Re: [Tinycc-devel] "identifier expected"? But not by GCC.

[KHMan]

lostgallifreyan wrote:
> Dave Dodge <address@hidden> wrote:
> (17/04/2009 20:41)
>
> > On Sat, Apr 18, 2009 at 03:23:31AM +0800, KHMan wrote:
> > >     test.c:10: error: a label can only be part of a statement
> > >     and a declaration is not a statement
> > >
> > > IIRC it's not ISO C90 compliant. Nothing wrong with tcc. I'll leave the 
> > > digging up the ISO C standard to others.
> > According to the C grammar, a declaration is not a statement, but both
> > are block items.  [snip snip]
> [snip]
> KHMan, the GCC I tested it with reports itself as "GCC.EXE (GCC) 3.4.5 (mingw-vista 
> special r3)". It's supplied with wxDev-C++. If it hadn't accepted my code I'd have 
> just carried on looking for my error, but when compilers disagree it leaves me wondering 
> if I should let them duke it out between themselves. :) Seriously, as it appears it's a 
> basic C error in hindsight, it's a very misleading event when it's not flagged by GCC at 
> all, and the complied program runs as expected too. I initially tried GCC purely to see 
> if its error reporting gave me further clues, so I hadn't expected that.

I tried the following on:
gcc version 3.4.5 (mingw-vista special r3)

$ cat test.c
#include <stdlib.h>
int main()
{
    switch (0) {
        case 1: int a = 10; break;
    }
    return 0;
}

$ gcc -c test.c
test.c: In function `main':
test.c:5: error: syntax error before "int"

Very strange. I've always added braces for such cases where the 
scope is appropriate and I don't believe gcc 3.4.x and above ever 
allow such things. (Which was why I explicitly tested it.) Are you 
compiling in C++ mode? That wouldn't be good practice if there is 
a commitment to code stuff in a platform-neutral and 
compiler-neutral way. Other compilers that assumes proper C code 
may break on such code snippets.

Anyway, this is getting to be OT, so I'll stop here.

--
Cheers,
Kein-Hong Man (esq.)
Kuala Lumpur, Malaysia