From: Andrew Morton <akpm@zip.com.au>
To: Linus Torvalds <torvalds@transmeta.com>
Cc: lkml <linux-kernel@vger.kernel.org>,
"reiserfs-dev@namesys.com" <reiserfs-dev@namesys.com>
Subject: [patch 5/15] reiserfs locking fix
Date: Sun, 19 May 2002 12:39:06 -0700 [thread overview]
Message-ID: <3CE7FF5A.FEDD69CB@zip.com.au> (raw)
reiserfs is using b_inode_buffers and fsync_buffers_list() for
attaching dependent buffers to its journal. For writeout prior to
commit.
This worked OK when a global lock was used everywhere, but the locking
is currently incorrect - try_to_free_buffers() is taking a different
lock when detaching buffers from their "foreign" inode. So list_head
corruption could occur on SMP.
The patch implements a reiserfs_releasepage() which holds the
journal-wide buffer lock while it runs try_to_free_buffers(), so all
those list_heads are protected. The lock is held across the
try_to_free_buffers() call as well, so nobody will attach one of this
page's buffers to a list while try_to_free_buffers() is running.
=====================================
--- 2.5.16/fs/reiserfs/inode.c~reiserfs_releasepage Sun May 19 11:49:47 2002
+++ 2.5.16-akpm/fs/reiserfs/inode.c Sun May 19 11:49:47 2002
@@ -2096,9 +2096,47 @@ static int reiserfs_commit_write(struct
return ret ;
}
+/*
+ * Returns 1 if the page's buffers were dropped. The page is locked.
+ *
+ * Takes j_dirty_buffers_lock to protect the b_assoc_buffers list_heads
+ * in the buffers at page_buffers(page).
+ *
+ * FIXME: Chris says the buffer list is not used with `mount -o notail',
+ * so in that case the fs can avoid the extra locking. Create a second
+ * address_space_operations with a NULL ->releasepage and install that
+ * into new address_spaces.
+ */
+static int reiserfs_releasepage(struct page *page, int unused_gfp_flags)
+{
+ struct inode *inode = page->mapping->host ;
+ struct reiserfs_journal *j = SB_JOURNAL(inode->i_sb) ;
+ struct buffer_head *head ;
+ struct buffer_head *bh ;
+ int ret = 1 ;
+
+ spin_lock(&j->j_dirty_buffers_lock) ;
+ head = page_buffers(page) ;
+ bh = head ;
+ do {
+ if (!buffer_dirty(bh) && !buffer_locked(bh)) {
+ list_del_init(&bh->b_assoc_buffers) ;
+ } else {
+ ret = 0 ;
+ break ;
+ }
+ bh = bh->b_this_page ;
+ } while (bh != head) ;
+ if (ret)
+ ret = try_to_free_buffers(page) ;
+ spin_unlock(&j->j_dirty_buffers_lock) ;
+ return ret ;
+}
+
struct address_space_operations reiserfs_address_space_operations = {
writepage: reiserfs_writepage,
readpage: reiserfs_readpage,
+ releasepage: reiserfs_releasepage,
sync_page: block_sync_page,
prepare_write: reiserfs_prepare_write,
commit_write: reiserfs_commit_write,
--- 2.5.16/fs/reiserfs/file.c~reiserfs_releasepage Sun May 19 11:49:47 2002
+++ 2.5.16-akpm/fs/reiserfs/file.c Sun May 19 11:49:47 2002
@@ -72,6 +72,12 @@ static void reiserfs_vfs_truncate_file(s
}
/* Sync a reiserfs file. */
+
+/*
+ * FIXME: sync_mapping_buffers() never has anything to sync. Can
+ * be removed...
+ */
+
static int reiserfs_sync_file(
struct file * p_s_filp,
struct dentry * p_s_dentry,
-
next reply other threads:[~2002-05-19 19:35 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2002-05-19 19:39 Andrew Morton [this message]
2002-05-21 12:29 ` [reiserfs-dev] [patch 5/15] reiserfs locking fix Chris Mason
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=3CE7FF5A.FEDD69CB@zip.com.au \
--to=akpm@zip.com.au \
--cc=linux-kernel@vger.kernel.org \
--cc=reiserfs-dev@namesys.com \
--cc=torvalds@transmeta.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).