summaryrefslogtreecommitdiffstats
path: root/llvm/lib/Target/X86/X86ISelLowering.cpp
Commit message (Expand)AuthorAgeFilesLines
* Fixed uninitialized stuff inside LegalizeDAG. Fortunately, the onlyAnton Korobeynikov2007-02-011-0/+4
* Finish off bug 680, allowing targets to custom lower frame and returnNate Begeman2007-01-291-19/+22
* Fix compile error "jump to case label crosses initialization".Nick Lewycky2007-01-281-5/+6
* Propagate changes from my local tree. This patch includes:Anton Korobeynikov2007-01-281-747/+349
* Make LABEL a builtin opcode.Jim Laskey2007-01-261-1/+1
* Linux GOT indirect reference is only necessary in PIC mode.Evan Cheng2007-01-221-1/+2
* * PIC codegen for X86/Linux has been implementedAnton Korobeynikov2007-01-121-37/+42
* - FCOPYSIGN custom lowering bug. Clear the sign bit of operand 0 first beforeEvan Cheng2007-01-051-14/+37
* TypoEvan Cheng2007-01-051-1/+1
* With SSE2, expand FCOPYSIGN to a series of SSE bitwise operations.Evan Cheng2007-01-051-4/+61
* Fix a comment that referred to the now defunct ubyte type.Reid Spencer2007-01-031-1/+1
* Really big cleanup.Anton Korobeynikov2007-01-031-3/+3
* For PR950:Reid Spencer2006-12-311-12/+20
* Refactored JIT codegen for mingw32. Now we're using standart relocationAnton Korobeynikov2006-12-221-21/+25
* f64 <-> i64 bit_convert using movq in 64-bit mode.Evan Cheng2006-12-141-4/+0
* Cleaned setjmp/longjmp lowering interfaces. Now we're producing rightAnton Korobeynikov2006-12-101-3/+12
* If we have ScalarSSE, we can select bitconvert into single instructions.Chris Lattner2006-12-051-2/+4
* Fix PR1033 and CodeGen/X86/bitcast.ll, by expanding bitcast to a load/store p...Chris Lattner2006-12-051-0/+5
* Fix typo noticed by Lauro Ramos Venancio, thanks!Chris Lattner2006-12-051-1/+1
* - Fix X86-64 JIT by temporarily disabling code that treats GV address as 32-bitEvan Cheng2006-12-051-34/+5
* Add support for 64-bit 'r' constraint. Patch by by Rafael ~Avila de Espín...Chris Lattner2006-12-041-0/+5
* - Use a different wrapper node for RIP-relative GV, etc.Evan Cheng2006-11-301-7/+36
* In PIC mode, GV not requiring an extra load can be used as address immediate.Evan Cheng2006-11-291-14/+7
* Fix for PR1018 - Better support for X86-64 Linux in small code model.Evan Cheng2006-11-291-13/+11
* Custom lower READCYCLECOUNTER for x86-64.Evan Cheng2006-11-291-6/+20
* Change MachineInstr ctor's to take a TargetInstrDescriptor reference insteadEvan Cheng2006-11-271-9/+13
* Refactored *GVRequiresExtraLoad() to Subtarget method.Anton Korobeynikov2006-11-211-115/+86
* We should perform extra load for dllimported callsAnton Korobeynikov2006-11-201-12/+26
* Matches MachineInstr changes.Evan Cheng2006-11-131-1/+1
* Add implicit use / def operands to created MI's.Evan Cheng2006-11-111-1/+1
* Don't dag combine floating point select to max and min intrinsics. ThoseEvan Cheng2006-11-101-27/+9
* Fixing PR990: http://llvm.org/PR990.Anton Korobeynikov2006-11-101-6/+6
* Fixed a bug which causes x86 be to incorrectly matchEvan Cheng2006-11-071-1/+17
* For PR786:Reid Spencer2006-11-021-7/+3
* allow the address of a global to be used with the "i" constraint when inChris Lattner2006-10-311-0/+30
* handle "st" as "st(0)"Chris Lattner2006-10-311-3/+12
* 1. Clean up code due to changes in SwitchTo*Section(2)Anton Korobeynikov2006-10-311-1/+3
* Add debug support for X86/ELF targets (Linux). This allows llvm-gcc4Reid Spencer2006-10-301-1/+1
* All targets expand BR_JT for now.Evan Cheng2006-10-301-0/+1
* Fixed a significant bug where unpcklpd is incorrectly used to extract element...Evan Cheng2006-10-271-0/+11
* Fix for PR968: expand vector sdiv, udiv, srem, urem.Evan Cheng2006-10-271-9/+16
* During vector shuffle lowering, we sometimes commute a vector shuffle to tryEvan Cheng2006-10-251-21/+21
* Remove -disable-x86-shuffle-optiEvan Cheng2006-10-251-61/+53
* Implement branch analysis/xform hooks required by the branch folding pass.Chris Lattner2006-10-201-65/+42
* Avoid getting into an infinite loop when -disable-x86-shuffle-opti is specified.Evan Cheng2006-10-161-8/+12
* Merge ISD::TRUNCSTORE to ISD::STORE. Switch to using StoreSDNode.Evan Cheng2006-10-131-39/+31
* Some X86ISD::CMP were created with wrong ValueType's.Evan Cheng2006-10-121-9/+9
* Don't convert to MOVLP if using shufps etc. may allow load folding.Evan Cheng2006-10-091-2/+6
* Reflects ISD::LOAD / ISD::LOADX / LoadSDNode changes.Evan Cheng2006-10-091-44/+33
* Make use of getStore().Evan Cheng2006-10-051-67/+62
OpenPOWER on IntegriCloud