https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83448
--- Comment #7 from Martin Sebor <msebor at gcc dot gnu.org> ---
(In reply to David Malcolm from comment #6)
> Martin: what is the above code attempting to do? It's not clear to me
> (though this may be -ENOCOFFEE on my part, sorry).
The purpose of the set_caret_index call is to do just what the comment says:
point the caret at the first character in the format string that doesn't fit in
the destination. But the only test that exercises the caret location,
builtin-sprintf-warn-4.c, doesn't hit that call and the only tests that I have
been able to construct where this particular call makes a difference show that
it only causes trouble. E.g., this:
char d[4];
void f (int i)
{
if (i < 0 || 3 < i)
i = 0;
__builtin_sprintf (d, "%*s1234", i, "");
}
results in this with the set_caret call:
a.c: In function ‘f’:
a.c:8:3: warning: ‘1234’ directive writing 4 bytes into a region of size
between 1 and 4 [-Wformat-overflow=]
__builtin_sprintf (d, "%*s1234", i, "");
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
while in this without it:
a.c: In function ‘f’:
a.c:8:29: warning: ‘1234’ directive writing 4 bytes into a region of size
between 1 and 4 [-Wformat-overflow=]
__builtin_sprintf (d, "%*s1234", i, "");
^~~~
a.c:8:3: note: ‘__builtin_sprintf’ output between 5 and 8 bytes into a
destination of size 4
__builtin_sprintf (d, "%*s1234", i, "");
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
So it looks like it can be removed.
