summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* [VPlan] Add moveAfter to VPRecipeBase.Florian Hahn2019-10-113-0/+36
| | | | | | | | | | | | | This patch adds a moveAfter method to VPRecipeBase, which can be used to move elements after other elements, across VPBasicBlocks, if necessary. Reviewers: dcaballe, hsaito, rengolin, hfinkel Reviewed By: dcaballe Differential Revision: https://reviews.llvm.org/D46825 llvm-svn: 374565
* [AIX] Use .space instead of .zero in assemblyDavid Tenty2019-10-112-0/+18
| | | | | | | | | | | | | | Summary: The AIX system assembler does not understand .zero, so we should prefer emitting .space. Subscribers: nemanjai, hiraditya, kbarton, MaskRay, jsji, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D68815 llvm-svn: 374564
* Reland r374450 with Richard Smith's comments and test fixed.Erich Keane2019-10-1113-41/+46
| | | | | | | | | | The behavior from the original patch has changed, since we're no longer allowing LLVM to just ignore the alignment. Instead, we're just assuming the maximum possible alignment. Differential Revision: https://reviews.llvm.org/D68824 llvm-svn: 374562
* [AMDGPU][MC][GFX9][GFX10] Corrected number of src operands for ↵Dmitry Preobrazhensky2019-10-115-63/+76
| | | | | | | | | | | | ds_[read/write]_addtid_b32 See https://bugs.llvm.org/show_bug.cgi?id=37941 Reviewers: arsenm, rampitec Differential Revision: https://reviews.llvm.org/D68787 llvm-svn: 374561
* gn build: Merge r374558GN Sync Bot2019-10-111-0/+1
| | | | llvm-svn: 374560
* [AMDGPU][MC][GFX6][GFX7][GFX10] Added instructions ↵Dmitry Preobrazhensky2019-10-114-14/+152
| | | | | | | | | | | | buffer_atomic_[fcmpswap/fmin/fmax]* See https://bugs.llvm.org/show_bug.cgi?id=28232 Reviewers: arsenm, rampitec Differential Revision: https://reviews.llvm.org/D68788 llvm-svn: 374559
* [libTooling] Move `RewriteRule` abstraction into its own header and impl.Yitzhak Mandelbaum2019-10-115-438/+473
| | | | | | | | | | | | | | | | | | | Summary: Move the `RewriteRule` class and related declarations into its own set of files (header, implementation). Only the `Transformer` class is left in the Transformer-named files. This change clarifies the distinction between the `RewriteRule` class, which is essential to the Transformer library, and the `Transformer` class, which is only one possible `RewriteRule` interpreter (compare to `TransformerClangTidyCheck`, a clang-tidy based interpreter). Reviewers: gribozavr Subscribers: jfb, cfe-commits Tags: #clang Differential Revision: https://reviews.llvm.org/D68795 llvm-svn: 374558
* [AMDGPU][MC][GFX10] Enabled null for 64-bit dst operandsDmitry Preobrazhensky2019-10-115-1/+38
| | | | | | | | | | See https://bugs.llvm.org/show_bug.cgi?id=43524 Reviewers: arsenm, rampitec Differential Revision: https://reviews.llvm.org/D68785 llvm-svn: 374557
* [llvm] [ocaml] Support linking against dylibMichal Gorny2019-10-111-14/+18
| | | | | | | | | Support linking OCaml modules against LLVM dylib when requested, rather than against static libs that might not be installed at all. Differential Revision: https://reviews.llvm.org/D68452 llvm-svn: 374556
* [DAGCombiner] fold vselect-of-constants to shiftSanjay Patel2019-10-113-28/+22
| | | | | | | | | | The diffs suggest that we are missing some more basic analysis/transforms, but this keeps the vector path in sync with the scalar (rL374397). This is again a preliminary step for introducing the reverse transform in IR as proposed in D63382. llvm-svn: 374555
* Fix compilation warnings. NFC.Michael Liao2019-10-112-2/+2
| | | | llvm-svn: 374554
* [AMDGPU][MC] Corrected parsing of optional operandsDmitry Preobrazhensky2019-10-112-12/+11
| | | | | | | | | | See https://bugs.llvm.org/show_bug.cgi?id=43486 Reviewers: artem.tamazov, arsenm Differential Revision: https://reviews.llvm.org/D68350 llvm-svn: 374553
* [libTooling] Change Stencil equality to use `toString()`Yitzhak Mandelbaum2019-10-113-112/+15
| | | | | | | | | | | | | | | | | | | | | | | | | Summary: Removes the `isEqual` method from StencilPartInterface and modifies equality to use the string representation returned by the `toString` method for comparison. This means the `run` and `selection` stencils return true by default, and clients should be cautious in relying on equality operator for comparison of stencils containing parts generated by these functions. It also means we no longer need the custom RTTI support (typeId() and down_cast()), so it has been removed. Patch by Harshal T. Lehri. Reviewers: gribozavr Subscribers: cfe-commits Tags: #clang Differential Revision: https://reviews.llvm.org/D68825 llvm-svn: 374552
* Updated add_new_check.py to create checker tests in the new directoryDmitri Gribenko2019-10-111-1/+1
| | | | llvm-svn: 374551
* [ClangTidy] Separate tests for infrastructure and checkers, fixupDmitri Gribenko2019-10-111-0/+0
| | | | | | Renamed a file that I missed in r374540. llvm-svn: 374549
* [mips] Follow-up to r374544. Fix test case.Simon Atanasyan2019-10-111-2/+2
| | | | llvm-svn: 374548
* [Tests] Output of od can be lower or upper case (llvm-objcopy/yaml2obj).Kai Nacke2019-10-1111-28/+28
| | | | | | | | | | | | | | | The command `od -t x` is used to dump data in hex format. The LIT tests assumes that the hex characters are in lowercase. However, there are also platforms which use uppercase letter. To solve this issue the tests are updated to use the new `--ignore-case` option of FileCheck. Reviewers: Bigcheese, jakehehrlich, rupprecht, espindola, alexshap, jhenderson Differential Revision: https://reviews.llvm.org/D68693 llvm-svn: 374547
* [lldb] Insert break to avoid unannotated fall-throughKadir Cetinkaya2019-10-111-0/+1
| | | | llvm-svn: 374545
* [mips] Fix loading "double" immediate into a GPR and FPRSimon Atanasyan2019-10-112-18/+36
| | | | | | | | | | | | | | | If a "double" (64-bit) value has zero low 32-bits, it's possible to load such value into a GP/FP registers as an instruction immediate. But now assembler loads only high 32-bits of the value. For example, if a target register is GPR the `li.d $4, 1.0` instruction converts into the `lui $4, 16368` one. As a result, we get `0x3FF00000` in the register. While a correct representation of the `1.0` value is `0x3FF0000000000000`. The patch fixes that. Differential Revision: https://reviews.llvm.org/D68776 llvm-svn: 374544
* [MS ABI]: Fix mangling function arguments for template types to be ↵Nico Weber2019-10-112-1/+18
| | | | | | | | | | | | | | | | | | | | | compatible with MSVC MS name mangling supports cache for first 10 distinct function arguments. The error was when non cached template type occurred twice (e.g. 11th and 12th). For such case in code there is another cache table TemplateArgStrings (for performance reasons). Then one '@' character at the end of the mangled name taken from this table was missing. For other cases the missing '@' character was added in the call to mangleSourceName(TemplateMangling) in the cache miss code, but the cache hit code didn't add it. This fixes a regression from r362560. Patch by Adam Folwarczny <adamf88@gmail.com>! Differential Revision: https://reviews.llvm.org/D68099 llvm-svn: 374543
* [LLD][ELF] - Update test cases after llvm-readobj change.George Rimar2019-10-1112-182/+139
| | | | | | https://reviews.llvm.org/D68704 changed the output format. llvm-svn: 374542
* [llvm-readobj] - Remove excessive fields when dumping "Version symbols".George Rimar2019-10-114-61/+42
| | | | | | | | | | | This removes a few fields that are not useful: "Section Name", "Address", "Offset" and "Link" (they duplicated the information available under the "Sections [" tag). Differential revision: https://reviews.llvm.org/D68704 llvm-svn: 374541
* [ClangTidy] Separate tests for infrastructure and checkersDmitri Gribenko2019-10-11564-0/+0
| | | | | | | | | | | | | | | | | | | | Summary: This change moves tests for checkers and infrastructure into separate directories, making it easier to find infrastructure tests. Tests for checkers are already easy to find because they are named after the checker. Tests for infrastructure were difficult to find because they were outnumbered by tests for checkers. Now they are in a separate directory. Reviewers: jfb, jdoerfert, lebedev.ri Subscribers: srhines, nemanjai, aheejin, kbarton, christof, mgrang, arphaman, jfb, lebedev.ri, cfe-commits Tags: #clang Differential Revision: https://reviews.llvm.org/D68807 llvm-svn: 374540
* Dead Virtual Function EliminationOliver Stannard2019-10-1134-83/+1428
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Currently, it is hard for the compiler to remove unused C++ virtual functions, because they are all referenced from vtables, which are referenced by constructors. This means that if the constructor is called from any live code, then we keep every virtual function in the final link, even if there are no call sites which can use it. This patch allows unused virtual functions to be removed during LTO (and regular compilation in limited circumstances) by using type metadata to match virtual function call sites to the vtable slots they might load from. This information can then be used in the global dead code elimination pass instead of the references from vtables to virtual functions, to more accurately determine which functions are reachable. To make this transformation safe, I have changed clang's code-generation to always load virtual function pointers using the llvm.type.checked.load intrinsic, instead of regular load instructions. I originally tried writing this using clang's existing code-generation, which uses the llvm.type.test and llvm.assume intrinsics after doing a normal load. However, it is possible for optimisations to obscure the relationship between the GEP, load and llvm.type.test, causing GlobalDCE to fail to find virtual function call sites. The existing linkage and visibility types don't accurately describe the scope in which a virtual call could be made which uses a given vtable. This is wider than the visibility of the type itself, because a virtual function call could be made using a more-visible base class. I've added a new !vcall_visibility metadata type to represent this, described in TypeMetadata.rst. The internalization pass and libLTO have been updated to change this metadata when linking is performed. This doesn't currently work with ThinLTO, because it needs to see every call to llvm.type.checked.load in the linkage unit. It might be possible to extend this optimisation to be able to use the ThinLTO summary, as was done for devirtualization, but until then that combination is rejected in the clang driver. To test this, I've written a fuzzer which generates random C++ programs with complex class inheritance graphs, and virtual functions called through object and function pointers of different types. The programs are spread across multiple translation units and DSOs to test the different visibility restrictions. I've also tried doing bootstrap builds of LLVM to test this. This isn't ideal, because only classes in anonymous namespaces can be optimised with -fvisibility=default, and some parts of LLVM (plugins and bugpoint) do not work correctly with -fvisibility=hidden. However, there are only 12 test failures when building with -fvisibility=hidden (and an unmodified compiler), and this change does not cause any new failures for either value of -fvisibility. On the 7 C++ sub-benchmarks of SPEC2006, this gives a geomean code-size reduction of ~6%, over a baseline compiled with "-O2 -flto -fvisibility=hidden -fwhole-program-vtables". The best cases are reductions of ~14% in 450.soplex and 483.xalancbmk, and there are no code size increases. I've also run this on a set of 8 mbed-os examples compiled for Armv7M, which show a geomean size reduction of ~3%, again with no size increases. I had hoped that this would have no effect on performance, which would allow it to awlays be enabled (when using -fwhole-program-vtables). However, the changes in clang to use the llvm.type.checked.load intrinsic are causing ~1% performance regression in the C++ parts of SPEC2006. It should be possible to recover some of this perf loss by teaching optimisations about the llvm.type.checked.load intrinsic, which would make it worth turning this on by default (though it's still dependent on -fwhole-program-vtables). Differential revision: https://reviews.llvm.org/D63932 llvm-svn: 374539
* [FileCheck] Implement --ignore-case option.Kai Nacke2019-10-116-2/+66
| | | | | | | | | | | | The FileCheck utility is enhanced to support a `--ignore-case` option. This is useful in cases where the output of Unix tools differs in case (e.g. case not specified by Posix). Reviewers: Bigcheese, jakehehrlich, rupprecht, espindola, alexshap, jhenderson, MaskRay Differential Revision: https://reviews.llvm.org/D68146 llvm-svn: 374538
* [LLDB] [lldb-server] Use llvm::InitLLVM for doing unicode conversion of ↵Martin Storsjo2019-10-111-0/+2
| | | | | | | | | | | | | | | | arguments for windows This should allow lldb-server to operate on files with non-ascii pathnames. I tried looking around in lldb/tools, and this seemed like the only other tool (other than the main lldb driver itself) that would be used (implicitly) by an end user (which could be working in non-ascii paths). Differential Revision: https://reviews.llvm.org/D68864 llvm-svn: 374537
* [SCEV] Add stricter verification option.Florian Hahn2019-10-111-5/+8
| | | | | | | | | | | | | | | | | | | | | | Currently -verify-scev only fails if there is a constant difference between two BE counts. This misses a lot of cases. This patch adds a -verify-scev-strict options, which fails for any non-zero differences, if used together with -verify-scev. With the stricter checking, some unit tests fail because of mis-matches, especially around IndVarSimplify. If there is no reason I am missing for just checking constant deltas, I am planning on looking into the various failures. Reviewers: efriedma, sanjoy.google, reames, atrick Reviewed By: sanjoy.google Differential Revision: https://reviews.llvm.org/D68592 llvm-svn: 374535
* [X86] isFNEG - add recursion depth limitSimon Pilgrim2019-10-111-5/+9
| | | | | | Now that its used by isNegatibleForFree we should try to avoid costly deep recursion llvm-svn: 374534
* [llvm-exegesis] Show noise cluster in analysis output.Clement Courbet2019-10-113-19/+91
| | | | | | | | | | | | Reviewers: gchatelet Subscribers: tschuett, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D68780 llvm-svn: 374533
* minidump: Use llvm memory info list parserPavel Labath2019-10-114-123/+26
| | | | llvm-svn: 374532
* ProcessInstanceInfoMatch: Don't match processes with no name if a name match ↵Pavel Labath2019-10-114-18/+57
| | | | | | | | | | | | | | | | | | | | | | | | | | was requested, take 2 Summary: The previous attempt at making nameless process not match when searching for a given name failed because the macos implementation was depending on this detail in its partial matching strategy. Doing partial matching to avoid expensive lookups is a perfectly valid thing to do, the way it was implemented seems somewhat unexpected. This patch implements it differently by providing special methods in the ProcessInstanceInfoMatch which match only a subset of fields, and changes mac host code to use those instead. Then, it re-applies r373925 to get make the ProcessInstanceInfoMatch with a name *not* match a nameless process. Reviewers: JDevlieghere, teemperor, jingham Subscribers: wallace, lldb-commits Differential Revision: https://reviews.llvm.org/D68631 llvm-svn: 374529
* [Windows] Use information from the PE32 exceptions directory to construct ↵Aleksandr Urakov2019-10-1119-11/+1149
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | unwind plans This patch adds an implementation of unwinding using PE EH info. It allows to get almost ideal call stacks on 64-bit Windows systems (except some epilogue cases, but I believe that they can be fixed with unwind plan disassembly augmentation in the future). To achieve the goal the CallFrameInfo abstraction was made. It is based on the DWARFCallFrameInfo class interface with a few changes to make it less DWARF-specific. To implement the new interface for PECOFF object files the class PECallFrameInfo was written. It uses the next helper classes: - UnwindCodesIterator helps to iterate through UnwindCode structures (and processes chained infos transparently); - EHProgramBuilder with the use of UnwindCodesIterator constructs EHProgram; - EHProgram is, by fact, a vector of EHInstructions. It creates an abstraction over the low-level unwind codes and simplifies work with them. It contains only the information that is relevant to unwinding in the unified form. Also the required unwind codes are read from the object file only once with it; - EHProgramRange allows to take a range of EHProgram and to build an unwind row for it. So, PECallFrameInfo builds the EHProgram with EHProgramBuilder, takes the ranges corresponding to every offset in prologue and builds the rows of the resulted unwind plan. The resulted plan covers the whole range of the function except the epilogue. Reviewers: jasonmolenda, asmith, amccarth, clayborg, JDevlieghere, stella.stamenova, labath, espindola Reviewed By: jasonmolenda Subscribers: leonid.mashinskiy, emaste, mgorny, aprantl, arichardson, MaskRay, lldb-commits, llvm-commits Tags: #lldb Differential Revision: https://reviews.llvm.org/D67347 llvm-svn: 374528
* Insert module constructors in a module passVitaly Buka2019-10-119-55/+102
| | | | | | | | | | | | | | | | | | | | | Summary: If we insert them from function pass some analysis may be missing or invalid. Fixes PR42877. Reviewers: eugenis, leonardchan Reviewed By: leonardchan Subscribers: hiraditya, cfe-commits, llvm-commits Tags: #clang, #llvm Differential Revision: https://reviews.llvm.org/D68832 > llvm-svn: 374481 Signed-off-by: Vitaly Buka <vitalybuka@google.com> llvm-svn: 374527
* [LLDB] [Driver] Use llvm::InitLLVM to do unicode argument conversion on WindowsMartin Storsjo2019-10-111-17/+3
| | | | | | | | | This avoids the currently MSVC specific codepath of using the wchar entry point and converting that to utf8. Differential Revision: https://reviews.llvm.org/D68770 llvm-svn: 374526
* [lldb] Fix crash in CxxModuleHandler when std module is emptyRaphael Isemann2019-10-117-4/+88
| | | | | | | | | | | | | | | | | | | | We currently don't handle the error in the Expected we get when searching for an equal local DeclContext. Usually this can't happen as this would require that we have a STL container and we can find libc++'s std module, but when we load the module in the expression parser the module doesn't even contain the 'std' namespace. The only way I see to test this is by having a fake 'std' module that requires a special define to actually provide its contents, while it will just be empty (that is, it doesn't even contain the 'std' namespace) without that define. LLDB currently doesn't know about that define in the expression parser, so it will load the wrong 'empty' module which should trigger this error. Also removed the 'auto' for that variable as the function name doesn't make it obvious that this is an expected and not just a optional/ptr (which is how this slipped in from the start). llvm-svn: 374525
* [TableGen] Fix a bug that MCSchedClassDesc is interfered between different ↵QingShan Zhang2019-10-113-1/+48
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | SchedModel Assume that, ModelA has scheduling resource for InstA and ModelB has scheduling resource for InstB. This is what the llvm::MCSchedClassDesc looks like: llvm::MCSchedClassDesc ModelASchedClasses[] = { ... InstA, 0, ... InstB, -1,... }; llvm::MCSchedClassDesc ModelBSchedClasses[] = { ... InstA, -1,... InstB, 0,... }; The -1 means invalid num of macro ops, while it is valid if it is >=0. This is what we look like now: llvm::MCSchedClassDesc ModelASchedClasses[] = { ... InstA, 0, ... InstB, 0,... }; llvm::MCSchedClassDesc ModelBSchedClasses[] = { ... InstA, 0,... InstB, 0,... }; And compiler hit the assertion here because the SCDesc is valid now for both InstA and InstB. Differential Revision: https://reviews.llvm.org/D67950 llvm-svn: 374524
* [X86] Add v8i64->v8i8 ssat/usat/packus truncate tests to ↵Craig Topper2019-10-111-0/+103
| | | | | | | | | | | min-legal-vector-width.ll I wonder if we should split the v8i8 stores in order to form two v4i8 saturating truncating stores. This would remove the unpckl needed to concatenated the v4i8 results to make a single store. llvm-svn: 374519
* [ADT][Statistics] Fix test after rL374490Kadir Cetinkaya2019-10-111-0/+2
| | | | llvm-svn: 374518
* Fix modules build for r374337Pavel Labath2019-10-111-0/+2
| | | | | | | | | | A modules build failed with the following error: call to function 'operator&' that is neither visible in the template definition nor found by argument-dependent lookup Fix that by declaring the appropriate operators in the llvm::minidump namespace. llvm-svn: 374517
* [X86] Always define the tzcnt intrinsics even when _MSC_VER is defined.Craig Topper2019-10-113-130/+144
| | | | | | | | | | | | These intrinsics use llvm.cttz intrinsics so are always available even without the bmi feature. We already don't check for the bmi feature on the intrinsics themselves. But we were blocking the include of the header file with _MSC_VER unless BMI was enabled on the command line. Fixes PR30506. llvm-svn: 374516
* [PowerPC] Remove assertion "Shouldn't overwrite a register before it is killed"Yi-Hong Lyu2019-10-112-8/+31
| | | | | | | | | | | | | | The assertion is everzealous and fail tests like: renamable $x3 = LI8 0 STD renamable $x3, 16, $x1 renamable $x3 = LI8 0 Remove the assertion since killed flag of $x3 is not mandentory. Differential Revision: https://reviews.llvm.org/D68344 llvm-svn: 374515
* [NFC] run specific pass instead of whole -O3 pipeline for popcount ↵Chen Zheng2019-10-111-5/+5
| | | | | | recoginzation testcase. llvm-svn: 374514
* [InstCombine] recognize popcount.Chen Zheng2019-10-113-5/+271
| | | | | | | | | This patch recognizes popcount intrinsic according to algorithm from website http://graphics.stanford.edu/~seander/bithacks.html#CountBitsSetParallel Differential Revision: https://reviews.llvm.org/D68189 llvm-svn: 374512
* [libc++] Fix linker script generationPetr Hosek2019-10-111-1/+3
| | | | | | | | | | Handle the case when libc++abi and libunwind are being built together with libc++ in the runtimes build. This logic was used in the previous implementation but dropped in r374116. Differential Revision: https://reviews.llvm.org/D68791 llvm-svn: 374510
* [X86] Add a DAG combine to turn v16i16->v16i8 VTRUNCUS+store into a ↵Craig Topper2019-10-112-2/+14
| | | | | | saturating truncating store. llvm-svn: 374509
* [X86] Add test case for trunc_packus_v16i32_v16i8_store to ↵Craig Topper2019-10-111-1/+21
| | | | | | | | min-legal-vector-width.ll We aren't folding the vpmovuswb into the store. llvm-svn: 374507
* [CVP] Remove a masking operation if range information implies it's a noopPhilip Reames2019-10-114-3/+155
| | | | | | | | | | This is really a known bits style transformation, but known bits isn't context sensitive. The particular case which comes up happens to involve a range which allows range based reasoning to eliminate the mask pattern, so handle that case specifically in CVP. InstCombine likes to generate the mask-by-low-bits pattern when widening an arithmetic expression which includes a zext in the middle. Differential Revision: https://reviews.llvm.org/D68811 llvm-svn: 374506
* [X86] Add more packus/ssat/usat truncate tests from legal vectors to less ↵Craig Topper2019-10-113-3/+4134
| | | | | | | | than 128-bit vectors. Some of these have sub-optimal codegen for avx512 relative to avx2. llvm-svn: 374505
* [Polly] Fix formatting violation. NFC.Volodymyr Sapsai2019-10-111-3/+1
| | | | llvm-svn: 374504
* Revert 374481 "[tsan,msan] Insert module constructors in a module pass"Nico Weber2019-10-119-102/+55
| | | | | | | CodeGen/sanitizer-module-constructor.c fails on mac and windows, see e.g. http://lab.llvm.org:8011/builders/clang-x64-windows-msvc/builds/11424 llvm-svn: 374503
OpenPOWER on IntegriCloud