| Commit message (Collapse) | Author | Age | Files | Lines | ||
|---|---|---|---|---|---|---|
| ... | ||||||
| * | Reset the debug location even if the instruction was a terminator. | Dan Gohman | 2010-04-14 | 1 | -5/+3 | |
| | | | | | llvm-svn: 101272 | |||||
| * | Refine #includes. | Dan Gohman | 2010-04-14 | 1 | -1/+1 | |
| | | | | | llvm-svn: 101269 | |||||
| * | Pull utility routines with no SelectionDAG dependence out of | Dan Gohman | 2010-04-14 | 3 | -76/+95 | |
| | | | | | | | | SelectionDAGBuilder. FunctionLoweringInfo isn't an ideal place for them to live, but it's better than SelectionDAGBuilder for now. llvm-svn: 101267 | |||||
| * | Fix typos in comments. | Dan Gohman | 2010-04-14 | 2 | -2/+2 | |
| | | | | | llvm-svn: 101266 | |||||
| * | performance: cache the dereferenced use_iterator | Gabor Greif | 2010-04-14 | 1 | -7/+8 | |
| | | | | | llvm-svn: 101265 | |||||
| * | Delete an obsolete comment. | Dan Gohman | 2010-04-14 | 1 | -1/+0 | |
| | | | | | llvm-svn: 101264 | |||||
| * | Delete an unused function. | Dan Gohman | 2010-04-14 | 1 | -4/+0 | |
| | | | | | llvm-svn: 101263 | |||||
| * | Clear the FunctionLoweringInfo object before doing other things that | Dan Gohman | 2010-04-14 | 1 | -4/+4 | |
| | | | | | | | don't need it. llvm-svn: 101262 | |||||
| * | Move this assert out of SelectionDAGISel into FunctionLoweringInfo, and | Dan Gohman | 2010-04-14 | 2 | -5/+3 | |
| | | | | | | | drop the redundant #ifndef NDEBUG. llvm-svn: 101261 | |||||
| * | Add a comment. | Dan Gohman | 2010-04-14 | 1 | -0/+2 | |
| | | | | | llvm-svn: 101260 | |||||
| * | Move the code for initialing the entry block livein set out of | Dan Gohman | 2010-04-14 | 2 | -5/+5 | |
| | | | | | | | SelectionDAGISel. llvm-svn: 101258 | |||||
| * | Reorgnaize this code to be more tidy and readable. | Dan Gohman | 2010-04-14 | 1 | -10/+5 | |
| | | | | | llvm-svn: 101256 | |||||
| * | Trim #includes. | Dan Gohman | 2010-04-14 | 1 | -9/+0 | |
| | | | | | llvm-svn: 101255 | |||||
| * | Move the code for emitting livein copies out of SelectionDAGISel. | Dan Gohman | 2010-04-14 | 2 | -105/+107 | |
| | | | | | llvm-svn: 101254 | |||||
| * | performance: cache the dereferenced use_iterator | Gabor Greif | 2010-04-14 | 1 | -3/+4 | |
| | | | | | llvm-svn: 101253 | |||||
| * | Sink landing-pad marking code out of | Dan Gohman | 2010-04-14 | 2 | -5/+5 | |
| | | | | | | | SelectionDAGISel::runOnMachineFunction into FunctionLowering. llvm-svn: 101252 | |||||
| * | It's not necessary to recompute EB here. | Dan Gohman | 2010-04-14 | 1 | -1/+1 | |
| | | | | | llvm-svn: 101251 | |||||
| * | performance: cache the dereferenced use_iterator | Gabor Greif | 2010-04-14 | 1 | -3/+5 | |
| | | | | | llvm-svn: 101250 | |||||
| * | Add a comment. | Dan Gohman | 2010-04-14 | 1 | -0/+2 | |
| | | | | | llvm-svn: 101248 | |||||
| * | Enable RTTI again. While this works fine for LLVM, it creates an ABI | Nick Lewycky | 2010-04-14 | 1 | -0/+1 | |
| | | | | | | | incompatibility with some clients covered by the buildbots, such as llvm-gcc. llvm-svn: 101237 | |||||
| * | Turn off RTTI for VMCore. Yay! | Nick Lewycky | 2010-04-14 | 1 | -1/+0 | |
| | | | | | llvm-svn: 101233 | |||||
| * | I don't know how, but I managed to goof the revert. Remove function that should | Nick Lewycky | 2010-04-14 | 1 | -37/+0 | |
| | | | | | | | have been removed in r101231. llvm-svn: 101232 | |||||
| * | Revert r101213. | Nick Lewycky | 2010-04-14 | 1 | -29/+14 | |
| | | | | | llvm-svn: 101231 | |||||
| * | Fix 80 column ruler. | Nick Lewycky | 2010-04-14 | 1 | -1/+1 | |
| | | | | | llvm-svn: 101229 | |||||
| * | implement mc asmparser support for '.', which gets the | Chris Lattner | 2010-04-14 | 2 | -0/+17 | |
| | | | | | | | | | | | | | | | | | | | | current PC. rdar://7834775 We now produce an identical .o file compared to the cctools assembler for something like this: _f0: L0: jmp L1 .long . - L0 L1: jmp A .long . - L1 .zerofill __DATA,_bss,A,0 llvm-svn: 101227 | |||||
| * | Remove tab. | Nick Lewycky | 2010-04-14 | 1 | -1/+1 | |
| | | | | | llvm-svn: 101223 | |||||
| * | While DAE can't modify the function signature of an externally visible function, | Nick Lewycky | 2010-04-14 | 1 | -14/+66 | |
| | | | | | | | | | | it can check whether the visible direct callers are passing in parameters to dead arguments and replace those with undef. This reinstates r94322 with bugs fixed. llvm-svn: 101213 | |||||
| * | Generalize this code to handle Instructions in addition to ConstantExprs. | Dan Gohman | 2010-04-14 | 1 | -3/+3 | |
| | | | | | llvm-svn: 101210 | |||||
| * | Reorder the methods of this class to be a little more organized. | Dan Gohman | 2010-04-14 | 1 | -22/+22 | |
| | | | | | llvm-svn: 101206 | |||||
| * | Fixed another assert exposed by fuzzing. Now, when an encoding error occurs | Johnny Chen | 2010-04-14 | 1 | -4/+14 | |
| | | | | | | | | involing getBFCInvMask() where lsb <= msb does not hold true, the disassembler just returns false, instead of assert, to indicate disassembly error. llvm-svn: 101205 | |||||
| * | Clear MachineInstr->MCSymbol maps at the end of a function. | Devang Patel | 2010-04-14 | 1 | -0/+2 | |
| | | | | | llvm-svn: 101202 | |||||
| * | Fixed an assert() exposed by fuzzing. Now, instead of assert when an invalid | Johnny Chen | 2010-04-14 | 2 | -10/+15 | |
| | | | | | | | | instruction encoding is encountered, we just return a NULL ARMBasicMCBuilder instance and the client just returns false to indicate disassembly error. llvm-svn: 101201 | |||||
| * | Unbreak CMake build by improving the EnhancedDisassembly makefile a | Douglas Gregor | 2010-04-13 | 2 | -0/+2 | |
| | | | | | | | | bit (we're not trying to build a shared library yet) and generating the X86GenEDInfo.inc and ARMGenEDInfo.inc files as necessary. llvm-svn: 101188 | |||||
| * | Fast path implicit_def check. | Evan Cheng | 2010-04-13 | 1 | -5/+5 | |
| | | | | | llvm-svn: 101183 | |||||
| * | Handle a v2f64 formal parameter that is split between registers and memory | Bob Wilson | 2010-04-13 | 1 | -7/+13 | |
| | | | | | | | such that the entire second half is in memory. Radar 7855014. llvm-svn: 101181 | |||||
| * | Fixed a nasty layering violation in the edis source | Sean Callanan | 2010-04-13 | 4 | -2/+30 | |
| | | | | | | | | | | | | | | | | code. It used to #include the enhanced disassembly information for the targets it supported straight out of lib/Target/{X86,ARM,...} but now it uses a new interface provided by MCDisassembler, and (so far) implemented by X86 and ARM. Also removed hacky #define-controlled initialization of targets in edis. If clients only want edis to initialize a limited set of targets, they can set --enable-targets on the configure command line. llvm-svn: 101179 | |||||
| * | Changed getSOImmValRotate()'s hunt retry logic to ignore the low order 6 bits, | Johnny Chen | 2010-04-13 | 1 | -3/+3 | |
| | | | | | | | instead of 7, because we are only looking for even rotate amount. llvm-svn: 101172 | |||||
| * | Do not include types without any definition in pubtypes list. | Devang Patel | 2010-04-13 | 1 | -2/+4 | |
| | | | | | llvm-svn: 101171 | |||||
| * | Avoid variable shadowing. | Evan Cheng | 2010-04-13 | 1 | -2/+2 | |
| | | | | | llvm-svn: 101170 | |||||
| * | Expand postra machine licm's capability a little more. If an instruction's ↵ | Evan Cheng | 2010-04-13 | 1 | -8/+24 | |
| | | | | | | | register operands are all loop invariants, then it's safe to hoist it. llvm-svn: 101167 | |||||
| * | Teach MachineSinking to handle easy critical edges. | Jakob Stoklund Olesen | 2010-04-13 | 1 | -2/+17 | |
| | | | | | | | | | | | | | | | Sometimes it is desirable to sink instructions along a critical edge: x = ... if (a && b) ... else use(x); The 'a && b' condition creates a critical edge to the else block, but we still want to sink the computation of x into the block. The else block is dominated by the parent block, so we are not pushing instructions into new code paths. llvm-svn: 101165 | |||||
| * | Re-apply 101075 and fix it properly. Just reuse the debug info of the branch ↵ | Evan Cheng | 2010-04-13 | 1 | -1/+44 | |
| | | | | | | | instruction being optimized. There is no need to --I which can deref off start of the BB. llvm-svn: 101162 | |||||
| * | Temporarily revert r101075, it's causing invalid iterator assertions | Eric Christopher | 2010-04-13 | 1 | -46/+1 | |
| | | | | | | | in a nightly tester. llvm-svn: 101158 | |||||
| * | Teach postra machine licm to hoist more obvious invariants, e.g. ↵ | Evan Cheng | 2010-04-13 | 1 | -17/+38 | |
| | | | | | | | instructions with no source operands. llvm-svn: 101154 | |||||
| * | Add a few comments. | Dan Gohman | 2010-04-13 | 1 | -0/+5 | |
| | | | | | llvm-svn: 101148 | |||||
| * | Eliminate MachineBasicBlock::const_livein_iterator and make | Dan Gohman | 2010-04-13 | 7 | -10/+11 | |
| | | | | | | | | | MachineBasicBlock::livein_iterator a const_iterator, because clients shouldn't ever be using the iterator interface to mutate the livein set. llvm-svn: 101147 | |||||
| * | Rename MachineFrameInfo variables to MFI, for consistency with | Dan Gohman | 2010-04-13 | 1 | -41/+41 | |
| | | | | | | | the rest of CodeGen. llvm-svn: 101146 | |||||
| * | Move MachineRegisterInfo's isLiveIn and isLiveOut out of line. | Dan Gohman | 2010-04-13 | 1 | -0/+14 | |
| | | | | | llvm-svn: 101145 | |||||
| * | Use MachineBasicBlock::isLiveIn. | Dan Gohman | 2010-04-13 | 1 | -8/+1 | |
| | | | | | llvm-svn: 101144 | |||||
| * | Delete an unused member variable. | Dan Gohman | 2010-04-13 | 1 | -1/+0 | |
| | | | | | llvm-svn: 101143 | |||||

