From: Izik Eidus <ieidus@redhat.com>
To: akpm@linux-foundation.org
Cc: hugh.dickins@tiscali.co.uk, aarcange@redhat.com,
chrisw@redhat.com, avi@redhat.com, riel@redhat.com,
linux-kernel@vger.kernel.org, linux-mm@kvack.org,
nickpiggin@yahoo.com.au, ieidus@redhat.com,
Wu Fengguang <fengguang.wu@intel.com>
Subject: [PATCH 06/10] ksm: identify PageKsm pages
Date: Fri, 17 Jul 2009 20:30:46 +0300 [thread overview]
Message-ID: <1247851850-4298-7-git-send-email-ieidus@redhat.com> (raw)
In-Reply-To: <1247851850-4298-6-git-send-email-ieidus@redhat.com>
From: Hugh Dickins <hugh.dickins@tiscali.co.uk>
KSM will need to identify its kernel merged pages unambiguously,
and /proc/kpageflags will probably like to do so too.
Since KSM will only be substituting anonymous pages, statistics are
best preserved by making a PageKsm page a special PageAnon page:
one with no anon_vma.
But KSM then needs its own page_add_ksm_rmap() - keep it in ksm.h near
PageKsm; and do_wp_page() must COW them, unlike singly mapped PageAnons.
Signed-off-by: Hugh Dickins <hugh.dickins@tiscali.co.uk>
Signed-off-by: Chris Wright <chrisw@redhat.com>
Signed-off-by: Izik Eidus <ieidus@redhat.com>
Cc: Wu Fengguang <fengguang.wu@intel.com>
---
fs/proc/page.c | 5 +++++
include/linux/ksm.h | 29 +++++++++++++++++++++++++++++
mm/memory.c | 3 ++-
3 files changed, 36 insertions(+), 1 deletions(-)
diff --git a/fs/proc/page.c b/fs/proc/page.c
index 2707c6c..2281c2c 100644
--- a/fs/proc/page.c
+++ b/fs/proc/page.c
@@ -2,6 +2,7 @@
#include <linux/compiler.h>
#include <linux/fs.h>
#include <linux/init.h>
+#include <linux/ksm.h>
#include <linux/mm.h>
#include <linux/mmzone.h>
#include <linux/proc_fs.h>
@@ -95,6 +96,8 @@ static const struct file_operations proc_kpagecount_operations = {
#define KPF_UNEVICTABLE 18
#define KPF_NOPAGE 20
+#define KPF_KSM 21
+
/* kernel hacking assistances
* WARNING: subject to change, never rely on them!
*/
@@ -137,6 +140,8 @@ static u64 get_uflags(struct page *page)
u |= 1 << KPF_MMAP;
if (PageAnon(page))
u |= 1 << KPF_ANON;
+ if (PageKsm(page))
+ u |= 1 << KPF_KSM;
/*
* compound pages: export both head/tail info
diff --git a/include/linux/ksm.h b/include/linux/ksm.h
index eb2a448..a485c14 100644
--- a/include/linux/ksm.h
+++ b/include/linux/ksm.h
@@ -10,6 +10,7 @@
#include <linux/bitops.h>
#include <linux/mm.h>
#include <linux/sched.h>
+#include <linux/vmstat.h>
#ifdef CONFIG_KSM
int ksm_madvise(struct vm_area_struct *vma, unsigned long start,
@@ -29,6 +30,27 @@ static inline void ksm_exit(struct mm_struct *mm)
if (test_bit(MMF_VM_MERGEABLE, &mm->flags))
__ksm_exit(mm);
}
+
+/*
+ * A KSM page is one of those write-protected "shared pages" or "merged pages"
+ * which KSM maps into multiple mms, wherever identical anonymous page content
+ * is found in VM_MERGEABLE vmas. It's a PageAnon page, with NULL anon_vma.
+ */
+static inline int PageKsm(struct page *page)
+{
+ return ((unsigned long)page->mapping == PAGE_MAPPING_ANON);
+}
+
+/*
+ * But we have to avoid the checking which page_add_anon_rmap() performs.
+ */
+static inline void page_add_ksm_rmap(struct page *page)
+{
+ if (atomic_inc_and_test(&page->_mapcount)) {
+ page->mapping = (void *) PAGE_MAPPING_ANON;
+ __inc_zone_page_state(page, NR_ANON_PAGES);
+ }
+}
#else /* !CONFIG_KSM */
static inline int ksm_madvise(struct vm_area_struct *vma, unsigned long start,
@@ -45,6 +67,13 @@ static inline int ksm_fork(struct mm_struct *mm, struct mm_struct *oldmm)
static inline void ksm_exit(struct mm_struct *mm)
{
}
+
+static inline int PageKsm(struct page *page)
+{
+ return 0;
+}
+
+/* No stub required for page_add_ksm_rmap(page) */
#endif /* !CONFIG_KSM */
#endif
diff --git a/mm/memory.c b/mm/memory.c
index 8b1922c..6707072 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -45,6 +45,7 @@
#include <linux/swap.h>
#include <linux/highmem.h>
#include <linux/pagemap.h>
+#include <linux/ksm.h>
#include <linux/rmap.h>
#include <linux/module.h>
#include <linux/delayacct.h>
@@ -1972,7 +1973,7 @@ static int do_wp_page(struct mm_struct *mm, struct vm_area_struct *vma,
* Take out anonymous pages first, anonymous shared vmas are
* not dirty accountable.
*/
- if (PageAnon(old_page)) {
+ if (PageAnon(old_page) && !PageKsm(old_page)) {
if (!trylock_page(old_page)) {
page_cache_get(old_page);
pte_unmap_unlock(page_table, ptl);
--
1.5.6.5
--
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:[~2009-07-17 17:28 UTC|newest]
Thread overview: 42+ messages / expand[flat|nested] mbox.gz Atom feed top
2009-07-17 17:30 [PATCH 00/10] ksm resend Izik Eidus
2009-07-17 17:30 ` [PATCH 01/10] ksm: add mmu_notifier set_pte_at_notify() Izik Eidus
2009-07-17 17:30 ` [PATCH 02/10] ksm: first tidy up madvise_vma() Izik Eidus
2009-07-17 17:30 ` [PATCH 03/10] ksm: define MADV_MERGEABLE and MADV_UNMERGEABLE Izik Eidus
2009-07-17 17:30 ` [PATCH 04/10] ksm: the mm interface to ksm Izik Eidus
2009-07-17 17:30 ` [PATCH 05/10] ksm: no debug in page_dup_rmap() Izik Eidus
2009-07-17 17:30 ` Izik Eidus [this message]
2009-07-17 17:30 ` [PATCH 07/10] ksm: Kernel SamePage Merging Izik Eidus
2009-07-17 17:30 ` [PATCH 08/10] ksm: prevent mremap move poisoning Izik Eidus
2009-07-17 17:30 ` [PATCH 09/10] ksm: change copyright message Izik Eidus
2009-07-17 17:30 ` [PATCH 10/10] ksm: change ksm nice level to be 5 Izik Eidus
2009-07-19 13:50 ` Hugh Dickins
2009-07-20 4:50 ` Balbir Singh
2009-07-20 11:48 ` Izik Eidus
2009-07-20 12:14 ` Balbir Singh
2009-07-20 18:38 ` Rik van Riel
2009-07-19 13:49 ` [PATCH 09/10] ksm: change copyright message Hugh Dickins
2009-07-20 18:37 ` Rik van Riel
2009-07-20 18:37 ` [PATCH 08/10] ksm: prevent mremap move poisoning Rik van Riel
2009-07-20 18:35 ` [PATCH 07/10] ksm: Kernel SamePage Merging Rik van Riel
2009-07-18 2:45 ` [PATCH 06/10] ksm: identify PageKsm pages Wu Fengguang
2009-07-20 18:32 ` Rik van Riel
2009-07-21 17:51 ` Andrea Arcangeli
2009-07-21 17:55 ` Rik van Riel
2009-07-21 18:01 ` Andrea Arcangeli
2009-07-21 18:17 ` Izik Eidus
2009-07-22 12:54 ` Hugh Dickins
2009-07-23 2:06 ` KAMEZAWA Hiroyuki
2009-07-23 11:43 ` Hugh Dickins
2009-07-23 11:49 ` Lee Schermerhorn
2009-07-22 12:45 ` Hugh Dickins
2009-07-22 16:52 ` Andrea Arcangeli
2009-07-23 11:36 ` Hugh Dickins
2009-07-20 18:11 ` [PATCH 05/10] ksm: no debug in page_dup_rmap() Rik van Riel
2009-07-21 7:16 ` Nick Piggin
2009-07-20 17:38 ` [PATCH 04/10] ksm: the mm interface to ksm Rik van Riel
2009-07-20 15:09 ` [PATCH 03/10] ksm: define MADV_MERGEABLE and MADV_UNMERGEABLE Rik van Riel
2009-07-20 15:44 ` Ralf Baechle
2009-07-20 15:05 ` [PATCH 02/10] ksm: first tidy up madvise_vma() Rik van Riel
2009-07-20 14:48 ` [PATCH 01/10] ksm: add mmu_notifier set_pte_at_notify() Rik van Riel
2009-07-21 17:59 ` [PATCH 00/10] ksm resend Andrea Arcangeli
2009-07-22 13:05 ` Hugh Dickins
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=1247851850-4298-7-git-send-email-ieidus@redhat.com \
--to=ieidus@redhat.com \
--cc=aarcange@redhat.com \
--cc=akpm@linux-foundation.org \
--cc=avi@redhat.com \
--cc=chrisw@redhat.com \
--cc=fengguang.wu@intel.com \
--cc=hugh.dickins@tiscali.co.uk \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=nickpiggin@yahoo.com.au \
--cc=riel@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).