summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
...
* [APFloat] APFloat::Storage::Storage - fix use after moveSimon Pilgrim2019-05-141-3/+3
| | | | | | This was mentioned both in https://www.viva64.com/en/b/0629/ and by scan-build checks llvm-svn: 360675
* [Driver][Windows] Add dependent lib argument for profile instr generateRussell Gallop2019-05-143-4/+24
| | | | | | | | | | This is needed so lld-link can find clang_rt.profile when self hosting on Windows with PGO. Using clang-cl as a linker knows to add the library but self hosting, using -DCMAKE_LINKER=<...>/lld-link.exe doesn't. Differential Revision: https://reviews.llvm.org/D61742 llvm-svn: 360674
* [libc++] [test] Use std::nextafter() instead of std::nexttoward()Michal Gorny2019-05-141-2/+2
| | | | | | | | | | | | Use std::nextafter() instead of std::nexttoward() in midpoint tests. In the context of this test, this should not cause any difference. Since nexttowardl() is not implemented on NetBSD 8, the latter function combined with 'long double' type caused test failure. nextafterl() does not have this problem. Differential Revision: https://reviews.llvm.org/D61748 llvm-svn: 360673
* [lit][tests]Add feature libcxx-used and use it in llvm-*-fuzzer testsXing Xue2019-05-145-0/+13
| | | | | | | | | | | | | | | | When a LLVM binary such as llvm-*-fuzzer is built with libc++, it has dependency on libc++. The path to find shared libraries specified in llvm-*-fuzzer is relative. As a result, these binaries cannot be copied to an arbitrary directory and launched from there. Changes in this patch add a LIT feature to indicate that libc++ is used to build and, based on the feature exclude test cases that test by copying llvm-*-fuzzer binaries to a directory. Reviewers: hubert.reinterpretcast, dberris, amyk, jasonliu, EricWF Reviewed By: hubert.reinterpretcast, amyk Subscribers: javed.absar, jsji, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D61265 llvm-svn: 360672
* Save the induction binary operator in IVDescriptors for non FP induction ↵Kit Barton2019-05-144-3/+115
| | | | | | | | | | | | | | | | | | | | | | | | | variables. Summary: Currently InductionBinOps are only saved for FP induction variables, the PR extends it with non FP induction variable, so user of IVDescriptors can query the InductionBinOps for integer induction variables. The changes in hasUnsafeAlgebra() and getUnsafeAlgebraInst() are required for the existing LIT test cases to pass. As described in the comment of the two functions, one of the requirement to return true is it is a FP induction variable. The checks was not needed because InductionBinOp was not set on non FP cases before. https://reviews.llvm.org/D60565 depends on the patch. Committed on behalf of @Whitney (Whitney Tsang). Reviewers: jdoerfert, kbarton, fhahn, hfinkel, dmgreen, Meinersbur Reviewed By: jdoerfert Subscribers: mgorny, hiraditya, jsji, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D61329 llvm-svn: 360671
* TableGen: support #ifndef in addition to #ifdef.Tim Northover2019-05-146-7/+33
| | | | | | | TableGen has a limited preprocessor, which only really supports easier. llvm-svn: 360670
* DWARF/NFC: Centralize DIERef conversionsPavel Labath2019-05-143-40/+24
| | | | | | | | | | | | | apple and manual indexing code were creating a DIERef in a bunch of places. Though the code itself is not much, it is also easy to simplify by factoring out the DIERef creation. In HashedNameToDIE I create a conversion operator from DIEInfo to DIERef, and in ManualDWARFIndex I just create the DIERef in a global variable up-front. This also reduces the diff in follow-up patches which change how DIERefs are constructed. llvm-svn: 360669
* Make getObjCEncodingForTypeImpl() take a bitmask instead of 8 boolsNico Weber2019-05-142-115/+122
| | | | | | | | | | | Slightly easier to read, uses slightly less stack space, and makes it impossible to mix up the order of all those bools. No behavior change. Differential Revision: https://reviews.llvm.org/D61788 llvm-svn: 360668
* Add a new language mode for C2x; enable [[attribute]] support by default in C2x.Aaron Ballman2019-05-1411-17/+37
| | | | llvm-svn: 360667
* Move SymbolTable::addFile to InputFiles.cpp.Rui Ueyama2019-05-145-86/+95
| | | | | | | | | | | | | | | | | The symbol table used to be a container of vectors of input files, but that's no longer the case because the vectors are moved out of SymbolTable and are now global variables. Therefore, addFile doesn't have to belong to any class. This patch moves the function out of the class. This patch is a preparation for my RFC [1]. [1] http://lists.llvm.org/pipermail/llvm-dev/2019-April/131902.html Differential Revision: https://reviews.llvm.org/D61854 llvm-svn: 360666
* Reinstate "FileCheck [5/12]: Introduce regular numeric variables"Thomas Preud'homme2019-05-1411-152/+803
| | | | | | | | | | | | | | | | | | | | | | | This reinstates r360578 (git e47362c1ec1ea31b626336cc05822035601c3e57), reverted in r360653 (git 004393681c25e34e921adccc69ae6378090dee54), with a fix for the list added in FileCheck.rst to build without error. Copyright: - Linaro (changes up to diff 183612 of revision D55940) - GraphCore (changes in later versions of revision D55940 and in new revision created off D55940) Reviewers: jhenderson, chandlerc, jdenny, probinson, grimar, arichardson, rnk Subscribers: hiraditya, llvm-commits, probinson, dblaikie, grimar, arichardson, tra, rnk, kristina, hfinkel, rogfer01, JonChesterfield Tags: #llvm Differential Revision: https://reviews.llvm.org/D60385 llvm-svn: 360665
* [X86] X86TargetLowering::LowerINTRINSIC_WO_CHAIN - ensure rounding control ↵Simon Pilgrim2019-05-141-7/+7
| | | | | | | | is initialized. NFCI. Fixes scan-build warnings llvm-svn: 360664
* AArch64: support binutils-like things on arm64_32.Tim Northover2019-05-1420-22/+162
| | | | | | | | This adds support for the arm64_32 watchOS ABI to LLVM's low level tools, teaching them about the specific MachO choices and constants needed to disassemble things. llvm-svn: 360663
* GlobalOpt: do not promote globals used atomically to constants.Tim Northover2019-05-142-2/+16
| | | | | | | | | | Some atomic loads are implemented as cmpxchg (particularly if large or floating), and that usually requires write access to the memory involved or it will segfault. We can still propagate the constant value to users we understand though. llvm-svn: 360662
* [llvm-objcopy] Cache gnu_debuglink's target CRCJames Henderson2019-05-145-20/+30
| | | | | | | | | | | | | | | | | | | | | | .gnu_debuglink section contains information regarding file with debugging symbols, identified by its CRC32. This target file is not intended to ever change or it would invalidate the stored checksum, yet the checksum is calculated over and over again for each of the objects inside the archive, usually hundreds of times. This patch precomputes the CRC32 of the target once and then reuses the value where required, saving lots of redundant I/O. The error message reported should stay the same, although now it might be reported earlier. Reviewed by: jhenderson, jakehehrlich, MaskRay Differential Revision: https://reviews.llvm.org/D61343 Patch by Michal Janiszewski llvm-svn: 360661
* [test]Make test work on WindowsJames Henderson2019-05-141-2/+1
| | | | | | | | | | | | | Previously, the test didn't work because '\' characters appeared in the sed string, causing bogus escape characters to form in the substituted string literal. Switching to using '%/p' causes the path to be emitted with '/' characters instead, so that there are are no escaping issues. Reviewed by: kzhuravl, grimar Differential Revision: https://reviews.llvm.org/D61856 llvm-svn: 360660
* [MemorySanitizer] getMMXVectorTy - assert valid element size. NFCI.Simon Pilgrim2019-05-141-0/+2
| | | | | | Fixes scan-build warnings llvm-svn: 360658
* Revert r360637 "PR41817: Fix regression in r359260 that caused the MS ↵Hans Wennborg2019-05-142-46/+6
| | | | | | | | | | | | compatibility" > extension allowing a "static" declaration to follow an "extern" > declaration to stop working. It introduced asserts for some "static-following-extern" cases, breaking the Chromium build. See the cfe-commits thread for reproducer. llvm-svn: 360657
* [IRTranslator] Don't hardcode GEP index typeDiana Picus2019-05-142-2/+25
| | | | | | | | | | | | | | | | | | When breaking up loads and stores of aggregates, the IRTranslator uses LLT::scalar(64) for the index type of the G_GEP instructions that compute the addresses. This is unnecessarily large for 32-bit targets. Use the int ptr type provided by the DataLayout instead. Note that we're already doing the right thing when translating getelementptr instructions from the IR. This is just an oversight when generating new ones while translating loads/stores. Both x86 and AArch64 already have tests confirming that the old behaviour is preserved for 64-bit targets. Differential Revision: https://reviews.llvm.org/D61852 llvm-svn: 360656
* TestMinidumpNew.py: Use yaml2obj where possiblePavel Labath2019-05-1410-79/+147
| | | | | | | | | Replace checked-in minidumps with their yaml forms now that yaml2obj supports the ThreadList stream. I delete the test_modules_in_mini_dump test altogether as this functionality is covered more systematically in TestMinidumpUUID.py. llvm-svn: 360655
* typedef enum -> enumFangrui Song2019-05-1448-202/+172
| | | | | | | | Reviewed By: labath Differential Revision: https://reviews.llvm.org/D61883 llvm-svn: 360654
* Revert "FileCheck [5/12]: Introduce regular numeric variables"Thomas Preud'homme2019-05-1411-803/+152
| | | | | | | | This reverts r360578 (git e47362c1ec1ea31b626336cc05822035601c3e57) to solve the sphinx build failure on http://lab.llvm.org:8011/builders/llvm-sphinx-docs buildbot. llvm-svn: 360653
* [CMake] Correctly reinstate LLDB_CAN_USE_LLDB_SERVERPavel Labath2019-05-141-2/+2
| | | | | | | | | | | | | | | | | r360631 introduced a "syntax error" which meant that cmake was still not honoring the value of LLDB_CAN_USE_LLDB_SERVER variable. The correct syntax for seting an internal cache variable is "set(VAR value CACHE INTERNAL)", but the patch omitted the "CACHE" keyword. The "syntax error" is in quotes because without the CACHE keyword this is still valid syntax for setting the value of LLDB_CAN_USE_LLDB_SERVER to "1 INTERNAL". There doesn't seem to be a need for this to be a cache variable so I'm reverting this variable to a plain one, as it was before r360621. This will hopefully fix the windows build. llvm-svn: 360652
* Add guidelines/recommendations for organizers of LLVM SocialsAlex Denisov2019-05-142-0/+97
| | | | | | Differential Revision: https://reviews.llvm.org/D61550 llvm-svn: 360651
* [X86] Prefer locked stack op over mfence for seq_cst 64-bit stores on 32-bit ↵Philip Reames2019-05-143-5/+6
| | | | | | | | | | targets This is a follow on to D58632, with the same logic. Given a memory operation which needs ordering, but doesn't need to modify any particular address, prefer to use a locked stack op over an mfence. Differential Revision: https://reviews.llvm.org/D61863 llvm-svn: 360649
* [Object] Change ObjectFile::getSectionContents to return ↵Fangrui Song2019-05-1417-59/+58
| | | | | | | | | | | | | | | | | | | Expected<ArrayRef<uint8_t>> Change std::error_code getSectionContents(DataRefImpl, StringRef &) const; to Expected<ArrayRef<uint8_t>> getSectionContents(DataRefImpl) const; Many object formats use ArrayRef<uint8_t> as the underlying type, which is generally better than StringRef to represent binary data, so change the type to decrease the number of type conversions. Reviewed By: ruiu, sbc100 Differential Revision: https://reviews.llvm.org/D61781 llvm-svn: 360648
* gn build: add Hexagon targetDavid L. Jones2019-05-146-0/+248
| | | | | | Differential Revision: https://reviews.llvm.org/D61819 llvm-svn: 360647
* ARMDefines.h: fix -Wimplicit-fallthrough in -DLLVM_ENABLE_ASSERTIONS=OFF buildsFangrui Song2019-05-141-2/+3
| | | | | | | | Reviewed By: JDevlieghere Differential Revision: https://reviews.llvm.org/D61844 llvm-svn: 360646
* gn build: add Sparc targetDavid L. Jones2019-05-146-0/+192
| | | | llvm-svn: 360645
* gn build: add Lanai targetDavid L. Jones2019-05-146-0/+190
| | | | llvm-svn: 360644
* [PowerPC][NFC] Fix typos in triplesJinsong Ji2019-05-144-7/+7
| | | | | | Found by bzEq (Kai Luo). llvm-svn: 360643
* [X86] Use X86 instead of X32 as a check prefix in atomic-idempotent.ll. NFCCraig Topper2019-05-141-149/+145
| | | | | | | | | | X32 can refer to a 64-bit ABI that uses 32-bit ints, longs, and pointers. I plan to add gnux32 command lines to this test so this prepares for that. Also remove some check lines that have a prefix that is not in any run lines. llvm-svn: 360642
* LLDB website: Change the title back to "The LLDB Debugger"Adrian Prantl2019-05-142-4/+9
| | | | | | This is what the old homepage also used. llvm-svn: 360641
* [SDAG] fix unused variable warning and unneeded indirection; NFCSanjay Patel2019-05-142-3/+3
| | | | llvm-svn: 360640
* [SDAG, x86] allow targets to override test for binop opcodesSanjay Patel2019-05-1410-79/+119
| | | | | | | | This follows the pattern of the existing isCommutativeBinOp(). x86 shows improvements from vector narrowing for the min/max opcodes. llvm-svn: 360639
* [Docs] Differentiate between public and private APIJonas Devlieghere2019-05-141-2/+3
| | | | | | | On the homepage we should have a clear distinction between the public and private C++ APIs. llvm-svn: 360638
* PR41817: Fix regression in r359260 that caused the MS compatibilityRichard Smith2019-05-142-6/+46
| | | | | | | extension allowing a "static" declaration to follow an "extern" declaration to stop working. llvm-svn: 360637
* [coroutines] Fix spills of static array allocasGor Nishanov2019-05-132-6/+111
| | | | | | | | | | | | | | | | | | | | | | | Summary: CoroFrame was not considering static array allocas, and was only ever reserving a single element in the coroutine frame. This meant that stores to the non-zero'th element would corrupt later frame data. Store static array allocas as field arrays in the coroutine frame. Added test. Committed by Gor Nishanov on behalf of ben-clayton Reviewers: GorNishanov, modocache Reviewed By: GorNishanov Subscribers: Orlando, capn, EricWF, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D61372 llvm-svn: 360636
* [c++20] P1064R0: Allow virtual function calls in constant expressionRichard Smith2019-05-1315-57/+456
| | | | | | | | | | | | | evaluation. This reinstates r360559, reverted in r360580, with a fix to avoid crashing if evaluation-for-overflow mode encounters a virtual call on an object of a class with a virtual base class, and to generally not try to resolve virtual function calls to objects whose (notional) vptrs are not readable. (The standard rules are unclear here, but this seems like a reasonable approach.) llvm-svn: 360635
* [AMDGPU] gfx1010 clang targetStanislav Mekhanoshin2019-05-139-3/+37
| | | | | | Differential Revision: https://reviews.llvm.org/D61875 llvm-svn: 360634
* [Docs] Fix code formattign in variable.rstJonas Devlieghere2019-05-131-0/+1
| | | | | | Fixes missing newline between :: and the actual code. llvm-svn: 360632
* [CMake] Reinstate LLDB_CAN_USE_LLDB_SERVERJonas Devlieghere2019-05-132-3/+8
| | | | | | | | | | We cannot manipulate the LLDB_TOOL_LLDB_SERVER_BUILD directly from LLDBConfig.cmake because this would set the variable before the option is defined in AddLLVM.cmake. Instead, we need to use the LLDB_CAN_USE_LLDB_SERVER variable to conditionally add the lldb-server subdirectory. This should ensure the variable doesn't get cleared. llvm-svn: 360631
* [X86] Make `x86intrin.h`, `immintrin.h` includable with `-fno-gnu-inline-asm`.Volodymyr Sapsai2019-05-134-2/+12
| | | | | | | | | | | | | | | | | | | | | | | | | Currently `immintrin.h` includes `pconfigintrin.h` and `sgxintrin.h` which contain inline assembly. It causes failures when building with the flag `-fno-gnu-inline-asm`. Fix by excluding functions with inline assembly when this extension is disabled. So far there was no need to support `_pconfig_u32`, `_enclu_u32`, `_encls_u32`, `_enclv_u32` on platforms that require `-fno-gnu-inline-asm`. But if developers start using these functions, they'll have compile-time undeclared identifier errors which is preferrable to runtime errors. rdar://problem/49540880 Reviewers: craig.topper, GBuella, rnk, echristo Reviewed By: rnk Subscribers: jkorous, dexonsmith, cfe-commits Differential Revision: https://reviews.llvm.org/D61621 llvm-svn: 360630
* [gn] Fix buildVitaly Buka2019-05-133-0/+4
| | | | llvm-svn: 360629
* Removing an unused member variable; NFC.Aaron Ballman2019-05-131-2/+1
| | | | llvm-svn: 360628
* [X86] Use ISD::MERGE_VALUES to return from lowerAtomicArith instead of ↵Craig Topper2019-05-131-4/+8
| | | | | | | | | | | | | | calling ReplaceAllUsesOfValueWith and returning SDValue(). Returning SDValue() makes the caller think that nothing happened and it will end up executing the Expand path. This generates extra nodes that will need to be pruned as dead code. Returning an ISD::MERGE_VALUES will tell the caller that we'd like to make a change and it will take care of replacing uses. This will prevent falling into the Expand path. llvm-svn: 360627
* [OpenMP][Clang][BugFix] Split declares and math functions inclusion.Gheorghe-Teodor Bercea2019-05-1314-21/+66
| | | | | | | | | | | | | | | | Summary: This patches fixes an issue in which the __clang_cuda_cmath.h header is being included even when cmath or math.h headers are not included. Reviewers: jdoerfert, ABataev, hfinkel, caomhin, tra Reviewed By: tra Subscribers: tra, mgorny, guansong, cfe-commits Tags: #clang Differential Revision: https://reviews.llvm.org/D61765 llvm-svn: 360626
* Make language option `GNUAsm` discoverable with `__has_extension` macro.Volodymyr Sapsai2019-05-133-0/+9
| | | | | | | | | | | | | | | | This can be used for better support of `-fno-gnu-inline-asm` builds. rdar://problem/49540880 Reviewers: aaron.ballman, rsmith Reviewed By: aaron.ballman Subscribers: eraman, jkorous, dexonsmith, craig.topper, cfe-commits Differential Revision: https://reviews.llvm.org/D61619 llvm-svn: 360625
* Disable TestEnvironment on WindowsJonas Devlieghere2019-05-131-0/+2
| | | | | | | | | | | | | | The input source file seems to be triggering an error in the Visual Studio headers. > xstddef:338:2: error: ''auto' return without trailing return type; > deduced return types are a C++14 extension I tried converting the test to use the %build stuff Zachary added, but that seems to be missing some Darwin support. Disabling the test on Windows in the meantime. llvm-svn: 360624
* [AMDGPU] gfx1010 SearchableTableEmitter patch for NSAStanislav Mekhanoshin2019-05-131-1/+1
| | | | | | | | This part was accidentally missing from NSA image support commit. Differential Revision: https://reviews.llvm.org/D61868 llvm-svn: 360623
OpenPOWER on IntegriCloud