When optimzing a mul by immediate into two, the resulting mul's should get a x86...
[oota-llvm.git] / lib / Target / X86 / X86InstrInfo.td
2009-03-30 Evan ChengWhen optimzing a mul by immediate into two, the resulti...
2009-03-13 Rafael Espindolaadd 8 and 16 bit TLS moves.
2009-03-13 Rafael EspindolaImprove sext and zext of TLS variables.
2009-03-12 Evan ChengRe-apply 66024 with fixes: 1. Fixed indirect call to...
2009-03-11 Dan GohmanRevert r66024. The JIT encoding for CALLpcrel32 is...
2009-03-11 Rafael Espindolaoptimize i8 and i16 tls values.
2009-03-05 Dan GohmanDon't use plain INC32 and DEC32 on x86-64; it needs
2009-03-04 Dan GohmanRe-apply 66008, now that the unfoldMemoryOperand bug...
2009-03-04 Evan ChengFix PR3666: isel calls to constant addresses.
2009-03-04 Dan GohmanRevert r66004 for now; it's causing a variety of test...
2009-03-04 Dan GohmanTeach the x86 backend to eliminate "test" instructions...
2009-03-03 Dan GohmanAdd '(implicit EFLAGS)' for AND, OR, XOR, NEG, INC...
2009-02-05 Evan ChengA few more isAsCheapAsAMove.
2009-01-27 Evan ChengImplement multiple with overflow by 2 with an add instr...
2009-01-26 Nate BegemanMap address space 256 to gs; similar mappings could...
2009-01-21 Evan ChengFavors generating "not" over "xor -1". For example.
2009-01-13 Dan GohmanDisable the register+memory forms of the bt instruction...
2009-01-13 Dan GohmanAdd bt instructions that take immediate operands.
2009-01-13 Dan GohmanFix a few more JIT encoding issues in the BT instructions.
2009-01-07 Dan GohmanAdd patterns to match conditional moves with loads...
2009-01-07 Dan GohmanDefine instructions for cmovo and cmovno.
2009-01-07 Dan GohmanX86_COND_C and X86_COND_NC are alternate mnemonics for
2009-01-02 Evan ChengDo not isel load folding bt instructions for pentium...
2008-12-25 Chris LattnerFix some JIT encodings.
2008-12-25 Chris LattnerBT memory operands load from their address operand.
2008-12-23 Dan GohmanAdd instruction patterns and encodings for the x86...
2008-12-12 Bill Wendling- Use patterns instead of creating completely new instr...
2008-12-12 Bill WendlingRedo the arithmetic with overflow architecture. I was...
2008-12-09 Bill WendlingAdd sub/mul overflow intrinsics. This currently doesn...
2008-12-07 Nick LewyckyFix typo, psuedo -> pseudo.
2008-12-03 Dan GohmanRename isSimpleLoad to canFoldAsLoad, to better reflect...
2008-12-02 Bill WendlingReapply r60382. This time, don't mark "ADC" nodes with...
2008-12-01 Bill WendlingTemporarily revert r60382. It caused CodeGen/X86/i2k...
2008-12-01 Bill Wendling- Have "ADD" instructions return an implicit EFLAGS.
2008-11-26 Bill WendlingGenerate something sensible for an [SU]ADDO op when...
2008-11-18 Dan GohmanDon't set neverHasSideEffects on x86's divide instructi...
2008-10-25 Nicolas GeoffrayGenerate code for TLS instructions.
2008-10-17 Evan ChengAdd implicit defs of XMM8 to XMM15 on 32-bit call instr...
2008-10-17 Dan GohmanFun x86 encoding tricks: when adding an immediate value...
2008-10-17 Dan GohmanDefine patterns for shld and shrd that match immediate
2008-10-15 Dan GohmanNow that predicates can be composed, simplify several of
2008-10-11 Chris LattnerChange CALLSEQ_BEGIN and CALLSEQ_END to take TargetCons...
2008-10-07 Dale JohannesenModel hardwired inputs & outputs of x86 8-bit divides...
2008-10-03 Dale JohannesenMake atomic Swap work, 64-bit on x86-32.
2008-10-03 Dale JohannesenPass MemOperand through for 64-bit atomics on 32-bit,
2008-10-02 Dale JohannesenHandle some 64-bit atomics on x86-32, some of the time.
2008-10-01 Dan GohmanSplit x86's ADJCALLSTACK instructions into 32-bit and...
2008-10-01 Dan GohmanMark CALL instructions as having a Use of ESP/RSP.
2008-09-29 Evan ChengFix PR2835. Do not change the width of a volatile load.
2008-09-26 Evan ChengImplement "punpckldq %xmm0, $xmm0" as "pshufd $0x50...
2008-09-24 Evan ChengFix patterns for SSE4.1 move and sign extend instructio...
2008-09-16 Bill WendlingReverting r56249. On further investigation, this functi...
2008-09-16 Bill Wendling- Change "ExternalSymbolSDNode" to "SymbolSDNode".
2008-09-12 Dan GohmanRename ConstantSDNode::getValue to getZExtValue, for...
2008-08-30 Evan ChengTransform (x << (y&31)) -> (x << y). This takes advanta...
2008-08-28 Dale JohannesenSplit the ATOMIC NodeType's to include the size, e.g.
2008-08-22 Bill WendlingReverting r55190, r55191, and r55192. They broke the...
2008-08-22 Dan GohmanAnyext tweaks for x86. When extloading a value to i32...
2008-08-20 Dan GohmanMove the handling of ANY_EXTEND, SIGN_EXTEND_INREG...
2008-08-20 Dan GohmanTablegen generated code already tests the opcode value...
2008-08-20 Bill WendlingRevert r55018 and apply the correct "fix" for the 64...
2008-08-19 Bill WendlingAdd support for the __sync_sub_and_fetch atomics and...
2008-08-19 Dale JohannesenAdd support for 8 and 16 bit forms of __sync
2008-08-06 Dan GohmanRe-introduce the 8-bit subreg zext-inreg patterns for...
2008-08-06 Dan Gohmanxchg does not modify FLAGS.
2008-07-30 Dan GohmanReapply r54147 with a constraint to only use the 8-bit
2008-07-29 Dan GohmanRevert 54147.
2008-07-28 Dan GohmanAdd x86 isel patterns to match what would be a ZERO_EXT...
2008-07-22 Anton KorobeynikovFix encoding of atomic compare and swap for i64
2008-06-25 Mon P WangAdded MemOperands to Atomic operations since Atomics...
2008-06-18 Evan ChengXOR32rr, etc. are not AsCheapAsMove, but MOV32ri, etc...
2008-06-14 Andrew Lenharthadd missing atomic intrinsic from gcc
2008-05-31 Dan GohmanTeach the DAGISelEmitter to not compute the variable_op...
2008-05-29 Dan GohmanAdd patterns for CALL32m and CALL64m. They aren't match...
2008-05-29 Dan GohmanFix a tblgen problem handling variable_ops in tblgen...
2008-05-29 Bill WendlingXOR?RI instructions aren't as cheap as moves.
2008-05-29 Bill WendlingImplement "AsCheapAsAMove" for some obviously cheap...
2008-05-14 Evan ChengDoh. Alignment is in bytes, not in bits.
2008-05-13 Evan Cheng- Fix the pasto in the fix for a previous pasto.
2008-05-13 Evan Cheng- Don't treat anyext 16-bit load as a 32-bit load if...
2008-05-13 Evan ChengOn x86, it's safe to treat i32 load anyext as a normal...
2008-05-12 Dan GohmanFix a copy+paste bug; pseudo-instructions shouldn't...
2008-05-05 Mon P WangAdded addition atomic instrinsics and, or, xor, min...
2008-05-04 Anton KorobeynikovAdd General Dynamic TLS model for X86-64. Some parts...
2008-04-30 Arnold SchwaighoferTail call optimization improvements:
2008-04-25 Evan ChengFix MMX_MOVQ2DQrr pattern. It's illegal to do a bitconv...
2008-04-19 Evan Chengxchg which references a memory operand does not need...
2008-04-18 Evan Cheng- Fix atomic operation JIT encoding.
2008-04-17 Evan ChengAlso support Intel asm syntax.
2008-04-17 Evan ChengFix assembly code for atomic operations.
2008-04-12 Nate Begeman80 col fix
2008-03-27 Evan ChengAllow certain lea instructions to be rematerialized.
2008-03-19 Arnold SchwaighoferDon't loose incoming argument registers. Fix documentat...
2008-03-15 Evan ChengReplace all target specific implicit def instructions...
2008-03-13 Christopher LambGet rid of a pseudo instruction and replace it with...
2008-03-11 Christopher LambRecommitting parts of r48130. These do not appear to...
2008-03-11 Chris LattnerChange the model for FP Stack return to use fp operands...
2008-03-10 Evan ChengRevert 48125, 48126, and 48130 for now to unbreak some...
2008-03-10 Christopher LambAllow insert_subreg into implicit, target-specific...
2008-03-05 Andrew Lenharth64bit CAS on 32bit x86.
next