From: Badari Pulavarty <pbadari@gmail.com>
To: Ken Chen <kenchen@google.com>
Cc: "Siddha, Suresh B" <suresh.b.siddha@intel.com>,
linux-mm <linux-mm@kvack.org>
Subject: Re: [rfc] more granular page table lock for hugepages
Date: Tue, 09 Oct 2007 14:05:57 -0700 [thread overview]
Message-ID: <1191963958.12131.43.camel@dyn9047017100.beaverton.ibm.com> (raw)
In-Reply-To: <b040c32a0710091323v7fab02b0vaab61f0ea12278d@mail.gmail.com>
On Tue, 2007-10-09 at 13:23 -0700, Ken Chen wrote:
> On 10/8/07, Siddha, Suresh B <suresh.b.siddha@intel.com> wrote:
> > Appended patch is a quick prototype which extends the concept of separate
> > spinlock per page table page to hugepages. More granular spinlock will
> > be used to guard the page table entries in the pmd page, instead of using the
> > mm's single page_table_lock.
>
> What path do you content on mm->page_table_lock?
>
> The major fault for hugetlb page is blanket by
> hugetlb_instantiation_mutex. So likelihood of contention on
> page_table spin lock is low. For minor fault, I would think
> mapping->i_mmap_lock will kick in before page table lock. That left
> follow_hugetlb_page path. Is it the case?
Yes. follow_hugetlb_page() is where our benchmark team has seen
contention with threaded workload.
Thanks,
Badari
--
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:[~2007-10-09 21:02 UTC|newest]
Thread overview: 30+ messages / expand[flat|nested] mbox.gz Atom feed top
2007-10-08 22:52 Siddha, Suresh B
2007-10-09 20:23 ` Ken Chen
2007-10-09 21:05 ` Badari Pulavarty [this message]
2007-10-10 0:15 ` Siddha, Suresh B
2007-10-10 6:10 ` Ken Chen
2007-10-10 7:50 ` Ken Chen
2007-10-11 11:39 ` Nick Piggin
2007-10-12 20:34 ` Siddha, Suresh B
2007-10-13 23:27 ` Nick Piggin
2007-10-14 1:01 ` [rfc] lockless get_user_pages for dio (and more) Nick Piggin
2007-10-14 18:19 ` Siddha, Suresh B
2007-10-15 4:15 ` Nick Piggin
2007-10-15 12:25 ` Nick Piggin
2007-10-15 17:03 ` Badari Pulavarty
2007-10-15 17:49 ` Siddha, Suresh B
2007-10-15 17:54 ` Siddha, Suresh B
2007-10-15 20:21 ` Ken Chen
2007-10-16 2:15 ` Nick Piggin
2007-10-16 0:14 ` Dave Hansen
2007-10-16 3:26 ` Nick Piggin
2007-10-16 3:32 ` Nick Piggin
2007-12-10 21:30 ` Dave Kleikamp
2007-12-12 4:57 ` Nick Piggin
2007-12-12 5:11 ` Dave Kleikamp
2007-12-12 5:40 ` Nick Piggin
2008-01-16 19:58 ` Dave Kleikamp
2008-01-17 6:34 ` Nick Piggin
2008-01-24 7:06 ` Nick Piggin
2007-10-14 15:42 ` [rfc] more granular page table lock for hugepages Siddha, Suresh B
2007-10-15 4:17 ` Nick Piggin
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=1191963958.12131.43.camel@dyn9047017100.beaverton.ibm.com \
--to=pbadari@gmail.com \
--cc=kenchen@google.com \
--cc=linux-mm@kvack.org \
--cc=suresh.b.siddha@intel.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