predictable diagnostic experience. The hash-of-DeclID order we were using
before gave different results on Win32 due to a different predefined
declaration of __builtin_va_list.
llvm-svn: 246521
GCC 4.8+ has a PowerPC-specific intrinsic, __builtin_ppc_get_timebase, to do
what Clang's __builtin_readcyclecounter does. For compatibility with code that
uses GCC's spelling (including glibc), support it as well.
Partially fixes PR23681.
llvm-svn: 246510
walking the loaded ModuleFiles looking for lookup tables for the context, store
them all in one place, and merge them together if we find we have too many
(currently, more than 4). If we do merge, include the merged form in our
serialized lookup table, so that downstream readers never need to look at our
imports' tables.
This gives a huge performance improvement to builds with very large numbers of
modules (in some cases, more than a 2x speedup was observed).
llvm-svn: 246497
This patch refactors the code to use the GCC installation detector
(modified so that it works in Solaris), and uses
ToolChain::GetFilePath everywhere once it works.
Patch by Xan López <xan@igalia.com>!
llvm-svn: 246473
A class without a name for linkage purposes gets a name along the lines
of <unnamed-type-foo> where foo is either the name of a declarator which
defined it (like a variable or field) or a
typedef-name (like a typedef or alias-declaration).
We handled the declarator case correctly but it would fall down during
template instantiation if the declarator didn't share the tag's type.
We failed to handle the typedef-name case at all.
Instead, keep track of the association between the two and keep it up to
date in the face of template instantiation.
llvm-svn: 246469
Also:
- Add a typedef to make working with the result easier.
- Update callers to use the new function.
- Make initFeatureMap out of line.
llvm-svn: 246468
Proper diagnostic and resolution of mangled names conflicts between C++ methods
and C functions. This patch implements support for functions/methods only;
support for variables is coming separately.
Differential Revision: http://reviews.llvm.org/D11297
llvm-svn: 246438
Added codegen for array section in 'depend' clause of 'task' directive. It emits to pointers, one for the begin of array section and another for the end of array section. Size of the section is calculated as (end + 1 - start) * sizeof(basic_element_type).
llvm-svn: 246422
This replaces the filtered generic iterator with a type-specfic one based
on dyn_cast instead of comparing the kind enum. This allows us to use
range-based for loops and eliminates casts. No functionality change
intended.
llvm-svn: 246384
const char pointers. In turn, push this through Clang APIs as well,
simplifying a number of bits of code that was handling the oddities of
nullptrs.
llvm-svn: 246375
We cannot tell if an object is past-the-end if its type is incomplete.
Zero sized objects satisfy past-the-end criteria and our object might
turn out to be such an object.
This fixes PR24622.
llvm-svn: 246359
Change the analyzer's modeling of memcpy to be more precise when copying into fixed-size
array fields. With this change, instead of invalidating the entire containing region the
analyzer now invalidates only offsets for the array itself when it can show that the
memcpy stays within the bounds of the array.
This addresses false positive memory leak warnings of the kind reported by
krzysztof in https://llvm.org/bugs/show_bug.cgi?id=22954
A patch by Pierre Gousseau!
Differential Revision: http://reviews.llvm.org/D11832
llvm-svn: 246345
This patch does two things:
1) Don't error about dllimport/export on thread-local static local variables.
We put those attributes on static locals in dllimport/export functions
implicitly in case the function gets inlined. Now, for TLS variables this
is a problem because we can't import such variables, but it's a benign
problem becase:
2) Make sure we never inline a dllimport function TLS static locals. In fact,
never inline a dllimport function that references a non-imported function
or variable (because these are not defined in the importing library). This
seems to match MSVC's behaviour.
Differential Revision: http://reviews.llvm.org/D12422
llvm-svn: 246338
Added codegen for array section in 'depend' clause of 'task' directive. It emits to pointers, one for the begin of array section and another for the end of array section. Size of the section is calculated as (end + 1 - start) * sizeof(basic_element_type).
llvm-svn: 246278
Sometimes, a macro that expands to another macro name will not be printed in
the macro backtrace. This patch finds the missed macro expansions and prints
them. Fixes PR16799
llvm-svn: 246237
Without this, 64-byte vector types (__m512), specified to be 64-byte
aligned in the AVX512 draft SysV ABI, will only be 32-byte aligned.
This is analoguous to AVX, for which we accept 32-byte max alignment.
Differential Revision: http://reviews.llvm.org/D10724
llvm-svn: 246230
There's no point in using a larger alignment if we have no instructions
that would benefit from it.
Differential Revision: http://reviews.llvm.org/D12389
llvm-svn: 246229
The ABI string only exists to communicate with TargetCodeGenInfo.
Concretely, since we only used "avx*" ABI strings on x86_64 (as AVX
doesn't affect the i386 ABIs), this meant that, when initializing
SimdDefaultAlign, we would ignore AVX/AVX512 on i386, for no good
reason.
Instead, directly check the features. A similar change for
MaxVectorAlign will follow.
Differential Revision: http://reviews.llvm.org/D12390
llvm-svn: 246228
There was linker problem, and it turns out that it is not always safe
to refer to vtable. If the vtable is used, then we can refer to it
without any problem, but because we don't know when it will be used or
not, we can only check if vtable is external or it is safe to to emit it
speculativly (when class it doesn't have any inline virtual functions).
It should be fixed in the future.
http://reviews.llvm.org/D12385
llvm-svn: 246214
Usually debug info is created on the fly while during codegen.
With this API it becomes possible to create standalone debug info
for types that are not referenced by any code, such as emitting debug info
for a clang module or for implementing something like -gfull.
Because on-the-fly debug info generation may still insert retained types
on top of them, all RetainedTypes are uniqued in CGDebugInfo::finalize().
llvm-svn: 246210
A couple of changes here:
a) Do less work in the case where we don't have a target attribute on the
function. We've already canonicalized the attributes for the function -
no need to do more work.
b) Use the newer canonicalized feature adding functions from TargetInfo
to do the work when we do have a target attribute. This enables us to diagnose
some warnings in the case of conflicting written attributes (only ppc does
this today) and also make sure to get all of the features for a cpu that's
listed rather than just change the cpu.
Updated all testcases accordingly and added a new testcase to verify that we'll
error out on ppc if we have some incompatible options using the existing diagnosis
framework there.
llvm-svn: 246195
to enable the use of external type references in the debug info
(a.k.a. module debugging).
The driver expands -gmodules to "-g -fmodule-format=obj -dwarf-ext-refs"
and passes that to cc1. All this does at the moment is set a flag
codegenopts.
http://reviews.llvm.org/D11958
llvm-svn: 246192
Current implementation tries to guess which Action will result in a
job which needs to incorporate device-side GPU binaries. The guessing
was attempting to work around the fact that multiple actions may be
combined into a single compiler invocation. If CudaHostAction ends up
being combined (and thus bypassed during action list traversal) no
device-side actions it pointed to were processed. The guessing worked
for most of the usual cases, but fell apart when external assembler
was used.
This change removes the guessing and makes sure we create and pass
device-side jobs regardless of how the jobs get combined.
* CudaHostAction is always inserted either at Compile phase or the
FinalPhase of current compilation, whichever happens first.
* If selectToolForJob combines CudaHostAction with other actions, it
passes info about CudaHostAction up to the caller
* When it sees that CudaHostAction got combined with other actions
(and hence will never be passed to BuildJobsForActions),
BuildJobsForActions creates device-side jobs the same way they would
be created if CudaHostAction was passed to BuildJobsForActions
directly.
* Added two more test cases to make sure GPU binaries are passed to
correct jobs.
Differential Revision: http://reviews.llvm.org/D11280
llvm-svn: 246174
with multiple uses of feature map construction.
Note: We could make this a static function on TargetInfo if we
fix the x86 port needing to check the triple in an isolated case.
llvm-svn: 246128
DeclarationName (because all ctor names are considered the same, and so on).
Reflect this in the type used as the lookup table key. As a side-effect, remove
one copy of the duplicated code used to compute the hash of the key.
llvm-svn: 246124
I stared at `false /*declaration*/` for quite some time before giving up
and checking the actual function to see what it meant. Replacing with
`/* isDefinition = */ false` to save myself effort later.
llvm-svn: 246095
Error out if -mfloat-abi=hard or -mhard-float is specified on the command
line and the target ABI is APCS. Previously clang issued no warnings or
errors and just passed the option to the backend, which had no effect on
code generation for targets using APCS.
This commit corrects the patch commited in r245866, which didn't take into
account the fact that not all darwin targets use APCS.
rdar://problem/22257950
http://reviews.llvm.org/D12344
llvm-svn: 246054
and CompilerInvocation::getFileSystemOpts by renaming it to getFileSystemOpts,
marking the const-returning access method const and adding a non-const version,
making the function prototypes identical to CompilerInstance::getFileSystemOpts.
llvm-svn: 246026
We tried to provide a very nice diagnostic when diagnosing an assignment
to a const int & produced by a function call. However, we cannot always
determine what function was called.
This fixes PR24568.
llvm-svn: 246014
Summary:
- Store the exception specification range's begin and end SourceLocation in DeclaratorChuck::FunctionTypeInfo. These SourceLocations can be used in a FixItHint Range.
- Add diagnostic; function concept having an exception specification.
Reviewers: hubert.reinterpretcast, fraggamuffin, faisalv, aaron.ballman, rsmith
Subscribers: cfe-commits
Differential Revision: http://reviews.llvm.org/D11789
llvm-svn: 246005
As discussed in PR23648 - the intrinsics _m_from_int, _m_to_int and _m_prefetch are defined in mmintrin.h and prfchwintrin.h so we don't need to in Intrin.h
Added tests for _m_from_int and _m_to_int
D11338 already added a test for _m_prefetch
Differential Revision: http://reviews.llvm.org/D12272
llvm-svn: 245975
GNU attributes can have a leading and trailing __ appended/prepended to
the attribute name. While the parser and AttributeList::getKind did the
right thing, AttributeList::getAttributeSpellingListIndex did not.
This fixes PR24565.
llvm-svn: 245953
Eventually, we will need sample profiles to be incorporated into the
inliner's cost models. To do this, we need the sample profile pass to
be a module pass.
This patch makes no functional changes beyond the mechanical adjustments
needed to run SampleProfile as a module pass.
llvm-svn: 245941
Adds parsing/sema analysis/serialization/deserialization for array sections in OpenMP constructs (introduced in OpenMP 4.0).
Currently it is allowed to use array sections only in OpenMP clauses that accepts list of expressions.
Differential Revision: http://reviews.llvm.org/D10732
llvm-svn: 245937
This involved specializing handleUserFeatures so that we could perform
diagnostics on -only- user supplied features and migrating the rest of
the initialization functions to set features based on enabling and disabling
full feature sets. No functional change intended.
llvm-svn: 245936
_rotl, _rotwl and _lrotl (and their right-shift counterparts) are official x86
intrinsics, and should be supported regardless of environment. This is in contrast
to _rotl8, _rotl16, and _rotl64 which are MS-specific.
Note that the MS documentation for _lrotl is different from the Intel
documentation. Intel explicitly documents it as a 64-bit rotate, while for MS,
since sizeof(unsigned long) for MSVC is always 4, a 32-bit rotate is implied.
Differential Revision: http://reviews.llvm.org/D12271
llvm-svn: 245923
Summary:
This is more consistent with other targets and also makes the -fuse-ld
flag work.
Reviewers: jvoung
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D10697
llvm-svn: 245908
Error out if the user tries to use float-abi="hard" since it isn't
supported on darwin platforms. Previously clang issued no warnings or
erros and just passed the option to the backend, which had no effect on
code generation for targets using apcs.
rdar://problem/22257950
Differential Revision: http://reviews.llvm.org/D12155
llvm-svn: 245866
Before:
DEPRECATED("Use NewClass::NewFunction instead.") int OldFunction(
const string ¶meter) {}
Could not be formatted at all, as clang-format would both require and
disallow the break before "int".
llvm-svn: 245846
The ACLE (ARM C Language Extensions) 2.0 defines that the predefined macro
__ARM_FP16_ARGS should be defined if __fp16 can be used as an argument and
result.
The support for __fp16 to be used as an argument and result is already
implemented for AArch64 so this change is just adding the missing macro.
Differential Revision: http://reviews.llvm.org/D12240
llvm-svn: 245833
If a function declaration is found inside a template function as in:
template<class T> void f() {
void g(int x = T::v) except(T::w);
}
it must be instantiated along with the enclosing template function,
including default arguments and exception specification.
Together with the patch committed in r240974 this implements DR1484.
Differential Revision: http://reviews.llvm.org/D11194
llvm-svn: 245810
Summary:
The signatures of the methods in LLVM for creating EH pads/rets are changing
to require token arguments on rets and assume token return type on pads.
Update creation code accordingly.
Reviewers: majnemer, rnk
Subscribers: cfe-commits
Differential Revision: http://reviews.llvm.org/D12109
llvm-svn: 245798
pick the most recent declaration, and we can often tell which declaration is
more recent without walking the redeclaration chain. Do so when possible.
llvm-svn: 245794
Instead of eagerly deserializing a list of DeclIDs when we load a module file
and doing a binary search to find the redeclarations of a decl, store a list of
redeclarations of each chain before the first declaration and load it directly.
llvm-svn: 245789
Summary:
According to CUDA documentation, global variables declared with __device__,
__constant__ can be initialized from host code, so mark them as
externally initialized. Because __shared__ variables cannot have an
initialization as part of their declaration and since the value maybe kept
across different kernel invocation, the value of __shared__ is effectively
undefined instead of zero initialized.
Wrongly using zero initializer may cause illegitimate optimization, e.g.
removing unused __constant__ variable because it's not updated in the device
code and the value is initialized with zero.
Test Plan: test/CodeGenCUDA/address-spaces.cu
Patch by Xuetian Weng
Reviewers: jholewinski, eliben, tra, jingyue
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D12241
llvm-svn: 245786
all modules and reduce the number of declarations we load when loading a
redeclaration chain.
The new approach is:
* when loading the first declaration of an entity within a module file, we
first load all declarations of the entity that were imported into that
module file, and then load all the other declarations of that entity from
that module file and build a suitable decl chain from them
* when loading any other declaration of an entity, we first load the first
declaration from the same module file
As before, we complete redecl chains through name lookup where necessary.
To make this work, I also had to change the way that template specializations
are stored -- it no longer suffices to track only canonical specializations; we
now emit all "first local" declarations when emitting a list of specializations
for a template.
On one testcase with several thousand imported module files, this reduces the
total runtime by 72%.
llvm-svn: 245779
a contextually-typed expression that semantic analysis will
probably need to invasively rewrite, don't include the
RHS OVE as a separate semantic expression, and check the
operation with the original RHS expression.
There are two contextually-typed expressions that can survive
to here: overloaded function references, which are at least
safe to double-emit, and C++11 initializer list expressions,
which are not at all safe to double-emit and which often
don't update the original syntactic InitListExpr with
implicit conversions to member types, etc.
This means that the original RHS may appear, undecorated by
an OVE, in the semantic expressions. Fortunately, it will
only ever be used in a single place there, and I don't
believe there are clients that rely on being able to pick
out the original RHS from the semantic expressions.
But this could be problematic if there are clients that do
visit the entire tree and rely on not seeing the same
expression multiple times, once in the syntactic and once
in the semantic expressions. This is a very fiddly part
of the compiler.
rdar://21801088
llvm-svn: 245771
We had "vcvt_f16" and "VCVT_HIGH_F16": for other FP types, this naming
is used for intrinsics with integer overloads. The FP->FP conversions,
on the other hand, use the full "vcvt_f32_f64" name instead.
Use the same naming convention for the f16<->f32 conversions.
While there, reorder the definitions a little bit.
llvm-svn: 245763
This is important in the case that the LLVM-inferred llvm-struct
alignment is not the same as the clang-known C-struct alignment.
Differential Revision: http://reviews.llvm.org/D12243
llvm-svn: 245719
Add emission of metadata for simd loops in presence of 'simdlen' clause.
If 'simdlen' clause is provided without 'safelen' clause, the vectorizer width for the loop is set to value of 'simdlen' clause + all read/write ops in loop are marked with '!llvm.mem.parallel_loop_access' metadata.
If 'simdlen' clause is provided along with 'safelen' clause, the vectorizer width for the loop is set to value of 'simdlen' clause + all read/write ops in loop are not marked with '!llvm.mem.parallel_loop_access' metadata.
If 'safelen' clause is provided without 'simdlen' clause, the vectorizer width for the loop is set to value of 'safelen' clause + all read/write ops in loop are not marked with '!llvm.mem.parallel_loop_access' metadata.
llvm-svn: 245697
This is a bit of a step back of what we did in r222531, as there are
some corner cases in C++, where this kind of formatting is really bad.
Example:
Before:
virtual aaaaaaaaaaaaaaaa(std::function<bool()> IsKindWeWant = [&]() {
return true;
}, aaaaa aaaaaaaaa);
After:
virtual aaaaaaaaaaaaaaaa(std::function<bool()> IsKindWeWant =
[&]() { return true; },
aaaaa aaaaaaaaa);
The block formatting logic in JavaScript will probably go some other changes,
too, and we'll potentially be able to make the rules more consistent again. For
now, this seems to be the best approach for C++.
llvm-svn: 245694
Add parsing/sema analysis for 'simdlen' clause in simd directives. Also add check that if both 'safelen' and 'simdlen' clauses are specified, the value of 'simdlen' parameter is less than the value of 'safelen' parameter.
llvm-svn: 245692
While working around a bug in certain standard library implementations,
we would try to diagnose the issue so that library implementors would
fix their code. However, we assumed an entity being initialized was
a non-static data member subobject when other circumstances are
possible.
This fixes PR24526.
llvm-svn: 245675
According to standard the 'uval' modifier declares the address of the original list item to have an invariant value for all iterations of the associated loop(s). Patch improves codegen for this qualifier by removing usage of the original reference variable and replacing by referenced l-value.
llvm-svn: 245674
For some reason, clang had been treating a command like:
clang -static -fPIC foo.c
as if it should be compiled without the PIC relocation model.
This was incorrect: -static should be affecting only the linking
model, and -fPIC only the compilation.
This new behavior also matches GCC.
This is a follow-up from a review comment on r245447.
Differential Revision: http://reviews.llvm.org/D12208
llvm-svn: 245667
Now that -Winfinite-recursion no longer uses recursive calls to before path
analysis, several bits of the code can be improved. The main changes:
1) Early return when finding a path to the exit block without a recursive call
2) Moving the states vector into checkForRecursiveFunctionCall instead of
passing it in by reference
3) Change checkForRecursiveFunctionCall to return a bool when the warning
should be emitted.
4) Use the State vector instead of storing it in the Stack vector.
llvm-svn: 245666
This lets us optimize them better. We agreed to remove the intrinsics,
instead of combining them later, as, at -O0, we generate the expected
instructions. Plus, it's a nice cleanup.
Differential Revision: http://reviews.llvm.org/D10556
llvm-svn: 245605
The problem is that the arguments are of TheCall are reset later
to the ones in Args, making TypoExpr put back. Some TypoExpr that have
already been diagnosed and will assert later in Sema::getTypoExprState
llvm-svn: 245560
Standard allows to use 'uval' and 'ref' modifiers in 'linear' clause for variables with reference types only. Added check for it and modified test.
llvm-svn: 245556
OpenMP 4.1 adds 3 optional modifiers to 'linear' clause.
Format of 'linear' clause has changed to:
```
linear(linear-list[ : linear-step])
```
where linear-list is one of the following
```
list
modifier(list)
```
where modifier is one of the following:
```
ref (C++)
val (C/C++)
uval (C++)
```
Patch adds parsing and sema analysis for these modifiers.
llvm-svn: 245550
alignment is ignored, and they always allocate a complete
storage unit.
Also, change the dumping of AST record layouts: use the more
readable C++-style dumping even in C, include bitfield offset
information in the dump, and don't print sizeof/alignof
information for fields of record type, since we don't do so
for bases or other kinds of field.
rdar://22275433
llvm-svn: 245514
Remove the assumption of a Boolean type by checking if an expression is known
to have a boolean value. Disable warning in two other tests.
llvm-svn: 245507
doing assembly-only, and unify the Driver's PIC argument parsing.
On a few architectures, parsing of assembly files annoyingly depends
on whether PIC is enabled or not. This was handled for external 'as'
already (passing -KPIC), but was missed for calls to the standalone
internal assembler.
The integrated-as.s test needed to be modified to not expect
-fsanitize=address to be unused, as now fsanitize *IS* used for
assembly, since -fsanitize=memory can sometimes imply -fPIE, which the
assembler needs to know (gack!!).
Differential Revision: http://reviews.llvm.org/D11845
llvm-svn: 245447
"generic" cpu was wrongly handled as exact real CPU name of ARMv8.1A architecture.
This has been fixed, now it is abstract name, suitable for any arch.
Reviewers: rengolin
Subscribers: cfe-commits
Differential Revision: http://reviews.llvm.org/D11640
llvm-svn: 245445
Adds libomp.lib for -fopenmp=libomp and libiomp5md.lib for -fopenmp=libiomp5 on Windows
Differential Revision: http://reviews.llvm.org/D11932
llvm-svn: 245414
This has been disabled for a long time, but:
1) Initializers work (and apparently they're re reason why this was disabled).
2) various tests happen to hit this code path and the invariant seems to be
always verified.
Differential Revision: http://reviews.llvm.org/D12110
Reviewed by: rsmith
llvm-svn: 245404
__builtin_object_size would return incorrect answers for many uses where
type=3. This fixes the inaccuracy by making us emit 0 instead of LLVM's
objectsize intrinsic.
Additionally, there are many cases where we would emit suboptimal (but
correct) answers, such as when arrays are involved. This patch fixes
some of these cases (please see new tests in test/CodeGen/object-size.c
for specifics on which cases are improved)
Resubmit of r245323 with PR24493 fixed.
Patch mostly by Richard Smith.
Differential Revision: http://reviews.llvm.org/D12000
This fixes PR15212.
llvm-svn: 245403
Maybe this and the NumDeclsFound member should just be a std::vector
instead. (it could be a std::dynarray, but that missed standardization)
llvm-svn: 245392
__builtin_object_size would return incorrect answers for many uses where
type=3. This fixes the inaccuracy by making us emit 0 instead of LLVM's
objectsize intrinsic.
Additionally, there are many cases where we would emit suboptimal (but
correct) answers, such as when arrays are involved. This patch fixes
some of these cases (please see new tests in test/CodeGen/object-size.c
for specifics on which cases are improved)
Patch mostly by Richard Smith.
Differential Revision: http://reviews.llvm.org/D12000
This fixes PR15212.
llvm-svn: 245323
OpenMP 4.1 allows to use variables with reference types in all private clauses (private, firstprivate, lastprivate, linear etc.). Patch allows to use such variables and fixes codegen for linear variables with reference types.
llvm-svn: 245268
context is the class itself but lookups should be performed starting with the
lookup parent of the class (class and base members don't shadow types from the
surrounding context because they have not been declared yet).
llvm-svn: 245236
Summary:
If a module was unavailable (either a missing requirement on the module
being imported, or a missing file anywhere in the top-level module (and
not dominated by an unsatisfied `requires`)), we would silently treat
inclusions as textual. This would cause all manner of crazy and
confusing errors (and would also silently "work" sometimes, making the
problem difficult to track down).
I'm really not a fan of the `M->isAvailable(getLangOpts(), getTargetInfo(),
Requirement, MissingHeader)` function; it seems to do too many things at
once, but for now I've done things in a sort of awkward way.
The changes to test/Modules/Inputs/declare-use/module.map
were necessitated because the thing that was meant to be tested there
(introduced in r197805) was predicated on silently falling back to textual
inclusion, which we no longer do.
The changes to test/Modules/Inputs/macro-reexport/module.modulemap
are just an overlooked missing header that seems to have been missing since
this code was committed (r213922), which is now caught.
Reviewers: rsmith, benlangmuir, djasper
Subscribers: cfe-commits
Differential Revision: http://reviews.llvm.org/D10423
llvm-svn: 245228
Added an additional ctor that takes a NumOccurrenceFlag parameter for the
SourcePaths option. This frees applications from always having to pass at least
one source file, e.g., -list-checks.
http://reviews.llvm.org/D12069
Patch by Don Hinton!
llvm-svn: 245204
This enables Clang to correctly handle code such as:
struct __declspec(dllexport) S {
int x = 42;
};
where it would otherwise error due to trying to generate the default
constructor before the in-class initializer for x has been parsed.
Differential Revision: http://reviews.llvm.org/D11850
llvm-svn: 245139
Summary: Adding check to emit diagnostic for invalid tag when concept is specified and associated tests.
Reviewers: rsmith, hubert.reinterpretcast, fraggamuffin, faisalv, aaron.ballman
Subscribers: aaron.ballman, cfe-commits
Differential Revision: http://reviews.llvm.org/D11916
llvm-svn: 245123
Add checkers that detect code-level localizability issues for OS X / iOS:
- A path sensitive checker that warns about uses of non-localized
NSStrings passed to UI methods expecting localized strings.
- A syntax checker that warns against not including a comment in
NSLocalizedString macros.
A patch by Kulpreet Chilana!
(This is the second attempt with the compilation issue on Windows and
the random test failures resolved.)
llvm-svn: 245093
Summary:
MSDN says that fastcall, stdcall, thiscall, and vectorcall are all
accepted but ignored on ARM and X64.
https://msdn.microsoft.com/en-us/library/984x0h58.aspx
MSDN also says cdecl is also accepted and typically ignored
This patch brings ARM in line with how we ignore them for X64
Reviewers: rnk
Subscribers: compnerd, cfe-commits
Differential Revision: http://reviews.llvm.org/D12034
llvm-svn: 245076
So, we now reject that. We also warn for any external-linkage global
variable named main in C, because it results in undefined behavior.
PR: 24309
Differential Revision: http://reviews.llvm.org/D11658
Reviewed by: rsmith
llvm-svn: 245051
blender uses statements expression in condition of the loop under control of the '#pragma omp parallel for'. This condition is used several times in different expressions required for codegen of the loop directive. If there are some variables defined in statement expression, it fires an assert during codegen because of redefinition of the same variables.
We have to rebuild several expression to be sure that all variables are unique.
llvm-svn: 245041
Currently, arguments are passed via the string attribute 'command',
assuming a shell-escaped / quoted command line to extract the original
arguments. This works well enough on Unix systems, but turns out to be
problematic for Windows tools to generate.
This CL adds a new attribute 'arguments', an array of strings, which
specifies the exact command line arguments. If 'arguments' is available
in the compilation database, it is preferred to 'commands'.
Currently there is no plan to retire 'commands': there are enough
different use cases where users want to create their own mechanism for
creating compilation databases, that it doesn't make sense to force them
all to implement shell command line parsing.
Patch by Daniel Dilts.
llvm-svn: 245036
file in the .pcm files. This allows a smaller set of files to be sent to a
remote build worker when building with explicit modules (for instance, module
map files need not be sent along with the corresponding precompiled modules).
This doesn't actually make the embedded files visible to header search, so
it's not useful as a packaging format for public header files.
llvm-svn: 245028
ConsumedBlockInfo objects were move assigned, but only in a state where
the dtor was a no-op anyway. Subtle and easily could've happened in ways
that wouldn't've been safe - so this change makes it safe no matter what
state the ConsumedBlockInfo object is in.
llvm-svn: 244998
Make the copy/move ctors protected and defaulted in the base, make the
derived classes final to avoid exposing any slicing-prone APIs.
Also, while I'm here, simplify the use of buildByrefHelpers by taking
the parameter by value instead of non-const ref. None of the callers
care aobut observing the state after the call.
llvm-svn: 244990
We risk iterator invalidation issues if we use a DenseMap to hold the
backing storage for an APValue. Instead, BumpPtrAllocate them and
use APValue * as our DenseMap value.
Also, don't assume that MaterializedGlobalTemporaryMap won't regrow
between when we initially perform a lookup and later on when we actually
try to insert into it.
This fixes PR24289.
Differential Revision: http://reviews.llvm.org/D11629
llvm-svn: 244989
via a module map found by -fmodule-map-file=, the home directory of the module
is the current working directory, even if that's a different directory on
reload.
llvm-svn: 244988
Make the copy/move ctors defaulted in the base class and make the
derived classes final to avoid any intermediate hierarchy slicing if
these types were further derived.
llvm-svn: 244979
(return by value is in ExprEngine::processPointerEscapedOnBind and any
other call to the scanReachableSymbols function template used there)
Protect the special members in the base class to avoid slicing, and make
derived classes final so these special members don't accidentally become
public on an intermediate base which would open up the possibility of
slicing again.
llvm-svn: 244975
The user-defined copy assignment looks like it was working around the
presence of a reference member (that probably doesn't change in the copy
assignment cases present in the program). Rather than continuing this - just
change the reference to a pointer and let all the special members be
defined implicitly.
llvm-svn: 244974
Turns out the one place that relied on the implicit copy ctor was safe
because it created an object in a state where the dtor was a no-op, but
that's more luck that good management.
Sure up the API by defining move construction and using it, which
implicitly disallows the unreliable copy operations.
llvm-svn: 244968
Partly addressed by r244843, but the explicit dtor in LambdaScopeInfo
was still thwarting the implicit copy ctor. This does remove the key
function from LambdaScopeInfo unfortunately, but it seems neater than
having to explicitly default any special members LambdaScopeInfo needs.
llvm-svn: 244957
Summary: Poisoning applied to only class members, and before dtors for base class invoked
Implement poisoning of only class members in dtor, as opposed to also
poisoning fields inherited from base classes. Members are poisoned
only once, by the last dtor for a class. Skip poisoning if class has
no fields.
Verify emitted code for derived class with virtual destructor sanitizes
its members only once.
Removed patch file containing extraneous changes.
Reviewers: eugenis, kcc
Differential Revision: http://reviews.llvm.org/D11951
Simplified test cases for use-after-dtor
Summary: Simplified test cases to focus on one feature at time.
Tests updated to align with new emission order for sanitizing
callback.
Reviewers: eugenis, kcc
Differential Revision: http://reviews.llvm.org/D12003
llvm-svn: 244933
After r244870 flush() will only compare two null pointers and return,
doing nothing but wasting run time. The call is not required any more
as the stream and its SmallString are always in sync.
Thanks to David Blaikie for reviewing.
llvm-svn: 244928
files: include the .pcm file itself in the .d output, rather than including its
own input files. Other forms of module file continue to be transparent for .d
output.
Arguably, the input files for the .pcm file are still inputs to the
compilation, but that's unnecessary for make-like build systems (where the
mtime of the .pcm file is sufficient) and harmful for smarter build systems
that know about module files and want to track only the local dependencies.
llvm-svn: 244923
-fno-rtti-data makes it so that vtables emitted in the current TU lack
RTTI data. This means that dynamic_cast usually fails at runtime. Users
of the existing cxx_rtti feature expect all of RTTI to work, not just
some of it.
Chromium bug for context: http://crbug.com/518191
llvm-svn: 244922
This preserves backwards compatibility for two hacks in the Darwin
system module map files:
1. The use of 'requires excluded' to make headers non-modular, which
should really be mapped to 'textual' now that we have this feature.
2. Silently removes a bogus cplusplus requirement from IOKit.avc.
Once we start diagnosing missing requirements and headers on
auto-imports these would have broken compatibility with existing Darwin
SDKs.
llvm-svn: 244912
Summary:
When we want to use mingw-w64 and clang with compiler-rt we should not
need to have libgcc installed. This fixes finding includes when libgcc
is not installed
Reviewers: yaron.keren
Subscribers: cfe-commits
Differential Revision: http://reviews.llvm.org/D11808
llvm-svn: 244902
Summary:
Clang sanitizers, such as AddressSanitizer, ThreadSanitizer, MemorySanitizer,
Control Flow Integrity and others, use blacklists to specify which types / functions
should not be instrumented to avoid false positives or suppress known failures.
This change adds the blacklist filenames to the list of dependencies of the rules,
generated with -M/-MM/-MD/-MMD. This lets CMake/Ninja recognize that certain
C/C++/ObjC files need to be recompiled (if a blacklist is updated).
Reviewers: pcc
Subscribers: rsmith, honggyu.kim, pcc, cfe-commits
Differential Revision: http://reviews.llvm.org/D11968
llvm-svn: 244867
Verify emitted code for derived class with virtual destructor sanitizes its members only once.
Changed emission order for dtor callback, so only the last dtor for a class emits the sanitizing callback, while ensuring that class members are poisoned before base class destructors are invoked.
Skip poisoning of members, if class has no fields.
Removed patch file containing extraneous changes.
Summary: Poisoning applied to only class members, and before dtors for base class invoked
Reviewers: eugenis, kcc
Differential Revision: http://reviews.llvm.org/D11951
llvm-svn: 244819
When displaying the macro backtrace, ignore some of the backtraces that do not
provide extra information to the diagnostic. Typically, if the problem is
entirely contained within a macro argument, the macro expansion is often not
needed. Also take into account SourceRange's attached to the diagnostic when
selecting which backtraces to ignore. Two previous test cases have also been
updated.
Patch by Zhengkai Wu, with minor formatting fixes.
Differential Revision: http://reviews.llvm.org/D11778
llvm-svn: 244788
emit lexical contents for a declaration for another module. Track which module
those contents came from, and ensure that we only grab the lexical contents
from a single such instantiation.
llvm-svn: 244682
Summary:
float_cast_overflow is the only UBSan check without a source location attached.
This patch propagates SourceLocations where necessary to get them to the
EmitCheck() call.
Reviewers: rsmith, ABataev, rjmccall
Subscribers: cfe-commits
Differential Revision: http://reviews.llvm.org/D11757
llvm-svn: 244568
Our implementations of these type trait intrinsics simply mapped them to
__has_trivial_destructor. Instead, flesh these intrinsics out with a
full implementation which matches the standard's description.
llvm-svn: 244564
This patche and a related llvm patch solve the problem of having to explicitly enable analysis when specifying a loop hint pragma to get the diagnostics. Passing AlwasyPrint as the pass name (see below) causes the front-end to print the diagnostic if the user has specified '-Rpass-analysis' without an '=<target-pass>’. Users of loop hints can pass that compiler option without having to specify the pass and they will get diagnostics for only those loops with loop hints.
llvm-svn: 244556
Following one of the appended options will allow the loop to be vectorized. We do not include a command line option for modifying the pointer checking threshold because there is no clang-level interface for this currently.
llvm-svn: 244526
This allows emitting kernels that were instantiated from the host code
and which would never be explicitly referenced otherwise.
Differential Revision: http://reviews.llvm.org/D11666
llvm-svn: 244501
With this patch clang appends the command line options that would allow vectorization when floating-point commutativity is required. Specifically those are enabling fast-math or specifying a loop hint.
llvm-svn: 244492
Summary:
The vtable takes its DLL storage class from the class, not the key
function. When they disagree, the vtable won't be exported by the DLL
that defines the key function. The easiest way to ensure that importers
of the class emit their own vtable is to say that the class has no key
function.
Reviewers: hans, majnemer
Subscribers: cfe-commits
Differential Revision: http://reviews.llvm.org/D11913
llvm-svn: 244488
These changes are for Android x86_64 targets to be compatible
with current Android g++ and conform to AMD64 ABI.
https://llvm.org/bugs/show_bug.cgi?id=23897
* Return type of long double (fp128) should be fp128, not x86_fp80.
* Vararg of long double (fp128) could be in register and overflowed to memory.
https://llvm.org/bugs/show_bug.cgi?id=24111
* Return value of long double (fp128) _Complex should be in memory like a structure of {fp128,fp128}.
Differential Revision: http://reviews.llvm.org/D11437
llvm-svn: 244468
This change adds the new unroll metadata "llvm.loop.unroll.enable" which directs
the optimizer to unroll a loop fully if the trip count is known at compile time, and
unroll partially if the trip count is not known at compile time. This differs from
"llvm.loop.unroll.full" which explicitly does not unroll a loop if the trip count is not
known at compile time
With this change "#pragma unroll" generates "llvm.loop.unroll.enable" rather than
"llvm.loop.unroll.full" metadata. This changes the semantics of "#pragma unroll" slightly
to mean "unroll aggressively (fully or partially)" rather than "unroll fully or not at all".
The motivating example for this change was some internal code with a loop marked
with "#pragma unroll" which only sometimes had a compile-time trip count depending
on template magic. When the trip count was a compile-time constant, everything works
as expected and the loop is fully unrolled. However, when the trip count was not a
compile-time constant the "#pragma unroll" explicitly disabled unrolling of the loop(!).
Removing "#pragma unroll" caused the loop to be unrolled partially which was desirable
from a performance perspective.
llvm-svn: 244467
Summary:
Add brace style `BS_WebKit` as described on https://www.webkit.org/coding/coding-style.html:
* Function definitions: place each brace on its own line.
* Other braces: place the open brace on the line preceding the code block; place the close brace on its own line.
Set brace style used in `getWebKitStyle()` to the newly added `BS_WebKit`.
Reviewers: djasper, klimek
Subscribers: klimek, cfe-commits
Differential Revision: http://reviews.llvm.org/D11837
llvm-svn: 244446
-mkernel enables -fno-builtin and -fno-common by default, but allows -fbuiltin
and -fcommon to override that. However "-fbuiltin -fno-builtin" is treated the
same as "-fbuiltin" which is wrong, so fix that. Also fixes similar behaviour
when -fno-common is default.
Differential Revision: http://reviews.llvm.org/D11459
llvm-svn: 244437
Original class was not marked with inheritance attribute and it causes a crash on codegen.
Differential Revision: http://reviews.llvm.org/D11828
llvm-svn: 244428
build process when we implicitly build a module. Previously, we'd create the
specified .d file once for each implicitly-built module and then finally
overwrite it with the correct contents after the requested build completes.
(This fails if you use stdout as a dependency file, which is what the provided
testcase does, and is how I discovered this brokenness.)
llvm-svn: 244412
created, rather than creating and attaching a new listener each time we load a
module file (yes, the old ones were kept around too!). No functionality change
intended, but a bit more sanity.
llvm-svn: 244411
arguments because the reloaded form might have become non-canonical across the
serialization/deserialization step (this particularly happens when the
canonical form of the type involves an expression).
llvm-svn: 244409
Implemented in MinGW::Linker::AddLibGCC since AddLibgcc is a logic puzzle even
before adding one more boolean. A first step towards simplification of AddLibgcc
would be to factor out the Android AddLibgcc code into its own routine.
llvm-svn: 244407
This reverts commit fc885033a30b6e30ccf82398ae7c30e646727b10.
Revert all localization checker commits until the proper fix is implemented.
llvm-svn: 244394
This reverts commit 57a46a75b408245cf4154a838fe13ad702065745.
Revert all localization checker commits until the proper fix is implemented.
llvm-svn: 244393
Add checkers that detect code-level localizability issues for OS X / iOS:
- A path sensitive checker that warns about uses of non-localized
NSStrings passed to UI methods expecting localized strings.
- A syntax checker that warns against not including a comment in
NSLocalizedString macros.
A patch by Kulpreet Chilana!
llvm-svn: 244389
The ObjCSuperCallChecker issues alarms for various Objective-C APIs that require
a subclass to call to its superclass's version of a method when overriding it.
So, for example, it raises an alarm when the -viewDidLoad method in a subclass
of UIViewController does not call [super viewDidLoad].
This patch fixes a false alarm where the analyzer erroneously required the
implementation of the superclass itself (e.g., UIViewController) to call
super.
rdar://problem/18416944
Differential Revision: http://reviews.llvm.org/D11842
llvm-svn: 244386
Function types without prototypes can arise when mangling a function type
within an overloadable function in C. We mangle these as the absence of
any parameter types (not even an empty parameter list).
Differential Revision: http://reviews.llvm.org/D11848
llvm-svn: 244374
... and add aarch32 to specifically refer to the 32-bit ones.
Previously, 'arm' meant only 32-bit architectures and there was no way
for a module to build with both 32 and 64 bit ARM architectures.
Now a module that is intended to work on both architectures can specify
requires arm
whereas a module only for 32-bit platforms can say
requires aarch32
and just like before, 64-bit only can say
requires aarch64
llvm-svn: 244306
so that we can populate it on a per-target basis with required features.
Future commits will start using this information for warnings.
llvm-svn: 244286
MinGW has some pretty strange behvaior around RTTI and
dllimport/dllexport:
- RTTI data is never imported
- RTTI data is only exported if the class has no key function.
llvm-svn: 244266
This initial commit serves as an example -- the remainder of the
classes using pointer arithmetic for trailing objects will be
converted in subsequent changes.
Differential Revision: http://reviews.llvm.org/D11298
llvm-svn: 244262
OpenMP 4.1 allows to use variables with reference types in private clauses and, therefore, in init expressions of the cannonical loop forms.
llvm-svn: 244209
When a thunk is generated with a call to the original adjusted function,
the thunk appears in the debugger call stack. We want the backend to perform
tail-call optimization on the call, to make it invisible to the debugger.
This fixes PR24235
Patch by: amjad.aboud@intel.com
Differential Revision: http://reviews.llvm.org/D11476
llvm-svn: 244207
determine the primary context, rather than sometimes registering the lookup
table on the wrong context.
This exposed a couple of bugs:
* the odr violation check didn't deal properly with mergeable declarations
if the declaration retained by name lookup wasn't in the canonical
definition of the class
* the (broken) RewriteDecl mechanism would emit two name lookup tables for
the same DeclContext into the same module file (one as part of the
rewritten declaration and one as a visible update for the old declaration)
These are both fixed too.
llvm-svn: 244192
useless return value. Switch to using it directly when completing the
redeclaration chain for an anonymous declaration, and reduce the set of
declarations that we load in the process to just those of the right kind.
llvm-svn: 244161
This patch adds flags -fno-profile-instr-generate and
-fno-profile-instr-use, and the GCC aliases -fno-profile-generate and
-fno-profile-use.
These flags are used in situations where users need to disable profile
generation or use for specific files in a build, without affecting other
files.
llvm-svn: 244153
Summary:
By default, 'clang' emits dwarf and 'clang-cl' emits codeview. You can
force emission of one or both by passing -gcodeview and -gdwarf to
either driver.
Reviewers: dblaikie, hans
Subscribers: cfe-commits
Differential Revision: http://reviews.llvm.org/D11742
llvm-svn: 244097
This seems preferable to printing two warnings per unsupported option-
one warning about not supporting it, and one about not using it.
It also makes the '-Wno-' option do what you mean.
Differential Revision: http://reviews.llvm.org/D11766
llvm-svn: 244079
noticed until now.
The code for setting up the driver's InstalledDir didn't respect
-no-canonical-prefixes. Because of this, there are a few places in the
driver where we would unexpectedly form absolute paths, notably when
searching for and finding GCC installations to use, etc. The fix is
straightforward, and I've added this path to '-v' both so we can test it
sanely and so that it will be substantially more obvious the next time
someone has to debug something here.
Note that there is another bug that we don't actually *canonicalize* the
installed directory! I don't really want to fix that because I don't
have a realistic way to test the usage of this mode. I suspect that
folks using the shared module cache would care about getting this right
though, and so they might want to address it. I've left the appropriate
FIXMEs so that it is clear what to change, and I've updated the test
code to make it clear what is happening here.
llvm-svn: 244065
Support for emitting libcalls for __atomic_fetch_nand and
__atomic_{add,sub,and,or,xor,nand}_fetch was missing; add it, and some
test cases.
Differential Revision: http://reviews.llvm.org/D10847
llvm-svn: 244063
SmallVector::set_size does not reallocate the vector. Sadly I have no
idea how to test this. The vector never has more than one member in all
of the regression tests.
Found by inspection.
llvm-svn: 243987
set_size only resets the end pointer and asserts if it is used to grow
the buffer. This would crash when mangling a float with more than 80 bits,
add a test with a ppc double double (128 bits).
Found by inspection.
llvm-svn: 243979
This brings ASTContext closer to LLVM's Allocator concept. Ideally we
would just derive ASTContext from llvm::AllocatorBase, but that does
not work because ASTContext's allocator is mutable and we allocate using
const ASTContext& everywhere.
llvm-svn: 243972
If a global variable is marked as private in OpenMP construct and then is used in of the private clauses of the same construct, it might cause compiler crash because of incorrect capturing.
llvm-svn: 243964
a BumpPtrAllocator. This at least now handles the case where there is no
concatentation without calling memcpy on a null pointer. It might be
interesting to handle the case where everything is empty without
round-tripping through the allocator, but it wasn't clear to me if the
pointer returned is significant in any way, so I've left it in
a conservatively more-correct state.
Again, found with UBSan.
llvm-svn: 243948
rather than forcing the bump pointer allocator to produce a viable
pointer. This also fixes UB when we would try to memcpy from the null
incoming StringRef.
llvm-svn: 243947
the nested name specifier code.
First, skip the entire thing when the input is empty.
Next, handle the case where we started off with a null buffer and a zero
capacity to skip copying and freeing.
This was found with UBSan.
llvm-svn: 243946
Summary: In addition to checking compiler flags, the front-end also examines the attributes of the destructor definition to ensure that the SanitizeMemory attribute is attached.
Reviewers: eugenis, kcc
Subscribers: cfe-commits
Differential Revision: http://reviews.llvm.org/D11727
refactored test into new file, revised how function attribute examined
modified test to examine default dtor with and without attribute
removed attribute check
llvm-svn: 243912
Summary: Add IsConcept bit to VarDecl::NonParmVarDeclBitfields and associated isConcept/setConcept member functions. Set IsConcept to true when 'concept' specifier is in variable declaration. Create diagnostic when variable concept is not initialized.
Reviewers: fraggamuffin, hubert.reinterpretcast, faisalv, aaron.ballman, rsmith
Subscribers: aemerson, cfe-commits
Differential Revision: http://reviews.llvm.org/D11600
llvm-svn: 243876
Compiler crashed when vector elements / global register vars were used in inline assembler with "m" restriction. This patch fixes this.
Differential Revision: http://reviews.llvm.org/D10476
llvm-svn: 243870
This patch fixes bug 23800 ( https://llvm.org/bugs/show_bug.cgi?id=23800#c2 ). There existed a case where the index operand from extractelement was directly used to create a shufflevector mask. Since the index can be of any integral type but the mask must only contain 32 bit integers a 64 bit index operand led to an assertion error later on.
Committed on behalf of mpflanzer (Moritz Pflanzer)
Differential Revision: http://reviews.llvm.org/D10838
llvm-svn: 243851
The new EH instructions make it possible for LLVM to generate .xdata
tables that the MSVC personality routines will be happy about. Because
this is experimental, hide it behind a -cc1 flag (-fnew-ms-eh).
Differential Revision: http://reviews.llvm.org/D11405
llvm-svn: 243767
Adjust to LLVM DIBuilder API changes in r243764, using
`createAutoVariable()` and `createParameterVariable()` in place of
`createLocalVariable()`. No real functionality change here.
llvm-svn: 243765
Parsing of pragmas followed by a class member declaration is outlined into a separate function Parser::ParseCXXClassMemberDeclarationWithPragmas().
llvm-svn: 243739
It doesn't make any sense to enable -gmlt with -gsplit-dwarf, since
-gmlt is designed for on-line symbolication (and -gsplit-dwarf normally
emits all the -gmlt data into the .o anyway - so there's nothing to
split out except redundant/duplicate info).
With this change they override each other, -gmlt -gsplit-dwarf is the
same as -gsplit-dwarf and -gsplit-dwarf -gmlt is the same as -gmlt.
llvm-svn: 243694
Summary:
Currently, if the argument to _Pragma is not a parenthesised string
literal, the bad token will be consumed, as well as the ')', if present.
If additional bad tokens are passed to the _Pragma, this results in
extra error messages which may distract from the true problem.
The proposed patch causes all tokens to be consumed until the closing
')' or a new line, whichever is reached first.
Reviewers: hfinkel, rsmith
Subscribers: hubert.reinterpretcast, fraggamuffin, rnk, cfe-commits
Differential Revision: http://reviews.llvm.org/D8308
Patch by Rachel Craik!
llvm-svn: 243692
In llvm commit r243581, a reverse range adapter was added which allows
us to change code such as
for (auto I = Fields.rbegin(), E = Fields.rend(); I != E; ++I) {
in to
for (const FieldDecl *I : llvm::reverse(Fields))
This commit changes a few of the places in clang which are eligible to use
this new adapter.
llvm-svn: 243663
Copying the already existing code for x86 to ARM to set the correct CPU
features when using -mcpu=native. We can already detect the CPU name
but we were not setting the correct feature bits.
Moving fpu/hwdiv down to make sure they override whatever we set the
default to be.
No tests because this is native detection, and not all ARM-enabled builds
will hapen at a specific CPU, or even ARM. I have tested locally and it
works as expected.
Fixes PR12794.
llvm-svn: 243656
The memory-sanitizer build bot reported:
==5574== WARNING: MemorySanitizer: use-of-uninitialized-value
#0 0x7f03089e15f6 in
clang::Parser::TryAltiVecTokenOutOfLine(clang::DeclSpec&,
clang::SourceLocation, char const*&, unsigned int&, bool&)
/mnt/b/sanitizer-buildbot3/sanitizer-x86_64-linux-fast/build/llvm/tools/clang/lib/Parse/ParseDecl.cpp:6290:11
This is because the "Ident_pixel" variable was uninitialized
in the getLangOpts().ZVector case, but we'd still call into
clang::Parser::TryAltiVecTokenOutOfLine, which uses the variable.
The simplest fix for this without sprinkling !getLangOpts().ZVector
checks all over the code seems to be to just initialize the variable
to nullptr; this will then do the right thing on ZVector.
Checked in to unbreak the build bots.
llvm-svn: 243644
This patch adds support for the System Z vector built-in functions.
The API-defined header file has the name vecintrin.h.
The user-level functions are defined in the same style as the clang
version of altivec.h, making heavy use of the __overloadable__ and
__always_inline__ attributes. Where possible the functions expand to
generic operations rather than specific built-in functions, in the hope
that that form can be optimised better.
Where a built-in routine is specified to require an immediate integer
argument, the __enable_if__ attribute is used to verify the argument is
in fact constant and in the appropriate range.
Based on a patch by Richard Sandiford.
llvm-svn: 243643
The z13 vector facility has an associated language extension,
closely modeled on AltiVec/VSX. The main differences are:
- vector long, vector float and vector pixel are not supported
- vector long long and vector double are supported (like VSX)
- comparison operators return a vector rather than a scalar integer
- shift operators behave like the OpenCL shift operators
- vector bool is only supported as argument to certain operators;
some operators allow mixing a bool with a non-bool vector
This patch adds clang support for the extension. It is closely modelled
on the AltiVec support. Similarly to the -faltivec option, there's a
new -fzvector option to enable the extensions (as well as an -mzvector
alias for compatibility with GCC). There's also a separate LangOpt.
The extension as implemented here is intended to be compatible with
the -mzvector extension recently implemented by GCC.
Based on a patch by Richard Sandiford.
Differential Revision: http://reviews.llvm.org/D11001
llvm-svn: 243642
OpenMP 4.1 introduces optional argument '(n)' for 'ordered' clause, where 'n' is a number of loops that immediately follow the directive.
'n' must be constant positive integer expressions and it must be less or equal than the number of the loops in the resulting loop nest.
Patch adds parsing and semantic analysis for this optional argument.
llvm-svn: 243635
Also fix completely broken and untested code which was hiding the
primary bug. The !LLVM_ON_UNIX branch of the ifdef was actually a no-op.
I ran into this in the wild. It was causing failures in our SDK build.
Ideally we'd have a perfect llvm::sys::fs::canonical, but at least this
is a step in the right direction, and fixes an obviously broken case.
In some sense the test case I've added here is an integration test. We
should have these routines thoroughly unit tested in llvm::sys::fs.
llvm-svn: 243597
Without DR1579 implemented, the only case for -Wredundant-move is for a
parameter being returned with the same type as the function return type. Also
include a check to verify that the move constructor will be used by matching
nodes in the AST dump.
llvm-svn: 243594
UsingShadowDecls over other declarations of the same entity in the lookup
results. This ensures that we build correct redeclaration chains for the
UsingShadowDecls (otherwise we could see assertions and other misbehavior in
modules builds, when merging combines multiple redeclaration chains for the
same entity from the same module into one chain).
llvm-svn: 243592
Dependent types can throw off the analysis for these warnings, possibly giving
conflicting warnings and fix-its. Disabling the warning in template
instantiations will prevent this problem, and will still catch the
non-dependent cases in templates.
llvm-svn: 243538
new GV (usually NAME.1) instead of the correct NAME of the old GV. Moving comdat
creation after GV replacement solves this. Patch + testcase.
Reviewed by Reid Kleckner.
http://reviews.llvm.org/D11594
llvm-svn: 243525
This commit changes the driver to save subtarget feature "+strict-align"
to the IR instead of using backend option "aarch64-strict-align". This is
needed for LTO.
rdar://problem/21529937
llvm-svn: 243518
This commit changes the driver to save subtarget feature "+strict-align" to the
IR instead of using backend option "arm-strict-align". This is needed for LTO.
Also, move the logic in ARM backend that was deciding whether strict alignment
should be forced to the front-end.
rdar://problem/21529937
http://reviews.llvm.org/D11472
llvm-svn: 243489
Host-only cuda compilation does produce valid host object
file and in some cases users do want to proceed on to the linking phase.
The change removes special case that stopped compilation pipeline at
the Assembly phase. Device-side compilation is still stopped early
by the types::getCompilationPhases().
Differential Revision: http://reviews.llvm.org/D11573
llvm-svn: 243478
Store the locations for a macro expansion in a vector, then iterate over them
instead of using recursion. This simplifies the logic around the backtrace
limit and gives easier access to the source locations. No functionality change.
Patch by Zhengkai Wu.
Differential Revision: http://reviews.llvm.org/D11542
llvm-svn: 243477
error.
If the object being moved has a move constructor and a deleted copy constructor,
std::move is required, otherwise Clang will give a deleted constructor error.
llvm-svn: 243463
This will be used for old targets like Android that do not
support ELF TLS models.
Differential Revision: http://reviews.llvm.org/D10524
llvm-svn: 243441
- Use cached LLVM types
- Turn SmallVectors into Arrays/ArrayRef if the size is static
- Use ConstantInt::get's implicit splatting for vector types
No functionality change intended.
llvm-svn: 243425
This was calling FD->hasBody(), meaning "Does the function that this
decl refers to have a body?", rather than
FD->doesThisDeclarationHaveABody(), meaning "Is this decl a
non-deleted definition?".
We might want to consider renaming these APIs :/
llvm-svn: 243360
This patch allows Clang to pass on -Wa,-mfpu, -Wa,-mhwdiv and
-Wa,-mcpu to the integrated assembler (via target-features), but
-march is still not being passed, but validated.
In case the command line has both -mxxx and -Wa,-mxxx, we warn
that the naked one will not be used in assembler mode.
llvm-svn: 243353
To be able to handle -Wa, options in the assembler (ClangAs), we need to
make the handling of options based on the value of the options, not direct
Arguments from the list, since the list is immutable.
No functional change in this patch, but this allows validating of -Wa,-mfpu
and friends in the same way we validate -mfpu and friends, *just* for the
assembler.
llvm-svn: 243352
When ‘#pragma clang loop vectorize(assume_safety)’ was specified on a loop other loop hints were lost. The problem is that CGLoopInfo attaches metadata differently than EmitCondBrHints in CGStmt. For do-loops CGLoopInfo attaches metadata to the br in the body block and for while and for loops, the inc block. EmitCondBrHints on the other hand always attaches data to the br in the cond block. When specifying assume_safety CGLoopInfo emits an empty llvm.loop metadata shadowing the metadata in the cond block. Loop transformations like rotate and unswitch would then eliminate the cond block and its non-empty metadata.
This patch unifies both approaches for adding metadata and modifies the existing safety tests to include non-assume_safety loop hints.
llvm-svn: 243315
The 3DNOW/PRFCHW cpu targets define both the PREFETCHW (set cache line modified) and PREFETCH (set cache line exclusive) instructions but only the _m_prefetchw (PREFETCHW) intrinsic is included in the header. This patch adds the missing _m_prefetch intrinsic.
I'm basing this off AMD documentation - the intel docs on the support for PREFETCHW isn't clear whether Silvermont/Broadwell properly support PREFETCH but given that the intrinsic implementation is a default __builtin_prefetch call, it is safe whatever.
Fix for PR23648
Differential Revision: http://reviews.llvm.org/D11338
llvm-svn: 243305
Also rename XCore (the toolchain) to XCoreToolChain since XCore is
also a namespace for its tools.
Differential Revision: http://reviews.llvm.org/D10609
llvm-svn: 243279
Currently trigger to select hard-float linker is only based of -gnueabihf
appearing in target triplet, but we should also select it when hardfloat
is requested via cmdline.
Patch by Khem Raj.
llvm-svn: 243262
We had multiple bugs here:
- We didn't support multiple optimization options in one argument.
e.g. -O2y-
- We didn't correctly expand -O[12dx] to their respective options.
- We treated -O1 as clang -O1 instead of clang -Os.
- We treated -Ox as clang -O3 instead of clang -O2. In fact, cl's -Ox
option is *less* powerful than cl's -O2 option despite -Ox described
as "Full Optimization".
This fixes PR24003.
llvm-svn: 243261
isWhitespaceExceptNL. If that's wasn't bad enough, we have an identically
named function in Basic/CharInfo.h.
While at it, remove unnecessary includes, .str() calls and #ifdef.
llvm-svn: 243260
chain and fix the cases where it fires.
* Handle the __va_list_tag as a predefined decl. Previously we failed to merge
sometimes it because it's not visible to name lookup. (In passing, remove
redundant __va_list_tag typedefs that we were creating for some ABIs. These
didn't affect the mangling or representation of the type.)
* For Decls derived from Redeclarable that are not in fact redeclarable
(implicit params, function params, ObjC type parameters), remove them from
the list of expected redeclarable decls.
llvm-svn: 243259
There is currently no support in MSVC for using i128 as an integer
literal suffix. In fact, there appears to be no evidence that they have
ever supported this feature in any of their compilers. This was an over
generalization of their actual feature and is a nasty source of bugs.
Why is it a source of bugs? Because most code in clang expects that
evaluation of an integer constant expression won't give them something
that 'long long' can't represent. Instead of providing a meaningful
feature, i128 gives us cute ways of exploding the compiler.
llvm-svn: 243243
RecordLayoutBuilder is an inaccruate name because it does not build all
records. It only builds layouts for targets using the Itanium C++ ABI.
llvm-svn: 243225
This lets us pass functors (and lambdas) without void * tricks. On the
downside we can't pass CXXRecordDecl's Find* members (which are now type
safe) to lookupInBases directly, but a lambda trampoline is a small
price to pay. No functionality change intended.
llvm-svn: 243217
__builtin_frame_address requires its argument to be a constant
expression which already implies that it cannot have undefined behavior.
However, we used EmitScalarExpr to emit the argument causing UBSan to
try to check for overflow.
Instead, use the constant expression emission system.
This fixes PR24256.
llvm-svn: 243206
option "-aarch64-reserve-x18".
This change is needed since backend options do not make it to the backend
when doing LTO and are not capable of changing the behavior of code-gen
passes on a per-function basis.
rdar://problem/21529937
Differential Revision: http://reviews.llvm.org/D11462
llvm-svn: 243185
Change `getOrCreateLimitedType()` to return a `DICompositeType` and
remove the casts from its callers. Inside, I've strengthened a `cast`
from `DICompositeTypeBase`, but the casts in the callers already prove
that this is safe. There should be no functionality change here.
llvm-svn: 243155
The outer nullability is transferred from the declaration to the implementation so including them is redundant.
The inner ones are not transferred so they are kept to match the exact types. When we transfer the inner ones as well
adding them in the implementation will become redundant and we should strip those as well.
rdar://21737451
llvm-svn: 243119
Address Richard Smith comments: remove the trailing seperator from the Arch
variable, implement six mingw_* trees under tools/clangtest/Driver/Inputs
and merge linux and Windows tests into a universal test that uses these trees.
llvm-svn: 243098
The flag allows users to specify that they do not want the object file
to have any implicit /defaultlib directives.
This fixes PR24236.
llvm-svn: 243097
We treated DependentSizedArrayTypes with the same element type but
differing size expressions as equivalently canonical. This would lead
to bizarre behavior during template instantiation.
This fixes PR24212.
llvm-svn: 243093
Generating available_externally vtables for optimizations purposes.
Unfortunatelly ItaniumABI doesn't guarantee that we will be able to
refer to virtual inline method by name.
But when we don't have any inline virtual methods, and key function is
not defined in this TU, we can generate that there will be vtable and
mark it as available_externally.
This is patch will help devirtualize better.
Differential Revision: http://reviews.llvm.org/D11441
llvm-svn: 243090
to consider the storage size of the vector instead of its
sizeof. In other words, ban <3 x int> to <4 x int> casts,
which produced invalid IR anyway.
Also, attempt to be a little more rigorous, or at least
explicit, about when enums are allowed in these casts.
rdar://21901132
llvm-svn: 243069
The catch keyword isn't really part of a region, so it's fairly
meaningless to extend into it. This was usually harmless, but it could
crash when catch blocks involved macros in strange ways.
llvm-svn: 243066
Large CFGs cause `checkForFunctionCall()` to overflow its stack. Break
the recursion by manually managing the call stack instead.
Patch by Vedant Kumar!
llvm-svn: 243039
Split out `hasRecursiveCallInPath()` from `checkForFunctionCall()` to
flatten nesting and clarify the code. This also simplifies a follow-up
patch that refactors the other logic in `checkForFunctionCall()`.
Patch by Vedant Kumar!
llvm-svn: 243038
Currently, for --rtlib=compiler-rt on GNU systems, we're assuming
that one has libgcc_s and libgcc_eh as low-level libraries, which
when used in conjunction with -lunwind or -lc++abi, breaks that
assumption.
My original fix was wrong, and this patch reverts it to prepare for
a new flag to choose the unwinder/C++ libraries. For the time being,
people can use "-lgcc_s -lgcc_eh" or "-lunwind -lc++abi" or any
combination they need explicitly.
llvm-svn: 243025
Richard Smith