Move isLoadFrom/StoreToStackSlot from MRegisterInfo to TargetInstrInfo,a far more...
[oota-llvm.git] / lib / Target /
2006-02-02 Chris LattnerMove isLoadFrom/StoreToStackSlot from MRegisterInfo...
2006-02-02 Chris Lattnerimplement isStoreToStackSlot
2006-02-02 Chris Lattneradd a method
2006-02-02 Chris Lattnermore notes
2006-02-02 Chris Lattneradd a note, I have no idea how important this is.
2006-02-02 Chris Lattner%fcc is not an alias for %fcc0
2006-02-02 Chris Lattnercorrect an opcode
2006-02-02 Chris Lattnernew example
2006-02-02 Nate BegemanUpdate the README
2006-02-02 Chris Lattnerimplemented, testcase here: test/Regression/CodeGen...
2006-02-02 Evan ChengUpdate.
2006-02-02 Evan ChengFix a erroneous comment.
2006-02-01 Chris Lattnermore notes
2006-02-01 Evan ChengTell codegen MOVAPSrr and MOVAPDrr are copies.
2006-02-01 Evan ChengAdded SSE entries to foldMemoryOperand().
2006-02-01 Evan ChengRearrange code to my liking. :)
2006-02-01 Chris Lattneradd a method
2006-02-01 Chris Lattneranother note
2006-02-01 Andrew LenharthAdd immediate forms of cmov and remove some cruft
2006-02-01 Chris LattnerFinegrainify namespacification
2006-02-01 Chris Lattneradd a note
2006-02-01 Nate BegemanFix some of the stuff in the PPC README file, and clean...
2006-02-01 Chris Lattneradd a note, I'll take care of this after nate commits...
2006-02-01 Evan Cheng- Use xor to clear integer registers (set R, 0).
2006-02-01 Evan ChengRemove another entry.
2006-02-01 Chris LattnerAnother regression from the pattern isel
2006-02-01 Evan ChengReturn's chain should be matching either the chain...
2006-02-01 Chris Lattneranother testcase.
2006-02-01 Evan ChengWhen folding a load into a return of SSE value, check...
2006-02-01 Evan ChengRemove an item. It's done.
2006-01-31 Evan ChengBe smarter about whether to store the SSE return value...
2006-01-31 Chris Lattnerturning these into 'adds' would require extra copies
2006-01-31 Evan Cheng- Allow XMM load (for scalar use) to be folded into...
2006-01-31 Evan ChengRemove entries on fabs and fneg. These are done.
2006-01-31 Evan ChengAllow the specification of explicit alignments for...
2006-01-31 Chris Lattner* Fix 80-column violations
2006-01-31 Chris Lattneradd info about the inline asm register constraints...
2006-01-31 Chris Lattneradd a missing break that caused a lot of failures last...
2006-01-31 Nate BegemanCodegen
2006-01-31 Chris Lattnerokay, one more
2006-01-31 Chris Lattneranother note
2006-01-31 Chris LattnerMore notes
2006-01-31 Chris Lattneranother one
2006-01-31 Chris Lattneradd a note
2006-01-31 Chris Lattneradd conditional moves of float and double values on...
2006-01-31 Chris Lattnerexample nate pointed out
2006-01-31 Chris Lattnertreat conditional branches the same way as conditional...
2006-01-31 Chris Lattnercompactify all of the integer conditional moves into...
2006-01-31 Chris LattnerAdd immediate forms of integer cmovs
2006-01-31 Chris LattnerShrinkify
2006-01-31 Chris LattnerAdd the full complement of conditional moves of integer...
2006-01-31 Chris LattnerCompile this:
2006-01-31 Evan ChengAdded custom lowering of fabs
2006-01-31 Chris Lattneradd the 'lucas' optimization
2006-01-31 Chris LattnerI don't see why this optimization isn't safe, but it...
2006-01-31 Chris LattnerAnother high-prio selection performance bug
2006-01-31 Chris Lattnermore mumbling
2006-01-31 Chris Lattneradd some notes
2006-01-30 Evan ChengDon't generate complex sequence for SETOLE, SETOLT...
2006-01-30 Chris LattnerFix FP constants, and the SparcV8/2006-01-22-BitConvert...
2006-01-30 Evan Chengi64 -> f32, f32 -> i64 and some clean up.
2006-01-30 Evan ChengAlways use FP stack instructions to perform i64 to...
2006-01-30 Chris LattnerRevamp the ICC/FCC reading instructions to be parameter...
2006-01-30 Chris LattnerCompile:
2006-01-30 Chris LattnerIf the target has V9 instructions, this pass is a noop...
2006-01-30 Chris LattnerWhen in v9 mode, emit fabsd/fnegd/fmovd
2006-01-30 Chris LattnerFirst step towards V9 instructions in the V8 backend...
2006-01-30 Chris LattnerTwo changes:
2006-01-30 Chris LattnerWhen lowering SELECT_CC, see if the input is a lowered...
2006-01-30 Chris LattnerMove MaskedValueIsZero from the DAGCombiner to the...
2006-01-30 Chris LattnerImplement isMaskedValueZeroForTargetNode for the variou...
2006-01-30 Chris Lattneradjust prototype
2006-01-29 Chris LattnerFunctions that are lazily streamed in from the .bc...
2006-01-29 Chris Lattneradd another note
2006-01-29 Chris Lattneradd some performance notes from looking at sgefa
2006-01-29 Chris Lattneradd a high-priority SSE issue from sgefa
2006-01-29 Chris Lattneradd a missed optimization
2006-01-29 Chris LattnerNow that OpActions is big enough, we can specify action...
2006-01-29 Chris Lattnerdisable this for now
2006-01-29 Reid SpencerAdd a note about lowering llvm.memset, llvm.memcpy...
2006-01-29 Chris Lattnerremove now-dead code, the legalizer takes care of this...
2006-01-29 Chris LattnerThe FP stack doesn't support UNDEF, ask the legalizer...
2006-01-29 Chris LattnerRequest expansion of ConstantVec nodes.
2006-01-29 Chris LattnerTargets all now request ConstantFP to be legalized...
2006-01-29 Chris LattnerUpdate alpha to reflect recent constantfp legalize...
2006-01-29 Chris Lattnercmovle != cmovlt
2006-01-29 Jeff CohenFix typo.
2006-01-28 Jeff CohenFlesh out AMD family/models.
2006-01-28 Jeff CohenCorrectly determine CPU vendor.
2006-01-28 Jeff CohenUse union instead of reinterpret_cast.
2006-01-28 Jeff CohenFix recognition of Intel CPUs.
2006-01-28 Chris LattnerIs64Bit reflects the capability of the chip, not an...
2006-01-28 Chris LattnerFix a bunch of JIT failures with the new isel
2006-01-28 Jeff CohenImprove X86 subtarget support for Windows and AMD.
2006-01-28 Chris Lattnersilence a warning
2006-01-28 Chris LattnerFix a bug in my elimination of ISD::CALL this morning...
2006-01-28 Chris Lattnermake this work on non-native hosts
2006-01-28 Chris Lattneradd a note about how we should implement this FIXME...
2006-01-28 Nate BegemanImplement Promote for VAARG, and allow it to be custom...
2006-01-28 Nate BegemanAdd a couple more things to the readme.
next