| Commit message (Expand) | Author | Age | Files | Lines |
* | Change LEA to have 5 operands for its memory operand, just | Chris Lattner | 2010-07-08 | 1 | -6/+7 |
* | Move getExtLoad() and (some) getLoad() DebugLoc argument after EVT argument f... | Evan Cheng | 2010-07-07 | 1 | -1/+1 |
* | Propagate debug loc. | Devang Patel | 2010-07-06 | 1 | -1/+2 |
* | When creating X86 MUL8 and DIV8 instructions, make sure we don't produce | Jakob Stoklund Olesen | 2010-06-26 | 1 | -37/+49 |
* | Change UpdateNodeOperands' operand and return value from SDValue to | Dan Gohman | 2010-06-18 | 1 | -3/+3 |
* | Don't maintain a set of deleted nodes; instead, use a HandleSDNode | Dan Gohman | 2010-06-18 | 1 | -55/+29 |
* | Add first pass at darwin tls compiler support. | Eric Christopher | 2010-06-03 | 1 | -1/+1 |
* | Rename X86 subregister indices to something shorter. | Jakob Stoklund Olesen | 2010-05-24 | 1 | -6/+6 |
* | Don't leave Base.FrameIndex uninitialized, so that it doesn't | Dan Gohman | 2010-04-29 | 1 | -34/+33 |
* | Enable i16 to i32 promotion by default. | Evan Cheng | 2010-04-28 | 1 | -0/+3 |
* | teach the x86 address matching stuff to handle | Chris Lattner | 2010-04-20 | 1 | -7/+22 |
* | Use const qualifiers with TargetLowering. This eliminates several | Dan Gohman | 2010-04-17 | 1 | -5/+4 |
* | Add const qualifiers to CodeGen's use of LLVM IR constructs. | Dan Gohman | 2010-04-15 | 1 | -4/+3 |
* | Delete unneeeded arguments. | Dan Gohman | 2010-04-14 | 1 | -5/+5 |
* | use DebugLoc default ctor instead of DebugLoc::getUnknownLoc() | Chris Lattner | 2010-04-02 | 1 | -1/+1 |
* | X86 address mode matching code MatchAddressRecursively does some aggressive h... | Evan Cheng | 2010-03-17 | 1 | -14/+60 |
* | Do not force indirect tailcall through fixed registers: eax, r11. Add support... | Evan Cheng | 2010-03-14 | 1 | -15/+22 |
* | add a comment. | Chris Lattner | 2010-03-04 | 1 | -0/+1 |
* | factor the 'sign extended from 8 bit' patterns better so | Chris Lattner | 2010-03-03 | 1 | -6/+6 |
* | merge two loops over all nodes in the graph into one. | Chris Lattner | 2010-03-02 | 1 | -57/+41 |
* | eliminate PreprocessForRMW now that isel handles it. | Chris Lattner | 2010-03-02 | 1 | -136/+27 |
* | Fix some issues in WalkChainUsers dealing with | Chris Lattner | 2010-03-02 | 1 | -45/+9 |
* | Sink InstructionSelect() out of each target into SDISel, and rename it | Chris Lattner | 2010-03-02 | 1 | -89/+17 |
* | remove a little hack I did for the old isel, not needed | Chris Lattner | 2010-03-01 | 1 | -16/+2 |
* | remove a terrible hack that disabled assertions from this file because of bui... | Chris Lattner | 2010-03-01 | 1 | -9/+0 |
* | no need to override IsLegalToFold, the base implementation | Chris Lattner | 2010-02-23 | 1 | -10/+0 |
* | fix and un-xfail X86/vec_ss_load_fold.ll | Chris Lattner | 2010-02-21 | 1 | -3/+2 |
* | rename SelectScalarSSELoad -> SelectScalarSSELoadXXX and rewrite | Chris Lattner | 2010-02-21 | 1 | -18/+28 |
* | rename and document some arguments so I don't have to keep | Chris Lattner | 2010-02-17 | 1 | -11/+18 |
* | fix rdar://7653908, a crash on a case where we would fold a load | Chris Lattner | 2010-02-16 | 1 | -9/+10 |
* | Split SelectionDAGISel::IsLegalAndProfitableToFold to | Evan Cheng | 2010-02-15 | 1 | -10/+26 |
* | Remove an assumption of default arguments. This is in anticipation of a | David Greene | 2010-02-15 | 1 | -2/+3 |
* | refactor the conditional jump instructions in the .td file to | Chris Lattner | 2010-02-11 | 1 | -2/+2 |
* | move target-independent opcodes out of TargetInstrInfo | Chris Lattner | 2010-02-09 | 1 | -1/+1 |
* | Re-implement the main strength-reduction portion of LoopStrengthReduction. | Dan Gohman | 2010-01-21 | 1 | -1/+1 |
* | When XDEBUG is enabled, check for SelectionDAG cycles at some key | David Greene | 2010-01-20 | 1 | -0/+1 |
* | Fix PR6019. A load has more than one use if it feeds a bitconvert that | David Greene | 2010-01-15 | 1 | -1/+4 |
* | Reapply the MOV64r0 patch, with a fix: MOV64r0 clobbers EFLAGS. | Dan Gohman | 2010-01-12 | 1 | -21/+4 |
* | Revert 93158. It's breaking quite a few x86_64 tests. | Evan Cheng | 2010-01-11 | 1 | -4/+21 |
* | Re-instate MOV64r0 and MOV16r0, with adjustments to work with the | Dan Gohman | 2010-01-11 | 1 | -21/+4 |
* | Change errs() to dbgs(). | David Greene | 2010-01-05 | 1 | -30/+30 |
* | Change SelectCode's argument from SDValue to SDNode *, to make it more | Dan Gohman | 2010-01-05 | 1 | -34/+33 |
* | Flags-producing add, and, or, etc. have the same profibility | Dan Gohman | 2010-01-04 | 1 | -0/+5 |
* | completely eliminate the MOV16r0 'instruction'. The only | Chris Lattner | 2009-12-23 | 1 | -7/+8 |
* | Re-apply 91623 now that I actually know what I was trying to do. | Evan Cheng | 2009-12-18 | 1 | -25/+1 |
* | Revert r91623 to unbreak the buildbots. | Jeffrey Yasskin | 2009-12-17 | 1 | -1/+4 |
* | Remove an unused option. | Evan Cheng | 2009-12-17 | 1 | -4/+1 |
* | Target-independent support for TargetFlags on BlockAddress operands, | Dan Gohman | 2009-11-20 | 1 | -4/+4 |
* | llvm-gcc/clang don't (won't?) need this hack. | Daniel Dunbar | 2009-11-11 | 1 | -1/+2 |
* | Add a monstrous hack to improve X86ISelDAGToDAG compile time. | Daniel Dunbar | 2009-11-10 | 1 | -0/+8 |