From: Mel Gorman <mgorman@suse.de>
To: Hugh Dickins <hughd@google.com>
Cc: Andrew Morton <akpm@linux-foundation.org>,
Petr Holasek <pholasek@redhat.com>,
Andrea Arcangeli <aarcange@redhat.com>,
Izik Eidus <izik.eidus@ravellosystems.com>,
linux-kernel@vger.kernel.org, linux-mm@kvack.org
Subject: Re: [PATCH 6/11] ksm: remove old stable nodes more thoroughly
Date: Thu, 14 Feb 2013 11:58:05 +0000 [thread overview]
Message-ID: <20130214115805.GC7367@suse.de> (raw)
In-Reply-To: <alpine.LNX.2.00.1302081057110.4233@eggly.anvils>
On Fri, Feb 08, 2013 at 11:33:40AM -0800, Hugh Dickins wrote:
> > > <SNIP>
> > >
> > > 2. __ksm_enter() has a nice little optimization, to insert the new mm
> > > just behind ksmd's cursor, so there's a full pass for it to stabilize
> > > (or be removed) before ksmd addresses it. Nice when ksmd is running,
> > > but not so nice when we're trying to unmerge all mms: we were missing
> > > those mms forked and inserted behind the unmerge cursor. Easily fixed
> > > by inserting at the end when KSM_RUN_UNMERGE.
> > >
> > > 3. It is possible for a KSM page to be faulted back from swapcache into
> > > an mm, just after unmerge_and_remove_all_rmap_items() scanned past it.
> > > Fix this by copying on fault when KSM_RUN_UNMERGE: but that is private
> > > to ksm.c, so dissolve the distinction between ksm_might_need_to_copy()
> > > and ksm_does_need_to_copy(), doing it all in the one call into ksm.c.
>
> What I found is that a 4th cause emerges once KSM migration
> is properly working: that interval during page migration when the old
> page has been fully unmapped but the new not yet mapped in its place.
>
For anyone else watching -- normal page migration expects to be protected
during that particular window with migration ptes. Any references to the
PTE mapping a page being migrated faults on a swap-like PTE and waits
in migration_entry_wait().
> The KSM COW breaking cannot see a page there then, so it ends up with
> a (newly migrated) KSM page left behind. Almost certainly has to be
> fixed in follow_page(), but I've not yet settled on its final form -
> the fix I have works well, but a different approach might be better.
>
follow_page() is one option. My guess is that you're thinking of adding
a FOLL_ flag that will cause follow_page() to check is_migration_entry()
and migration_entry_wait() if the flag is present.
Otherwise you would need to check for migration ptes in a number of places
under page lock and then hold the lock for long periods of time to prevent
migration starting. I did not check this option in depth because it quickly
looked like it would be a mess, with long page lock hold times and might
not even be workable.
> > > +static int remove_all_stable_nodes(void)
> > > +{
> > > + struct stable_node *stable_node;
> > > + int nid;
> > > + int err = 0;
> > > +
> > > + for (nid = 0; nid < nr_node_ids; nid++) {
> > > + while (root_stable_tree[nid].rb_node) {
> > > + stable_node = rb_entry(root_stable_tree[nid].rb_node,
> > > + struct stable_node, node);
> > > + if (remove_stable_node(stable_node)) {
> > > + err = -EBUSY;
> > > + break; /* proceed to next nid */
> > > + }
> >
> > If remove_stable_node() returns an error then it's quite possible that it'll
> > go boom when that page is encountered later but it's not guaranteed. It'd
> > be best effort to continue removing as many of the stable nodes anyway.
> > We're in trouble either way of course.
>
> If it returns an error, then indeed something we don't yet understand
> has occurred, and we shall want to debug it. But unless it's due to
> corruption somewhere, we shouldn't be in much trouble, shouldn't go boom:
> remove_all_stable_nodes() error is ignored at the end of unmerging, it
> will be tried again when changing merge_across_nodes, and an error
> then will just prevent changing merge_across_nodes at that time. So
> the mysteriously unremovable stable nodes remain the same kind of tree.
>
Ok.
--
Mel Gorman
SUSE Labs
--
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>
next prev parent reply other threads:[~2013-02-14 11:58 UTC|newest]
Thread overview: 69+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-01-26 1:53 [PATCH 0/11] ksm: NUMA trees and page migration Hugh Dickins
2013-01-26 1:54 ` [PATCH 1/11] ksm: allow trees per NUMA node Hugh Dickins
2013-01-27 1:14 ` Simon Jeons
2013-01-27 2:54 ` Hugh Dickins
2013-01-27 3:16 ` Simon Jeons
2013-01-27 21:55 ` Hugh Dickins
2013-01-28 23:03 ` Andrew Morton
2013-01-29 1:17 ` Hugh Dickins
2013-01-28 23:08 ` Andrew Morton
2013-01-29 1:38 ` Hugh Dickins
2013-02-05 16:41 ` Mel Gorman
2013-02-07 23:57 ` Hugh Dickins
2013-01-26 1:56 ` [PATCH 2/11] ksm: add sysfs ABI Documentation Hugh Dickins
2013-01-26 1:58 ` [PATCH 3/11] ksm: trivial tidyups Hugh Dickins
2013-01-28 23:11 ` Andrew Morton
2013-01-29 1:44 ` Hugh Dickins
2013-01-26 1:59 ` [PATCH 4/11] ksm: reorganize ksm_check_stable_tree Hugh Dickins
2013-02-05 16:48 ` Mel Gorman
2013-02-08 0:07 ` Hugh Dickins
2013-02-14 11:30 ` Mel Gorman
2013-01-26 2:00 ` [PATCH 5/11] ksm: get_ksm_page locked Hugh Dickins
2013-01-27 2:36 ` Simon Jeons
2013-01-27 22:08 ` Hugh Dickins
2013-01-28 0:36 ` Simon Jeons
2013-01-28 3:35 ` Hugh Dickins
2013-01-27 2:48 ` Simon Jeons
2013-01-27 22:10 ` Hugh Dickins
2013-02-05 17:18 ` Mel Gorman
2013-02-08 0:33 ` Hugh Dickins
2013-02-14 11:34 ` Mel Gorman
2013-01-26 2:01 ` [PATCH 6/11] ksm: remove old stable nodes more thoroughly Hugh Dickins
2013-01-27 4:55 ` Simon Jeons
2013-01-27 23:05 ` Hugh Dickins
2013-01-28 1:42 ` Simon Jeons
2013-01-28 4:14 ` Hugh Dickins
2013-01-28 2:12 ` Simon Jeons
2013-01-28 4:19 ` Hugh Dickins
2013-01-28 6:36 ` Simon Jeons
2013-01-28 23:44 ` Andrew Morton
2013-01-29 2:03 ` Hugh Dickins
2013-02-05 17:55 ` Mel Gorman
2013-02-08 19:33 ` Hugh Dickins
2013-02-14 11:58 ` Mel Gorman [this message]
2013-02-14 22:19 ` Hugh Dickins
2013-01-26 2:03 ` [PATCH 7/11] ksm: make KSM page migration possible Hugh Dickins
2013-01-27 5:47 ` Simon Jeons
2013-01-27 23:12 ` Hugh Dickins
2013-01-28 0:41 ` Simon Jeons
2013-01-28 3:44 ` Hugh Dickins
2013-02-05 19:11 ` Mel Gorman
2013-02-08 20:52 ` Hugh Dickins
2013-01-26 2:05 ` [PATCH 8/11] ksm: make !merge_across_nodes migration safe Hugh Dickins
2013-01-27 8:49 ` Simon Jeons
2013-01-27 23:25 ` Hugh Dickins
2013-01-28 3:44 ` Simon Jeons
2013-01-26 2:06 ` [PATCH 9/11] ksm: enable KSM page migration Hugh Dickins
2013-01-26 2:07 ` [PATCH 10/11] mm: remove offlining arg to migrate_pages Hugh Dickins
2013-01-26 2:10 ` [PATCH 11/11] ksm: stop hotremove lockdep warning Hugh Dickins
2013-01-27 6:23 ` Simon Jeons
2013-01-27 23:35 ` Hugh Dickins
2013-02-08 18:45 ` Gerald Schaefer
2013-02-11 22:13 ` Hugh Dickins
2013-01-28 23:54 ` [PATCH 0/11] ksm: NUMA trees and page migration Andrew Morton
2013-01-29 0:49 ` Izik Eidus
2013-01-29 2:26 ` Izik Eidus
2013-01-29 16:51 ` Andrea Arcangeli
2013-01-31 0:05 ` Ric Mason
2013-01-29 1:07 ` Hugh Dickins
2013-01-29 10:45 ` Gleb Natapov
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=20130214115805.GC7367@suse.de \
--to=mgorman@suse.de \
--cc=aarcange@redhat.com \
--cc=akpm@linux-foundation.org \
--cc=hughd@google.com \
--cc=izik.eidus@ravellosystems.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=pholasek@redhat.com \
/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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).