From: Peter Zijlstra <peterz@infradead.org>
To: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>,
"Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>,
Johannes Weiner <hannes@cmpxchg.org>, Jan Kara <jack@suse.cz>,
Mel Gorman <mgorman@techsingularity.net>,
Rik van Riel <riel@redhat.com>, linux-mm <linux-mm@kvack.org>
Subject: Re: page_waitqueue() considered harmful
Date: Tue, 27 Sep 2016 09:30:55 +0200 [thread overview]
Message-ID: <20160927073055.GM2794@worktop> (raw)
In-Reply-To: <CA+55aFwVSXZPONk2OEyxcP-aAQU7-aJsF3OFXVi8Z5vA11v_-Q@mail.gmail.com>
On Mon, Sep 26, 2016 at 01:58:00PM -0700, Linus Torvalds wrote:
> Why is the page_waitqueue() handling so expensive? Let me count the ways:
> (b) It's cache miss heaven. It takes a cache miss on three different
> things:looking up the zone 'wait_table', then looking up the hash
> queue there, and finally (inside __wake_up_bit) looking up the wait
> queue itself (which will effectively always be NULL).
> Is there really any reason for that incredible indirection? Do we
> really want to make the page_waitqueue() be a per-zone thing at all?
> Especially since all those wait-queues won't even be *used* unless
> there is actual IO going on and people are really getting into
> contention on the page lock.. Why isn't the page_waitqueue() just one
> statically sized array?
I suspect the reason is to have per node hash tables, just like we get
per node page-frame arrays with sparsemem.
> Also, if those bitlock ops had a different bit that showed contention,
> we could actually skip *all* of this, and just see that "oh, nobody is
> waiting on this page anyway, so there's no point in looking up those
> wait queues". We don't have that many "__wait_on_bit()" users, maybe
> we could say that the bitlocks do have to haev *two* bits: one for the
> lock bit itself, and one for "there is contention".
That would be fairly simple to implement, the difficulty would be
actually getting a page-flag to use for this. We're running pretty low
in available bits :/
--
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:[~2016-09-27 7:31 UTC|newest]
Thread overview: 42+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-09-26 20:58 page_waitqueue() considered harmful Linus Torvalds
2016-09-26 21:23 ` Rik van Riel
2016-09-26 21:30 ` Linus Torvalds
2016-09-26 23:11 ` Kirill A. Shutemov
2016-09-27 1:01 ` Rik van Riel
2016-09-27 7:30 ` Peter Zijlstra [this message]
2016-09-27 8:54 ` Mel Gorman
2016-09-27 9:11 ` Kirill A. Shutemov
2016-09-27 9:42 ` Mel Gorman
2016-09-27 9:52 ` Minchan Kim
2016-09-27 12:11 ` Kirill A. Shutemov
2016-09-29 8:01 ` Peter Zijlstra
2016-09-29 12:55 ` Nicholas Piggin
2016-09-29 13:16 ` Peter Zijlstra
2016-09-29 13:54 ` Nicholas Piggin
2016-09-29 15:05 ` Rik van Riel
2016-09-27 8:03 ` Jan Kara
2016-09-27 8:31 ` Mel Gorman
2016-09-27 14:34 ` Peter Zijlstra
2016-09-27 15:08 ` Nicholas Piggin
2016-09-27 16:31 ` Linus Torvalds
2016-09-27 16:49 ` Peter Zijlstra
2016-09-28 10:45 ` Mel Gorman
2016-09-28 11:11 ` Peter Zijlstra
2016-09-28 16:10 ` Linus Torvalds
2016-09-29 13:08 ` Peter Zijlstra
2016-10-03 10:47 ` Mel Gorman
2016-09-27 14:53 ` Nicholas Piggin
2016-09-27 15:17 ` Nicholas Piggin
2016-09-27 16:52 ` Peter Zijlstra
2016-09-27 17:06 ` Nicholas Piggin
2016-09-28 7:05 ` Peter Zijlstra
2016-09-28 11:05 ` Paul E. McKenney
2016-09-28 11:16 ` Peter Zijlstra
2016-09-28 12:58 ` Paul E. McKenney
2016-09-29 1:31 ` Nicholas Piggin
2016-09-29 2:12 ` Paul E. McKenney
2016-09-29 6:21 ` Peter Zijlstra
2016-09-29 6:42 ` Nicholas Piggin
2016-09-29 7:14 ` Peter Zijlstra
2016-09-29 7:43 ` Peter Zijlstra
2016-09-28 7:40 ` Mel Gorman
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=20160927073055.GM2794@worktop \
--to=peterz@infradead.org \
--cc=akpm@linux-foundation.org \
--cc=hannes@cmpxchg.org \
--cc=jack@suse.cz \
--cc=kirill.shutemov@linux.intel.com \
--cc=linux-mm@kvack.org \
--cc=mgorman@techsingularity.net \
--cc=riel@redhat.com \
--cc=torvalds@linux-foundation.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 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).