summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* Add link to polyhedral.infoTobias Grosser2014-02-111-0/+2
| | | | | | This website provides general information about polyhedral compilation. llvm-svn: 201192
* Remove MayAliasSet classTobias Grosser2014-02-115-201/+1
| | | | | | | The MayAliasSet class is currently not used and just confuses people. We can reintroduce it in case need a more precise tracking of alias sets. llvm-svn: 201191
* Reapply r201180 with an additional error path.Adrian Prantl2014-02-114-5/+180
| | | | | | | | | | | | Debug info: Emit values in subregisters that do not have a separate DWARF register number by emitting a super-register + DW_OP_bit_piece. This is necessary because on x86_64, there are no DWARF register numbers for i386-style subregisters. Fixes a bunch of FIXMEs. rdar://problem/16015314 llvm-svn: 201190
* Fix build error caused by r201186 (GCC's "declaration of X changes meaning ↵Alexander Kornienko2014-02-111-8/+7
| | | | | | of Y"). llvm-svn: 201189
* Revert "Debug info: Emit values in subregisters that do not have a separate"Adrian Prantl2014-02-114-175/+5
| | | | | | This reverts commit r201179 for buildbot breakage. llvm-svn: 201188
* DebugInfo: Don't include the name of the CU file in the line table file list ↵David Blaikie2014-02-112-3/+15
| | | | | | | | | | when it's unneeded This comes up in empty files or files containing #file directives that never reference the actual source file name. Came up in a small test of line tables I was playing with. llvm-svn: 201187
* Expose the name of the checker producing each diagnostic message.Alexander Kornienko2014-02-1169-453/+629
| | | | | | | | | | | | | | | | | | | | | | | | | | Summary: In clang-tidy we'd like to know the name of the checker producing each diagnostic message. PathDiagnostic has BugType and Category fields, which are both arbitrary human-readable strings, but we need to know the exact name of the checker in the form that can be used in the CheckersControlList option to enable/disable the specific checker. This patch adds the CheckName field to the CheckerBase class, and sets it in the CheckerManager::registerChecker() method, which gets them from the CheckerRegistry. Checkers that implement multiple checks have to store the names of each check in the respective registerXXXChecker method. Reviewers: jordan_rose, krememek Reviewed By: jordan_rose CC: cfe-commits Differential Revision: http://llvm-reviews.chandlerc.com/D2557 llvm-svn: 201186
* DebugInfo: Remove dependence on file numbering in the line table.David Blaikie2014-02-112-2/+4
| | | | | | | | These tests were unnecessarily sensitive to the presence and ordering of elements in the line table file_names list which will break on a future change I'm working on. llvm-svn: 201185
* Require CMake 2.8.12.1 or later on WindowsHans Wennborg2014-02-111-1/+7
| | | | | | | That version knows to use the /FS flag, which is needed for building lib/interception with VS2013. llvm-svn: 201184
* Remove bad debug info test.David Blaikie2014-02-111-20/+0
| | | | | | | | | | Another test that's testing through assembly from Clang which is problematic. An attempt to upgrade this to just be an IR-generation test was unsuccessful (I was unable to get this test to fail) but it looks like there's other, better, test coverage in this area (test/CodeGenObjC/debuginfo-properties.m) anyway. llvm-svn: 201183
* update of the clang version (should probably managed in the configure)Sylvestre Ledru2014-02-111-1/+1
| | | | llvm-svn: 201182
* whitespaceAdrian Prantl2014-02-111-1/+1
| | | | llvm-svn: 201181
* Debug info: Emit values in subregisters that do not have a separateAdrian Prantl2014-02-114-5/+175
| | | | | | | | | | | DWARF register number by emitting a super-register + DW_OP_bit_piece. This is necessary because on x86_64, there are no DWARF register numbers for i386-style subregisters. Fixes a bunch of FIXMEs. rdar://problem/16015314 llvm-svn: 201180
* make llvm-dwarfdump a little more resilient when parsing .debug_locAdrian Prantl2014-02-111-1/+3
| | | | | | | | sections. The call to data.getUnsigned(&Offset, AddressSize) only increments Offset if the read succeeds, which will result in an infinite loop. llvm-svn: 201179
* Remove bad debug info test.David Blaikie2014-02-111-36/+0
| | | | | | | | | | | | | | This test case doesn't belong in Clang (it's testing IndVarSimplify) but in an effort to reproduce the test case this was intended to cover (by essentially reverting r134441) I wasn't able to reproduce the failure this test case should've produced. So I haven't ported this down to LLVM, instead I'm just deleting it. I suspect the test is just underconstrained, but I've no great interest in trying hard to fix it right now - if anyone else wants to, I'd be more than welcome to that. llvm-svn: 201178
* R600/SI: Fix assertion on infinite loops.Matt Arsenault2014-02-112-2/+21
| | | | | | | This isn't the most useful case to fix in the real world, but bugpoint runs into this. llvm-svn: 201177
* InstCombine: Teach icmp merging about the equivalence of bit tests and ↵Benjamin Kramer2014-02-112-23/+76
| | | | | | | | | UGE/ULT with a power of 2. This happens in bitfield code. While there reorganize the existing code a bit. llvm-svn: 201176
* MS ABI: Add support for the -vm{b,g,s,m,v} flagsDavid Majnemer2014-02-1113-37/+157
| | | | | | | | | These flags control the inheritance model initially used by the translation unit. Differential Revision: http://llvm-reviews.chandlerc.com/D2741 llvm-svn: 201175
* Tidy up a bit. Formatting only.Jim Grosbach2014-02-111-8/+8
| | | | llvm-svn: 201174
* ARM: Thumb2 LDR(literal) can target SP.Jim Grosbach2014-02-112-1/+6
| | | | | | | | | Fix a slightly overzealous destination register restriction for the 'without .w' alias. Add some explicit testcases. rdar://16033140 llvm-svn: 201173
* Test commit -- Remove an extra newline from Toolchains.cppJonathan Roelofs2014-02-111-1/+0
| | | | llvm-svn: 201172
* When a multiline expression produces output, the multi-line help message is ↵Greg Clayton2014-02-112-15/+7
| | | | | | | | printed twice. <rdar://problem/16031890> llvm-svn: 201171
* SCEV: Cast switched values to make -Wswitch more useful.Benjamin Kramer2014-02-111-16/+16
| | | | llvm-svn: 201170
* Remove decorator for llvm.org/pr17233Ed Maste2014-02-111-1/+0
| | | | | | | This test was skipped as it used to segfault on FreeBSD. It seems the original issue has since been fixed, so have the test run again. llvm-svn: 201169
* Remove FreeBSD decorator for llvm.org/pr17499 (no longer fails)Ed Maste2014-02-112-2/+0
| | | | llvm-svn: 201168
* Remove decorators for llvm.org/pr17920 fixed by r200646Ed Maste2014-02-112-8/+0
| | | | llvm-svn: 201167
* Update decorator for FreeBSD failure pr18066Ed Maste2014-02-111-1/+1
| | | | | | | | The test expects the inferior to have exited, but it is still stopped. llvm.org/pr18066 llvm-svn: 201166
* [Sema] Revert the change in r200622 that allowed integer casts to silence ↵Argyrios Kyrtzidis2014-02-112-5/+2
| | | | | | | | | | -Wnon-literal-null-conversion in C code. It is actually useful to warn in such cases, thanks to Dmitri for pushing on this and making us see the light! Related to rdar://15925483 and rdar://15922612. The latter radar is where the usefulness of the warning is most clear. llvm-svn: 201165
* Don’t leak memory when reading memory and we do an early return for error ↵Greg Clayton2014-02-111-4/+3
| | | | | | conditions. llvm-svn: 201164
* Improve the declaration when LDPO_PIE is not available.Sylvestre Ledru2014-02-111-7/+6
| | | | | | Thanks to İsmail Dönmez for the better declaration. llvm-svn: 201163
* 'nonnull(1)' on a block parameter should apply to the block's argument.Jordan Rose2014-02-113-17/+36
| | | | | | | | | | | | | | | Thanks to r199467, __attribute__((nonnull)) (without arguments) can apply directly to parameters, instead of being applied to the whole function. However, the old form of nonnull (with an argument index) could also apply to the arguments of function and block pointers, and both of these can be passed as parameters. Now, if 'nonnull' with an argument is found on a parameter, /and/ the parameter is a function or block pointer, it is handled the old way. PR18795 llvm-svn: 201162
* If LDPO_PIE is not defined (before binutils 2.23 version), Sylvestre Ledru2014-02-111-0/+7
| | | | | | | use the hardcoded declaration 3 See r201110 for the initial change llvm-svn: 201161
* Add a note about using "Differential Revision:" in commit messagesMark Seaborn2014-02-111-0/+20
| | | | | | | | | | I noticed this convention from the commit logs. It seems like it would be useful to document it, to encourage other committers to link back to code reviews in their commits. Differential Revision: http://llvm-reviews.chandlerc.com/D2678 llvm-svn: 201160
* ScalarEvolution: Analyze trip count of loops with a switch guarding the exit.Benjamin Kramer2014-02-113-15/+90
| | | | llvm-svn: 201159
* Teach the DAGCombiner how to fold concat_vector nodes when the input is twoRobert Lougher2014-02-113-0/+141
| | | | | | | | | | | | | BUILD_VECTOR nodes, e.g.: (concat_vectors (BUILD_VECTOR a1, a2, a3, a4), (BUILD_VECTOR b1, b2, b3, b4)) -> (BUILD_VECTOR a1, a2, a3, a4, b1, b2, b3, b4) This fixes an issue with AVX, where a sequence was not recognized as a 256-bit vbroadcast due to the concat_vectors. llvm-svn: 201158
* Rename some PCH-related errors to have 'err_' as their prefixDmitri Gribenko2014-02-112-7/+7
| | | | llvm-svn: 201157
* Fix a typo in the function name that I just introduced (r201155)Dmitri Gribenko2014-02-118-26/+26
| | | | llvm-svn: 201156
* libclang: audit all APIs that accept a CXTranslationUnit to make sure thatDmitri Gribenko2014-02-118-26/+123
| | | | | | | checks for invalid translation unit are in place. Also, while there, add log output for this case. llvm-svn: 201155
* [asan] Fix Windows build.Evgeniy Stepanov2014-02-111-0/+5
| | | | llvm-svn: 201154
* Add decorator for llvm.org/pr18805 until root cause is identifiedEd Maste2014-02-111-0/+1
| | | | llvm-svn: 201153
* [sanitizer] Fix build.Evgeniy Stepanov2014-02-113-15/+15
| | | | llvm-svn: 201152
* [sanitizer] Use system unwinder in signal handlers on Android.Evgeniy Stepanov2014-02-1111-26/+111
| | | | | | | | | | Because of the way Bionic sets up signal stack frames, libc unwinder is unable to step through it, resulting in broken SEGV stack traces. Luckily, libcorkscrew.so on Android implements an unwinder that can start with a signal context, thus sidestepping the issue. llvm-svn: 201151
* [Sanitizer] External symbolizer refactoring: split protocol for ↵Alexey Samsonov2014-02-111-129/+164
| | | | | | | | | | communicating with llvm-symbolizer binary and external process handling into separate classes. No functionality change. llvm-svn: 201150
* [AArch64] Add missing PCRel relocations for AArch64 in RuntimeDyldELFBradley Smith2014-02-111-0/+42
| | | | llvm-svn: 201149
* [LPM] Switch LICM to actively use LCSSA in addition to preserving it.Chandler Carruth2014-02-113-169/+172
| | | | | | | | | | | | | | | | | | | | | | | Fixes PR18753 and PR18782. This is necessary for LICM to preserve LCSSA correctly and efficiently. There is still some active discussion about whether we should be using LCSSA, but we can't just immediately stop using it and we *need* LICM to preserve it while we are using it. We can restore the old SSAUpdater driven code if and when there is a serious effort to remove the reliance on LCSSA from all of the loop passes. However, this also serves as a great example of why LCSSA is very nice to have. This change significantly simplifies the process of sinking instructions for LICM, and makes it quite a bit less expensive. It wouldn't even be as complex as it is except that I had to start the process of removing the big recursive LCSSA formation hammer in order to switch even this much of the re-forming code to asserting that LCSSA was preserved. I'll fully remove that next just to tidy things up until the LCSSA debate settles one way or the other. llvm-svn: 201148
* Remove unused config.llvm_symbolizer_pathAlexey Samsonov2014-02-111-3/+0
| | | | llvm-svn: 201147
* libclang: use namespace qualifiers in the macro, so that it does not requireDmitri Gribenko2014-02-111-1/+2
| | | | | | 'using namespace clang::cxindex'. llvm-svn: 201146
* AArch64: share slgihtly more NEON implementation with ARM.Tim Northover2014-02-111-14/+10
| | | | | | | | The s64/u64 vcvt conversion operations are actually pretty much identical to the s32/u32 ones in implementation, and can be shared with just one extra variable. llvm-svn: 201145
* XCore target: fix const section handlingRobert Lytton2014-02-118-48/+107
| | | | | | | | | | | | Xcore target ABI requires const data that is externally visible to be handled differently if it has C-language linkage rather than C++ language linkage. Clang now emits ".cp.rodata" section information. All other externally visible constant data will be placed in the DP section. llvm-svn: 201144
* XCore target: Lower ATOMIC_LOAD & ATOMIC_STORERobert Lytton2014-02-113-0/+146
| | | | llvm-svn: 201143
OpenPOWER on IntegriCloud