[9fans] why does this compile?

[erik quanstrom]

this is rejected by most other compilers because
rlogn is not initialized to a constant. the value
returned is bizarre.

- erik

#include <u.h>
#include <libc.h>
static char rlog[24];
static char *rloge = rlog + sizeof rlog;
static char *rlogl = rlog + sizeof rlog - 1;
static char *rlogn = rloge - 1;
void
main(int, char **)
{
print("%p %p %p %p\n", rlog, rlogl, rlogn, rloge);
exits(nil);
}

[Charles Forsyth]

Post by erik quanstrom
the value
returned is bizarre.

it's the (lvalue of rloge)-1, not bizarre.

[erik quanstrom]

Post by Charles Forsyth

Post by erik quanstrom
the value
returned is bizarre.

it's the (lvalue of rloge)-1, not bizarre.

is that a legal thing to do?

- erik

[erik quanstrom]

No, but the value at least isn't bizarre. It would be correct for rloge as
an array, and i assume it gets through because of the way the
array->pointer change is done. Even so, if it had been GCC, by now it would
be an essential extension, and have been used to implement an elisp
interpreter, shave micro cycles from the time of a Java vm , and by GCC
itself. I might look at nipping all that in the bud tomorrow, although I
think I have looked at this once before.

thanks.

- erik

[Charles Forsyth]

No, but the value at least isn't bizarre. It would be correct for rloge as
an array, and i assume it gets through because of the way the
array->pointer change is done. Even so, if it had been GCC, by now it would
be an essential extension, and have been used to implement an elisp
interpreter, shave micro cycles from the time of a Java vm , and by GCC
itself. I might look at nipping all that in the bud tomorrow, although I
think I have looked at this once before.

Post by erik quanstrom

Post by Charles Forsyth

Post by erik quanstrom
the value
returned is bizarre.

it's the (lvalue of rloge)-1, not bizarre.

is that a legal thing to do?
- erik