Commit Graph

31154 Commits

Author SHA1 Message Date
Chandler Carruth 64d376a8fa Switch all local code and comments from '[Ii]nstantiat(ion|ed)' to
'[Ee]xpan(sion|ded)' in the TextDiagnosticPrinter.

No functionality changed.

llvm-svn: 135136
2011-07-14 08:20:31 +00:00
Chandler Carruth cd7a02ec1b Switch the diagnostic messages about macros to use the terms 'expanded'
and 'expansions' rather than 'instantiated' and 'contexts'.

This is the first of several patches migrating Clang's terminology
surrounding macros from 'instantiation' to 'expansion'.

llvm-svn: 135135
2011-07-14 08:20:28 +00:00
Ted Kremenek ba84cf53bc Reapply r135075, but modify format-strings.c and format-strings-fixit.c test cases to be more portable with an explicit target triple.
llvm-svn: 135134
2011-07-14 06:49:52 +00:00
NAKAMURA Takumi 91d831bc2d Revert r135075, "format string checking: long and int have the same widths on 32-bit, so we shouldn't warn about using"
It fails on freebsd, mingw and msvc10.

llvm-svn: 135129
2011-07-14 05:16:18 +00:00
Douglas Gregor 5cf0e1534a Add a hackaround to avoid the crash in PR10355. However, our recovery
is still terrible here because typo correction is not behaving well in
the presence of overloaded functions.

llvm-svn: 135128
2011-07-14 04:54:23 +00:00
Argyrios Kyrtzidis e2e40b4dc7 [arcmt] Make sure migrating to ARC works even if '-fobjc-arc' is included in command-line flags. rdar://9567824
llvm-svn: 135115
2011-07-14 00:17:54 +00:00
John McCall 7c3ed3d0c5 Document the behavior of property modifiers in ARC.
rdar://problem/9768338

llvm-svn: 135104
2011-07-13 23:15:32 +00:00
Richard Smith 48d2464c3f PR8800: When building a conversion to A& using a member operatorA&(), do not require A to be a complete type.
llvm-svn: 135101
2011-07-13 22:53:21 +00:00
Bruno Cardoso Lopes 98154a76fd Reapply r134946 with fixes. Tested on Benjamin testcase and other test-suite failures.
llvm-svn: 135091
2011-07-13 21:58:55 +00:00
Douglas Gregor ad69e65a39 Add 'mutable' to the function declarator chunk, to be used when
parsing lambda expressions, from John Freeman!

llvm-svn: 135090
2011-07-13 21:47:47 +00:00
Douglas Gregor 385cfc073e Fix an incorrect namespace typo-correction diagnostic, from Kaelyn
Uhrain! Fixes PR10318.

llvm-svn: 135086
2011-07-13 21:36:26 +00:00
Devang Patel f0335ce632 Emit debug info for extended vectors.
llvm-svn: 135083
2011-07-13 21:23:30 +00:00
Ted Kremenek 60654d697a format string checking: long and int have the same widths on 32-bit, so we shouldn't warn about using
an "int" format specifier with a "long" type in 32-bit.

llvm-svn: 135075
2011-07-13 20:20:58 +00:00
John McCall 928a25714e Enforce access control for the destructor in a new[] expression and mark
it as used.  Otherwise, we can fail to instantiate or validate the destructor,
which can lead to crashes in IR gen like PR10351.

llvm-svn: 135073
2011-07-13 20:12:57 +00:00
Argyrios Kyrtzidis de223c3a7b [arcmt] Add weak/unsafe_unretained for "@property (readonly)" when we are @synthesizing it.
llvm-svn: 135067
2011-07-13 19:47:57 +00:00
Argyrios Kyrtzidis c8b3619773 [arcmt] For properties rewrite 'assign' -> 'weak or unsafe_unretained', 'retain' -> 'strong', and add
'weak or unsafe_unretained' when 'assign' is missing. rdar://9496219&9602589.

llvm-svn: 135065
2011-07-13 19:22:00 +00:00
Fariborz Jahanian 9a881019a5 With -Wselector, don't warn about unimplemented optional method
used in @selector expression because, well, their implementation 
is optional. // rdar://9545564

llvm-svn: 135057
2011-07-13 19:05:43 +00:00
John McCall dffafded6c Don't crash if defining -dealloc in a category.
llvm-svn: 135054
2011-07-13 18:26:47 +00:00
John McCall 3f4138c5d0 In debugger mode, make ObjC message sends to unknown selectors return
__unknown_anytype, and rewrite such message sends correctly.

I had to bite the bullet and actually add a debugger support mode for this
one, which is a bit unfortunate, but there really isn't anything else
I could imagine doing;  this is clearly just debugger-specific behavior.

llvm-svn: 135051
2011-07-13 17:56:40 +00:00
Fariborz Jahanian 17585e7940 objc++: Some level of covariance is allowed in ObjC properties.
Make it also available in ObjC++ propeties. Use common code for
objc and objc++ so they don't diverge. // rdar://9740328

llvm-svn: 135050
2011-07-13 17:55:01 +00:00
Ted Kremenek cc47e0fad1 Re-relax conversion specifier checking for printf format strings and conversion specifiers. My recent change was a mistake.
llvm-svn: 135048
2011-07-13 17:35:14 +00:00
Ted Kremenek 1d3fb8af68 Fix inversion in argument type checking for format strings with conversion specifiers for character types.
llvm-svn: 135046
2011-07-13 17:25:47 +00:00
John McCall 97eab0a271 Okay, that rule about zero-length arrays applies to destroying
them, too.

llvm-svn: 135038
2011-07-13 08:09:46 +00:00
John McCall 6549b31fe9 Arrays are permitted to be zero-length in some situations.
llvm-svn: 135036
2011-07-13 07:37:11 +00:00
John McCall c368838b20 Make the integer-range analysis recognize ^= correctly,
and (while I'm at it) teach it to grok the results of simple
assignments.

The first is PR10336.

llvm-svn: 135034
2011-07-13 06:35:24 +00:00
John McCall f677a8e99e Convert the standard default-construction loops to use phis and
partial destruction.

llvm-svn: 135033
2011-07-13 06:10:41 +00:00
Chris Lattner 46ca3272e4 per john's advice, speculatively lower uses of forward-declared enums to
i32.  They almost always end up this way in the end anyway, and if we get
lucky, this avoids generating some bitcasts.

llvm-svn: 135032
2011-07-13 05:31:19 +00:00
Chris Lattner 3517f14219 PR10337 reminds me that calls return values, lets handle them just
like arguments.  Thanks PR10337! :)

llvm-svn: 135030
2011-07-13 03:59:32 +00:00
John McCall f47c069162 Aggressive dead code elimination.
llvm-svn: 135029
2011-07-13 03:03:51 +00:00
John McCall 98de3d74d2 Generalize the routine for destroying an object with static
storage duration, then explicitly exempt ownership-qualified
types from it.

llvm-svn: 135028
2011-07-13 03:01:35 +00:00
Douglas Gregor d29fde248e Eliminate an incomplete/incorrect attempt to provide support for C++0x
unrestricted unions, which ended up attempting to initialize objects
in a union (which CodeGen isn't prepared for). Fixes PR9683.

llvm-svn: 135027
2011-07-13 02:14:02 +00:00
Eli Friedman 609ada27ce Silliness with commas, as reported at http://blog.regehr.org/archives/558 . As it turns out, this is my fault for not noticing this was an issue when I was looking at this a long time ago. :(
llvm-svn: 135026
2011-07-13 02:05:57 +00:00
John McCall ca2c56f20b Switch delete[] IR-generation over to the destroy framework,
which implicitly makes it EH-safe as well.

llvm-svn: 135025
2011-07-13 01:41:37 +00:00
Douglas Gregor 1c2e20d73d When compiling ::delete for a class with a virtual destructor, call
the complete destructor and then invoke the global delete
operator. Previously, we would invoke the deleting destructor, which
calls the wrong delete operator. Fixes PR10341.

llvm-svn: 135021
2011-07-13 00:54:47 +00:00
Richard Smith 26935e60b6 Correctly set up the list of virtual base classes for a CXXRecordDecl. Previously we got the source range wrong for everything in the virtual bases list.
llvm-svn: 135011
2011-07-12 23:49:11 +00:00
Fariborz Jahanian c87c8799ec revert fix for // rdar://9740328
llvm-svn: 135010
2011-07-12 23:20:13 +00:00
Bruno Cardoso Lopes 0aadf83f80 Revert r134946
llvm-svn: 135004
2011-07-12 22:30:58 +00:00
Argyrios Kyrtzidis a200973ead [arcmt] Also avoid 'weak' for forward references to objc classes.
llvm-svn: 135003
2011-07-12 22:16:25 +00:00
Argyrios Kyrtzidis ce9b739999 [arcmt] Before applying '__weak' check whether the objc class is annotated with objc_arc_weak_reference_unavailable
or is in a list of classes not supporting 'weak'.

rdar://9489367.

llvm-svn: 135002
2011-07-12 22:05:17 +00:00
Fariborz Jahanian c0f6af2103 objc++: Some level of covariance is allowed in ObjC properties.
Make it also available in ObjC++ propeties. // rdar://9740328

llvm-svn: 135001
2011-07-12 22:05:16 +00:00
David Chisnall c2800aab7b Fix another typo (int -> id for captured_obj).
llvm-svn: 134998
2011-07-12 20:34:06 +00:00
John McCall 30317fda63 Generalize Cleanup::Emit's "isForEH" parameter into a set
of flags.  No functionality change.

llvm-svn: 134997
2011-07-12 20:27:29 +00:00
Peter Collingbourne 77b0e7f296 Implement -MG. Fixes PR9613
llvm-svn: 134996
2011-07-12 19:35:15 +00:00
Benjamin Kramer ae2d344384 Add more compiler workarounds. Should fix the build with old GCCs and MSVC.
llvm-svn: 134995
2011-07-12 18:37:23 +00:00
Douglas Gregor 7e6bfb4a0d In ARC mode, consider Objective-C lifetime types (object pointers and
block pointers) that don't have any qualification to be POD types. We
were previously considering them to be non-POD types, because this was
convenient in C++ for is_pod-like traits. However, we now end up
inferring lifetime in such cases (template arguments infer __strong),
so it is not necessary.

Moreover, we want rvalues of object type (which have their lifetime
stripped) to be PODs to allow, e.g., va_arg(arglist, id) to function
properly. Fixes <rdar://problem/9758798>.

llvm-svn: 134993
2011-07-12 17:28:52 +00:00
Fariborz Jahanian 7f4427fc60 Fix a bug where a local variable named 'self' is causing
implicit ivar accesses to go through the 'self' variable
rather than the real 'self' for the method. // rdar://9730771

llvm-svn: 134992
2011-07-12 17:16:56 +00:00
John McCall 5fa8e20878 Ugh, use this compiler workaround again.
llvm-svn: 134989
2011-07-12 16:53:04 +00:00
John McCall 4bd0fb1f09 Switch field destruction over to use the new destroyer-based API
and kill a lot of redundant code.

llvm-svn: 134988
2011-07-12 16:41:08 +00:00
Douglas Gregor 2207ec273a Improve name mangling for instantiation-dependent types that are not
dependent. This covers an odd class of types such as

  int (&)[sizeof(sizeof(T() + T()))];

which involve template parameters but, because of some trick typically
involving a form of expression that is never type-dependent, resolve
down to a non-dependent type. Such types need to be mangled
essentially as they were written in the source code (involving
template parameters), rather than via their canonical type.

In general, instantiation-dependent types should be mangled as
they were written in the source. However, since we can't do that now
without non-trivial refactoring of the AST (see the new FIXME), I've
gone for this partial solution: only use the as-written-in-the-source
mangling for these strange types that are instantiation-dependent but
not dependent. This provides better compatibility with previous
incarnations of Clang and with GCC. In the future, we'd like to get
this right.

Fixes <rdar://problem/9663282>.

llvm-svn: 134984
2011-07-12 15:18:55 +00:00
Benjamin Kramer a4fb836f06 Pop block scope after reading from it.
Found by valgrind.

llvm-svn: 134983
2011-07-12 14:11:05 +00:00