Dan Albert [Tue, 15 Aug 2017 22:10:44 +0000 (22:10 +0000)]
Add a target for new LTO plugin name tests.
Not all targets will use -plugin with -flto. Pick a fixed target so
this works regardless of the default target (regardless of host OS,
the toolchain should be picking the correct LTO plugin for a target
that supports it).
John McCall [Tue, 15 Aug 2017 21:42:52 +0000 (21:42 +0000)]
Extract IRGen's constant-emitter into its own helper class and clean up
the interface.
The ultimate goal here is to make it easier to do some more interesting
things in constant emission, like emit constant initializers that have
ignorable side-effects, or doing the majority of an initialization
in-place and then patching up the last few things with calls. But for
now this is mostly just a refactoring.
Dan Albert [Tue, 15 Aug 2017 21:31:17 +0000 (21:31 +0000)]
Revert "Revert "Fix LLVMgold plugin name/path for non-Linux.""
Summary:
Relanding https://reviews.llvm.org/D35739 which was reverted because
it broke the tests on non-Linux. The tests have been fixed to be
platform agnostic, and additional tests have been added to make sure
that the plugin has the correct extension on each platform
(%pluginext doesn't work in CHECK lines).
Stephen Hines [Tue, 15 Aug 2017 21:20:41 +0000 (21:20 +0000)]
Switch to consumeError(), since this can crash otherwise.
Summary:
If assertions are disabled, but LLVM_ABI_BREAKING_CHANGES is enabled,
this will cause an issue with an unchecked Success. Switching to
consumeError() is the correct way to bypass the check.
Kuba Mracek [Tue, 15 Aug 2017 19:47:06 +0000 (19:47 +0000)]
[clang] Get rid of "%T" expansions
The %T lit expansion expands to a common directory shared between all the tests in the same directory, which is unexpected and unintuitive, and more importantly, it's been a source of subtle race conditions and flaky tests. In https://reviews.llvm.org/D35396, it was agreed that it would be best to simply ban %T and only keep %t, which is unique to each test. When a test needs a temporary directory, it can just create one using mkdir %t.
[index] Update indexing to handle CXXDeductionGuideDecls properly
CXXDeductionGuideDecls can't be referenced so there's no need to output a symbol occurrence for them.
Also handle DeducedTemplateSpecializationTypeLocs in the TypeIndexer so we don't miss the symbol occurrences of the corresponding template decls.
Alexey Bataev [Tue, 15 Aug 2017 14:34:04 +0000 (14:34 +0000)]
[OPENMP] Fix compiler crash on argument translate for NVPTX.
When translating arguments for NVPTX target it is not taken into account
that function may have variable number of arguments. Patch fixes this
problem.
Reid Kleckner [Tue, 15 Aug 2017 01:17:47 +0000 (01:17 +0000)]
Avoid PointerIntPair of constexpr EvalInfo structs
They are stack allocated, so their alignment is not to be trusted.
32-bit MSVC only guarantees 4 byte stack alignment, even though alignof
would tell you otherwise. I tried fixing this with __declspec align, but
that apparently upsets GCC. Hopefully this version will satisfy all
compilers.
This diff fixes a crash (triggered assert) on the newly added test case.
In the method Simplifier::VisitSymbolData we check the type of S and return
Loc/NonLoc accordingly.
Alexey Bataev [Mon, 14 Aug 2017 17:56:13 +0000 (17:56 +0000)]
[OPENMP] Fix for PR33922: New ident_t flags for
__kmpc_for_static_init().
OpenMP 5.0 will include OpenMP Tools interface that requires distinguishing different worksharing constructs.
Since the same entry point (__kmp_for_static_init(ident_t *loc,
kmp_int32 global_tid,........)) is called in case static
loop/sections/distribute it is suggested using 'flags' field of the
ident_t structure to pass the type of the construct.
Alex Lorenz [Mon, 14 Aug 2017 16:19:24 +0000 (16:19 +0000)]
[rename] Introduce symbol occurrences
Symbol occurrences store the results of local rename and will also be used for
the global, indexed rename results. Their kind is used to determine whether they
should be renamed automatically or not. They can be converted to a set of
AtomicChanges as well.
Martin Probst [Mon, 14 Aug 2017 16:09:08 +0000 (16:09 +0000)]
clang-format: [JS] wrap optional properties in type aliases.
Summary:
clang-format wraps object literal keys in an object literal if they are
marked as `TT_SelectorName`s and/or the colon is marked as
`TT_DictLiteral`. Previously, clang-format would accidentally work
because colons in type aliases were marked as `TT_DictLiteral`. r310367
fixed this to assing `TT_JsTypeColon`, which broke wrapping in certain
situations. However the root cause was that clang-format incorrectly
didn't skip questionmarks when detecting selector name.
This change fixes both locations to (1) assign `TT_SelectorName` and (2)
treat `TT_JsTypeColon` like `TT_DictLiteral`.
Martin Probst [Mon, 14 Aug 2017 16:08:16 +0000 (16:08 +0000)]
clang-format: [JS] do not insert whitespace in call positions.
Summary:
In JavaScript, may keywords can be used in method names and thus call sites:
foo.delete();
foo.instanceof();
clang-format would previously insert whitespace after the `instanceof`. This
change generically skips inserting whitespace between a keyword and a
parenthesis if preceded by a dot, i.e. in a callsite.
clang-format: Fix left pointer alignment after delctype/typeof
Change 272124* introduced a regression in spaceRequiredBetween for left aligned pointers to decltype and typeof expressions. This fix adds logic to fix this. The test added is based on a related test in determineStarAmpUsage. Also add test cases for the regression.
Richard Smith [Sat, 12 Aug 2017 01:46:03 +0000 (01:46 +0000)]
PR34163: Don't cache an incorrect key function for a class if queried between
the class becoming complete and its inline methods being parsed.
This replaces the hack of using the "late parsed template" flag to track member
functions with bodies we've not parsed yet; instead we now use the "will have
body" flag, which carries the desired implication that the function declaration
*is* a definition, and that we've just not parsed its body yet.
George Karpenkov [Fri, 11 Aug 2017 17:22:58 +0000 (17:22 +0000)]
Add -fsanitize=fuzzer-no-link flag to the driver.
The flag will perform instrumentation necessary to the fuzzing,
but will NOT link libLLVMFuzzer.a library.
Necessary when modifying CFLAGS for projects which may produce
executables as well as a fuzzable target.
Alex Lorenz [Fri, 11 Aug 2017 12:06:52 +0000 (12:06 +0000)]
[modules] Set the lexical DC for dummy tag decls that refer to hidden
declarations that are made visible after the dummy is parsed and ODR verified
Prior to this commit the
"(getContainingDC(DC) == CurContext && "The next DeclContext should be lexically contained in the current one."),"
assertion failure was triggered during semantic analysis of the dummy
tag declaration that was declared in another tag declaration because its
lexical context did not point to the outer tag decl.
This is causing failures when compiling clang with -O3
as one of the structures used by clang is passed by
value and uses the fastcc calling convention.
Richard Smith [Fri, 11 Aug 2017 02:04:19 +0000 (02:04 +0000)]
PR33489: A function-style cast to a deduced class template specialization type is type-dependent if it can't be resolved due to a type-dependent argument.
Akira Hatanaka [Fri, 11 Aug 2017 00:06:49 +0000 (00:06 +0000)]
[Sema][ObjC] Fix spurious -Wcast-qual warnings.
We do not meaningfully track object const-ness of Objective-C object
types. Silence the -Wcast-qual warning that is issued when casting to or
from Objective-C object types results in losing const qualification.
Bob Wilson [Thu, 10 Aug 2017 16:42:46 +0000 (16:42 +0000)]
Add a getName accessor for ModuleMacros.
Swift would like to be able to access the name of a ModuleMacro.
There was some discussion of this in
https://github.com/apple/swift-clang/pull/93, suggesting that it makes
sense to have this accessor in Clang.
Momchil Velikov [Thu, 10 Aug 2017 15:43:06 +0000 (15:43 +0000)]
Place implictly declared functions at block scope
Such implicitly declared functions behave as if the enclosing block
contained the declaration extern int name() (C90, 6.3.3.2 Function calls),
thus their names should have block scope (C90, 6.1.2.1 Scope of identifiers).
This patch fixes https://bugs.llvm.org/show_bug.cgi?id=33224
Simon Atanasyan [Thu, 10 Aug 2017 15:42:25 +0000 (15:42 +0000)]
[mips] Show warning in case of mixing -mlong-calls and -mabicalls options
While we do not support `-mshared / -mno-shared` properly, show warning
and ignore `-mlong-calls` option in case of implicitly or explicitly
provided `-mabicalls` option.
Simon Atanasyan [Thu, 10 Aug 2017 15:42:16 +0000 (15:42 +0000)]
[mips] Notify user that `-mabicalls` is ignored on non-PIC N64 ABI
The -mabicalls option does not make sense in the case of non position
independent code for the N64 ABI. After this change the driver shows a
warning that -mabicalls is ignored in that case.
[Modules] Prevent #import to reenter header if not building a module.
When non-modular headers are imported while not building a module but
in -fmodules mode, be conservative and preserve the default #import
semantic: do not reenter headers.
[sanitizer-coverage] Change cmp instrumentation to distinguish const operands
This implementation of SanitizerCoverage instrumentation inserts different
callbacks depending on constantness of operands:
1. If both operands are non-const, then a usual
__sanitizer_cov_trace_cmp[1248] call is inserted.
2. If exactly one operand is const, then a
__sanitizer_cov_trace_const_cmp[1248] call is inserted. The first
argument of the call is always the constant one.
3. If both operands are const, then no callback is inserted.
This separation comes useful in fuzzing when tasks like "find one operand
of the comparison in input arguments and replace it with the other one"
have to be done. The new instrumentation allows us to not waste time on
searching the constant operands in the input.
Nikolai Bozhenov [Thu, 10 Aug 2017 11:22:52 +0000 (11:22 +0000)]
[ValueTracking] Enabling ValueTracking patch by default (recommit). Part 1.
The original patch was an improvement to IR ValueTracking on non-negative
integers. It has been checked in to trunk (D18777, r284022). But was disabled by
default due to performance regressions.
Perf impact has improved. The patch would be enabled by default.
Alex Lorenz [Thu, 10 Aug 2017 10:34:46 +0000 (10:34 +0000)]
Revert r310489 and follow-up commits r310505, r310519, r310537 and r310549
Commit r310489 caused 'openmp-offload.c' test failures on Darwin and other
platforms:
http://lab.llvm.org:8080/green/job/clang-stage1-cmake-RA-incremental_check/39230/testReport/junit/Clang/Driver/openmp_offload_c/
The follow-up commits tried to fix the test, but the test is still failing.
Petr Hosek [Thu, 10 Aug 2017 04:16:38 +0000 (04:16 +0000)]
[Driver] Search compiler-rt paths in -print-file-name=
This makes it possible to print the name of compiler-rt libraries
by using simply clang -print-file-name=libclang_rt.${runtime}-${arch}.so
same as other libraries, without having to know the details of the
resource directory organization.
[clang-format] let PointerAlignment dictate spacing of function ref qualifiers
Summary: The original changes for ref qualifiers in rL272537 and rL272548 allowed function const+ref qualifier spacing to diverge from the spacing used for variables. It seems more consistent for `T const& x;` to match `void foo() const&;`.
clang-format: Fix bug with ENAS_DontAlign and empty lines
This fixes a bug in `ENAS_DontAlign` (introduced in D32733) where blank lines had an EscapedNewlineColumn of 0, causing a subtraction to overflow when converted back to unsigned and leading to runaway memory allocation.
Coby Tayree [Wed, 9 Aug 2017 21:50:22 +0000 (21:50 +0000)]
[X86][Asm] Allow negative immediate to appear before bracketed expression
Currently, only non-negative immediate is allowed prior to a brac expression (memory reference).
MASM / GAS does not have any problem cope with the left side of the real line, so we should be able to as well.
Matt Arsenault [Wed, 9 Aug 2017 21:44:58 +0000 (21:44 +0000)]
AMDGPU: Use direct struct returns and arguments
This is an improvement over always using byval for
structs.
This will use registers until ~16 are used, and then
switch back to byval. This needs more work, since I'm
not sure it ever really makes sense to use byval. If
the register limit is exceeded, the arguments still
end up passed on the stack, but with a different ABI.
It also may make sense to base this on number of
registers used for non-struct arguments, rather than
just arguments that appear first in the argument list.
Hans Wennborg [Wed, 9 Aug 2017 20:12:53 +0000 (20:12 +0000)]
Make -std=c++17 an alias of -std=c++1z
As suggested on PR33912.
Trying to keep this small to make it easy to merge to the 5.0 branch. We
can do a follow-up with more thorough renaming (diagnostic text,
options, ids, etc.) later.
(For C++14 this was done in r215982, and I think a smaller patch for the
3.5 branch:
http://lists.llvm.org/pipermail/cfe-commits/Week-of-Mon-20140818/113013.html)
[OpenMP] Enable executable lookup into driver directory.
Summary: Invoking the compiler inside a script causes the clang-offload-bundler executable to not be found. This patch enables the lookup for executables in the driver directory where the clang-offload-bundler resides.
Alexey Bataev [Wed, 9 Aug 2017 19:38:53 +0000 (19:38 +0000)]
[OPENMP] Emit non-debug version of outlined functions with original
name.
If the host code is compiled with the debug info, while the target
without, there is a problem that the compiler is unable to find the
debug wrapper. Patch fixes this problem by emitting special name for the
debug version of the code.
[OpenMP] Add flag for disabling the default generation of relocatable OpenMP target code for NVIDIA GPUs.
Summary: Previously we have added the "-c" flag which gets passed to PTXAS by default to generate relocatable OpenMP target code by default. This set of flags exposes control over this behaviour.
Erich Keane [Wed, 9 Aug 2017 15:27:36 +0000 (15:27 +0000)]
Fix broken getAttributeSpellingListIndex for pragma attributes
We noticed when implementing a new pragma that the TableGen-generated function
getAttributeSpellingListIndex() did not work for pragma attributes. It relies
on the values in the enum AttributeList::Syntax and a new value
AS_ContextSensitiveKeyword was added changing the value for AS_Pragma.
Apparently no tests failed since no pragmas currently make use of the
generated function.
To fix this we can move AS_Pragma back to the value that TableGen code expects.
Also to prevent changes in the enum from breaking that routine again I added
calls to getAttributeSpellingListIndex() in the unroll pragma code. That will
cause some lit test failures if the order is changed. I added a comment to
remind of this issue in the future.
This assumes we don’t need/want full TableGen support for
AS_ContextSensitiveKeyword. It currently only appears in getAttrKind and no
other TableGen-generated routines.
projects / clang / log