Calculating array sizes in C - was: Re: Build regressions/improvements in v6.2-rc1

[David Laight]

From: Segher Boessenkool
> Sent: 20 January 2023 10:54
...
> > > I suggest to file a bug against gcc complaining about a "spurious
> > > warning", and using "-Werror -Wno-error-sizeof-pointer-div" until gcc is
> > > adapted to not emit the warning about the pointer division if the result
> > > is not used.
> 
> Yeah.  If the first operand of a conditional operator is non-zero, the
> second operand is not evaluated, and if the first is zero, the third
> operand is not evaluated.  It is better if we do not warn about
> something we do not evaluate.  In cases like here where it is clear at
> compile time which branch is taken, that shouldn't be too hard.
> 
> Can someone please file a GCC PR?  With reduced testcase preferably.

It's not a bug.
All the operands of the conditional operator have to be valid.
It might be that the optimiser can discard one, but that happens
much later on.
Even the operands of choose_expr() have to be valid - but can
have different types.

I'm not sure what the code is trying to do or why it is failing.
Was it a fail in userspace - where the option to allow sizeof (void)
isn't allowed.

FWIW you can check for a compile-time NULL (or 0) with:
#define is_null(x) (sizeof *(0 : (void *)(x) ? (int *)0) != 1)

Although that is a compile-time error for non-NULL unless
'void *' arithmetic is allowed.

	David

-
Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK
Registration No: 1397386 (Wales)