summaryrefslogtreecommitdiffstats
path: root/llvm/lib/Transforms
Commit message (Collapse)AuthorAgeFilesLines
* [LoopSimplify] Set proper debug location in loop backedge blocks.Alexey Samsonov2015-06-291-1/+2
| | | | | | | | | | | | | Set debug location for terminator instruction in loop backedge block (which is an unconditional jump to loop header). We can't copy debug location from original backedges, as there can be several of them, with different debug info locations. So, we follow the approach of SplitBlockPredecessors, and copy the debug info from first non-PHI instruction in the header (i.e. destination block). This is yet another change for PR23837. llvm-svn: 240999
* Tidy comment.Diego Novillo2015-06-291-1/+1
| | | | llvm-svn: 240987
* [SLSR] S's basis must have the same type as SJingyue Wu2015-06-281-2/+4
| | | | llvm-svn: 240910
* [LoopVectorize] Pointer indicies may be wider than the pointerDavid Majnemer2015-06-271-3/+10
| | | | | | | | | | | If we are dealing with a pointer induction variable, isInductionPHI gives back a step value of Stride / size of pointer. However, we might be indexing with a legal type wider than the pointer width. Handle this by inserting casts where appropriate instead of crashing. This fixes PR23954. llvm-svn: 240877
* [PruneEH] A naked, noinline function can return via InlineAsmDavid Majnemer2015-06-271-25/+38
| | | | | | | | | | The PruneEH pass tries to annotate functions as 'noreturn' if it doesn't see a ReturnInst. However, a naked function containing inline assembly can contain control flow leaving the function. This fixes PR23971. llvm-svn: 240876
* LowerBitSets: Ignore bitset entries that do not directly refer to a global.Peter Collingbourne2015-06-271-3/+7
| | | | | | | | | | | | | | | | | It is possible for a global to be substituted with another global of a different type or a different kind (i.e. an alias) at IR link time. One example of this scenario is when a Microsoft ABI vtable is substituted with an alias referring to a larger vtable containing an RTTI reference. This will cause the global to be RAUW'd with a possibly bitcasted reference to the other global. This will of course also affect any references to the global in bitset metadata. The right way to handle such metadata is simply to ignore it. This is sound because the linked module should contain another copy of the bitset entries as applied to the new global. llvm-svn: 240866
* [RewriteStatepointsForGC] Generalized vector phi/select handling for base ↵Philip Reames2015-06-261-55/+124
| | | | | | | | | | | | | | | pointers This change extends the detection of base pointers for vector constructs to handle arbitrary phi and select nodes. The existing non-vector code already handles those, so this is basically just extending the vector special case to be less special cased. It still isn't generalized vector handling since we can't handle arbitrary vector instructions (e.g. shufflevectors), but it's a lot closer. The general structure of the change is as follows: * Extend the base defining value relation over a subset of vector instructions and vector typed phi & select instructions. * Move scalarization from before base pointer rewriting to after base pointer rewriting. The extension of the BDV relation is sufficient to find vector base phis for vector inputs. * Preserve the existing special case logic for when the base of a vector element is locally obvious. This general idea could be extended to the scalar case as well. Differential Revision: http://reviews.llvm.org/D10461#inline-84275 llvm-svn: 240850
* Move VectorUtils from Transforms to Analysis to correct layering violationDavid Blaikie2015-06-264-216/+2
| | | | llvm-svn: 240804
* Fix ODR violation waiting to happen by making static function definitions in ↵David Blaikie2015-06-262-0/+214
| | | | | | | | | | VectorUtils.h non-static and defined out of line Patch by Ashutosh Nema Differential Revision: http://reviews.llvm.org/D10682 llvm-svn: 240794
* [ASan] Use llvm::getDISubprogram() to get function entry debug location.Alexey Samsonov2015-06-261-7/+3
| | | | | | | | It can be more robust than copying debug info from first non-alloca instruction in the entry basic block. We use the same strategy in coverage instrumentation. llvm-svn: 240738
* [asan] Do not instrument special purpose LLVM sections.Anna Zaks2015-06-251-0/+2
| | | | | | | | | | | | | | Do not instrument globals that are placed in sections containing "__llvm" in their name. This fixes a bug in ASan / PGO interoperability. ASan interferes with LLVM's PGO, which places its globals into a special section, which is memcpy-ed by the linker as a whole. When those goals are instrumented, ASan's memcpy wrapper reports an issue. http://reviews.llvm.org/D10541 llvm-svn: 240723
* [asan] Don't run stack malloc on functions containing inline assembly.Anna Zaks2015-06-251-4/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | It makes LLVM run out of registers even on 64-bit platforms. For example, the following test case fails on darwin. clang -cc1 -O0 -triple x86_64-apple-macosx10.10.0 -emit-obj -fsanitize=address -mstackrealign -o ~/tmp/ex.o -x c ex.c error: inline assembly requires more registers than available void TestInlineAssembly(const unsigned char *S, unsigned int pS, unsigned char *D, unsigned int pD, unsigned int h) { unsigned int sr = 4, pDiffD = pD - 5; unsigned int pDiffS = (pS << 1) - 5; char flagSA = ((pS & 15) == 0), flagDA = ((pD & 15) == 0); asm volatile ( "mov %0, %%"PTR_REG("si")"\n" "mov %2, %%"PTR_REG("cx")"\n" "mov %1, %%"PTR_REG("di")"\n" "mov %8, %%"PTR_REG("ax")"\n" : : "m" (S), "m" (D), "m" (pS), "m" (pDiffS), "m" (pDiffD), "m" (sr), "m" (flagSA), "m" (flagDA), "m" (h) : "%"PTR_REG("si"), "%"PTR_REG("di"), "%"PTR_REG("ax"), "%"PTR_REG("cx"), "%"PTR_REG("dx"), "memory" ); } http://reviews.llvm.org/D10719 llvm-svn: 240722
* Use foreach loop over constant operands. NFC.Pete Cooper2015-06-252-7/+5
| | | | | | | A number of places had explicit loops over Constant::operands(). Just use foreach loops where possible. llvm-svn: 240694
* [InstCombine] call SimplifyICmpInst with correct contextJingyue Wu2015-06-251-2/+4
| | | | | | | | | | | | | | | | | | | | | Summary: Fixes PR23809. Without passing the context to SimplifyICmpInst, we would use the assume to prove that the condition feeding the assume is trivially true (see isValidAssumeForContext in ValueTracking.cpp), causing the removal of the assume which may be useful for later optimizations. Test Plan: pr23800.ll Reviewers: hfinkel, majnemer Reviewed By: hfinkel Subscribers: henryhu, llvm-commits, wengxt, broune, meheff, eliben Differential Revision: http://reviews.llvm.org/D10695 llvm-svn: 240683
* Rangify for loop in Inliner.cpp. NFC.Yaron Keren2015-06-251-8/+5
| | | | llvm-svn: 240678
* GVN: If a branch has two identical successors, we cannot declare either dead.Peter Collingbourne2015-06-251-0/+4
| | | | | | | | | This previously caused miscompilations as a result of phi nodes receiving undef incoming values from blocks dominated by such successors. Differential Revision: http://reviews.llvm.org/D10726 llvm-svn: 240670
* Teach LLVM about the PPC64 memory sanitizer implementation.Jay Foad2015-06-251-0/+17
| | | | | | | | | | | | | | | | Summary: This is the LLVM part of the PPC memory sanitizer implementation in D10648. Reviewers: kcc, samsonov, willschm, wschmidt, eugenis Reviewed By: eugenis Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D10649 llvm-svn: 240627
* Add simplify_type<const WeakVH>; simplify IndVarSimplifyDuncan P. N. Exon Smith2015-06-241-4/+3
| | | | | | | | r240214 fixed some UB in IndVarSimplify, and it needed a temporary `WeakVH` to do it. Add `simplify_type<const WeakVH>` so that this temporary isn't necessary. llvm-svn: 240599
* [GVN] Intersect the IR flags when CSE'ing two instructionsDavid Majnemer2015-06-241-7/+3
| | | | | | | | | We performed a simple, but incomplete, intersection when it came time to CSE instructions. It didn't handle, for example, the 'exact' flag. This fixes PR23922. llvm-svn: 240595
* [Reassociate] Don't propogate flags when creating negationsDavid Majnemer2015-06-241-0/+10
| | | | | | | | | Reassociate mutated existing instructions in order to form negations which would create additional reassociate opportunities. This fixes PR23926. llvm-svn: 240593
* fix typos; NFCSanjay Patel2015-06-241-3/+2
| | | | llvm-svn: 240592
* don't repeat function names in comments; NFCSanjay Patel2015-06-242-106/+91
| | | | llvm-svn: 240591
* fix typos; NFCSanjay Patel2015-06-241-2/+2
| | | | llvm-svn: 240585
* [LoopVectorizer] Fix bailing-out condition for OptForSize case.Michael Zolotukhin2015-06-241-4/+3
| | | | | | | | | | With option OptForSize enabled, the Loop Vectorizer is not supposed to create tail loop. The condition checking that was invalid and was not matching to the comment above. Patch by Marianne Mailhot-Sarrasin. llvm-svn: 240556
* fix typo; NFCSanjay Patel2015-06-231-1/+1
| | | | llvm-svn: 240480
* don't repeat function names in comments; NFCSanjay Patel2015-06-231-18/+15
| | | | llvm-svn: 240478
* Let llvm::ReplaceInstWithInst copy debug location from old to new instruction.Alexey Samsonov2015-06-231-2/+5
| | | | | | | | | | | | | | Currently some users of this function do this explicitly, and all the rest forget to do this. ThreadSanitizer was one of such users, and had missing debug locations for calls into TSan runtime handling atomic operations, eventually leading to poorly symbolized stack traces and malfunctioning suppressions. This is another change relevant to PR23837. llvm-svn: 240460
* This change fixes three bugs in loop unswitching. This change causes an 81% ↵Mark Heffernan2015-06-231-40/+65
| | | | | | | | | | | | | | | speed-up on a benchmark that is based on EigenConvolutionKernel2D from Eigen3, where the lack of loop unswitching blocks hoisting of loads out of a nested loop (see bug 23816 for how loop unswitching and load hoisting are related). Change 1: Unswitching on trivial conditions should always happen regardless of the computed unswitching cost, as really the cost is zero. While there is code to make that happen, the logic that checks the unswitching cost against a threshold was moved to an earlier point (revision 147935) than the point where trivial unswitching is detected, so trivial unswitching is currently blocked by the cost threshold. This change fixes that. Change 2: Before revision 147935 (from 2012-01-11), the threshold parameter was a per-loop threshold. So an unswitching happened only if the cost of the unswitching was less than the threshold. In an indirect way (and I believe unintentionally), the logic for this since then has been that the threshold is an over-all budget across all loops for all loop unswitching done by a given LoopUnswitch loop pass object. So if an unswitching with cost 100 happens in one function, that in effect reduces the threshold from 100 to 0 for the loops even in another function. This persists for the lifetime of that loop pass object. This makes no difference for most small examples but it is important for large examples. This revision fixes that. Change 3: The cost is currently calculated as std::min(NumInstructions, 5 * NumBlocks). So a loop with 2 blocks and a million instructions will have an unswitching cost of 10. I changed this to just NumInstructions, as it were before revision 147935, though I'm open to e.g. instead replacing std::min with std::max. I've tried to make the change minimally invasive while staying with what I think was the original intent of the code. Submitted on behalf of broune@. llvm-svn: 240438
* Revert r240137 (Fixed/added namespace ending comments using clang-tidy. NFC)Alexander Kornienko2015-06-2386-131/+131
| | | | | | Apparently, the style needs to be agreed upon first. llvm-svn: 240390
* Fix PR13851: Preserve metadata for the unswitched branchWeiming Zhao2015-06-231-20/+67
| | | | | | | This patch copies the metadata of the unswitched branch to the newly crreated branch in loop unswitch pass. llvm-svn: 240378
* [InstCombine] Optimize subtract of selects into a select of a subDavid Majnemer2015-06-231-0/+26
| | | | | | | This came up when examining some code generated by clang's IRGen for certain member pointers. llvm-svn: 240369
* [LoopDist] Improve variable names and comments in LoopVersioning class, NFCAdam Nemet2015-06-221-25/+29
| | | | | | | | As with the previous patch, the goal is to turn the class into a general loop-versioning class. This patch removes any references to loop distribution. llvm-svn: 240352
* SafeStack: Create the unsafe stack pointer on demand.Peter Collingbourne2015-06-221-3/+4
| | | | | | | This avoids creating an unnecessary undefined reference on targets such as NVPTX that require such references to be declared in asm output. llvm-svn: 240321
* [PM/AA] Hoist the AliasResult enum out of the AliasAnalysis class.Chandler Carruth2015-06-223-15/+14
| | | | | | | | | | | | | | | | | | | | | | | | | | | This will allow classes to implement the AA interface without deriving from the class or referencing an internal enum of some other class as their return types. Also, to a pretty fundamental extent, concepts such as 'NoAlias', 'MayAlias', and 'MustAlias' are first class concepts in LLVM and we aren't saving anything by scoping them heavily. My mild preference would have been to use a scoped enum, but that feature is essentially completely broken AFAICT. I'm extremely disappointed. For example, we cannot through any reasonable[1] means construct an enum class (or analog) which has scoped names but converts to a boolean in order to test for the possibility of aliasing. [1]: Richard Smith came up with a "solution", but it requires class templates, and lots of boilerplate setting up the enumeration multiple times. Something like Boost.PP could potentially bundle this up, but even that would be quite painful and it doesn't seem realistically worth it. The enum class solution would probably work without the need for a bool conversion. Differential Revision: http://reviews.llvm.org/D10495 llvm-svn: 240255
* [SwitchLowering] Remove quadratic vector removal.Benjamin Kramer2015-06-201-12/+12
| | | | | | | This can be triggered with giant switches. No functionality change intended. llvm-svn: 240221
* Rangify for loops in Inliner::runOnSCC(), NFC.Yaron Keren2015-06-201-11/+9
| | | | llvm-svn: 240215
* IndVarSimplify: Avoid UB from binding a reference to a null pointerJustin Bogner2015-06-201-3/+4
| | | | | | | | | | Calling operator* on a WeakVH whose Value is null hits undefined behaviour, since we bind the value to a reference. Instead, go through `operator Value*` so that we work with the pointer itself. Found by ubsan. llvm-svn: 240214
* LowerSwitch: Avoid some undefined behaviourJustin Bogner2015-06-201-1/+2
| | | | | | | | | | | | When a case of INT64_MIN was followed by a case that was greater than zero, we were overflowing a signed integer here. Since we've sorted the cases here anyway (and thus currentValue must be greater than nextValue) it's simple enough to avoid this by using addition rather than subtraction. Found by UBSAN on existing tests. llvm-svn: 240201
* [LoopDist] Rename RuntimeCheckEmitter to LoopVersioning, NFCAdam Nemet2015-06-191-8/+8
| | | | llvm-svn: 240165
* [LoopDist] Move pointer-to-partition computation out of RuntimeCheckEmitter, NFCAdam Nemet2015-06-191-18/+12
| | | | | | | This starts preparing the class to become a (more) general LoopVersioning utility class. llvm-svn: 240164
* [SLP] Vectorize for all-constant entries.Michael Zolotukhin2015-06-191-2/+4
| | | | | | Differential Revision: http://reviews.llvm.org/D10531 llvm-svn: 240144
* Fixed/added namespace ending comments using clang-tidy. NFCAlexander Kornienko2015-06-1986-131/+131
| | | | | | | | | | | | | The patch is generated using this command: tools/clang/tools/extra/clang-tidy/tool/run-clang-tidy.py -fix \ -checks=-*,llvm-namespace-comment -header-filter='llvm/.*|clang/.*' \ llvm/lib/ Thanks to Eugene Kosov for the original patch! llvm-svn: 240137
* [ASan] Initial support for Kernel AddressSanitizerAlexander Potapenko2015-06-191-42/+61
| | | | | | | | | This patch adds initial support for the -fsanitize=kernel-address flag to Clang. Right now it's quite restricted: only out-of-line instrumentation is supported, globals are not instrumented, some GCC kasan flags are not supported. Using this patch I am able to build and boot the KASan tree with LLVMLinux patches from github.com/ramosian-glider/kasan/tree/kasan_llvmlinux. To disable KASan instrumentation for a certain function attribute((no_sanitize("kernel-address"))) can be used. llvm-svn: 240131
* Fix "the the" in comments.Eric Christopher2015-06-194-4/+4
| | | | llvm-svn: 240112
* [EliminateDuplicatePHINodes] Replace custom hash map with DenseSet.Benjamin Kramer2015-06-181-51/+34
| | | | | | | While there use hash_combine instead of hand-rolled hashing. No functionality change intended. llvm-svn: 240023
* [NFC] more comments in SLSRJingyue Wu2015-06-181-0/+9
| | | | llvm-svn: 239984
* Move the personality function from LandingPadInst to FunctionDavid Majnemer2015-06-175-33/+27
| | | | | | | | | | | | | | | | | | | The personality routine currently lives in the LandingPadInst. This isn't desirable because: - All LandingPadInsts in the same function must have the same personality routine. This means that each LandingPadInst beyond the first has an operand which produces no additional information. - There is ongoing work to introduce EH IR constructs other than LandingPadInst. Moving the personality routine off of any one particular Instruction and onto the parent function seems a lot better than have N different places a personality function can sneak onto an exceptional function. Differential Revision: http://reviews.llvm.org/D10429 llvm-svn: 239940
* LowerBitSets: Do not assign names to aliases of unnamed bitset element objects.Peter Collingbourne2015-06-171-3/+2
| | | | | | | The restriction on unnamed aliases was removed in r239921. Mostly reverts r239590, but we keep the test. llvm-svn: 239923
* AVX-512: cvtusi2ss/d intrinsics.Igor Breger2015-06-171-0/+2
| | | | | | | | | Change builtin function name and signature ( add third parameter - rounding mode ). Added tests for intrinsics. Differential Revision: http://reviews.llvm.org/D10473 llvm-svn: 239888
* [PM/AA] Remove the UnknownSize static member from AliasAnalysis.Chandler Carruth2015-06-173-8/+8
| | | | | | | | This is now living in MemoryLocation, which is what it pertains to. It is also an enum there rather than a static data member which is left never defined. llvm-svn: 239886
OpenPOWER on IntegriCloud