From: Benjamin Herrenschmidt <benh@kernel.crashing.org>
To: Linus Torvalds <torvalds@osdl.org>
Cc: "David S. Miller" <davem@redhat.com>,
wesolows@foobazco.org, willy@debian.org,
Andrea Arcangeli <andrea@suse.de>, Andrew Morton <akpm@osdl.org>,
Linux Kernel list <linux-kernel@vger.kernel.org>,
mingo@elte.hu, bcrl@kvack.org, linux-mm@kvack.org,
Linux Arch list <linux-arch@vger.kernel.org>
Subject: Re: [PATCH] ppc64: Fix possible race with set_pte on a present PTE
Date: Wed, 26 May 2004 08:05:36 +1000 [thread overview]
Message-ID: <1085522735.14969.130.camel@gaston> (raw)
In-Reply-To: <Pine.LNX.4.58.0405251452590.9951@ppc970.osdl.org>
On Wed, 2004-05-26 at 07:54, Linus Torvalds wrote:
> On Wed, 26 May 2004, Benjamin Herrenschmidt wrote:
> >
> > Well, just setting one of those 2 bits doesn't require a hash table
> > invalidate as long as nothing else changes.
>
> Ok. And nothing ever writes to the SW page tables outside the page table
> lock, right? So on ppc64, we could just do
>
> #define ptep_update_dirty_accessed(ptep, entry, dirty) \
> *(ptep) = (entry)
>
> and be done with it. No?
>
> I'm not going to do it without a big ack from you.
No. The hash fault path will update the PTE dirty/accessed on a hash miss
exception without holding the page table lock (acts a bit like a HW TLB
as far as linux is concerned). That's why it needs to be atomic.
Ben.
WARNING: multiple messages have this Message-ID (diff)
From: Benjamin Herrenschmidt <benh@kernel.crashing.org>
To: Linus Torvalds <torvalds@osdl.org>
Cc: "David S. Miller" <davem@redhat.com>,
wesolows@foobazco.org, willy@debian.org,
Andrea Arcangeli <andrea@suse.de>, Andrew Morton <akpm@osdl.org>,
Linux Kernel list <linux-kernel@vger.kernel.org>,
mingo@elte.hu, bcrl@kvack.org, linux-mm@kvack.org,
Linux Arch list <linux-arch@vger.kernel.org>
Subject: Re: [PATCH] ppc64: Fix possible race with set_pte on a present PTE
Date: Wed, 26 May 2004 08:05:36 +1000 [thread overview]
Message-ID: <1085522735.14969.130.camel@gaston> (raw)
In-Reply-To: <Pine.LNX.4.58.0405251452590.9951@ppc970.osdl.org>
On Wed, 2004-05-26 at 07:54, Linus Torvalds wrote:
> On Wed, 26 May 2004, Benjamin Herrenschmidt wrote:
> >
> > Well, just setting one of those 2 bits doesn't require a hash table
> > invalidate as long as nothing else changes.
>
> Ok. And nothing ever writes to the SW page tables outside the page table
> lock, right? So on ppc64, we could just do
>
> #define ptep_update_dirty_accessed(ptep, entry, dirty) \
> *(ptep) = (entry)
>
> and be done with it. No?
>
> I'm not going to do it without a big ack from you.
No. The hash fault path will update the PTE dirty/accessed on a hash miss
exception without holding the page table lock (acts a bit like a HW TLB
as far as linux is concerned). That's why it needs to be atomic.
Ben.
--
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:"aart@kvack.org"> aart@kvack.org </a>
next prev parent reply other threads:[~2004-05-25 22:09 UTC|newest]
Thread overview: 153+ messages / expand[flat|nested] mbox.gz Atom feed top
2004-05-24 3:29 [PATCH] ppc64: Fix possible race with set_pte on a present PTE Benjamin Herrenschmidt
2004-05-24 3:47 ` Linus Torvalds
2004-05-24 4:13 ` Benjamin Herrenschmidt
2004-05-24 4:36 ` Linus Torvalds
2004-05-24 4:44 ` Benjamin Herrenschmidt
2004-05-24 5:10 ` Linus Torvalds
2004-05-24 5:10 ` Linus Torvalds
2004-05-24 5:34 ` Benjamin Herrenschmidt
2004-05-24 5:34 ` Benjamin Herrenschmidt
2004-05-24 5:38 ` Benjamin Herrenschmidt
2004-05-24 5:38 ` Benjamin Herrenschmidt
2004-05-24 5:52 ` Benjamin Herrenschmidt
2004-05-24 5:52 ` Benjamin Herrenschmidt
2004-05-24 7:39 ` Ingo Molnar
2004-05-24 7:39 ` Ingo Molnar
2004-05-24 5:39 ` Benjamin Herrenschmidt
2004-05-24 5:39 ` Benjamin Herrenschmidt
2004-05-25 3:43 ` Andrea Arcangeli
2004-05-25 3:43 ` Andrea Arcangeli
2004-05-25 4:00 ` Linus Torvalds
2004-05-25 4:00 ` Linus Torvalds
2004-05-25 4:17 ` Benjamin Herrenschmidt
2004-05-25 4:17 ` Benjamin Herrenschmidt
2004-05-25 4:37 ` Andrea Arcangeli
2004-05-25 4:37 ` Andrea Arcangeli
2004-05-25 4:40 ` Benjamin Herrenschmidt
2004-05-25 4:40 ` Benjamin Herrenschmidt
2004-05-25 4:20 ` Andrea Arcangeli
2004-05-25 4:20 ` Andrea Arcangeli
2004-05-25 4:39 ` Linus Torvalds
2004-05-25 4:39 ` Linus Torvalds
2004-05-25 4:44 ` Linus Torvalds
2004-05-25 4:44 ` Linus Torvalds
2004-05-25 4:59 ` Andrea Arcangeli
2004-05-25 4:59 ` Andrea Arcangeli
2004-05-25 5:09 ` Andrea Arcangeli
2004-05-25 5:09 ` Andrea Arcangeli
2004-05-25 4:50 ` Andrea Arcangeli
2004-05-25 4:50 ` Andrea Arcangeli
2004-05-25 4:59 ` Linus Torvalds
2004-05-25 4:59 ` Linus Torvalds
2004-05-25 4:43 ` David Mosberger
2004-05-25 4:43 ` David Mosberger
2004-05-25 4:53 ` Andrea Arcangeli
2004-05-25 4:53 ` Andrea Arcangeli
2004-05-27 21:56 ` David Mosberger
2004-05-27 21:56 ` David Mosberger
2004-05-27 22:00 ` Benjamin Herrenschmidt
2004-05-27 22:00 ` Benjamin Herrenschmidt
2004-05-27 22:12 ` David Mosberger
2004-05-27 22:12 ` David Mosberger
2004-05-25 11:44 ` Matthew Wilcox
2004-05-25 11:44 ` Matthew Wilcox
2004-05-25 14:48 ` Linus Torvalds
2004-05-25 14:48 ` Linus Torvalds
2004-05-25 15:35 ` Keith M Wesolowski
2004-05-25 15:35 ` Keith M Wesolowski
2004-05-25 16:19 ` Linus Torvalds
2004-05-25 16:19 ` Linus Torvalds
2004-05-25 17:25 ` David S. Miller
2004-05-25 17:25 ` David S. Miller
2004-05-25 17:49 ` Linus Torvalds
2004-05-25 17:49 ` Linus Torvalds
2004-05-25 17:54 ` David S. Miller
2004-05-25 17:54 ` David S. Miller
2004-05-25 18:05 ` Linus Torvalds
2004-05-25 18:05 ` Linus Torvalds
2004-05-25 20:30 ` Linus Torvalds
2004-05-25 20:30 ` Linus Torvalds
2004-05-25 20:35 ` David S. Miller
2004-05-25 20:35 ` David S. Miller
2004-05-25 20:35 ` David S. Miller
2004-05-25 20:49 ` Linus Torvalds
2004-05-25 20:49 ` Linus Torvalds
2004-05-25 20:57 ` David S. Miller
2004-05-25 20:57 ` David S. Miller
2004-05-26 6:20 ` Keith M Wesolowski
2004-05-26 6:20 ` Keith M Wesolowski
2004-05-25 21:40 ` Benjamin Herrenschmidt
2004-05-25 21:40 ` Benjamin Herrenschmidt
2004-05-25 21:54 ` Linus Torvalds
2004-05-25 21:54 ` Linus Torvalds
2004-05-25 22:00 ` Linus Torvalds
2004-05-25 22:00 ` Linus Torvalds
2004-05-25 22:07 ` Benjamin Herrenschmidt
2004-05-25 22:07 ` Benjamin Herrenschmidt
2004-05-25 22:14 ` Linus Torvalds
2004-05-25 22:14 ` Linus Torvalds
2004-05-26 0:21 ` Benjamin Herrenschmidt
2004-05-26 0:21 ` Benjamin Herrenschmidt
2004-05-26 0:50 ` Linus Torvalds
2004-05-26 0:50 ` Linus Torvalds
2004-05-26 3:25 ` Benjamin Herrenschmidt
2004-05-26 3:25 ` Benjamin Herrenschmidt
2004-05-26 4:08 ` Linus Torvalds
2004-05-26 4:08 ` Linus Torvalds
2004-05-26 4:12 ` Benjamin Herrenschmidt
2004-05-26 4:12 ` Benjamin Herrenschmidt
2004-05-26 4:18 ` Benjamin Herrenschmidt
2004-05-26 4:18 ` Benjamin Herrenschmidt
2004-05-26 4:50 ` Linus Torvalds
2004-05-26 4:50 ` Linus Torvalds
2004-05-26 4:49 ` Benjamin Herrenschmidt
2004-05-26 4:49 ` Benjamin Herrenschmidt
2004-05-26 4:28 ` Linus Torvalds
2004-05-26 4:28 ` Linus Torvalds
2004-05-26 4:46 ` Benjamin Herrenschmidt
2004-05-26 4:46 ` Benjamin Herrenschmidt
2004-05-26 4:54 ` Linus Torvalds
2004-05-26 4:54 ` Linus Torvalds
2004-05-26 4:55 ` Benjamin Herrenschmidt
2004-05-26 4:55 ` Benjamin Herrenschmidt
2004-05-26 5:41 ` Benjamin Herrenschmidt
2004-05-26 5:41 ` Benjamin Herrenschmidt
2004-05-26 5:59 ` [PATCH] (signoff) " Benjamin Herrenschmidt
2004-05-26 5:59 ` Benjamin Herrenschmidt
2004-05-26 6:55 ` Benjamin Herrenschmidt
2004-05-26 6:55 ` Benjamin Herrenschmidt
2004-05-26 7:11 ` [PATCH] ppc32 implementation of ptep_set_access_flags Benjamin Herrenschmidt
2004-05-26 15:22 ` Linus Torvalds
2004-05-26 18:49 ` David S. Miller
2004-05-26 21:43 ` Benjamin Herrenschmidt
2004-05-28 1:29 ` David Mosberger
2004-05-25 22:05 ` Benjamin Herrenschmidt [this message]
2004-05-25 22:05 ` [PATCH] ppc64: Fix possible race with set_pte on a present PTE Benjamin Herrenschmidt
2004-05-25 22:09 ` Linus Torvalds
2004-05-25 22:09 ` Linus Torvalds
2004-05-25 22:19 ` Benjamin Herrenschmidt
2004-05-25 22:19 ` Benjamin Herrenschmidt
2004-05-25 22:24 ` Linus Torvalds
2004-05-25 22:24 ` Linus Torvalds
2004-05-25 21:27 ` Andrea Arcangeli
2004-05-25 21:27 ` Andrea Arcangeli
2004-05-25 21:43 ` Linus Torvalds
2004-05-25 21:43 ` Linus Torvalds
2004-05-25 21:55 ` Andrea Arcangeli
2004-05-25 21:55 ` Andrea Arcangeli
2004-05-25 22:01 ` Linus Torvalds
2004-05-25 22:01 ` Linus Torvalds
2004-05-25 22:18 ` Ivan Kokshaysky
2004-05-25 22:18 ` Ivan Kokshaysky
2004-05-25 22:42 ` Andrea Arcangeli
2004-05-25 22:42 ` Andrea Arcangeli
2004-05-26 2:26 ` Linus Torvalds
2004-05-26 2:26 ` Linus Torvalds
2004-05-26 7:06 ` Andrea Arcangeli
2004-05-26 7:06 ` Andrea Arcangeli
2004-05-25 21:44 ` Andrea Arcangeli
2004-05-25 21:44 ` Andrea Arcangeli
-- strict thread matches above, loose matches on Subject: below --
2004-06-01 12:04 Martin Schwidefsky
2004-06-01 12:04 ` Martin Schwidefsky
2004-06-01 12:10 Martin Schwidefsky
2004-06-01 12:10 ` Martin Schwidefsky
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1085522735.14969.130.camel@gaston \
--to=benh@kernel.crashing.org \
--cc=akpm@osdl.org \
--cc=andrea@suse.de \
--cc=bcrl@kvack.org \
--cc=davem@redhat.com \
--cc=linux-arch@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mingo@elte.hu \
--cc=torvalds@osdl.org \
--cc=wesolows@foobazco.org \
--cc=willy@debian.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.