summaryrefslogtreecommitdiffstats
path: root/llvm/lib/CodeGen
Commit message (Collapse)AuthorAgeFilesLines
...
* Fix updateTerminator to be resiliant to degenerate terminators whereChandler Carruth2012-04-161-7/+29
| | | | | | | | | | | | | | | | | | | | | both fallthrough and a conditional branch target the same successor. Gracefully delete the conditional branch and introduce any unconditional branch needed to reach the actual successor. This fixes memory corruption in 2009-06-15-RegScavengerAssert.ll and possibly other tests. Also, while I'm here fix a latent bug I spotted by inspection. I never applied the same fundamental fix to this fallthrough successor finding logic that I did to the logic used when there are no conditional branches. As a consequence it would have selected landing pads had they be aligned in just the right way here. I don't have a test case as I spotted this by inspection, and the previous time I found this required have of TableGen's source code to produce it. =/ I hate backend bugs. ;] Thanks to Jim Grosbach for helping me reason through this and reviewing the fix. llvm-svn: 154867
* Flip the new block-placement pass to be on by default.Chandler Carruth2012-04-161-6/+7
| | | | | | | | | | | | | | | | | | | | | | | | | | This is mostly to test the waters. I'd like to get results from FNT build bots and other bots running on non-x86 platforms. This feature has been pretty heavily tested over the last few months by me, and it fixes several of the execution time regressions caused by the inlining work by preventing inlining decisions from radically impacting block layout. I've seen very large improvements in yacr2 and ackermann benchmarks, along with the expected noise across all of the benchmark suite whenever code layout changes. I've analyzed all of the regressions and fixed them, or found them to be impossible to fix. See my email to llvmdev for more details. I'd like for this to be in 3.1 as it complements the inliner changes, but if any failures are showing up or anyone has concerns, it is just a flag flip and so can be easily turned off. I'm switching it on tonight to try and get at least one run through various folks' performance suites in case SPEC or something else has serious issues with it. I'll watch bots and revert if anything shows up. llvm-svn: 154816
* Add a somewhat hacky heuristic to do something different from whole-loopChandler Carruth2012-04-161-3/+78
| | | | | | | | | | | | | rotation. When there is a loop backedge which is an unconditional branch, we will end up with a branch somewhere no matter what. Try placing this backedge in a fallthrough position above the loop header as that will definitely remove at least one branch from the loop iteration, where whole loop rotation may not. I haven't seen any benchmarks where this is important but loop-blocks.ll tests for it, and so this will be covered when I flip the default. llvm-svn: 154812
* Tweak the loop rotation logic to check whether the loop is naturallyChandler Carruth2012-04-161-11/+51
| | | | | | | | | | | laid out in a form with a fallthrough into the header and a fallthrough out of the bottom. In that case, leave the loop alone because any rotation will introduce unnecessary branches. If either side looks like it will require an explicit branch, then the rotation won't add any, do it to ensure the branch occurs outside of the loop (if possible) and maximize the benefit of the fallthrough in the bottom. llvm-svn: 154806
* Remove dead SD nodes after the combining pass. Fixes PR12201.Hal Finkel2012-04-161-0/+1
| | | | llvm-svn: 154786
* Rewrite how machine block placement handles loop rotation.Chandler Carruth2012-04-161-66/+70
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This is a complex change that resulted from a great deal of experimentation with several different benchmarks. The one which proved the most useful is included as a test case, but I don't know that it captures all of the relevant changes, as I didn't have specific regression tests for each, they were more the result of reasoning about what the old algorithm would possibly do wrong. I'm also failing at the moment to craft more targeted regression tests for these changes, if anyone has ideas, it would be welcome. The first big thing broken with the old algorithm is the idea that we can take a basic block which has a loop-exiting successor and a looping successor and use the looping successor as the layout top in order to get that particular block to be the bottom of the loop after layout. This happens to work in many cases, but not in all. The second big thing broken was that we didn't try to select the exit which fell into the nearest enclosing loop (to which we exit at all). As a consequence, even if the rotation worked perfectly, it would result in one of two bad layouts. Either the bottom of the loop would get fallthrough, skipping across a nearer enclosing loop and thereby making it discontiguous, or it would be forced to take an explicit jump over the nearest enclosing loop to earch its successor. The point of the rotation is to get fallthrough, so we need it to fallthrough to the nearest loop it can. The fix to the first issue is to actually layout the loop from the loop header, and then rotate the loop such that the correct exiting edge can be a fallthrough edge. This is actually much easier than I anticipated because we can handle all the hard parts of finding a viable rotation before we do the layout. We just store that, and then rotate after layout is finished. No inner loops get split across the post-rotation backedge because we check for them when selecting the rotation. That fix exposed a latent problem with our exitting block selection -- we should allow the backedge to point into the middle of some inner-loop chain as there is no real penalty to it, the whole point is that it *won't* be a fallthrough edge. This may have blocked the rotation at all in some cases, I have no idea and no test case as I've never seen it in practice, it was just noticed by inspection. Finally, all of these fixes, and studying the loops they produce, highlighted another problem: in rotating loops like this, we sometimes fail to align the destination of these backwards jumping edges. Fix this by actually walking the backwards edges rather than relying on loopinfo. This fixes regressions on heapsort if block placement is enabled as well as lots of other cases where the previous logic would introduce an abundance of unnecessary branches into the execution. llvm-svn: 154783
* When emulating vselect using OR/AND/XOR make sure to bitcast the result back ↵Nadav Rotem2012-04-151-1/+2
| | | | | | to the original type. llvm-svn: 154764
* misched: Added CanHandleTerminators.Andrew Trick2012-04-131-3/+3
| | | | | | | | | This is a special flag for targets that really want their block terminators in the DAG. The default scheduler cannot handle this correctly, so it becomes the specialized scheduler's responsibility to schedule terminators. llvm-svn: 154712
* Reduce malloc traffic in DwarfAccelTableBenjamin Kramer2012-04-132-58/+28
| | | | | | | | | - Don't copy offsets into HashData, the underlying vector won't change once the table is finalized. - Allocate HashData and HashDataContents in a BumpPtrAllocator. - Allocate string map entries in the same allocator. - Random cleanups. llvm-svn: 154694
* HexagonPacketizer patch.Sirish Pande2012-04-121-55/+34
| | | | llvm-svn: 154616
* Reapply 154397. Original message:Nadav Rotem2012-04-111-11/+18
| | | | | | | | Fix a dagcombine optimization which assumes that the vsetcc result type is always of the same size as the compared values. This is ture for SSE/AVX/NEON but not for all targets. llvm-svn: 154490
* Fix an overly indented line. Remove an 'else' after an 'if' that returns.Craig Topper2012-04-111-4/+3
| | | | llvm-svn: 154479
* Inline implVisitAluOverflow by introducing a nested switch to convert the ↵Craig Topper2012-04-112-20/+18
| | | | | | intrinsic to an nodetype. llvm-svn: 154478
* Optimize code a bit by calling push_back only once in some loops. Reduces ↵Craig Topper2012-04-111-26/+24
| | | | | | compiled code size a bit. llvm-svn: 154473
* Tweak MachineLICM heuristics for cheap instructions.Jakob Stoklund Olesen2012-04-111-69/+89
| | | | | | | | | | | Allow cheap instructions to be hoisted if they are register pressure neutral or better. This happens if the instruction is the last loop use of another virtual register. Only expensive instructions are allowed to increase loop register pressure. llvm-svn: 154455
* Only check for PHI uses inside the current loop.Jakob Stoklund Olesen2012-04-111-27/+51
| | | | | | | | | | | Hoisting a value that is used by a PHI in the loop will introduce a copy because the live range is extended to cross the PHI. The same applies to PHIs in exit blocks. Also use this opportunity to make HasLoopPHIUse() non-recursive. llvm-svn: 154454
* Move the constant-folding support for FP_ROUND in SelectionDAG from the ↵Owen Anderson2012-04-101-1/+10
| | | | | | | | one-operand version of getNode() to the two-operand version, since it became a two-operand node at sound point. Zap a testcase that this allows us to completely fold away. llvm-svn: 154447
* Add a comment noting that the fdiv -> fmul conversion won't generateDuncan Sands2012-04-101-3/+3
| | | | | | multiplication by a denormal, and some tests checking that. llvm-svn: 154431
* To ensure that we have more accurate line information for a blockEric Christopher2012-04-101-2/+5
| | | | | | | | | don't elide the branch instruction if it's the only one in the block, otherwise it's ok. PR9796 and rdar://11215207 llvm-svn: 154417
* Revert r154397, which was causing make check failures on the buildbots.Owen Anderson2012-04-101-13/+6
| | | | llvm-svn: 154414
* Fix a dagcombine optimization which assumes that the vsetcc result type is ↵Nadav Rotem2012-04-101-6/+13
| | | | | | | | | always of the same size as the compared values. This is ture for SSE/AVX/NEON but not for all targets. llvm-svn: 154397
* Make a somewhat subtle change in the logic of block placement. SometimesChandler Carruth2012-04-101-0/+12
| | | | | | | | | | | | | | | | the loop header has a non-loop predecessor which has been pre-fused into its chain due to unanalyzable branches. In this case, rotating the header into the body of the loop in order to place a loop exit at the bottom of the loop is a Very Bad Idea as it makes the loop non-contiguous. I'm working on a good test case for this, but it's a bit annoynig to craft. I should get one shortly, but I'm submitting this now so I can begin the (lengthy) performance analysis process. An initial run of LNT looks really, really good, but there is too much noise there for me to trust it much. llvm-svn: 154395
* Transform div to mul with reciprocal only when fp imm is legal.Anton Korobeynikov2012-04-101-2/+9
| | | | | | This fixes PR12516 and uncovers one weird problem in legalize (workarounded) llvm-svn: 154394
* Make the code slightly more palatable.Evan Cheng2012-04-101-1/+5
| | | | llvm-svn: 154378
* Fix a long standing tail call optimization bug. When a libcall is emittedEvan Cheng2012-04-102-9/+10
| | | | | | | | | | | | | legalizer always use the DAG entry node. This is wrong when the libcall is emitted as a tail call since it effectively folds the return node. If the return node's input chain is not the entry (i.e. call, load, or store) use that as the tail call input chain. PR12419 rdar://9770785 rdar://11195178 llvm-svn: 154370
* Don't try to zExt just to check if an integer constant is zero, it mightRafael Espindola2012-04-101-2/+2
| | | | | | not fit in a i64. llvm-svn: 154364
* Have TargetLowering::getPICJumpTableRelocBase return a node that points to theAkira Hatanaka2012-04-091-1/+5
| | | | | | GOT if jump table uses 64-bit gp-relative relocation. llvm-svn: 154341
* Patch r153892 for PR11861 apparently broke an external project (see PR12493).Lang Hames2012-04-091-16/+17
| | | | | | | | | | This patch restores TwoAddressInstructionPass's pre-r153892 behaviour when rescheduling instructions in TryInstructionTransform. Hopefully this will fix PR12493. To refix PR11861, lowering of INSERT_SUBREGS is deferred until after the copy that unties the operands is emitted (this seems to be a more appropriate fix for that issue anyway). llvm-svn: 154338
* Pattern match a setcc of boolean value with 0 as a truncate.Rafael Espindola2012-04-091-9/+48
| | | | llvm-svn: 154322
* Remove unnecessary type check when combining and/or/xor of swizzles. Move ↵Craig Topper2012-04-091-13/+12
| | | | | | some checks to allow better early out. llvm-svn: 154309
* Remove unnecessary 'else' on an 'if' that always returnsCraig Topper2012-04-091-1/+2
| | | | llvm-svn: 154308
* Optimize code slightly. No functionality change.Craig Topper2012-04-091-6/+7
| | | | llvm-svn: 154307
* Replace some explicit checks with asserts for conditions that should never ↵Craig Topper2012-04-091-14/+7
| | | | | | happen. llvm-svn: 154305
* Optimize code a bit. No functional change intended.Craig Topper2012-04-081-9/+9
| | | | llvm-svn: 154299
* Silence sign-compare warning.Benjamin Kramer2012-04-081-1/+1
| | | | llvm-svn: 154297
* Only have codegen turn fdiv by a constant into fmul by the reciprocalDuncan Sands2012-04-081-5/+3
| | | | | | | | when -ffast-math, i.e. don't just always do it if the reciprocal can be formed exactly. There is already an IR level transform that does that, and it does it more carefully. llvm-svn: 154296
* Simplify code that tries to do vector extracts for shuffles when the mask ↵Craig Topper2012-04-081-12/+8
| | | | | | width and the input vector widths don't match. No need to check the min and max are in range before calculating the start index. The range check after having the start index is sufficient. Also no need to check for an extract from the beginning differently. llvm-svn: 154295
* Move the TLSModel information into the TargetMachine rather than hidingChandler Carruth2012-04-081-22/+0
| | | | | | | | in TargetLowering. There was already a FIXME about this location being odd. The interface is simplified as a consequence. This will also make it easier to change TLS models when compiling with PIE. llvm-svn: 154292
* Remove an over zealous assert. The assert was trying to catch placesChandler Carruth2012-04-081-1/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | where a chain outside of the loop block-set ended up in the worklist for scheduling as part of the contiguous loop. However, asserting the first block in the chain is in the loop-set isn't a valid check -- we may be forced to drag a chain into the worklist due to one block in the chain being part of the loop even though the first block is *not* in the loop. This occurs when we have been forced to form a chain early due to un-analyzable branches. No test case here as I have no idea how to even begin reducing one, and it will be hopelessly fragile. We have to somehow end up with a loop header of an inner loop which is a successor of a basic block with an unanalyzable pair of branch instructions. Ow. Self-host triggers it so it is unlikely it will regress. This at least gets block placement back to passing selfhost and the test suite. There are still a lot of slowdown that I don't like coming out of block placement, although there are now also a lot of speedups. =[ I'm seeing swings in both directions up to 10%. I'm going to try to find time to dig into this and see if we can turn this on for 3.1 as it does a really good job of cleaning up after some loops that degraded with the inliner changes. llvm-svn: 154287
* Add a debug-only 'dump' method to the BlockChain structure to easeChandler Carruth2012-04-081-0/+8
| | | | | | debugging. llvm-svn: 154286
* Turn avx2 vinserti128 intrinsic calls into INSERT_SUBVECTOR DAG nodes and ↵Craig Topper2012-04-071-1/+2
| | | | | | remove patterns for selecting the intrinsic. Similar was already done for avx1. llvm-svn: 154272
* Remove 'else' after 'if' that ends in return.Craig Topper2012-04-071-1/+1
| | | | llvm-svn: 154267
* 1. Remove the part of r153848 which optimizes shuffle-of-shuffle into a newNadav Rotem2012-04-071-12/+14
| | | | | | | | | | shuffle node because it could introduce new shuffle nodes that were not supported efficiently by the target. 2. Add a more restrictive shuffle-of-shuffle optimization for cases where the second shuffle reverses the transformation of the first shuffle. llvm-svn: 154266
* Convert floating point division by a constant into multiplication by theDuncan Sands2012-04-071-0/+13
| | | | | | | | reciprocal if converting to the reciprocal is exact. Do it even if inexact if -ffast-math. This substantially speeds up ac.f90 from the polyhedron benchmarks. llvm-svn: 154265
* Patch to set is_stmt a little better for prologue lines in a function.Eric Christopher2012-04-051-2/+5
| | | | | | | | | This enables debuggers to see what are interesting lines for a breakpoint rather than any line that starts a function. rdar://9852092 llvm-svn: 154120
* Don't break the IV update in TLI::SimplifySetCC().Jakob Stoklund Olesen2012-04-051-15/+27
| | | | | | | | | | | | | | | | | | | LSR always tries to make the ICmp in the loop latch use the incremented induction variable. This allows the induction variable to be kept in a single register. When the induction variable limit is equal to the stride, SimplifySetCC() would break LSR's hard work by transforming: (icmp (add iv, stride), stride) --> (cmp iv, 0) This forced us to use lea for the IC update, preventing the simpler incl+cmp. <rdar://problem/7643606> <rdar://problem/11184260> llvm-svn: 154119
* Treat f16 the same as f80/f128 for the purposes of generating constants ↵Owen Anderson2012-04-051-1/+2
| | | | | | during instruction selection. llvm-svn: 154113
* REG_SEQUENCE expansion to COPY instructions wasn't taking account of sub ↵Pete Cooper2012-04-041-1/+2
| | | | | | register indices on the source registers. No simple test case llvm-svn: 154051
* f16 FREM can now be legalized by promoting to f32Pete Cooper2012-04-041-0/+1
| | | | llvm-svn: 154039
* Remove spurious debug output.Jakob Stoklund Olesen2012-04-041-1/+0
| | | | llvm-svn: 154032
OpenPOWER on IntegriCloud