, see... |
tree | commitdiff |
2010-06-04 |
Evan Cheng | Re-apply 105308 with fix. |
tree | commitdiff |
2010-06-04 |
Dan Gohman | Don't track users of undef values; they aren't interest... |
tree | commitdiff |
2010-06-04 |
Jim Grosbach | Make if-conversion ignore dbg_value instructions in... |
tree | commitdiff |
2010-06-04 |
Stuart Hastings | Support for nested functions/classes in debug output... |
tree | commitdiff |
2010-06-04 |
Devang Patel | Copy location info for current function argument from... |
tree | commitdiff |
2010-06-04 |
Jim Grosbach | Skip dbg_value instructions when scanning instructions... |
tree | commitdiff |
2010-06-04 |
Dan Gohman | Fix normalization and de-normalization of non-affine... |
tree | commitdiff |
2010-06-04 |
Jakob Stoklund Olesen | Keep track of the call instructions whose clobber lists... |
tree | commitdiff |
2010-06-04 |
Dale Johannesen | More thoroughly disable tails calls by default. |
tree | commitdiff |
2010-06-04 |
Jim Grosbach | Another fix to prevent debug info from affecting codege... |
tree | commitdiff |
2010-06-04 |
Jim Grosbach | more dbg_value adjustments so debug info doesn't affect... |
tree | commitdiff |
2010-06-04 |
Mon P Wang | Fixed a bug during widening where we would avoid legali... |
tree | commitdiff |
2010-06-04 |
Jim Grosbach | fix typo |
tree | commitdiff |
2010-06-04 |
Bob Wilson | For NEON vectors with 32- or 64-bit elements, select... |
tree | commitdiff |
2010-06-03 |
Bob Wilson | Add some missing checks in TwoAddressInstructionPass... |
tree | commitdiff |
2010-06-03 |
Jim Grosbach | Cleanup 80-column and trim trailing whitespace |
tree | commitdiff |
2010-06-03 |
Jim Grosbach | Teach the ARM load-store optimizer to deal with dbg_val... |
tree | commitdiff |
2010-06-03 |
Dale Johannesen | Early implementation of tail call for ARM. |
tree | commitdiff |
2010-06-03 |
Dan Gohman | Fix SimplifyDemandedBits' AssertZext logic to demand... |
tree | commitdiff |
2010-06-03 |
Bob Wilson | Revert 105308. |
tree | commitdiff |
2010-06-03 |
Bill Wendling | Machine sink could potentially sink instructions into... |
tree | commitdiff |
2010-06-03 |
Eric Christopher | Add first pass at darwin tls compiler support. |
tree | commitdiff |
2010-06-03 |
Eric Christopher | One underscore, not two. |
tree | commitdiff |
2010-06-03 |
Eli Friedman | Implement expansion in type legalization for add/sub... |
tree | commitdiff |
2010-06-03 |
Eli Friedman | Remove some already-fixed README entries. |
tree | commitdiff |
2010-06-03 |
Eli Friedman | Remove README entry which no longer compiles to somethi... |
tree | commitdiff |
2010-06-03 |
Eli Friedman | Remove a fixed item, update a couple partially-fixed... |
tree | commitdiff |
2010-06-03 |
Jakob Stoklund Olesen | Use the fast register allocator by default for -O0... |
tree | commitdiff |
2010-06-03 |
Jakob Stoklund Olesen | Use readsWritesVirtualRegister instead of counting... |
tree | commitdiff |
2010-06-02 |
Jakob Stoklund Olesen | Add full register <imp-def> operands when the coalescer... |
tree | commitdiff |
2010-06-02 |
Devang Patel | Speedup bitcode writer. Do not walk all values for... |
tree | commitdiff |
2010-06-02 |
Bill Wendling | Compulsive reformating. No functionalitical changes. |
tree | commitdiff |
2010-06-02 |
Jakob Stoklund Olesen | Slightly change the meaning of the reMaterialize target... |
tree | commitdiff |
2010-06-02 |
Jim Grosbach | Clean up 80 column violations. No functional change. |
tree | commitdiff |
2010-06-02 |
Rafael Espindola | Remove the TargetRegisterClass member from CalleeSavedInfo |
tree | commitdiff |
2010-06-02 |
Eli Friedman | Fix comment so it doesn't include comments which are... |
tree | commitdiff |
2010-06-02 |
Dan Gohman | Use comments to document non-obvious code rather than |
tree | commitdiff |
2010-06-02 |
Devang Patel | Skip identical instruction while calculating DBG_VALUE... |
tree | commitdiff |
2010-06-02 |
Bob Wilson | Rename canCombinedSubRegIndex method to something more... |
tree | commitdiff |
2010-06-02 |
Rafael Espindola | Replace ARM's getCalleeSavedRegClasses with a simpler... |
tree | commitdiff |
2010-06-02 |
Devang Patel | Use local small vector. |
tree | commitdiff |
2010-06-02 |
Rafael Espindola | Remove unused function. |
tree | commitdiff |
2010-06-02 |
Jim Grosbach | Not all entries in the range will have an SUnit. Check... |
tree | commitdiff |
2010-06-02 |
Rafael Espindola | cleanup |
tree | commitdiff |
2010-06-02 |
Rafael Espindola | Remove uses of getCalleeSavedRegClasses from outside the |
tree | commitdiff |
2010-06-02 |
Evan Cheng | Enable machine cse of instructions which define physica... |
tree | commitdiff |
2010-06-02 |
Eli Friedman | Don't try to custom-lower 64-bit add-with-overflow... |
tree | commitdiff |
2010-06-02 |
Bob Wilson | Fix an obvious mistake: don't change the operands until... |
tree | commitdiff |
2010-06-02 |
Eli Friedman | Remove outdated README entries. |
tree | commitdiff |
2010-06-01 |
Jim Grosbach | Update debug information when breaking anti-dependencie... |
tree | commitdiff |
2010-06-01 |
Jakob Stoklund Olesen | Properly compose subregister indices when coalescing. |
tree | commitdiff |
2010-06-01 |
Jim Grosbach | Remove unused code |
tree | commitdiff |
2010-06-01 |
Devang Patel | Ignore line number of debug value in undefined register. |
tree | commitdiff |
2010-06-01 |
Jim Grosbach | fix think-o |
tree | commitdiff |
2010-06-01 |
Dan Gohman | Fix the allocation of shadow space for the Win64 callin... |
tree | commitdiff |
2010-06-01 |
Jim Grosbach | Simplify things a bit more. Fix prototype to use SmallV... |
tree | commitdiff |
2010-06-01 |
Dan Gohman | Fix the noalias checking so that it doesn't worry about |
tree | commitdiff |
2010-06-01 |
Devang Patel | Keep track of incoming debug value of unused argument. |
tree | commitdiff |
2010-06-01 |
Dan Gohman | Fill in missing support for ISD::FEXP, ISD::FPOWI,... |
tree | commitdiff |
2010-06-01 |
Jim Grosbach | Add a FIXME |
tree | commitdiff |
2010-06-01 |
Jim Grosbach | mirror of r105280 changes for LowerInvoke, which uses... |
tree | commitdiff |
2010-06-01 |
Jim Grosbach | When processing function arguments when splitting live... |
tree | commitdiff |
2010-06-01 |
Jim Grosbach | Use SmallVector instead of std::vector. |
tree | commitdiff |
2010-06-01 |
Bruno Cardoso Lopes | Refactor some SSE 2 unpack instructions |
tree | commitdiff |
2010-06-01 |
Dan Gohman | Don't call flush() at a library level which isn't check... |
tree | commitdiff |
2010-06-01 |
Kalle Raiskila | Fix handling of 'load' nodes. |
tree | commitdiff |
2010-05-31 |
Duncan Sands | Fix PR7272: when inlining through a callsite with byval... |
tree | commitdiff |
2010-05-31 |
Chris Lattner | fix PR6623: when optimizing for size, don't inline... |
tree | commitdiff |
next |