Commit Graph

29035 Commits

Author SHA1 Message Date
Devang Patel d3a6b0f184 Emit a stop point before a call expression so that debugger has some chance of getting some footing when user wants to stop at 2nd bar() in following expression when all function calls are inlined.
= bar() + ... + bar() + ...

clang keeps track of column numbers, so we could put location entries for all subexpressions but that will significantly bloat debug info in general, but a location for call expression is helpful here.

llvm-svn: 127018
2011-03-04 18:54:42 +00:00
Douglas Gregor 23648d7e3b When constructing source-location information for a
DependentTemplateSpecializationType during tree transformation, retain
the NestedNameSpecifierLoc as it was used to translate the template
name, rather than reconstructing it from the template name.

Fixes PR9401.

llvm-svn: 127015
2011-03-04 18:53:13 +00:00
Douglas Gregor 3c41bf740f *Recursively* set the context of a template parameter, so that we also
capture the template parameters of template template parameters.

llvm-svn: 127012
2011-03-04 18:32:38 +00:00
Douglas Gregor fd7c225530 Make sure to put template parameters into their owning template's
DeclContext once we've created it. This mirrors what we do for
function parameters, where the parameters start out with
translation-unit context and then are adopted by the appropriate
DeclContext when it is created. Also give template parameters public
access and make sure that they don't show up for the purposes of name
lookup.

Fixes PR9400, a regression introduced by r126920, which implemented
substitution of default template arguments provided in template
template parameters (C++ core issue 150).

How on earth could the DeclContext of a template parameter affect the
handling of default template arguments?

I'm so glad you asked! The link is
Sema::getTemplateInstantiationArgs(), which determines the outer
template argument lists that correspond to a given declaration. When
we're instantiating a default template argument for a template
template parameter within the body of a template definition (not it's
instantiation, per core issue 150), we weren't getting any outer
template arguments because the context of the template template
parameter was the translation unit. Now that the context of the
template template parameter is its owning template, we get the
template arguments from the injected-class-name of the owning
template, so substitution works as it should.

llvm-svn: 127004
2011-03-04 17:52:15 +00:00
Abramo Bagnara b9893d66bc Fixed source range for function template specializations.
template <class T> void foo();
template <> void foo<int>(); /* Spec 1 */
template <> void foo<int>(); /* Spec 2 */

If we look at the main location of the first explicit specialization (Spec 1) it can be seen that it points to the name of the *second* explicit specialization (Spec 2), which is a redeclaration of Spec1.
Hence, the source range obtained for Spec1 is not only inaccurate, but also invalid (the end location comes before the start location).

llvm-svn: 127002
2011-03-04 17:20:30 +00:00
Abramo Bagnara a0935267dc Fixed source range for ClassTemplateSpecializationDecl.
llvm-svn: 126999
2011-03-04 14:20:30 +00:00
Benjamin Kramer 024e619f2a Move private structs into anonymous namespaces.
llvm-svn: 126997
2011-03-04 13:12:48 +00:00
Abramo Bagnara 23485e04be Improved TemplateTypeParmDecl end location.
llvm-svn: 126996
2011-03-04 12:42:03 +00:00
Abramo Bagnara e15d553663 Fixed end location of NonTypeTemplateParamDecl.
llvm-svn: 126994
2011-03-04 11:03:48 +00:00
Abramo Bagnara 36a95be4ea Fixed end location of FriendDecl.
llvm-svn: 126993
2011-03-04 10:59:18 +00:00
John McCall 8bc6d5b919 Don't consider visibility from template parameter lists if we're
computing for a nested decl with explicit visibility.  This is all part
of the general philosophy of explicit visibility attributes, where
any information that was obviously available at the attribute site
should probably be ignored.  Fixes PR9371.

llvm-svn: 126992
2011-03-04 10:39:25 +00:00
John McCall 74186ab391 Access ivars with inbounds GEPs, even in -fwrapv. It's unlikely that
this can have any optimization effect, given the opacity of objects pointers,
but you never know.

llvm-svn: 126990
2011-03-04 08:25:59 +00:00
John McCall 6647584951 Let's do super message sends with static allocas instead of dynamic ones.
llvm-svn: 126989
2011-03-04 08:00:29 +00:00
John McCall 4223a9ee7a Make AttributedTypes for GC-qualified types and fix some miscellaneous
bugs with such types.  Not sure this is quite how I want the desugaring
and a.k.a. logic to go, but it suffices.

llvm-svn: 126986
2011-03-04 04:00:19 +00:00
Ted Kremenek be52871b1a Correctly handle nested switch statements in CFGBuilder when on switch statement has a condition that evaluates to a constant.
llvm-svn: 126977
2011-03-04 01:03:41 +00:00
Argyrios Kyrtzidis 7a1778eba1 -Woverloaded-virtual shouldn't warn for static functions. Fixes rdar://9083431 & http://llvm.org/PR9396.
llvm-svn: 126966
2011-03-03 22:58:57 +00:00
Douglas Gregor 2b1ca9eaad Fix PR9390 in not one, but two ways:
1) When we do an instantiation of the injected-class-name type,
  provide a proper source location. This is just plain good hygiene.

  2) When we're building a NestedNameSpecifierLoc from a CXXScopeSpec,
  only return an empty NestedNameSpecifierLoc if there's no
  representation.

Both problems contributed to the horrible test case in PR9390 that I
couldn't reduce down to something palatable.

llvm-svn: 126961
2011-03-03 21:48:55 +00:00
Devang Patel 68a1525290 Encode argument numbering in debug info so that code generator can emit them in order.
This fixes few blocks.exp regressions.

llvm-svn: 126960
2011-03-03 20:13:15 +00:00
Abramo Bagnara 124fdf6dd4 Fixed source range for LabelDecl.
llvm-svn: 126952
2011-03-03 18:24:14 +00:00
Argyrios Kyrtzidis e02831423a Don't emit unused warning for deleted functions. Fixes rdar://8365684 & http://llvm.org/PR9391.
llvm-svn: 126950
2011-03-03 17:47:42 +00:00
Douglas Gregor bd6b4447a7 Finish updated test
llvm-svn: 126948
2011-03-03 17:22:06 +00:00
Ted Kremenek 916c8357db Add coverage test for CFGImplicitDtor::getDestructorDecl() when handling typedefs.
llvm-svn: 126947
2011-03-03 17:07:11 +00:00
Douglas Gregor e29139c77d Eliminate redundant nested-name-specifiers on
TemplateSpecializationTypes, which also fixes PR9388.

llvm-svn: 126946
2011-03-03 17:04:51 +00:00
Abramo Bagnara 66a35d765f Removed left brace location from LinkageSpecDecl.
llvm-svn: 126945
2011-03-03 16:52:29 +00:00
Abramo Bagnara 4a8cda8556 Fixed end source location for LinkageSpecDecl.
llvm-svn: 126943
2011-03-03 14:52:38 +00:00
Abramo Bagnara 348823aa36 Fixed source range for FileScopeAsmDecl. Others source range fixes will follow.
llvm-svn: 126939
2011-03-03 14:20:18 +00:00
Douglas Gregor 739b107af8 When we use the default template arguments of a template template
parameter, save the instantiated default template arguments along with
the explicitly-specified template argument list. That way, we prefer
the default template template arguments corresponding to the template
template parameter rather than those of its template template argument.

This addresses the likely direction of C++ core issue 150, and fixes
PR9353/<rdar://problem/9069136>, bringing us closer to the behavior of
EDG and GCC.

llvm-svn: 126920
2011-03-03 02:41:12 +00:00
Douglas Gregor ce449ab16e Add a missing break, from John Wiegley
llvm-svn: 126919
2011-03-03 02:20:19 +00:00
Ted Kremenek 8cfe207861 Teach CFGImplicitDtor::getDestructorDecl() about arrays of objects with destructors.
llvm-svn: 126910
2011-03-03 01:21:32 +00:00
Ted Kremenek 1676a042e3 Teach CFGImplicitDtor::getDestructorDecl() about reference types.
llvm-svn: 126909
2011-03-03 01:01:03 +00:00
Matt Beaumont-Gay 86b900baca Let's go with John and Ted's preferred fix.
llvm-svn: 126907
2011-03-03 00:48:05 +00:00
Matt Beaumont-Gay 0532897f37 Keep GCC from complaining about falling off the end of the function.
llvm-svn: 126897
2011-03-02 23:25:06 +00:00
Dan Gohman 6ca9982757 Fix a typo.
llvm-svn: 126890
2011-03-02 22:39:34 +00:00
Douglas Gregor 978d07a5d7 Remove cruft
llvm-svn: 126888
2011-03-02 21:57:24 +00:00
Douglas Gregor 6471db0749 Force CaseStmt to store its child statements in source-code order,
which is important for libclang's token-annotation and
where's-my-cursor functionality. Fixes <rdar://problem/9004439>.

llvm-svn: 126887
2011-03-02 21:45:00 +00:00
Tilmann Scheller 99cc30c371 Revert "Add CC_Win64ThisCall and set it in the necessary places."
This reverts commit 126863.

llvm-svn: 126886
2011-03-02 21:36:49 +00:00
Rafael Espindola c4eb31e5e5 Apply sysroot to system c++ headers too. Fixes PR9372.
llvm-svn: 126884
2011-03-02 21:30:07 +00:00
Benjamin Kramer 0f074df979 Only emit string initializers in-place if types match. Fixes PR9373.
llvm-svn: 126883
2011-03-02 21:27:44 +00:00
Douglas Gregor 25b7e05b72 Fix the source range for a member access expression that includes a
nested-name-specifier and improve the detection of implicit 'this'
bases. Fixes <rdar://problem/8750392>.

llvm-svn: 126880
2011-03-02 21:06:53 +00:00
John McCall fb55f851a5 Work around a misdesigned GCC warning.
llvm-svn: 126879
2011-03-02 21:01:41 +00:00
Tilmann Scheller e60baf731b Revert "Add preliminary support for MSVC-style vtables."
This reverts commit 126865.

llvm-svn: 126876
2011-03-02 20:40:50 +00:00
Ted Kremenek e06a55c6b9 Introduce CFGImplicitDtor::isNoReturn() to query whether a destructor actually returns. Use this for -Wreturn-type to prune false positives reported in PR 6884.
llvm-svn: 126875
2011-03-02 20:32:29 +00:00
Devang Patel bd6f7f9770 revert r126858.
llvm-svn: 126874
2011-03-02 20:31:22 +00:00
Fariborz Jahanian ff9890302c IRGen. Fix IR when message returns reference type.
// rdar://8604515.

llvm-svn: 126869
2011-03-02 20:09:49 +00:00
Tilmann Scheller a6c995dbb2 Add preliminary support for MSVC-style vtables.
llvm-svn: 126865
2011-03-02 19:38:28 +00:00
Tilmann Scheller 454464b491 Add CC_Win64ThisCall and set it in the necessary places.
llvm-svn: 126863
2011-03-02 19:36:23 +00:00
Douglas Gregor 64f38ae7ee Teach libclang how to visit the children of a C++ base-class specifier
(i.e., the TypeLoc describing the base class type).

llvm-svn: 126861
2011-03-02 19:17:03 +00:00
Devang Patel 31e5fb52d1 Encode argument numbering in debug info so that code generator can emit them in order.
This fixes few blocks.exp regressions.

Reapply r126795 with a fix (one character change) for gdb testsuite regressions.

llvm-svn: 126858
2011-03-02 19:11:22 +00:00
Douglas Gregor cf256c1297 Eliminate an unnecessary TemporaryBase object from TreeTransform
llvm-svn: 126857
2011-03-02 19:02:55 +00:00
Douglas Gregor 1ee2406369 Kill off the
TreeTransform::TransformDependentTemplateSpecializationType() with
poor source-location information handling. All of the
CXXScopeSpec::MakeTrivial() and
NestedNameSpecifierLocBuilder::MakeTrivial() callers actually make
sense now.

llvm-svn: 126856
2011-03-02 18:57:38 +00:00
Douglas Gregor 1abdf36a2c Eliminate TreeTransform::RebuildNestedNameSpecifier(), all four of
them, which are no longer used.

llvm-svn: 126855
2011-03-02 18:54:26 +00:00
Douglas Gregor d4c41bcd7f Eliminate an unnecessary use of CXXScopeSpec::MakeTrivial. We have proper nested-name-specifier source-location information in DependentTemplateSpecializationTypeLocs now
llvm-svn: 126854
2011-03-02 18:52:42 +00:00
Douglas Gregor fd35cde35e Kill off TreeTransform::TransformNestedNameSpecifier() in favor of the
source-location-preserving
TreeTransform::TranformNestedNameSpecifierLoc(). No functionality
change: the victim had no callers (that themselves had callers) anyway.

llvm-svn: 126853
2011-03-02 18:50:38 +00:00
Douglas Gregor df846d11c1 Kill off the TreeTransform::TransformTemplateName overload that has
poor source-location information.

llvm-svn: 126852
2011-03-02 18:46:51 +00:00
Douglas Gregor 579c15f5d6 Kill off one of the TreeTransform::TransformTypeInObjectScope()
overloads (the one with the poor source-location information).

llvm-svn: 126851
2011-03-02 18:32:08 +00:00
Douglas Gregor 5f9591d3e5 Fix a thinko found by Craig Silverstein
llvm-svn: 126850
2011-03-02 18:10:05 +00:00
Douglas Gregor 9db535035d Start migrating TreeTransform's TransformTemplateName over to version
that preserve source-location information. This commit adds more
redundancy than it removes; WIP.

llvm-svn: 126849
2011-03-02 18:07:45 +00:00
Devang Patel a54696de8a Revert r126794.
llvm-svn: 126848
2011-03-02 17:54:58 +00:00
Douglas Gregor 9d80212115 Push nested-name-specifier source location information into template
template arguments. I believe that this is the last place in the AST
where we were storing a source range for a nested-name-specifier
rather than a proper nested-name-specifier location structure. (Yay!)

There is still a lot of cleanup to do in the TreeTransform, which
doesn't take advantage of nested-name-specifiers with source-location
information everywhere it could.

llvm-svn: 126844
2011-03-02 17:09:35 +00:00
Douglas Gregor 1f2f796690 Eliminate some unused getQualifierRange() member functions. We deal in nested-name-specifier locations now.
llvm-svn: 126842
2011-03-02 15:13:50 +00:00
John McCall 5fca7eaf85 Don't hard-code these constants, or at least don't hard-code them so hard.
llvm-svn: 126841
2011-03-02 12:29:23 +00:00
John McCall 80ee5963fd Pretty up the wrong-number-of-arguments-for-attribute diagnostic by
using a custom plural form.  Split out the range diagnostics as their
own message.

llvm-svn: 126840
2011-03-02 12:15:05 +00:00
John McCall 86bc21ffcf Provide an attribute, objc_method_family, to allow the inferred family
of an Objective-C method to be overridden on a case-by-case basis.  This
is a higher-level tool than ns_returns_retained &c.;  it lets users specify
that not only does a method have different retain/release semantics, but
that it semantically acts differently than one might assume from its name.
This in turn is quite useful to static analysis.

llvm-svn: 126839
2011-03-02 11:33:24 +00:00
John McCall 79adb25b86 Split out a subgroup of -Wconversion called -Wsign-conversion. Note that
unlike GCC, this is simply a subset of -Wconversion in all languages;  it
is not disabled by defualt in C++.

llvm-svn: 126836
2011-03-02 07:36:25 +00:00
John McCall 81a325e038 Hack in something so that we emit better debug information about
captured __block variables in the block-literal type.

llvm-svn: 126834
2011-03-02 06:57:14 +00:00
John McCall f79e87d727 Support a new InheritableAttr subclass, InheritableParamAttr, which is
used for attributes that are okay to inherit when written on a parameter.
Dependent on LLVM r126827.

llvm-svn: 126828
2011-03-02 04:00:57 +00:00
John McCall 8feeb49662 Semantic checking for exception specifications should be triggered by
whether C++ exceptions are enabled, not exceptions in general.  PR9358.

llvm-svn: 126820
2011-03-02 02:04:40 +00:00
Douglas Gregor c52264e719 When we're substituting into a parameter-type-list nested inside the pattern
of an expansion, and we have a paramameter that is not a parameter
pack, don't suppress substitution of parameter packs within this
context.

llvm-svn: 126819
2011-03-02 02:04:06 +00:00
John McCall b4526252db Move some of the logic about classifying Objective-C methods into
conventional categories into Basic and AST.  Update the self-init checker
to use this logic;  CFRefCountChecker is complicated enough that I didn't
want to touch it.

llvm-svn: 126817
2011-03-02 01:50:55 +00:00
Eli Friedman 846ded2e53 PR9350: increment/decrement of char (and anything else narrower than int)
can't overflow due to promotion rules; emit a wrapping add for those cases.

llvm-svn: 126816
2011-03-02 01:49:12 +00:00
Daniel Dunbar 908b48565d build: Allow disabling movt/movw from build, for testing purposes.
llvm-svn: 126810
2011-03-02 00:55:57 +00:00
Douglas Gregor e7c2065379 Push nested-name-specifier source-location information into dependent
template specialization types. This also required some parser tweaks,
since we were losing track of the nested-name-specifier's source
location information in several places in the parser. Other notable
changes this required:

  - Sema::ActOnTagTemplateIdType now type-checks and forms the
    appropriate type nodes (+ source-location information) for an
    elaborated-type-specifier ending in a template-id. Previously, we
    used a combination of ActOnTemplateIdType and
    ActOnTagTemplateIdType that resulted in an ElaboratedType wrapped
    around a DependentTemplateSpecializationType, which duplicated the
    keyword ("class", "struct", etc.) and nested-name-specifier
    storage.

  - Sema::ActOnTemplateIdType now gets a nested-name-specifier, which
    it places into the returned type-source location information.

  - Sema::ActOnDependentTag now creates types with source-location
    information.

llvm-svn: 126808
2011-03-02 00:47:37 +00:00
Ted Kremenek eff9a7ff91 Teach CFGBuilder to prune trivially unreachable case statements.
llvm-svn: 126797
2011-03-01 23:12:55 +00:00
Devang Patel 3bc2dedb40 Encode argument numbering in debug info so that code generator can emit them in order.
This fixes few blocks.exp regressions.

llvm-svn: 126795
2011-03-01 22:59:40 +00:00
Benjamin Kramer 51477bd0d0 Since getDriver().getInstalledDir() returns a const char *, don't try to
compare it with getDriver().Dir.c_str(), since that is a pointer
comparison, not a "are these strings equal" comparison.

Instead, just compare with getDriver().Dir directly, so both sides will
get promoted to std::string, and the regular std::string comparison
operator applies.

Patch by Dimitry Andric!

llvm-svn: 126791
2011-03-01 22:50:47 +00:00
Ted Kremenek e96dad9544 Don't warn about unused values in ternary ?: expressions unless both the LHS and RHS are "unused" (side-effect free).
Patch by Justin Bogner!  Fixes PR 8282.

llvm-svn: 126779
2011-03-01 20:34:48 +00:00
Douglas Gregor a7a795bed1 Push nested-name-specifier source-location information into dependent
template specialization types. There are still a few rough edges to
clean up with some of the parser actions dropping
nested-name-specifiers too early.

llvm-svn: 126776
2011-03-01 20:11:18 +00:00
Douglas Gregor 1bbd2d2ece Kill off more names to fix this test
llvm-svn: 126775
2011-03-01 19:55:40 +00:00
Dan Gohman 72346e97b3 Revert 123553, as sys::fs::unique_file is not finished yet.
llvm-svn: 126772
2011-03-01 19:50:49 +00:00
Ted Kremenek 26bbc3d494 Don't warning about shifting by too many bits in dead code.
llvm-svn: 126770
2011-03-01 19:13:22 +00:00
Roman Divacky 5e1af659e9 Fix the test.
llvm-svn: 126768
2011-03-01 18:50:59 +00:00
Daniel Dunbar 733b0f87df Driver/Darwin: Be bug compatible with GCC and ignore -pg on ARM.
llvm-svn: 126767
2011-03-01 18:49:30 +00:00
Ted Kremenek df26df726e For C++, enhance -Warray-bounds to recursively analyze array subscript accesses in ?: expressions.
llvm-svn: 126766
2011-03-01 18:41:00 +00:00
Douglas Gregor 844cb50266 Reinstate the introduction of source-location information for
nested-name-speciciers within elaborated type names, e.g.,
 
  enum clang::NestedNameSpecifier::SpecifierKind

Fixes in this iteration include:

  (1) Compute the type-source range properly for a dependent template
  specialization type that starts with "template template-id ::", as
  in a member access expression

    dep->template f<T>::f()

  This is a latent bug I triggered with this change (because now we're
  checking the computed source ranges for dependent template
  specialization types). But the real problem was...

  (2) Make sure to set the qualifier range on a dependent template
  specialization type appropriately. This will go away once we push
  nested-name-specifier locations into dependent template
  specialization types, but it was the source of the
  valgrind errors on the buildbots.
  

llvm-svn: 126765
2011-03-01 18:12:44 +00:00
Roman Divacky 432f10df68 The default CPU on FreeBSD for i386 should be i486, not pentium4.
Patch by Dimitry Andric!

llvm-svn: 126763
2011-03-01 18:11:37 +00:00
Ted Kremenek 63657fe995 Don't wanr about "negative shifts" in code that is unreachable. Fixes PR 5544.
llvm-svn: 126762
2011-03-01 18:09:31 +00:00
Roman Divacky 315c1675e8 On FreeBSD, make sure /usr/local/include is *not* in the default include
path.  This avoids accidentally including the wrong headers.

Patch by Dimitry Andric!

llvm-svn: 126761
2011-03-01 18:08:03 +00:00
Roman Divacky 2e5065d889 On FreeBSD, we don't want 'getDriver().Dir + "/../lib"' added to the
ToolChain's FilePaths.  If clang is installed as a port in /usr/local,
it is *not* supposed to use /usr/local/lib by default, for example.

Additionally, there are no clang-related executables in either
/usr/libexec, or getDriver().Dir + "/../libexec", anymore, so remove
that from the ToolChain's ProgramPaths.

Patch by Dimitry Andric!

llvm-svn: 126760
2011-03-01 18:03:28 +00:00
Roman Divacky ee8188a23b For linking on FreeBSD, don't add a hardcoded "-L/usr/lib", but
retrieve the library paths from the ToolChain object instead.

Copy the relevant code from linuxtools::Link::ConstructJob(), and
replace the std::string stuff with llvm::StringRef, while we're here.

Patch by Dimitry Andric!

llvm-svn: 126757
2011-03-01 17:53:14 +00:00
Roman Divacky 65b88cdb3b Implement -mrtd which sets the StdCall calling convention to be the default
one.

llvm-svn: 126756
2011-03-01 17:40:53 +00:00
Roman Divacky dc1f68d0a5 Add missing options.
llvm-svn: 126755
2011-03-01 17:36:40 +00:00
Fariborz Jahanian cf7f66f16f objc IRGen for Next runtime message API.
The prototype for objc_msgSend() is technically variadic - 
`id objc_msgSend(id, SEL, ...)`. 
But all method calls should use a prototype that matches the method, 
not the prototype for objc_msgSend itself().
// rdar://9048030

llvm-svn: 126754
2011-03-01 17:28:13 +00:00
Douglas Gregor b3a58b08e0 Revert r126748, my second attempt at nested-name-specifier source
location information for elaborated types. *sigh*

llvm-svn: 126753
2011-03-01 17:25:47 +00:00
Douglas Gregor 48e6bbffa1 Implement comparison of C++0x scoped enumeration types. Fixes PR9333.
llvm-svn: 126752
2011-03-01 17:16:20 +00:00
Douglas Gregor 9760a666d6 When digging into a cv-qualified return type that is a pointer type to
diagnose ignored qualifiers on return types, only assume that there is
a pointer chunk if the type is *structurally* a pointer type, not if
it's a typedef of a pointer type. Fixes PR9328/<rdar://problem/9055428>.

llvm-svn: 126751
2011-03-01 17:04:42 +00:00
Douglas Gregor 84a6a0a3ce When building a type for a typename specifier, check specifically for
a dependent template name rather than (indirectly and incorrectly)
trying to determine whether we can compute a context for the
nested-name-specifier. Fixes a GCC testsuite regression,
<rdar://problem/9068589>.

llvm-svn: 126749
2011-03-01 16:44:30 +00:00
Douglas Gregor bf5fe47b12 Reinstate r126737, extending the generation of type-source location
information for qualifier type names throughout the parser to address
several problems.

The commit message from r126737:

Push nested-name-specifier source location information into elaborated
name types, e.g., "enum clang::NestedNameSpecifier::SpecifierKind".

Aside from the normal changes, this also required some tweaks to the
parser. Essentially, when we're looking at a type name (via
getTypeName()) specifically for the purpose of creating an annotation
token, we pass down the flag that asks for full type-source location
information to be stored within the returned type. That way, we retain
source-location information involving nested-name-specifiers rather
than trying to reconstruct that information later, long after it's
been lost in the parser.

With this change, test/Index/recursive-cxx-member-calls.cpp is showing
much improved results again, since that code has lots of
nested-name-specifiers.

llvm-svn: 126748
2011-03-01 16:31:39 +00:00
Douglas Gregor 62a60c50f4 Revert r126737, the most recent nested-name-specifier location change, for buildbot breakage.
llvm-svn: 126746
2011-03-01 15:34:37 +00:00
Rafael Espindola d95a81293b Disable more warnings so that it is safe to use
CC="clang -use-gold-plugin -emit-llvm"
CXX="clang++ -use-gold-plugin -emit-llvm"

llvm-svn: 126740
2011-03-01 05:25:27 +00:00
Chandler Carruth 477a999788 Move the bool-conversions behind the DiagRuntimeBehavior logic. It's
possible for these to show up due to metaprogramming both in unevaluated
contexts and compile-time dead branches.

Those aren't the bugs we're looking for.

llvm-svn: 126739
2011-03-01 03:29:37 +00:00
Ted Kremenek 96a7a59119 In preparation for fixing PR 6884, rework CFGElement to have getAs<> return pointers instead of fresh CFGElements.
- Also, consoldiate getDtorKind() and getKind() into one "kind".
- Add empty getDestructorDecl() method to CFGImplicitDtor.

llvm-svn: 126738
2011-03-01 03:15:10 +00:00
Douglas Gregor 9720642c68 Push nested-name-specifier source location information into elaborated
name types, e.g., "enum clang::NestedNameSpecifier::SpecifierKind".

Aside from the normal changes, this also required some tweaks to the
parser. Essentially, when we're looking at a type name (via
getTypeName()) specifically for the purpose of creating an annotation
token, we pass down the flag that asks for full type-source location
information to be stored within the returned type. That way, we retain
source-location information involving nested-name-specifiers rather
than trying to reconstruct that information later, long after it's
been lost in the parser.

With this change, test/Index/recursive-cxx-member-calls.cpp is showing
much improved results again, since that code has lots of
nested-name-specifiers.

llvm-svn: 126737
2011-03-01 03:11:17 +00:00
Argyrios Kyrtzidis a9a0f5d206 [analyzer] Also make sure that the parameter is coming from the current stack frame.
llvm-svn: 126735
2011-03-01 01:59:41 +00:00
Argyrios Kyrtzidis 456b18c279 [analyzer] Remove SVal::getAsVarDecl() and reason about MemRegions, not Decls. Suggestion by Ted!
llvm-svn: 126734
2011-03-01 01:47:48 +00:00
Ken Dyck dbe37f38d0 Use CharUnits for TypeAlign variable in LayoutWideBitField(). No change in
functionality intended.

llvm-svn: 126730
2011-03-01 01:36:00 +00:00
Douglas Gregor 3d0da5f5dd Push nested-name-specifier source location information into
DependentNameTypeLoc. Teach the recursive AST visitor and libclang how to
walk DependentNameTypeLoc nodes.

Also, teach libclang about TypedefDecl source ranges, so that we get
those. The massive churn in test/Index/recursive-cxx-member-calls.cpp
is a good thing: we're annotating a lot more of this test correctly
now.

llvm-svn: 126729
2011-03-01 01:34:45 +00:00
Argyrios Kyrtzidis 918c847890 [analyzer] Move lib/StaticAnalyzer/Checkers/ExprEngine.cpp -> lib/StaticAnalyzer/Core
and hope the wrath of the buildbots will not descend upon me.

llvm-svn: 126728
2011-03-01 01:24:23 +00:00
Ken Dyck aab635e67a Change the return type of GetVirtualPointersSize() to CharUnits. No change
in functionality intended.

llvm-svn: 126727
2011-03-01 01:22:45 +00:00
Argyrios Kyrtzidis 6a5674ffa6 [analyzer] Rename CheckerV2 -> Checker.
llvm-svn: 126726
2011-03-01 01:16:21 +00:00
Argyrios Kyrtzidis 0036b19abd [analyzer] Remove Checker V1.
llvm-svn: 126725
2011-03-01 01:16:08 +00:00
Argyrios Kyrtzidis d665807901 [analyzer] Remove checker V1 registration and running from ExprEngine.
llvm-svn: 126724
2011-03-01 01:16:03 +00:00
Zhanyong Wan 5ad574c096 Improves the coding style in SValBuilder. This patch:
- renames evalCastNL and evalCastL to evalCastFromNonLoc and
  evalCastFromLoc (avoid abbreviations that aren't well known).

- makes all function parameter names start with a lower case letter
  for consistency and distinction from member variables.

- avoids abbreviations in function parameter names.

Reviewed by kremenek@apple.com.

llvm-svn: 126722
2011-03-01 00:45:32 +00:00
Douglas Gregor ff42ec5895 Fix some regressions in the GCC testsuite due to bad souce-location
information. Fixes <rdar://problem/9063643>.

llvm-svn: 126721
2011-03-01 00:34:57 +00:00
Chris Lattner 2e72da947c -fwrapv should turn off the inbounds markers from geps used for pointer
arithmetic.  This is part of PR9256, it would be great if someone else
wired up -fno-strict-overflow in the driver to -fwrapv.

llvm-svn: 126718
2011-03-01 00:03:48 +00:00
Douglas Gregor 9b27251e58 Refactor the construction of nested-name-specifiers with
source-location information into a NestedNameSpecifierLocBuilder
class, which lives within the AST library and centralize all knowledge
of the format of nested-name-specifier location information here.

No functionality change.

llvm-svn: 126716
2011-02-28 23:58:31 +00:00
Rafael Espindola 22f60303f2 Don't warn about "clang -use-gold-plugin -c ...". With this users can say
CC="clang -use-gold-plugin" CFLAGS=-emit-llvm ../configure...
and not be hit with a warning for each .c file.

llvm-svn: 126713
2011-02-28 23:29:45 +00:00
Douglas Gregor 9cbc22ba2d Teach Sema::CheckTypenameType to use nested-name-specifiers with
source-location information. We don't actually preserve this
information in any of the resulting TypeLocs (yet), so it doesn't
matter.

llvm-svn: 126693
2011-02-28 22:42:13 +00:00
Devang Patel bbe4ff99e0 Use single code path to handle c++ structs and classes while generating debug info.
llvm-svn: 126691
2011-02-28 22:32:45 +00:00
Argyrios Kyrtzidis 29b8656935 [analyzer] Remove unused checker stuff from AnalysisConsumer and some unused headers.
llvm-svn: 126690
2011-02-28 22:30:38 +00:00
Douglas Gregor ea972d3faa Push nested-name-specifier location information into DeclRefExpr and
MemberExpr, the last of the expressions with qualifiers!

llvm-svn: 126688
2011-02-28 21:54:11 +00:00
Fariborz Jahanian 557c1ed522 Revert r126678.
llvm-svn: 126685
2011-02-28 21:19:34 +00:00
Douglas Gregor 0da1d43e16 Push nested-name-specifier source location information into
UnresolvedLookupExpr and UnresolvedMemberExpr.

Also, improve the computation that checks whether the base of a member
expression (either unresolved or dependent-scoped) is implicit. The
previous check didn't cover all of the cases we use in our
representation, which threw off source-location information for these
expressions (which, in turn, caused some breakage in libclang's token
annotation). 

llvm-svn: 126681
2011-02-28 20:01:57 +00:00
Fariborz Jahanian 79246322a6 objc IRGen for Next runtime message API.
The prototype for objc_msgSend() is technically variadic - 
`id objc_msgSend(id, SEL, ...)`. 
But all method calls should use a prototype that matches the method, 
not the prototype for objc_msgSend itself().
// rdar://9048030

llvm-svn: 126678
2011-02-28 19:55:59 +00:00
Argyrios Kyrtzidis 9eb02dfa89 [analyzer] Remove '-analyzer-check-objc-mem' flag, the nominee for best misnomer award.
llvm-svn: 126676
2011-02-28 19:49:42 +00:00
Argyrios Kyrtzidis af181bb19b Move test/SemaObjC/method-arg-decay.m -> test/Analysis/method-arg-decay.m
llvm-svn: 126675
2011-02-28 19:49:21 +00:00
Argyrios Kyrtzidis 6a1c760760 [analyzer] Run the ExprEngine depending on the CheckerManager having path-sensitive checkers.
llvm-svn: 126674
2011-02-28 19:49:17 +00:00
Argyrios Kyrtzidis 20f5caa518 [analyzer] The current UninitializedValuesChecker will go away, remove '-warn-uninit-values'.
llvm-svn: 126673
2011-02-28 19:49:12 +00:00
Peter Collingbourne 938f7d0e46 Kill 18 month old #if 0'd code in StmtXML.cpp
llvm-svn: 126670
2011-02-28 19:42:30 +00:00
Douglas Gregor dd733fb041 Remove redundant friend declaration
llvm-svn: 126665
2011-02-28 19:17:05 +00:00
Douglas Gregor e16af53619 Push nested-name-specifier source location information into
CXXDependentScopeMemberExpr, and clean up instantiation of
nested-name-specifiers with dependent template specialization types in
the process.

llvm-svn: 126663
2011-02-28 18:50:33 +00:00
Argyrios Kyrtzidis 2c49ec7f1d [analyzer] Migrate NSErrorChecker and DereferenceChecker to CheckerV2.
They cooperate in that NSErrorChecker listens for ImplicitNullDerefEvent events that
DereferenceChecker can dispatch.
ImplicitNullDerefEvent is when we dereferenced a location that may be null.

llvm-svn: 126659
2011-02-28 17:36:18 +00:00
Argyrios Kyrtzidis a15dfec3f5 [analyzer] Introduce "event" mechanism in CheckerManager.
A checker can register as receiver/listener of "events" (basically it registers a callback
with a function getting called with an argument of the event type) and other checkers can
register as "dispatchers" and can pass an event object to all the listeners.
This allows cooperation amongst checkers but with very loose coupling.

llvm-svn: 126658
2011-02-28 17:36:09 +00:00
Argyrios Kyrtzidis 98b570ecad [analyzer] Run AST checkers for ObjCMethodDecls.
llvm-svn: 126657
2011-02-28 17:36:04 +00:00
Douglas Gregor 5a064725d6 Eliminate the last remains of TemplateSpecializationTypes with
dependent template names. There is still a lot of redundant code in
TreeTransform to cope with TemplateSpecializationTypes, which I'll
remove in stages.

llvm-svn: 126656
2011-02-28 17:23:35 +00:00
David Chisnall da20991a1b Make -fobjc-nonfragile-abi and -fgnu-runtime imply -fblocks (unless -fno-blocks is specified), because this combination of flags defines an Objective-C runtime that includes a blocks runtime.
llvm-svn: 126654
2011-02-28 17:11:43 +00:00
Douglas Gregor cce5c97ac3 Unbreak CMake build
llvm-svn: 126653
2011-02-28 14:32:11 +00:00
Axel Naumann aa627ba724 From Vassil Vassilev:
* Add default implementations (no-op) for ExternalASTSource's pure virtual functions. There are valid use cases that can live with these defaults.
* Move ExternalASTSource's out of line implementations into separate source file.
* Whitespace, forward decl, #include cleanup.

llvm-svn: 126648
2011-02-28 11:22:50 +00:00
NAKAMURA Takumi f5ea88b8ab test: Add the feature "shell". Frontend/dependency-gen.c would be executable with shell.
The feature "shell" is implemented in llvm/test.

llvm-svn: 126646
2011-02-28 09:41:07 +00:00
Chandler Carruth 306f98c241 Add a test case that would have caught the bug fixed in r126640.
I've only covered C++'s flags here, we should probably do something
similar for ObjC flags.

llvm-svn: 126641
2011-02-28 07:38:09 +00:00
Chandler Carruth 74f8711802 Fix a tiny goof in the Driver's logic which caused the explicit presence
of -fexceptions to disably C++ exceptions. The correct code was in the
ObjC branch, this just mirrors that logic on the C++ side of things.
Thanks to John Wiegley for pointing this out.

llvm-svn: 126640
2011-02-28 07:25:18 +00:00
Chris Lattner 35d3ac5e9e Make skipping of vardecls more precise: it's ok to skip a decl if the entire
compound stmt containing the decl is skipped.

llvm-svn: 126639
2011-02-28 07:22:44 +00:00
Chris Lattner 62208c395a make switch constant folding a bit stronger, handling a missed case.
llvm-svn: 126638
2011-02-28 07:16:14 +00:00
NAKAMURA Takumi 8fbf64f420 [PR9321] "make install" does not need install c-index-test. It is used for testing.
llvm-svn: 126634
2011-02-28 05:21:34 +00:00
NAKAMURA Takumi f7c666c38a CMake: [PR9321] Don't install include/clang/**/*.td.
llvm-svn: 126633
2011-02-28 05:21:27 +00:00
Peter Collingbourne 2f1e36bfd0 Rename tok::eom to tok::eod.
The previous name was inaccurate as this token in fact appears at
the end of every preprocessing directive, not just macro definitions.
No functionality change, except for a diagnostic tweak.

llvm-svn: 126631
2011-02-28 02:37:51 +00:00
Anders Carlsson e96ab55b28 Add a -fcxx-exceptions flag to the frontend, which can be used to enable
C++ exceptions, even when exceptions have been turned off using -fno-exceptions.
Make the -fobjc-exceptions flag do the same thing, but for Objective-C exceptions.

C++ and Objective-C exceptions can also be disabled using -fno-cxx-excptions and
-fno-objc-exceptions.

llvm-svn: 126630
2011-02-28 02:27:16 +00:00
Anders Carlsson 64965ad75d Fix CMake build.
llvm-svn: 126629
2011-02-28 02:13:22 +00:00
Ken Dyck 1b4420ed42 Retry r126357. Use CharUnits for the Size and DataSize calculations when
they are known to be exact multiples of the width of the char type. Add a
test case to CodeGen/union.c that would have caught the problem with the
previous attempt. No change in functionality intended.

llvm-svn: 126628
2011-02-28 02:01:38 +00:00
Argyrios Kyrtzidis 4b45f6628e [analyzer] Introduce SVal::getAsVarDecl().
llvm-svn: 126627
2011-02-28 01:28:18 +00:00
Argyrios Kyrtzidis 6d6801c5c7 [analzyer] Migrate CallAndMessageChecker to CheckerV2.
llvm-svn: 126626
2011-02-28 01:28:13 +00:00
Argyrios Kyrtzidis 64fe456a6b [analyzer] ExprEngine should not depend on checkers for not crashing.
llvm-svn: 126625
2011-02-28 01:28:08 +00:00
Argyrios Kyrtzidis 0a5a41d799 [analyzer] Migrate AdjustedReturnValueChecker to CheckerV2.
llvm-svn: 126624
2011-02-28 01:28:05 +00:00
Argyrios Kyrtzidis 6fff2e3d36 [analyzer] Migrate AttrNonNullChecker to CheckerV2.
llvm-svn: 126623
2011-02-28 01:28:01 +00:00
Argyrios Kyrtzidis 4f7745a3b1 [analyzer] ExprEngine should not depend on checkers for not crashing.
llvm-svn: 126622
2011-02-28 01:27:57 +00:00
Argyrios Kyrtzidis 68ed625bd3 [analyzer] Migrate VLASizeChecker to CheckerV2.
llvm-svn: 126621
2011-02-28 01:27:54 +00:00
Argyrios Kyrtzidis ae468f77fa [analyzer] Migrate DivZeroChecker to CheckerV2.
llvm-svn: 126620
2011-02-28 01:27:50 +00:00
Argyrios Kyrtzidis 4dc7fb37cb [analyzer] Migrate ReturnUndefChecker to CheckerV2.
llvm-svn: 126619
2011-02-28 01:27:46 +00:00
Argyrios Kyrtzidis da6c7568e2 [analyzer] Migrate UndefinedArraySubscriptChecker to CheckerV2.
llvm-svn: 126618
2011-02-28 01:27:41 +00:00
Argyrios Kyrtzidis 098874a2f8 [analyzer] Migrate UndefinedAssignmentChecker to CheckerV2.
llvm-svn: 126617
2011-02-28 01:27:37 +00:00
Argyrios Kyrtzidis 753b3ca32f [analyzer] Migrate UndefBranchChecker to CheckerV2.
llvm-svn: 126616
2011-02-28 01:27:33 +00:00
Argyrios Kyrtzidis 60b6da721f [analyzer] Migrate UndefCapturedBlockVarChecker to CheckerV2.
llvm-svn: 126615
2011-02-28 01:27:26 +00:00
Argyrios Kyrtzidis d4d3cee6e4 [analyzer] Migrate UndefResultChecker to CheckerV2.
llvm-svn: 126614
2011-02-28 01:27:22 +00:00
Argyrios Kyrtzidis 142dbbfcd8 [analyzer] Migrate NoReturnFunctionChecker to CheckerV2.
llvm-svn: 126613
2011-02-28 01:27:17 +00:00
Argyrios Kyrtzidis 3e7ab19863 [analyzer] Move the DeadStores checker out of the 'core' package.
-Now it gets enabled with '-analyzer-checker=DeadStores'.
-The driver passes the above flag by default.

llvm-svn: 126612
2011-02-28 01:27:12 +00:00
Argyrios Kyrtzidis f3ed8b631d [analyzer] Migrate BuiltinFunctionChecker to CheckerV2.
llvm-svn: 126611
2011-02-28 01:27:07 +00:00
Argyrios Kyrtzidis b2cf708395 [analyzer] Migrate OSAtomicChecker to CheckerV2.
llvm-svn: 126610
2011-02-28 01:27:02 +00:00
Argyrios Kyrtzidis 0a9ce3ec8f [analyzer] Migrate ArrayBoundCheckerV2 to CheckerV2.
Turns -analyzer-check-buffer-overflows into -analyzer-checker=core.experimental.Overflow

llvm-svn: 126609
2011-02-28 01:26:57 +00:00
Argyrios Kyrtzidis 560bbb1241 [analyzer] Turn -analyzer-stats into -analyzer-checker=debug.Stats
llvm-svn: 126608
2011-02-28 01:26:50 +00:00
Argyrios Kyrtzidis 21c9423ef4 [analyzer] Remove '-analyzer-experimental-checks' flag.
llvm-svn: 126607
2011-02-28 01:26:43 +00:00
Argyrios Kyrtzidis 183f0fb4cf [analyzer] Migrate MallocChecker to CheckerV2.
llvm-svn: 126606
2011-02-28 01:26:35 +00:00
Argyrios Kyrtzidis 4e23dbd0fd [analyzer] Have CheckerManager::registerChecker return a pointer to the checker object and
only allow a checker to be registered once.

llvm-svn: 126605
2011-02-28 01:26:28 +00:00
Chris Lattner dc2cc67e57 remove a bogus assertion, add a comment.
llvm-svn: 126603
2011-02-28 01:06:02 +00:00
Chris Lattner 0709542628 make switch condition constant folding much more aggressive, handling
compound statements and break statements.  This implements enough to 
handle PR9322 and rdar://6970405.

llvm-svn: 126602
2011-02-28 01:02:29 +00:00
Anders Carlsson 246ff3f19f Factor code out into a helper function, shouldUseExceptionTablesForObjCExceptions.
llvm-svn: 126601
2011-02-28 00:44:51 +00:00
Chris Lattner 30db828ce7 fix my containsBreak predicate.
llvm-svn: 126600
2011-02-28 00:42:31 +00:00
Anders Carlsson 6774b1f1c1 Add -fcxx-exceptions to all tests that use C++ exceptions.
llvm-svn: 126599
2011-02-28 00:40:07 +00:00
Anders Carlsson 6dc07d4e56 Get rid of the areExceptionsEnabled() getter from LangOptions.
llvm-svn: 126598
2011-02-28 00:33:03 +00:00
Chris Lattner 0725a8b653 First tiny step to implementing PR9322: build infrastructure for only emitting the
live case of a switch statement when switching on a constant.  This is terribly
limited, but enough to handle the trivial example included.  Before we would 
emit:

define void @test1(i32 %i) nounwind {
entry:
  %i.addr = alloca i32, align 4
  store i32 %i, i32* %i.addr, align 4
  switch i32 1, label %sw.epilog [
    i32 1, label %sw.bb
  ]

sw.bb:                                            ; preds = %entry
  %tmp = load i32* %i.addr, align 4
  %inc = add nsw i32 %tmp, 1
  store i32 %inc, i32* %i.addr, align 4
  br label %sw.epilog

sw.epilog:                                        ; preds = %sw.bb, %entry
  switch i32 0, label %sw.epilog3 [
    i32 1, label %sw.bb1
  ]

sw.bb1:                                           ; preds = %sw.epilog
  %tmp2 = load i32* %i.addr, align 4
  %add = add nsw i32 %tmp2, 2
  store i32 %add, i32* %i.addr, align 4
  br label %sw.epilog3

sw.epilog3:                                       ; preds = %sw.bb1, %sw.epilog
  ret void
}

now we emit:

define void @test1(i32 %i) nounwind {
entry:
  %i.addr = alloca i32, align 4
  store i32 %i, i32* %i.addr, align 4
  %tmp = load i32* %i.addr, align 4
  %inc = add nsw i32 %tmp, 1
  store i32 %inc, i32* %i.addr, align 4
  ret void
}

This improves -O0 compile time (less IR to generate and shove through the code 
generator) and the clever linux kernel people found a way to fail to build if we 
don't do this optimization.  This step isn't enough to handle the kernel case
though.

llvm-svn: 126597
2011-02-28 00:22:07 +00:00
Chris Lattner 29911cc2f5 Add some helper methods that will be used in my next patch.
llvm-svn: 126596
2011-02-28 00:18:40 +00:00
Chris Lattner 80d51eb03a tidy up
llvm-svn: 126595
2011-02-28 00:18:06 +00:00
Chris Lattner 589b989713 add a const version of this accessor.
llvm-svn: 126594
2011-02-28 00:17:41 +00:00
Douglas Gregor 6e068014f0 When we encounter a dependent template name within a
nested-name-specifier, e.g., 

  T::template apply<U>::

represent the dependent template name specialization as a
DependentTemplateSpecializationType, rather than a
TemplateSpecializationType with a dependent TemplateName.

llvm-svn: 126593
2011-02-28 00:04:36 +00:00
Chris Lattner 41c6ab538a Change the interface to ConstantFoldsToSimpleInteger to not encode
a bool + success into one tri-state integer, simplifying things.

llvm-svn: 126592
2011-02-27 23:02:32 +00:00
Douglas Gregor b09518c551 Eliminate a silly little Parse/Sema dance when parsing typename
specifiers such as

  typename T::template apply<U>

Previously, we would turn T::template apply<U> into a
TemplateSpecializationType. Then, we'd reprocess that
TemplateSpecializationType and turn it into either a
TemplateSpecializationType wrapped in an ElaboratedType (when we could
resolve "apply" to a template declaration) or a
DependentTemplateSpecializationType. We now produce the same ASTs but
without generating the intermediate TemplateSpecializationType.

The end goal here is to avoid generating TemplateSpecializationTypes
with dependent template-names, ever. We're not there yet.

llvm-svn: 126589
2011-02-27 22:46:49 +00:00
Benjamin Kramer 6e152156d5 Fix an obvious typo.
GCC -Waddress warns about this but clang doesn't (PR9043).

llvm-svn: 126577
2011-02-27 18:07:41 +00:00
Oscar Fuentes 6401523405 CMake: updated list of installable header files. PR9321.
llvm-svn: 126572
2011-02-27 13:33:31 +00:00
Frits van Bommel ba6ce03e46 Fix this test case for CMake builds after r126502, which sneakily changed the actual executable name to clang-<version>.
llvm-svn: 126560
2011-02-27 01:17:12 +00:00
Daniel Dunbar 160fc36896 cc1: Fix stats printing by default when using -mllvm -stats.
llvm-svn: 126559
2011-02-26 23:17:25 +00:00
Alexis Hunt c5575cced8 Implement delegating constructors partially.
This successfully performs constructor lookup and verifies that a
delegating initializer is the only initializer present.

This does not perform loop detection in the initialization, but it also
doesn't codegen delegating constructors at all, so this won't cause
runtime infinite loops yet.

llvm-svn: 126552
2011-02-26 19:13:13 +00:00
John McCall c57994467b Don't assume that whoever is asking for a message send is going
to give us a non-null return slot.

llvm-svn: 126544
2011-02-26 09:48:59 +00:00
John McCall 71309be80a Zero-initialize the struct-return slot of an Objective-C message
send before making the call.  Fixes rdar://problem/7854674

llvm-svn: 126543
2011-02-26 09:12:15 +00:00
John McCall 8a3c55567e A much-delayed response to Ken Dyck's review: make this somewhat more
char-width agnostic.

llvm-svn: 126542
2011-02-26 08:41:59 +00:00
John McCall 4b13633082 Don't crash during constant-evaluation of 1/(1/0). PR9262.
llvm-svn: 126541
2011-02-26 08:27:17 +00:00
John McCall 53fcbd2718 Pretty up the emission of field l-values and use volatile and TBAA when
loading references as part of that.  Use 'char' TBAA when accessing
(immediate!) fields of a may_alias struct;  fixes PR9307.

llvm-svn: 126540
2011-02-26 08:07:02 +00:00
John McCall bebede4d33 Provide a bit saying that a builtin undergoes custom type-checking, then
don't let calls to such functions go down the normal type-checking path.
Test this out with __builtin_classify_type and __builtin_constant_p.

llvm-svn: 126539
2011-02-26 05:39:39 +00:00
Fariborz Jahanian 8f4903397d Fix objc rewriting bug casting to qualified objective-c pointetr.
// rdar://9056351

llvm-svn: 126536
2011-02-26 01:31:36 +00:00
Fariborz Jahanian f2890fc93f Fix a rewriter bug involving call to property's
block. // rdar://9055596

llvm-svn: 126535
2011-02-26 00:33:41 +00:00
Matt Beaumont-Gay 531c079de8 ASTStmtReader is already a friend of DependentScopeDeclRefExpr, and GCC
doesn't like re-friending.

llvm-svn: 126524
2011-02-25 22:36:42 +00:00
Ted Kremenek e925322f41 Update test cases.
llvm-svn: 126523
2011-02-25 22:19:14 +00:00
Ted Kremenek 5eb4b60734 Tidy up help text in Checkers.td, and rename StackAddrLeakChecker to StackAddrEscapeChecker.
llvm-svn: 126522
2011-02-25 22:00:43 +00:00