r280553 introduced an issue where we'd emit ambiguity errors for code
like:
```
void foo(int *, int);
void foo(unsigned int *, unsigned int);
void callFoo() {
unsigned int i;
foo(&i, 0); // ambiguous: int->unsigned int is worse than int->int,
// but unsigned int*->unsigned int* is better than
// int*->int*.
}
```
This patch fixes this issue by changing how we handle ill-formed (but
valid) implicit conversions. Candidates with said conversions now always
rank worse than candidates without them, and two candidates are
considered to be equally bad if they both have these conversions for
the same argument.
Additionally, this fixes a case in C++11 where we'd complain about an
ambiguity in a case like:
```
void f(char *, int);
void f(const char *, unsigned);
void g() { f("abc", 0); }
```
...Since conversion to char* from a string literal is considered
ill-formed in C++11 (and deprecated in C++03), but we accept it as an
extension.
llvm-svn: 280847
Summary:
C++1z 6.4.1/p2:
If the if statement is of the form if constexpr, the value of the
condition shall be a contextually converted constant expression of type
bool [...]
C++1z 5.20/p4:
[...] A contextually converted constant expression of type bool is an
expression, contextually converted to bool (Clause4), where the
converted expression is a constant expression and the conversion
sequence contains only the conversions above. [...]
Contextually converting result of an expression `e` to a Boolean value
requires `bool t(e)` to be well-formed.
An explicit conversion function is only considered as a user-defined
conversion for direct-initialization, which is essentially what
//contextually converted to bool// requires.
Also, fixes PR28470.
Reviewers: rsmith
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D24158
llvm-svn: 280838
We have invariants we like to guarantee for the
`ImplicitConversionKind`s in a `StandardConversionSequence`. These
weren't being upheld in code that r280553 touched, so Richard suggested
that we should fix that. See D24113.
I'm not entirely sure how to go about testing this, so no test case is
included. Suggestions welcome.
llvm-svn: 280562
This patch allows us to perform incompatible pointer conversions when
resolving overloads in C. So, the following code will no longer fail to
compile (though it will still emit warnings, assuming the user hasn't
opted out of them):
```
void foo(char *) __attribute__((overloadable));
void foo(int) __attribute__((overloadable));
void callFoo() {
unsigned char bar[128];
foo(bar); // selects the char* overload.
}
```
These conversions are ranked below all others, so:
A. Any other viable conversion will win out
B. If we had another incompatible pointer conversion in the example
above (e.g. `void foo(int *)`), we would complain about
an ambiguity.
Differential Revision: https://reviews.llvm.org/D24113
llvm-svn: 280553
Summary:
Some function calls in CUDA are allowed to appear in
semantically-correct programs but are an error if they're ever
codegen'ed. Specifically, a host+device function may call a host
function, but it's an error if such a function is ever codegen'ed in
device mode (and vice versa).
Previously, clang made no attempt to catch these errors. For the most
part, they would be caught by ptxas, and reported as "call to unknown
function 'foo'".
Now we catch these errors and report them the same as we report other
illegal calls (e.g. a call from a host function to a device function).
This has a small change in error-message behavior for calls that were
previously disallowed (e.g. calls from a host to a device function).
Previously, we'd catch disallowed calls fairly early, before doing
additional semantic checking e.g. of the call's arguments. Now we catch
these illegal calls at the very end of our semantic checks, so we'll
only emit a "illegal CUDA call" error if the call is otherwise
well-formed.
Reviewers: tra, rnk
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D23242
llvm-svn: 278759
Currently, when trying to evaluate an enable_if condition, we try to
evaluate all arguments a user passes to a function. Given that we can't
use variadic arguments from said condition anyway, not converting them
is a reasonable thing to do. So, this patch makes us ignore any varargs
when attempting to check an enable_if condition.
We'd crash because, in order to convert an argument, we need its
ParmVarDecl. Variadic arguments don't have ParmVarDecls.
llvm-svn: 278471
Summary:
When we emit err_ref_bad_target, we should emit a "'method' declared
here" note. We already do so in most places, just not in
BuildCallToMemberFunction.
Reviewers: tra
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D23240
llvm-svn: 278195
Summary:
I want to reuse "CheckCUDAFoo" in a later patch. Also, I think
IsAllowedCUDACall gets the point across more clearly.
Reviewers: tra
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D23238
llvm-svn: 278193
I'm told that some optimizers like lambdas a lot more than mem_fn.
Given that the readability difference is basically nil, and we seem to
use lambdas basically everywhere else, it seems sensible to just use
lambdas.
llvm-svn: 276577
It's a patch for PR28050. Seems like overloading resolution wipes out
the first standard conversion sequence (before user-defined conversion)
in case of deprecated string literal conversion.
Differential revision: https://reviews.llvm.org/D21228
Patch by Alexander Makarov
llvm-svn: 275970
Summary:
You can overload a destructor in CUDA, and SemaOverload needs to be
tweaked not to crash when it sees an explicit call to an overloaded
destructor.
Reviewers: rsmith
Subscribers: cfe-commits, tra
Differential Revision: http://reviews.llvm.org/D21912
llvm-svn: 275231
We didn't assign an inheritance model for 'Foo' if the event an
exrepssion like '&Foo::Bar' occured if 'Bar' could resolve to multiple
functions.
Once the overload set is resolved to a particular member, we enforce a
specific inheritance model.
This fixes PR28360.
llvm-svn: 274202
Replace inheriting constructors implementation with new approach, voted into
C++ last year as a DR against C++11.
Instead of synthesizing a set of derived class constructors for each inherited
base class constructor, we make the constructors of the base class visible to
constructor lookup in the derived class, using the normal rules for
using-declarations.
For constructors, UsingShadowDecl now has a ConstructorUsingShadowDecl derived
class that tracks the requisite additional information. We create shadow
constructors (not found by name lookup) in the derived class to model the
actual initialization, and have a new expression node,
CXXInheritedCtorInitExpr, to model the initialization of a base class from such
a constructor. (This initialization is special because it performs real perfect
forwarding of arguments.)
In cases where argument forwarding is not possible (for inalloca calls,
variadic calls, and calls with callee parameter cleanup), the shadow inheriting
constructor is not emitted and instead we directly emit the initialization code
into the caller of the inherited constructor.
Note that this new model is not perfectly compatible with the old model in some
corner cases. In particular:
* if B inherits a private constructor from A, and C uses that constructor to
construct a B, then we previously required that A befriends B and B
befriends C, but the new rules require A to befriend C directly, and
* if a derived class has its own constructors (and so its implicit default
constructor is suppressed), it may still inherit a default constructor from
a base class
llvm-svn: 274049
Given the following C++:
```
void foo();
void foo() __attribute__((enable_if(false, "")));
bool bar() {
auto P = foo;
return P == foo;
}
```
We'll currently happily (and correctly) resolve `foo` to the `foo`
overload without `enable_if` when assigning to `P`. However, we'll
complain about an ambiguous overload on the `P == foo` line, because
`Sema::CheckPlaceholderExpr` doesn't recognize that there's only one
`foo` that could possibly work here.
This patch teaches `Sema::CheckPlaceholderExpr` how to properly deal
with such cases.
Grepping for other callers of things like
`Sema::ResolveAndFixSingleFunctionTemplateSpecialization`, it *looks*
like this is the last place that needed to be fixed up. If I'm wrong,
I'll see if there's something we can do that beats what amounts to
whack-a-mole with bugs.
llvm-svn: 272080
This is in preparation for C++ P0136R1, which switches the model for inheriting
constructors over from synthesizing a constructor to finding base class
constructors (via using shadow decls) when looking for derived class
constructors.
llvm-svn: 269231
This patch implements __unaligned (MS extension) as a proper type qualifier
(before that, it was implemented as an ignored attribute).
It also fixes PR27367 and PR27666.
Differential Revision: http://reviews.llvm.org/D20103
llvm-svn: 269220
This patch fixes a bug where we would assume all value-dependent
enable_if conditions give successful results.
Instead, we consider value-dependent enable_if conditions to always
fail. While this isn't ideal, this is the best we can realistically do
without changing both enable_if's semantics and large parts of Sema
(specifically, all of the parts that don't expect type dependence to
come out of nowhere, and that may interact with overload resolution).
Differential Revision: http://reviews.llvm.org/D20130
llvm-svn: 269154
Currently, if clang::isBetterOverloadCandidate encounters an enable_if
attribute on either candidate that it's inspecting, it will ignore all
lower priority attributes (e.g. pass_object_size). This is problematic
in cases like:
```
void foo(char *c) __attribute__((enable_if(1, "")));
void foo(char *c __attribute__((pass_object_size(0))))
__attribute__((enable_if(1, "")));
```
...Because we would ignore the pass_object_size attribute in the second
`foo`, and consider any call to `foo` to be ambiguous.
This patch makes overload resolution consult further tiebreakers (e.g.
pass_object_size) if two candidates have equally good enable_if
attributes.
llvm-svn: 269005
This patch corresponds to reviews:
http://reviews.llvm.org/D15120http://reviews.llvm.org/D19125
It adds support for the __float128 keyword, literals and target feature to
enable it. Based on the latter of the two aforementioned reviews, this feature
is enabled on Linux on i386/X86 as well as SystemZ.
This is also the second attempt in commiting this feature. The first attempt
did not enable it on required platforms which caused failures when compiling
type_traits with -std=gnu++11.
If you see failures with compiling this header on your platform after this
commit, it is likely that your platform needs to have this feature enabled.
llvm-svn: 268898
This patch implements __unaligned (MS extension) as a proper type qualifier
(before that, it was implemented as an ignored attribute).
It also fixes PR27367.
Differential Revision: http://reviews.llvm.org/D19654
llvm-svn: 268727
With this patch compiler emits warning if it tries to make implicit instantiation
of a template but cannot find the template definition. The warning can be suppressed
by explicit instantiation declaration or by command line options
-Wundefined-var-template and -Wundefined-func-template. The implementation follows
the discussion of http://reviews.llvm.org/D12326.
Differential Revision: http://reviews.llvm.org/D16396
llvm-svn: 266719
Since this patch provided support for the __float128 type but disabled it
on all platforms by default, some platforms can't compile type_traits with
-std=gnu++11 since there is a specialization with __float128.
This reverts the patch until D19125 is approved (i.e. we know which platforms
need this support enabled).
llvm-svn: 266460
This patch corresponds to review:
http://reviews.llvm.org/D15120
It adds support for the __float128 keyword, literals and a target feature to
enable it. This support is disabled by default on all targets and any target
that has support for this type is free to add it.
Based on feedback that I've received from target maintainers, this appears to
be the right thing for most targets. I have not heard from the maintainers of
X86 which I believe supports this type. I will subsequently investigate the
impact of enabling this on X86.
llvm-svn: 266186
Instead of searching the global pool multiple times: in
LookupFactoryMethodInGlobalPool, LookupInstanceMethodInGlobalPool,
CollectMultipleMethodsInGlobalPool, and AreMultipleMethodsInGlobalPool,
we now collect the method candidates in CollectMultipleMethodsInGlobalPool
only, and other functions will use the collected method set.
This commit adds parameter "Methods" to AreMultipleMethodsInGlobalPool,
and SelectBestMethod. It also changes the implementation of
CollectMultipleMethodsInGlobalPool to collect the desired kind first, if none is
found, to collect the other kind. This avoids the need to call both
LookupFactoryMethodInGlobalPool and LookupInstanceMethodInGlobalPool.
llvm-svn: 265711
With this patch, by a constexpr function is implicitly host+device
unless:
a) it's a variadic function (variadic functions are not allowed on the
device side), or
b) it's preceeded by a __device__ overload in a system header.
The restriction on overloading __host__ __device__ functions on the
basis of their CUDA attributes remains in place, but we use (b) to allow
us to define __device__ overloads for constexpr functions in cmath,
which would otherwise be __host__ __device__ and thus not overloadable.
You can disable this behavior with -fno-cuda-host-device-constexpr.
Reviewers: tra, rnk, rsmith
Subscribers: cfe-commits
Differential Revision: http://reviews.llvm.org/D18380
llvm-svn: 264964
Summary:
IsOverload has a param named UseUsingDeclRules. But as far as I can
tell, it should be called UseMemberUsingDeclRules. That is, it only
applies to "using" declarations inside classes or structs.
Reviewers: rsmith
Subscribers: cfe-commits
Differential Revision: http://reviews.llvm.org/D18538
llvm-svn: 264920
Summary:
* -fcuda-target-overloads
Previously unconditionally set to true by the driver. Necessary for
correct functioning of the compiler -- our CUDA headers wrapper won't
compile without this.
* -fcuda-disable-target-call-checks
Previously unconditionally set to true by the driver. Necessary to
compile almost any external CUDA code -- almost all libraries assume
that host+device code can call host or device functions.
* -fcuda-allow-host-calls-from-host-device
No effect when target overloading is enabled.
Reviewers: tra
Subscribers: rsmith, cfe-commits
Differential Revision: http://reviews.llvm.org/D18416
llvm-svn: 264739
Also includes a minor ``enable_if`` docs update.
Currently, our address-of overload machinery will only allow implicit
conversions of overloaded functions to void* in C. For example:
```
void f(int) __attribute__((overloadable));
void f(double) __attribute__((overloadable, enable_if(0, "")));
void *fp = f; // OK. This is C and the target is void*.
void (*fp2)(void) = f; // Error. This is C, but the target isn't void*.
```
This patch makes the assignment of `fp2` select the `f(int)` overload,
rather than emitting an error (N.B. you'll still get a warning about the
`fp2` assignment if you use -Wincompatible-pointer-types).
Differential Revision: http://reviews.llvm.org/D13704
llvm-svn: 264132
Some functions can't have their address taken. If we encounter an
overload set where only one of the candidates can have its address
taken, we should automatically select that candidate in cast
expressions.
Differential Revision: http://reviews.llvm.org/D17701
llvm-svn: 263887
Similar to the template cases in r262050, when a C++ method in an
unavailable struct/class calls unavailable API, don't diagnose an error.
I.e., this case was failing:
void foo() __attribute__((unavailable));
struct __attribute__((unavailable)) A {
void bar() { foo(); }
};
Since A is unavailable, A::bar is allowed to call foo. However, we were
emitting a diagnostic here. This commit checks up the context chain
from A::bar, in a manner inspired by SemaDeclAttr.cpp:isDeclUnavailable.
I expected to find other related issues but failed to trigger them:
- I wondered if DeclBase::getAvailability should check for
`TemplateDecl` instead of `FunctionTemplateDecl`, but I couldn't find
a way to trigger this. I left behind a few extra tests to make sure
we don't regress.
- I wondered if Sema::isFunctionConsideredUnavailable should be
symmetric, checking up the context chain of the callee (this commit
only checks up the context chain of the caller). However, I couldn't
think of a testcase that didn't require first referencing the
unavailable type; this, we already diagnose.
rdar://problem/25030656
llvm-svn: 262921
to allow arbitrary data to be associated with a parameter.
Also, fix a bug where we apparently haven't been serializing
this information for the last N years.
llvm-svn: 262278
__global__ functions are present on both host and device side,
so providing __host__ or __device__ overloads is not going to
do anything useful.
llvm-svn: 261778
This is an artefact of split-mode CUDA compilation that we need to
mimic. HD functions are sometimes allowed to call H or D functions. Due
to split compilation mode device-side compilation will not see host-only
function and thus they will not be considered at all. For clang both H
and D variants will become function overloads visible to
compiler. Normally target attribute is considered only if C++ rules can
not determine which function is better. However in this case we need to
ignore functions that would not be present during current compilation
phase before we apply normal overload resolution rules.
Changes:
* introduced another level of call preference to better describe
possible call combinations.
* removed WrongSide functions from consideration if the set contains
SameSide function.
* disabled H->D, D->H and G->H calls. These combinations are
not allowed by CUDA and we were reluctantly allowing them to work
around device-side calls to math functions in std namespace.
We no longer need it after r258880.
Differential Revision: http://reviews.llvm.org/D16870
llvm-svn: 260697
For an explicit specialization, we first build a FunctionDecl, and then
we call SubstDecl() on it to build a second FunctionDecl, which has the
first FunctionDecl as canonical decl.
The address of an explicit specialization of function template used to be the
canonical decl of the FunctionDecl. This is different from all the other
DeduceTemplateArguments() calls in SemaOverload, and since the canonical decl
isn't visited by ParentMap while the redecl is, it also made ParentMap assert
when computing the parent of a address-of-explicit-specialization-fun-template.
To fix, remove the getCanonicalDecl() call. No behavior difference for clang,
but it fixes an assert in ParentMap (which is e.g. used by libTooling).
llvm-svn: 260159
-Wdelete-non-virtual-dtor warns if A is a type with virtual functions but
without virtual dtor has its constructor called via `delete a`. This makes the
warning also fire if the dtor is called via `a->~A()`. This would've found a
security bug in Chromium at compile time. Fixes PR26137.
To fix the warning, add a virtual destructor, make the class final, or remove
its other virtual methods. If you want to silence the warning, there's also
a fixit that shows how:
test.cc:12:3: warning: destructor called on 'B' ... [-Wdelete-non-virtual-dtor]
b->~B();
^
test.cc:12:6: note: qualify call to silence this warning
b->~B();
^
B::
http://reviews.llvm.org/D16206
llvm-svn: 257939
We were emitting diagnostics from our shiny new C-only overload
resolution mode. This patch attempts to silence all such diagnostics.
This fixes PR26085.
Differential Revision: http://reviews.llvm.org/D16159
llvm-svn: 257710
George Burgess IV