This is the mail archive of the
mailing list for the GCC project.
Re: [patch] lno branch merge -- new iv & simple loop analysis
On Tue, Feb 17, 2004 at 01:34:55AM +0100, Zdenek Dvorak wrote:
> I have added check for this to iv_number_of_iterations and abort
> to get_mode_bounds; there are several places where I rely on that
> the returned value is a const_int, so I would rather postpone handling
> it till it shows up to be a problem.
I would expect
long long i, n;
for (i = 0; i < n; ++i)
to abort on i386 then, since HOST_BITS_PER_WIDE_INT == 32.
You're going to have to have some way to quit early if the
mode doesn't fit your expectations.
Unless you have a Really Good Reason, I think all of your
uses of HOST_WIDEST_INT are wrong.