From: Nick Piggin <nickpiggin@yahoo.com.au>
To: Linus Torvalds <torvalds@osdl.org>
Cc: Andrew Morton <akpm@osdl.org>,
linux-kernel@vger.kernel.org, linux-mm@kvack.org, mason@suse.com,
andrea@suse.de, hugh@veritas.com, axboe@suse.de
Subject: Re: [rfc][patch] remove racy sync_page?
Date: Tue, 30 May 2006 15:36:14 +1000 [thread overview]
Message-ID: <447BD9CE.2020505@yahoo.com.au> (raw)
In-Reply-To: <447BD31E.7000503@yahoo.com.au>
[-- Attachment #1: Type: text/plain, Size: 419 bytes --]
Nick Piggin wrote:
> Anyway, clearly a plug friendly bugfix needs to be implemented
> regardless.
I actually think it would be cleaner to introduce a new
mapping_lock_page(mapping, page) or something that does the
unplug on the _mapping_, and have lock_page be callable with
no ref on the inode. Should be a mechanical conversion for
most of the safe callers.
But for 2.6.17, how's this?
--
SUSE Labs, Novell Inc.
[-- Attachment #2: mm-non-syncing-lock_page.patch --]
[-- Type: text/plain, Size: 2455 bytes --]
lock_page needs the caller to have a reference on the page->mapping inode
due to sync_page. So set_page_dirty_lock is obviously buggy according to
its comments.
Solve it by introducing a new lock_page_nosync which does not do a sync_page.
Signed-off-by: Nick Piggin <npiggin@suse.de>
Index: linux-2.6/include/linux/pagemap.h
===================================================================
--- linux-2.6.orig/include/linux/pagemap.h 2006-05-19 12:49:06.000000000 +1000
+++ linux-2.6/include/linux/pagemap.h 2006-05-30 15:12:54.000000000 +1000
@@ -168,6 +168,7 @@ static inline pgoff_t linear_page_index(
}
extern void FASTCALL(__lock_page(struct page *page));
+extern void FASTCALL(__lock_page_nosync(struct page *page));
extern void FASTCALL(unlock_page(struct page *page));
static inline void lock_page(struct page *page)
@@ -176,6 +177,13 @@ static inline void lock_page(struct page
if (TestSetPageLocked(page))
__lock_page(page);
}
+
+static inline void lock_page_nosync(struct page *page)
+{
+ might_sleep();
+ if (TestSetPageLocked(page))
+ __lock_page_nosync(page);
+}
/*
* This is exported only for wait_on_page_locked/wait_on_page_writeback.
Index: linux-2.6/mm/filemap.c
===================================================================
--- linux-2.6.orig/mm/filemap.c 2006-05-30 15:07:58.000000000 +1000
+++ linux-2.6/mm/filemap.c 2006-05-30 15:31:46.000000000 +1000
@@ -544,6 +544,23 @@ void fastcall __lock_page(struct page *p
}
EXPORT_SYMBOL(__lock_page);
+static int __sleep_on_page_lock(void *word)
+{
+ io_schedule();
+ return 0;
+}
+
+/*
+ * Variant of lock_page that does not require the caller to hold a reference
+ * on the page's mapping.
+ */
+void fastcall __lock_page_nosync(struct page *page)
+{
+ DEFINE_WAIT_BIT(wait, &page->flags, PG_locked);
+ __wait_on_bit_lock(page_waitqueue(page), &wait, __sleep_on_page_lock,
+ TASK_UNINTERRUPTIBLE);
+}
+
/*
* a rather lightweight function, finding and getting a reference to a
* hashed page atomically.
Index: linux-2.6/mm/page-writeback.c
===================================================================
--- linux-2.6.orig/mm/page-writeback.c 2006-05-19 12:48:01.000000000 +1000
+++ linux-2.6/mm/page-writeback.c 2006-05-30 15:13:59.000000000 +1000
@@ -702,7 +702,7 @@ int set_page_dirty_lock(struct page *pag
{
int ret;
- lock_page(page);
+ lock_page_nosync(page);
ret = set_page_dirty(page);
unlock_page(page);
return ret;
next prev parent reply other threads:[~2006-05-30 5:36 UTC|newest]
Thread overview: 54+ messages / expand[flat|nested] mbox.gz Atom feed top
2006-05-29 9:34 Nick Piggin
2006-05-29 19:15 ` Andrew Morton
2006-05-30 0:08 ` Nick Piggin
2006-05-30 1:32 ` Andrew Morton
2006-05-30 2:54 ` Nick Piggin
2006-05-30 3:14 ` Andrew Morton
2006-05-30 4:13 ` Nick Piggin
2006-05-30 9:05 ` Jens Axboe
2006-05-31 13:43 ` Nick Piggin
2006-05-31 15:09 ` Hugh Dickins
2006-05-31 15:22 ` Nick Piggin
2006-05-31 17:51 ` Jens Axboe
2006-05-31 17:50 ` Jens Axboe
2006-05-30 4:20 ` Linus Torvalds
2006-05-30 5:07 ` Nick Piggin
2006-05-30 5:21 ` Nick Piggin
2006-05-30 6:12 ` Neil Brown
2006-05-30 7:10 ` Nick Piggin
2006-05-31 4:34 ` Neil Brown
2006-05-30 8:24 ` Nikita Danilov
2006-05-30 17:55 ` Linus Torvalds
2006-05-31 0:32 ` Nick Piggin
2006-05-31 0:56 ` Linus Torvalds
2006-05-31 1:33 ` Mark Lord
2006-05-31 6:11 ` Jens Axboe
2006-05-31 12:55 ` Mark Lord
2006-05-31 13:02 ` Jens Axboe
2006-06-01 13:19 ` NCQ performance (was Re: [rfc][patch] remove racy sync_page?) Jens Axboe
2006-06-01 14:56 ` Avi Kivity
2006-06-01 15:03 ` Jens Axboe
2006-06-01 18:04 ` Jens Axboe
2006-06-05 5:30 ` Avi Kivity
2006-06-05 7:59 ` Jens Axboe
2006-05-31 12:31 ` [rfc][patch] remove racy sync_page? Helge Hafting
2006-05-31 12:36 ` Arjan van de Ven
2006-05-31 13:29 ` Nick Piggin
2006-05-31 13:41 ` Jens Axboe
2006-05-31 13:54 ` Nick Piggin
2006-05-31 14:43 ` Linus Torvalds
2006-05-31 14:57 ` Nick Piggin
2006-05-31 15:13 ` Linus Torvalds
2006-05-31 15:09 ` Linus Torvalds
2006-05-31 18:13 ` Jens Axboe
2006-05-31 18:26 ` Linus Torvalds
2006-05-30 5:36 ` Nick Piggin [this message]
2006-05-30 18:31 ` Hugh Dickins
2006-05-31 0:21 ` Nick Piggin
2006-05-31 3:06 ` Hugh Dickins
2006-05-31 14:30 ` Hugh Dickins
2006-05-31 17:56 ` Jens Axboe
2006-05-30 5:51 ` Josef Sipek
2006-05-30 6:44 ` Nick Piggin
2006-05-30 6:50 ` Nick Piggin
2006-05-30 13:12 ` Josef Sipek
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=447BD9CE.2020505@yahoo.com.au \
--to=nickpiggin@yahoo.com.au \
--cc=akpm@osdl.org \
--cc=andrea@suse.de \
--cc=axboe@suse.de \
--cc=hugh@veritas.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mason@suse.com \
--cc=torvalds@osdl.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