On Tue, Sep 14, 2021 at 12:46 PM Linus Torvalds
[off-list ref] wrote:

On Tue, Sep 14, 2021 at 12:10 PM Nick Desaulniers
[off-list ref] wrote:

quoted

Isn't the parameter `blksize` of `nbd_set_size` declared as `loff_t`?

So?

I'm not seeing your point.

We've checked the range of it - in loff_t.

So the *value* is already checked, and most definitely fits in 'unsigned long'.

So __ffs() is perfectly fine. It will truncate that loff_t to a sane type.

What is the problem you're trying to solve?

Just making sure __ffs() works as expected should blksize > LONG_MAX
on 32b targets.  I don't see the range check you're referring to.
loff_t is a long long, yeah?
-- 
Thanks,
~Nick Desaulniers