Re: [PATCH v11 06/21] vfs: Add copy_to_iter(), copy_from_iter() and iov_iter_zero()

[Matthew Wilcox <willy@linux.intel.com>]

On Thu, Oct 16, 2014 at 03:33:55PM +0200, Mathieu Desnoyers wrote:
> > +static size_t copy_to_iter_iovec(void *from, size_t bytes, struct iov_iter *i)
> > +{
[...]
> > +	left = __copy_to_user(buf, from, copy);
> 
> How comes this function uses __copy_to_user without any access_ok()
> check ? This has security implications.

The access_ok() check is done higher up the call-chain if it's appropriate.
These functions can be (intentionally) called to access kernel addresses,
so it wouldn't be appropriate to do that here.

> > +static size_t copy_page_to_iter_bvec(struct page *page, size_t offset,
> > +					size_t bytes, struct iov_iter *i)
> > +{
> > +	void *kaddr = kmap_atomic(page);
> > +	size_t wanted = copy_to_iter_bvec(kaddr + offset, bytes, i);
> 
> missing newline.
> 
> > +	kunmap_atomic(kaddr);
> > +	return wanted;
> > +}

Are you seriously suggesting that:

static size_t copy_page_to_iter_bvec(struct page *page, size_t offset,
                                        size_t bytes, struct iov_iter *i)
{
        void *kaddr = kmap_atomic(page);
        size_t wanted = copy_to_iter_bvec(kaddr + offset, bytes, i);

        kunmap_atomic(kaddr);
        return wanted;
}

is more readable than without the newline?  I can see the point of the
rule for functions with a lot of variables, or a lot of lines, but I
don't see the point of it for such a small function.

In any case, this patch is now upstream, so I shan't be proposing any
stylistic changes for it.

--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org.  For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>