https://gcc.gnu.org/bugzilla/show_bug.cgi?id=118136
Bug ID: 118136
Summary: Linking start code built with -flto with application
where main signature does not match causes warning
Product: gcc
Version: 15.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: lto
Assignee: unassigned at gcc dot gnu.org
Reporter: azoff at gcc dot gnu.org
Target Milestone: ---
Linking an application where there are function signatures that does not match
should generally generate a warning like:
foo.c:8:12: warning: type of 'baz' does not match original declaration
[-Wlto-type-mismatch]
bar.c:3:5: note: type mismatch in parameter 1
bar.c:3:5: note: type 'void' should match type 'int'
bar.c:3:5: note: 'main' was previously declared here
But, I think the main symbol is a little bit different.
In the C23 spec, the following can be found:
> 5.1.2.3.2 Program startup
> The function called at program startup is named main. The implementation
> declares no prototype
> for this function. It shall be defined with a return type of int and with no
> > parameters:
> int main(void) { /* ... */ }
> or with two parameters (referred to here as argc and argv, though any names
> may >be used, as they
> are local to the function in which they are declared):
> int main(int argc, char *argv[]) { /* ... */ }
> or equivalent;6) or in some other implementation-defined manner.
> (footnote 6 permits some type equivalence for the arguments in the second
> case, eg via typedef).
This suggest that the call site can use main(int,char**) and the target may be
one of:
int main() {}
int main(int, char **) {}
int main(void) {}
Based on this, it looks like the warning issued when using -flto for the "main"
symbol is bogus.
Attached is a small reproducer that I think is generic enough to work on any
target.
The use case where the startup code is built with -flto comes from embedded
where it's common to try to reduce the code size as much as possible, for
example using a custom _start routine.
