Make x86-64 membarriers work without sse and clean up some of the
[oota-llvm.git] / lib / Target / X86 / X86ISelLowering.cpp
2010-08-04 Eric ChristopherMake x86-64 membarriers work without sse and clean...
2010-07-30 Bruno Cardoso LopesSupport all 128-bit AVX vector intrinsics. Most part...
2010-07-29 Jakob Stoklund OlesenRevert r109652, and remove the offending assert in...
2010-07-28 Jakob Stoklund OlesenCreate a fixed stack object for varargs that is as...
2010-07-28 Nate BegemanImplement a vectorized algorithm for <16 x i8> << ...
2010-07-27 Nate Begeman~40% faster vector shl <4 x i32> on SSE 4.1 Larger...
2010-07-26 Evan ChengOn x86, f32 / f64 nodes share the same registers as...
2010-07-24 Evan ChengAdd an ILP scheduler. This is a register pressure aware...
2010-07-23 Dale JohannesenThe only supported calling convention for X86-64 uses
2010-07-22 Eric ChristopherCustom lower the memory barrier instructions and add...
2010-07-22 Eric Christopher80-columns.
2010-07-21 Nate BegemanFix a couple issues with Win64 ABI
2010-07-21 Eric ChristopherPulling out previous patch, must've run the tests in
2010-07-21 Eric ChristopherLower MEMBARRIER on x86 and support processors without...
2010-07-15 Evan ChengSplit -enable-finite-only-fp-math to two options:
2010-07-14 Jakob Stoklund OlesenUse TargetOpcode::COPY instead of X86-native register...
2010-07-14 Evan ChengFix for PR7193 was overly conservative. The only case...
2010-07-10 Dan GohmanReapply bottom-up fast-isel, with several fixes for...
2010-07-10 Jakob Stoklund OlesenAn x86 function returns a floating point value in st...
2010-07-09 Bob Wilson--- Reverse-merging r107947 into '.':
2010-07-09 Dan GohmanFix the memoperand offsets in code generated for va_start.
2010-07-09 Dan GohmanRe-apply bottom-up fast-isel, with fixes. Be very caref...
2010-07-08 Chris LattnerChange LEA to have 5 operands for its memory operand...
2010-07-08 Chris Lattneradd some long-overdue enums to refer to the parts of...
2010-07-08 Dan GohmanRevert 107840 107839 107813 107804 107800 107797 107791.
2010-07-07 Evan ChengMove getExtLoad() and (some) getLoad() DebugLoc argumen...
2010-07-07 Dan GohmanAdd X86FastISel support for return statements. This...
2010-07-07 Dan GohmanSimplify FastISel's constructor by giving it a Function...
2010-07-07 Dan GohmanSplit the SDValue out of OutputArg so that SelectionDAG...
2010-07-06 Dale JohannesenAccept RIP-relative symbols with 'i' constraint, and
2010-07-06 Dan GohmanCanLowerReturn doesn't need a SelectionDAG; it just...
2010-07-06 Devang PatelPropagate debug loc.
2010-07-06 Dan GohmanReapply r107655 with fixes; insert the pseudo instructi...
2010-07-06 Dan GohmanRevert r107655.
2010-07-06 Dan GohmanFix a bunch of custom-inserter functions to handle...
2010-07-06 Eric ChristopherFix up -fstack-protector on linux to use the segment
2010-07-05 Eric ChristopherHave the X86 backend use Triple instead of a string...
2010-07-05 Chris Lattnermore tidying.
2010-07-04 Chris LattnerJust rip v2f32 support completely out of the X86 backen...
2010-07-04 Chris Lattnerfix PR7518 - terrible codegen of <2 x float>, by only...
2010-07-03 Evan ChengRemove isSS argument from CreateFixedObject. Fixed...
2010-06-30 Gabor Greifuse ArgOperand API
2010-06-29 Duncan SandsRemove pointless and unused variables.
2010-06-28 Bill WendlingReduce indentation via early exit. NFC.
2010-06-26 Gabor Greifuse ArgOperand API
2010-06-25 Dale JohannesenThe hasMemory argument is irrelevant to how the argument
2010-06-25 Bill Wendling- Reapply r106066 now that the bzip2 build regression...
2010-06-24 Dale JohannesenDisallow matching "i" constraint to symbol addresses...
2010-06-24 Dan GohmanReapply r106634, now that the bug it exposed is fixed.
2010-06-24 Dan GohmanFix a bug in the code which determines when it's safe...
2010-06-23 Daniel DunbarRevert r106263, "Fold the ShrinkDemandedOps pass into...
2010-06-23 Jim GrosbachThe generic DAG combiner can now fold atomic fences...
2010-06-23 Daniel DunbarRevert r106066, "Create a more targeted fix for not...
2010-06-22 Jim Grosbachfix typo
2010-06-20 Nick LewyckyFix warning in no-asserts build.
2010-06-18 Dan GohmanChange UpdateNodeOperands' operand and return value...
2010-06-18 Dan GohmanDelete unused variables.
2010-06-18 Dan GohmanEliminate unnecessary uses of getZExtValue().
2010-06-18 Dan GohmanisValueValidForType can be a static member function.
2010-06-18 Dan GohmanFold the ShrinkDemandedOps pass into the regular DAGCom...
2010-06-15 Bill WendlingCreate a more targeted fix for not sinking instructions...
2010-06-15 Eric ChristopherFor 32-bit non-pic tlv mach-o addressing we don't need...
2010-06-08 Eric ChristopherEnsure that mov and not lea are used to stick the addre...
2010-06-05 Dale JohannesenFix some liveout handling related to tail calls, see...
2010-06-03 Eric ChristopherAdd first pass at darwin tls compiler support.
2010-06-02 Eli FriedmanFix comment so it doesn't include comments which are...
2010-06-02 Dan GohmanUse comments to document non-obvious code rather than
2010-06-02 Eli FriedmanDon't try to custom-lower 64-bit add-with-overflow...
2010-05-29 Evan ChengFix PR7193: if sibling call address can take a register...
2010-05-28 Dale JohannesenFix comment typos.
2010-05-27 Dale JohannesenMark some math lib intrinsic nodes Legal on SSE4.1.
2010-05-27 Dan GohmanFastISel doesn't yet handle callee-pop functions.
2010-05-26 Zhongxing XuSRetReturnReg was set in LowerFormalArguments(). So...
2010-05-22 Evan ChengImplement @llvm.returnaddress. rdar://8015977.
2010-05-21 Dale JohannesenPrevious commit message should refer to 104308.
2010-05-21 Dale JohannesenFix two bugs in 104348:
2010-05-21 Dale JohannesenFix i64->f64 conversion, x86-64, -no-sse. A bit
2010-05-19 Evan ChengCode refactoring: pull SchedPreference enum from Target...
2010-05-16 Dale JohannesenRevert 103911; it broke a test that expects bitconvert
2010-05-16 Dale JohannesenMake x86-64 64-bit bitconvert work when SSE is not...
2010-05-16 Anton KorobeynikovAdd support for thiscall calling convention.
2010-05-15 Dale JohannesenFix uint64->{float, double} conversion to do rounding...
2010-05-14 Bill WendlingRename "HasCalls" in MachineFrameInfo to "AdjustsStack...
2010-05-14 Dan GohmanLowering of atomic instructions can result in operands...
2010-05-11 Dan GohmanImplement a bunch more TargetSelectionDAGInfo infrastru...
2010-05-01 Dan GohmanGet rid of the EdgeMapping map. Instead, just check...
2010-04-30 Dan GohmanMake this code less confusing. Instead of reassigning...
2010-04-30 Dan GohmanRemove the -disable-16bit command-line option, which...
2010-04-30 Evan ChengAnother sibcall bug. If caller and callee calling conve...
2010-04-28 Evan ChengEnable i16 to i32 promotion by default.
2010-04-28 Evan ChengUnbreak the build. Only form shld / shrd after legaliza...
2010-04-28 Evan ChengRather than having a ton of patterns for double shift...
2010-04-28 Stuart HastingsTweak x86 INC/DEC generation to look for CopyToReg...
2010-04-27 Evan ChengSRA promotion is also not free.
2010-04-26 Evan ChengPromoting 16-bit cmp / test aren't free. Don't do it.
2010-04-26 Evan Cheng- Move TargetLowering::EmitTargetCodeForFrameDebugValue...
2010-04-25 Dale JohannesenStop abusing EmitInstrWithCustomInserter for target...
2010-04-24 Evan ChengAvoid promoting a i16 node if it would eliminate a...
2010-04-23 Evan ChengFix X86ISD::CMP i16 to i32 promotion.
2010-04-22 Dan GohmanMove HandlePHINodesInSuccessorBlocks functions out...
next