summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* Sink AddrMode back into TargetLowering, removing one of the mostChandler Carruth2013-01-075-45/+19
| | | | | | | | | peculiar headers under include/llvm. This struct still doesn't make a lot of sense, but it makes more sense down in TargetLowering than it did before. llvm-svn: 171739
* Remove LSR's use of the random AddrMode struct. These variables wereChandler Carruth2013-01-071-68/+76
| | | | | | | | | | | | already in a class, just inline the four of them. I suspect that this class could be simplified some to not always keep distinct variables for these things, but it wasn't clear to me how given the usage so I opted for a trivial and mechanical translation. This removes one of the two remaining users of a header in include/llvm which does nothing more than define a 4 member struct. llvm-svn: 171738
* s/parseStatement/parseStructuralElement/g in the UnwrappedLineParser.Manuel Klimek2013-01-072-9/+9
| | | | llvm-svn: 171737
* Make clang-format binary understand C++11.Daniel Jasper2013-01-071-0/+2
| | | | llvm-svn: 171736
* Switch the SCEV expander and LoopStrengthReduce to useChandler Carruth2013-01-0710-186/+164
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | TargetTransformInfo rather than TargetLowering, removing one of the primary instances of the layering violation of Transforms depending directly on Target. This is a really big deal because LSR used to be a "special" pass that could only be tested fully using llc and by looking at the full output of it. It also couldn't run with any other loop passes because it had to be created by the backend. No longer is this true. LSR is now just a normal pass and we should probably lift the creation of LSR out of lib/CodeGen/Passes.cpp and into the PassManagerBuilder. =] I've not done this, or updated all of the tests to use opt and a triple, because I suspect someone more familiar with LSR would do a better job. This change should be essentially without functional impact for normal compilations, and only change behvaior of targetless compilations. The conversion required changing all of the LSR code to refer to the TTI interfaces, which fortunately are very similar to TargetLowering's interfaces. However, it also allowed us to *always* expect to have some implementation around. I've pushed that simplification through the pass, and leveraged it to simplify code somewhat. It required some test updates for one of two things: either we used to skip some checks altogether but now we get the default "no" answer for them, or we used to have no information about the target and now we do have some. I've also started the process of removing AddrMode, as the TTI interface doesn't use it any longer. In some cases this simplifies code, and in others it adds some complexity, but I think it's not a bad tradeoff even there. Subsequent patches will try to clean this up even further and use other (more appropriate) abstractions. Yet again, almost all of the formatting changes brought to you by clang-format. =] llvm-svn: 171735
* Fix a mistaken commit that included some debugging code.David Tweed2013-01-071-1/+1
| | | | llvm-svn: 171734
* There was a switch fall-through in the parser for textual LLVM that causedDavid Tweed2013-01-074-6/+6
| | | | | | | | bogus comparison operands to default to eq/oeq. Fix that, fix a couple of tests that accidentally passed and test for bogus comparison opeartors explicitly. llvm-svn: 171733
* Reformat clang-formats source code.Daniel Jasper2013-01-075-74/+66
| | | | | | All changes done by clang-format itself. No functional changes. llvm-svn: 171732
* Prefer not to break after assignments.Daniel Jasper2013-01-072-18/+44
| | | | | | | | | | | | | | | | | | | | | | | | | | This addresses llvm.org/PR14830. Before: unsigned Cost = TTI.getMemoryOpCost(I->getOpcode(), VectorTy, SI->getAlignment(), SI->getPointerAddressSpace()); CharSourceRange LineRange = CharSourceRange::getTokenRange(TheLine.Tokens.front().Tok.getLocation(), TheLine.Tokens.back().Tok.getLocation()); After: unsigned Cost = TTI.getMemoryOpCost(I->getOpcode(), VectorTy, SI->getAlignment(), SI->getPointerAddressSpace()); CharSourceRange LineRange = CharSourceRange::getTokenRange( TheLine.Tokens.front().Tok.getLocation(), TheLine.Tokens.back().Tok.getLocation()); This required rudimentary changes to static initializer lists, but we are not yet formatting them in a reasonable way. That will be done in a subsequent patch. llvm-svn: 171731
* Make the MergeGlobals pass correctly handle the address space qualifiers of ↵Silviu Baranga2013-01-072-11/+36
| | | | | | the global variables. We partition the set of globals by their address space, and apply the same the trasnformation as before to merge them. llvm-svn: 171730
* Documentation: add a note that Clang on Mageia 2 can not find libstdc++ headersDmitri Gribenko2013-01-071-0/+3
| | | | llvm-svn: 171729
* R600/SIISelLowering.cpp: Suppress a warning. [-Wunused-variable]NAKAMURA Takumi2013-01-071-3/+2
| | | | llvm-svn: 171728
* CallingConvLower.h: Use bitfields like unsigned:1 instead of bool:1 .NAKAMURA Takumi2013-01-071-2/+2
| | | | | | Some compilers might be confused if bool were potentially signed integer. In my case, g++-4.7.0 miscompiled CodeGen/ARM. llvm-svn: 171727
* Simplify LoopVectorize to require target transform info and rely on itChandler Carruth2013-01-071-55/+43
| | | | | | | | | being present. Make a member of one of the helper classes a reference as part of this. Reformatting goodness brought to you by clang-format. llvm-svn: 171726
* Add style option for number of spaces before trailing comments.Daniel Jasper2013-01-073-5/+12
| | | | | | | In LLVM style, a single space should be enough. In Google style, two spaces are required. llvm-svn: 171725
* Do not break before "0" in pure virtual function declarations.Daniel Jasper2013-01-072-108/+144
| | | | | | | | | | | | | | | | | | | Before: virtual void write(ELFWriter *writer, OwningPtr<FileOutputBuffer> &buffer) = 0 After: virtual void write(ELFWriter *writerrr, OwningPtr<FileOutputBuffer> &buffer) = 0; This addresses llvm.org/PR14815. To implement this I introduced a line type during parsing and moved the definition of TokenType out of the struct for increased readability. Should have done the latter in a separate patch, but it would be hard to pull apart now. llvm-svn: 171724
* Merge the unused header file for LoopVectorizer into the source file.Chandler Carruth2013-01-072-538/+519
| | | | | | | | | | | | | | This makes the loop vectorizer match the pattern followed by roughly all other passses. =] Notably, this header file was braken in several regards: it contained a using namespace directive, global #define's that aren't globaly appropriate, and global constants defined directly in the header file. As a side benefit, lots of the types in this file become internal, which will cause the optimizer to chew on this pass more effectively. llvm-svn: 171723
* Switch BBVectorize to directly depend on having a TTI analysis.Chandler Carruth2013-01-0711-19/+19
| | | | | | | | | | | | | This could be simplified further, but Hal has a specific feature for ignoring TTI, and so I preserved that. Also, I needed to use it because a number of tests fail when switching from a null TTI to the NoTTI nonce implementation. That seems suspicious to me and so may be something that you need to look into Hal. I worked it by preserving the old behavior for these tests with the flag that ignores all target info. llvm-svn: 171722
* Add LICENSE.TXT covering contributions made by ARM.Tim Northover2013-01-072-0/+48
| | | | | | | | | | Absent a Contributor's License Agreement (CLA) with an LLVM legal entity and as reviewed and agreed with Chris Lattner, add a patent license covering future contributions from ARM until there is a CLA. This is to make explicit ARM's grant of patent rights to recipients of LLVM containing ARM-contributed material. llvm-svn: 171721
* Fix incorrect FIXME.Manuel Klimek2013-01-072-2/+4
| | | | | | | | The case that we wanted to write a test for cannot happen, as the UnwrappedLineParser already protects against it. Added an assert to prevent regressions of that assumption. llvm-svn: 171720
* Fix a slew of indentation and parameter naming style issues. This 80% ofChandler Carruth2013-01-071-42/+36
| | | | | | | | | | | | | | | | | | | | | | | | | | | | this patch brought to you by the tool clang-format. I wanted to fix up the names of constructor parameters because they followed a bit of an anti-pattern by naming initialisms with CamelCase: 'Tti', 'Se', etc. This appears to have been in an attempt to not overlap with the names of member variables 'TTI', 'SE', etc. However, constructor arguments can very safely alias members, and in fact that's the conventional way to pass in members. I've fixed all of these I saw, along with making some strang abbreviations such as 'Lp' be simpler 'L', or 'Lgl' be the word 'Legal'. However, the code I was touching had indentation and formatting somewhat all over the map. So I ran clang-format and fixed them. I also fixed a few other formatting or doxygen formatting issues such as using ///< on trailing comments so they are associated with the correct entry. There is still a lot of room for improvement of the formating and cleanliness of this code. ;] At least a few parts of the coding standards or common practices in LLVM's code aren't followed, the enum naming rules jumped out at me. I may mix some of these while I'm here, but not all of them. llvm-svn: 171719
* CFG.cpp: Fix wrapping logic when printing block preds/succs.Will Dietz2013-01-072-2/+39
| | | | | | | | | First check only wrapped with i==8, second wrapped at i==2,8,18,28,... This fix restores the intended behavior: i==8,18,28,... Found with -fsanitize=integer. llvm-svn: 171718
* Remove outdated FIXME and add explanation for error handling strategyManuel Klimek2013-01-071-2/+6
| | | | | | while parsing #define's. llvm-svn: 171717
* Remove outdated fixme.Manuel Klimek2013-01-071-3/+0
| | | | llvm-svn: 171716
* Do not ever allow using the full line in preprocessor directives.Manuel Klimek2013-01-072-5/+14
| | | | | | | | | | | | | | | | We would format: #define A \ int f(a); int i; as #define A \ int f(a);\ int i The fix will break up macro definitions that could fit a line, but hit the last column; fixing that is more involved, though, as it requires looking at the following line. llvm-svn: 171715
* Switch LoopIdiom pass to directly require target transform information.Chandler Carruth2013-01-071-3/+3
| | | | | | | | I'm sorry for duplicating bad style here, but I wanted to keep consistency. I've pinged the code review thread where this style was reviewed and changes were requested. llvm-svn: 171714
* Fix layouting of single-line-comments preceded by an escaped newline.Manuel Klimek2013-01-072-7/+10
| | | | | | | | | | | | | Previously, we'd format int i;\ // comment as int i; // comment The problem is that the escaped newline is part of the next token, and thus the raw token text of the comment doesn't start with "//". llvm-svn: 171713
* Rough out a new c'tor for the AttrBuilder class.Bill Wendling2013-01-073-109/+141
| | | | | | | | This c'tor takes the AttributeSet class as the parameter. It will eventually grab the attributes from the specified index and create a new attribute builder with those attributes. llvm-svn: 171712
* [ELF] Properly handle constant data.Michael J. Spencer2013-01-073-0/+8
| | | | | | With this hello world works with dietlibc on x86-64 Linux. llvm-svn: 171711
* [lld] Add -emit-yaml option. This outputs yaml instead of a binary.Michael J. Spencer2013-01-075-5/+45
| | | | llvm-svn: 171710
* [ELF] Emit the file header and program headers at the beginning of the first ↵Michael J. Spencer2013-01-074-51/+55
| | | | | | | | | | segment. This is required for the loader to correctly set AT_PHDR. With this lld can correctly link against dietlibc and get to main. llvm-svn: 171709
* [ELF] Change Segment to hold Chunks instead of Sections. No functionality ↵Michael J. Spencer2013-01-071-4/+7
| | | | | | change. llvm-svn: 171708
* [ELF] Remove dead variable.Michael J. Spencer2013-01-071-2/+0
| | | | llvm-svn: 171707
* Fix layouting of tokens with a leading escaped newline.Manuel Klimek2013-01-073-8/+28
| | | | | | | | | | | If a token follows directly on an escaped newline, the escaped newline is stored with the token. Since we re-layout escaped newlines, we need to treat them just like normal whitespace - thus, we need to increase the whitespace-length of the token, while decreasing the token length (otherwise the token length contains the length of the escaped newline and we double-count it while indenting). llvm-svn: 171706
* Put a higher penalty on breaking before "." or "->".Daniel Jasper2013-01-072-2/+7
| | | | | | | | | | | | | | This fixes llvm.org/PR14823. Before: local_state->SetString(prefs::kApplicationLocale, parent_local_state ->GetString(prefs::kApplicationLocale)); After: local_state->SetString( prefs::kApplicationLocale, parent_local_state->GetString(prefs::kApplicationLocale)); llvm-svn: 171705
* [ELF] Emit zeroFill as STT_OBJECT in the final executable. ld does this.Michael J. Spencer2013-01-072-1/+5
| | | | llvm-svn: 171704
* [ELF] Add common x86_64 relocation types.Michael J. Spencer2013-01-073-1/+53
| | | | llvm-svn: 171703
* Remove more unnecessary # operators with nothing to paste proceeding them.Craig Topper2013-01-072-40/+40
| | | | llvm-svn: 171702
* Documentation updates for pointer-to-member debug info added in r171698.David Blaikie2013-01-071-9/+11
| | | | llvm-svn: 171701
* PR14759: Improve/correct support for debug info for C++ member pointers.David Blaikie2013-01-072-32/+7
| | | | | | | | | | Using added LLVM functionality in r171698. This works in GDB for member variable pointers but not member function pointers. See the LLVM commit and GDB bug 14998 for details. Un-xfailing cases in the GDB 7.5 test suite will follow. llvm-svn: 171699
* PR14759: Debug info support for C++ member pointers.David Blaikie2013-01-076-0/+67
| | | | | | | | This works fine with GDB for member variable pointers, but GDB's support for member function pointers seems to be quite unrelated to DW_TAG_ptr_to_member_type. (see GDB bug 14998 for details) llvm-svn: 171698
* Remove # from the beginning and end of def names. The # is a paste operator ↵Craig Topper2013-01-074-93/+93
| | | | | | and should only be used with something to paste on either side. llvm-svn: 171697
* Remove # from the beginning and end of def names.Craig Topper2013-01-074-283/+283
| | | | llvm-svn: 171696
* Revert r171140. We don't actually need to support #NAME. Because NAME by ↵Craig Topper2013-01-071-5/+1
| | | | | | itself is interpreted just fine. llvm-svn: 171695
* Remove unnecessary # tokens at the beginning and end of defm names.Craig Topper2013-01-071-10/+10
| | | | llvm-svn: 171694
* [Driver] x86-64 should use ELF 64.Michael J. Spencer2013-01-072-2/+5
| | | | llvm-svn: 171693
* [ELF] Give a slightly better message for unknown relocation types.Michael J. Spencer2013-01-071-1/+4
| | | | llvm-svn: 171692
* Make SimplifyCFG simply depend upon TargetTransformInfo and pass itChandler Carruth2013-01-073-48/+54
| | | | | | | | | | | | | through as a reference rather than a pointer. There is always *some* implementation of this available, so this simplifies code by not having to test for whether it is available or not. Further, it turns out there were piles of places where SimplifyCFG was recursing and not passing down either TD or TTI. These are fixed to be more pedantically consistent even though I don't have any particular cases where it would matter. llvm-svn: 171691
* Move the initialization to the Analysis library as well as the pass.Chandler Carruth2013-01-072-1/+1
| | | | | | | This was (somewhat distressingly) only caught be the ocaml bindings tests... llvm-svn: 171690
* Update comment.Eric Christopher2013-01-071-1/+1
| | | | llvm-svn: 171689
OpenPOWER on IntegriCloud