| Commit message (Expand) | Author | Age | Files | Lines |
* | System: Fix getMagicNumber on windows. | Michael J. Spencer | 2010-08-31 | 1 | -3/+2 |
* | Offset is not always unsigned number. | Devang Patel | 2010-08-31 | 2 | -2/+2 |
* | Simplify. | Devang Patel | 2010-08-31 | 2 | -15/+15 |
* | Switch to DenseSet, simplifying much more code. We now have a single iteration | Nick Lewycky | 2010-08-31 | 1 | -78/+83 |
* | Remove r111665, which implemented store-narrowing in InstCombine. Chris disc... | Owen Anderson | 2010-08-31 | 1 | -47/+0 |
* | zap unused method. x86 is the only user and already has a more powerfull version | Bruno Cardoso Lopes | 2010-08-31 | 1 | -29/+0 |
* | Use X86ISD::MOVSS and MOVSD to represent the movl mask pattern, also fix the ... | Bruno Cardoso Lopes | 2010-08-31 | 1 | -7/+14 |
* | Rewrite slightly so we can expand for floating point types easier. | Eric Christopher | 2010-08-31 | 1 | -17/+25 |
* | Add experimental -disable-physical-join command line option. | Jakob Stoklund Olesen | 2010-08-31 | 1 | -0/+10 |
* | Fix a typo. | Owen Anderson | 2010-08-30 | 1 | -1/+1 |
* | If we have an unhandled type then assert, we shouldn't get here for | Eric Christopher | 2010-08-30 | 1 | -1/+3 |
* | Cleanups suggested by Chris. | Owen Anderson | 2010-08-30 | 1 | -20/+18 |
* | Re-apply r112539, being more careful to respect the return values of the cons... | Owen Anderson | 2010-08-30 | 1 | -22/+25 |
* | Expand MOVi32imm in ARM mode after regalloc. This provides | Anton Korobeynikov | 2010-08-30 | 1 | -2/+7 |
* | Add statistics to evaluate this pass. | Owen Anderson | 2010-08-30 | 1 | -0/+8 |
* | Revert r112539. It accidentally introduced a miscompilation. | Owen Anderson | 2010-08-30 | 1 | -20/+16 |
* | Fixes and cleanups pointed out by Chris. In general, be careful to handle 0 ... | Owen Anderson | 2010-08-30 | 1 | -16/+20 |
* | Use the existing T2I_bin_s_irs pattern instead of creating T2I_bin_sw_irs, which | Bill Wendling | 2010-08-30 | 2 | -69/+2 |
* | Fix a comment. | NAKAMURA Takumi | 2010-08-30 | 1 | -1/+1 |
* | Remember to clear the shadow kill flag at the same time as clearing the real | Jakob Stoklund Olesen | 2010-08-30 | 1 | -0/+1 |
* | Add comments explaining why it's not necessary to include the | Dan Gohman | 2010-08-30 | 1 | -1/+6 |
* | Remove NEON vmovn intrinsic, replacing it with vector truncate operations. | Bob Wilson | 2010-08-30 | 2 | -3/+34 |
* | Make ARM add rN, sp, #imm instructions rematerializable. That's how the addre... | Jim Grosbach | 2010-08-30 | 3 | -0/+12 |
* | two changes: | Chris Lattner | 2010-08-30 | 1 | -5/+0 |
* | When expanding NEON VST pseudo instructions, if the original super-register | Bob Wilson | 2010-08-30 | 1 | -4/+6 |
* | MCELF: The value of all common symbols is the offset from the start of the se... | Benjamin Kramer | 2010-08-30 | 1 | -3/+5 |
* | It is possible to try to merge a not-constant with a constantrage, when deali... | Owen Anderson | 2010-08-30 | 1 | -2/+3 |
* | Don't print two "0x" prefixes. Use a raw_ostream overload instead of llvm::fo... | Benjamin Kramer | 2010-08-30 | 1 | -2/+1 |
* | EE/JIT: Do not invoke parent's ctors/dtors from main()! (PR3897) | NAKAMURA Takumi | 2010-08-30 | 1 | -0/+12 |
* | The value is offset from the start of the section for non-common symbols, sub... | Benjamin Kramer | 2010-08-30 | 1 | -0/+2 |
* | Index external symbols by symbol table instead of parent section, by Roman Di... | Benjamin Kramer | 2010-08-30 | 1 | -1/+1 |
* | Mark all common symbols external. This is not exactly correct but it lets apps | Benjamin Kramer | 2010-08-30 | 1 | -2/+3 |
* | Remove a hack that tries to understand incorrect triples from the | Duncan Sands | 2010-08-30 | 1 | -16/+0 |
* | Revert r112461. It was failing on PPC... | Bill Wendling | 2010-08-30 | 1 | -4/+2 |
* | Create Thumb2sI_cpsr and T2sI_cpsr. These new classes indicate that CPSR is the | Bill Wendling | 2010-08-30 | 2 | -11/+71 |
* | When adding a register, we should mark it as "def" if it can optionally define | Bill Wendling | 2010-08-30 | 1 | -2/+4 |
* | revert 112457, it looks like it broke selfhost. | Chris Lattner | 2010-08-29 | 1 | -133/+23 |
* | rewrite DwarfEHPrepare to use SSAUpdater to promote its allocas | Chris Lattner | 2010-08-29 | 2 | -24/+134 |
* | inline function into its only caller. | Chris Lattner | 2010-08-29 | 1 | -13/+6 |
* | two changes: 1) make AliasSet hold the list of call sites with an | Chris Lattner | 2010-08-29 | 2 | -14/+21 |
* | LICM does get dead instructions input to it. Instead of sinking them | Chris Lattner | 2010-08-29 | 1 | -1/+12 |
* | use moveBefore instead of remove+insert, it avoids some | Chris Lattner | 2010-08-29 | 1 | -9/+3 |
* | revert 112448 for now. | Chris Lattner | 2010-08-29 | 1 | -5/+6 |
* | optimize LICM::hoist to use moveBefore. Correct its updating | Chris Lattner | 2010-08-29 | 1 | -6/+5 |
* | fix some bugs (found by inspection) where LICM would not update | Chris Lattner | 2010-08-29 | 1 | -1/+3 |
* | rework the ownership of subloop alias information: instead of | Chris Lattner | 2010-08-29 | 1 | -16/+25 |
* | apparently unswitch had the same "Feature". Stop its | Chris Lattner | 2010-08-29 | 1 | -7/+2 |
* | now that loop passes don't use DomFrontier, there is no reason | Chris Lattner | 2010-08-29 | 1 | -8/+2 |
* | Make IVUsers iterative instead of recursive. | Dan Gohman | 2010-08-29 | 1 | -78/+89 |
* | Optionally rerun dedicated-register filtering after applying | Dan Gohman | 2010-08-29 | 1 | -0/+20 |