From: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
To: Dave Hansen <dave@sr71.net>
Cc: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>,
Andrea Arcangeli <aarcange@redhat.com>,
Andrew Morton <akpm@linux-foundation.org>,
Al Viro <viro@zeniv.linux.org.uk>,
Hugh Dickins <hughd@google.com>,
Wu Fengguang <fengguang.wu@intel.com>, Jan Kara <jack@suse.cz>,
Mel Gorman <mgorman@suse.de>,
linux-mm@kvack.org, Andi Kleen <ak@linux.intel.com>,
Matthew Wilcox <matthew.r.wilcox@intel.com>,
"Kirill A. Shutemov" <kirill@shutemov.name>,
Hillf Danton <dhillf@gmail.com>,
linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCHv4 04/39] radix-tree: implement preload for multiple contiguous elements
Date: Wed, 22 May 2013 15:03:56 +0300 (EEST) [thread overview]
Message-ID: <20130522120356.9CB12E0090@blue.fi.intel.com> (raw)
In-Reply-To: <519BC3BE.3070702@sr71.net>
Dave Hansen wrote:
> On 05/11/2013 06:23 PM, Kirill A. Shutemov wrote:
> > This patch introduces radix_tree_preload_count(). It allows to
> > preallocate nodes enough to insert a number of *contiguous* elements.
>
> Would radix_tree_preload_contig() be a better name, then?
Yes. Will rename.
> ...
> > On 64-bit system:
> > For RADIX_TREE_MAP_SHIFT=3, old array size is 43, new is 107.
> > For RADIX_TREE_MAP_SHIFT=4, old array size is 31, new is 63.
> > For RADIX_TREE_MAP_SHIFT=6, old array size is 21, new is 30.
> >
> > On 32-bit system:
> > For RADIX_TREE_MAP_SHIFT=3, old array size is 21, new is 84.
> > For RADIX_TREE_MAP_SHIFT=4, old array size is 15, new is 46.
> > For RADIX_TREE_MAP_SHIFT=6, old array size is 11, new is 19.
> >
> > On most machines we will have RADIX_TREE_MAP_SHIFT=6.
>
> Thanks for adding that to the description. The array you're talking
> about is just pointers, right?
>
> 107-43 = 64. So, we have 64 extra pointers * NR_CPUS, plus 64 extra
> radix tree nodes that we will keep around most of the time. On x86_64,
> that's 512 bytes plus 64*560 bytes of nodes which is ~35k of memory per CPU.
>
> That's not bad I guess, but I do bet it's something that some folks want
> to configure out. Please make sure to call out the actual size cost in
> bytes per CPU in future patch postings, at least for the common case
> (64-bit non-CONFIG_BASE_SMALL).
I will add this to the commit message:
On most machines we will have RADIX_TREE_MAP_SHIFT=6. In this case,
on 64-bit system the per-CPU feature overhead is
for preload array:
(30 - 21) * sizeof(void*) = 72 bytes
plus, if the preload array is full
(30 - 21) * sizeof(struct radix_tree_node) = 9 * 560 = 5040 bytes
total: 5112 bytes
on 32-bit system the per-CPU feature overhead is
for preload array:
(19 - 11) * sizeof(void*) = 32 bytes
plus, if the preload array is full
(19 - 11) * sizeof(struct radix_tree_node) = 8 * 296 = 2368 bytes
total: 2400 bytes
---
Is it good enough?
I probably, will add !BASE_SMALL dependency to
TRANSPARENT_HUGEPAGE_PAGECACHE config option.
>
> > Since only THP uses batched preload at the , we disable (set max preload
> > to 1) it if !CONFIG_TRANSPARENT_HUGEPAGE_PAGECACHE. This can be changed
> > in the future.
>
> "at the..." Is there something missing in that sentence?
at the moment :)
> No major nits, so:
>
> Acked-by: Dave Hansen <dave.hansen@linux.intel.com>
Thanks!
--
Kirill A. Shutemov
--
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:[~2013-05-22 12:01 UTC|newest]
Thread overview: 121+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-05-12 1:22 [PATCHv4 00/39] Transparent huge page cache Kirill A. Shutemov
2013-05-12 1:22 ` [PATCHv4 01/39] mm: drop actor argument of do_generic_file_read() Kirill A. Shutemov
2013-05-21 18:22 ` Dave Hansen
2013-05-12 1:22 ` [PATCHv4 02/39] block: implement add_bdi_stat() Kirill A. Shutemov
2013-05-21 18:25 ` Dave Hansen
2013-05-22 11:06 ` Kirill A. Shutemov
2013-05-12 1:23 ` [PATCHv4 03/39] mm: implement zero_huge_user_segment and friends Kirill A. Shutemov
2013-05-23 10:32 ` Hillf Danton
2013-05-23 11:32 ` Kirill A. Shutemov
2013-05-12 1:23 ` [PATCHv4 04/39] radix-tree: implement preload for multiple contiguous elements Kirill A. Shutemov
2013-05-21 18:58 ` Dave Hansen
2013-05-22 12:03 ` Kirill A. Shutemov [this message]
2013-05-22 14:20 ` Dave Hansen
2013-05-12 1:23 ` [PATCHv4 05/39] memcg, thp: charge huge cache pages Kirill A. Shutemov
2013-05-21 19:04 ` Dave Hansen
2013-05-12 1:23 ` [PATCHv4 06/39] thp, mm: avoid PageUnevictable on active/inactive lru lists Kirill A. Shutemov
2013-05-21 19:17 ` Dave Hansen
2013-05-22 12:34 ` Kirill A. Shutemov
2013-05-12 1:23 ` [PATCHv4 07/39] thp, mm: basic defines for transparent huge page cache Kirill A. Shutemov
2013-05-23 10:36 ` Hillf Danton
2013-05-23 15:49 ` Dave Hansen
2013-05-12 1:23 ` [PATCHv4 08/39] thp: compile-time and sysfs knob for thp pagecache Kirill A. Shutemov
2013-05-22 11:19 ` Hillf Danton
2013-05-12 1:23 ` [PATCHv4 09/39] thp, mm: introduce mapping_can_have_hugepages() predicate Kirill A. Shutemov
2013-05-21 19:28 ` Dave Hansen
2013-05-22 13:51 ` Kirill A. Shutemov
2013-05-22 15:31 ` Dave Hansen
2013-05-12 1:23 ` [PATCHv4 10/39] thp: account anon transparent huge pages into NR_ANON_PAGES Kirill A. Shutemov
2013-05-21 19:32 ` Dave Hansen
2013-05-12 1:23 ` [PATCHv4 11/39] thp: represent file thp pages in meminfo and friends Kirill A. Shutemov
2013-05-21 19:34 ` Dave Hansen
2013-05-12 1:23 ` [PATCHv4 12/39] thp, mm: rewrite add_to_page_cache_locked() to support huge pages Kirill A. Shutemov
2013-05-21 19:59 ` Dave Hansen
2013-05-23 14:36 ` Kirill A. Shutemov
2013-05-23 16:00 ` Dave Hansen
2013-05-28 11:59 ` Kirill A. Shutemov
2013-05-12 1:23 ` [PATCHv4 13/39] mm: trace filemap: dump page order Kirill A. Shutemov
2013-05-21 19:35 ` Dave Hansen
2013-05-12 1:23 ` [PATCHv4 14/39] thp, mm: rewrite delete_from_page_cache() to support huge pages Kirill A. Shutemov
2013-05-21 20:14 ` Dave Hansen
2013-05-28 12:28 ` Kirill A. Shutemov
2013-06-07 15:10 ` Kirill A. Shutemov
2013-06-07 15:56 ` Dave Hansen
2013-06-10 17:41 ` Kirill A. Shutemov
2013-05-12 1:23 ` [PATCHv4 15/39] thp, mm: trigger bug in replace_page_cache_page() on THP Kirill A. Shutemov
2013-05-21 20:17 ` Dave Hansen
2013-05-28 12:53 ` Kirill A. Shutemov
2013-05-28 16:33 ` Dave Hansen
2013-05-12 1:23 ` [PATCHv4 16/39] thp, mm: locking tail page is a bug Kirill A. Shutemov
2013-05-21 20:18 ` Dave Hansen
2013-05-22 14:12 ` Kirill A. Shutemov
2013-05-22 14:53 ` Dave Hansen
2013-05-12 1:23 ` [PATCHv4 17/39] thp, mm: handle tail pages in page_cache_get_speculative() Kirill A. Shutemov
2013-05-21 20:49 ` Dave Hansen
2013-06-27 12:40 ` Kirill A. Shutemov
2013-05-12 1:23 ` [PATCHv4 18/39] thp, mm: add event counters for huge page alloc on write to a file Kirill A. Shutemov
2013-05-21 20:54 ` Dave Hansen
2013-05-12 1:23 ` [PATCHv4 19/39] thp, mm: allocate huge pages in grab_cache_page_write_begin() Kirill A. Shutemov
2013-05-21 21:14 ` Dave Hansen
2013-05-30 13:20 ` Kirill A. Shutemov
2013-05-12 1:23 ` [PATCHv4 20/39] thp, mm: naive support of thp in generic read/write routines Kirill A. Shutemov
2013-05-21 21:28 ` Dave Hansen
2013-06-07 15:17 ` Kirill A. Shutemov
2013-06-07 15:29 ` Dave Hansen
2013-05-12 1:23 ` [PATCHv4 21/39] thp, libfs: initial support of thp in simple_read/write_begin/write_end Kirill A. Shutemov
2013-05-21 21:49 ` Dave Hansen
2013-05-12 1:23 ` [PATCHv4 22/39] thp: handle file pages in split_huge_page() Kirill A. Shutemov
2013-05-12 1:23 ` [PATCHv4 23/39] thp: wait_split_huge_page(): serialize over i_mmap_mutex too Kirill A. Shutemov
2013-05-21 22:05 ` Dave Hansen
2013-06-03 15:02 ` Kirill A. Shutemov
2013-06-03 15:53 ` Dave Hansen
2013-06-03 16:09 ` Kirill A. Shutemov
2013-05-12 1:23 ` [PATCHv4 24/39] thp, mm: truncate support for transparent huge page cache Kirill A. Shutemov
2013-05-21 22:39 ` Dave Hansen
2013-05-12 1:23 ` [PATCHv4 25/39] thp, mm: split huge page on mmap file page Kirill A. Shutemov
2013-05-12 1:23 ` [PATCHv4 26/39] ramfs: enable transparent huge page cache Kirill A. Shutemov
2013-05-21 22:43 ` Dave Hansen
2013-05-22 14:22 ` Kirill A. Shutemov
2013-05-22 14:55 ` Dave Hansen
2013-05-12 1:23 ` [PATCHv4 27/39] x86-64, mm: proper alignment mappings with hugepages Kirill A. Shutemov
2013-05-21 22:56 ` Dave Hansen
2013-06-25 14:56 ` Kirill A. Shutemov
2013-06-25 16:46 ` Dave Hansen
2013-05-21 23:20 ` Dave Hansen
2013-05-12 1:23 ` [PATCHv4 28/39] thp: prepare zap_huge_pmd() to uncharge file pages Kirill A. Shutemov
2013-05-22 7:26 ` Hillf Danton
2013-05-12 1:23 ` [PATCHv4 29/39] thp: move maybe_pmd_mkwrite() out of mk_huge_pmd() Kirill A. Shutemov
2013-05-21 23:23 ` Dave Hansen
2013-05-22 14:37 ` Kirill A. Shutemov
2013-05-22 14:56 ` Dave Hansen
2013-05-21 23:23 ` Dave Hansen
2013-05-12 1:23 ` [PATCHv4 30/39] thp: do_huge_pmd_anonymous_page() cleanup Kirill A. Shutemov
2013-05-22 11:45 ` Hillf Danton
2013-05-12 1:23 ` [PATCHv4 31/39] thp: consolidate code between handle_mm_fault() and do_huge_pmd_anonymous_page() Kirill A. Shutemov
2013-05-21 23:38 ` Dave Hansen
2013-05-22 6:51 ` Hillf Danton
2013-05-12 1:23 ` [PATCHv4 32/39] mm: cleanup __do_fault() implementation Kirill A. Shutemov
2013-05-21 23:57 ` Dave Hansen
2013-05-12 1:23 ` [PATCHv4 33/39] thp, mm: implement do_huge_linear_fault() Kirill A. Shutemov
2013-05-22 12:47 ` Hillf Danton
2013-05-22 15:13 ` Kirill A. Shutemov
2013-05-22 12:56 ` Hillf Danton
2013-05-22 15:14 ` Kirill A. Shutemov
2013-05-22 13:24 ` Hillf Danton
2013-05-22 15:26 ` Kirill A. Shutemov
2013-05-12 1:23 ` [PATCHv4 34/39] thp, mm: handle huge pages in filemap_fault() Kirill A. Shutemov
2013-05-22 11:37 ` Hillf Danton
2013-05-22 15:34 ` Kirill A. Shutemov
2013-05-12 1:23 ` [PATCHv4 35/39] mm: decomposite do_wp_page() and get rid of some 'goto' logic Kirill A. Shutemov
2013-05-12 1:23 ` [PATCHv4 36/39] mm: do_wp_page(): extract VM_WRITE|VM_SHARED case to separate function Kirill A. Shutemov
2013-05-12 1:23 ` [PATCHv4 37/39] thp: handle write-protect exception to file-backed huge pages Kirill A. Shutemov
2013-05-23 11:57 ` Hillf Danton
2013-05-23 12:08 ` Kirill A. Shutemov
2013-05-23 12:12 ` Hillf Danton
2013-05-23 12:33 ` Kirill A. Shutemov
2013-05-12 1:23 ` [PATCHv4 38/39] thp: vma_adjust_trans_huge(): adjust file-backed VMA too Kirill A. Shutemov
2013-05-23 11:01 ` Hillf Danton
2013-05-12 1:23 ` [PATCHv4 39/39] thp: map file-backed huge pages on fault Kirill A. Shutemov
2013-05-23 11:36 ` Hillf Danton
2013-05-23 11:48 ` Kirill A. Shutemov
2013-05-21 18:37 ` [PATCHv4 00/39] Transparent huge page cache Dave Hansen
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=20130522120356.9CB12E0090@blue.fi.intel.com \
--to=kirill.shutemov@linux.intel.com \
--cc=aarcange@redhat.com \
--cc=ak@linux.intel.com \
--cc=akpm@linux-foundation.org \
--cc=dave@sr71.net \
--cc=dhillf@gmail.com \
--cc=fengguang.wu@intel.com \
--cc=hughd@google.com \
--cc=jack@suse.cz \
--cc=kirill@shutemov.name \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=matthew.r.wilcox@intel.com \
--cc=mgorman@suse.de \
--cc=viro@zeniv.linux.org.uk \
/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