All of lore.kernel.org
 help / color / mirror / Atom feed
From: Kirill Korotaev <dev@openvz.org>
To: Andrew Morton <akpm@osdl.org>,
	viro@ftp.linux.org.uk, linux-kernel@vger.kernel.org,
	Andrey Savochkin <saw@saw.sw.com.sg>,
	devel@openvz.org
Subject: [PATCH] ext3: ext3_symlink should use GFP_NOFS allocations inside
Date: Fri, 10 Mar 2006 11:08:29 +0300	[thread overview]
Message-ID: <441133FD.2070808@openvz.org> (raw)

[-- Attachment #1: Type: text/plain, Size: 597 bytes --]

Andrew,

Resending a patch with a separate helper as requested by Al Viro:

This patch fixes illegal __GFP_FS allocation inside ext3
transaction in ext3_symlink().
Such allocation may re-enter ext3 code from
try_to_free_pages. But JBD/ext3 code keeps a pointer to current
journal handle in task_struct and, hence, is not reentrable.

This bug led to "Assertion failure in journal_dirty_metadata()" messages.

http://bugzilla.openvz.org/show_bug.cgi?id=115

Signed-Off-By: Andrey Savochkin <saw@saw.sw.com.sg>
Signed-Off-By: Kirill Korotaev <dev@openvz.org>

Thanks,
Kirill
P.S. against 2.6.16-rc5

[-- Attachment #2: diff-ext3-nofssymlink-20060210 --]
[-- Type: text/plain, Size: 2462 bytes --]

--- ./fs/ext3/namei.c.symlnkfix	2006-03-10 10:24:05.000000000 +0300
+++ ./fs/ext3/namei.c	2006-03-10 10:24:49.000000000 +0300
@@ -2141,7 +2141,7 @@ retry:
 		 * We have a transaction open.  All is sweetness.  It also sets
 		 * i_size in generic_commit_write().
 		 */
-		err = page_symlink(inode, symname, l);
+		err = __page_symlink(inode, symname, l, GFP_NOFS);
 		if (err) {
 			ext3_dec_count(handle, inode);
 			ext3_mark_inode_dirty(handle, inode);
--- ./fs/namei.c.symlnkfix	2006-03-10 10:24:05.000000000 +0300
+++ ./fs/namei.c	2006-03-10 10:34:58.000000000 +0300
@@ -2613,13 +2613,16 @@ void page_put_link(struct dentry *dentry
 	}
 }
 
-int page_symlink(struct inode *inode, const char *symname, int len)
+int __page_symlink(struct inode *inode, const char *symname, int len,
+		gfp_t gfp_mask)
 {
 	struct address_space *mapping = inode->i_mapping;
-	struct page *page = grab_cache_page(mapping, 0);
+	struct page *page;
 	int err = -ENOMEM;
 	char *kaddr;
 
+	page = find_or_create_page(mapping, 0,
+			mapping_gfp_mask(mapping) | gfp_mask);
 	if (!page)
 		goto fail;
 	err = mapping->a_ops->prepare_write(NULL, page, 0, len-1);
@@ -2654,6 +2657,11 @@ fail:
 	return err;
 }
 
+int page_symlink(struct inode *inode, const char *symname, int len)
+{
+	return __page_symlink(inode, symname, len, GFP_KERNEL);
+}
+
 struct inode_operations page_symlink_inode_operations = {
 	.readlink	= generic_readlink,
 	.follow_link	= page_follow_link_light,
@@ -2672,6 +2680,7 @@ EXPORT_SYMBOL(lookup_one_len);
 EXPORT_SYMBOL(page_follow_link_light);
 EXPORT_SYMBOL(page_put_link);
 EXPORT_SYMBOL(page_readlink);
+EXPORT_SYMBOL(__page_symlink);
 EXPORT_SYMBOL(page_symlink);
 EXPORT_SYMBOL(page_symlink_inode_operations);
 EXPORT_SYMBOL(path_lookup);
--- ./include/linux/fs.h.symlnkfix	2006-03-10 10:24:05.000000000 +0300
+++ ./include/linux/fs.h	2006-03-10 10:27:40.000000000 +0300
@@ -1669,6 +1669,8 @@ extern int vfs_follow_link(struct nameid
 extern int page_readlink(struct dentry *, char __user *, int);
 extern void *page_follow_link_light(struct dentry *, struct nameidata *);
 extern void page_put_link(struct dentry *, struct nameidata *, void *);
+extern int __page_symlink(struct inode *inode, const char *symname, int len,
+		gfp_t gfp_mask);
 extern int page_symlink(struct inode *inode, const char *symname, int len);
 extern struct inode_operations page_symlink_inode_operations;
 extern int generic_readlink(struct dentry *, char __user *, int);

             reply	other threads:[~2006-03-10  8:03 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2006-03-10  8:08 Kirill Korotaev [this message]
2006-03-10  8:31 ` [PATCH] ext3: ext3_symlink should use GFP_NOFS allocations inside Al Viro
2006-03-10  8:46   ` Kirill Korotaev
  -- strict thread matches above, loose matches on Subject: below --
2006-03-09 17:19 Kirill Korotaev
2006-03-09 21:04 ` Al Viro
2006-03-10 13:02 ` Michael Tokarev
2006-03-10 20:21   ` Andrew Morton

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=441133FD.2070808@openvz.org \
    --to=dev@openvz.org \
    --cc=akpm@osdl.org \
    --cc=devel@openvz.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=saw@saw.sw.com.sg \
    --cc=viro@ftp.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.