summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* Don't crash on code where the user put __attribute__((constructor)) onDan Gohman2012-04-181-1/+5
| | | | | | a function with arguments. This fixes rdar://11265785. llvm-svn: 155073
* Factor out core linking options from Platform in a new ResolverOptionsNick Kledzik2012-04-1821-626/+669
| | | | | | | | | | class. Change Resolver to no longer use Platform. Core linking now issues errors directly. We need to factor that out later. Rework how Darwin executable writer finds "main" atom. It now adds to core linking an Atom which has a Reference to "main". llvm-svn: 155060
* Add a flag to rebranch if we need to.Bill Wendling2012-04-181-4/+14
| | | | llvm-svn: 155049
* Revert some Hexagon builtin commits to match reverts done to LLVM inChandler Carruth2012-04-185-1628/+829
| | | | | | | | | | | | | | | | | | | r155047. See the LLVM log for the primary motivation: http://llvm.org/viewvc/llvm-project?rev=155047&view=rev Primary commit r154828: - Several issues were raised in review, and fixed in subsequent commits. - Follow-up commits also reverted, and which should be folded into the original before reposting: - r154837: Re-add the 'undef BUILTIN' thing to fix the build. - r154928: Fix build warnings, re-add (and correct) header and license - r154937: Typo fix. Please resubmit this patch with the relevant LLVM resubmission. llvm-svn: 155048
* This reverts a long string of commits to the Hexagon backend. TheseChandler Carruth2012-04-1839-13677/+2375
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | commits have had several major issues pointed out in review, and those issues are not being addressed in a timely fashion. Furthermore, this was all committed leading up to the v3.1 branch, and we don't need piles of code with outstanding issues in the branch. It is possible that not all of these commits were necessary to revert to get us back to a green state, but I'm going to let the Hexagon maintainer sort that out. They can recommit, in order, after addressing the feedback. Reverted commits, with some notes: Primary commit r154616: HexagonPacketizer - There are lots of review comments here. This is the primary reason for reverting. In particular, it introduced large amount of warnings due to a bad construct in tablegen. - Follow-up commits that should be folded back into this when reposting: - r154622: CMake fixes - r154660: Fix numerous build warnings in release builds. - Please don't resubmit this until the three commits above are included, and the issues in review addressed. Primary commit r154695: Pass to replace transfer/copy ... - Reverted to minimize merge conflicts. I'm not aware of specific issues with this patch. Primary commit r154703: New Value Jump. - Primarily reverted due to merge conflicts. - Follow-up commits that should be folded back into this when reposting: - r154703: Remove iostream usage - r154758: Fix CMake builds - r154759: Fix build warnings in release builds - Please incorporate these fixes and and review feedback before resubmitting. Primary commit r154829: Hexagon V5 (floating point) support. - Primarily reverted due to merge conflicts. - Follow-up commits that should be folded back into this when reposting: - r154841: Remove unused variable (fixing build warnings) There are also accompanying Clang commits that will be reverted for consistency. llvm-svn: 155047
* Make sure EmulateInstructionARM doesn't have to have "armv4", "armv6", ↵Greg Clayton2012-04-181-4/+4
| | | | | | "armv7" as the exact architecture name, the arch name can just start with any of these strings. We need to be able to recognize different variants that might come along and not fail to backtrace completely (which happens when we aren't able to find an architecture that matches) when we don't have exact matches. llvm-svn: 155045
* Richard Smith: __int128 is a GNU keyword which clang 3.1 supports. This ↵Howard Hinnant2012-04-181-2/+2
| | | | | | usage of it as an identifier will need to be renamed. llvm-svn: 155044
* LiveIntervalUpdate validators weren't recorded after the calls to ↵Pete Cooper2012-04-181-6/+6
| | | | | | std::for_each. Turns out std::for_each doesn't update the variable passed in for the functor but instead copy constructs a new one. llvm-svn: 155041
* Document that StringMap iteration order is non-deterministic.Jim Grosbach2012-04-181-0/+3
| | | | llvm-svn: 155040
* Revert "Replace some uses of std:map<std::string,...> with StringMap."Jim Grosbach2012-04-181-6/+6
| | | | | | | StringMap iterators are not deterministic, and that's more important here than speed or memory. llvm-svn: 155039
* tblgen: remove duplicated newlines.Benjamin Kramer2012-04-183-6/+6
| | | | llvm-svn: 155038
* Replace some uses of std:map<std::string,...> with StringMap.Jim Grosbach2012-04-181-6/+6
| | | | llvm-svn: 155037
* objective-c: Issue diagnostic when an implicitFariborz Jahanian2012-04-182-3/+50
| | | | | | | property accessor (getter) missing, instead of crashing. // rdar://11273060 llvm-svn: 155036
* SourceMgr: Colorize diagnostics.Benjamin Kramer2012-04-182-9/+46
| | | | | | Same color scheme as clang uses. The colors are only enabled if the output is a tty. llvm-svn: 155035
* Use SmallVector for the requirements on an InstAlias.Jim Grosbach2012-04-181-1/+1
| | | | llvm-svn: 155034
* Tidy up. Formatting.Jim Grosbach2012-04-181-2/+3
| | | | llvm-svn: 155032
* Mark instruction classes ArithLogicR, ArithLogicI and LoadUpper as ↵Akira Hatanaka2012-04-181-1/+5
| | | | | | isRematerializable. llvm-svn: 155031
* Delete blank line.Akira Hatanaka2012-04-181-1/+0
| | | | llvm-svn: 155030
* Move a few more warnings to use PrintWarning().Jim Grosbach2012-04-182-7/+9
| | | | llvm-svn: 155027
* Tidy up. No need for a Twine here, as it's just constants.Jim Grosbach2012-04-181-4/+4
| | | | llvm-svn: 155026
* Formatting.Jim Grosbach2012-04-181-3/+2
| | | | llvm-svn: 155025
* Tidy up. Add a '.' at the end of the sentence.Jim Grosbach2012-04-181-1/+1
| | | | llvm-svn: 155024
* Fix copy/paste-o.Jim Grosbach2012-04-181-1/+1
| | | | llvm-svn: 155016
* Clean up warning text. Remove extraneous prefix.Jim Grosbach2012-04-181-4/+4
| | | | llvm-svn: 155015
* TableGen use PrintWarning rather than fprintf(stderr,...) for warnings.Jim Grosbach2012-04-181-12/+13
| | | | | | That way we get source line number information from the diagnostics. llvm-svn: 155014
* Tidy up. Trailing whitespace.Jim Grosbach2012-04-181-2/+2
| | | | llvm-svn: 155013
* TableGen add warning diagnostic helper functions.Jim Grosbach2012-04-182-0/+21
| | | | llvm-svn: 155012
* Suppress -Wunused-variable warning in -Asserts buildMatt Beaumont-Gay2012-04-181-1/+1
| | | | llvm-svn: 155011
* FileCheckizeJakob Stoklund Olesen2012-04-181-2/+59
| | | | llvm-svn: 155010
* Nobody likes shifty instructions, but that was a bit strong.Jakob Stoklund Olesen2012-04-181-1/+1
| | | | llvm-svn: 155009
* VerifyICE: Pass PartialDiagnostics by reference.Benjamin Kramer2012-04-182-7/+8
| | | | llvm-svn: 155005
* Added support for disassembling unpredictable swp/swpb ARM instructions.Silviu Baranga2012-04-184-2/+33
| | | | llvm-svn: 155004
* Fix the bahavior of the disassembler when decoding unpredictable mrs ↵Silviu Baranga2012-04-182-5/+33
| | | | | | instructions on ARM. Now the diasassembler emmits warnings instead of errors. llvm-svn: 155002
* Added support for unpredictable mcrr/mcrr2/mrrc/mrrc2 ARM instruction in the ↵Silviu Baranga2012-04-184-4/+53
| | | | | | disassembler. Since the upredicability conditions are complex, C++ code was added to handle them. llvm-svn: 155001
* Fixed decoding for the ARM cdp2 instruction. The restriction on the ↵Silviu Baranga2012-04-183-1/+10
| | | | | | coprocessor number was removed for this instruction. llvm-svn: 155000
* Add suport for unpredicatble cases of the cmp, tst, teq and cmnz ARM ↵Silviu Baranga2012-04-182-2/+40
| | | | | | instructions in the disassembler. llvm-svn: 154999
* MIPS: Followup to r154606. Expand list of accepted MIPS target features in ↵Simon Atanasyan2012-04-181-1/+4
| | | | | | the MipsTargetInfoBase::setFeatureEnabled() routine. llvm-svn: 154998
* fix GCC version comparison wrt. equal patch suffices (and squash two typos)Gabor Greif2012-04-181-2/+3
| | | | llvm-svn: 154997
* SmallPtrSet: Reuse DenseMapInfo's pointer hash function instead of inventing ↵Benjamin Kramer2012-04-182-4/+2
| | | | | | | | | | a bad one ourselves. DenseMap's hash function uses slightly more entropy and reduces hash collisions significantly. I also experimented with Hashing.h, but it didn't gave a lot of improvement while being much more expensive to compute. llvm-svn: 154996
* Move the JIT flags from llc to lli. These flags showed up as part of movingNick Lewycky2012-04-182-26/+28
| | | | | | backend flags in TargetOptions.h into their own class in r145714. llvm-svn: 154993
* A bit of cleanup in the TestingGuide doc - dg.exp files no longer existEli Bendersky2012-04-181-9/+9
| | | | llvm-svn: 154992
* Some formatting and grammar fixes in the FileCheck documentationEli Bendersky2012-04-181-29/+29
| | | | llvm-svn: 154990
* Adds a FixedCompilationDatabase to be able to specify tool parametersManuel Klimek2012-04-185-9/+170
| | | | | | at the command line. llvm-svn: 154989
* Use a heavy hammer to fix PR12573.Bill Wendling2012-04-181-0/+9
| | | | | | | | | | | If the loop contains invoke instructions, whose unwind edge escapes the loop, then don't try to unswitch the loop. Doing so may cause the unwind edge to be split, which not only is non-trivial but doesn't preserve loop simplify information. Fixes PR12573 llvm-svn: 154987
* Fix bad typo reported by I-Jui Sung.Ted Kremenek2012-04-181-1/+1
| | | | llvm-svn: 154986
* Remove AVX vpermil intrinsics. I removed their uses from clang headers and ↵Craig Topper2012-04-183-25/+38
| | | | | | builtins a while back. llvm-svn: 154985
* Added an iOS local debugging optimization when reading the __LINKEDIT ↵Greg Clayton2012-04-181-11/+51
| | | | | | section data for files in the dyld shared cache. llvm-svn: 154984
* loop-reduce: Add an early bailout to catch extremely large loops.Andrew Trick2012-04-181-0/+17
| | | | | | | | | | | | | | This introduces a threshold of 200 IV Users, which is very conservative but should be sufficient to avoid serious compile time sink or stack overflow. The llvm test-suite with LTO never exceeds 190 users per loop. The bug doesn't relate to a specific type of loop. Checking in an arbitrary giant loop as a unit test would be silly. Fixes rdar://11262507. llvm-svn: 154983
* As per John McCall comment:Francois Pichet2012-04-181-4/+4
| | | | | | Follow up to r154924: check that we are in a static CMethodDecl to enable the Microsoft bug emulation regarding access to protected member during PTM creation. Not just any static function. llvm-svn: 154982
* fix display of source lines with null charactersSeth Cantrell2012-04-182-2/+5
| | | | llvm-svn: 154981
OpenPOWER on IntegriCloud