llvm-svn: 177716

llvm-svn: 177715

llvm-svn: 177714

llvm-svn: 177713

InstCombine: Improve the result bitvect type when folding (cmp pred (load (gep GV, i)) C) to a bit test.

The original code used i32, and i64 if legal. This introduced unneeded
casts when they aren't legal, or when the index variable i has another
type. In order of preference: try to use i's type; use the smallest
fitting legal type (using an added DataLayout method); default to i32.
A testcase checks that this works when the index gep operand is i16.

Patch by : Ahmed Bougacha <ahmed.bougacha@gmail.com>
Reviewed by : Duncan

llvm-svn: 177712

llvm-svn: 177711

[ASan] Fix an error on invalid deallocation in ASan allocator. When ASan checks if memory freed by user was indeed previously allocated, it first does an atomic write to presumed location of chunk header. This is wrong, as if the free is invalid, we may overwrite some valuable data (like other fields of the chunk header). Fix this by using atomic_compare_exchange instead.

llvm-svn: 177710

llvm-svn: 177709

ScavengedRC was a dead private variable (set, but not otherwise used). No
functionality change intended.

llvm-svn: 177708

llvm-svn: 177707

More modules documentation, including the straw-man import declaration syntax and "how to modularize a platform".

llvm-svn: 177706

llvm-svn: 177705

llvm-svn: 177704

llvm-svn: 177703

llvm-svn: 177702

llvm-svn: 177701

the ARM build bots, and it adds a weird case to the test suite where
a test uses as inputs files in the parent directory.

Talked about this with Dave on IRC and he's fine with this approach even
though it isn't optimal.

llvm-svn: 177700

Introduce LLVM_STATIC_ASSERT macro, which expands to C/C++'s static_assert on compilers which support it.

llvm-svn: 177699

hierarchy-related at a possibly nonzero offset.

Patch by Alexander Zinenko!

llvm-svn: 177698

More cleanup to remove the CoreFoundation classes out of mainstream code (CFCMutableDictionary, CFCMutableArray, CFCString, etc). Now it is only used in the Results.cpp file for Apple builds only.

llvm-svn: 177697

- Added new abtract Results class to keep CoreFoundation out of the tests. There are many subclasses for different settings:
    Results::Result::Dictionary
    Results::Result::Array
    Results::Result::Unsigned
    Results::Result::Double
    Results::Result::String
- Gauge<T> can now write themselves out via a templatized write to results function:
    template <class T>
    Results::ResultSP GetResult (const char *description, T value);
  
- There are four specializations of this so far:
    template <>
    Results::ResultSP GetResult (const char *description, double value);

    template <>
    Results::ResultSP GetResult (const char *description, uint64_t value);

    template <>
    Results::ResultSP GetResult (const char *description, std::string value);

    template <>
    Results::ResultSP GetResult (const char *description, MemoryStats value);
- Don't emit the virtual memory reading from the task info call as it really doesn't mean much as it includes way too much (shared cache + other stuff we don't have control over)
- Fixed other test cases to build correctly and use the new classes

llvm-svn: 177696

-time-ir-parsing flag

This breaks the layering of the Support library. We can't add an
implementation side to IRReader because it refers directly to entities
only accessible as part of the IR, AsmParser, and BitcodeReader
libraries. It can only be used in a context where all of those libraries
will be available.

We'll need to find some other way to get this functionality, and
hopefully solve the long-standing layering problem of IRReader.h...

llvm-svn: 177695

llvm-svn: 177694

Fix undefined behavior in syntax_option_type::operator~ and match_flag_type::operator./a.out  Found by UBSan

llvm-svn: 177693

picking up cleanups from earlier in the statement.  Also fix a
crash-on-invalid where a reference to an invalid decl from an
enclosing scope was causing an expression to fail to build, but
only *after* a cleanup was registered from that statement,
causing an assertion downstream.

The crash-on-valid is rdar://13459289.

llvm-svn: 177692

Change the AppleObjCTrampolineHandler to always run all threads when resolving the target of an ObjC method call.
Add a StopOthers method to AppleThreadPlanStepThroughObjCTrampoline, don't rely on the setting in the ThreadPlanToCallFunction, since that
gets pushed too late to determine which threads will continue.

<rdar://problem/13447638>

llvm-svn: 177691

value argument. If not, be sure we don't accidentally use a dynamic alloca.

llvm-svn: 177690

value argument.

llvm-svn: 177689

It won't have one and it isn't needed.

llvm-svn: 177688

For mips a branch an 18-bit signed offset (the 16-bit 
offset field shifted left 2 bits) is added to the 
address of the instruction following the branch 
(not the branch itself), in the branch delay slot, 
to form a PC-relative effective target address. 

Previously, the code generator did not perform the 
shift of the immediate branch offset which resulted 
in wrong instruction opcode. This patch fixes the issue.

Contributor: Vladimir Medic
llvm-svn: 177687

<rdar://problem/13479214> Make Clang's <stddef.h> robust against system headers defining size_t/ptrdiff_t/wchar_t.

Clang's <stddef.h> provides definitions for the C standard library
types size_t, ptrdiff_t, and wchar_t. However, the system's C standard
library headers tend to provide the same typedefs, and the two
generally avoid each other using the macros
_SIZE_T/_PTRDIFF_T/_WCHAR_T. With modules, however, we need to see
*all* of the places where these types are defined, so provide the
typedefs (ignoring the macros) when modules are enabled.

llvm-svn: 177686

This patch uses the generated instruction info tables to 
identify memory/load store instructions.
After successful matching and based on the operand type 
and size, it generates additional instructions to the output.

Contributor: Vladimir Medic
llvm-svn: 177685

Don't forget to link in the C++-specific parts of the ubsan runtime when using the Makefile build system on Darwin.

llvm-svn: 177684

As Jakob pointed out in his review of r177423, having a shared ZERO
register between the 32- and 64-bit register classes causes this
odd G8RC_NOX0_and_GPRC_NOR0 class to be created. As recommended,
this adds a ZERO8 register which differentiates the 32- and 64-bit
zeros.

No functionality change intended.

llvm-svn: 177683

To use this in conjunction with exuberant ctags to generate a single
combined tags file, run tblgen first and then
  $ ctags --append [...]

Since some identifiers have corresponding definitions in C++ code,
it can be useful (if using vim) to also use cscope, and
  :set cscopetagorder=1
so that
  :tag X
will preferentially select the tablegen symbol, while
  :cscope find g X
will always find the C++ symbol.

Patch by Kevin Schoedel!

(a couple small formatting changes courtesy of clang-format)

llvm-svn: 177682

How did this ever work?

Basically, if you have a function that's inlined into the caller, it may not
have any 'call' instructions, but any 'resume' instructions it may have should
still be forwarded to the outer (caller's) landing pad. This requires that all
of the 'landingpad' instructions in the callee have their clauses merged with
the caller's outer 'landingpad' instruction (hence the bit of ugly code in the
`forwardResume' method).

Testcase in a follow commit to the test-suite repository.

<rdar://problem/13360379> & PR15555

llvm-svn: 177680

Thanks to Jakob for isolating the underlying problem from the
test case in r177423. The original commit had introduced
asymmetric copy operations, but these turned out to be a work-around
to the real problem (the use of == instead of hasSubClassEq in PPCCTRLoops).

llvm-svn: 177679

PR14963

llvm-svn: 177678

Refactor the filename/directory information in DISubprogram to refer directly to the pair rather than the DIFile.

llvm-svn: 177677

Switching the DIFile field in DISubprogram to refer to the raw
filename/directory pair instead of a DIFile.

llvm-svn: 177676