Re: kiobugs in 2.3.48

From: Stephen C. Tweedie (sct@redhat.com)
Date: Thu Mar 02 2000 - 09:16:55 EST


Hi,

On Thu, 2 Mar 2000 11:53:19 +0530 (IST), V Ganesh <ganesh@veritas.com>
said:

> we do
> map = follow_page(ptr);
> to convert the userspace vaddr to a struct page *. right ?

Yes.

> where is it converted to a kernel virtual address ?

There is a one-to-one equivalence between kernel addresses and struct
page *s.

> then we do
> map = get_page_map(map, ptr);
> and now we pass this to get_page_map() which promptly returns 0,

Yikes. Yes, somebody snuck a completely bogus patch into 2.3.48. The
_only_ reason vaddr should be used in get_page_map() is to act as an
argument to ZERO_PAGE() for architectures in which the zero page is
cache-coloured.

This bug is not present in the patch I sent out last week. I'll submit
the fixes to Linus shortly.

--Stephen

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/



This archive was generated by hypermail 2b29 : Tue Mar 07 2000 - 21:00:12 EST