JustinStitt wrote:
Hi, gentle ping.
@efriedma-quic, do you still have concerns with the current implementation (as
an attribute attached to typedefs)? After limiting `wraps` to C and adding
warnings to let users know when the attribute may be stripped, I think it's in
a good spot.
```c
typedef int __attribute__((wraps)) wrap_int;
extern void foo(int b);
int main(void) {
wrap_int a;
foo(a); // wraps.c:12:7: warning: 'wraps' attribute may be implicitly
discarded \
// when converted to 'int' [-Wimplicitly-discarded-wraps-attribute]
}
```
This feature is obviously niche and requires opting in with actual source
modifications, this is not something people will use accidentally. However,
Eli, I hear your concerns about some of the non-obvious ways attributes can get
lost. I've tried to remedy this with some warnings (as seen above) but, of
course, the semantics of attribute persistence are what they are. Eli, if you
think this is not good enough then I will try and redesign the feature as an
extended type. Although, I reckon the C type system has just as many quirks and
foot-guns as the Clang attribute system.
https://github.com/llvm/llvm-project/pull/86618
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
