Patch by chris.wailes@gmail.com. The following functionality was added:
* The same functionality is now supported for both CXXOperatorCallExprs and CXXMemberCallExprs.
* Factored out some code in StmtVisitor.
* Removed variables from the state map when their destructors are encountered.
* Started adding documentation for the consumed analysis attributes.
llvm-svn: 189059
DFSan changes the ABI of each function in the module. This makes it possible
for a function with the native ABI to be called with the instrumented ABI,
or vice versa, thus possibly invoking undefined behavior. A simple way
of statically detecting instances of this problem is to prepend the prefix
"dfs$" to the name of each instrumented-ABI function.
This will not catch every such problem; in particular function pointers passed
across the instrumented-native barrier cannot be used on the other side.
These problems could potentially be caught dynamically.
Differential Revision: http://llvm-reviews.chandlerc.com/D1373
llvm-svn: 189052
Summary:
These typically come from static data members of class template
specializations. This accomplishes two things:
1. May expose GlobalOpt optimizations for Itanium C++ ABI code.
2. Works toward fixing double initialization in the Microsoft C++ ABI.
CC: cfe-commits
Differential Revision: http://llvm-reviews.chandlerc.com/D1475
llvm-svn: 189051
Test-time dependencies now live within test/clang-replace/Inputs which is more
in line with llvm and clang test suites.
Added 'Inputs' to the lit config's 'exclude' list as with llvm's lit.cfg since
test-time dependencies should not have any lit scripts within.
llvm-svn: 189047
A single metadata will not span multiple lines. This also helps me with
my script to automatic update the testing cases.
A debug info testing case should have a llvm.dbg.cu.
Do not use hard-coded id for debug nodes.
llvm-svn: 189033
Summary:
Refactor VariantMatcher to use an interface underneath.
It supports "Single" and "Polymorphic". Will support more in the future.
Reviewers: klimek
CC: cfe-commits, revane
Differential Revision: http://llvm-reviews.chandlerc.com/D1446
llvm-svn: 189032
Thanks for pointing this out, Stephen. I think this is right now -- I
attempted to try all four valid combinations with both the autoconf and
CMake builds.
See also LLVM changes to the configure script.
llvm-svn: 189027
I don't actually have a version of autoconf so I edited configure directly
as well. It's copy-pasted so I think there was little margin for error.
See also Clang-side dependency graph changes.
llvm-svn: 189026
This might be able to be optimized further by only doing this in the
absence of a key function, but it doesn't look like GCC is doing that so
I'm not rushing to do it just yet.
llvm-svn: 189022
Subdirectories of test/clang-replace contain test-time dependencies and not
more LIT tests. Tell LIT to ignore these directories from the test suite.
llvm-svn: 189019
With this patch, braced lists (with more than 3 elements are formatted in a
column layout if possible). E.g.:
static const uint16_t CallerSavedRegs64Bit[] = {
X86::RAX, X86::RDX, X86::RCX, X86::RSI, X86::RDI,
X86::R8, X86::R9, X86::R10, X86::R11, 0
};
Required other changes:
- FormatTokens can now have a special role that contains extra data and can do
special formattings. A comma separated list is currently the only
implementation.
- Move penalty calculation entirely into ContinuationIndenter (there was a last
piece still in UnwrappedLineFormatter).
Review: http://llvm-reviews.chandlerc.com/D1457
llvm-svn: 189018
The IncludeDirectives class helps with detecting and modifying #include
directives. For now it allows the users to add angled-includes in a source file.
This is a start for this class that will evolve in the future to add more
functionality.
This should fix the reverted commit r188791 (buildbot failures on Windows).
llvm-svn: 189017
Functionality for clang-replace completed with the addition of the ability to
write merged replacements to disk.
Test added.
Differential Revision: http://llvm-reviews.chandlerc.com/D1460
llvm-svn: 189014
The instruction to convert between floating point and fixed point representations
takes an immediate operand for the number of fractional bits of the fixed point
value. ARMARM specifies that when that number of bits is zero, the assembler
should encode floating point/integer conversion instructions.
This patch adds the necessary instruction aliases to achieve this behaviour.
llvm-svn: 189009
Introducing new tool 'clang-replace' that finds files containing
serialized Replacements and applies those changes after deduplication
and detecting conflicts.
Currently the tool does not apply changes. It stops just after the
deduplication and conflict report phase. Forthcoming patches will
complete functionality.
Both build systems updated for new tool.
Includes a conflict test case.
clang-replace added to Doxygen build.
Differential Revision: http://llvm-reviews.chandlerc.com/D1424
llvm-svn: 189008
using GEPs. Previously, it used a number of different heuristics for
analyzing the GEPs. Several of these were conservatively correct, but
failed to fall back to SCEV even when SCEV might have given a reasonable
answer. One was simply incorrect in how it was formulated.
There was good code already to recursively evaluate the constant offsets
in GEPs, look through pointer casts, etc. I gathered this into a form
code like the SLP code can use in a previous commit, which allows all of
this code to become quite simple.
There is some performance (compile time) concern here at first glance as
we're directly attempting to walk both pointers constant GEP chains.
However, a couple of thoughts:
1) The very common cases where there is a dynamic pointer, and a second
pointer at a constant offset (usually a stride) from it, this code
will actually not do any unnecessary work.
2) InstCombine and other passes work very hard to collapse constant
GEPs, so it will be rare that we iterate here for a long time.
That said, if there remain performance problems here, there are some
obvious things that can improve the situation immensely. Doing
a vectorizer-pass-wide memoizer for each individual layer of pointer
values, their base values, and the constant offset is likely to be able
to completely remove redundant work and strictly limit the scaling of
the work to scrape these GEPs. Since this optimization was not done on
the prior version (which would still benefit from it), I've not done it
here. But if folks have benchmarks that slow down it should be straight
forward for them to add.
I've added a test case, but I'm not really confident of the amount of
testing done for different access patterns, strides, and pointer
manipulation.
llvm-svn: 189007