aboutsummaryrefslogtreecommitdiff
path: root/lib/Target/X86/X86ISelPattern.cpp
AgeCommit message (Expand)Author
2005-04-13Fix some mysteriously missing {}'s which cause the miscompilation ofChris Lattner
2005-04-13Z_E_I is goneChris Lattner
2005-04-09Use live out sets for return values instead of imp_defs, which is cleaner and...Chris Lattner
2005-04-09This target does not support/want ISD::BRCONDTWOWAYChris Lattner
2005-04-07X86 zero extends setcc resultsChris Lattner
2005-04-06Add (untested) support for MULHS and MULHU.Chris Lattner
2005-04-02add support for FABS and FNEGChris Lattner
2005-04-02This target doesn't support fabs/fneg yet.Chris Lattner
2005-04-02Add support for 64-bit shifts.Chris Lattner
2005-04-01Add support for ISD::UNDEF to the X86 beChris Lattner
2005-03-30don't depend on the cfg being set up yetChris Lattner
2005-03-26Change interface to LowerCallTo to take a boolean isVarArg argument.Nate Begeman
2005-03-24Remove comments that are now meaningless from the pattern ISels, at Chris'sNate Begeman
2005-03-17Don't emit two comparisons when comparing a FP value against zero!Chris Lattner
2005-03-15This mega patch converts us from using Function::a{iterator|begin|end} toChris Lattner
2005-02-23Silence some uninit variable warnings.Chris Lattner
2005-01-25We can fold promoted and non-promoted loads into divs also!Chris Lattner
2005-01-25Fold promoted loads into binary ops for FP, allowing us to generate m32 formsChris Lattner
2005-01-23Silence a warning.Chris Lattner
2005-01-21Speed up folding operations into loads.Chris Lattner
2005-01-21The ever-important vanity pass name :)Chris Lattner
2005-01-21Fix a FIXME: realize that argument stores are all independent (don't alias)Chris Lattner
2005-01-20Implement ADD_PARTS/SUB_PARTS so that 64-bit integer add/sub work. ThisChris Lattner
2005-01-20Fix a crash compiling 134.perl.Chris Lattner
2005-01-19Fix a problem where were were literally selecting for INCREASED registerChris Lattner
2005-01-19Implement Regression/CodeGen/X86/rotate.ll: emit rotate instructions (whichChris Lattner
2005-01-19Match 16-bit shld/shrd instructions as well, implementing shift-double.llx:test5Chris Lattner
2005-01-19Codegen long >> 2 to this:Chris Lattner
2005-01-19X86 shifts mask the amount.Chris Lattner
2005-01-18Code to handle FP_EXTEND is dead now. X86 doesn't support any data types toChris Lattner
2005-01-18Remove more dead code.Chris Lattner
2005-01-18The selection dag code handles the promotions from F32 to F64 for us, so weChris Lattner
2005-01-18Fix 124.m88ksim.Chris Lattner
2005-01-18Do not emit loads multiple times, potentially in the wrong places.Chris Lattner
2005-01-18Eliminate bad assertions.Chris Lattner
2005-01-18* Eliminate the TokenSet and just use the ExprMap for both tokens and values.Chris Lattner
2005-01-18Rely on the code in MatchAddress to do this work. Otherwise we fail toChris Lattner
2005-01-18Fix a problem where probing for addressing modes caused expressions to beChris Lattner
2005-01-17Don't bother using max here.Chris Lattner
2005-01-17Do not give token factor nodes outrageous weightsChris Lattner
2005-01-17Two changes:Chris Lattner
2005-01-17Refactor load/op/store folding into it's own method, no functionality changes.Chris Lattner
2005-01-17Fix a major regression last night that prevented us from producing [mem] op= regChris Lattner
2005-01-17Codegen this:Chris Lattner
2005-01-17Fix test/Regression/CodeGen/X86/2005-01-17-CycleInDAG.ll and 132.ijpeg.Chris Lattner
2005-01-17Do not fold a load into a comparison that is used by more than one place.Chris Lattner
2005-01-17Do not codegen 'xor bool, true' as 'not reg'. not reg inverts the upper bitsChris Lattner
2005-01-17Set up the shift and setcc types.Chris Lattner
2005-01-16* Adjust to changes in TargetLowering interfaces.Chris Lattner
2005-01-15Add support for truncstore and *extload.Chris Lattner