Commit Graph
994 Commits
Author SHA1 Message Date
Richard Smith b44ba47cf3 Don't treat dependent types as having a copy value representation. (#6055)
Add `Dependent` value and initializing representations for types whose
representations are unknown because they are dependent. When generating
SemIR in such cases, use a worst-case initializing representation that
both provides a destination address and also propagates a potential
result value.

Use this to fix incorrect lowering and lowering crashes for specific
functions involving generic types that don't use a copy value
representation.

In lowering, be careful to distinguish between whether the initializing
representation for the generic return type uses a return slot (which
affects whether the SemIR declaration and call have one) and whether the
initializing representation for the specific return type uses a return
slot (which affects whether the LLVM IR declaration and call have one).
2025-09-15 23:59:00 +00:00
Richard Smith ca40e9d693 Support making method calls to C++ overload sets. (#6069)
Fixes #6059
2025-09-15 23:41:15 +00:00
Dana Jansens 95b5cce9b4 Add tests that show .X and .Self.X are treated the same on the RHS of a rewrite constraint (#6056) 2025-09-15 18:11:05 +00:00
Jon Ross-PerkinsandDana Jansens 5e3bb523f8 Add builtin functions for destroy, with special requirements in facet types (#6035)
This is in support of a goal of changing the blanket `destroy` impl to
use (roughly):

```
private fn CanAggregateDestroy() -> type = "type.can_aggregate_destroy";

// Handles aggregate type destruction.
impl forall [AggregateDestroyT:! CanAggregateDestroy()] AggregateDestroyT as Destroy {
  fn Op[addr self: Self*]() = "type.aggregate_destroy";
}
```

That isn't done here because there's still other issues that migrating
raises. What this *does* do is add the builtin functions, and in
particular, support to `FacetTypeInfo` to make `CanAggregateDestroy`
work.

The "special requirement" approach in `FacetTypeInfo` allows us to
support restricting a blanket impl under the current approach of impls.
Maybe we'll find a cleaner approach that can work in the future, but
this fits into the current model by propagating similar to other
requirements. I'm using an enum mask because we have a number of similar
things to add (e.g. copy, move) but I'm not sure we need a full vector.

A few alternatives considered were:

- Supporting syntax more like `where .Self impls
TypeCanAggregateDestroy(.Self, SupportedInterface,
UnsupportedInterface)`. I think it'd be a little cleaner, but requires
better compile-time evaluation in order to assess the type of the call.
Right now it's expected to be a `FacetType` too early to make this work,
and I was concerned about pouring too much more time down this route.
- Providing an actual interface, in particular doing name lookup back
into `Core.` for an interface. This would've added name lookup overhead,
and the question of whether an `impl` exists.
- Generating an interface. This avoids the name lookup, but would still
raise the question of whether an `impl` should also be generated. Work
I've previously done generating interfaces for class destruction also
feels complex to both write and understand (an unfortunate issue).
- Still modeling as an `ImplsConstraint`, for example by defining a
special `InterfaceId::CanAggregateDestroy = -2` similar to what we do on
other ids. I was hesitant because of how this expands the number of
modes of `InterfaceId`, and things for consuming code to watch out for,
for what feels like a relatively niche set of use-cases that are only
interface-like.

---------

Co-authored-by: Dana Jansens <danakj@orodu.net>
2025-09-15 17:03:43 +00:00
Ivana Ivanovska 12ddfb9c7c [Carbon/C++ interop] Add support for C++ overloaded functions (#5891)
As proposed in [Carbon: C++ interop for overloaded functions and
function
templates](https://docs.google.com/document/d/1KUxumZtNe3mY3TsjW2s_ZADOlAaFlrtsLKHVILtqIaM/edit?tab=t.0),
Clang is used to perform the overload resolution using C++ rules, when
an overloaded C++ set is called from Carbon. Once a function is
selected, it's converted into a Carbon function and called using the
Carbon rules including argument conversions.

A single non-templated function is treated the same way as an overload
set and the same rules apply for its call.
Template functions are not supported yet.

Demo:

a) Non-templated function calls:

```c++
// --- overloads.h

auto foo(int a, short b) -> void;
auto foo(double a) -> void;
auto foo(int a) -> void;
```

```c++
// overloads.cpp

#include "overloads.h"
#include <cstdio>

auto foo(int a, short b) -> void {
  printf("hello from foo_int_short(%d, %d) \n", a, b);
}
auto foo(double a) -> void { printf("hello from foo_double(%f) \n", a); }
auto foo(int a) -> void { printf("hello from foo_int(%d) \n", a); }
```
```c++
library "Main";

import Cpp library "overloads.h";

fn Run() -> i32 {
  Cpp.foo(1.1 as f64);
  return 0;
}
```
```
$ clang -c overloads.cpp 
$ bazel-bin/toolchain/carbon compile main.carbon
$ bazel-bin/toolchain/carbon link overloads.o main.o --output=demo
$ ./demo
hello from foo_double(1.100000) 
```

b) Constructors:
```c++
// --- constructor_overloads.h
class C {
 public:
  C();
  C(int a, int b);
};
```

```c++
// constructor_overloads.cpp
#include "constructor_overloads.h"
#include <cstdio>

C::C() { printf("hello from C() \n"); }
C::C(int a, int b) { printf("hello from C(%d, %d) \n", a, b); }
```
```c++
library "Main";

import Cpp library "constructor_overloads.h";

fn Run() -> i32 {
  let c1: Cpp.C = Cpp.C.C();
  let c2: Cpp.C = Cpp.C.C(1, 2);
  return 0;
}
```

```
$ clang -c constructor_overloads.cpp 
$ bazel-bin/toolchain/carbon compile main.carbon
$ bazel-bin/toolchain/carbon link constructor_overloads.o main.o \--output=demo
$ ./demo
hello from C() 
hello from C(1, 2) 
```


Follow-ups:

- `Cpp.foo({})` - proper handling of struct literals as call args.
- Fix access for overloaded sets.
- Fix tests:
- Method calls: `error: missing object argument in method call
[MissingObjectInMethodCall]` in tests.
    - Fix `toolchain/check/testdata/interop/cpp/import.carbon` test.
    - Fix `enums` support.
    - Fix `str` -> `std::string_view` mapping.


Part of #5915
2025-09-15 12:29:49 +00:00
Elliott KaltandRichard Smith f4bd6e42f9 Replace impl fn with override fn (#6008)
This proposal renames the syntax used to mark an overriding definition
of a virtual method from `impl fn` to `override fn` to avoid ambiguity:
besides indicating an overriding virtual function, it can be parsed as
an "impl" declaration when the construct following "impl" begins with a
lambda introduced by "fn".

Closes #5711

---------

Co-authored-by: Richard Smith <richard@metafoo.co.uk>
2025-09-12 23:27:02 +00:00
Boaz Brickner 508a88e2a9 C++ inteop: Set type source info for a generated C++ thunk function (#6049)
This prevents a null pointer access crash when generating a thunk with
an automatically deduced trivial return type.
In this case, Clang calls `Sema::DeduceFunctionTypeFromReturnExpr()`
which calls `Sema::getReturnTypeLoc()`, which requires this information.

Part of #5514.
2025-09-12 07:11:36 +00:00
Richard Smith d60900cbeb Remove special case for returning value expressions by copy (#6052)
When returning a value from a function whose return type has a by-copy
initializing representation, perform initialization like we do when the
return type has an in-place initializing representation. This makes our
SemIR representation more uniform, as the return expression will now
always be an initializing expression rather than a value expression, but
more importantly it means that attempts to return a non-copyable type by
value now fail, even if the type has a by-copy initializing
representation.

This catches a bunch of places where we were returning a value of an
unconstrained template parameter `T:! type`, which we were incorrectly
allowing because we didn't notice it was not copyable. Unfortunately
this then requires quite a few test updates.

Like #6034, this exposes a lowering issue where lowering crashes when
attempting to lower a specific copy operation for certain types; a
couple more tests are temporarily disabled here. An upcoming PR
dependent on this one will fix the issue and re-enable those tests.
2025-09-12 00:13:33 +00:00
Richard SmithandGeoff Romer 1ec8ac7ef9 Add Copy interface and use it for making copies. (#6034)
Instead of hardcoding which types are copyable, add a `Core.Copy`
interface to perform copying. Move almost all the current copy support
to that interface. Some remaining pieces are still using builtin logic
after this PR:

* For tuples and structs, builtin logic is used to perform elementwise
copies. This also supports copying *adapters of* tuples and structs,
which seems like it may not be desirable, especially for non-extending
adapters. A `Copy` impl is provided for tuples of at most 2 elements, so
that `Core.Copy` constraints are satisfied, but we can't implement this
generally until we have variadics support, and don't yet have a
mechanism to generalize this to structs.
* For `enum` types imported from C++, builtin logic is used to perform a
copy. This is temporary until we have a mechanism to identify these
types from an impl in the prelude.

One lowering test in `toolchain/lower/testdata/class/generic.carbon` is
disabled for now, as it causes a crash in the lowering code due to an
ABI mismatch between the call signature in the lowered declaration of a
specific function and the call that is generated in the specific callee.
Fixing this is a little involved, and will be done in a separate PR.

---------

Co-authored-by: Geoff Romer <gromer@google.com>
2025-09-10 23:55:55 +00:00
Richard Smith e8cd229e74 When performing an impl lookup, only import impls for related interfaces. (#6040)
This avoids impl lookups involving, say, `Core.Int` pulling in all ~65
impls in "prelude/types/int", which resulted in a lot of unnecessary
importing work, followed by a lot of unnecessary inst namer and inst
formatter work.

Before:
```
Ran 1335 tests in 6186 ms wall time, 146818 ms across threads
  Slowest tests:
  - toolchain/check/testdata/interop/cpp/function/arithmetic_types_bridged.carbon: 5611 ms, 5532 ms in Run
  - toolchain/check/testdata/interop/cpp/function/operators.carbon: 2034 ms, 1981 ms in Run
  - toolchain/check/testdata/primitives/import_symbolic.carbon: 1796 ms, 1786 ms in Run
  - toolchain/lower/testdata/operators/arithmetic.carbon: 1729 ms, 1728 ms in Run
  - toolchain/lower/testdata/function/generic/call_recursive_sccs_deep.carbon: 1700 ms, 1697 ms in Run
[==========] 1335 tests from 1 test suite ran. (682 ms total)
```

After:
```
Ran 1335 tests in 2419 ms wall time, 109587 ms across threads
  Slowest tests:
  - toolchain/check/testdata/interop/cpp/function/arithmetic_types_bridged.carbon: 1748 ms, 1665 ms in Run
  - toolchain/check/testdata/interop/cpp/function/operators.carbon: 1106 ms, 1057 ms in Run
  - toolchain/lower/testdata/function/generic/call_recursive_diamond.carbon: 1044 ms, 1041 ms in Run
  - toolchain/lower/testdata/function/generic/call_recursive_sccs_deep.carbon: 1015 ms, 1012 ms in Run
  - toolchain/lower/testdata/operators/arithmetic.carbon: 998 ms, 997 ms in Run
[==========] 1335 tests from 1 test suite ran. (652 ms total)
```

That's still slower than it should be, but a large improvement
nonetheless.

Fixes #6029
2025-09-10 21:40:27 +00:00
Dana Jansens ed43fd2c1c Give the BindSymbolicName for .Self in a binding pattern a FacetType type (#6036)
If it's just `TypeType` then the `BindSymbolicName` appears directly in
type positions, but if it is replaced with another facet value, then we
would need to insert a `FacetAccessType` around it. By giving it a
`FacetType` type, like other `BindSymbolicName`s we make it consistent
and avoid having to introduce extra instructions.
2025-09-10 17:57:08 +00:00
Boaz Brickner d6fbe3c663 C++ interop: Support importing operators defined in namespaces (#6024)
C++ Interop Demo:

```c++
// my_number.h

namespace MyNamespace {

class MyNumber {
 public:
  explicit MyNumber(int value) : value_(value) {}
  auto value() const -> int { return value_; }

 private:
  int value_;
};

auto operator+(MyNumber lhs, MyNumber rhs) -> MyNumber;

}  // namespace MyNamespace
```

```c++
// my_number.cpp

#include "my_number.h"

namespace MyNamespace {

auto operator+(MyNumber lhs, MyNumber rhs) -> MyNumber {
  return MyNumber(lhs.value() + rhs.value());
}

}  // namespace MyNamespace
```

```carbon
// main.carbon

library "Main";

import Core library "io";
import Cpp library "my_number.h";

fn Run() -> i32 {
  let n1: Cpp.MyNamespace.MyNumber = Cpp.MyNamespace.MyNumber.MyNumber(5);
  Core.Print(n1.value());
  let n2: Cpp.MyNamespace.MyNumber = Cpp.MyNamespace.MyNumber.MyNumber(7);
  Core.Print(n2.value());
  let n3: Cpp.MyNamespace.MyNumber = n1 + n2;
  Core.Print(n3.value());
  return 0;
}
```

Before this change:
```
$ bazel-bin/toolchain/carbon compile main.carbon
main.carbon:13:38: error: cannot access member of interface `Core.AddWith(Cpp.MyNamespace.MyNumber)` in type `Cpp.MyNamespace.MyNumber` that does not implement that interface
  let n3: Cpp.MyNamespace.MyNumber = n1 + n2;
                                     ^~~~~~~
```

With this change:

```shell
$ clang -c my_number.cpp
$ bazel-bin/toolchain/carbon compile main.carbon
$ bazel-bin/toolchain/carbon link my_number.o main.o --output=demo
$ ./demo
5
7
12
```

Part of https://github.com/carbon-language/carbon-lang/issues/5995.
2025-09-10 14:03:09 +00:00
Jon Ross-Perkins 0518fdebbc Fix potential fingerprint conflict in constraints (#6033)
This uses each vector's size as a barrier between lists, to eliminate
the possibility of incidental collisions between entries of different
lists. This is the same as is done inside `AddBlock`.
2025-09-09 21:44:03 +00:00
Boaz Brickner 56adfa20ce C++ interop: Add a test for calling an operator on an inner class (#6030)
This currently works and I'd like to keep it that way when adding
operators in namespace support.

Part of #5995.
2025-09-09 13:06:55 +00:00
Boaz Brickner 471b394c6d C++ interop: Support unary operators (#6020)
Newly supported: `-`.
Partially supported due to lack of reference support: `++` (prefix),
`--` (prefix).
Not supported due to lack of Carbon support to call them correctly: `+`,
`++` (postfix), `--` (postfix), `~`, `!`, `&`, `*`, `->`.

Also (for consistency):
* Add the operator declarations to unsupported binary operators tests.
* Logical operators and the unary `operator&` (address of) are expected
to be called by explicitly calling `operatorX`.

C++ Interop Demo:

```c++
// my_number.h

class MyNumber {
 public:
  explicit MyNumber(int value) : value_(value) {}
  auto value() const -> int { return value_; }

 private:
  int value_;
};

auto operator++(MyNumber operand) -> MyNumber;
auto operator--(MyNumber operand) -> MyNumber;
auto operator-(MyNumber operand) -> MyNumber;
```

```c++
// my_number.cpp

#include "my_number.h"

auto operator++(MyNumber operand) -> MyNumber {
  return MyNumber(operand.value() + 1);
}

auto operator--(MyNumber operand) -> MyNumber {
  return MyNumber(operand.value() - 1);
}

auto operator-(MyNumber operand) -> MyNumber {
  return MyNumber(-operand.value());
}
```

```carbon
// main.carbon

library "Main";

import Core library "io";
import Cpp library "my_number.h";

fn Run() -> i32 {
  var num: Cpp.MyNumber = Cpp.MyNumber.MyNumber(14);
  Core.Print(num.value());
  ++num;
  Core.Print(num.value());
  --num;
  Core.Print(num.value());
  num = -num;
  Core.Print(num.value());
  return 0;
}
```

```shell
$ clang -c my_number.cpp
$ bazel-bin/toolchain/carbon compile main.carbon
$ bazel-bin/toolchain/carbon link my_number.o main.o --output=demo
$ ./demo
14
14
14
-14
```

Part of https://github.com/carbon-language/carbon-lang/issues/5995.
2025-09-08 15:39:08 +00:00
Boaz Brickner b88b53e7e3 C++ interop: Add support for importing globals (#6005)
This supports importing globals in the global scope and within
namespaces, and support for class static data members.

C++ Interop Demo:

```c++
// my_global.h

extern int my_global;
void inc_my_global();
```

```c++
// my_global.cpp

#include "my_global.h"

int my_global = 5;

void inc_my_global() {
  ++my_global;
}
```

```carbon
// main.carbon

library "Main";

import Core library "io";
import Cpp library "my_global.h";

fn Run() -> i32 {
  Core.Print(Cpp.my_global);
  Cpp.inc_my_global();
  Core.Print(Cpp.my_global);
  return 0;
}
```

```shell
$ clang -c my_global.cpp
$ bazel-bin/toolchain/carbon compile main.carbon
$ bazel-bin/toolchain/carbon link my_global.o main.o --output=demo
$ ./demo
5
6
```

Part of #6006.
2025-09-08 08:34:55 +00:00
Boaz Brickner ee42b2db93 C++ interop: Support more binary operators (#6017)
Already supported: `+`.
Newly supported: `-`, `*`, `/`, `%`, `&`, `|`, `^`, `<<`, `>>`, `==`,
`!=`, `<`, `>`, `<=`, `>=`.
Partially supported due to lack of reference support: `+=`, `-=`, `*=`,
`/=`, `%=`, `&=`, `|=`, `^=`.
Not supported due to lack of reference support: `<<=`, `>>=`.
Not supported (I think Carbon doesn't want overloading these): `&&`,
`||`.

C++ Interop Demo:

```c++
// my_number.h

class MyNumber {
 public:
  explicit MyNumber(int value) : value_(value) {}
  auto value() const -> int { return value_; }
  void set_value(int value) { value_ = value; }

 private:
  int value_;
};

// Arithmetic
auto operator+(MyNumber lhs, MyNumber rhs) -> MyNumber;
auto operator-(MyNumber lhs, MyNumber rhs) -> MyNumber;
auto operator*(MyNumber lhs, MyNumber rhs) -> MyNumber;
auto operator/(MyNumber lhs, MyNumber rhs) -> MyNumber;
auto operator%(MyNumber lhs, MyNumber rhs) -> MyNumber;

// Bitwise
auto operator&(MyNumber lhs, MyNumber rhs) -> MyNumber;
auto operator|(MyNumber lhs, MyNumber rhs) -> MyNumber;
auto operator^(MyNumber lhs, MyNumber rhs) -> MyNumber;
auto operator<<(MyNumber lhs, int shift) -> MyNumber;
auto operator>>(MyNumber lhs, int shift) -> MyNumber;

// Compound Arithmetic
auto operator+=(MyNumber* _Nonnull lhs, MyNumber rhs) -> MyNumber* _Nonnull;
auto operator-=(MyNumber* _Nonnull lhs, MyNumber rhs) -> MyNumber* _Nonnull;
auto operator*=(MyNumber* _Nonnull lhs, MyNumber rhs) -> MyNumber* _Nonnull;
auto operator/=(MyNumber* _Nonnull lhs, MyNumber rhs) -> MyNumber* _Nonnull;
auto operator%=(MyNumber* _Nonnull lhs, MyNumber rhs) -> MyNumber* _Nonnull;

// Compound Bitwise
auto operator&=(MyNumber* _Nonnull lhs, MyNumber rhs) -> MyNumber* _Nonnull;
auto operator|=(MyNumber* _Nonnull lhs, MyNumber rhs) -> MyNumber* _Nonnull;
auto operator^=(MyNumber* _Nonnull lhs, MyNumber rhs) -> MyNumber* _Nonnull;

// Relational
auto operator==(MyNumber lhs, MyNumber rhs) -> bool;
auto operator!=(MyNumber lhs, MyNumber rhs) -> bool;
auto operator<(MyNumber lhs, MyNumber rhs) -> bool;
auto operator>(MyNumber lhs, MyNumber rhs) -> bool;
auto operator<=(MyNumber lhs, MyNumber rhs) -> bool;
auto operator>=(MyNumber lhs, MyNumber rhs) -> bool;
```

```c++
// my_number.cpp

#include "my_number.h"

// Arithmetic
auto operator+(MyNumber lhs, MyNumber rhs) -> MyNumber {
  return MyNumber(lhs.value() + rhs.value());
}
auto operator-(MyNumber lhs, MyNumber rhs) -> MyNumber {
  return MyNumber(lhs.value() - rhs.value());
}
auto operator*(MyNumber lhs, MyNumber rhs) -> MyNumber {
  return MyNumber(lhs.value() * rhs.value());
}
auto operator/(MyNumber lhs, MyNumber rhs) -> MyNumber {
  return MyNumber(lhs.value() / rhs.value());
}
auto operator%(MyNumber lhs, MyNumber rhs) -> MyNumber {
  return MyNumber(lhs.value() % rhs.value());
}

// Bitwise
auto operator&(MyNumber lhs, MyNumber rhs) -> MyNumber {
  return MyNumber(lhs.value() & rhs.value());
}
auto operator|(MyNumber lhs, MyNumber rhs) -> MyNumber {
  return MyNumber(lhs.value() | rhs.value());
}
auto operator^(MyNumber lhs, MyNumber rhs) -> MyNumber {
  return MyNumber(lhs.value() ^ rhs.value());
}
auto operator<<(MyNumber lhs, int shift) -> MyNumber {
  return MyNumber(lhs.value() << shift);
}
auto operator>>(MyNumber lhs, int shift) -> MyNumber {
  return MyNumber(lhs.value() >> shift);
}

// Compound Arithmetic
auto operator+=(MyNumber* _Nonnull lhs, MyNumber rhs) -> MyNumber* _Nonnull {
  return &(*lhs = *lhs + rhs);
}
auto operator-=(MyNumber* _Nonnull lhs, MyNumber rhs) -> MyNumber* _Nonnull {
  return &(*lhs = *lhs - rhs);
}
auto operator*=(MyNumber* _Nonnull lhs, MyNumber rhs) -> MyNumber* _Nonnull {
  return &(*lhs = *lhs * rhs);
}
auto operator/=(MyNumber* _Nonnull lhs, MyNumber rhs) -> MyNumber* _Nonnull {
  return &(*lhs = *lhs / rhs);
}
auto operator%=(MyNumber* _Nonnull lhs, MyNumber rhs) -> MyNumber* _Nonnull {
  return &(*lhs = *lhs % rhs);
}

// Compound Bitwise
auto operator&=(MyNumber* _Nonnull lhs, MyNumber rhs) -> MyNumber* _Nonnull {
  return &(*lhs = *lhs & rhs);
}
auto operator|=(MyNumber* _Nonnull lhs, MyNumber rhs) -> MyNumber* _Nonnull {
  return &(*lhs = *lhs | rhs);
}
auto operator^=(MyNumber* _Nonnull lhs, MyNumber rhs) -> MyNumber* _Nonnull {
  return &(*lhs = *lhs ^ rhs);
}

// Relational
auto operator==(MyNumber lhs, MyNumber rhs) -> bool {
  return lhs.value() == rhs.value();
}
auto operator!=(MyNumber lhs, MyNumber rhs) -> bool {
  return lhs.value() != rhs.value();
}
auto operator<(MyNumber lhs, MyNumber rhs) -> bool {
  return lhs.value() < rhs.value();
}
auto operator>(MyNumber lhs, MyNumber rhs) -> bool {
  return lhs.value() > rhs.value();
}
auto operator<=(MyNumber lhs, MyNumber rhs) -> bool {
  return lhs.value() <= rhs.value();
}
auto operator>=(MyNumber lhs, MyNumber rhs) -> bool {
  return lhs.value() >= rhs.value();
}
```

```carbon
// main.carbon

library "Main";

import Core library "io";
import Cpp library "my_number.h";

fn PrintBool(b: bool) {
  if (b) {
    Core.Print(1);
  } else {
    Core.Print(0);
  }
}

fn Run() -> i32 {
  // Arithmetic
  var num1: Cpp.MyNumber = Cpp.MyNumber.MyNumber(14);
  var num2: Cpp.MyNumber = Cpp.MyNumber.MyNumber(5);
  Core.Print(num1.value());
  Core.Print(num2.value());
  Core.Print((num1 + num2).value());
  Core.Print((num1 - num2).value());
  Core.Print((num1 * num2).value());
  Core.Print((num1 / num2).value());
  Core.Print((num1 % num2).value());

  // Bitwise
  var bits1: Cpp.MyNumber = Cpp.MyNumber.MyNumber(12);
  var bits2: Cpp.MyNumber = Cpp.MyNumber.MyNumber(10);
  Core.Print(bits1.value());
  Core.Print(bits2.value());
  Core.Print((bits1 & bits2).value());
  Core.Print((bits1 | bits2).value());
  Core.Print((bits1 ^ bits2).value());
  Core.Print((bits1 << 2).value());
  Core.Print((bits1 >> 1).value());

  // Compound Arithmetic
  var c: Cpp.MyNumber = Cpp.MyNumber.MyNumber(100);
  Core.Print(c.value());
  &c += Cpp.MyNumber.MyNumber(10);
  Core.Print(c.value());
  &c -= Cpp.MyNumber.MyNumber(20);
  Core.Print(c.value());
  &c *= Cpp.MyNumber.MyNumber(2);
  Core.Print(c.value());
  &c /= Cpp.MyNumber.MyNumber(6);
  Core.Print(c.value());
  &c %= Cpp.MyNumber.MyNumber(9);
  Core.Print(c.value());

  // Compound Bitwise
  &c |= Cpp.MyNumber.MyNumber(12);
  Core.Print(c.value());
  &c &= Cpp.MyNumber.MyNumber(7);
  Core.Print(c.value());
  &c ^= Cpp.MyNumber.MyNumber(10);
  Core.Print(c.value());

  // Relational
  var rel1: Cpp.MyNumber = Cpp.MyNumber.MyNumber(20);
  var rel2: Cpp.MyNumber = Cpp.MyNumber.MyNumber(30);
  var rel3: Cpp.MyNumber = Cpp.MyNumber.MyNumber(20);
  Core.Print(rel1.value());
  Core.Print(rel2.value());
  Core.Print(rel3.value());
  PrintBool(rel1 == rel3);
  PrintBool(rel1 != rel2);
  PrintBool(rel1 < rel2);
  PrintBool(rel2 > rel1);
  PrintBool(rel1 <= rel3);
  PrintBool(rel1 >= rel2);

  return 0;
}
```

```shell
$ clang -c my_number.cpp
$ bazel-bin/toolchain/carbon compile main.carbon
$ bazel-bin/toolchain/carbon link my_number.o main.o --output=demo
$ ./demo
14
5
19
9
70
2
4
12
10
8
14
6
48
6
100
110
90
180
30
3
15
7
13
20
30
20
1
1
1
1
1
0
```

Part of https://github.com/carbon-language/carbon-lang/issues/5995.
2025-09-05 13:03:10 +00:00
Richard Smith db0a00d713 Fix double-destruction of temporaries. (#6010)
Attach the cleanup to the `Temporary` instruction instead of to the
`TemporaryStorage` instruction. We create `TemporaryStorage`
instructions speculatively when creating an initializing expression, and
may overwrite those instructions with other instructions if it turns out
that a temporary is not required. Instead, wait until we finalize the
temporary and create a `Temporary` instruction to register the cleanup.
2025-09-04 19:19:59 +00:00
Richard Smith f943f31e41 Allow a value of type MaybeUnformed(T) to convert to T with unsafe as (#6014)
We already allowed this for reference expressions; this extends the
support to also cover value expressions. This requires a little more
work because the value representation of `T` and `MaybeUnformed(T)`
don't necessarily match in general.
2025-09-04 19:10:28 +00:00
Dana JansensandJon Ross-Perkins 70f104aa40 Always import canonical instructions except for exceptional Decl cases (#6009)
There are a few instructions that import in multiple phases, which
receive the `const_id` and use it to construct multiple constants until
building the final constant value. These include
`AssociatedConstantDecl`, `FunctionDecl`, and `InterfaceDecl`.

Other instructions just construct a constant value in a single attempt,
once all their dependencies are imported. For these instruction types,
avoid importing the non-canonical instruction. Always get the canonical
constant instruction and import that.

Since the constant value of an instruction can have a very different
structure than its non-canonical value, this ensures import has a
consistent structure to work with, by only working with canonical values
as much as possible.

The `VtableDecl` and `VtablePtr` were set up to pass along `const_id`
but do not actually require multiple phases, so they have been changed
to stop passing along the unused (and always empty) `const_id`.

---------

Co-authored-by: Jon Ross-Perkins <jperkins@google.com>
2025-09-03 22:36:27 +00:00
Richard Smith cb5e2e1597 Improve support for qualification conversions. (#5999)
* Treat `MaybeUnformed` and `partial` as qualifiers, like `const`.
* Allow pointer conversions to add qualifiers.
* Allow unsafe pointer conversions to remove qualifiers.
* Allow conversions on non-reference expressions to drop `const`.
* Allow unsafe conversions on any expression to drop `const`.
* Allow unsafe conversions on non-initializing expressions to drop
  `partial`. For initializing expressions, we should initialize the
  vptr when dropping `partial`; this is not yet supported so we reject.
* Allow conversions on reference expressions to add `MaybeUnformed`.
* Allow unsafe conversions on reference expressions to drop
  `MaybeUnformed`. For non-reference expressions, additional work is
  required, because the value / initializing representation may not
  match between `T` and `MaybeUnformed(T)`, so those are rejected for
  now.
2025-09-03 21:00:12 +00:00
Boaz Brickner 870c5380a0 C++ interop: Support importing binary operator+ (#5996)
Triggered by calling a binary operator with LHS being an imported C++
class type.

Not supported (yet):
* Multiple overloads.
* Other operators.

C++ Interop Demo:

```c++
// hello_world.h

class C {
 public:
  C(int x) : x_(x) {}
  auto x() const -> int { return x_; }

 private:
  int x_ = 0; 
};

auto operator+ (C c1, C c2) -> C;
```

```c++
// hello_world.cpp

#include "hello_world.h"

#include <cstdio>

auto operator+ (C c1, C c2) -> C {
  printf("Adding %d with %d\n", c1.x(), c2.x());
  return C(c1.x() + c2.x());
}
```

```carbon
// main.carbon

library "Main";

import Cpp library "hello_world.h";

fn Run() -> i32 {
  let c1 : Cpp.C = Cpp.C.C(7);
  let c2 : Cpp.C = Cpp.C.C(8);
  let c3 : Cpp.C = c1 + c2;
  let c4 : Cpp.C = c3 + c2;
  return 0;
}

```

```shell
$ clang -c hello_world.cpp
$ bazel-bin/toolchain/carbon compile main.carbon
$ bazel-bin/toolchain/carbon link hello_world.o main.o --output=demo
$ ./demo
Adding 7 with 8
Adding 15 with 8
```

Part of #5995.
2025-09-03 12:01:20 +00:00
Elliott Kalt 58de34e534 Decouple associated constants from let (#5973)
Decouples associated constants from being special cased in let handlers.
Enforces associated constant grammar restrictions in parsing instead of
checking.

Closes #5411
2025-09-02 23:15:26 +00:00
Richard Smith 0e6dd7e701 Add MaybeUnformed(T) type. (#5989)
This type has the same object representation as `T`, but always uses a
pointer type as its value representation. No other semantics are
provided for it yet.
2025-09-02 20:50:49 +00:00
Richard SmithandDana Jansens 4483d1e5a7 Recover better from invalid C++ classes. (#5992)
When importing a class definition, don't ask for the class layout if the
definition is invalid. Avoids an assertion failure in Clang.

---------

Co-authored-by: Dana Jansens <danakj@orodu.net>
2025-09-02 19:32:35 +00:00
Boaz Brickner b5d86fdb6f Properly dump SemIR for inline C++ imports (#6001)
Dumping SemIR crashed on inline C++ imports and this outputs `import Cpp
inline` instead.
Followup of #5904.
2025-08-29 21:36:10 +00:00
Boaz Brickner 8adb3570ac C++ Interop: Add support for char (#5988)
Added tests for different character types.
`char` is currently not in primitives prelude, so had to use full
prelude.

C++ Interop Demo:

```carbon
// main.carbon

library "Main";

import Cpp inline '''
auto output_char(char c) -> void {
  printf("%c", c);
}
''';

fn Run() -> i32 {
  let msg: array(Core.Char, 13) =
      ('H', 'e', 'l', 'l', 'o', ' ', 'w', 'o', 'r', 'l', 'd', '!', '\n');
  for (c: Core.Char in msg) {
    Cpp.output_char(c);
  }
  return 0;
}
```

```shell
$ bazel-bin/toolchain/carbon compile main.carbon
$ bazel-bin/toolchain/carbon link main.o --output=demo
$ ./demo
Hello world!
```

Part of https://github.com/carbon-language/carbon-lang/issues/5263.
2025-08-29 01:11:24 +00:00
Richard Smith 50e5476ee3 Add support for unsafe as operator to the toolchain. (#5993)
Following the direction of #5913, add support for parsing an `unsafe as`
operator. For now, we allow one additional conversion using `unsafe as`
beyond the conversions supported by `as`: we permit pointer conversions
that remove qualifiers, such as `const T*` -> `T*`.
2025-08-28 23:00:53 +00:00
David Blaikie 1d0f30df3a Fix a use of an imported InstId used where a local InstId is required (#5998)
Found by WIP validation for this type of issue ongoing in #5997

I'm not entirely sure how the one test update falls out of this change -
but it is from the same test that I originally reduced the problem from,
which is reassuring.

The reduced test case I investigated the issue with was this:
`a.carbon`:
```
library "lib";
interface I1(Other:! type) {
   let Result:! type;
}
```
`b.carbon`:
```
import library "lib";
class T1 { }
impl T1 as I1(Self) where .Result = Self { }
```
The SemIR dump diff looked like this:
```
89c89
<   %Main.import_ref.b6f = import_ref Main//lib, inst28 [no loc], unloaded
---
>   %Main.import_ref.b6f = import_ref Main//lib, inst27 [no loc], unloaded
96c96
<   %Main.import_ref.f7b: @I1.%I1.type (%I1.type.e87) = import_ref Main//lib, inst28 [no loc], loaded [symbolic = @I1.%Self (constants.%Self.c47)]
---
>   %Main.import_ref.f7b: @I1.%I1.type (%I1.type.e87) = import_ref Main//lib, inst27 [no loc], loaded [symbolic = @I1.%Self (constants.%Self.c47)]
```
Which is a difference, but given the `inst28`/`inst27` don't appear
anywhere else than these two lines, it doesn't give a terribly
meaningful diff/story about what changed - but perhaps it's
sufficient...

Not sure if this test ^ is sufficiently more interesting than the diff
update already in this patch. If so, happy to add the above as a new
test case.

Open to ideas.
2025-08-28 20:02:26 +00:00
Richard SmithandChandler Carruth bd90fe1d9b Interop: map C++ std::string_view into Carbon str when importing. (#5985)
We assume these types have the same representation. For now, that will
only be the case for libc++ on 64-bit targets, because libc++ puts the
size field first, and `Core.String` always uses a 64-bit size field even
on 32-bit targets.

---------

Co-authored-by: Chandler Carruth <chandlerc@gmail.com>
2025-08-27 01:02:39 +00:00
Richard Smith 82ba1a43a1 Support for importing C++ enum types. (#5978)
We import C++ enum types as Carbon class types as adapters for the
corresponding builtin integer type, and we import enumerator constants
as integer constants of that class type.

No operators are supported on such values for now; eventually once we
start asking Clang to implement operators on C++-owned types, these
types should be handled in the same way. However, they can be converted
to the corresponding integer type with `as` via adapter conversion, and
integer builtin functions can operate on them.
2025-08-26 23:11:35 +00:00
Richard Smith 3533668186 Support for building thunks for C++ constructors. (#5977) 2025-08-26 21:49:11 +00:00
Richard Smith d37f1ae6b5 Add return value support to C++ thunks. (#5976)
Based on #5948. A couple of tricky parts:

* When generating the C++ side of the thunk, we are given a pointer to
the location to emplace the return value. The only mechanism C++
provides to perform this emplacement is using placement `operator new`,
which requires a library function in the `<new>` header. We handle this
by declaring that library function ourselves, and rely on Clang not
actually needing a definition for it (which the standard library owns).

* On the Carbon side of the thunk, we want to form an initializing
expression as the result of the call. We don't have a way of expressing
in SemIR that an initializing expression performs its initialization by
storing through a pointer, so this PR adds a new initializing
instruction, `InPlaceInit`, to model an initialization that's performed
opaquely in-place.
2025-08-26 02:23:38 +00:00
Richard Smith 1331ade57f Attempt to complete the source type in a conversion. (#5984)
This is necessary if the source type is an adapter, as we would not
otherwise be able to determine what type it adapts and hence could be
converted to.
2025-08-26 00:33:12 +00:00
Richard Smith 8c9080801c Support for building thunks for C++ methods. (#5972)
Also tweak how we import C++ methods to properly handle C++23's explicit
object parameters.
2025-08-25 22:33:51 +00:00
Richard Smith ddafbc9331 Use direct passing for 32- and 64-bit unsigned integers. (#5980)
Previously we only avoided creating a thunk for signed integers. But the
same logic also applies to the unsigned 32-bit and 64-bit types.
2025-08-25 21:49:03 +00:00
Richard Smithandjosh11b 30b8a93fde Support conversion from T* to const T*. (#5971)
Also support conversion from Derived* to const Base*.

---------

Co-authored-by: josh11b <15258583+josh11b@users.noreply.github.com>
2025-08-20 23:27:03 +00:00
Richard Smith 2352e93bb5 Remove now-unused StringType instruction. (#5964) 2025-08-20 18:30:11 +00:00
Richard Smith b72c11e94a Support importing nested types from C++. (#5955)
Fix the algorithm for importing declarations in dependency order to
properly walk the dependency graph. Add the parent declaration of a
declaration to the dependency set so that we have a parent declaration
context to import a declaration into.

Fixes a crash when attempting to import a class whose parent is not
imported.
2025-08-20 07:39:12 +00:00
Richard Smith 41ed82e033 Add basic support for strings to core, check, and lower. (#5963)
Add a `Core.String` class to the prelude representing a string view, and
rename the `String` keyword to `str` and make it evaluate to
`Core.String`.

`Core.String` is represented as a pair of a pointer to a character
(actually, to the first character of a string, but we don't have a way
of modeling that yet) and a size (which should be pointer-width, but is
currently always a `u64` as we don't have a `usize` equivalent yet).
`Core.String` values are generated directly by the toolchain for string
literal expressions.

This follows the direction established at the recent summit, but the
design implemented here has not been through the proposal process yet.
2025-08-20 00:18:50 +00:00
Jon Ross-Perkins 8d08e774fc Add a feature to explicitly include a file's SemIR (#5961)
Trying to figure out an easy way to debug semir in the prelude, #5703
removed an option to set `--exclude-dump-file-prefix` to empty. But,
this is probably an improvement over that flow... With this change, it's
possible to add `//@dump-sem-ir-file` to a specific prelude file, and
its full IR will be printed. Additionally, it becomes an option with the
default `--dump-sem-ir-ranges=only` to add `//@dump-sem-ir-file` and get
the full file's IR.
2025-08-15 18:53:43 +00:00
David BlaikieandRichard Smith 3f9fc633fe Add a vtableDecl inst and use that in classes instead of VtablePtr (#5945)
This addresses/avoids the duplicate import of vtables.

I went through a few iterations/etc along the way and left them in the
commit
history for the PR in case any of them are useful to illustrate how I
got here,
or worth revisiting.

Essentially I ended up with a circularity in importing - importing the
class
imported the vtable_decl which imported the virtual functions - and then
pending
specifics of the virtual functions needed the self specific of the
enclosing
class which wasn't ready yet.

Adding ImportRef to the vtable_decl to break the cycle caused me trouble
when
naming the vtable_decl instructions - so I tried making the functions in
the
vtable unloaded ImportRefs instead. That worked, but meant that
importing a
class still was doing O(number of vtable entries) even if the vtable
wasn't
used.

So I revisited the lazy vtable_decl - figured out how to make the naming
work
(when building the vtable_ptr, even though the vtable_decl doesn't have
to be
loaded for the vtable_ptr, I force it to be loaded anyway, to load the
vtable so
it's usable by lowering, etc). And then I could go back to the old
non-lazy
loaded vtable entries (using some loaded ImportRefs in the cases where
we needed
them/had already adopted them).

Then thinking about the VtablePtr instruction, went back/forth on
exactly what
it needed - went from VtablePtr's member being a VtableDecl InstId, to a
ClassId, then back to a VtableId as it was before this patch.

Naming the instructions has one oddity, that the VtableDecl and
VtablePtr
instructions seem to need to add the pending name for the VtableId -
despite not
using the VtableId in their own name - should the inst namer be doing
this work
for parameters of instructions rather than requiring the inst to do it
deliberately? (or am I holding it wrong in some way?)

---------

Co-authored-by: Richard Smith <richard@metafoo.co.uk>
2025-08-15 18:36:54 +00:00
Richard Smith 7727c62880 Enforce a couple of char literal restrictions from #1964: (#5960)
* `\x` escapes are not permitted in character literals
* ASCII control characters (U+0000 .. U+001F) are not permitted in
character literals unless specified with escape sequences.
2025-08-15 00:45:16 +00:00
Richard Smith ae54873441 Change tests in check/testdata/builtins/char to be tests for the builtin char functions. (#5956)
Follow the pattern used by other tests in check/testdata/builtins.
2025-08-14 20:58:09 +00:00
Richard Smith b851e8c423 Add support for f16, f64, f128. (#5952)
Generalize the f64 support to support other sizes. Also provide interop
support for `float`, `_Float16`, and `__float128`.

Also lay some groundwork for non-standard floating-point types, though
we don't have any syntax to name them yet.
2025-08-14 01:14:40 +00:00
Jon Ross-Perkins cfe5599144 Support imports of more literal values (#5954)
I noticed while trying to set up an associated constant in the prelude
that we weren't supporting bool value imports; this goes through and
addresses support for simple builtin types.

Array initialization fails on declaration, which seems like a bug but
I'm only documenting it here.

Also fix missing export of `Core.FloatLiteral`

I checked and this doesn't seem to affect #5952, which is doing more
float changes.
2025-08-13 23:36:40 +00:00
Richard SmithandJon Ross-Perkins b2b0b4a73f Improve recovery from bad type imports. (#5953)
The main change here is that a bad type appearing somewhere within a
field or base class of a class shouldn't cause an import of that class
to fail. Instead, only that field or base class becomes inaccessible
from Carbon.

Also improve the way that type importing errors are diagnosed. While we
lose the precision of a diagnostic saying why a type is not supported,
we gain a useful source location for where the type was mentioned in C++
code.

---------

Co-authored-by: Jon Ross-Perkins <jperkins@google.com>
2025-08-13 22:07:54 +00:00
Richard Smith 629f77eb61 Switch to representing FloatLiteralType as a RealId. (#5944)
Don't convert to f64 until we know that's the type that we actually
want. Also reimplement the conversion from RealId to FloatId to perform
an exact conversion with a real check for overflow, rather than
performing an approximate conversion via the host `double` type.

Unfortunately, LLVM doesn't expose its integer mantissa and exponent to
APFloat conversion, so we convert the RealId back to a string for now.

The LLVM conversion also detects overflow only if the literal would
round to having an out-of-range exponent, not if the literal is outside
the range of values of the type as the Carbon design expects. It's not
clear to me which rule we actually want here, so for simplicitly I'm
using the LLVM rule for now.

In preparation for adding other floating-point types beyond f64.
2025-08-12 22:08:07 +00:00
Richard Smith 28103b8f2e Convert LegacyFloatType into FloatLiteralType. (#5939)
* Rename the type.
* Change lowering to lower FloatLiteralType values as the placeholder
  `{}` value we use for literals instead of as an LLVM f64.
* Change eval to convert the type as part of a floating point
  conversion, so that lowering can lower converted constants properly.

For now we still represent a value of FloatLiteralType as a
double-precision APFloat. (That will need to change so that we can
losslessly convert literals to f80 / f128 values, and so that we can
convert literals to f32 values without double-rounding.)
2025-08-12 18:55:38 +00:00
Dana JansensandJon Ross-Perkins 4b0e2b03b6 Add the .Self name for the type expression of a compile time binding (#5937)
We add a virtual node (`CompileTimeBindingPatternStart`) as the first
child of `CompileTimeBindingPattern` which holds the identifier
underneath it, so that it is checked just before the type expression of
the `CompileTimeBindingPattern`. When we reach this virtual node during
check, we add `.Self` as a name in the current scope, and when we reach
`CompileTimeBindingPattern` we remove it from scope, which ensures it's
present during only the checking of the type expression for the compile
time pattern.

At the moment the `.Self` has a different type (it's a `TypeType`) than
other `.Self` in the facet type (which are a single `FacetType`), but
the intention is to immediately substitute it out of the facet type
entirely, replacing it with a reference to the compile time binding (a
`BindSymbolicName`) itself. A TODO has been added for this.

---------

Co-authored-by: Jon Ross-Perkins <jperkins@google.com>
2025-08-12 16:27:07 +00:00