diff mbox series

[UBUNTU,SAUCE,SRU,B/F/G/H/I] ext4: fix directory index node split corruption

Message ID 20210706133602.30706-2-colin.king@canonical.com
State New
Headers show
Series [UBUNTU,SAUCE,SRU,B/F/G/H/I] ext4: fix directory index node split corruption | expand

Commit Message

Colin Ian King July 6, 2021, 1:36 p.m. UTC 
From: Colin Ian King <colin.king@canonical.com>

BugLink: https://bugs.launchpad.net/bugs/1933074

[ Note: This was originally sent upstream to linux-ext4@vger.kernel.org
  and Ted Tso but didn't seem to make it to the list, presumably as it
  was marked as SPAM

  See: http://www.voxelsoft.com/2021/ext4_large_dir_corruption.html ]

Following commit b5776e7, a trivial sequential write of empty files to
an empty ext4 file system (with large_dir enabled) fails after just
over 26 million files. Depending on luck, file creation will give error
EEXIST or EUCLEAN.

Commit b5776e7 fixed the no-restart condition so that
ext4_handle_dirty_dx_node is always called, but it also broke the
restart condition. This is because when restart=1, the original
implementation correctly skipped do_split() but b5776e7 clobbered the
"if(restart)goto journal_error;" logic.

This complementary change protects do_split() from restart condition,
making it safe from both current and future ordering of goto statements
in earlier sections of the code.

Tested on 5.11.20 with handy testing script:

i = 0
while i <= 32000000:
    print (i)
    with open('tmpmnt/%d' % i, 'wb') as fout:
        i += 1

Google-Bug-Id: 176345532
Fixes: b5776e7 ("ext4: fix potential htree index checksum corruption")
Fixes: e08ac99 ("ext4: add largedir feature")
Signed-off-by: denis@voxelsoft.com
Signed-off-by: Colin Ian King <colin.king@canonical.com>
---
 fs/ext4/namei.c | 12 +++++++-----
 1 file changed, 7 insertions(+), 5 deletions(-)

Comments

Dimitri John Ledkov July 6, 2021, 1:58 p.m. UTC | #1 
Hi,

On Tue, Jul 6, 2021 at 2:36 PM Colin King <colin.king@canonical.com> wrote:
>
> From: Colin Ian King <colin.king@canonical.com>
>
> BugLink: https://bugs.launchpad.net/bugs/1933074
>
> [ Note: This was originally sent upstream to linux-ext4@vger.kernel.org
>   and Ted Tso but didn't seem to make it to the list, presumably as it
>   was marked as SPAM
>
>   See: http://www.voxelsoft.com/2021/ext4_large_dir_corruption.html ]
>
> Following commit b5776e7, a trivial sequential write of empty files to
> an empty ext4 file system (with large_dir enabled) fails after just
> over 26 million files. Depending on luck, file creation will give error
> EEXIST or EUCLEAN.
>
> Commit b5776e7 fixed the no-restart condition so that
> ext4_handle_dirty_dx_node is always called, but it also broke the
> restart condition. This is because when restart=1, the original
> implementation correctly skipped do_split() but b5776e7 clobbered the
> "if(restart)goto journal_error;" logic.
>

Interesting how that was missed in reviews. It is not normal to drop
goto without explanation, and that commit didn't explain anything wrt
goto change.

> This complementary change protects do_split() from restart condition,
> making it safe from both current and future ordering of goto statements
> in earlier sections of the code.
>
> Tested on 5.11.20 with handy testing script:
>
> i = 0
> while i <= 32000000:
>     print (i)
>     with open('tmpmnt/%d' % i, 'wb') as fout:
>         i += 1
>
> Google-Bug-Id: 176345532
> Fixes: b5776e7 ("ext4: fix potential htree index checksum corruption")
> Fixes: e08ac99 ("ext4: add largedir feature")
> Signed-off-by: denis@voxelsoft.com
> Signed-off-by: Colin Ian King <colin.king@canonical.com>
> ---
>  fs/ext4/namei.c | 12 +++++++-----
>  1 file changed, 7 insertions(+), 5 deletions(-)
>

Acked-by: Dimitri John Ledkov <dimitri.ledkov@canonical.com>

> diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c
> index 4c37abe..9ad5cc6 100644
> --- a/fs/ext4/namei.c
> +++ b/fs/ext4/namei.c
> @@ -2432,13 +2432,15 @@ static int ext4_dx_add_entry(handle_t *handle, struct ext4_filename *fname,
>                         goto journal_error;

Just above this line there is "restart = 1;" which with this patch
makes "goto journal_error" kind of redundant, since the code between
`restart =1;` and journal_error is disabled by setting restart = 1.
I would have dropped this goto journal_error, although in a way it
documents what should happen next if any new code gets added below
that is run with and without restart.

>                 }
>         }
> -       de = do_split(handle, dir, &bh, frame, &fname->hinfo);
> -       if (IS_ERR(de)) {
> -               err = PTR_ERR(de);
> +       if (!restart) {
> +               de = do_split(handle, dir, &bh, frame, &fname->hinfo);
> +               if (IS_ERR(de)) {
> +                       err = PTR_ERR(de);
> +                       goto cleanup;
> +               }
> +               err = add_dirent_to_buf(handle, fname, dir, inode, de, bh);
>                 goto cleanup;
>         }
> -       err = add_dirent_to_buf(handle, fname, dir, inode, de, bh);
> -       goto cleanup;
>
>  journal_error:
>         ext4_std_error(dir->i_sb, err); /* this is a no-op if err == 0 */
> --
> 2.7.4
Stefan Bader July 7, 2021, 7:34 a.m. UTC | #2 
On 06.07.21 15:36, Colin King wrote:
> From: Colin Ian King <colin.king@canonical.com>
> 
> BugLink: https://bugs.launchpad.net/bugs/1933074
> 
> [ Note: This was originally sent upstream to linux-ext4@vger.kernel.org
>    and Ted Tso but didn't seem to make it to the list, presumably as it
>    was marked as SPAM
> 
>    See: http://www.voxelsoft.com/2021/ext4_large_dir_corruption.html ]
> 
> Following commit b5776e7, a trivial sequential write of empty files to
> an empty ext4 file system (with large_dir enabled) fails after just
> over 26 million files. Depending on luck, file creation will give error
> EEXIST or EUCLEAN.
> 
> Commit b5776e7 fixed the no-restart condition so that
> ext4_handle_dirty_dx_node is always called, but it also broke the
> restart condition. This is because when restart=1, the original
> implementation correctly skipped do_split() but b5776e7 clobbered the
> "if(restart)goto journal_error;" logic.
> 
> This complementary change protects do_split() from restart condition,
> making it safe from both current and future ordering of goto statements
> in earlier sections of the code.
> 
> Tested on 5.11.20 with handy testing script:
> 
> i = 0
> while i <= 32000000:
>      print (i)
>      with open('tmpmnt/%d' % i, 'wb') as fout:
>          i += 1
> 
> Google-Bug-Id: 176345532
> Fixes: b5776e7 ("ext4: fix potential htree index checksum corruption")
> Fixes: e08ac99 ("ext4: add largedir feature")
> Signed-off-by: denis@voxelsoft.com
> Signed-off-by: Colin Ian King <colin.king@canonical.com>
Acked-by: Stefan Bader <stefan.bader@canonical.com>
> ---

As this is not anywhere upstream, yet, we would need to prefix this with 
"UBUNTU: SAUCE" when we apply it. The change does what it explains but more than 
that there is good testing with this.

-Stefan

>   fs/ext4/namei.c | 12 +++++++-----
>   1 file changed, 7 insertions(+), 5 deletions(-)
> 
> diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c
> index 4c37abe..9ad5cc6 100644
> --- a/fs/ext4/namei.c
> +++ b/fs/ext4/namei.c
> @@ -2432,13 +2432,15 @@ static int ext4_dx_add_entry(handle_t *handle, struct ext4_filename *fname,
>   			goto journal_error;
>   		}
>   	}
> -	de = do_split(handle, dir, &bh, frame, &fname->hinfo);
> -	if (IS_ERR(de)) {
> -		err = PTR_ERR(de);
> +	if (!restart) {
> +		de = do_split(handle, dir, &bh, frame, &fname->hinfo);
> +		if (IS_ERR(de)) {
> +			err = PTR_ERR(de);
> +			goto cleanup;
> +		}
> +		err = add_dirent_to_buf(handle, fname, dir, inode, de, bh);
>   		goto cleanup;
>   	}
> -	err = add_dirent_to_buf(handle, fname, dir, inode, de, bh);
> -	goto cleanup;
>   
>   journal_error:
>   	ext4_std_error(dir->i_sb, err); /* this is a no-op if err == 0 */
>
diff mbox series

Patch

diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c
index 4c37abe..9ad5cc6 100644
--- a/fs/ext4/namei.c
+++ b/fs/ext4/namei.c
@@ -2432,13 +2432,15 @@  static int ext4_dx_add_entry(handle_t *handle, struct ext4_filename *fname,
 			goto journal_error;
 		}
 	}
-	de = do_split(handle, dir, &bh, frame, &fname->hinfo);
-	if (IS_ERR(de)) {
-		err = PTR_ERR(de);
+	if (!restart) {
+		de = do_split(handle, dir, &bh, frame, &fname->hinfo);
+		if (IS_ERR(de)) {
+			err = PTR_ERR(de);
+			goto cleanup;
+		}
+		err = add_dirent_to_buf(handle, fname, dir, inode, de, bh);
 		goto cleanup;
 	}
-	err = add_dirent_to_buf(handle, fname, dir, inode, de, bh);
-	goto cleanup;
 
 journal_error:
 	ext4_std_error(dir->i_sb, err); /* this is a no-op if err == 0 */