linux-fsdevel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: "Josef 'Jeff' Sipek" <jsipek@cs.sunysb.edu>
To: linux-kernel@vger.kernel.org, linux-fsdevel@vger.kernel.org
Cc: akpm@linux-foundation.org, Erez Zadok <ezk@cs.sunysb.edu>,
	"Josef 'Jeff' Sipek" <jsipek@cs.sunysb.edu>
Subject: [PATCH 21/21] Unionfs: Correctly decrement refcounts of mnt's upon branch management
Date: Wed, 23 May 2007 20:36:11 -0400	[thread overview]
Message-ID: <11799669742009-git-send-email-jsipek@cs.sunysb.edu> (raw)
In-Reply-To: <11799669712090-git-send-email-jsipek@cs.sunysb.edu>

From: Erez Zadok <ezk@cs.sunysb.edu>

The old logic was broken in one place, which another place tried to "fix"
incorrectly.  Also added detailed comments to explain the new/correct logic.

Signed-off-by: Erez Zadok <ezk@cs.sunysb.edu>
Signed-off-by: Josef 'Jeff' Sipek <jsipek@cs.sunysb.edu>
---
 fs/unionfs/commonfops.c |   17 ++++++++++++-----
 fs/unionfs/dentry.c     |   18 +++++++++++++-----
 2 files changed, 25 insertions(+), 10 deletions(-)

diff --git a/fs/unionfs/commonfops.c b/fs/unionfs/commonfops.c
index 69d9c87..83001aa 100644
--- a/fs/unionfs/commonfops.c
+++ b/fs/unionfs/commonfops.c
@@ -135,14 +135,21 @@ static void cleanup_file(struct file *file)
 				       "file %p\n", file);
 			else {
 				unionfs_read_lock(sb);
+				/* decrement count of open files */
 				branchput(sb, i);
 				unionfs_read_unlock(sb);
-				/* XXX: is it OK to use sb->s_root here? */
-				unionfs_mntput(sb->s_root, i);
-				/* mntget b/c fput below will call mntput */
-				unionfs_mntget(sb->s_root, bindex);
+				/*
+				 * fput will perform an mntput for us on the
+				 * correct branch.  Although we're using the
+				 * file's old branch configuration, bindex,
+				 * which is the old index, correctly points
+				 * to the right branch in the file's branch
+				 * list.  In other words, we're going to
+				 * mntput the correct branch even if
+				 * branches have been added/removed.
+				 */
+				fput(unionfs_lower_file_idx(file, bindex));
 			}
-			fput(unionfs_lower_file_idx(file, bindex));
 		}
 	}
 
diff --git a/fs/unionfs/dentry.c b/fs/unionfs/dentry.c
index 95cea3b..db0ef43 100644
--- a/fs/unionfs/dentry.c
+++ b/fs/unionfs/dentry.c
@@ -258,14 +258,22 @@ out_this:
 	/* finally, lock this dentry and revalidate it */
 	verify_locked(dentry);
 	dgen = atomic_read(&UNIONFS_D(dentry)->generation);
-	saved_bstart = dbstart(dentry);
-	saved_bend = dbend(dentry);
 	valid = __unionfs_d_revalidate_one(dentry, nd);
 
-	if (valid && chain_len > 0 && sbgen != dgen) {
-		for (bindex = saved_bstart; bindex <= saved_bend; bindex++)
+	/*
+	 * If __unionfs_d_revalidate_one() succeeded above, then it will
+	 * have incremented the refcnt of the mnt's, but also the branch
+	 * indices of the dentry will have been updated (to take into
+	 * account any branch insertions/deletion.  So the current
+	 * dbstart/dbend match the current, and new, indices of the mnts
+	 * which __unionfs_d_revalidate_one has incremented.  Note: the "if"
+	 * test below does not depend on whether chain_len was 0 or greater.
+	 */
+	if (valid && sbgen != dgen)
+		for (bindex = dbstart(dentry);
+		     bindex <= dbend(dentry);
+		     bindex++)
 			unionfs_mntput(dentry, bindex);
-	}
 
 out_free:
 	/* unlock/dput all dentries in chain and return status */
-- 
1.5.2.rc1.165.gaf9b


      parent reply	other threads:[~2007-05-24  0:36 UTC|newest]

Thread overview: 22+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2007-05-24  0:35 [GIT PULL -mm] Unionfs cleanups and fixes Josef 'Jeff' Sipek
2007-05-24  0:35 ` [PATCH 01/21] Unionfs: Tiny documentation fixups Josef 'Jeff' Sipek
2007-05-24  0:35 ` [PATCH 02/21] Unionfs: Coding style fixes Josef 'Jeff' Sipek
2007-05-24  0:35 ` [PATCH 03/21] Unionfs: Every printk should prefix with "unionfs: " consistently Josef 'Jeff' Sipek
2007-05-24  0:35 ` [PATCH 04/21] Unionfs: Add missing copyright notices Josef 'Jeff' Sipek
2007-05-24  0:35 ` [PATCH 05/21] Unionfs: Cleanup of strings and comments Josef 'Jeff' Sipek
2007-05-24  0:35 ` [PATCH 06/21] Unionfs: Added numerous comments Josef 'Jeff' Sipek
2007-05-24  0:35 ` [PATCH 07/21] Unionfs: Consistent pointer declaration spacing Josef 'Jeff' Sipek
2007-05-24  0:35 ` [PATCH 08/21] Unionfs: Rename Unionfs's double_lock_dentry to avoid confusion Josef 'Jeff' Sipek
2007-05-24  0:35 ` [PATCH 09/21] Unionfs: Rename our "do_rename" to __unionfs_rename Josef 'Jeff' Sipek
2007-05-24  0:36 ` [PATCH 10/21] Unionfs: Move unionfs_query_file to commonfops.c Josef 'Jeff' Sipek
2007-05-24  0:36 ` [PATCH 11/21] Unionfs: Combine unionfs_write with __unionfs_write Josef 'Jeff' Sipek
2007-05-24  0:36 ` [PATCH 12/21] Unionfs: Prefix external functions with 'extern' properly Josef 'Jeff' Sipek
2007-05-24  0:36 ` [PATCH 13/21] Unionfs: Don't leak resources when copyup fails partially Josef 'Jeff' Sipek
2007-05-24  0:36 ` [PATCH 14/21] Unionfs: Call realloc unconditionally Josef 'Jeff' Sipek
2007-05-24  0:36 ` [PATCH 15/21] Unionfs: Use krealloc instead of open-coding the functionality Josef 'Jeff' Sipek
2007-05-24  0:36 ` [PATCH 16/21] Unionfs: Disallow setting leftmost branch to readonly Josef 'Jeff' Sipek
2007-05-24  0:36 ` [PATCH 17/21] Unionfs: Documentation update regarding overlapping branches and new lookup code Josef 'Jeff' Sipek
2007-05-24  0:36 ` [PATCH 18/21] Unionfs: Remove defunct unionfs_put_inode super op Josef 'Jeff' Sipek
2007-05-24  0:36 ` [PATCH 19/21] Unionfs: Actually catch bad use of unionfs_mnt{get,put} Josef 'Jeff' Sipek
2007-05-24  0:36 ` [PATCH 20/21] Unionfs: Removed a trailing whitespace Josef 'Jeff' Sipek
2007-05-24  0:36 ` Josef 'Jeff' Sipek [this message]

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=11799669742009-git-send-email-jsipek@cs.sunysb.edu \
    --to=jsipek@cs.sunysb.edu \
    --cc=akpm@linux-foundation.org \
    --cc=ezk@cs.sunysb.edu \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.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;
as well as URLs for NNTP newsgroup(s).