| Commit message (Expand) | Author | Age | Files | Lines |
|---|
| * | When fast iseling a GEP, accumulate the offset rather than emitting a series of | Chad Rosier | 2011-11-17 | 1 | -11/+35 |
| * | Make sure to replace the chain properly when DAGCombining a LOAD+EXTRACT_VECT... | Eli Friedman | 2011-11-16 | 1 | -4/+17 |
| * | Add fast-isel stats to determine who's doing all the work, the | Chad Rosier | 2011-11-16 | 1 | -0/+7 |
| * | Fix the stats collection for fast-isel. The failed count was only accounting | Chad Rosier | 2011-11-16 | 1 | -5/+18 |
| * | CONCAT_VECTORS can have more than two operands. PR11389. | Eli Friedman | 2011-11-16 | 1 | -22/+12 |
| * | Add a couple asserts so it will be easier to debug if we accidentally pass in... | Eli Friedman | 2011-11-16 | 1 | -0/+4 |
| * | Rename MVT::untyped to MVT::Untyped to match similar nomenclature. | Owen Anderson | 2011-11-16 | 1 | -2/+2 |
| * | GEPs with all zero indices are trivially coalesced by fast-isel. For example, | Chad Rosier | 2011-11-15 | 1 | -0/+5 |
| * | Added custom lowering for load->dec->store sequence in x86 when the EFLAGS re... | Pete Cooper | 2011-11-15 | 1 | -0/+5 |
| * | Remove all remaining uses of Value::getNameStr(). | Benjamin Kramer | 2011-11-15 | 1 | -2/+2 |
| * | Twinify GraphWriter a little bit. | Benjamin Kramer | 2011-11-15 | 1 | -1/+1 |
| * | Remove some unnecessary includes of PseudoSourceValue.h. | Jay Foad | 2011-11-15 | 7 | -7/+0 |
| * | Don't try to form pre/post-indexed loads/stores until after LegalizeDAG runs.... | Eli Friedman | 2011-11-12 | 2 | -11/+11 |
| * | Some cleanup and bulletproofing for node replacement in LegalizeDAG. To main... | Eli Friedman | 2011-11-11 | 1 | -57/+59 |
| * | Use a bigger hammer to fix PR11314 by disabling the "forcing two-address | Evan Cheng | 2011-11-10 | 1 | -1/+5 |
| * | Add check so we don't try to perform an impossible transformation. Fixes issu... | Eli Friedman | 2011-11-09 | 1 | -1/+2 |
| * | Speculatively revert commit 144124 (djg) in the hope that the 32 bit | Duncan Sands | 2011-11-09 | 1 | -13/+3 |
| * | Add a hack to the scheduler to disable pseudo-two-address dependencies in | Dan Gohman | 2011-11-08 | 1 | -3/+13 |
| * | Lower mem-ops to unaligned i32/i16 load/stores on ARM where supported. | Lang Hames | 2011-11-08 | 1 | -0/+10 |
| * | Added invariant field to the DAG.getLoad method and changed all calls. | Pete Cooper | 2011-11-08 | 10 | -64/+85 |
| * | Add a bunch of calls to RemoveDeadNode in LegalizeDAG, so legalization doesn'... | Eli Friedman | 2011-11-08 | 2 | -3/+25 |
| * | Revert r144034 while I try to track down a crash. | Eli Friedman | 2011-11-07 | 1 | -19/+3 |
| * | Add a bunch of calls to RemoveDeadNode in LegalizeDAG, so legalization doesn'... | Eli Friedman | 2011-11-07 | 1 | -3/+19 |
| * | Don't introduce custom nodes after legalization in TargetLowering::BuildSDIV() | Richard Osborne | 2011-11-07 | 2 | -10/+16 |
| * | Reapply r143206, with fixes. Disallow physical register lifetimes | Dan Gohman | 2011-11-03 | 4 | -518/+454 |
| * | build: Add initial cut at LLVMBuild.txt files. | Daniel Dunbar | 2011-11-03 | 1 | -0/+23 |
| * | An array of chars of length 8 will also cause the stack protector to be inserted | Bill Wendling | 2011-11-02 | 1 | -1/+1 |
| * | Cleanup. Document. Make sure that this build_vector optimization only runs be... | Nadav Rotem | 2011-10-31 | 1 | -16/+29 |
| * | Silence compiler warning. | Benjamin Kramer | 2011-10-30 | 1 | -4/+4 |
| * | Add a new DAGCombine optimization for BUILD_VECTOR. | Nadav Rotem | 2011-10-29 | 1 | -0/+83 |
| * | Revert r143206, as there are still some failing tests. | Dan Gohman | 2011-10-29 | 4 | -436/+518 |
| * | Reapply r143177 and r143179 (reverting r143188), with scheduler | Dan Gohman | 2011-10-28 | 4 | -518/+436 |
| * | Speculatively disable Dan's commits 143177 and 143179 to see if | Duncan Sands | 2011-10-28 | 4 | -407/+516 |
| * | Delete #if 0 code accidentally left in. | Dan Gohman | 2011-10-28 | 1 | -17/+0 |
| * | Eliminate LegalizeOps' LegalizedNodes map and have it just call RAUW | Dan Gohman | 2011-10-28 | 4 | -515/+423 |
| * | Don't crash on 128-bit sdiv by constant. Found by inspection. | Eli Friedman | 2011-10-27 | 1 | -9/+6 |
| * | Rename NonScalarIntSafe to something more appropriate. | Lang Hames | 2011-10-26 | 1 | -4/+4 |
| * | Simplify SplitVecRes_UnaryOp by removing all the code that is | Duncan Sands | 2011-10-26 | 1 | -43/+5 |
| * | Don't use floating point to do an integer's job. | Jakob Stoklund Olesen | 2011-10-26 | 1 | -4/+7 |
| * | Remove a couple redundant checks. | Eli Friedman | 2011-10-25 | 1 | -2/+0 |
| * | Really unbreak CMake build | Douglas Gregor | 2011-10-24 | 1 | -3/+1 |
| * | Unbreak CMake build | Douglas Gregor | 2011-10-24 | 1 | -0/+1 |
| * | Delete the top-down "Latency" scheduler. Top-down scheduling doesn't handle | Dan Gohman | 2011-10-24 | 1 | -265/+0 |
| * | Delete the Latency scheduling preference. | Dan Gohman | 2011-10-24 | 1 | -2/+0 |
| * | Change this overloaded use of Sched::Latency to be an overloaded | Dan Gohman | 2011-10-24 | 1 | -4/+4 |
| * | Change the default scheduler from Latency to ILP, since Latency | Dan Gohman | 2011-10-24 | 1 | -1/+1 |
| * | Fix pr11194. When promoting and splitting integers we need to use | Nadav Rotem | 2011-10-21 | 1 | -3/+12 |
| * | 1. Fix the widening of SETCC in WidenVecOp_SETCC. Use the correct return CC t... | Nadav Rotem | 2011-10-21 | 3 | -14/+17 |
| * | Remove a now dead function, fixing -Wunused-function warnings from | Chandler Carruth | 2011-10-21 | 1 | -20/+0 |
| * | Delete the list-tdrr scheduler. Top-down schedulers are going away | Dan Gohman | 2011-10-20 | 1 | -203/+11 |
