https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66110
Richard Biener <rguenth at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |jsm28 at gcc dot gnu.org
--- Comment #9 from Richard Biener <rguenth at gcc dot gnu.org> ---
(In reply to Kevin OConnor from comment #8)
> Thanks! I can confirm the latest trunk performs the desired optimization.
>
> However, this test case still isn't fully optimized:
>
> void f2(struct s1 *ps1, uint8_t *pi8)
> {
> ps1->f1 = 3;
> *pi8 = 8;
> ps1->f1 += 2;
> }
>
> That is, an "uint8_t*" still aliases with every other type. The "struct"
> optimization is more important for my usage, but it is unfortunate that
> uint8_t*/int8_t* are pessimized. (In particular, there does not appear to
> be any way to declare a pointer to an 8 bit integer that doesn't alias every
> other type.)
>
> I can open a separate bugzilla entry on the above.
Well, so it ultimiatively boils down to GCC defining
#define __UINT8_TYPE__ unsigned char
and that being a character type. I'm not sure whether the C or C++ standards
mandate that uint8_t be a 'character type' as far as type-based aliasing goes
but I would expect that user-code may rely on that property. So yes, there
would be no 1-byte type that doesn't get alias-set zero. Well - you could
use
struct byte { uint8_t b; };
and make sure to use aggregate assignments everywhere, only ever extracting
the actual value from temporary aggregates...
