| 2008-07-12 |
Evan Cheng | Typos. |
tree | commitdiff |
| 2008-07-12 |
Evan Cheng | Fix PR2536: a nasty spiller bug. If a two-address instr... |
tree | commitdiff |
| 2008-07-12 |
Evan Cheng | Back out 53476 and 53480 for now. Somehow they cause... |
tree | commitdiff |
| 2008-07-11 |
Dan Gohman | Include a frame index in the "fixed stack" pseudo sourc... |
tree | commitdiff |
| 2008-07-11 |
Dan Gohman | Fix an obsolete top-level comment. |
tree | commitdiff |
| 2008-07-11 |
Dan Gohman | Factor out debugging code into the common base class. |
tree | commitdiff |
| 2008-07-11 |
Dan Gohman | Add support for putting NamedRegionTimers in TimerGroup... |
tree | commitdiff |
| 2008-07-11 |
Dan Gohman | Trim unnecessary #includes. |
tree | commitdiff |
| 2008-07-11 |
Duncan Sands | Remove an apparently useless routine: there should |
tree | commitdiff |
| 2008-07-11 |
Duncan Sands | It is pointless to turn a UINT_TO_FP into an |
tree | commitdiff |
| 2008-07-11 |
Duncan Sands | Add two missing SINT_TO_FP libcalls. |
tree | commitdiff |
| 2008-07-11 |
Duncan Sands | Port a shift-by-1 optimization from LegalizeDAG: it |
tree | commitdiff |
| 2008-07-11 |
Duncan Sands | Add support for 128 bit shifts and 32 bit shifts |
tree | commitdiff |
| 2008-07-10 |
Chris Lattner | Fix a bug in the soft-float handling of FCOPYSIGN that... |
tree | commitdiff |
| 2008-07-10 |
Chris Lattner | make legalize types be a command line option: -enable... |
tree | commitdiff |
| 2008-07-10 |
Dan Gohman | Make stack slot coloring's debug output more consistent... |
tree | commitdiff |
| 2008-07-10 |
Evan Cheng | Change StackSlotForVirtReg (which maps vregs to frame... |
tree | commitdiff |
| 2008-07-10 |
Duncan Sands | Add support for 128 bit multiplicative operations. |
tree | commitdiff |
| 2008-07-10 |
Duncan Sands | Add a mysteriously missing libcall, FPTOSINT_F80_I32. |
tree | commitdiff |
| 2008-07-10 |
Duncan Sands | Fix a FIXME: use an apint in CTTZ legalization. |
tree | commitdiff |
| 2008-07-10 |
Duncan Sands | Remove PromoteIntRes_FP_ROUND - not sure what it |
tree | commitdiff |
| 2008-07-10 |
Duncan Sands | Make sure the alignment of the temporary created |
tree | commitdiff |
| 2008-07-10 |
Duncan Sands | Make the LegalizeType method naming scheme more regular. |
tree | commitdiff |
| 2008-07-10 |
Duncan Sands | Don't barf when dumping a constant that contains |
tree | commitdiff |
| 2008-07-10 |
Evan Cheng | - Change the horrible N^2 isRegReDefinedByTwoAddr.... |
tree | commitdiff |
| 2008-07-10 |
Owen Anderson | Use DenseMap instead of std::map in local register... |
tree | commitdiff |
| 2008-07-10 |
Owen Anderson | Fix 403.gcc. Finally got the check for two-address... |
tree | commitdiff |
| 2008-07-09 |
Owen Anderson | Revert r53367, which was breaking things. |
tree | commitdiff |
| 2008-07-09 |
Dan Gohman | Simplify hasNUsesOfValue and hasAnyUsesOfValue even... |
tree | commitdiff |
| 2008-07-09 |
Dan Gohman | hasAnyUseOfValue can check SDUse nodes of its users... |
tree | commitdiff |
| 2008-07-09 |
Dan Gohman | Move MemoryVT out of LSBaseNode into MemSDNode, allowin... |
tree | commitdiff |
| 2008-07-09 |
Evan Cheng | Avoid creating expensive comment string if it's not... |
tree | commitdiff |
| 2008-07-09 |
Owen Anderson | Loosen our check here. Local regalloc only cares that... |
tree | commitdiff |
| 2008-07-09 |
Dan Gohman | Move the IsVolatile and SVOffset fields into the MemSDN... |
tree | commitdiff |
| 2008-07-09 |
Owen Anderson | Don't use an expensive check for two-address-ness when... |
tree | commitdiff |
| 2008-07-09 |
Owen Anderson | Factor local liveness computation out into its own... |
tree | commitdiff |
| 2008-07-09 |
Dan Gohman | Reuse the MO variable instead of recomputing it in... |
tree | commitdiff |
| 2008-07-09 |
Dan Gohman | Give RegAllocSimple a TargetInstrInfo member to keep... |
tree | commitdiff |
| 2008-07-09 |
Dan Gohman | RegAllocLocal has a TargetInstrInfo data member. Use... |
tree | commitdiff |
| 2008-07-09 |
Dan Gohman | Use find with std::map, when that's what's needed,... |
tree | commitdiff |
| 2008-07-09 |
Anton Korobeynikov | Switch to new section name handling facility |
tree | commitdiff |
| 2008-07-09 |
Duncan Sands | Remove some unneeded includes. |
tree | commitdiff |
| 2008-07-09 |
Duncan Sands | Redo LegalizeTypes soft float support for |
tree | commitdiff |
| 2008-07-09 |
Duncan Sands | Forgot to update the chain result when softening |
tree | commitdiff |
| 2008-07-09 |
Duncan Sands | LegalizeTypes soft float support for FP_TO_SINT and |
tree | commitdiff |
| 2008-07-09 |
Duncan Sands | LegalizeTypes support for powi soft float. |
tree | commitdiff |
| 2008-07-09 |
Duncan Sands | Make the role of MVT::i32 clearer here, and add a |
tree | commitdiff |
| 2008-07-09 |
Evan Cheng | Missed alignment argument on stores lowered from memcpy. |
tree | commitdiff |
| 2008-07-09 |
Bill Wendling | Make the DICountVisitor not a visitor. This keeps us... |
tree | commitdiff |
| 2008-07-09 |
Dan Gohman | const-ify SelectionDAG::getNodeValueTypes. |
tree | commitdiff |
| 2008-07-08 |
Dan Gohman | It's no longer necessary to test if a MachineBasicBlock's |
tree | commitdiff |
| 2008-07-08 |
Dan Gohman | Verify that MachineMemOperand alignment is a non-zero... |
tree | commitdiff |
| 2008-07-08 |
Dan Gohman | Factor out the code for computing an alignment value... |
tree | commitdiff |
| 2008-07-08 |
Owen Anderson | Fix the build. Apparently MachineInstr& is no longer... |
tree | commitdiff |
| 2008-07-08 |
Owen Anderson | Make the local register allocator compute (purely local... |
tree | commitdiff |
| 2008-07-08 |
Dale Johannesen | Remove some dead code. |
tree | commitdiff |
| 2008-07-08 |
Evan Cheng | Do not CSE DEBUG_LOC, DBG_LABEL, DBG_STOPPOINT, DECLARE... |
tree | commitdiff |
| 2008-07-08 |
Duncan Sands | Remove custom expansion from LegalizeTypes when doing |
tree | commitdiff |
| 2008-07-08 |
Duncan Sands | Add missing select_cc libcall line, somehow omitted |
tree | commitdiff |
| 2008-07-08 |
Evan Cheng | Unbreak C++ tests on x86 Darwin. |
tree | commitdiff |
| 2008-07-08 |
Duncan Sands | LegalizeTypes support for FP_ROUND and FP_EXTEND |
tree | commitdiff |
| 2008-07-08 |
Evan Cheng | Avoid unnecessary string construction during asm printing. |
tree | commitdiff |
| 2008-07-07 |
Dan Gohman | Pool-allocation for MachineInstrs, MachineBasicBlocks... |
tree | commitdiff |
| 2008-07-07 |
Dan Gohman | Pool-allocation for SDNodes. The pool is allocated... |
tree | commitdiff |
| 2008-07-07 |
Bill Wendling | Use the canonical way to get an empty structure. |
tree | commitdiff |
| 2008-07-07 |
Bill Wendling | Use StringMap for greater justice! |
tree | commitdiff |
| 2008-07-07 |
Dan Gohman | Fix SDNode::MorphNodeTo (a function used by by SelectNo... |
tree | commitdiff |
| 2008-07-07 |
Dan Gohman | Move MachineMemOperand's constructor out of line, to... |
tree | commitdiff |
| 2008-07-07 |
Dan Gohman | Use of operator* is redundant and confusing here. |
tree | commitdiff |
| 2008-07-07 |
Dan Gohman | Minor const-correctness fixes. |
tree | commitdiff |
| 2008-07-07 |
Dan Gohman | Assert that all MachineInstrs update PhysRegUseDefLists in |
tree | commitdiff |
| 2008-07-07 |
Dan Gohman | Remove most of the uses of SDOperandPtr, usually replac... |
tree | commitdiff |
| 2008-07-07 |
Dan Gohman | Add explicit keywords. |
tree | commitdiff |
| 2008-07-07 |
Dan Gohman | Make DenseMap's insert return a pair, to more closely... |
tree | commitdiff |
| 2008-07-07 |
Evan Cheng | LegalizeSetCCOperands should legalize the result of... |
tree | commitdiff |
| 2008-07-07 |
Bill Wendling | Prevent option name conflict. |
tree | commitdiff |
| 2008-07-07 |
Duncan Sands | LegalizeTypes soft-float support for stores of a |
tree | commitdiff |
| 2008-07-05 |
Mon P Wang | Fixed generating incorrect aligned stores that I backou... |
tree | commitdiff |
| 2008-07-04 |
Duncan Sands | Rather than having a different custom legalization |
tree | commitdiff |
| 2008-07-04 |
Duncan Sands | Linux also does not require exception handling |
tree | commitdiff |
| 2008-07-03 |
Bill Wendling | Don't return std::vector by value, but pass it in by... |
tree | commitdiff |
| 2008-07-03 |
Bill Wendling | Revert my previous check-in that split up MachineModule... |
tree | commitdiff |
| 2008-07-03 |
Evan Cheng | Backed out 53031. |
tree | commitdiff |
| 2008-07-03 |
Evan Cheng | - Remove calls to copyKillDeadInfo which is an N^2... |
tree | commitdiff |
| 2008-07-03 |
Dan Gohman | Reapply r52988, "Simplify addRegisterKilled and addRegi... |
tree | commitdiff |
| 2008-07-03 |
Dan Gohman | Avoid unnecessarily copying APInt objects. |
tree | commitdiff |
| 2008-07-03 |
Evan Cheng | Use std::replace instead of std::find and push_back. |
tree | commitdiff |
| 2008-07-03 |
Evan Cheng | - Add LiveVariables::replaceKillInstruction. This does... |
tree | commitdiff |
| 2008-07-02 |
Owen Anderson | Make LiveVariables even more optional, by making it... |
tree | commitdiff |
| 2008-07-02 |
Dan Gohman | Replace a few uses of SelectionDAG::getTargetNode with |
tree | commitdiff |
| 2008-07-02 |
Dan Gohman | Revert r52988. It broke 254.gap on x86-64. |
tree | commitdiff |
| 2008-07-02 |
Owen Anderson | TwoAddressInstructionPass doesn't really require LiveVa... |
tree | commitdiff |
| 2008-07-02 |
Duncan Sands | Add a new getMergeValues method that does not need |
tree | commitdiff |
| 2008-07-02 |
Mon P Wang | Fixed problem in EmitStackConvert where the source... |
tree | commitdiff |
| 2008-07-02 |
Chris Lattner | instead of aborting on shifts of i1, just implicitly... |
tree | commitdiff |
| 2008-07-02 |
Duncan Sands | Fix typo compounded by a cut-and-pasto. |
tree | commitdiff |
| 2008-07-02 |
Duncan Sands | Let AnalyzeNewNode take care of calling ExpungeNode. |
tree | commitdiff |
| 2008-07-02 |
Evan Cheng | - Use a faster priority comparison function if -fast. |
tree | commitdiff |
| 2008-07-02 |
Chris Lattner | Add a new (simple) StringMap::clear method, patch by... |
tree | commitdiff |
| 2008-07-02 |
Bill Wendling | Use the canonical form for getting an empty structure. |
tree | commitdiff |
| next |
projects / oota-llvm.git / history