Commit Graph

227 Commits

Author SHA1 Message Date
Chris Lattner fd65291aa3 Fix ASTContext::typesAreCompatible when analyzing a function type with
proto and function type without proto.  It would never call 
'functionTypesAreCompatible' because they have different type classes.

llvm-svn: 45952
2008-01-14 05:45:46 +00:00
Steve Naroff 090353191c Rewrite Expr::isNullPointerConstant() to deal with multiple levels of explicit casts.
Now, isNullPointerConstant() will return true for the following: "(void*)(double*)0"

llvm-svn: 45951
2008-01-14 02:53:34 +00:00
Steve Naroff 826e91ae04 Change Sema::CheckAddressOfOperation() to respect C99-only addressof rules.
Remove diagnostics from Sema::CheckIndirectionOperand(). C89/C99 allow dereferencing an incomplete type. clang appears to be emulating some incorrect gcc behavior (see below).

void
foo (void)
{
 struct b;
 struct b* x = 0;
 struct b* y = &*x; // gcc produces an error ("dereferencing pointer to incomplete type")
}

With this patch, the above is now allowed.

Bug/Patch by Eli Friedman!

llvm-svn: 45933
2008-01-13 17:10:08 +00:00
Chris Lattner 317e6ba07f Tighten up handling of __func__ and friends: it should be an array
of const char, and it should error if it occurs outside a function.

Is it valid in an objc method?  If so we should handle that too.

llvm-svn: 45910
2008-01-12 18:39:25 +00:00
Chris Lattner a81a0279cc Fix the type of predefined identifiers like __func__. Patch by
Eli Friedman!

llvm-svn: 45906
2008-01-12 08:14:25 +00:00
Fariborz Jahanian db701b47d2 Pass rewritten output to 'clang' for verification.
llvm-svn: 45794
2008-01-10 00:30:24 +00:00
Fariborz Jahanian 82ae0152a7 Allow messaging expression as foreach's collection expression.
llvm-svn: 45793
2008-01-10 00:24:29 +00:00
Steve Naroff e6b0ec8b5e Fix Sema::ActOnDeclarator() to call MergeFunctionDecl for function decls that aren't in scope. Since C functions are in a flat namespace, we need to give them special treatment (when compared with variables and typedefs).
llvm-svn: 45789
2008-01-09 23:34:55 +00:00
Steve Naroff c6edcbdb5d Fix ASTContext::typesAreCompatible to allow for int/enum compatibility (C99 6.7.2.2p4).
Fix Sema::MergeFunctionDecl to allow for function type compatibility (by using the predicate on ASTContext). Function types don't have to be identical to be compatible...

llvm-svn: 45784
2008-01-09 22:43:08 +00:00
Steve Naroff 08ddb8c55c Teach Sema::ActOnCompoundLiteral about constraint C99 6.5.2.5p3.
llvm-svn: 45782
2008-01-09 20:58:06 +00:00
Fariborz Jahanian 6fa7516bc9 Type-cast RHS of assignment to prevent warning compiling rewritten foreach code.
llvm-svn: 45777
2008-01-09 18:15:42 +00:00
Fariborz Jahanian 22f2347791 Fixed a bug whereby a parethesized collection expression was not being rewritten correctly.
llvm-svn: 45776
2008-01-09 17:50:00 +00:00
Fariborz Jahanian 05d2876ec3 Another test case for testing rewriteing of nested foreach-statement.
llvm-svn: 45769
2008-01-09 00:47:02 +00:00
Fariborz Jahanian a305a5609e Remove dependency on objc.h
llvm-svn: 45767
2008-01-09 00:33:05 +00:00
Steve Naroff 66a26044f8 Teach Expr::isConstantExpr() about CompoundLiterals.
llvm-svn: 45764
2008-01-09 00:05:37 +00:00
Fariborz Jahanian 965a8961c7 Patch to rewrite ObjC2's foreach-stmt.
llvm-svn: 45760
2008-01-08 22:06:28 +00:00
Steve Naroff 039ad3cf90 Fix Sema::CheckConditionalOperands(). The null pointer constant checks need to precede the check for two pointer operands.
llvm-svn: 45732
2008-01-08 01:11:38 +00:00
Chris Lattner 940cfebf90 add comments for the various AssignConvertType's, and split int->pointer from pointer->int.
llvm-svn: 45591
2008-01-04 18:22:42 +00:00
Chris Lattner 9bad62c72a Merge all the 'assignment' diagnostic code into one routine, decloning
it from several places.  This merges the diagnostics, making them more
uniform and fewer in number. This also simplifies and cleans up the code.

Some highlights:
1. This removes a bunch of very-similar diagnostics.
2. This renames AssignmentCheckResult -> AssignConvertType
3. This merges PointerFromInt + IntFromPointer which were always treated the same.
4. This updates a bunch of test cases that have minor changes to the produced diagnostics.

llvm-svn: 45589
2008-01-04 18:04:52 +00:00
Chris Lattner 816dea2fc5 generalize some of the conversion warnings.
llvm-svn: 45560
2008-01-03 23:38:43 +00:00
Chris Lattner 166ae81be1 update for changes in diagnostic strings.
llvm-svn: 45559
2008-01-03 23:36:08 +00:00
Chris Lattner 0a788433eb give better diagnostics for converting between function pointer and void*.
llvm-svn: 45556
2008-01-03 22:56:36 +00:00
Fariborz Jahanian 5cc21a7e02 Fixed a bug reported by Chris, involving assiging 0 to a qualified object type.
llvm-svn: 45542
2008-01-03 18:46:52 +00:00
Chris Lattner 4f20351142 When promoting array to pointer for argument, don't lose type qualifiers.
llvm-svn: 45510
2008-01-02 22:50:48 +00:00
Nate Begeman 16a3a9d5f9 Rename stats to print-stats to avoid conflicting with llvm's
stats statistic when clang is built as a dylib.

llvm-svn: 45441
2007-12-30 01:38:50 +00:00
Fariborz Jahanian 4f41b272c9 Another test case, testing a variety of objective-c type comparisons.
llvm-svn: 45302
2007-12-22 00:17:49 +00:00
Fariborz Jahanian 24e1a0bd35 Another test for objective-c's type comparison.
llvm-svn: 45301
2007-12-21 23:48:59 +00:00
Fariborz Jahanian a7c705ddca This patch implements some of the more obscure type-checking involving
'id' quallified with protocols and static types which have categories and
inheritance which implement these protocols.

llvm-svn: 45294
2007-12-21 22:22:33 +00:00
Fariborz Jahanian 56b5c96ef7 Patch to compare to objective-c static types where one or the other
(but not both) may be a protocol qualified static type.

llvm-svn: 45283
2007-12-21 17:34:43 +00:00
Fariborz Jahanian fd6f3a2460 Test case for my last patch.
llvm-svn: 45277
2007-12-21 00:35:35 +00:00
Fariborz Jahanian 76fcf32632 Another test for objective-c type comparison.
llvm-svn: 45271
2007-12-20 22:53:29 +00:00
Fariborz Jahanian 4368af07a7 More objective-c type checking. This time comparing objective-c known objects.
llvm-svn: 45269
2007-12-20 22:37:58 +00:00
Fariborz Jahanian 63b19f1b6e Patch to do type-checking for objctive-c's object types.
More is yet to come.

llvm-svn: 45263
2007-12-20 19:24:10 +00:00
Chris Lattner 2da14fb84f implement semantic analysis for __builtin_islessequal and friends.
llvm-svn: 45239
2007-12-20 00:26:33 +00:00
Fariborz Jahanian 6a128ae962 Minor test twik.
llvm-svn: 45208
2007-12-19 19:07:24 +00:00
Fariborz Jahanian a1e3420c49 This patch concludes implementation of dynamic objective-c type qualified by
protocol list.

llvm-svn: 45203
2007-12-19 17:45:58 +00:00
Chris Lattner ea72f449fe Implement C99 6.7.5.3p1
llvm-svn: 45188
2007-12-19 05:31:29 +00:00
Fariborz Jahanian 9f0e310861 Refactoring work. ObjcQualifiedIdType is now derived from 'Type'.
llvm-svn: 45174
2007-12-18 21:33:44 +00:00
Chris Lattner 20aad334c7 Fix the location we emit the "not a constant" error for this:
int foo() {
  typedef int x[foo()];
  static int y = sizeof(x);
}

previously we'd emit it on the typedef, which made not sense at all.

llvm-svn: 45154
2007-12-18 07:15:40 +00:00
Chris Lattner dcb7cc59ae Fix an nice and subtle parser bug reported by Nico Weber.
llvm-svn: 45149
2007-12-18 06:06:23 +00:00
Steve Naroff b74406ba98 Fixe bogus error for variable argument methods. Sema::ObjcGetTypeForMethodDefinition() wasn't preserving the isVariadic boolean. Another fix is to avoid synthsizing the function decl entirely, however this is a separate issue that I don't want to deal with now. Also added a FIXME to Sema::CheckFunctionCall(), which is currently emitting a bogus warning.
llvm-svn: 45146
2007-12-18 03:41:15 +00:00
Steve Naroff 1d2538cb4d Improve how we find private method decls. This involved:
- Changed Sema::ObjcActOnStartOfMethodDef() to register the methods with the global pools.
- Changed Sema::ActOnInstanceMessage() to look in global pools (should be much less error prone).
- Added a test case to message.m (for lookup that was broken).

Misc changes while I was investigating this...

- Changed Sema::ActOnAtEnd() to call AddFactoryMethodToGlobalPool (this looked like a cut/paste error).
- Added a comment and tweaked another where I was using the first person.

llvm-svn: 45142
2007-12-18 01:30:32 +00:00
Fariborz Jahanian 24cb52c9b0 Patch to implemented objective-c's dynamic object pointer qualified with
the protocol list (id<P,...> types).

llvm-svn: 45121
2007-12-17 21:03:50 +00:00
Ted Kremenek 3fbeaea7ee Modified format-string checking to not emit a warning when all of the
following hold:

(1) A vprintf-like function is called that takes the argument list via a
    via_list argument.

(2) The format string is a non-literal that is the parameter value of
    the enclosing function, e.g:
    
    void logmessage(const char *fmt,...) { 
      va_list ap;
      va_start(ap,fmt);
      fprintf(fmt,ap);  // Do not emit a warning.
    }
    
In the future this special case will be enhanced to consult the "format"
attribute attached to a function declaration instead of just allowing a blank
check for all function parameters to be used as format strings to vprintf-like
functions. This will happen when more support for attributes becomes
available.

llvm-svn: 45114
2007-12-17 19:03:13 +00:00
Steve Naroff eaaae467c7 Sema::ActOnMemberReferenceExpr() needs to perform the default conversions.
Bug and test case provided by Carl Lewis.

llvm-svn: 45078
2007-12-16 21:42:28 +00:00
Chris Lattner 3c1334013d Don't do integer promotions of LHS for compound shift assignment. The LHS has to be a modifiable lvalue.
llvm-svn: 44993
2007-12-13 07:28:16 +00:00
Fariborz Jahanian 56637d127f Fixed test to match the new diagnostic text.
llvm-svn: 44966
2007-12-13 00:16:34 +00:00
Fariborz Jahanian 4d1288a67e Concatenation of objc strings.
llvm-svn: 44964
2007-12-12 23:55:49 +00:00
Chris Lattner 6d513f37c9 Unbreak -stats on cocoa.h
llvm-svn: 44919
2007-12-12 06:43:05 +00:00
Chris Lattner b011825453 add run lines.
llvm-svn: 44918
2007-12-12 06:22:14 +00:00