Hi,
I'm looking at PR C++/25156 for which I have a fix (not the
incomplete one posted in the audit trail). Everyting goes well,
except it makes the testcase g++.dg/expr/return1.C fail for excess
error
return1.C:8: error: return-statement with a value, in function returning
'void'
// PR c++/18545
struct A;
A foo() // { dg-error "" }
{
A a; // { dg-error "" }
return a;
}
The resason here is that, after we complained that A is incomplete
(therefore cannot be used as return type in the function definition),
cp/decl.c:check_function_type() changes the return type to void, thus
giving misleading diagnostic later.
My question is: why don't we leave the type as is and refrain from
going through in-depth semantics processing -- like gimplification and
all that (we know the code is in error so that effort is wasted. --
that require that the compiler internal structure be in a sound state?
-- Gaby
