oota-llvm.git
2005-10-23 Andrew Lenharthadd TargetExternalSymbol
2005-10-23 Jeff CohenAdd new files to Visual Studio.
2005-10-22 Andrew LenharthWell, the Constant matching pattern works. Can't say...
2005-10-22 Chris Lattnerdocument this as experimental
2005-10-22 Chris LattnerThis file is entirely ifdef'd out
2005-10-22 Chris LattnerBuildSDIV and BuildUDIV only work for i32/i64, but...
2005-10-22 Jim LaskeyAdd g3 back to the mix and reorder to irritate them...
2005-10-22 Jim LaskeySort the features and processor lists for the sake...
2005-10-21 Chris Lattner64-bit reg support should not be enabled by default...
2005-10-21 Chris Lattneradd a case missing from the dag combiner that exposed...
2005-10-21 Chris LattnerInstead of aborting if not a case we can handle special...
2005-10-21 Chris LattnerNew testcase that crashes the ppc backend
2005-10-21 Jim LaskeyPlugin new subtarget backend into the build.
2005-10-21 Jim LaskeyNew TableGen backends for subtarget information. Only...
2005-10-21 Chris Lattnersilence a release mode warning
2005-10-21 Chris Lattneradd missing prototype
2005-10-21 Chris LattnerMake the coallescer a bit smarter, allowing it to join...
2005-10-21 Chris LattnerFix LiveInterval::getOverlapingRanges to take things...
2005-10-21 Nate BegemanMatch rotate. This does actually match the rotates...
2005-10-21 Chris LattnerMy previous patch was too conservative. Reject FP...
2005-10-21 Nate BegemanDon't generate operations that aren't yet supported
2005-10-21 Nate BegemanKill some now-dead code.
2005-10-21 Nate BegemanFix a typo in the dag combiner, so that this can work...
2005-10-21 Andrew Lenharthbyte zap not immediate goodness
2005-10-21 Chris LattnerMake tblgen emit:
2005-10-21 Nate BegemanInvert the TargetLowering flag that controls divide...
2005-10-20 Andrew LenharthInst cleanup. As a bonus, operands are in the correct...
2005-10-20 Chris LattnerUse a literal to define ineg instead of immzero
2005-10-20 Chris LattnerFix a conditional so we don't access past the end of...
2005-10-20 Andrew Lenharthadded a few 1 operand form stuff. Seems to break regal...
2005-10-20 Andrew Lenharthadd cttz and ctpop
2005-10-20 Nate BegemanFix a couple bugs in the const div stuff where we'd...
2005-10-20 Chris Lattnerdon't use llabs with apparently VC++ doesn't have
2005-10-20 Chris LattnerFix order of eval problem from when I refactored this...
2005-10-20 Andrew LenharthSounds good, finish the intop conversion.
2005-10-20 Nate BegemanAdd some more patterns for i64 on ppc
2005-10-20 Chris Lattneradd a new method, play around with some code.
2005-10-20 Chris Lattneradd a new method
2005-10-20 Chris LattnerRefactor some code, pulling it out into a function...
2005-10-20 Chris LattnerDo NOT touch FP ops with LSR. This fixes a testcase...
2005-10-20 Chris LattnerAdd some pattern fragments to simplify the repetitive...
2005-10-20 Chris Lattnersimplify this a bit by using immediates
2005-10-20 Nate BegemanMove the target constant divide optimization up into...
2005-10-20 Nate BegemanEnable targets to say that integer divide is expensive...
2005-10-20 Andrew Lenharthforgot this one
2005-10-20 Andrew Lenharthret 0; works, not much else
2005-10-19 John CriswellThis fixes PR638:
2005-10-19 Jim LaskeyAdded InstrSchedClass to each of the PowerPC Instructions.
2005-10-19 John CriswellMoved to Regression/Codegen/Generic.
2005-10-19 John CriswellThis is a generic test for all code generators. It...
2005-10-19 Nate BegemanWrite patterns for the various shl and srl patterns...
2005-10-19 John CriswellForce i386 code generation for an i386 specific test.
2005-10-19 Jim LaskeyPush processor descriptions to the top of target and...
2005-10-19 Chris Lattneradd support for literal immediates in patterns to match...
2005-10-19 Chris Lattnernow that tblgen is smarter, use integers directly....
2005-10-19 Chris LattnerAdd basic support for integer constants in pattern...
2005-10-19 Chris LattnerFix some checking that was causing duraid to get a...
2005-10-19 Chris LattnerAdd support for patterns that have physical registers...
2005-10-19 Chris LattnerAsserting here is to violent
2005-10-19 Chris Lattnerteach ppc backend these are copies
2005-10-19 Chris LattnerMake this work with the internalize change
2005-10-19 Chris LattnerConvert these cases to patterns
2005-10-19 Chris LattnerNate wants to define 'Pat's which turn into instruction...
2005-10-19 Nate BegemanWoo, it kinda works. We now generate this atrociously...
2005-10-19 Chris LattnerTestcase that crashes llvmgcc in type layout
2005-10-19 Chris Lattnerapply some tblgen majik to simplify the X register...
2005-10-19 Nate BegemanTeach Legalize how to do something with EXTRACT_ELEMENT...
2005-10-19 Nate BegemanMake a new reg class for 64 bit regs that aliases the...
2005-10-18 Nate BegemanAdd the ability to lower return instructions to TargetL...
2005-10-18 Chris LattnerFix Generic/2005-10-18-ZeroSizeStackObject.ll by not...
2005-10-18 Chris Lattnernew testcase
2005-10-18 Chris Lattnerremove hack
2005-10-18 Chris LattnerThis was never updated for the project makefile changes
2005-10-18 Chris LattnerThis never got updated to reflect the project makefile...
2005-10-18 Jim LaskeySimple edits; remove unimplimented cases and clarify...
2005-10-18 Chris LattnerFix the JIT encoding of LWA, LD, STD, and STDU.
2005-10-18 Jim LaskeyChecking in first round of scheduling tablegen files...
2005-10-18 Chris Lattneradd a case
2005-10-18 Chris LattnerFix PR637
2005-10-18 Chris LattnerAdd an option to this pass. If it is set, we are allow...
2005-10-18 Chris Lattneradd an option to the internalize pass
2005-10-18 Chris LattnerFold (select C, load A, load B) -> load (select C,...
2005-10-18 Chris LattnerDuraid pointed out that it is impolite to emit PPC...
2005-10-18 Nate BegemanDo the right thing and enable 64 bit regs under the...
2005-10-18 Nate BegemanFirst bits of 64 bit PowerPC stuff, currently disabled...
2005-10-18 Nate BegemanImplement some feedback from Chris re: constant canonic...
2005-10-18 Nate BegemanLegalize BUILD_PAIR appropriately for upcoming 64 bit...
2005-10-17 John CriswellUse %s instead of hard coding the input filename.
2005-10-17 Nate Begemanfold fmul X, +2.0 -> fadd X, X;
2005-10-17 Chris LattnerMake this work for FP constantexprs
2005-10-17 Chris LattnerOops, X+0.0 isn't foldable, but X+-0.0 is.
2005-10-17 Chris Lattnerrelax this a bit, as we only support the default roundi...
2005-10-17 Chris LattnerApparently, people object to floating pointers. Picky...
2005-10-17 Jim LaskeyAs requested, a blurb on sub-targets.
2005-10-17 Chris Lattnerfix some grammar-o's I noticed
2005-10-17 Chris LattnerFeedback from Nate.
2005-10-17 Chris LattnerDocument -view-isel-dags and -view-sched-dags
2005-10-17 Chris LattnerAdd notes about MF.viewCFG() and friends.
2005-10-17 Chris Lattneradd a trivial fold
2005-10-16 Chris LattnerAdd a bunch of info about the isel autogenerator. ...
next