This adds diagnostics for conflicting attributes on the same
declarataion, conflicting attributes on a forward and final
declaration, and defines a more narrowly scoped HLSLEntry attribute
target.
Big shout out to @aaron.ballman for the great feedback and review on
this!
The memory_scope enum is not available before OpenCL 2.0, so ensure
the sub_group_barrier overload with a memory_scope argument is
restricted to OpenCL 2.0 and above. This is already the case in
opencl-c.h.
Fixes the issue revealed by https://reviews.llvm.org/D120254
Reported-by: Harald van Dijk (hvdijk)
Index of vset/vget must be a constant integer and be
located in right range.
Reviewed By: kito-cheng
Differential Revision: https://reviews.llvm.org/D122629
We started diagnosing this situation with a more clear diagnostic
message, but it was pointed out that unevaluated contexts don't really
have the undefined behavior property as there is no runtime access
involved.
This augments the changes in https://reviews.llvm.org/D122656 to not
diagnose in an unevaluated context.
HLSL uses Microsoft-style attributes `[attr]`, which clang mostly
ignores. For HLSL we need to handle known Microsoft attributes, and to
maintain C/C++ as-is we ignore unknown attributes.
To utilize this new code path, this change adds the HLSL `numthreads`
attribute.
Reviewed By: rnk
Differential Revision: https://reviews.llvm.org/D122627
Member access for an atomic structure or union is unconditional
undefined behavior (C11 6.5.2.3p5). However, we would issue a confusing
error message about the base expression not being a structure or union
type.
GCC issues a warning for this case. Clang now warns as well, but the
warning is defaulted to an error because the actual access is still
unsafe.
This fixes Issue 54563.
This builtin returns the address of a global instance of the
`std::source_location::__impl` type, which must be defined (with an
appropriate shape) before calling the builtin.
It will be used to implement std::source_location in libc++ in a
future change. The builtin is compatible with GCC's implementation,
and libstdc++'s usage. An intentional divergence is that GCC declares
the builtin's return type to be `const void*` (for
ease-of-implementation reasons), while Clang uses the actual type,
`const std::source_location::__impl*`.
In order to support this new functionality, I've also added a new
'UnnamedGlobalConstantDecl'. This artificial Decl is modeled after
MSGuidDecl, and is used to represent a generic concept of an lvalue
constant with global scope, deduplicated by its value. It's possible
that MSGuidDecl itself, or some of the other similar sorts of things
in Clang might be able to be refactored onto this more-generic
concept, but there's enough special-case weirdness in MSGuidDecl that
I gave up attempting to share code there, at least for now.
Finally, for compatibility with libstdc++'s <source_location> header,
I've added a second exception to the "cannot cast from void* to T* in
constant evaluation" rule. This seems a bit distasteful, but feels
like the best available option.
Reviewers: aaron.ballman, erichkeane
Differential Revision: https://reviews.llvm.org/D120159
Calling an ObjC method from a C function marked with the 'enforce_tcb'
attribute did not produce a warning. Now it does, and on top of that
Objective-C methods can participate in TCBs themselves.
Differential Revision: https://reviews.llvm.org/D122343
Previously, we would instantiate the UDL by marking the function as
referenced and potentially binding to a temporary; this skipped
transforming the call when the UDL was dependent on a template
parameter.
Now, we defer all the work to instantiating the call expression for the
UDL. This ensures that constant evaluation occurs at compile time
rather than deferring until runtime.
Fixes Issue 54578.
We wish to support emitting a pre-processed output for an importable
header unit, that can be consumed to produce the same header units as
the original source.
This means that ee need to find the original filename used to produce
the re-preprocessed output, so that it can be assigned as the module
name. This is peeked from the first line of the pre-processed source
when the action sets up the files.
Differential Revision: https://reviews.llvm.org/D121098
In C, assignment expressions result in an rvalue whose type is the type
of the lhs of the assignment after it undergoes lvalue to rvalue
conversion. lvalue to rvalue conversion in C strips all qualifiers
including _Atomic.
We used getUnqualifiedType() which does not strip the _Atomic qualifier
when we should have used getAtomicUnqualifiedType(). This corrects the
usage and adds some comments to getUnqualifiedType() to make it more
clear that it does not strip _Atomic and that's on purpose (see C11
6.2.5p27).
This addresses Issue 48742.
Some functions can end up non-externally visible despite not being
declared "static" or in an unnamed namespace in C++ - such as by having
parameters that are of non-external types.
Such functions aren't mistakenly intended to be defining some function
that needs a declaration. They could be maybe more legible (except for
the operator new example) with an explicit static, but that's a
stylistic thing outside what should be addressed by a warning.
This reapplies 275c56226d7fbd6a4d554807374f78d323aa0c1c - once we figure
out what to do about the change in behavior for -Wnon-c-typedef-for-linkage
(this reverts the revert commit 85ee1d3ca1d06b6bd3477515b8d0c72c8df7c069)
Differential Revision: https://reviews.llvm.org/D121328
When adding the support for modules partitions we added an assert that the
actual status of Global Module Fragments matches the state machine that is
driven by the module; keyword.
That does not apply to the modules-ts case, where there is an implicit GMF.
Differential Revision: https://reviews.llvm.org/D122394
There seems to be more than one way to get to that state. I included to
example cases in the test, both were noticed recently.
There is room for improvement, for example by creating RecoveryExpr in
place of the bad initializer, but for now let's stop the crashes.
Differential Revision: https://reviews.llvm.org/D121824
This is the first in a series of patches that introduce C++20 importable
header units.
These differ from clang header modules in that:
(a) they are identifiable by an internal name
(b) they represent the top level source for a single header - although
that might include or import other headers.
We name importable header units with the path by which they are specified
(although that need not be the absolute path for the file).
So "foo/bar.h" would have a name "foo/bar.h". Header units are made a
separate module type so that we can deal with diagnosing places where they
are permitted but a named module is not.
Differential Revision: https://reviews.llvm.org/D121095
Update `WeakUndeclaredIdentifiers` to hold a collection of weak
aliases per identifier instead of only one.
This also allows the "used" state to be removed from `WeakInfo`
because it is really only there as an alternative to removing
processed map entries, and we can represent that using an empty set
now. The serialization code is updated for the removal of the field.
Additionally, a PCH test is added for the new functionality.
The records are grouped by the "target" identifier, which was already
being used as a key for lookup purposes. We also store only one record
per alias name; combined, this means that diagnostics are grouped by
the "target" and limited to one per alias (which should be acceptable).
Fixes PR28611.
Fixesllvm/llvm-project#28985.
Reviewed By: aaron.ballman, cebowleratibm
Differential Revision: https://reviews.llvm.org/D121927
Co-authored-by: Rachel Craik <rcraik@ca.ibm.com>
Co-authored-by: Jamie Schmeiser <schmeise@ca.ibm.com>
A previous patch removed the compiler generating offloading entries
for variables that were declared on the device but were internal or
hidden. This allowed us to compile programs but turns any attempt to run
'#pragma omp target update' on one of those variables a silent failure.
This patch adds a check in the semantic analysis for if the user is
attempting the update a variable on the device from the host that is not
externally visible.
Reviewed By: jdoerfert
Differential Revision: https://reviews.llvm.org/D122403
CUDA/HIP determines whether a function can be called based on
the device/host attributes of callee and caller. Clang assumes the
caller is CurContext. This is correct in most cases, however, it is
not correct in OpenMP parallel region when CUDA/HIP program
is compiled with -fopenmp. This causes incorrect overloading
resolution and missed diagnostics.
To get the correct caller, clang needs to chase the parent chain
of DeclContext starting from CurContext until a function decl
or a lambda decl is reached. Sema API is adapted to achieve that
and used to determine the caller in hostness check.
Reviewed by: Artem Belevich, Richard Smith
Differential Revision: https://reviews.llvm.org/D121765
Adds basic parsing/sema/serialization support for the
#pragma omp target parallel loop directive.
Differential Revision: https://reviews.llvm.org/D122359
Move the SourceRange from the old ParsedAttributesWithRange into
ParsedAttributesView, so we have source range information available
everywhere we use attributes.
This also removes ParsedAttributesWithRange (replaced by simply using
ParsedAttributes) and ParsedAttributesVieWithRange (replaced by using
ParsedAttributesView).
Differential Revision: https://reviews.llvm.org/D121201
Clang fails to diagnose:
```
void test() {
int j = 0;
for (int i = 0; i < 1000; i++)
j++;
return;
}
```
Reason: Missing support for UnaryOperator.
We should not warn with volatile variables... so add check for it.
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D122271
Currently, Clang handles some qualifiers correctly for __auto_type, but
it does not handle the restrict or _Atomic qualifiers in the same way
that GCC does. This patch handles those qualifiers so that they attach
to the deduced type the same as const and volatile already do.
This fixes https://github.com/llvm/llvm-project/issues/53652
This patch extends the support for C/C++ operators for SVE
types to allow one of the arguments to be a scalar, in which
case a vector splat is performed.
Differential Revision: https://reviews.llvm.org/D121829
This reverts commit 56d46b36fc231a0beb518602503035bba92043e0.
The LIT test SemaCXX/attr-trivial-abi.cpp is failing with 32bit build on
Windows. All the lines with the ifdef WIN32 are asserting but they are
not expected to. It looks like the LIT test was not tested on a 32bit
build of the compiler.
GCC supports power-of-2 size structures for the arguments. Clang supports fewer than GCC. But Clang always crashes for the unsupported cases.
This patch adds sema checks to do the diagnosts to solve these crashes.
Reviewed By: jyu2
Differential Revision: https://reviews.llvm.org/D107141
Allow goto, labelled statements as well as `static`, `thread_local`, and
non-literal variables in `constexpr` functions.
As specified. for all of the above (except labelled statements) constant
evaluation of the construct still fails.
For `constexpr` bodies, the proposal is implemented with diagnostics as
a language extension in older language modes. For determination of
whether a lambda body satisfies the requirements for a constexpr
function, the proposal is implemented only in C++2b mode to retain the
semantics of older modes for programs conforming to them.
Reviewed By: aaron.ballman, hubert.reinterpretcast, erichkeane
Differential Revision: https://reviews.llvm.org/D111400
@mgehre-amd pointed out the following post-commit review feedback on
the changes in 8cba72177dcd8de5d37177dbaf2347e5c1f0f1e8:
As an example, the paper says 3wb /* Yields an _BitInt(3); two value
bits, one sign bit */.
So I would expect that 0xFwb gives _BitInt(5); four value bits, one
sign bit, but with this implementation I get _BitInt(2).
This is because ResultVal as 4 bits, and getMinSignedBits() inteprets
it as negative and thus says that 1 bit is enough to represent -1.
This corrects the behavior for calculating the bit-width and adds some
test coverage.
For the following code,
void test() {
volatile int j = 0;
for (int i = 0; i < 1000; i++)
j += 1;
return;
}
If compiled with
clang -g -Wall -Werror -S -emit-llvm test.c
we will see the following error:
test.c:2:6: error: variable 'j' set but not used [-Werror,-Wunused-but-set-variable]
volatile int j = 0;
^
This is not quite right since 'j' is indeed used due to '+=' operator.
gcc doesn't emit error either in this case.
Also if we change 'j += 1' to 'j++', the warning will disappear
with latest clang.
Note that clang will issue the warning if the volatile declaration
involves only simple assignment (var = ...).
To fix the issue, in function MaybeDecrementCount(), if the
operator is a compound assignment (i.e., +=, -=, etc.) and the
variable is volatile, the count for RefsMinusAssignments will be
decremented, similar to 'j++' case.
Differential Revision: https://reviews.llvm.org/D121715
Previously, OpenMP variant declarations for a function declaration that included
the 'cpu_dispatch', 'cpu_specific', or 'target' attributes was diagnosed, but
one with the 'target_clones' attribute was not. Now fixed.
Reviewed By: erichkeane, jdoerfert
Differential Revision: https://reviews.llvm.org/D121963
Previously, an attempt to declare an overload of a multiversion function
in C was not properly diagnosed. In some cases, diagnostics were simply
missing. In other cases the following assertion failure occured...
```
Assertion `(Previous.empty() || llvm::any_of(Previous, [](const NamedDecl *ND) { return ND->hasAttr(); })) && "Non-redecls shouldn't happen without overloadable present"' failed.
```
... or the following diagnostic was spuriously issued.
```
error: at most one overload for a given name may lack the 'overloadable' attribute
```
The diagnostics issued in some cases could be improved. When the function
type of a redeclaration does not match the prior declaration, it would be
preferable to diagnose the type mismatch before diagnosing mismatched
attributes. Diagnostics are also missing for some cases.
Reviewed By: erichkeane
Differential Revision: https://reviews.llvm.org/D121959
Checking of multiversion function declarations performed by various functions
in clang/lib/Sema/SemaDecl.cpp previously forced the valus of a passed in
'MergeTypeWithPrevious' reference argument in several scenarios. This was
unnecessary and possibly incorrect in the one case that the value
was forced to 'true' (though seemingly unobservably so).
Reviewed By: erichkeane
Differential Revision: https://reviews.llvm.org/D121958
This change removes redundant code in the definition of
CheckTargetCausesMultiVersioning() in SemaDecl.cpp. The removed code checked
for multiversion function support. The code immediately following the removed
code is a call to CheckMultiVersionAdditionalRules(); that function performs
the same check on entry. In both cases, the consequences of missing multiversion
function support results in the same diagnostic message being issued and the
applicable function declaration being marked as invalid.
Reviewed By: erichkeane, aaron.ballman
Differential Revision: https://reviews.llvm.org/D121957
This change removes dead code in the definition of CheckMultiVersionFunction()
in clang/lib/Sema/SemaDecl.cpp. The removed code was made dead by commit
fc53eb69c26cdd7efa6b629c187d04326f0448ca: "Reapply 'Implement target_clones multiversioning'".
See the added code just above the code being deleted; it contains the same
return statement with the previous condition now distributed across an if
statement and a switch statement.
Reviewed By: erichkeane, aaron.ballman
Differential Revision: https://reviews.llvm.org/D121955
