| 2008-10-22 |
Duncan Sands | Get this working with LegalizeTypes: (1) don't |
tree | commitdiff |
| 2008-10-22 |
Oscar Fuentes | CMake: Turned some libraries into partially linked... |
tree | commitdiff |
| 2008-10-22 |
Dale Johannesen | Adjust comments for pedantic satisfaction. |
tree | commitdiff |
| 2008-10-21 |
Dale Johannesen | Add comments to explain uint64->f64 algorithm, |
tree | commitdiff |
| 2008-10-21 |
Dale Johannesen | Add an SSE2 algorithm for uint64->f64 conversion. |
tree | commitdiff |
| 2008-10-21 |
Dan Gohman | Implement the optimized FCMP_OEQ/FCMP_UNE code for... |
tree | commitdiff |
| 2008-10-21 |
Dan Gohman | Don't create TargetGlobalAddress nodes with offsets... |
tree | commitdiff |
| 2008-10-21 |
Dan Gohman | Optimized FCMP_OEQ and FCMP_UNE for x86. |
tree | commitdiff |
| 2008-10-21 |
Dan Gohman | When the coalescer is doing rematerializing, have it... |
tree | commitdiff |
| 2008-10-20 |
Duncan Sands | Have X86 custom lowering for LegalizeTypes use |
tree | commitdiff |
| 2008-10-18 |
Dan Gohman | Teach DAGCombine to fold constant offsets into GlobalAd... |
tree | commitdiff |
| 2008-10-17 |
Dan Gohman | This is done. |
tree | commitdiff |
| 2008-10-17 |
Evan Cheng | Add implicit defs of XMM8 to XMM15 on 32-bit call instr... |
tree | commitdiff |
| 2008-10-17 |
Chris Lattner | add support for 128 bit inputs on both x86-64 and x86-32. |
tree | commitdiff |
| 2008-10-17 |
Chris Lattner | Fix a bug where the x86 backend would reject 64-bit... |
tree | commitdiff |
| 2008-10-17 |
Evan Cheng | Fix lfence and mfence encoding. These look like MRM5r... |
tree | commitdiff |
| 2008-10-17 |
Evan Cheng | getX86RegNum has long been moved to X86RegisterInfo. |
tree | commitdiff |
| 2008-10-17 |
Dan Gohman | Fun x86 encoding tricks: when adding an immediate value... |
tree | commitdiff |
| 2008-10-17 |
Dan Gohman | Define patterns for shld and shrd that match immediate |
tree | commitdiff |
| 2008-10-16 |
Dan Gohman | Trim #includes. |
tree | commitdiff |
| 2008-10-16 |
Chris Lattner | fix typo noticed by sdt |
tree | commitdiff |
| 2008-10-16 |
Chris Lattner | add some notes |
tree | commitdiff |
| 2008-10-16 |
Chris Lattner | add some notes and a file to collect unimplemented... |
tree | commitdiff |
| 2008-10-16 |
Dan Gohman | Const-ify several TargetInstrInfo methods. |
tree | commitdiff |
| 2008-10-16 |
Dan Gohman | Remove an unused variable. |
tree | commitdiff |
| 2008-10-16 |
Dan Gohman | Fix the predicate for memop64 to be a regular load... |
tree | commitdiff |
| 2008-10-15 |
Dan Gohman | Now that predicates can be composed, simplify several of |
tree | commitdiff |
| 2008-10-15 |
Chris Lattner | add support for folding immediates into stores when... |
tree | commitdiff |
| 2008-10-15 |
Chris Lattner | fold immediates into stores in simple cases, this produ... |
tree | commitdiff |
| 2008-10-15 |
Chris Lattner | fold compare of null pointer into compare with 0. |
tree | commitdiff |
| 2008-10-15 |
Chris Lattner | Some minor cleanups: |
tree | commitdiff |
| 2008-10-15 |
Chris Lattner | Use switch on VT instead of Type* comparisons. |
tree | commitdiff |
| 2008-10-15 |
Chris Lattner | Use X86FastEmitCompare for FCMP_OEQ and FCMP_UNE: it... |
tree | commitdiff |
| 2008-10-15 |
Chris Lattner | refactor compare emission out into a new X86FastEmitCom... |
tree | commitdiff |
| 2008-10-15 |
Chris Lattner | Fold immediates into compares when possible, producing... |
tree | commitdiff |
| 2008-10-15 |
Chris Lattner | more minor refactoring of X86SelectBranch, no functiona... |
tree | commitdiff |
| 2008-10-15 |
Chris Lattner | factor buildmi calls in X86SelectBranch |
tree | commitdiff |
| 2008-10-15 |
Chris Lattner | factor some more BuildMI's in X86SelectCmp |
tree | commitdiff |
| 2008-10-15 |
Chris Lattner | factor some BuildMI calls, no functionality change. |
tree | commitdiff |
| 2008-10-15 |
Evan Cheng | - Add target lowering hooks that specify which setcc... |
tree | commitdiff |
| 2008-10-14 |
Dan Gohman | FastISel support for exception-handling constructs. |
tree | commitdiff |
| 2008-10-14 |
Dale Johannesen | Accept -march=i586, because gcc does (a synonym |
tree | commitdiff |
| 2008-10-14 |
Evan Cheng | Rename LoadX to LoadExt. |
tree | commitdiff |
| 2008-10-14 |
Dan Gohman | Fix command-line option printing to print two spaces... |
tree | commitdiff |
| 2008-10-14 |
Evan Cheng | Fix indentation. |
tree | commitdiff |
| 2008-10-13 |
Dan Gohman | When doing the very-late shift-and address-mode optimiz... |
tree | commitdiff |
| 2008-10-12 |
Anton Korobeynikov | Update size of inst correctly with segment override. |
tree | commitdiff |
| 2008-10-11 |
Chris Lattner | Change CALLSEQ_BEGIN and CALLSEQ_END to take TargetCons... |
tree | commitdiff |
| 2008-10-11 |
Duncan Sands | Fix comment typo. |
tree | commitdiff |
| 2008-10-11 |
Anton Korobeynikov | Add ability to override segment (mostly for code emitte... |
tree | commitdiff |
| 2008-10-10 |
Dale Johannesen | Fix SSE4.1 roundss, roundsd. While the instructions... |
tree | commitdiff |
| 2008-10-09 |
Dale Johannesen | Add a "loses information" return value to APFloat:... |
tree | commitdiff |
| 2008-10-07 |
Dan Gohman | Add MBB successors and physreg Uses in the same order... |
tree | commitdiff |
| 2008-10-07 |
Dan Gohman | Instead of emitting an implicit use for the super-regis... |
tree | commitdiff |
| 2008-10-07 |
Dale Johannesen | Model hardwired inputs & outputs of x86 8-bit divides... |
tree | commitdiff |
| 2008-10-07 |
Anders Carlsson | Certain patterns involving the "movss" instruction... |
tree | commitdiff |
| 2008-10-06 |
Evan Cheng | Fix PR2850 and PR2863. Only generate movddup for 128... |
tree | commitdiff |
| 2008-10-06 |
Devang Patel | It is possible that all functions in one module are... |
tree | commitdiff |
| 2008-10-06 |
Devang Patel | Remove unncessary isDeclaration() checks. |
tree | commitdiff |
| 2008-10-05 |
Anton Korobeynikov | Fix weird think-o and unbreak build on all gcc-3.4... |
tree | commitdiff |
| 2008-10-04 |
Anton Korobeynikov | Revert r56675 - it breaks unwinding runtime everywhere. |
tree | commitdiff |
| 2008-10-03 |
Dale Johannesen | Make atomic Swap work, 64-bit on x86-32. |
tree | commitdiff |
| 2008-10-03 |
Dale Johannesen | Pass MemOperand through for 64-bit atomics on 32-bit, |
tree | commitdiff |
| 2008-10-03 |
Dan Gohman | Avoid creating two TargetLowering objects for each... |
tree | commitdiff |
| 2008-10-03 |
Dan Gohman | Remove an unused field. |
tree | commitdiff |
| 2008-10-03 |
Dan Gohman | Switch the MachineOperand accessors back to the short... |
tree | commitdiff |
| 2008-10-03 |
Dan Gohman | Fix X86FastISel to handle dynamic allocas that have... |
tree | commitdiff |
| 2008-10-02 |
Dan Gohman | Optimize conditional branches in X86FastISel. This... |
tree | commitdiff |
| 2008-10-02 |
Dale Johannesen | Handle some 64-bit atomics on x86-32, some of the time. |
tree | commitdiff |
| 2008-10-02 |
Dan Gohman | Work around an interaction between fast-isel and regall... |
tree | commitdiff |
| 2008-10-02 |
Bill Wendling | "The original bug was a complaint that _mm_srli_si128... |
tree | commitdiff |
| 2008-10-01 |
Devang Patel | Remove OptimizeForSize global. Use function attribute... |
tree | commitdiff |
| 2008-10-01 |
Dan Gohman | Split x86's ADJCALLSTACK instructions into 32-bit and... |
tree | commitdiff |
| 2008-10-01 |
Dan Gohman | Mark CALL instructions as having a Use of ESP/RSP. |
tree | commitdiff |
| 2008-10-01 |
Bill Wendling | Implement the -fno-builtin option in the front-end... |
tree | commitdiff |
| 2008-09-30 |
Bill Wendling | Just don't transform this memset into "bzero" if no... |
tree | commitdiff |
| 2008-09-30 |
Bill Wendling | Add the new `-no-builtin' flag. This flag is meant... |
tree | commitdiff |
| 2008-09-30 |
Dan Gohman | Optimize SelectionDAG's AssignTopologicalOrder even... |
tree | commitdiff |
| 2008-09-30 |
Dan Gohman | Fix X86FastISel's output for x86-32 PIC constant pool... |
tree | commitdiff |
| 2008-09-30 |
Dan Gohman | Move the GlobalBaseReg field out of X86ISelDAGToDAG.cpp |
tree | commitdiff |
| 2008-09-30 |
Dan Gohman | Disable all x87 usage, including f32 and f64 when the... |
tree | commitdiff |
| 2008-09-29 |
Dale Johannesen | Remove misuse of ReplaceNodeResults for atomics with |
tree | commitdiff |
| 2008-09-29 |
Dan Gohman | Fix an over-pessimization about GlobalVariable addresse... |
tree | commitdiff |
| 2008-09-29 |
Evan Cheng | Fix PR2835. Do not change the width of a volatile load. |
tree | commitdiff |
| 2008-09-29 |
Duncan Sands | Rename isWeakForLinker to mayBeOverridden. Use it |
tree | commitdiff |
| 2008-09-27 |
Evan Cheng | Re-apply 56683 with fixes. |
tree | commitdiff |
| 2008-09-27 |
Daniel Dunbar | Unbreak build. |
tree | commitdiff |
| 2008-09-26 |
Devang Patel | Implement function notes as function attributes. |
tree | commitdiff |
| 2008-09-26 |
Evan Cheng | Implement "punpckldq %xmm0, $xmm0" as "pshufd $0x50... |
tree | commitdiff |
| 2008-09-26 |
Bill Wendling | Temporarily reverting r56683. This is causing a failure... |
tree | commitdiff |
| 2008-09-26 |
Dan Gohman | Rename ConstantSDNode's getSignExtended to getSExtValue... |
tree | commitdiff |
| 2008-09-26 |
Dan Gohman | Mark lea fi# as being really rematerializable. |
tree | commitdiff |
| 2008-09-26 |
Evan Cheng | unpckhps requires sse1, punpckhdq requires sse2. |
tree | commitdiff |
| 2008-09-26 |
Dan Gohman | Fix X86FastISel's address folding to check displacement |
tree | commitdiff |
| 2008-09-26 |
Evan Cheng | Fix @llvm.frameaddress codegen. FP elimination optimiza... |
tree | commitdiff |
| 2008-09-26 |
Dale Johannesen | Add "inreg" field to CallSDNode (doesn't increase |
tree | commitdiff |
| 2008-09-26 |
Dan Gohman | Factor out the code for determining when symblic addresses |
tree | commitdiff |
| 2008-09-26 |
Evan Cheng | Avoid spilling EBP / RBP twice in the prologue. |
tree | commitdiff |
| 2008-09-26 |
Evan Cheng | X86 address displacement field must be interpreted... |
tree | commitdiff |
| 2008-09-26 |
Oscar Fuentes | CMake: Builds all targets. |
tree | commitdiff |
| next |
projects / oota-llvm.git / history