This allows us to keep from chaining LazyCompoundVals in cases like this:
CGRect r = CGRectMake(0, 0, 640, 480);
CGRect r2 = r;
CGRect r3 = r2;
Previously we only made this optimization if the struct did not begin with
an aggregate member, to make sure that we weren't picking up an LCV for
the first field of the struct. But since LazyCompoundVals are typed, we can
make that inference directly by comparing types.
This is a pure optimization; the test changes are to guard against possible
future regressions.
llvm-svn: 174211
r173593 made us a little too eager to associate all code at the end of a
function with the user-written 'return' line. This caused problems with
breakpoints as they'd be set in exception handling code preceeding the
actual non-exception return handling code, leading to the breakpoint never
being hit in non-exceptional execution.
This change restores the pre-r173593 exception handling line information where
the cleanup code is associated with the '}' not the return line.
llvm-svn: 174206
Prior to the patch, Clang does not properly promote types when a complex
integer operand is combined with an integer via a binary operator, or when
one is assigned to the other in either order. This patch detects when
promotion is needed (and permissible) and generates the necessary code.
The test assmes no target has the same size operands for "char" and
"long long," and that no target performs arithmetic on char operands without
extending them to a larger format first. If there are any targets for
which this is not the case, they should be XFAILed.
llvm-svn: 174181
Introduces these negation forms explicitly and uses them to control a new
"altivec" target feature for PowerPC. This allows avoiding generating
Altivec instructions on processors that support Altivec.
The new test case verifies that the Altivec "lvx" instruction is not
used when -fno-altivec is present on the command line.
llvm-svn: 174140
Remove "IsMSDeclspec" argument from Align attribute since the arguments in Attr.td should
only model those appear in source code. Introduce attribute Accessor, and teach TableGen
to generate syntax kind accessors for Align attribute, and use those accessors to decide
if an alignment attribute is a declspec attribute.
llvm-svn: 174133
This matches our behavior for autorelease pools created by +alloc. Some
people like to create autorelease pools in one method and release them
somewhere else.
If you want safe autorelease pool semantics, use the new ARC-compatible
syntax: @autoreleasepool { ... }
<rdar://problem/13121353>
llvm-svn: 174096
This allows people to use Unicode in their #pragma mark and in macros
that exist only to be string-ized.
<rdar://problem/13107323&13121362>
llvm-svn: 174081
It's causing hangs on our internal analyzer buildbot. Will restore after
investigating.
This reverts r173951 / baa7ca1142990e1ad6d4e9d2c73adb749ff50789.
llvm-svn: 174069
In cooperation with the LLVM patch, this should implement all scalar front-end
parts of the C and C++ ABIs for AArch64.
This patch excludes the NEON support also reviewed due to an outbreak of
batshit insanity in our legal department. That will be committed soon bringing
the changes to precisely what has been approved.
Further reviews would be gratefully received.
llvm-svn: 174055
have a direct mismatch between some component of the template and some
component of the argument. The diagnostic now says what the mismatch was, but
doesn't yet say which part of the template doesn't match.
llvm-svn: 174039
Objective-C message receiver, the user is as likely to want to write a
type name as any other declaration, so give types the same priority as
other declarations. Fixes <rdar://problem/12480600>.
llvm-svn: 174038
This is a hack to work around the fact that we don't track extents for our
default bindings:
CGPoint p;
p.x = 0.0;
p.y = 0.0;
rectParam.origin = p;
use(rectParam.size); // warning: uninitialized value in rectParam.size.width
In this case, the default binding for 'p' gets copied into 'rectParam',
because the 'origin' field is at offset 0 within CGRect. From then on,
rectParam's old default binding (in this case a symbol) is lost.
This patch silences the warning by pretending that lazy bindings are never
made from uninitialized memory, but not only is that not true, the original
default binding is still getting overwritten (see FIXME test cases).
The long-term solution is tracked in <rdar://problem/12701038>
PR14765 and <rdar://problem/12875012>
llvm-svn: 174031
them the same if they are actually the same; having the same name isn't enough.
Similar to r174013, template template arguments were also mistakenly considered
the same when they had the same name but were in different namespaces.
In addition, when printing template template arguments, use the qualified name
if the regular name is the same.
llvm-svn: 174029
Indents were given the color blue when outputting with color.
AST dumping now looks like this:
Node
|-Node
| `-Node
`-Node
`-Node
Compared to the previous:
(Node
(Node
(Node))
(Node
(Node)))
llvm-svn: 174022
positives.
The includeSuffix was only set on the first iteration through the
function, resulting in invalid regions being produced by getLazyBinding
(ex: zoomRegion.y).
llvm-svn: 174016
Specifically, don't print snippets, caret diagnostics, or ranges for
lines over 4096 characters. We copy the line around a few times in our
diagnostics machinery, and we have to print a caret line that's just as
long. This uses a lot of memory just to create a poor user experience as
we print out a line much too long for anyone to read...or spend extra
energy trying to fit it to -fmessage-length.
<rdar://problem/13106850>
llvm-svn: 173976
Redefine the shallow mode to inline all functions for which we have a
definite definition (ipa=inlining). However, only inline functions that
are up to 4 basic blocks large and cut the max exploded nodes generated
per top level function in half.
This makes shallow faster and allows us to keep inlining small
functions. For example, we would keep inlining wrapper functions and
constructors/destructors.
With the new shallow, it takes 104s to analyze sqlite3, whereas
the deep mode is 658s and previous shallow is 209s.
llvm-svn: 173958
This is faster for the analyzer to process than inlining the constructor
and performing a member-wise copy, and it also solves the problem of
warning when a partially-initialized POD struct is copied.
Before:
CGPoint p;
p.x = 0;
CGPoint p2 = p; <-- assigned value is garbage or undefined
After:
CGPoint p;
p.x = 0;
CGPoint p2 = p; // no-warning
This matches our behavior in C, where we don't see a field-by-field copy.
<rdar://problem/12305288>
llvm-svn: 173951
__fp16 isn't covered by the standard, but this resolves the oddity that float
gets promoted when passed variadically, but not the smaller type. This is
required by the AArch64 ABI, and a sane action elsewhere.
llvm-svn: 173918
People use the C preprocessor for things other than C files. Some of them
have Unicode characters. We shouldn't warn about Unicode characters
appearing outside of identifiers in this case.
There's not currently a way for the preprocessor to tell if it's in -E mode,
so I added a new flag, derived from the PreprocessorOutputOptions. This is
only used by the Unicode warnings for now, but could conceivably be used by
other warnings or even behavioral differences later.
<rdar://problem/13107323>
llvm-svn: 173881
If the member has an initializer, assume it was probably intended to be static
and suggest/recover with that.
If the member doesn't have an initializer, assume it was probably intended to
be const instead of constexpr and suggest that.
(if the attempt to apply these changes fails, don't make any suggestion &
produce the same diagnostic experience as before. The only case where this can
come up that I know of is with a mutable constexpr with an initializer, since
mutable is incompatible with static (but it's already incompatible with
const anyway))
llvm-svn: 173873
argument to be memset, check for its type to be complete
before calling Context.getTypeSize(PointeeTy) to prevent
crash. // rdar://13081751.
llvm-svn: 173872
implementation; this is much more inline with the original implementation
(i.e., pre-ubsan) and does not require run-time library support.
The trapping implementation can be invoked using either '-fcatch-undefined-behavior'
or '-fsanitize=undefined-trap -fsanitize-undefined-trap-on-error', with the latter
being preferred. Eventually, the -fcatch-undefined-behavior' flag will be removed.
llvm-svn: 173848
if they were already concatenated in source using the spelling locations
even if they came from a macro expansion.
This fixes an issue where a GUID passed as macro argument ends up
malformed after preprocessing because we added spaces inside it.
rdar://13016645
llvm-svn: 173826
subsequent commands from being executed.
The diagnostics generation isn't designed for this use case, so add a note to
fix this in the very near future. For now, just generated the diagnostics for
the first failing command.
Part of rdar://12984531
llvm-svn: 173825
the diagnostic's warn_ name. Switch some places (notably C++11 attributes)
which really wanted an error over to a different diagnostic. Finally, suppress
the diagnostic entirely for __ptr32, __ptr64 and __w64, to avoid producing
diagnostics in important system headers.
llvm-svn: 173788
It turns out that there's no correctness bug here (because we can't have a type
definition in this location), but there was a diagnostic bug.
llvm-svn: 173766
working, and add the missing attribute spellings. This brings _pascal,
_fastcall, _stdcall and _cdecl to life in -fborland-extensions mode.
llvm-svn: 173749
as a keyword. Rationalize existing attributes to use it as appropriate, and to
not lie about some __declspec attributes being GNU attributes. In passing,
remove a gross hack which was discarding attributes which we could handle. This
results in us actually respecting the __pascal keyword again.
llvm-svn: 173746
This required plumbing through a new flag to determine whether a ParmVarDecl is
actually a parameter of a function declaration (as opposed to a function
typedef etc, where the attribute is prohibited). Weirdly, this attribute (just
like [[noreturn]]) cannot be applied to a function type, just to a function
declaration (and its parameters).
llvm-svn: 173726
-fno-modules-global-index -cc1 option to allow one to disable the
index for performance testing purposes, but with a 10% win in
-fsyntax-only time, there is no reason a user would do this.
llvm-svn: 173707
1. Use a hanging ident for function calls nested in binary expressions.
E.g.:
int aaaaa = aaaaaaaaa && aaaaaaaaaa(
aaaaaaaaaa);
2. Slightly improve heuristic for builder type expressions and reduce
penalty for breaking before "." and "->" in those.
3. Remove mostly obsolete metric of decreasing indent level. This
fixes: llvm.org/PR14931.
Changes #1 and #2 were necessary to keep tests passing after #3.
llvm-svn: 173680
The -E output from clang did not produce the correct indentation on the first line.
This is because MoveToLine returned false, and when this happens,
the regular process for producing initial indentation is skipped.
Thanks to Eli for suggesting a way to simplify this to a one-line change.
llvm-svn: 173657
For example,
cur) unittests/ADT/Release/ADTTests
new) unittests/ADT/ADTTests
RUNTIME_BUILD_MODE can be substituted to CMAKE_CFG_INTDIR.
With Make and Ninja, the tree is not built with multiple configurations.
Then, including the build type in target directory doesn't make sense.
See also "How can I build multiple modes without switching?"
http://www.cmake.org/Wiki/CMake_FAQ
CMAKE_CFG_INTDIR is set to "."
With multiple-configuration-aware build system, like Visual Studio, each unittest is built on appropriate directory, for example,
unittests/ADT/Release/ADTTests.exe
CMAKE_CFG_INTDIR is set to build system's variable, like "$(Configuration)" or "$(OutDir)".
Thus, "--param build_config" is also deprecated.
llvm-svn: 173616
One of the gotchas (see changes to CodeGenFunction) was due to the fix in
r139416 (for PR10829). This only worked previously because the top level
lexical block would set the location to the end of the function, the debug
location would be updated (as per r139416), the location would be set to
the end of the function again (but that would no-op, since it was the same
as the previous location), then the return instruction would be emitted using
the debug location.
Once the top level lexical block was no longer emitted, the end-of-function
location change was causing the debug loc to be updated, regressing that bug.
llvm-svn: 173593
When the analyzer sees an initializer, it checks if the initializer
contains a CXXConstructExpr. If so, it trusts that the CXXConstructExpr
does the necessary work to initialize the object, and performs no further
initialization.
This patch looks through any implicit wrapping expressions like
ExprWithCleanups to find the CXXConstructExpr inside.
Fixes PR15070.
llvm-svn: 173557
on a type. Currently, it gives a generic "expected unqualified-id" error.
The new error message is "cannot use (dot|arrow) operator on a type".
llvm-svn: 173556
-f(no-)color-diagnostics. In addition, dumpColor() function calls are added
to force color printing. No structural changes to -ast-dump.
llvm-svn: 173548
The expression 'a->b.c()' contains a call to the 'c' method of 'a->b'.
We emit an error if 'a' is NULL, but previously didn't actually track
the null value back through the 'a->b' expression, which caused us to
miss important false-positive-suppression cases, including
<rdar://problem/12676053>.
llvm-svn: 173547
Title: [PR9027] volatile struct bug: member is not loaded at -O;
This is caused by last flag passed to @llvm.memcpy being false,
not honoring that aggregate has at least one 'volatile' data member
(even though aggregate itself has not been qualified as 'volatile'.
As a result, optimization optimizes away the memcpy altogether.
Patch review by John MaCall (I still need to fix up a test though).
llvm-svn: 173535
You may see such a message on non-posix system;
ignoring nonexistent directory "/usr/local/include"
ignoring nonexistent directory "/usr/include"
llvm-svn: 173525
never key functions. We did not implement that rule for the
iOS ABI, which was driven by what was implemented in gcc-4.2.
However, implement it now for other ARM-based platforms.
llvm-svn: 173515
- This just scratches the surface, We have pretty horrible test coverage in
this area it seems like, but this at least covers the change in r173410.
llvm-svn: 173464
with -target i686-win32, you will see;
debug-info-static-member.cpp:11:22: error: in-class initializer for static data member of type 'const float' requires 'constexpr' specifier [-Wstatic-float-init]
const static float const_b = 3.14;
^ ~~~~
constexpr
llvm-svn: 173418
and split it out of -Wgnu into its own warning flag.
* In C++11, this is now a hard error (GCC has no extension here in C++11 mode).
The error can be disabled with -Wno-static-float-init, and has a fixit to
add 'constexpr'.
* In C++98, this is still an ExtWarn, but is now controlled by
-Wstatic-float-init as well as -Wgnu.
llvm-svn: 173414
AST reader.
The global module index tracks all of the identifiers known to a set
of module files. Lookup of those identifiers looks first in the global
module index, which returns the set of module files in which that
identifier can be found. The AST reader only needs to look into those
module files and any module files not known to the global index (e.g.,
because they were (re)built after the global index), reducing the
number of on-disk hash tables to visit. For an example source I'm
looking at, we go from 237844 total identifier lookups into on-disk
hash tables down to 126817.
Unfortunately, this does not translate into a performance advantage.
At best, it's a wash once the global module index has been built, but
that's ignore the cost of building the global module index (which
is itself fairly large). Profiles show that the global module index
code is far less efficient than it should be; optimizing it might give
enough of an advantage to justify its continued inclusion.
llvm-svn: 173405
The idea is to introduce a higher level "user mode" option for
different use scenarios. For example, if one wants to run the analyzer
for a small project each time the code is built, they would use
the "shallow" mode.
The user mode option will influence the default settings for the
lower-level analyzer options. For now, this just influences the ipa
modes, but we plan to find more optimal settings for them.
llvm-svn: 173386
The idea is to eventually place all analyzer options under
"analyzer-config". In addition, this lays the ground for introduction of
a high-level analyzer mode option, which will influence the
default setting for IPAMode.
llvm-svn: 173385
This is a missing piece for C99 conformance.
This patch handles UCNs by adding a '\\' case to LexTokenInternal and
LexIdentifier -- if we see a backslash, we tentatively try to read in a UCN.
If the UCN is not syntactically well-formed, we fall back to the old
treatment: a backslash followed by an identifier beginning with 'u' (or 'U').
Because the spelling of an identifier with UCNs still has the UCN in it, we
need to convert that to UTF-8 in Preprocessor::LookUpIdentifierInfo.
Of course, valid code that does *not* use UCNs will see only a very minimal
performance hit (checks after each identifier for non-ASCII characters,
checks when converting raw_identifiers to identifiers that they do not
contain UCNs, and checks when getting the spelling of an identifier that it
does not contain a UCN).
This patch also adds basic support for actual UTF-8 in the source. This is
treated almost exactly the same as UCNs except that we consider stray
Unicode characters to be mistakes and offer a fixit to remove them.
llvm-svn: 173369
Introduce a spelling index to Attr class, which is an index into the attribute spelling list of an attribute defined in Attr.td.
This index will determine the actual spelling used by an attribute, as it incorporates both the syntax and naming of the attribute.
When constructing an attribute AST node, the spelling index is computed based on attribute kind, scope (if it's a C++11 attribute), and
name, then passed to Attr that will use the index to print itself.
Thanks to Richard Smith for the idea and review.
llvm-svn: 173358
It had program scope variables that were not in the constant address space,
make them to be function scope variables instead.
Also move the test to the SemaOpenCL directory.
llvm-svn: 173352
The global module index is a "global" index for all of the module
files within a particular subdirectory in the module cache, which
keeps track of all of the "interesting" identifiers and selectors
known in each of the module files. One can perform a fast lookup in
the index to determine which module files will have more information
about entities with a particular name/selector. This information can
help eliminate redundant lookups into module files (a serious
performance problem) and help with creating auto-import/auto-include
Fix-Its.
The global module index is created or updated at the end of a
translation unit that has triggered a (re)build of a module by
scraping all of the .pcm files out of the module cache subdirectory,
so it catches everything. As with module rebuilds, we use the file
system's atomicity to synchronize.
llvm-svn: 173301
GCC implements -Wvla as "warn on every VLA" (this is useful to find every VLA,
for example, if they are forbidden by coding guidelines). Currently Clang
implements -Wvla as "warn on VLA when it is an extension".
The attached patch makes our behavior match GCC. The existing vla extwarn is
moved under -Wvla-extension and is still included into -Wgnu.
This fixes PR5953.
llvm-svn: 173286
that redefined a macro without undef'ing it first.
Proper reconstruction of the macro info history from modules will properly fix this in subsequent commits.
rdar://13016031
llvm-svn: 173281
This allows users to promote -Wincompatible-pointer-type warnings to
errors but keep those for "discard qualifiers" as warnings (if they
so desire).
Addresses <rdar://problem/13062738>.
llvm-svn: 173184
This change also makes the serialisation store the required semantics,
fixing an issue where PPC128 was always assumed when re-reading a
128-bit value.
llvm-svn: 173139
We were previously hard-coding a particular field index. This was
fine before (because we were obviously guaranteed the presence
of a copy/dispose member) except for (1) alignment padding and
(2) future extensions adding extra members to the header, such
as the extended-layout pointer.
Note that we only introduce the extended-layout pointer in the
presence of structs. (We also seem to be introducing it even
in the presence of an all-non-object layout, but that's a
different potential issue.)
llvm-svn: 173122
We now only put empty blocks into a single line, if all of:
- all tokens of the structural element fit into a single line
- we're not in a control flow statement
Note that we usually don't put record definitions into a single line, as
there's usually at least one more token (the semicolon) after the
closing brace. This doesn't hold when we are in a context where there is
no semicolon, like "enum E {}".
There were some missing tests around joining lines around the corner
cases of the allowed number of columns, so this patch adds some.
llvm-svn: 173055
Before: template <template <typename T>, typename P > class X;
After: template <template <typename T>, typename P> class X;
More importantly, the token annotations for the second ">" are now computed
correctly.
llvm-svn: 173047
Looks like r161368 fixed this for one case but not all. This change generalizes
the solution over all the unwrapping cases. Now that preserving the qualifiers
is done independent of the particular type being unwrapped I won't bother
adding test cases for each one but at least demonstrate that this change was
necessary & sufficient to fix the bug.
llvm-svn: 173002
Before:
Calling implicit default constructor for 'Foo' (where Foo is constructed)
Entered call from 'test' (at "=default" or 'Foo' declaration)
Calling default constructor for 'Bar' (at "=default" or 'Foo' declaration)
After:
Calling implicit default constructor for 'Foo' (where Foo is constructed)
Calling default constructor for 'Bar' (at "=default" or 'Foo' declaration)
This only affects the plist diagnostics; this note is never shown in the
other diagnostics.
llvm-svn: 172915
Adding the pseudo first parameter to a member function pointer's function type
and mark it as artificial.
Combined with a fix to GDB ( http://sourceware.org/bugzilla/show_bug.cgi?id=14998 )
this fixes gdb.cp/member-ptr.exp with Clang.
llvm-svn: 172911
Suppress the warning by just not emitting the report. The sink node
would get generated, which is fine since we did reach a bad state.
Motivation
Due to the way code is structured in some of these macros, we do not
reason correctly about it and report false positives. Specifically, the
following loop reports a use-after-free. Because of the way the code is
structured inside of the macro, the analyzer assumes that the list can
have cycles, so you end up with use-after-free in the loop, that is
safely deleting elements of the list. (The user does not have a way to
teach the analyzer about shape of data structures.)
SLIST_FOREACH_SAFE(item, &ctx->example_list, example_le, tmpitem) {
if (item->index == 3) { // if you remove each time, no complaints
assert((&ctx->example_list)->slh_first == item);
SLIST_REMOVE(&ctx->example_list, item, example_s, example_le);
free(item);
}
}
llvm-svn: 172883
partially-substituted parameter pack in a template, forget about the
partially-substituted parameter pack: it is now completed. Fixes
<rdar://problem/12176336>.
llvm-svn: 172859
r159549 / r159164 regressed clang to reject
struct s {};
struct s
operator++(struct s a)
{ return a; }
This fixes the regression. Richard, pleas check if this looks right.
llvm-svn: 172834
').' is likely part of a builder pattern statement.
This is based upon a patch developed by Nico Weber. Thank you!
Before:
int foo() {
return llvm::StringSwitch<Reference::Kind>(name).StartsWith(
".eh_frame_hdr", ORDER_EH_FRAMEHDR).StartsWith(
".eh_frame", ORDER_EH_FRAME).StartsWith(".init", ORDER_INIT).StartsWith(
".fini", ORDER_FINI).StartsWith(".hash", ORDER_HASH).Default(ORDER_TEXT);
}
After:
int foo() {
return llvm::StringSwitch<Reference::Kind>(name)
.StartsWith(".eh_frame_hdr", ORDER_EH_FRAMEHDR)
.StartsWith(".eh_frame", ORDER_EH_FRAME)
.StartsWith(".init", ORDER_INIT).StartsWith(".fini", ORDER_FINI)
.StartsWith(".hash", ORDER_HASH).Default(ORDER_TEXT);
}
Probably not ideal, but makes many cases much more readable.
The changes to overriding-ftemplate-comments.cpp don't seem better or
worse. We should address those soon.
llvm-svn: 172804
return type of a function by canonicalizing them away. They are
useless anyway, and conflict with our rules for template argument
deduction and __strong. Fixes <rdar://problem/12367446>.
llvm-svn: 172768
Presumably, if the printf format has the sign explicitly requested, the user
wants to treat the data as signed.
This is a fix-up for r172739, and also includes several test changes that
didn't make it into that commit.
llvm-svn: 172762
unsequenced operations in the RHS. We don't compare the RHS with the rest of
the expression yet; such checks will need care to avoid diagnosing unsequenced
operations which are both in conditionally-evaluated subexpressions which
actually can't occur together, such as in '(b && ++x) + (!b && ++x)'.
llvm-svn: 172760
AT_producer. Which includes clang's version information so we can tell
which version of the compiler was used.
This is second of the two steps to allow us to do this. The first was a
change to llvm-mc with revision 172630 to provide a method to set the
AT_producer string. This second step has the clang driver passing the value
of getClangFullVersion() via the new flag -dwarf-debug-producer when invoking
the integrated assembler on assembly source files. Then using the new
setDwarfDebugProducer() method to set the AT_producer string.
rdar://12888242
llvm-svn: 172758
-fopenmp in the link step on Linux. There is probably more tweaking that
will need to take place to get good support for linking the relevant
libraries on all Linux distributions and/or on other platforms, but this
get's the ball moving and allows Clang to build programs which contain
OpenMP pragmas that can be safely ignored by a compiler that doesn't
implement them, and yet makes direct calls into the OpenMP runtime.
llvm-svn: 172715
This is inspired by a number of false positives in real code, including
PR14968. I've added test cases reduced from these false positives to
test/Sema/unused-expr.c, as well as corresponding test cases that pass the
offending expressions as arguments to a no-op macro to ensure that we do warn
there.
This also removes my previous tweak from r166522/r166534, so that we warn on
unused cast expressions in macro arguments.
There were several test cases that were using -Wunused-value to test general
diagnostic emission features; I changed those to use other warnings or warn on
a macro argument expression. I stared at the test case for PR14399 for a while
with Richard Smith and we believe the new test case exercises the same
codepaths as before.
llvm-svn: 172696
it apart from [[gnu::noreturn]] / __attribute__((noreturn)), since their
semantics are not equivalent (for instance, we treat [[gnu::noreturn]] as
affecting the function type, whereas [[noreturn]] does not).
llvm-svn: 172691
expressions which have undefined behavior due to multiple unsequenced
modifications or an unsequenced modification and use of a variable.
llvm-svn: 172690
if we can see the elements of the arrays.
for example:
NSDictionary *dict = [NSDictionary dictionaryWithObjects:[NSArray arrayWithObjects:@"1", @"2", nil] forKeys:[NSArray arrayWithObjects:@"A", @"B", nil]];
-->
NSDictionary *dict = @{ @"A" : @"1", @"B" : @"2" };
rdar://12428166
llvm-svn: 172679
This fixes pr14946. The problem was that the linkage computation was done too
early, so things like "extern int a;" would be given external linkage, even if
a previous declaration was static.
llvm-svn: 172667
consider (sub)module visibility.
The bulk of this change replaces myriad hand-rolled loops over the
linked list of Objective-C categories/extensions attached to an
interface declaration with loops using one of the four new category
iterator kinds:
visible_categories_iterator: Iterates over all visible categories
and extensions, hiding any that have their "hidden" bit set. This is
by far the most commonly used iterator.
known_categories_iterator: Iterates over all categories and
extensions, ignoring the "hidden" bit. This tends to be used for
redeclaration-like traversals.
visible_extensions_iterator: Iterates over all visible extensions,
hiding any that have their "hidden" bit set.
known_extensions_iterator: Iterates over all extensions, whether
they are visible to normal name lookup or not.
The effect of this change is that any uses of the visible_ iterators
will respect module-import visibility. See the new tests for examples.
Note that the old accessors for categories and extensions are gone;
there are *Raw() forms for some of them, for those (few) areas of the
compiler that have to manipulate the linked list of categories
directly. This is generally discouraged.
Part two of <rdar://problem/10634711>.
llvm-svn: 172665
Previously we would serialize the macro redefinitions as a list, part of
the identifier, and try to chain them together across modules individually
without having the info that they were already chained at definition time.
Change this by serializing the macro redefinition chain and then try
to synthesize the chain parts across modules. This allows us to correctly
pinpoint when 2 different definitions are ambiguous because they came from
unrelated modules.
Fixes bogus "ambiguous expansion of macro" warning when a macro in a PCH
is redefined without undef'ing it first.
rdar://13016031
llvm-svn: 172620
Fariborz Jahanian