Date: Sat, 27 Jul 2013 01:18:44 +0400
From: Cyrill Gorcunov <gorcunov@gmail.com>
To: Andy Lutomirski <luto@amacapital.net>
Cc: Linux MM <linux-mm@kvack.org>, LKML <linux-kernel@vger.kernel.org>,
        Pavel Emelyanov <xemul@parallels.com>,
        Andrew Morton <akpm@linux-foundation.org>,
        Matt Mackall <mpm@selenic.com>,
        Xiao Guangrong <xiaoguangrong@linux.vnet.ibm.com>,
        Marcelo Tosatti <mtosatti@redhat.com>,
        KOSAKI Motohiro <kosaki.motohiro@gmail.com>,
        Stephen Rothwell <sfr@canb.auug.org.au>
Subject: Re: [PATCH] mm: Save soft-dirty bits on file pages
Message-ID: <20130726211844.GB8508@moon>
References: <20130726201807.GJ8661@moon>
 <CALCETrUJa-Y40vnb6YOPry0dCXb3zCQ0y19i2yHWdzKR75HUzg@mail.gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <CALCETrUJa-Y40vnb6YOPry0dCXb3zCQ0y19i2yHWdzKR75HUzg@mail.gmail.com>
User-Agent: Mutt/1.5.21 (2010-09-15)
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1500
Lines: 31

On Fri, Jul 26, 2013 at 01:55:04PM -0700, Andy Lutomirski wrote:
> On Fri, Jul 26, 2013 at 1:18 PM, Cyrill Gorcunov <gorcunov@gmail.com> wrote:
> > Andy reported that if file page get reclaimed we loose soft-dirty bit
> > if it was there, so save _PAGE_BIT_SOFT_DIRTY bit when page address
> > get encoded into pte entry. Thus when #pf happens on such non-present
> > pte we can restore it back.
> >
> 
> Unless I'm misunderstanding this, it's saving the bit in the
> non-present PTE.  This sounds wrong -- what happens if the entire pmd

It's the same as encoding pgoff in pte entry (pte is not present),
but together with pgoff we save soft-bit status, later on #pf we decode
pgoff and restore softbit back if it was there, pte itself can't disappear
since it holds pgoff information.

> (or whatever the next level is called) gets zapped?  (Also, what
> happens if you unmap a file and map a different file there?)

If file pages are remapped to a new place we remember softdity
bit status previously has and propagate it to a new pte (as in
install_file_pte, old ptes are cleared).

If file unmapped then new one mapped, pmd/ptes are cleared
(including softbit) and it remains clear until new write
happens, if only i've not missed something obvious.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/