summaryrefslogtreecommitdiffstats
path: root/llvm/lib/Transforms
Commit message (Collapse)AuthorAgeFilesLines
...
* [IPO/LowerTypesTest] Skip blockaddress(es) when replacing uses.Davide Italiano2017-11-071-1/+1
| | | | | | | | | | | | | Blockaddresses refer to the function itself, therefore replacing them would cause an assertion in doRAUW. Fixes https://bugs.llvm.org/show_bug.cgi?id=35201 This was found when trying CFI on a proprietary kernel by Dmitry Mikulin. Differential Revision: https://reviews.llvm.org/D39695 llvm-svn: 317527
* InstCombine: salvage the debug info of DCE'ed add instructions.Adrian Prantl2017-11-061-12/+23
| | | | | | rdar://problem/31209283 llvm-svn: 317522
* Revert r317510 "[InstCombine] Pull shifts through a select plus binop with ↵Hans Wennborg2017-11-061-82/+27
| | | | | | | | | | | | | | | | constant" This broke the CodeGen/Hexagon/loop-idiom/pmpy-mod.ll test on a bunch of buildbots. > This pulls shifts through a select+binop with a constant where the select conditionally executes the binop. We already do this for just the binop, but not with the select. > > This can allow us to get the select closer to other selects to enable removing one. > > Differential Revision: https://reviews.llvm.org/D39222 > > git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@317510 91177308-0d34-0410-b5e6-96231b3b80d8 llvm-svn: 317518
* Fix comment /NFCXinliang David Li2017-11-061-3/+4
| | | | llvm-svn: 317514
* [InstCombine] Pull shifts through a select plus binop with constantCraig Topper2017-11-061-27/+82
| | | | | | | | | | This pulls shifts through a select+binop with a constant where the select conditionally executes the binop. We already do this for just the binop, but not with the select. This can allow us to get the select closer to other selects to enable removing one. Differential Revision: https://reviews.llvm.org/D39222 llvm-svn: 317510
* Include already promoted counts when computing SUM for VP.Dehao Chen2017-11-061-14/+12
| | | | | | | | | | | | | | Summary: When computing the SUM for indirect call promotion, if the callsite is already promoted in the profile, it will be promoted before ICP. In the current implementation, ICP only sees remaining counts in SUM. This may cause extra indirect call targets being promoted. This patch updates the SUM to include the counts already promoted earlier. This way we do not end up promoting too many indirect call targets. Reviewers: tejohnson Reviewed By: tejohnson Subscribers: llvm-commits, sanjoy Differential Revision: https://reviews.llvm.org/D38763 llvm-svn: 317502
* [IR] redefine 'UnsafeAlgebra' / 'reassoc' fast-math-flags and add 'trans' ↵Sanjay Patel2017-11-068-47/+45
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | fast-math-flag As discussed on llvm-dev: http://lists.llvm.org/pipermail/llvm-dev/2016-November/107104.html and again more recently: http://lists.llvm.org/pipermail/llvm-dev/2017-October/118118.html ...this is a step in cleaning up our fast-math-flags implementation in IR to better match the capabilities of both clang's user-visible flags and the backend's flags for SDNode. As proposed in the above threads, we're replacing the 'UnsafeAlgebra' bit (which had the 'umbrella' meaning that all flags are set) with a new bit that only applies to algebraic reassociation - 'AllowReassoc'. We're also adding a bit to allow approximations for library functions called 'ApproxFunc' (this was initially proposed as 'libm' or similar). ...and we're out of bits. 7 bits ought to be enough for anyone, right? :) FWIW, I did look at getting this out of SubclassOptionalData via SubclassData (spacious 16-bits), but that's apparently already used for other purposes. Also, I don't think we can just add a field to FPMathOperator because Operator is not intended to be instantiated. We'll defer movement of FMF to another day. We keep the 'fast' keyword. I thought about removing that, but seeing IR like this: %f.fast = fadd reassoc nnan ninf nsz arcp contract afn float %op1, %op2 ...made me think we want to keep the shortcut synonym. Finally, this change is binary incompatible with existing IR as seen in the compatibility tests. This statement: "Newer releases can ignore features from older releases, but they cannot miscompile them. For example, if nsw is ever replaced with something else, dropping it would be a valid way to upgrade the IR." ( http://llvm.org/docs/DeveloperPolicy.html#ir-backwards-compatibility ) ...provides the flexibility we want to make this change without requiring a new IR version. Ie, we're not loosening the FP strictness of existing IR. At worst, we will fail to optimize some previously 'fast' code because it's no longer recognized as 'fast'. This should get fixed as we audit/squash all of the uses of 'isFast()'. Note: an inter-dependent clang commit to use the new API name should closely follow commit. Differential Revision: https://reviews.llvm.org/D39304 llvm-svn: 317488
* [LTO][ThinLTO] Use the linker resolutions to mark global values as dso_local.Sean Fertile2017-11-041-0/+17
| | | | | | | | | | | | | Now that we have a way to mark GlobalValues as local we can use the symbol resolutions that the linker plugin provides as part of lto/thinlto link step to refine the compilers view on what symbols will end up being local. Originally commited as r317374, but reverted in r317395 to update some missed tests. Differential Revision: https://reviews.llvm.org/D35702 llvm-svn: 317408
* Revert "[LTO][ThinLTO] Use the linker resolutions to mark global values ..."Sean Fertile2017-11-041-17/+0
| | | | | | | | | Changes more tests then expected on one of the build bots. reverting to investigate. This reverts https://llvm.org/svn/llvm-project/llvm/trunk@317374 llvm-svn: 317395
* [CallSiteSplitting] clang-format my last commit. NFCI.Davide Italiano2017-11-041-3/+2
| | | | | | Thanks to Rui for pointing out. llvm-svn: 317393
* [CallSiteSplitting] Silence GCC's -Wparentheses. NFCI.Davide Italiano2017-11-031-2/+2
| | | | llvm-svn: 317385
* Invoke salvageDebugInfo from CodeGenPrepare's SinkCast()Adrian Prantl2017-11-031-1/+1
| | | | | | | | | | This preserves the debug info for the cast operation in the original location. rdar://problem/33460652 Reapplied r317340 with the test moved into an ARM-specific directory. llvm-svn: 317375
* [LTO][ThinLTO] Use the linker resolutions to mark global values as dso_local.Sean Fertile2017-11-031-0/+17
| | | | | | | | | | Now that we have a way to mark GlobalValues as local we can use the symbol resolutions that the linker plugin provides as part of lto/thinlto link step to refine the compilers view on what symbols will end up being local. Differential Revision: https://reviews.llvm.org/D35702 llvm-svn: 317374
* [SimplifyCFG] When merging conditional stores, don't count the store we're ↵Craig Topper2017-11-031-1/+3
| | | | | | | | | | | | merging against the PHINodeFoldingThreshold Merging conditional stores tries to check to see if the code is if convertible after the store is moved. But the store hasn't been moved yet so its being counted against the threshold. The patch adds 1 to the threshold comparison to make sure we don't count the store. I've adjusted a test to use a lower threshold to ensure we still do that conversion with the lower threshold. Differential Revision: https://reviews.llvm.org/D39570 llvm-svn: 317368
* Recommit r317351 : Add CallSiteSplitting passJun Bum Lim2017-11-034-0/+501
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This recommit r317351 after fixing a buildbot failure. Original commit message: Summary: This change add a pass which tries to split a call-site to pass more constrained arguments if its argument is predicated in the control flow so that we can expose better context to the later passes (e.g, inliner, jump threading, or IPA-CP based function cloning, etc.). As of now we support two cases : 1) If a call site is dominated by an OR condition and if any of its arguments are predicated on this OR condition, try to split the condition with more constrained arguments. For example, in the code below, we try to split the call site since we can predicate the argument (ptr) based on the OR condition. Split from : if (!ptr || c) callee(ptr); to : if (!ptr) callee(null ptr) // set the known constant value else if (c) callee(nonnull ptr) // set non-null attribute in the argument 2) We can also split a call-site based on constant incoming values of a PHI For example, from : BB0: %c = icmp eq i32 %i1, %i2 br i1 %c, label %BB2, label %BB1 BB1: br label %BB2 BB2: %p = phi i32 [ 0, %BB0 ], [ 1, %BB1 ] call void @bar(i32 %p) to BB0: %c = icmp eq i32 %i1, %i2 br i1 %c, label %BB2-split0, label %BB1 BB1: br label %BB2-split1 BB2-split0: call void @bar(i32 0) br label %BB2 BB2-split1: call void @bar(i32 1) br label %BB2 BB2: %p = phi i32 [ 0, %BB2-split0 ], [ 1, %BB2-split1 ] llvm-svn: 317362
* Add llvm::for_each as a range-based extensions to <algorithm> and make use ↵Aaron Ballman2017-11-031-9/+9
| | | | | | of it in some cases where it is a more clear alternative to std::for_each. llvm-svn: 317356
* Revert "Add CallSiteSplitting pass"Jun Bum Lim2017-11-034-500/+0
| | | | | | | | Revert due to Buildbot failure. This reverts commit r317351. llvm-svn: 317353
* Add CallSiteSplitting passJun Bum Lim2017-11-034-0/+500
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Summary: This change add a pass which tries to split a call-site to pass more constrained arguments if its argument is predicated in the control flow so that we can expose better context to the later passes (e.g, inliner, jump threading, or IPA-CP based function cloning, etc.). As of now we support two cases : 1) If a call site is dominated by an OR condition and if any of its arguments are predicated on this OR condition, try to split the condition with more constrained arguments. For example, in the code below, we try to split the call site since we can predicate the argument (ptr) based on the OR condition. Split from : if (!ptr || c) callee(ptr); to : if (!ptr) callee(null ptr) // set the known constant value else if (c) callee(nonnull ptr) // set non-null attribute in the argument 2) We can also split a call-site based on constant incoming values of a PHI For example, from : BB0: %c = icmp eq i32 %i1, %i2 br i1 %c, label %BB2, label %BB1 BB1: br label %BB2 BB2: %p = phi i32 [ 0, %BB0 ], [ 1, %BB1 ] call void @bar(i32 %p) to BB0: %c = icmp eq i32 %i1, %i2 br i1 %c, label %BB2-split0, label %BB1 BB1: br label %BB2-split1 BB2-split0: call void @bar(i32 0) br label %BB2 BB2-split1: call void @bar(i32 1) br label %BB2 BB2: %p = phi i32 [ 0, %BB2-split0 ], [ 1, %BB2-split1 ] Reviewers: davidxl, huntergr, chandlerc, mcrosier, eraman, davide Reviewed By: davidxl Subscribers: sdesmalen, ashutosh.nema, fhahn, mssimpso, aemerson, mgorny, mehdi_amini, kristof.beyls, llvm-commits Differential Revision: https://reviews.llvm.org/D39137 llvm-svn: 317351
* The patch fixes PR35131Evgeny Stupachenko2017-11-031-3/+3
| | | | | | | | | | | | | Summary: Fix a misprint which led to false CTLZ recognition. Reviewers: craig.topper Differential Revision: https://reviews.llvm.org/D39585 From: Evgeny Stupachenko <evstupac@gmail.com> llvm-svn: 317348
* Revert "Invoke salvageDebugInfo from CodeGenPrepare's SinkCast()"Adrian Prantl2017-11-031-1/+1
| | | | | | This reverts commit 317342 while investigating bot breakage. llvm-svn: 317345
* Invoke salvageDebugInfo from CodeGenPrepare's SinkCast()Adrian Prantl2017-11-031-1/+1
| | | | | | | | This preserves the debug info for the cast operation in the original location. rdar://problem/33460652 llvm-svn: 317340
* [LICM] sink through non-trivially replicable PHIJun Bum Lim2017-11-031-56/+140
| | | | | | | | | | | | | | | | | | | Summary: The current LICM allows sinking an instruction only when it is exposed to exit blocks through a trivially replacable PHI of which all incoming values are the same instruction. This change enhance LICM to sink a sinkable instruction through non-trivially replacable PHIs by spliting predecessors of loop exits. Reviewers: hfinkel, majnemer, davidxl, bmakam, mcrosier, danielcdh, efriedma, jtony Reviewed By: efriedma Subscribers: nemanjai, dberlin, llvm-commits Differential Revision: https://reviews.llvm.org/D37163 llvm-svn: 317335
* [LoopPredication] NFC: Refactored code to separate out functions being reusedAnna Thomas2017-11-031-62/+92
| | | | | | | | | | | | Summary: Refactored the code to separate out common functions that are being reused. This is to reduce the changes for changes coming up wrt loop predication with reverse loops. This refactoring is what we have in our downstream code. llvm-svn: 317324
* [ADCE] Use MapVector for BlockInfo to make iteration order deterministicMikael Holmen2017-11-031-1/+3
| | | | | | | | | | | | | | | | | | | | | | | | | Summary: Also added a reserve() method to MapVector since we want to use that from ADCE. DenseMap does not provide deterministic iteration order so with that we will handle the members of BlockInfo in random order, eventually leading to random order of the blocks in the predecessor lists. Without this change, I get the same predecessor order in about 90% of the time when I compile a certain reproducer and in 10% I get a different one. No idea how to make a proper test case for this. Reviewers: kuhar, david2050 Reviewed By: kuhar Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D39593 llvm-svn: 317323
* [PartialInliner] Skip call sites where inlining fails.Florian Hahn2017-11-031-7/+9
| | | | | | | | | | | | | | | | | | Summary: InlineFunction can fail, for example when trying to inline vararg fuctions. In those cases, we do not want to bump partial inlining counters or set AnyInlined to true, because this could leave an unused function hanging around. Reviewers: davidxl, davide, gyiu Reviewed By: davide Subscribers: llvm-commits, eraman Differential Revision: https://reviews.llvm.org/D39581 llvm-svn: 317314
* [LSR] Clarify a comment. NFC.Vedant Kumar2017-11-031-1/+1
| | | | llvm-svn: 317295
* IndVarSimplify: preserve debug information attached to widened PHI nodes.Adrian Prantl2017-11-021-0/+10
| | | | | | | | | | This fixes PR35015. https://bugs.llvm.org/show_bug.cgi?id=35015 Differential Revision: https://reviews.llvm.org/D39345 llvm-svn: 317282
* Irreducible loop metadata for more accurate block frequency under PGO.Hiroshi Yamauchi2017-11-021-2/+26
| | | | | | | | | | | | | | | | | | | | | | | Summary: Currently the block frequency analysis is an approximation for irreducible loops. The new irreducible loop metadata is used to annotate the irreducible loop headers with their header weights based on the PGO profile (currently this is approximated to be evenly weighted) and to help improve the accuracy of the block frequency analysis for irreducible loops. This patch is a basic support for this. Reviewers: davidxl Reviewed By: davidxl Subscribers: mehdi_amini, llvm-commits, eraman Differential Revision: https://reviews.llvm.org/D39028 llvm-svn: 317278
* [LoopPredication] Enable predication when latchCheckIV is wider than rangeCheckAnna Thomas2017-11-021-10/+96
| | | | | | | | | | | | | | | | | | | | | | Summary: This patch allows us to predicate range checks that have a type narrower than the latch check type. We leverage SCEV analysis to identify a truncate for the latchLimit and latchStart. There is also safety checks in place which requires the start and limit to be known at compile time. We require this to make sure that the SCEV truncate expr for the IV corresponding to the latch does not cause us to lose information about the IV range. Added tests show the loop predication over range checks that are of various types and are narrower than the latch type. This enhancement has been in our downstream tree for a while. Reviewers: apilipenko, sanjoy, mkazantsev Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D39500 llvm-svn: 317269
* Strip off invariant.start because memory locations arent invariantAnna Thomas2017-11-021-9/+33
| | | | | | | | | | | | The original change was reverted in rL317217 because of the failure in the RS4GC testcase. I couldn't reproduce the failure on my local machine (macbook) but could reproduce it on a linux box. The failure was around removing the uses of invariant.start. The fix here is to just RAUW undef (which was the first implementation in D39388). This is perfectly valid IR as discussed in the review. llvm-svn: 317225
* Revert "[RS4GC] Strip off invariant.start because memory locations arent ↵Anna Thomas2017-11-021-39/+9
| | | | | | | | invariant" This reverts commit r317215, investigating the test failure. llvm-svn: 317217
* [RS4GC] Strip off invariant.start because memory locations arent invariantAnna Thomas2017-11-021-9/+39
| | | | | | | | | | | | | | | | | | | | Summary: Invariant.start on memory locations has the property that the memory location is unchanging. However, this is not true in the face of rewriting statepoints for GC. Teach RS4GC about removing invariant.start so that optimizations after RS4GC does not incorrect sink a load from the memory location past a statepoint. Added test showcasing the issue. Reviewers: reames, apilipenko, dneilson Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D39388 llvm-svn: 317215
* Revert "[ExpandMemCmp] Split ExpandMemCmp from CodeGen into its own pass."Clement Courbet2017-11-023-830/+0
| | | | | | | | | undefined reference to `llvm::TargetPassConfig::ID' on clang-ppc64le-linux-multistage This reverts commit eea333c33fa73ad225ef28607795984829f65688. llvm-svn: 317213
* [ExpandMemCmp] Split ExpandMemCmp from CodeGen into its own pass.Clement Courbet2017-11-023-0/+830
| | | | | | | | | | | | | | | | | Summary: This is mostly a noop (most of the test diffs are renamed blocks). There are a few temporary register renames (eax<->ecx) and a few blocks are shuffled around. See the discussion in PR33325 for more details. Reviewers: spatel Subscribers: mgorny Differential Revision: https://reviews.llvm.org/D39456 llvm-svn: 317211
* [SimplifyCFG] Discard speculated dbg intrinsicsBjorn Pettersson2017-11-021-1/+11
| | | | | | | | | | | | | | | | | | | | | | | | | | Summary: SpeculativelyExecuteBB can flatten the CFG by doing speculative execution followed by a select instruction. When the speculatively executed BB contained dbg intrinsics the result could be a little bit weird, since those dbg intrinsics were inserted before the select in the flattened CFG. So when single stepping in the debugger, printing the value of the variable referenced in the dbg intrinsic, it could happen that it looked like the variable had values that never actually were assigned to the variable. This patch simply discards all dbg intrinsics that were found in the speculatively executed BB. Reviewers: aprantl, chandlerc, craig.topper Reviewed By: aprantl Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D39494 llvm-svn: 317198
* loop-unroll: teach remapInstruction to update dbg.value intrinsics.Adrian Prantl2017-11-011-1/+15
| | | | | | | | Fixes PR35112. https://bugs.llvm.org/show_bug.cgi?id=35112 llvm-svn: 317138
* loop-rotate: avoid duplicating dbg.value intrinsics in the entry block.Adrian Prantl2017-11-011-0/+24
| | | | | | | | This fixes the second half of PR35113. This reapplies r317106 without modifications. llvm-svn: 317121
* loop-rotate: eliminate duplicate debug intrinsics after splicing.Adrian Prantl2017-11-011-1/+26
| | | | | | | | | Fixes part of PR35113. This reapplies r317105 with an additional check for isa<Instruction> as found by the bots. llvm-svn: 317120
* Include GUIDs from the same module when computing GUIDs that needs to be ↵Dehao Chen2017-11-011-15/+16
| | | | | | | | | | | | | | | | imported. Summary: In the compile phase of SamplePGO+ThinLTO, ICP is not invoked. Instead, indirect call targets will be included as function metadata for ThinIndex to buidl the call graph. This should not only include functions defined in other modules, but also functions defined in the same module, otherwise ThinIndex may find the callee dead and eliminate it, while ICP in backend will revive the symbol, which leads to undefined symbol. Reviewers: tejohnson Reviewed By: tejohnson Subscribers: sanjoy, llvm-commits, mehdi_amini Differential Revision: https://reviews.llvm.org/D39480 llvm-svn: 317118
* Revert 317016 and 317048Philip Reames2017-11-011-44/+50
| | | | | | The former appears to have introduced a miscompile in a stage2 clang build. Revert so I can investigate offline. llvm-svn: 317116
* Revert r317105 to investigate bot breakage.Adrian Prantl2017-11-011-23/+1
| | | | llvm-svn: 317110
* Revert r317106 to facilitate reverting r317105.Adrian Prantl2017-11-011-24/+0
| | | | llvm-svn: 317109
* LTO: Apply global DCE to ThinLTO modules at LTO opt level 0.Peter Collingbourne2017-11-011-0/+8
| | | | | | | | | | | | | | | | | | | | | | | | | This is necessary because DCE is applied to full LTO modules. Without this change, a reference from a dead ThinLTO global to a dead full LTO global will result in an undefined reference at link time. This problem is only observable when --gc-sections is disabled, or when targeting COFF, as the COFF port of lld requires all symbols to have a definition even if all references are dead (this is consistent with link.exe). This change also adds an EliminateAvailableExternally pass at -O0. This is necessary to handle the situation on Windows where a non-prevailing copy of a linkonce_odr function has an SEH filter function; any such filters must be DCE'd because they will contain a call to the llvm.localrecover intrinsic, passing as an argument the address of the function that the filter belongs to, and llvm.localrecover requires this function to be defined locally. Fixes PR35142. Differential Revision: https://reviews.llvm.org/D39484 llvm-svn: 317108
* loop-rotate: avoid duplicating dbg.value intrinsics in the entry block.Adrian Prantl2017-11-011-0/+24
| | | | | | This fixes the second half of PR35113. llvm-svn: 317106
* loop-rotate: eliminate duplicate debug intrinsics after splicing.Adrian Prantl2017-11-011-1/+23
| | | | | | Fixes part of PR35113. llvm-svn: 317105
* Revert rL311205 "[IRCE] Fix buggy behavior in Clamp"Max Kazantsev2017-11-011-2/+1
| | | | | | | | | | | | | This patch reverts rL311205 that was initially a wrong fix. The real problem was in intersection of signed and unsigned ranges (see rL316552), and the patch being reverted masked the problem instead of fixing it. By now, the test against which rL311205 was made works OK even without this code. This revert patch also contains a test case that demonstrates incorrect behavior caused by rL311205: it is caused by incorrect choise of signed max instead of unsigned. llvm-svn: 317088
* [CodeExtractor] Fix iterator invalidation in findOrCreateBlockForHoisting.Florian Hahn2017-11-011-1/+3
| | | | | | | | | | | | | | | | | | | Summary: By replacing branches to CommonExitBlock, we remove the node from CommonExitBlock's predecessors, invalidating the iterator. The problem is exposed when the common exit block has multiple predecessors and needs to sink lifetime info. The modification in the test case trigger the issue. Reviewers: davidxl, davide, wmi Reviewed By: davidxl Subscribers: llvm-commits Differential Revision: https://reviews.llvm.org/D39112 llvm-svn: 317084
* [SimplifyIndVar] Inline makIVComparisonInvariant to eleminate code ↵Philip Reames2017-10-311-51/+29
| | | | | | | | duplication [NFC] This formulation might be slightly slower since I eagerly compute the cheap replacements. If anyone sees this having a compile time impact, let me know and I'll use lazy population instead. llvm-svn: 317048
* loop-rotate: simplify code by using llvm::findDbgValues(). (NFC)Adrian Prantl2017-10-311-31/+23
| | | | llvm-svn: 317037
* [coro] Make Spill a proper struct instead of deriving from pair.Benjamin Kramer2017-10-311-12/+10
| | | | | | No functionality change. llvm-svn: 317027
OpenPOWER on IntegriCloud