Fix a really horrible problem that causes the JIT to miscompile any program
[oota-llvm.git] / lib /
2005-05-12 Chris LattnerFix a really horrible problem that causes the JIT to...
2005-05-12 Chris LattnerFix a problem where early legalization can cause token...
2005-05-12 Chris LattnerThese targets don't like setcc
2005-05-12 Chris LattnerMake legalize a bit more efficient, and canonicalize...
2005-05-11 Nate BegemanNecessary changes to codegen cttz efficiently on PowerPC
2005-05-11 Tanya LattnerFixed issue that broke ssa.
2005-05-11 Chris Lattnerfix some GCC 4 warnings
2005-05-11 John CriswellAdded support for decomposing constant expressions...
2005-05-11 Chris LattnerFix lowering of ctlz, so now UnitTests/2005-05-11-Popco...
2005-05-11 Chris LattnerFix lowering of cttz to work with signed values
2005-05-11 Chris Lattnerfix and concisify intinsic lowering for ctpop. Unfortu...
2005-05-11 Chris LattnerFix the last remaining bug preventing us from switching...
2005-05-11 Chris LattnerDo not memoize ADJCALLSTACKDOWN nodes, provide a method...
2005-05-11 Chris Lattnerwrap long line
2005-05-11 Chris LattnerMake sure to legalize generated ctpop nodes, convert...
2005-05-11 Duraid Madinaexpand count-leading/trailing-zeros; the test 2005...
2005-05-11 Chris LattnerAdd some notes for expanding clz/ctz
2005-05-11 Chris LattnerSimplify this code, use the proper shift amount
2005-05-11 Duraid Madinaadd the popcount instruction and support this in the...
2005-05-11 Chris LattnerLegalize this correctly
2005-05-11 Chris LattnerNo really IA*64* :)
2005-05-11 Chris LattnerX86 has more than just 32-bit registers
2005-05-11 Chris Lattnerimplement expansion of ctpop nodes, implementing CodeGe...
2005-05-11 Chris LattnerPrint bit count nodes correctly
2005-05-10 Chris LattnerDo not use "" as a sentinal for a missing argument...
2005-05-10 Misha BrukmanWhy output multiple strings, let the compiler concatena...
2005-05-10 Misha Brukman* Convert tabs to spaces, fix code alignment
2005-05-10 Chris LattnerConvert feature of the simple isel over for the pattern...
2005-05-10 Chris LattnerFix Reassociate/shifttest.ll
2005-05-10 Jeff CohenSilence some VC++ warnings
2005-05-09 Chris LattnerIf a function contains no allocas, all of the calls...
2005-05-09 Chris LattnerThe semantics of cast X to bool are a comparison agains...
2005-05-09 Chris LattnerImplement READPORT/WRITEPORT, implementing the last...
2005-05-09 Chris Lattnerdo not emit illegal instructions
2005-05-09 Chris LattnerFix the syntax of the i/o instructions, these are obvio...
2005-05-09 Chris Lattnerlegalize readio/writeio into load/stores, fixing CodeGe...
2005-05-09 Chris Lattnerlegalize readio/writeio into a load/store if requested
2005-05-09 Chris Lattnerlegalize READPORT, WRITEPORT, READIO, WRITEIO, at least...
2005-05-09 Chris LattnerAdd support for matching the READPORT, WRITEPORT, READI...
2005-05-09 Chris LattnerAdd support for READPORT, WRITEPORT, READIO, WRITEIO
2005-05-09 Chris Lattnerrestore some non-dead code I removed last night breakin...
2005-05-09 Chris Lattnerfold and (shl X, C1), C2 -> rlwinm when possible. ...
2005-05-09 Chris LattnerFold shifts into subsequent SHL's. These shifts often...
2005-05-09 Duraid Madinafix and cleanup constmul code a bit, this fixes mediabe...
2005-05-09 Chris LattnerWrap long lines, remove dead code that is now handled...
2005-05-09 Chris LattnerFix FP -> bool casts
2005-05-09 Chris Lattnerimplement and.ll:test33
2005-05-09 Chris LattnerDon't use the load/store instruction as the source...
2005-05-09 Chris Lattnermemoize all nodes, even null Value* nodes. Do not...
2005-05-09 Chris Lattnerwrap long lines
2005-05-09 Chris LattnerPrint SrcValue nodes correctly
2005-05-09 Chris LattnerFix X86/2005-05-08-FPStackifierPHI.ll: ugly gross hack.
2005-05-09 Chris LattnerPreserve CC's when linking modules
2005-05-09 Chris LattnerPreserve calling conventions when doing IPO
2005-05-09 Chris Lattnerwrap long lines, preserve calling conventions when...
2005-05-08 Chris LattnerBy definition, 'tail' calls cannot access the stack...
2005-05-08 Chris LattnerVerify that varargs functions all have ccc
2005-05-08 Chris LattnerConvert non-address taken functions with C calling...
2005-05-08 Chris LattnerImplement Reassociate/mul-neg-add.ll
2005-05-08 Chris LattnerBail out earlier
2005-05-08 Chris LattnerTeach reassociate that 0-X === X*-1
2005-05-08 Chris LattnerFix PR557 and basictest[34].ll.
2005-05-08 Chris LattnerAdd debugging information
2005-05-08 Chris Lattnereliminate gotos
2005-05-08 Chris LattnerWrap long lines. Fix "warning: conflicting types for...
2005-05-08 Chris LattnerImprove reassociation handling of inverses, implementin...
2005-05-08 Chris Lattnerclean up and modernize this pass.
2005-05-08 Chris LattnerStrength reduce SAR into SHR if there is no way sign...
2005-05-08 Chris LattnerRefactor some code
2005-05-08 Chris LattnerHandle some simple cases where we can see that values...
2005-05-07 Chris LattnerFix a miscompilation of crafty by clobbering the "A...
2005-05-07 Chris LattnerRewrite the guts of the reassociate pass to be more...
2005-05-07 Reid Spencer* Add two strlen optimizations:
2005-05-07 Reid SpencerDon't increment the counter unless the debug flag is...
2005-05-07 Chris LattnerConvert shifts to muls to assist reassociation. This...
2005-05-07 Chris LattnerSimplify the code and rearrange it. No major functiona...
2005-05-07 Jeff CohenSilence VC++ warnings about unsafe mixing of ints and...
2005-05-06 Chris Lattnerremove some dead (always dynamically false) flags
2005-05-06 Chris Lattnerencode calling conventions for call/invoke instructions.
2005-05-06 Chris Lattnerencode function calling convs in the bytecode file...
2005-05-06 Chris Lattnerparse new calling conv specifiers
2005-05-06 Chris Lattnerwrap a longline
2005-05-06 Chris Lattneradd support for explicit calling conventions
2005-05-06 Chris Lattneruse splice instead of remove/insert for a minor speedup
2005-05-06 Chris Lattnerremove some ugly hacks that are no longer needed since...
2005-05-06 Chris LattnerBAD typeo which caused many testsuite failures last...
2005-05-06 Chris Lattnerclean up the CBE output a bit
2005-05-06 Chris Lattneradd tail marker as a comment
2005-05-06 Chris LattnerMake the stub functions be tail calls
2005-05-06 Chris LattnerPreserve tail marker
2005-05-06 Chris LattnerImplement Transforms/Inline/inline-tail.ll
2005-05-06 Chris Lattnerpreserve the tail marker
2005-05-06 Chris Lattnerlex tail
2005-05-06 Chris Lattneradd bytecode reader support for tail calls
2005-05-06 Chris LattnerAdd a 'tail' marker for call instructions, patch contri...
2005-05-06 Chris LattnerWrap long lines
2005-05-06 Chris LattnerDCE intrinsic instructions without side effects.
2005-05-06 Chris LattnerThese intrinsics do not access memory
2005-05-06 Chris LattnerTeach instcombine propagate zeroness through shl instru...
2005-05-06 Chris LattnerImplement shift.ll:test23. If we are shifting right...
next