summaryrefslogtreecommitdiffstats
path: root/clang
Commit message (Collapse)AuthorAgeFilesLines
* Fix implementation of _mm_pause.Anders Carlsson2008-12-261-1/+1
| | | | llvm-svn: 61441
* OK, all tests pass. Let's start using the SSE2 headerAnders Carlsson2008-12-262-2/+3
| | | | llvm-svn: 61440
* Add two more builtinsAnders Carlsson2008-12-261-0/+2
| | | | llvm-svn: 61439
* Complete the test after adding handling of merged attributes on decls.Anton Korobeynikov2008-12-261-3/+1
| | | | llvm-svn: 61438
* Add full dllimport / dllexport support: both sema checks and codegen.Anton Korobeynikov2008-12-267-15/+135
| | | | | | Patch by Ilya Okonsky llvm-svn: 61437
* Testcase for stdcall/fastcall sema checks.Anton Korobeynikov2008-12-261-0/+12
| | | | | | Patch by Ilya Okonsky! llvm-svn: 61436
* Fix formattingAnders Carlsson2008-12-261-13/+13
| | | | llvm-svn: 61435
* Add _mm_pause and _MM_SHUFFLEAnders Carlsson2008-12-261-0/+7
| | | | llvm-svn: 61434
* More SSE2 intrinsicsAnders Carlsson2008-12-261-0/+137
| | | | llvm-svn: 61433
* More SSE2 intrinsicsAnders Carlsson2008-12-251-2/+344
| | | | llvm-svn: 61432
* Add parser support for __forceinline, __w64, __ptr64.Steve Naroff2008-12-254-3/+21
| | | | llvm-svn: 61431
* Add parser support for __cdecl, __stdcall, and __fastcall.Steve Naroff2008-12-254-3/+48
| | | | | | Change preprocessor implementation of _cdecl to reference __cdecl. llvm-svn: 61430
* Add Fedora 10 GCC paths.Zhongxing Xu2008-12-251-0/+10
| | | | llvm-svn: 61429
* Make _mm_add_si64 inlinedAnders Carlsson2008-12-251-1/+1
| | | | llvm-svn: 61428
* Add explicit "fuzzy" parse support for Microsoft declspec.Steve Naroff2008-12-245-1/+29
| | | | | | Remove previous __declspec macro that would effectively erase the construct prior to parsing. llvm-svn: 61422
* Fix try statement deserialization.Sebastian Redl2008-12-241-0/+3
| | | | llvm-svn: 61421
* Add a bunch of comments and FIXMEs.Ted Kremenek2008-12-241-13/+90
| | | | llvm-svn: 61419
* set region default value if there are not enough init values for array and ↵Zhongxing Xu2008-12-241-5/+5
| | | | | | struct. llvm-svn: 61418
* Add serialization support for TypeOfType.Zhongxing Xu2008-12-242-1/+31
| | | | llvm-svn: 61417
* Add serialization support for TypeOfExpr.Zhongxing Xu2008-12-242-0/+28
| | | | llvm-svn: 61416
* Keep track of template arguments when we parse them. Right now, we don't ↵Douglas Gregor2008-12-2415-87/+284
| | | | | | actually do anything with the template arguments, but they'll be used to create template declarations llvm-svn: 61413
* Add integer arithmetic intrinsicsAnders Carlsson2008-12-241-0/+156
| | | | llvm-svn: 61412
* Add store/set/load intrisicsAnders Carlsson2008-12-241-0/+103
| | | | llvm-svn: 61411
* Start implementing some SSE2 instructionsAnders Carlsson2008-12-241-0/+371
| | | | llvm-svn: 61410
* Comment and fixup GDM entries for RegionStore to use unique 'tag classes' to ↵Ted Kremenek2008-12-241-40/+66
| | | | | | identify GDM entries. llvm-svn: 61409
* Correct the order in which we cope with end-of-class-definitionDouglas Gregor2008-12-2416-36/+124
| | | | | | | | | | | | | | | | | | | | | | semantics and improve our handling of default arguments. Specifically, we follow this order: - As soon as the see the '}' in the class definition, the class is complete and we add any implicit declarations (default constructor, copy constructor, etc.) to the class. - If there are any default function arguments, parse them - If there were any inline member function definitions, parse them As part of this change, we now keep track of the the fact that we've seen unparsed default function arguments within the AST. See the new ParmVarDecl::hasUnparsedDefaultArg member. This allows us to properly cope with calls inside default function arguments to other functions where we're making use of the default arguments. Made some C++ error messages regarding failed initializations more specific. llvm-svn: 61406
* Sema for fastcall/stdcall stuff. Tests will follow.Anton Korobeynikov2008-12-232-6/+37
| | | | | | Patch by Ilya Okonsky! llvm-svn: 61394
* Fix misguided type selectionDouglas Gregor2008-12-231-1/+1
| | | | llvm-svn: 61393
* When determining whether a class type has a const copy constructor, beDouglas Gregor2008-12-231-14/+5
| | | | | | | sure to look at all of the results returned by name lookup. Fixes <rdar://problem/6465262> llvm-svn: 61388
* Last patch, for now, to privde ObjC's encoding of types.Fariborz Jahanian2008-12-232-1/+20
| | | | | | We now pass all gcc's encoding compatibility tests. llvm-svn: 61387
* Don't push OverloadedFunctionDecls onto the chain of declarationsDouglas Gregor2008-12-237-176/+154
| | | | | | | | attached to an identifier. Instead, all overloaded functions will be pushed into scope, and we'll synthesize an OverloadedFunctionDecl on the fly when we need it. llvm-svn: 61386
* Fix <rdar://problem/6465284> clang ObjC rewriter: objc_super messed up again.Steve Naroff2008-12-232-16/+36
| | | | llvm-svn: 61384
* Lot more encoding work. We are closing the gap toFariborz Jahanian2008-12-232-7/+55
| | | | | | gcc compatibilty in all aspects of encoding now. llvm-svn: 61383
* Remove old PTH token-generation test harness.Ted Kremenek2008-12-231-71/+0
| | | | llvm-svn: 61382
* PTH: Remove some methods and simplify some conditions in PTHLexer::Lex(). ↵Ted Kremenek2008-12-233-63/+32
| | | | | | No big functionality change. llvm-svn: 61381
* PTH: Use 3 bytes instead of 4 bytes to encode the persistent ID for a token.Ted Kremenek2008-12-232-10/+16
| | | | | | | - This reduces the PTH size for Cocoa.h by 7%. - The increases PTH -Eonly speed for Cocoa.h by 0.8%. llvm-svn: 61377
* Cosmetics: rename a variable and tighten spacing. No functionality change.Ted Kremenek2008-12-231-4/+2
| | | | llvm-svn: 61375
* Merge pr-3188.cpp into constructor.cpp.Sebastian Redl2008-12-232-3/+10
| | | | llvm-svn: 61370
* Add a setter for CVR qualifiers, patch by Lukasz Janyst!Chris Lattner2008-12-231-0/+1
| | | | llvm-svn: 61367
* PTH:Ted Kremenek2008-12-232-6/+10
| | | | | | | | - Encode the token length with 2 bytes instead of 4. - This reduces the size of the .pth file for Cocoa.h by 12%. - This speeds up PTH time (-Eonly) on Cocoa.h by 1.6%. llvm-svn: 61364
* PTH:Ted Kremenek2008-12-232-32/+41
| | | | | | | | | | | | | - In PTHLexer::Lex read all of the token data from PTH file before constructing the token. The idea is to enhance locality. - Do not use Read8/Read32 in PTHLexer::Lex. Inline these operations manually. - Change PTHManager::ReadIdentifierInfo() to PTHManager::GetIdentifierInfo(). They are functionally the same except that PTHLexer::Lex() reads the persistent id. These changes result in a 3.3% speedup for PTH on Cocoa.h (-Eonly). llvm-svn: 61363
* PTH:Ted Kremenek2008-12-234-137/+144
| | | | | | | | | | | | - Embed 'eom' tokens in PTH file. - Use embedded 'eom' tokens to not lazily generate them in the PTHLexer. This means that PTHLexer can always advance to the next token after reading a token (instead of buffering tokens using a copy). - Moved logic of 'ReadToken' into Lex. GetToken & ReadToken no longer exist. - These changes result in a 3.3% speedup (-Eonly) on Cocoa.h. - The code is a little gross. Many cleanups are possible and should be done. llvm-svn: 61360
* Add some block-pointer conversions in C++Douglas Gregor2008-12-233-11/+50
| | | | llvm-svn: 61359
* Don't explicitly represent OverloadedFunctionDecls withinDouglas Gregor2008-12-2311-275/+267
| | | | | | | | | | | | | | | | | | | | | | | | DeclContext. Instead, just keep the list of currently-active declarations and only build the OverloadedFunctionDecl when we absolutely need it. This is a half-step toward eliminating the need to explicitly build OverloadedFunctionDecls that store sets of overloaded functions. This was suggested by Argiris a while back, and it's a good thing for several reasons: first, it eliminates the messy logic that currently tries to keep the OverloadedFunctionDecl in sync with the declarations that are being added. Second, it will (eventually) eliminate the need to allocate memory for overload sets, which could help performance. Finally, it helps set us up for when name lookup can return multiple (possibly ambiguous) results, as can happen with lookup of class members in C++. Next steps: make the IdentifierResolver store overloads as separate entries in its list rather than replacing them with an OverloadedFunctionDecl now, then see how far we can go toward eliminating OverloadedFunctionDecl entirely. llvm-svn: 61357
* More encoding support; in this case, encoding ofFariborz Jahanian2008-12-223-4/+16
| | | | | | outer-most const of pointer types. llvm-svn: 61355
* Fix <rdar://problem/6463613> clang ObjC rewriter: assertion failure ↵Steve Naroff2008-12-221-0/+2
| | | | | | rewriting @selector?. llvm-svn: 61351
* Guard against the return of PR3188Sebastian Redl2008-12-221-0/+3
| | | | llvm-svn: 61348
* Full AST support and better Sema support for C++ try-catch.Sebastian Redl2008-12-2210-7/+131
| | | | llvm-svn: 61346
* Support conversion from a null pointer constant o any Objective-C object ↵Douglas Gregor2008-12-222-0/+10
| | | | | | pointer type. Fixes rdar://problem/6463298 llvm-svn: 61340
* Fixed a bug showed up the meta-data for protocol Fariborz Jahanian2008-12-221-2/+10
| | | | | | | instance methods by building print-class-info.m, whose output is now identical to what gcc puts out. llvm-svn: 61339
OpenPOWER on IntegriCloud