- 84b9a56 remove the ability to get memdep info for vaarg. I don't think the by Chris Lattner · 16 years ago
- aea5a2a make getDependencyFrom private. by Chris Lattner · 16 years ago
- 630c99f improve a note. by Chris Lattner · 16 years ago
- 6a09a74 some more PRE/GVN/DSE related notes. by Chris Lattner · 16 years ago
- f96ca79 add a note by Chris Lattner · 16 years ago
- 78a7e7c some random notes. by Chris Lattner · 16 years ago
- 4a31364 Minor cleanup. Use dyn_cast, not isa/cast pairs. No functionality change. by Nick Lewycky · 16 years ago
- c3ccc1a Clean up some ARM GV asm printing out; minor fixes to match what gcc does. by Evan Cheng · 16 years ago
- 5e600e6 Reimplement the inner loop of DSE. It now uniformly uses getDependence(), by Chris Lattner · 16 years ago
- dd17b25 Don't use plain %x to print pointer values. I had changed it from %p by Dan Gohman · 16 years ago
- 3b95481 Fix test to pass on Linux. by Dale Johannesen · 16 years ago
- 368eb2b Forgot a file. by Dale Johannesen · 16 years ago
- 203af58 Make LoopStrengthReduce smarter about hoisting things out of by Dale Johannesen · 16 years ago
- b51deb9 Make a few major changes to memdep and its clients: by Chris Lattner · 16 years ago
- 56d9b6d Demangle and pretty-print symbols in internal backtraces. Patch by by Dan Gohman · 16 years ago
- 64b5356 Revert invalid r60393. It causes llvm-gcc bootstrap fails in release builds. by Anton Korobeynikov · 16 years ago
- 9b052cf This test also requires -mattr=+sse41. by Evan Cheng · 16 years ago
- 6951381 Make it illegal to call getDependency* on non-memory instructions by Chris Lattner · 16 years ago
- 650d7f3 Reason #3 from 60595 doesn't hold true. If we can fold a PIC load from constpool into a use, the rewrite happens at time of spill (not in VirtRegMap). Later on, if the GlobalBaseReg is spilled, the spiller can see the use uses GlobalBaseReg and do the right thing. by Evan Cheng · 16 years ago
- 84100ba Effectively undo 60461 in PIC mode which simply transform V_SET0 / V_SETALLONES into a load from constpool in order to fold into restores. This is not safe to do when PIC base is being used for a number of reasons: by Evan Cheng · 16 years ago
- 7f7c736 Fix test/Transforms/GVN/pre-load.ll by Chris Lattner · 16 years ago
- 72eeb94 Fix comment. by Evan Cheng · 16 years ago
- 8f241b0 The use of the construct: by Cedric Venet · 16 years ago
- 72bc70d Make IsValueFullyAvailableInBlock safe. by Chris Lattner · 16 years ago
- 7f50863 add a new pop_back_val method which returns the value popped. This is by Chris Lattner · 16 years ago
- 2ce7f20 Drop the reg argument to isRegReDefinedByTwoAddr, which was redundant. by Dan Gohman · 16 years ago
- 4c0d951 Update comments. There is no getArgumentAccesses. by Dan Gohman · 16 years ago
- 9bf23b5 Teach StackSlotColoring to update MachineMemOperands when by Dan Gohman · 16 years ago
- 490b183 Ignore IMPLICIT_DEF instructions when computing physreg liveness. by Dan Gohman · 16 years ago
- ae94e59 Re-did 60519. It turns out Darwin's handling of hidden visibility symbols are a bit more complicate than I expected. Both declarations and weak definitions still need a stub indirection. However, the stubs are in data section and they contain the addresses of the actual symbols. by Evan Cheng · 16 years ago
- cd730fa CellSPU: Add new directory under tests/CodeGen/CellSPU to retain tests that by Scott Michel · 16 years ago
- d75ba1c Have raw_fd_ostream keep track of the position in the file to make tell() go faster by not requiring a flush(). by Ted Kremenek · 16 years ago
- 4d6e233 Enable LoopIndexSplit pass. by Devang Patel · 16 years ago
- 3831005 Rewrite code that 1) filters loops and 2) calculates new loop bounds. by Devang Patel · 16 years ago
- 6002e99 Factor out some common code. by Owen Anderson · 16 years ago
- cc18827 CellSPU: Fix bug 3055 by Scott Michel · 16 years ago
- 9304f2c When allocating a stack temporary, use the correct by Duncan Sands · 16 years ago
- 23f2ff7 Missing closing brace and reverse conditional condition on NDEBUG by Scott Michel · 16 years ago
- 3cda3cd Start simplifying a switch that has a successor that is a switch. by Chris Lattner · 16 years ago
- d021901 This code is apparently quite confused. In the meantime, by Chris Lattner · 16 years ago
- a8103da Temporarily revert r60519. It was causing a bootstrap failure: by Bill Wendling · 16 years ago
- 30ee7df CellSPU: by Scott Michel · 16 years ago
- 80e201b Use register names instead of numbers in debug output. by Dan Gohman · 16 years ago
- b29ffc8 Make debug output more informative. by Dan Gohman · 16 years ago
- eb83dfd Visibility hidden GVs do not require extra load of symbol address from the GOT or non-lazy-ptr. by Evan Cheng · 16 years ago
- 6a9041e Add minimal support for disambiguating memory references. Currently by Dan Gohman · 16 years ago
- 77ee977 add a debugging option to help track down j-t problems. by Chris Lattner · 16 years ago
- cef874a Rewrite the liveness bookkeeping code to fix a bunch of by Dan Gohman · 16 years ago
- 1bdd93a Make the debugging dump be a full line. by Dale Johannesen · 16 years ago
- 308f24d Remove an unused field. by Dale Johannesen · 16 years ago
- d71703d Have PseudoSourceValue override Value::dump, so that it works by Dan Gohman · 16 years ago
- 544e0d0 Fix a misspelled function name. by Dale Johannesen · 16 years ago
- 29874e0 Factor some code into a new FoldSingleEntryPHINodes method. by Chris Lattner · 16 years ago
- e96cc77 Fix an inconsistency in a comment. by Dan Gohman · 16 years ago
- 62fb4f2 Use mmx (punpckldq VR64, (mmx_v_set0)) to clear high 32-bits of a VR64 register. by Evan Cheng · 16 years ago
- e561751 Don't charge the full latency for anti and output dependencies. This is by Dan Gohman · 16 years ago
- f94a327 A step towards geting linux ppc to work (see PR 3099) by Dale Johannesen · 16 years ago
- 0dba0e5 When looking for anti-dependences on the critical path, don't bother by Dan Gohman · 16 years ago
- ebb0a31 Add a comment about callee-saved registers. by Dan Gohman · 16 years ago
- 67c7989 Fix a really wrong comment. by Dale Johannesen · 16 years ago
- 7c99f49 fix a really incorrect comment. by Chris Lattner · 16 years ago
- c54baa2 Split foldMemoryOperand into public non-virtual and protected virtual by Dan Gohman · 16 years ago
- 15511cf Rename isSimpleLoad to canFoldAsLoad, to better reflect its meaning. by Dan Gohman · 16 years ago
- 8cf7713 Extend X86's addFrameReference to add a MachineMemOperand for by Dan Gohman · 16 years ago
- 3eda88e Fix some tests. The grep for "il" was matching "file". by Rafael Espindola · 16 years ago
- f033b5a Update a comment. by Dan Gohman · 16 years ago
- 8b8e5a7 Only check that the result of the mapping was not by Duncan Sands · 16 years ago
- 952b839 Fix bug 3140. Print a single parameter .file directive if we have an ELF target. by Rafael Espindola · 16 years ago
- 269bc00 Add support for ISD::TRAP to the XCore backend by Richard Osborne · 16 years ago
- 370e534 Refactor code. No functionality change. by Evan Cheng · 16 years ago
- 0ea25cb CC should only be a ConstantSDNode at this point. Just use 'cast' instead of 'dyn_cast'. by Bill Wendling · 16 years ago
- c190651 Fix test. by Evan Cheng · 16 years ago
- 2a5dc33 testcase for br undef folding. by Chris Lattner · 16 years ago
- 421fa9e Teach jump threading some more simple tricks: by Chris Lattner · 16 years ago
- 2973a25 third time is the charm. by Chris Lattner · 16 years ago
- 15efa1a fix assertion. by Chris Lattner · 16 years ago
- 1ea7771 don't spew tons of stuff to the output. This testcase is *not* for by Chris Lattner · 16 years ago
- 71af9b0 Rename DeleteBlockIfDead to DeleteDeadBlock and make it by Chris Lattner · 16 years ago
- 2b1ba24 Factor some code out of SimplifyCFG, forming a new DeleteBlockIfDead method. by Chris Lattner · 16 years ago
- 62c939d Mark x86's V_SET0 and V_SETALLONES with isSimpleLoad, and teach X86's by Dan Gohman · 16 years ago
- b3cf03c Change label to 'carry' for unsigned adds. by Bill Wendling · 16 years ago
- 41474ba Add a sanity-check to tablegen to catch the case where isSimpleLoad by Dan Gohman · 16 years ago
- c70f329 Fix a missing #include. by Dan Gohman · 16 years ago
- 03e7d83 Add an explicit keyword. by Dan Gohman · 16 years ago
- 92faff2 Replace a #include with a forward-declaration. by Dan Gohman · 16 years ago
- c5a1a22 Fix this comment to reflect that it applies to types other than just i32. by Dan Gohman · 16 years ago
- e4300e2 Fix byval arguments in the fastcc calling convention. The fastcc convention by Dan Gohman · 16 years ago
- 7f9b352 Add nounwind attributes to this test. by Dan Gohman · 16 years ago
- 3cc1b8d testcases for recent dag combiner changes by Dale Johannesen · 16 years ago
- 1d93b2e Fix isIntN to work with APInts > 64 bits. This method is only by Chris Lattner · 16 years ago
- 96da041 Remove a (what appears to be) overly strict assertion. Here is what happened: by Evan Cheng · 16 years ago
- fb10cd4 Minor rewrite per review feedback. by Dale Johannesen · 16 years ago
- 546d7b5 Non-functional change: make custom lowering for truncate stylistically by Scott Michel · 16 years ago
- b30e8f6 CellSPU: by Scott Michel · 16 years ago
- d118342 Fix a typo in a comment. by Dan Gohman · 16 years ago
- 7b9d67c Add support for folding spills into preceding defs when doing pre-alloc splitting. by Owen Anderson · 16 years ago
- 34d7985 One more transformation. by Dale Johannesen · 16 years ago
- eccdd08 Make the code do what the comment says it does. by Dale Johannesen · 16 years ago
- 01c8e02 Comment typeo fix, thanks Duncan! by Chris Lattner · 16 years ago
- b0a5cdd make it possible to custom lower TRUNCATE (needed for the CellSPU target) by Tilmann Scheller · 16 years ago