aboutsummaryrefslogtreecommitdiff
path: root/lib/CodeGen
AgeCommit message (Expand)Author
2012-02-22Initialize SUnits before DAG building.Andrew Trick
2012-02-22Make all pointers to TargetRegisterClass const since they are all pointers to...Craig Topper
2012-02-22Use SparseSet for the RAFast live virtual register map.Jakob Stoklund Olesen
2012-02-21Add API "handleMoveIntoBundl" for updating liveness when moving instructions ...Lang Hames
2012-02-21There's no need for a DW_AT_byte_size on a pointer type.Eric Christopher
2012-02-21Clear virtual registers after they are no longer referenced.Andrew Trick
2012-02-21StackSlotColoring does not use a VirtRegMapAndrew Trick
2012-02-21Fix some bugs in HMEditor's moveAllOperandsInto logic.Lang Hames
2012-02-20Fix machine-cp by having it to check sub-register indicies. e.g.Evan Cheng
2012-02-20Teach the DAGCombiner that certain loadext nodes followed by ANDs can be conv...James Molloy
2012-02-20Make post-ra tail duplication bundle safe. No test case as recent codegenEvan Cheng
2012-02-19Silence operator precedence warning.Benjamin Kramer
2012-02-19Remove dead code. Improve llvm_unreachable text. Simplify some control flow.Ahmed Charles
2012-02-19Add machinery for pushing live ranges onto bundle starts while bundling.Lang Hames
2012-02-19Simplify moveEnteringDownFrom rules.Lang Hames
2012-02-19Skip through instructions rather than operands when looking for last use slot.Lang Hames
2012-02-19Fix TODO and trailing whitespace.Lang Hames
2012-02-19Defer sanity checks on live intervals until after all have been updated. Hold...Lang Hames
2012-02-17Bring HMEditor into line with LLVM coding standards.Lang Hames
2012-02-17Ignore the lifetime intrinsics in fast-isel.Eric Christopher
2012-02-17Don't print out pointer values in SUnit::dump().Jakob Stoklund Olesen
2012-02-17Sink variable into assertMatt Beaumont-Gay
2012-02-17Add support for regmask slots to HMEditor. Also fixes a comment error.Lang Hames
2012-02-17Transfer regmasks to MRI.Jakob Stoklund Olesen
2012-02-17Refactor 'handleMove' code in live intervals. Clients of LiveIntervals won't seeLang Hames
2012-02-17Tidy up.Jim Grosbach
2012-02-17Revert r150288, "Allow Post-RA LICM to hoist reserved register reads."Jakob Stoklund Olesen
2012-02-17... and it's probably best to use the correct alignment, rather than just gue...David Chisnall
2012-02-17It turns out that putting an 8-byte symbol in a 4-byte section makes Solaris ...David Chisnall
2012-02-17Reverse iterator - should be incrementing rather than decrementing.Lang Hames
2012-02-17MachineScheduler shouldn't use/preserve LiveDebugVariables.Lang Hames
2012-02-17Oops - isRegLiveIntoSuccessor is used in non-assert builds now. Remove NDEBUG...Lang Hames
2012-02-17Re-enable 150652 and 150654 - Make FPSCR non-reserved, and make MachineCSE ba...Lang Hames
2012-02-17Turn off assertion, conservatively compute liveness for live-in un-allocatabl...Lang Hames
2012-02-16Disable machine copy propagation for now. It's known to be buggy (PR11940) an...Benjamin Kramer
2012-02-16Remove extraneous #include and spelling mistake introduced in r150669.James Molloy
2012-02-16Modify the algorithm when traversing the DAGCombiner's worklist to be O(log N...James Molloy
2012-02-16Oop - r150653 + r150654 broke one of my test cases. Backing out for now...Lang Hames
2012-02-16MachineCSE shouldn't extend the live ranges of reserved or allocatable regist...Lang Hames
2012-02-15Handle register masks in branch folding.Jakob Stoklund Olesen
2012-02-15Fix library visibility problems with VLIWPacketizer.Andrew Trick
2012-02-15Make LiveIntervals::handleMove() bundle aware.Lang Hames
2012-02-15Use 'getDataNoRel' for the section kind.Bill Wendling
2012-02-15Fix assertion condition.Lang Hames
2012-02-15Modify the code that emits the module flags to use the new module flags accessorBill Wendling
2012-02-15Don't expose DefaultVLIWSchedulerAndrew Trick
2012-02-15Remove overly conservative assert.Lang Hames
2012-02-15Generic "VLIW" packetizer based on a DFA generated from target itinerary.Andrew Trick
2012-02-15Revert r150565 again. Appears to be a stage2 failure with dragonegg.Andrew Trick
2012-02-15Reapply r150565 with the typo fix properly merged.Andrew Trick