summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* MIR: Support MachineMemOperands without associated valueMatthias Braun2016-06-044-34/+28
| | | | | | | This is allowed (though used rarely) and useful to keep your tests short. llvm-svn: 271752
* Replace hard coded probability threshold with parameter /NFCXinliang David Li2016-06-032-8/+13
| | | | llvm-svn: 271751
* CodeGen: correct assertionSaleem Abdulrasool2016-06-031-2/+3
| | | | | | | The assertion added earlier was overly strict. We need to strip the pointer casts (as when constructing the GV). Correct the types (Function or Variable). llvm-svn: 271750
* [CMake] Support constructing output paths from LLVM variablesChris Bieneman2016-06-031-1/+7
| | | | | | | | This tweak to constructing output paths allows compiler-rt to use LLVM output variables if they are set regardless of whether or not the build is in-tree. llvm-svn: 271749
* [llvm-profdata] Fix use-after-free from discarded MemoryBuffer (NFC)Vedant Kumar2016-06-031-8/+13
| | | | | | | | Thanks to Justin Bogner for pointing this out! Caught by ASAN. llvm-svn: 271748
* Bump libclang API minor version after r271351.Manman Ren2016-06-031-2/+2
| | | | | | Also use the next enum value for CXObjCPropertyAttr_class. llvm-svn: 271747
* [pgo] extend r271532 to darwin platformXinliang David Li2016-06-031-4/+4
| | | | llvm-svn: 271746
* Reapply r271728 after adding move cobstructor for ProfileSummaryInfoEaswaran Raman2016-06-038-2/+322
| | | | llvm-svn: 271745
* [esan] Initialize runtime during early interceptorsDerek Bruening2016-06-033-4/+18
| | | | | | | | | | | | | | | | | Summary: Adds initialization of esan's runtime library during any early interceptors that are sometimes called prior to the official __esan_init() invocation (we see this with apps using tcmalloc). Adds handling of interceptors called during interceptor initialization. Reviewers: aizatsky Subscribers: vitalybuka, zhaoqin, kcc, eugenis, llvm-commits, kubabrecka Differential Revision: http://reviews.llvm.org/D20976 llvm-svn: 271744
* [esan|wset] Optionally assume intra-cache-line accessesDerek Bruening2016-06-033-12/+185
| | | | | | | | | | | | | | | | | | Summary: Adds an option -esan-assume-intra-cache-line which causes esan to assume that a single memory access touches just one cache line, even if it is not aligned, for better performance at a potential accuracy cost. Experiments show that the performance difference can be 2x or more, and accuracy loss is typically negligible, so we turn this on by default. This currently applies just to the working set tool. Reviewers: aizatsky Subscribers: vitalybuka, zhaoqin, kcc, eugenis, llvm-commits Differential Revision: http://reviews.llvm.org/D20978 llvm-svn: 271743
* Add a FIXME note in the release notes about documenting ThinLTOMehdi Amini2016-06-031-0/+3
| | | | llvm-svn: 271742
* [libcxx] Fix c++98 test failures.Asiri Rathnayake2016-06-0322-23/+23
| | | | | | | | | Adds XFAIL/UNSUPPORTED lit tags as appropriate. Gets a clean test run for -std=c++98 on Fedora 20. NFC. llvm-svn: 271741
* [libfuzzer] hiding custom mutator handling in MutationDispatcher.Mike Aizatsky2016-06-033-29/+60
| | | | | | | | Summary: Refactoring, no functional changes. Differential Revision: http://reviews.llvm.org/D20975 llvm-svn: 271740
* [clang-tidy] modernize-use-auto: don't remove stars by defaultAlexander Kornienko2016-06-035-37/+175
| | | | | | | | | | | | | | | Summary: By default, modernize-use-auto check will retain stars when replacing an explicit type with `auto`: `MyType *t = new MyType;` will be changed to `auto *t = new MyType;`, thus resulting in more consistency with the recommendations to use `auto *` for iterating over pointers in range-based for loops: http://llvm.org/docs/CodingStandards.html#beware-unnecessary-copies-with-auto The new `RemoveStars` option allows to revert to the old behavior: with the new option turned on the check will change `MyType *t = new MyType;` to `auto t = new MyType;`. Reviewers: aaron.ballman, sbenza Subscribers: Eugene.Zelenko, cfe-commits Differential Revision: http://reviews.llvm.org/D20917 llvm-svn: 271739
* Revert r271728 as it breaks Windows buildEaswaran Raman2016-06-038-320/+2
| | | | llvm-svn: 271738
* [libunwind] Remove unused code.Asiri Rathnayake2016-06-031-32/+0
| | | | | | | | | | The whole file is guarded with #if _LIBUNWIND_ARM_EHABI, and then in the middle we have these two blocks, which render them pretty unused. An artefact of a refactoring it seems. NFC. llvm-svn: 271737
* pdbdump: print out TPI hashes.Rui Ueyama2016-06-038-26/+91
| | | | | | Differential Revision: http://reviews.llvm.org/D20945 llvm-svn: 271736
* [esan|cfrag] Compute the struct field access difference ratioQin Zhao2016-06-034-23/+122
| | | | | | | | | | | | | | | | | | Summary: Computes the struct field access variation based on each field access count. Adds a flag to control the report thresholds. Updates struct-simple.cpp with variance report output. Reviewers: aizatsky Subscribers: kubabrecka, zhaoqin, llvm-commits, eugenis, vitalybuka, kcc, bruening Differential Revision: http://reviews.llvm.org/D20914 llvm-svn: 271734
* Re-apply "SDAG: Update ChainNodesMatched as nodes are deleted"Justin Bogner2016-06-032-4/+17
| | | | | | | | | | | | | | | My first attempt at this had an overly aggressive assert - chain nodes will only be removed, but we could hit the assert if a non-chain node was CSE'd (NodeToMatch, for instance). This reapplies r271706 by reverting r271713 and fixing an assert. Original message: Avoid relying on UB by looking into deleted nodes for a marker value. Instead, update the list of chain nodes as we go. llvm-svn: 271733
* [ThinLTO/gold] Pass import lists by reference (NFC)Teresa Johnson2016-06-031-1/+1
| | | | | | | In my earlier patch r271690 I missed passing a map by reference in one place. llvm-svn: 271732
* Analysis pass to access profile summary infoEaswaran Raman2016-06-038-2/+320
| | | | | | Differential Revision: http://reviews.llvm.org/D20648 llvm-svn: 271728
* Fix non-Windows build when inserting a move only type into a mapReid Kleckner2016-06-031-1/+1
| | | | llvm-svn: 271727
* [cpu-detection] Naming conventionAlina Sbirlea2016-06-031-28/+27
| | | | | | | | | | | | Summary: Follow-up to D20926 (committed as r271595, r271596). This patch is in preparation for a substantial refactoring of the code. No functionality changed. Differential Revision: http://reviews.llvm.org/D20970 llvm-svn: 271726
* [Symbolize] Check if the PE file has a PDB and emit an error if we can't load itReid Kleckner2016-06-039-86/+161
| | | | | | | | | | | | | | | | | | | | | Summary: Previously we would try to load PDBs for every PE executable we tried to symbolize. If that failed, we would fall back to DWARF. If there wasn't any DWARF, we'd print mostly useless symbol information using the export table. With this change, we only try to load PDBs for executables that claim to have them. If that fails, we can now print an error rather than falling back silently. This should make it a lot easier to diagnose and fix common symbolization issues, such as not having DIA or not having a PDB. Reviewers: zturner, eugenis Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D20982 llvm-svn: 271725
* [AArch64] Move tests from r271677 to a more appropriately named file. NFC.Chad Rosier2016-06-032-88/+88
| | | | llvm-svn: 271718
* [AArch64] Spot SBFX-compatible code expressed with sign_extend.Chad Rosier2016-06-032-0/+40
| | | | | | | This is very similar to r271677, but for extracts from i32 with the SIGN_EXTEND acting on a arithmetic shift. llvm-svn: 271717
* Fix a printf warning.Greg Clayton2016-06-031-2/+1
| | | | llvm-svn: 271716
* [esan] Specify which tool via a global variableDerek Bruening2016-06-031-0/+13
| | | | | | | | | | | | | | | Summary: Adds a global variable to specify the tool, to support handling early interceptors that invoke instrumented code and require shadow memory to be initialized prior to __esan_init() being invoked. Reviewers: aizatsky Subscribers: vitalybuka, zhaoqin, kcc, eugenis, llvm-commits Differential Revision: http://reviews.llvm.org/D20973 llvm-svn: 271715
* [esan] Specify which tool via a global variableDerek Bruening2016-06-034-20/+24
| | | | | | | | | | | | | | | Summary: Adds a global variable to specify the tool, to support handling early interceptors that invoke instrumented code, thus requiring shadow memory to be initialized prior to __esan_init() being invoked. Reviewers: aizatsky Subscribers: vitalybuka, zhaoqin, kcc, eugenis, llvm-commits, kubabrecka Differential Revision: http://reviews.llvm.org/D20974 llvm-svn: 271714
* Revert "SDAG: Update ChainNodesMatched as nodes are deleted"Justin Bogner2016-06-032-17/+4
| | | | | | | | | Seeing failures in CodeGen/Generic/icmp-illegal.ll on quite a few bots. This reverts r271706. llvm-svn: 271713
* [llvm-pdbdump] Introduce an abstraction for the output style.Zachary Turner2016-06-036-653/+788
| | | | | | | | | | | This opens the door to introducing a YAML outputter which can be used for machine consumption. Currently the yaml output style is unimplemented and returns an error if you try to use it. Reviewed By: rnk, ruiu Differential Revision: http://reviews.llvm.org/D20967 llvm-svn: 271712
* Test commit. Removes some spaces. No functionality changed.Alina Sbirlea2016-06-031-22/+22
| | | | | | | | | | | | Summary: Test commit. Removes some spaces. No functionality changed. Reviewers: llvm-commits Differential Revision: http://reviews.llvm.org/D20972 llvm-svn: 271711
* [llvm-profdata] Fix option description (NFC)Vedant Kumar2016-06-031-1/+1
| | | | llvm-svn: 271710
* [llvm-profdata] Add option to ingest filepaths from a fileVedant Kumar2016-06-033-4/+70
| | | | | | Differential Revision: http://reviews.llvm.org/D20980 llvm-svn: 271709
* Use the name of the file on disk to issue a new diagnostic about ↵Taewook Oh2016-06-0314-13/+162
| | | | | | | | | | | non-portable #include and #import paths. Differential Revision: http://reviews.llvm.org/D19843 Corresponding LLVM change: http://reviews.llvm.org/D19842 Patch by Eric Niebler llvm-svn: 271708
* Fix modulo compared to zero.Michael Kruse2016-06-032-1/+49
| | | | | | | | | | | | | | | | In case of modulo compared to zero, we need to do signed modulo operation as unsigned can give different results based on whether the dividend is negative or not. This addresses llvm.org/PR27707 Contributed-by: Chris Jenneisch <chrisj@codeaurora.org> Reviewers: _jdoerfert, grosser, Meinersbur Differential Revision: http://reviews.llvm.org/D20145 llvm-svn: 271707
* SDAG: Update ChainNodesMatched as nodes are deletedJustin Bogner2016-06-032-4/+17
| | | | | | | Avoid relying on UB by looking into deleted nodes for a marker value. Instead, update the list of chain nodes as we go. llvm-svn: 271706
* [FIX] Fix potential issue related to subtraction from an unsigned 0 in ↵Roman Gareev2016-06-031-1/+3
| | | | | | | | | | | circularShiftOutputDims Reported-by: Mehdi Amini <mehdi.amini@apple.com> Contributed-by: Michael Kruse <llvm@meinersbur.de> Differential Revision: http://reviews.llvm.org/D20969 llvm-svn: 271705
* In openFileForRead, attempt to fetch the actual name of the file on disk -- ↵Taewook Oh2016-06-034-3/+196
| | | | | | | | | | including case -- so that clang can later warn about non-portable #include and #import directives. Differential Revision: http://reviews.llvm.org/D19842 Patch by Eric Niebler llvm-svn: 271704
* [WebAssembly] Emit type signatures for declared functionsDerek Schuff2016-06-034-10/+129
| | | | | | | | | | | | | | | | | | | | | Under emscripten, C code can take the address of a function implemented in Javascript (which is exposed via an import in wasm). Because imports do not have linear memory address in wasm, we need to generate a thunk to be the target of the indirect call; it call the import directly. To make this possible, LLVM needs to emit the type signatures for these functions, because they may not be called directly or referred to other than where the address is taken. This uses s new .s directive (.functype) which specifies the signature. Differential Revision: http://reviews.llvm.org/D20891 Re-apply r271599 but instead of bailing with an error when a declared function has multiple returns, replace it with a pointer argument. Also add the test case I forgot to 'git add' last time around. llvm-svn: 271703
* [Title] Revert test commitTaewook Oh2016-06-031-1/+0
| | | | | | | | | | | | | | | | | | | | | | | | | | | | Summary: Revert test commit Trac Bug: # Blame Rev: Reviewed By: Test Plan: Revert Plan: Database Impact: Memcache Impact: Other Notes: EImportant: - begin *PUBLIC* platform impact section - Bugzilla: # - end platform impact - llvm-svn: 271702
* Test commit Taewook Oh2016-06-031-0/+1
| | | | llvm-svn: 271701
* SDAG: Replace some unreachable code with an assert. NFCJustin Bogner2016-06-031-6/+3
| | | | | | | The current node shouldn't be (and isn't) removed partway through selection. llvm-svn: 271699
* [sancov] Disable these tests if there is no X86 backendReid Kleckner2016-06-031-0/+4
| | | | | | Copied from test/CodeGen/X86 llvm-svn: 271698
* [libfuzzer] splitting fuzzer.testMike Aizatsky2016-06-0311-97/+99
| | | | llvm-svn: 271697
* Add support in debug LLDB builds (if LLDB_CONFIGURATION_DEBUG is defined) ↵Greg Clayton2016-06-031-0/+40
| | | | | | | | | | where we can set an environment variable named LLDB_DWARF_DONT_COMPLETE_TYPENAMES that can contain one or more typenames separated by ';' characters. This will cause us to not complete any types whose names match and can help us to try and reproduce issues we see in bugs. So you can launch LLDB with the environment variable: % LLDB_DWARF_DONT_COMPLETE_TYPENAMES=Foo;Bar;Baz lldb llvm-svn: 271696
* [sancov] Run sancov tests on more platformsReid Kleckner2016-06-037-5/+10
| | | | | | | | The only tests that need to be run on Linux are the ones that use C++ demangling. I'm assuming they will fail on Mac, since __cxa_demangle there won't handle the non-double-underscore prefixed mangled names. llvm-svn: 271695
* Doxygen for FoldingSet::reserve and FoldingSet::capacityBen Craig2016-06-031-0/+5
| | | | llvm-svn: 271694
* [CMake] detect_target_arch needs to be moved to UtilsChris Bieneman2016-06-032-34/+34
| | | | | | This macro is called from the base config, so it can't live in config-ix, it needs to be in the utils. I suspect the only reason this hasn't caused problems is that nobody is building the Android builtins from the builtins subdirectory. llvm-svn: 271693
* Don't pass --build-id to ld by default.Rafael Espindola2016-06-033-4/+8
| | | | | | | | | | | | We now have a cmake option to change the default: ENABLE_LINKER_BUILD_ID. The reason is that build-id is fairly expensive, so we shouldn't impose it in the regular edit/build cycle. This is similar to gcc, that has an off by default --enable-linker-build-id option. llvm-svn: 271692
OpenPOWER on IntegriCloud