lkml.org 
[lkml]   [2009]   [Jun]   [4]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: [PATCH] [6/16] HWPOISON: Add various poison checks in mm/memory.c
On Thu, Jun 04, 2009 at 01:19:15PM +0800, Andi Kleen wrote:
> On Thu, Jun 04, 2009 at 12:26:03PM +0800, Wu Fengguang wrote:
> > On Thu, Jun 04, 2009 at 02:46:38AM +0800, Andi Kleen wrote:
> > >
> > > Bail out early when hardware poisoned pages are found in page fault handling.
> >
> > I suspect this patch is also not absolutely necessary: the poisoned
> > page will normally have been isolated already.
>
> It's needed to prevent new pages comming in when there is a parallel
> fault while the memory failure handling is in process.
> Otherwise the pages could get remapped in that small window.

This patch makes no difference at least for file pages, including tmpfs.

In filemap_fault(), it will first do find_lock_page(), which will lock
the page and then double check if the page->mapping is NULL. If so,
it drops that page and re-find/re-create one in the radix tree.

That logic automatically avoids the poisoned page that is being
processed, because the poisoned page is now being locked, and when
find_lock_page() eventually is able to lock it, its page->mapping
will be NULL. So the PageHWPoison(vmf.page) test will never be true
for filemap_fault.

shmem_fault() calls shmem_getpage() to get the page, which will return
EIO on !uptodate page. It then returns VM_FAULT_SIGBUS on its own.

> > > --- linux.orig/mm/memory.c 2009-06-03 19:36:23.000000000 +0200
> > > +++ linux/mm/memory.c 2009-06-03 19:36:23.000000000 +0200
> > > @@ -2797,6 +2797,9 @@
> > > if (unlikely(ret & (VM_FAULT_ERROR | VM_FAULT_NOPAGE)))
> > > return ret;
> > >
> > > + if (unlikely(PageHWPoison(vmf.page)))
> > > + return VM_FAULT_HWPOISON;
> > > +
> >
> > Direct return with locked page could lockup someone later.
> > Either drop this patch or fix it with this check?
>
> Fair point. Fixed.

OK, thanks.

Fengguang


\
 
 \ /
  Last update: 2009-06-04 13:59    [W:0.103 / U:0.168 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site