aboutsummaryrefslogtreecommitdiff
path: root/lib/CodeGen
AgeCommit message (Collapse)Author
2011-06-01Recommit 132404 with fixes. rdar://problem/5993888Stuart Hastings
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132424 91177308-0d34-0410-b5e6-96231b3b80d8
2011-06-01Allow bitcasts between valid types of the same size and vectorEric Christopher
types if the vector type is legal. Fixes rdar://9306086 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132420 91177308-0d34-0410-b5e6-96231b3b80d8
2011-06-01Refactor LegalizeTypes: Erase LegalizeAction and make the type legalizer useNadav Rotem
the TargetLowering enum. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132418 91177308-0d34-0410-b5e6-96231b3b80d8
2011-06-01Revert r132358 "Simplify the eviction policy by making the failsafe explicit."Jakob Stoklund Olesen
This commit caused regressions in i386 flops-[568], matrix, salsa20, 256.bzip2, and enc-md5. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132413 91177308-0d34-0410-b5e6-96231b3b80d8
2011-06-01Fix double FGETSIGN to work on x86_32; followup to 132396.Stuart Hastings
rdar://problem/5660695 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132411 91177308-0d34-0410-b5e6-96231b3b80d8
2011-06-01Turn on FGETSIGN for x86. Followup to 132388. rdar://problem/5660695Stuart Hastings
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132396 91177308-0d34-0410-b5e6-96231b3b80d8
2011-06-01This patch is another step in the direction of adding vector select. In thisNadav Rotem
patch we add a flag to enable a new type legalization decision - to promote integer elements in vectors. Currently, the rest of the codegen does not support this kind of legalization. This flag will be removed when the transition is complete. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132394 91177308-0d34-0410-b5e6-96231b3b80d8
2011-06-01Add an issue width check to the postRA scheduler. Patch by Max Kazakov!Andrew Trick
For targets with no itinerary (x86) it is a nop by default. For targets with issue width already expressed in the itinerary (ARM) it bypasses a scoreboard check but otherwise does not affect the schedule. It does make the code more consistent and complete and allows new targets to specify their issue width in an arbitrary way. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132385 91177308-0d34-0410-b5e6-96231b3b80d8
2011-06-01The ARM stuff already calls the Resume function, not the Resume_or_Rethrow. ItBill Wendling
turns out that it could cause an infinite loop in some situations. If this code is triggered and it converts a cleanup into a catchall, but that cleanup was in already in a cleanup, then the _Unwind_SjLj_Resume could infinite loop. I.e., the code doesn't consume the exception object and passes it on to _Unwind_SjLj_Resume. But _USjLjR expects it to be consumed (since it's landing at a catchall instead of a cleanup). So it uses the values that are presently there, which are the values that tell it to jump to the fake landing pad. <rdar://problem/9508402> git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132381 91177308-0d34-0410-b5e6-96231b3b80d8
2011-06-01Incomplete type may not have corresponding DIE, so do not check DIEEntry ↵Devang Patel
eagerly. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132377 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-31Refactor.Devang Patel
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132373 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-31Include global types, that are referenced through local variables, in ↵Devang Patel
debug_pubtypes list. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132371 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-31Simplify the eviction policy by making the failsafe explicit.Jakob Stoklund Olesen
When assigned ranges are evicted, they are put in the RS_Evicted stage and are not allowed to evict anything else. That prevents looping automatically. When evicting ranges just to get a cheaper register, use only spill weights to find the possible candidates. Avoid breaking hints for this purpose, it is not worth it. Start implementing more complex eviction heuristics, guarded by the temporary -complex-eviction flag. The initial version permits a heavier range to be evicted if it doesn't have any uses where the evicting range is live. This makes it a good candidate for live ranfge splitting. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132358 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-30Reapply r132245 with a fix for the bug that broke the darwin9/i386 build.Jakob Stoklund Olesen
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132309 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-30Emit the handler's data area. For GCC-style exceptions under Win64, theCharles Davis
handler's data area starts with a 4-byte reference to the personality function, followed by the DWARF LSDA. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132302 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-29Revert r132245, "Create two BlockInfo entries when a live range is ↵Jakob Stoklund Olesen
discontinuous through a block." This commit seems to have broken a darwin 9 tester. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132299 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-29Fix PR10046 by updating LiveVariables kill info when splitting live ranges.Jakob Stoklund Olesen
This only affects targets like Mips where branch instructions may kill virtual registers. Most other targets branch on flag values, so virtual registers are not involved. The problem is that MachineBasicBlock::updateTerminator deletes branches and inserts new ones while LiveVariables keeps a list of pointers to instructions that kill virtual registers. That list wasn't properly updated in MBB::SplitCriticalEdge. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132298 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-29When generating against the Win64 EH scheme, set the handler to the GCC-specificCharles Davis
handler. At this moment, only GCC-style exceptions are supported. Other kinds of exceptions, including "traditional" SEH and Microsoft Visual C++ exceptions, need more work--and an compiler exception model that isn't specific to GCC-style exceptions! In particular, I imagine that it would be possible to mix "traditional" SEH with GCC-style EH or Microsoft C++ EH. Currently LLVM has no way (beyond some target-specific defaults and whole-module compiler switches) of knowing which scheme to use when. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132283 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-28Refactor the type legalizer. Switch TargetLowering to a new enum - ↵Nadav Rotem
LegalizeTypeAction. This patch does not change the behavior of the type legalizer. The codegen produces the same code. This infrastructural change is needed in order to enable complex decisions for vector types (needed by the vector-select patch). git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132263 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-28Erase instructions _after_ checking their type.Benjamin Kramer
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132256 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-28Implement and document the llvm.eh.resume intrinsic, which isJohn McCall
transformed by the inliner into a branch to the enclosing landing pad (when inlined through an invoke). If not so optimized, it is lowered DWARF EH preparation into a call to _Unwind_Resume (or _Unwind_SjLj_Resume as appropriate). Its chief advantage is that it takes both the exception value and the selector value as arguments, meaning that there is zero effort in recovering these; however, the frontend is required to pass these down, which is not actually particularly difficult. Also document the behavior of landing pads a bit better, and make it clearer that it's okay that personality functions don't always land at landing pads. This is just a fact of life. Don't write optimizations that rely on pushing things over an unwind edge. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132253 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-28When generating code for Win64 EH, emit StartProc and EndProc directives.Charles Davis
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132250 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-28Create two BlockInfo entries when a live range is discontinuous through a block.Jakob Stoklund Olesen
Delete the Kill and Def markers in BlockInfo. They are no longer necessary when BlockInfo describes a continuous live range. This only affects the relatively rare kind of basic block where a live range looks like this: |---x o---| Now live range splitting can pretend that it is looking at two blocks: |---x o---| This allows the code to be simplified a bit. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132245 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-28Add SplitAnalysis::getNumLiveBlocks().Jakob Stoklund Olesen
It is important that this function returns the same number of live blocks as countLiveBlocks(CurLI) because live range splitting uses the number of live blocks to ensure it is making progress. This is in preparation of supporting duplicate UseBlock entries for basic blocks that have a virtual register live-in and live-out, but not live-though. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132244 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-28Select DW_AT_const_value size based on global variable size.Devang Patel
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132239 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-28Fix the root cause of the bootstrap failure:Rafael Espindola
There was no way to check if a given register/mode pair was valid. We now return an error code (-2) instead of asserting. If anyone thinks that an assert at this point is really needed, we can autogen a hasValidDwarfRegNum instead. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132236 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-27Stub out support for Win64-style exceptions. Note that this is merely usingCharles Davis
the Win64 EH mechanism to implement GCC-style exceptions. LLVM supports hardly anything else at this point! git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132234 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-27Fix a regression I recently introduced by removing DwarfRegNum ofRafael Espindola
subregisters: When a value is in a subregister, at least report the location as being the superregister. We should extend the .td files to encode the bit range so that we can produce a DW_OP_bit_piece. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132224 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-27Make size computation less brittle.Rafael Espindola
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132222 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-27Add the suffix to the Win64 EH data sections' names if given. Add a test forCharles Davis
this. XFAIL'd, because the COFF AsmParser can't handle .section yet. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132220 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-27Refactor getActionType and getTypeToTransformTo ; place all of the 'decision'Nadav Rotem
code in one place. Re-apply 131534 and fix the multi-step promotion of integers. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132217 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-27Keep this simple. Use DIType to get signness and size of a type. Based on ↵Devang Patel
size, select appropraite form. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132206 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-27Handle signed types gracefully.Devang Patel
This fixes regressions reported by buildbots as a fallout of r132193. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132197 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-27Select DW_AT_const_value size based on variable size.Devang Patel
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132193 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-27Fix PR10029 - VerifyCoalescing failure on patterns_dfa.c of 445.gobmk.Cameron Zwarich
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132181 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-26Add comment.Devang Patel
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132149 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-26During branch folding avoid inserting redundant DBG_VALUE machine instructions.Devang Patel
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132148 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-26Revert r132111. I built Release (without Asserts), so I didn't know about theCharles Davis
assert that prevented setting alignment on section creation. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132113 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-26Align Win64 EH Table sections to 4 bytes.Charles Davis
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132111 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-26Reverting 132105: it broke some LLVM-GCC DejaGNU tests.Stuart Hastings
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132108 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-26Correctly handle a one-word struct passed byval on x86_64.Stuart Hastings
rdar://problem/6920088 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132105 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-25Add a RAGreedy::canEvict function.Jakob Stoklund Olesen
This doesn't change functionality (much), but it allows for a more fine-grained eviction policy. The current policy only compares spill weights, and that is not always the best thing to do. Spill weights are designed to serve linear scan, and they don't consider live range splitting. Add a mechanism so canEvict() can request that a live range be evicted and split/spilled. This is to avoid infinite eviction loops. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132101 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-25Rewrite fast-isel integer cast handling to handle more cases, and to be ↵Eli Friedman
simpler and more consistent. The practical effects here are that x86-64 fast-isel can now handle trunc from i8 to i1, and ARM fast-isel can handle many more constructs involving integers narrower than 32 bits (including loads, stores, and many integer casts). rdar://9437928 . git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132099 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-25Remove unused statistical counter.Devang Patel
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132087 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-25Replace the -unwind-tables option with a per function flag. This is moreRafael Espindola
LTO friendly as we can now correctly merge files compiled with or without -fasynchronous-unwind-tables. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@132033 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-24Remove dead code.Devang Patel
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@131974 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-24Explain FIXME.Rafael Espindola
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@131952 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-24Fix the defaults for .eh_frame. We were marking it as writable.Rafael Espindola
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@131951 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-24- Teach SelectionDAG::isKnownNeverZero to return true (op x, c) when c isEvan Cheng
non-zero. - Teach X86 cmov optimization to eliminate the cmov from ctlz, cttz extension when the source of X86ISD::BSR / X86ISD::BSF is proven to be non-zero. rdar://9490949 git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@131948 91177308-0d34-0410-b5e6-96231b3b80d8
2011-05-24Fix debug info for blocks' variable.Devang Patel
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@131940 91177308-0d34-0410-b5e6-96231b3b80d8