| Commit message (Collapse) | Author | Age | Files | Lines | |
|---|---|---|---|---|---|
| * | Implement expansion in type legalization for add/sub with overflow. The | Eli Friedman | 2010-06-03 | 3 | -0/+115 |
| | | | | | | | | | | | expansion is the same as that used by LegalizeDAG. The resulting code sucks in terms of performance/codesize on x86-32 for a 64-bit operation; I haven't looked into whether different expansions might be better in general. llvm-svn: 105378 | ||||
| * | Remove some already-fixed README entries. | Eli Friedman | 2010-06-03 | 1 | -86/+1 |
| | | | | | llvm-svn: 105377 | ||||
| * | Remove README entry which no longer compiles to something sane. | Eli Friedman | 2010-06-03 | 1 | -56/+0 |
| | | | | | llvm-svn: 105376 | ||||
| * | Remove a fixed item, update a couple partially-fixed items. | Eli Friedman | 2010-06-03 | 1 | -81/+22 |
| | | | | | llvm-svn: 105375 | ||||
| * | Use the fast register allocator by default for -O0 builds. | Jakob Stoklund Olesen | 2010-06-03 | 1 | -1/+1 |
| | | | | | | | This affects both llvm-gcc and clang. llvm-svn: 105372 | ||||
| * | Use readsWritesVirtualRegister instead of counting uses and defs when inserting | Jakob Stoklund Olesen | 2010-06-03 | 1 | -46/+8 |
| | | | | | | | | | | | | | | spills and reloads. This means that a partial define of a register causes a reload so the other parts of the register are preserved. The reload can be prevented by adding an <imp-def> operand for the full register. This is already done by the coalescer and live interval analysis where relevant. llvm-svn: 105369 | ||||
| * | Add full register <imp-def> operands when the coalescer is creating partial | Jakob Stoklund Olesen | 2010-06-02 | 1 | -50/+50 |
| | | | | | | | | | | register updates. These operands tell the spiller that the other parts of the partially defined register are don't-care, and a reload is not necessary. llvm-svn: 105361 | ||||
| * | Speedup bitcode writer. Do not walk all values for all functions to emit ↵ | Devang Patel | 2010-06-02 | 3 | -9/+16 |
| | | | | | | | function local metadata. In one testcase, probably worst case scenario, the 70x speed up is seen. llvm-svn: 105360 | ||||
| * | Compulsive reformating. No functionalitical changes. | Bill Wendling | 2010-06-02 | 1 | -11/+19 |
| | | | | | llvm-svn: 105359 | ||||
| * | Slightly change the meaning of the reMaterialize target hook when the original | Jakob Stoklund Olesen | 2010-06-02 | 14 | -40/+48 |
| | | | | | | | | | | | | | | | | | | | | | | | | instruction defines subregisters. Any existing subreg indices on the original instruction are preserved or composed with the new subreg index. Also substitute multiple operands mentioning the original register by using the new MachineInstr::substituteRegister() function. This is necessary because there will soon be <imp-def> operands added to non read-modify-write partial definitions. This instruction: %reg1234:foo = FLAP %reg1234<imp-def> will reMaterialize(%reg3333, bar) like this: %reg3333:bar-foo = FLAP %reg333:bar<imp-def> Finally, replace the TargetRegisterInfo pointer argument with a reference to indicate that it cannot be NULL. llvm-svn: 105358 | ||||
| * | Update Readme and Makefiles for the new gtest. | Benjamin Kramer | 2010-06-02 | 4 | -17/+2 |
| | | | | | llvm-svn: 105355 | ||||
| * | Merge gtest-1.5.0. | Benjamin Kramer | 2010-06-02 | 24 | -989/+1497 |
| | | | | | llvm-svn: 105354 | ||||
| * | Merge gtest-1.4.0. | Benjamin Kramer | 2010-06-02 | 21 | -2039/+4137 |
| | | | | | llvm-svn: 105353 | ||||
| * | Merge gtest-1.3.0. | Benjamin Kramer | 2010-06-02 | 20 | -468/+2089 |
| | | | | | | | | OSX users: make sure that CrashReporter is disabled when running unit tests. Death tests are enabled now so you'll get a ton of message boxes. llvm-svn: 105352 | ||||
| * | Clean up 80 column violations. No functional change. | Jim Grosbach | 2010-06-02 | 19 | -50/+64 |
| | | | | | llvm-svn: 105350 | ||||
| * | arm_neon.h emitter now mostly complete for the purposes of initial testing. | Nate Begeman | 2010-06-02 | 1 | -10/+81 |
| | | | | | llvm-svn: 105349 | ||||
| * | Remove the TargetRegisterClass member from CalleeSavedInfo | Rafael Espindola | 2010-06-02 | 10 | -59/+65 |
| | | | | | llvm-svn: 105344 | ||||
| * | Fix comment so it doesn't include comments which are irrelevant to the x86 | Eli Friedman | 2010-06-02 | 1 | -9/+5 |
| | | | | | | | backend. Add a FIXME noting what can be fixed here. llvm-svn: 105342 | ||||
| * | Use comments to document non-obvious code rather than | Dan Gohman | 2010-06-02 | 1 | -0/+11 |
| | | | | | | | mailing list archives. llvm-svn: 105341 | ||||
| * | Skip identical instruction while calculating DBG_VALUE range. | Devang Patel | 2010-06-02 | 1 | -1/+4 |
| | | | | | llvm-svn: 105340 | ||||
| * | Rename canCombinedSubRegIndex method to something more grammatically correct | Bob Wilson | 2010-06-02 | 4 | -19/+21 |
| | | | | | | | and tidy up the comment describing it. llvm-svn: 105339 | ||||
| * | Replace ARM's getCalleeSavedRegClasses with a simpler solution | Rafael Espindola | 2010-06-02 | 2 | -94/+40 |
| | | | | | llvm-svn: 105335 | ||||
| * | remove trailing whitespace | Jim Grosbach | 2010-06-02 | 1 | -75/+75 |
| | | | | | llvm-svn: 105333 | ||||
| * | Use local small vector. | Devang Patel | 2010-06-02 | 2 | -4/+3 |
| | | | | | llvm-svn: 105332 | ||||
| * | Remove unused function. | Rafael Espindola | 2010-06-02 | 2 | -56/+0 |
| | | | | | llvm-svn: 105325 | ||||
| * | Not all entries in the range will have an SUnit. Check for that when looking | Jim Grosbach | 2010-06-02 | 2 | -0/+2 |
| | | | | | | | for debug information. llvm-svn: 105324 | ||||
| * | Update the documentation, getCalleeSavedRegClasses is not required anymore. | Rafael Espindola | 2010-06-02 | 1 | -3/+0 |
| | | | | | llvm-svn: 105323 | ||||
| * | cleanup | Rafael Espindola | 2010-06-02 | 22 | -352/+0 |
| | | | | | llvm-svn: 105322 | ||||
| * | Remove uses of getCalleeSavedRegClasses from outside the | Rafael Espindola | 2010-06-02 | 3 | -12/+26 |
| | | | | | | | | backends and removes the virtual declaration. With that out of the way I should be able to cleanup one backend at a time. llvm-svn: 105321 | ||||
| * | Pacify recent gcc: remove a pointless const qualifier. | Duncan Sands | 2010-06-02 | 1 | -1/+1 |
| | | | | | llvm-svn: 105318 | ||||
| * | Checkpoint; handle 'int' and 'void' correctly | Nate Begeman | 2010-06-02 | 1 | -25/+52 |
| | | | | | llvm-svn: 105316 | ||||
| * | Emit full function prototypes. Definitions & typedefs to come. | Nate Begeman | 2010-06-02 | 1 | -16/+72 |
| | | | | | llvm-svn: 105315 | ||||
| * | Enable machine cse of instructions which define physical registers. | Evan Cheng | 2010-06-02 | 2 | -4/+19 |
| | | | | | llvm-svn: 105308 | ||||
| * | Checkpoint arm_neon.h generation with tablegen | Nate Begeman | 2010-06-02 | 1 | -13/+241 |
| | | | | | llvm-svn: 105307 | ||||
| * | Don't try to custom-lower 64-bit add-with-overflow and friends on x86-32; the | Eli Friedman | 2010-06-02 | 1 | -5/+7 |
| | | | | | | | | | | x86 backend currently doesn't know how to handle them. This doesn't really fix anything because LegalizeTypes doesn't know how to handle them either. We do get a better error message, though. llvm-svn: 105305 | ||||
| * | Fix an obvious mistake: don't change the operands until all of them have been | Bob Wilson | 2010-06-02 | 1 | -2/+0 |
| | | | | | | | checked and it is safe to proceed with the changes. llvm-svn: 105304 | ||||
| * | Remove outdated README entries. | Eli Friedman | 2010-06-02 | 1 | -140/+0 |
| | | | | | llvm-svn: 105303 | ||||
| * | Update debug information when breaking anti-dependencies. rdar://7759363 | Jim Grosbach | 2010-06-01 | 2 | -1/+31 |
| | | | | | llvm-svn: 105300 | ||||
| * | Fix comment | Alexis Hunt | 2010-06-01 | 1 | -1/+1 |
| | | | | | llvm-svn: 105297 | ||||
| * | DwarfWrite is now smart enough to drop debug value pointing to undefined ↵ | Devang Patel | 2010-06-01 | 1 | -40/+35 |
| | | | | | | | | | register. Update this test to avoid this. iSel not properly lowring argument into a well formed DBG_VALUE in some cases is a separate issue and not related to the test in this testcase. llvm-svn: 105295 | ||||
| * | Properly compose subregister indices when coalescing. | Jakob Stoklund Olesen | 2010-06-01 | 2 | -14/+3 |
| | | | | | | | | The comment about ordering of subreg indices is no longer true. This exposed a bug in the new substVirtReg method that is also fixed. llvm-svn: 105294 | ||||
| * | Remove unused code | Jim Grosbach | 2010-06-01 | 1 | -64/+0 |
| | | | | | llvm-svn: 105293 | ||||
| * | Ignore line number of debug value in undefined register. | Devang Patel | 2010-06-01 | 1 | -0/+1 |
| | | | | | llvm-svn: 105292 | ||||
| * | fix think-o | Jim Grosbach | 2010-06-01 | 1 | -6/+2 |
| | | | | | llvm-svn: 105291 | ||||
| * | Fix the allocation of shadow space for the Win64 calling convention | Dan Gohman | 2010-06-01 | 1 | -0/+6 |
| | | | | | | | in X86FastISel. Patch by Jan Sjodin. llvm-svn: 105290 | ||||
| * | Simplify things a bit more. Fix prototype to use SmallVectorImpl and | Jim Grosbach | 2010-06-01 | 1 | -17/+17 |
| | | | | | | | change a few SmallVectors to vanilla C arrays. llvm-svn: 105289 | ||||
| * | Fix the noalias checking so that it doesn't worry about | Dan Gohman | 2010-06-01 | 1 | -12/+12 |
| | | | | | | | an argument aliasing itself. Thanks Duncan! llvm-svn: 105288 | ||||
| * | Keep track of incoming debug value of unused argument. | Devang Patel | 2010-06-01 | 3 | -1/+72 |
| | | | | | | | Radar 7927666. llvm-svn: 105285 | ||||
| * | Fill in missing support for ISD::FEXP, ISD::FPOWI, and friends. | Dan Gohman | 2010-06-01 | 4 | -4/+55 |
| | | | | | llvm-svn: 105283 | ||||
| * | Add a FIXME | Jim Grosbach | 2010-06-01 | 1 | -0/+2 |
| | | | | | llvm-svn: 105282 | ||||

