From: Nick Piggin <nickpiggin@yahoo.com.au>
To: Andrew Morton <akpm@linux-foundation.org>
Cc: clameter@sgi.com, linux-mm@kvack.org, linux-kernel@vger.kernel.org
Subject: Re: [QUICKLIST 0/4] Arch independent quicklists V2
Date: Tue, 13 Mar 2007 22:30:19 +1100 [thread overview]
Message-ID: <45F68B4B.9020200@yahoo.com.au> (raw)
In-Reply-To: <20070313041551.565891b5.akpm@linux-foundation.org>
Andrew Morton wrote:
>>On Tue, 13 Mar 2007 22:06:46 +1100 Nick Piggin <nickpiggin@yahoo.com.au> wrote:
>>Andrew Morton wrote:
>>
>>>>On Tue, 13 Mar 2007 19:03:38 +1100 Nick Piggin <nickpiggin@yahoo.com.au> wrote:
>>
>>...
>>
>>
>>>>Page allocator still requires interrupts to be disabled, which this doesn't.
>
>
>>>>it is worthwhile.
>>>
>>>
>>>If you want a zeroed page for pagecache and someone has just stuffed a
>>>known-zero, cache-hot page into the pagetable quicklists, you have good
>>>reason to be upset.
>>
>>The thing is, pagetable pages are the one really good exception to the
>>rule that we should keep cache hot and initialise-on-demand. They
>>typically are fairly sparsely populated and sparsely accessed. Even
>>for last level page tables, I think it is reasonable to assume they will
>>usually be pretty cold.
>
>
> eh? I'd have thought that a pte page which has just gone through
> zap_pte_range() will very often have a _lot_ of hot cachelines, and
> that's a common case.
>
> Still. It's pretty easy to test.
Well I guess that would be the case if you had just unmapped a 4MB
chunk that was pretty dense with pages.
My malloc seems to allocate and free in blocks of 128K, so that's
only going to give us 3% of the last level pte being cache hot when
it gets freed. Not sure what common mmap(file) access patterns
look like.
The majority of programs I run have a smattering of llpt pages
pretty sparsely populated, covering text, libraries, heap, stack,
vdso.
We don't actually have to zap_pte_range the entire page table in
order to free it (IIRC we used to have to, before the 4lpt patches).
But yeah let's see some tests. I would definitely want to avoid this
extra layer of complexity if it is just as good to return the pages
to the pcp lists.
>>>Maybe, dunno. It was apparently a win on powerpc many years ago. I had a
>>>fiddle with it 5-6 years ago on x86 using a cache-disabled mapping of the
>>>page. But it needed too much support in core VM to bother. Since then
>>>we've grown per-cpu page magazines and __GFP_ZERO. Plus I'm not aware of
>>>anyone having tried doing it on x86 with non-temporal stores.
>>
>>You can win on specifically constructed benchmarks, easily.
>>
>>But considering all the other problems you're going to introduce, we'd need
>>a significant win on a significant something, IMO.
>>
>>You waste memory bandwidth. You also use more CPU and memory cycles
>>speculatively, ergo you waste more power.
>
>
> Yeah, prezeroing in idle is probably pointless. But I'm not aware of
> anyone having tried it properly...
--
SUSE Labs, Novell Inc.
Send instant messages to your online friends http://au.messenger.yahoo.com
next prev parent reply other threads:[~2007-03-13 11:30 UTC|newest]
Thread overview: 35+ messages / expand[flat|nested] mbox.gz Atom feed top
2007-03-13 7:13 [QUICKLIST 0/4] Arch independent quicklists V2 Christoph Lameter
2007-03-13 7:13 ` [QUICKLIST 1/4] Generic quicklist implementation Christoph Lameter
2007-03-13 9:05 ` Paul Mundt
2007-03-15 20:51 ` Christoph Lameter
2007-03-13 7:13 ` [QUICKLIST 2/4] Quicklist support for i386 Christoph Lameter
2007-03-13 7:13 ` [QUICKLIST 3/4] Quicklist support for x86_64 Christoph Lameter
2007-03-13 7:13 ` [QUICKLIST 4/4] Quicklist support for sparc64 Christoph Lameter
2007-03-13 8:53 ` [QUICKLIST 0/4] Arch independent quicklists V2 Andrew Morton
2007-03-13 8:03 ` Nick Piggin
2007-03-13 11:52 ` Andrew Morton
2007-03-13 11:06 ` Nick Piggin
2007-03-13 12:15 ` Andrew Morton
2007-03-13 11:20 ` Christoph Lameter
2007-03-13 12:30 ` Andrew Morton
2007-03-15 20:23 ` Christoph Lameter
2007-03-13 11:30 ` Nick Piggin [this message]
2007-03-13 12:47 ` Andrew Morton
2007-03-13 12:01 ` Nick Piggin
2007-03-13 13:11 ` Andrew Morton
2007-03-13 12:18 ` Nick Piggin
2007-03-13 17:30 ` Jeremy Fitzhardinge
2007-03-13 20:03 ` Matt Mackall
2007-03-13 20:17 ` Jeremy Fitzhardinge
2007-03-13 20:21 ` Matt Mackall
2007-03-13 21:07 ` David Miller
2007-03-13 21:14 ` Matt Mackall
2007-03-13 21:36 ` Jeremy Fitzhardinge
2007-03-13 21:46 ` Peter Chubb
2007-03-13 21:48 ` David Miller
2007-03-14 1:12 ` William Lee Irwin III
2007-03-15 23:12 ` William Lee Irwin III
2007-03-13 23:58 ` Paul Mackerras
2007-03-13 11:17 ` Christoph Lameter
2007-03-13 12:27 ` Andrew Morton
2007-03-15 20:28 ` Christoph Lameter
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=45F68B4B.9020200@yahoo.com.au \
--to=nickpiggin@yahoo.com.au \
--cc=akpm@linux-foundation.org \
--cc=clameter@sgi.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.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).