[OpenCL] Implemented restrictions for pointer conversions specified in OpenCL v2.0.
OpenCL v2.0 s6.5.5 restricts conversion of pointers to different address spaces:
- the named address spaces (__global, __local, and __private) => __generic - implicitly converted;
- __generic => named - with an explicit cast;
- named <=> named - disallowed;
- __constant <=> any other - disallowed.
[OpenCL] Generic address space has been added in OpenCL v2.0.
To support it in the frontend, the following has been added:
- generic address space type attribute;
- documentation for the OpenCL address space attributes;
- parsing of __generic(generic) keyword;
- test code for the parser and diagnostics.
Richard Smith [Wed, 26 Nov 2014 03:26:53 +0000 (03:26 +0000)]
[c++1z] Most of N4268 (allow constant evaluation for non-type template arguments).
We don't yet support pointer-to-member template arguments that have undergone
pointer-to-member conversions, mostly because we don't have a mangling for them yet.
Change to original: ifndef out tests in Windows due to /-separated
paths.
Summary:
Often one is only interested in matches within the main-file or matches
that are not within a system-header, for which this patch adds
isInMainFile and isInSystemFile. They take no arguments and narrow down
the matches.
The isInFileMatchingName is mainly thought for interactive
clang-query-sessions, to make a matcher more specific without restarting
the session with the files you are interested in for that moment. It
takes a string that will be used as regular-expression to match the
filename of where the matched node is expanded.
Reid Kleckner [Mon, 24 Nov 2014 22:05:42 +0000 (22:05 +0000)]
Remove unnecessary environment switch
All supported environments on x86 Windows return structs in EAX:EDX.
This removes code added in r204978 that had to get updated in r222680.
We should now have the same behavior we had before r204978.
Samuel Benzaquen [Mon, 24 Nov 2014 21:21:09 +0000 (21:21 +0000)]
Filter the toplevel matchers by kind.
Summary:
Filter the toplevel matchers by kind.
Decl and Stmt matchers are tied to a specific node kind and trying to
match incompatible nodes is a waste.
Precalculate a filtered list of matchers that have a chance of matching
the node and ignore the rest.
Speeds up our clang-tidy benchmark by ~10%
Paul Robinson [Mon, 24 Nov 2014 20:51:42 +0000 (20:51 +0000)]
Correctly remove OptimizeForSize from functions marked OptimizeNone.
This allows using __attribute__((optnone)) and the -Os/-Oz options.
Fixes PR21604.
Aaron Ballman [Mon, 24 Nov 2014 15:42:34 +0000 (15:42 +0000)]
Reverting r222638; it broke the MSVC build bots because Visual Studio 2012 does not support variadic templates. Also reverting r222641 because it was relying on 222638.
Summary:
Often one is only interested in matches within the main-file or matches
that are not within a system-header, for which this patch adds
isInMainFile and isInSystemFile. They take no arguments and narrow down
the matches.
The isInFileMatchingName is mainly thought for interactive
clang-query-sessions, to make a matcher more specific without restarting
the session with the files you are interested in for that moment. It
takes a string that will be used as regular-expression to match the
filename of where the matched node is expanded.
Cygwin and MinGW fail to conform to the underlying system's structure passing
ABI. Make the check more precise to ensure that we correctly generate code for
the itanium environment.
Daniel Sanders [Sat, 22 Nov 2014 16:21:18 +0000 (16:21 +0000)]
Support matching signext attribute in the parallel_num_threads_codegen test to appease clang-cmake-mips builder.
The Mips target adds the signext attribute to signed 32-bit integers in order
to support the N32/N64 correctly. Integers must be promoted to 64-bit bit on
these ABI's.
Daniel Sanders [Sat, 22 Nov 2014 16:08:56 +0000 (16:08 +0000)]
Support matching signext attribute in vla-lambda-capturing test to appease clang-cmake-mips builder.
The Mips target adds the signext attribute to signed 32-bit integers in order
to support the N32/N64 correctly. Integers must be promoted to 64-bit bit on
these ABI's.
David Majnemer [Fri, 21 Nov 2014 21:09:12 +0000 (21:09 +0000)]
Sema: Don't permit variably modified types in typeid
GCC and ICC both reject this and the 'Runtime-sized arrays with
automatic storage duration' (N3639) paper forbade this as well.
Previously, we would crash on our way to mangling.
David Majnemer [Fri, 21 Nov 2014 19:57:25 +0000 (19:57 +0000)]
MS ABI: Mangle char16_t and char32_t string literals
We previously had support for char and wchar_t string literals. VS 2015
added support for char16_t and char32_t.
String literals must be mangled in the MS ABI in order for them to be
deduplicated across translation units: their linker has no notion of
mergeable section. Instead, they use the mangled name to make a COMDAT
for the string literal; the COMDAT will merge with other COMDATs in
other object files.
This allows strings in object files generated by clang to get merged
with strings in object files generated by MSVC.
Dmitry Vyukov [Fri, 21 Nov 2014 12:19:01 +0000 (12:19 +0000)]
clang: do not add -pie for tsan
Revision 220571 removes the requirement to use -pie for tsan binaries. So remove -pie from driver.
Also s/hasZeroBaseShadow/requiresPIE/ because that is what it is used for. Msan does not have zero-based shadow, but requires pie. And in general the relation between zero-based shadow and pie is unclear.
Bob Wilson [Fri, 21 Nov 2014 06:52:52 +0000 (06:52 +0000)]
Fix missing diagnostic for unsupported TLS for some thread_local variables.
Clang r181627 moved a check for block-scope variables into this code for
handling thread storage class specifiers, but in the process, it broke the
logic for checking if the target supports TLS. Fix this with some simple
restructuring of the code. rdar://problem/18796883
Richard Smith [Fri, 21 Nov 2014 05:37:20 +0000 (05:37 +0000)]
[modules] When explicitly importing a module, it's fine for the imported module
to be newer than we were expecting. That happens if .pcm's get moved between
file systems during a distributed build. (It's still not OK for them to actually
be different, though, so we still check the size and signature matches.)
Adrian Prantl [Fri, 21 Nov 2014 00:35:25 +0000 (00:35 +0000)]
Debug info for blocks: Fix a bug caught by the Verifier.
When emitting nested block definitions, the insert-at-point variant of
DIBuilder::insertDeclare() could be called with the insertion point set
to the end-of-BasicBlock sentinel, causing the parent pointer of the
CallInst to be set to the intentionally bogus value of the sentinel.
Fixed by conditionally invoking the correct version of insertDeclare().
rdar://problem/19034882
Richard Smith [Thu, 20 Nov 2014 23:53:14 +0000 (23:53 +0000)]
Remove code that tries to avoid transforming non-dependent call operator types
for lambda expressions. That can't ever work; we need to transform the
parameters in order to create new ones in the new call operator context.
Fixes a rejects-valid when transforming a context containing a
lambda-expression that uses its function parameters in C++14 mode.
Richard Smith [Thu, 20 Nov 2014 22:32:11 +0000 (22:32 +0000)]
PR21565: Further refine the conditions for enabling eager parsing of
std::X::swap exception specifications (allowing parsing of non-conforming code
in libstdc++). The old conditions also matched the functions in MSVC's STL,
which were relying on deferred parsing here.
Kaelyn Takata [Thu, 20 Nov 2014 22:06:44 +0000 (22:06 +0000)]
Ensure all TypoExprs are diagnosed by the tree transform.
If there is more than one TypoExpr within the expr being transformed and
any but the last TypoExpr seen don't have any viable candidates, the
tree transform will be aborted early and the remaining TypoExprs are
never seen and hence never diagnosed. This adds a simple
RecursiveASTVisitor to find all of the TypoExprs to be diagnosed in the
case where typo correction of the entire expr fails (and the result of
the tree transform is an ExprError).
Kaelyn Takata [Thu, 20 Nov 2014 22:06:30 +0000 (22:06 +0000)]
Factor out the default recovery handling.
The default handling is extended to properly create member expressions
and Objective-C ivar references.
Also detect and reject cases where multiple corrections have identical
correction distances and are valid, instead of suggesting the first one
that is found.
Aaron Ballman [Thu, 20 Nov 2014 21:34:56 +0000 (21:34 +0000)]
The Syntax enumeration requires at least three bits to store all of its values. Increase the number of bits for SyntaxUsed. Decrease the number of expression arguments allowed by a single bit so that the bit fields continue to add up to 32-bits evenly.
There is no test for this fix because I could find no reasonable way to trigger a visible failure from it.
Michael Ilseman [Thu, 20 Nov 2014 19:32:48 +0000 (19:32 +0000)]
Bring PostOrderCFGView's insert API up to date with other API changes.
r222334 updates LLVM data structure's insert API to return a
pair. This change updates PostOrderCFGView accordingly, so that it can
be used interchangably with other sets.
Frederic Riss [Thu, 20 Nov 2014 15:46:59 +0000 (15:46 +0000)]
Prepare for the DIBuilder populating DISubprogram's Variables with null.
This is a followup to r222373. A better solution to the problem solved
there is to not create the leaked nodes at all (we know that they will
never be used for forward declared functions anyway). To avoid bot
breakage in the interval between the cfe and llvm commits, add a check
that the nMDNode is not null before deleting it. This code can completely
go away after the LLVM part is in.
Samuel Benzaquen [Thu, 20 Nov 2014 15:45:53 +0000 (15:45 +0000)]
Replace variadic operator function pointer with an enum value.
Summary:
Replace variadic operator function pointer with an enum value.
Hiding the implementation of the variadic matcher will allow to specialize them for the operation performed.
In particular, it will allow for a more efficient allOf() matcher.
Alexey Bataev [Thu, 20 Nov 2014 04:34:54 +0000 (04:34 +0000)]
[OPENMP] Codegen for "omp flush" directive.
For each "omp flush" directive a call to "void kmpc_flush(ident_t *, ...)" function is generated.
Directive "omp flush" may have an associated list of variables to flush, but currently runtime function ignores them. So the patch generates just "call kmpc_flush(ident_t *<loc>, i32 0)".
Differential Revision: http://reviews.llvm.org/D6292
Jordan Rose [Thu, 20 Nov 2014 03:57:45 +0000 (03:57 +0000)]
[CMake] Use .svn/wc.db as our VC dependency if it's available.
In order to keep SVNVersion.inc from being regenerated on every build, we need
to specify as a dependency a file that only changes when the VC state updates.
I previously just had .svn/entries, but that's only used by Subversion 1.6.
1.7 uses .svn/wc.db instead; prefer that if it's present.