1This is a collection of things that test suites have 2said were "wrong" with GCC--but that I don't agree with. 3 4First, test suites sometimes test for compatibility with 5traditional C. GCC with -traditional is not completely 6compatible with traditional C, and in some ways I think it 7should not be. 8 9* K&R C allowed \x to appear in a string literal (or character 10literal?) even in cases where it is *not* followed by a sequence of 11hex digits. I'm not convinced this is desirable. 12 13* K&R compilers allow comments to cross over an inclusion boundary (i.e. 14started in an include file and ended in the including file). 15I think this would be quite ugly and can't imagine it could 16be needed. 17 18Sometimes tests disagree with GCC's interpretation of the ANSI standard. 19 20* One test claims that this function should return 1. 21 22 enum {A, B} foo; 23 24 func (enum {B, A} arg) 25 { 26 return B; 27 } 28 29I think it should return 0, because the definition of B that 30applies is the one in func. 31 32* Some tests report failure when the compiler does not produce 33an error message for a certain program. 34 35ANSI C requires a "diagnostic" message for certain kinds of invalid 36programs, but a warning counts as a diagnostic. If GCC produces 37a warning but not an error, that is correct ANSI support. 38When test suites call this "failure", the tests are broken. 39 40