2021-04-08 11:29:05

by Zhang Yi

[permalink] [raw]
Subject: [PATCH 0/3] ext4: fix two issue about bdev_try_to_free_page()

Hi,

This first patch fix a potential filesystem inconsistency problem and
other two fix a use after free problem.

Zhang Yi (3):
jbd2: protect buffers release with j_checkpoint_mutex
jbd2: do not free buffers in jbd2_journal_try_to_free_buffers()
ext4: add rcu to prevent use after free when umount filesystem

fs/ext4/inode.c | 6 ++++--
fs/ext4/super.c | 41 +++++++++++++++++++++++++++++------------
fs/jbd2/journal.c | 30 +++++++++++++++++++++++++++---
fs/jbd2/transaction.c | 20 ++++++++++----------
include/linux/jbd2.h | 11 ++++++++++-
5 files changed, 80 insertions(+), 28 deletions(-)

--
2.25.4


2021-04-08 11:30:22

by Zhang Yi

[permalink] [raw]
Subject: [PATCH 3/3] ext4: add rcu to prevent use after free when umount filesystem

There is a race between bdev_try_to_free_page() and
jbd2_journal_destroy() that could end up triggering a use after free
issue about journal.

drop cache umount filesystem
bdev_try_to_free_page()
get journal
jbd2_journal_try_to_free_buffers() ext4_put_super()
kfree(journal)
access journal <-- lead to UAF

The above race also could happens between the bdev_try_to_free_page()
and the error path of ext4_fill_super(). This patch avoid this race by
add rcu protection around accessing sbi->s_journal in
bdev_try_to_free_page() and destroy the journal after an rcu grace
period.

Signed-off-by: Zhang Yi <[email protected]>
---
fs/ext4/super.c | 33 ++++++++++++++++++++++++---------
fs/jbd2/journal.c | 30 +++++++++++++++++++++++++++---
include/linux/jbd2.h | 11 ++++++++++-
3 files changed, 61 insertions(+), 13 deletions(-)

diff --git a/fs/ext4/super.c b/fs/ext4/super.c
index 02ba47a5bc70..6bbaadc5357b 100644
--- a/fs/ext4/super.c
+++ b/fs/ext4/super.c
@@ -1150,6 +1150,21 @@ static inline void ext4_quota_off_umount(struct super_block *sb)
}
#endif

+static int ext4_journal_release(struct ext4_sb_info *sbi)
+{
+ journal_t *journal = sbi->s_journal;
+ int ret;
+
+ ret = jbd2_journal_release(journal);
+ sbi->s_journal = NULL;
+ /*
+ * Call rcu to prevent racing with bdev_try_to_free_page()
+ * accessing the journal at the same time.
+ */
+ call_rcu(&journal->j_rcu, jbd2_journal_release_rcu);
+ return ret;
+}
+
static void ext4_put_super(struct super_block *sb)
{
struct ext4_sb_info *sbi = EXT4_SB(sb);
@@ -1174,11 +1189,9 @@ static void ext4_put_super(struct super_block *sb)

if (sbi->s_journal) {
aborted = is_journal_aborted(sbi->s_journal);
- err = jbd2_journal_destroy(sbi->s_journal);
- sbi->s_journal = NULL;
- if ((err < 0) && !aborted) {
+ err = ext4_journal_release(sbi);
+ if ((err < 0) && !aborted)
ext4_abort(sb, -err, "Couldn't clean up the journal");
- }
}

ext4_es_unregister_shrinker(sbi);
@@ -1449,14 +1462,18 @@ static int ext4_nfs_commit_metadata(struct inode *inode)
static int bdev_try_to_free_page(struct super_block *sb, struct page *page,
gfp_t wait)
{
- journal_t *journal = EXT4_SB(sb)->s_journal;
+ journal_t *journal;
int ret = 0;

WARN_ON(PageChecked(page));
if (!page_has_buffers(page))
return 0;
+
+ rcu_read_lock();
+ journal = READ_ONCE(EXT4_SB(sb)->s_journal);
if (journal)
ret = jbd2_journal_try_to_free_buffers(journal, page);
+ rcu_read_unlock();
if (!ret)
return try_to_free_buffers(page);
return 0;
@@ -5146,10 +5163,8 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
ext4_xattr_destroy_cache(sbi->s_ea_block_cache);
sbi->s_ea_block_cache = NULL;

- if (sbi->s_journal) {
- jbd2_journal_destroy(sbi->s_journal);
- sbi->s_journal = NULL;
- }
+ if (sbi->s_journal)
+ ext4_journal_release(sbi);
failed_mount3a:
ext4_es_unregister_shrinker(sbi);
failed_mount3:
diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c
index 2dc944442802..071caaaa9de1 100644
--- a/fs/jbd2/journal.c
+++ b/fs/jbd2/journal.c
@@ -76,6 +76,8 @@ EXPORT_SYMBOL(jbd2_journal_check_available_features);
EXPORT_SYMBOL(jbd2_journal_set_features);
EXPORT_SYMBOL(jbd2_journal_load);
EXPORT_SYMBOL(jbd2_journal_destroy);
+EXPORT_SYMBOL(jbd2_journal_release);
+EXPORT_SYMBOL(jbd2_journal_release_rcu);
EXPORT_SYMBOL(jbd2_journal_abort);
EXPORT_SYMBOL(jbd2_journal_errno);
EXPORT_SYMBOL(jbd2_journal_ack_err);
@@ -1951,14 +1953,14 @@ int jbd2_journal_load(journal_t *journal)
}

/**
- * jbd2_journal_destroy() - Release a journal_t structure.
+ * jbd2_journal_release() - Release a journal_t structure.
* @journal: Journal to act on.
*
* Release a journal_t structure once it is no longer in use by the
* journaled object.
* Return <0 if we couldn't clean up the journal.
*/
-int jbd2_journal_destroy(journal_t *journal)
+int jbd2_journal_release(journal_t *journal)
{
int err = 0;

@@ -2021,11 +2023,33 @@ int jbd2_journal_destroy(journal_t *journal)
crypto_free_shash(journal->j_chksum_driver);
kfree(journal->j_fc_wbuf);
kfree(journal->j_wbuf);
- kfree(journal);

return err;
}

+/**
+ * jbd2_journal_release_rcu() - Free a journal_t structure.
+ * @rcu: rcu list node relate to the journal want to free.
+ *
+ * Freeing a journal_t structure after a rcu grace period.
+ */
+void jbd2_journal_release_rcu(struct rcu_head *rcu)
+{
+ kfree(container_of(rcu, journal_t, j_rcu));
+}
+
+/**
+ * jbd2_journal_destroy() - Release and free a journal_t structure.
+ * @journal: Journal to act on.
+ *
+ * Release and free a journal_t structure once it is no longer in use
+ * by the journaled object.
+ */
+void jbd2_journal_destroy(journal_t *journal)
+{
+ jbd2_journal_release(journal);
+ kfree(journal);
+}

/**
* jbd2_journal_check_used_features() - Check if features specified are used.
diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h
index 99d3cd051ac3..39a8d04596a2 100644
--- a/include/linux/jbd2.h
+++ b/include/linux/jbd2.h
@@ -1238,6 +1238,13 @@ struct journal_s
*/
__u32 j_csum_seed;

+ /**
+ * @j_rcu:
+ *
+ * Prevent racing between accessing and destroy at the same time.
+ */
+ struct rcu_head j_rcu;
+
#ifdef CONFIG_DEBUG_LOCK_ALLOC
/**
* @j_trans_commit_map:
@@ -1509,7 +1516,9 @@ extern int jbd2_journal_set_features
extern void jbd2_journal_clear_features
(journal_t *, unsigned long, unsigned long, unsigned long);
extern int jbd2_journal_load (journal_t *journal);
-extern int jbd2_journal_destroy (journal_t *);
+extern void jbd2_journal_destroy (journal_t *);
+extern int jbd2_journal_release (journal_t *);
+extern void jbd2_journal_release_rcu (struct rcu_head *rcu);
extern int jbd2_journal_recover (journal_t *journal);
extern int jbd2_journal_wipe (journal_t *, int);
extern int jbd2_journal_skip_recovery (journal_t *);
--
2.25.4

2021-04-08 11:30:35

by Zhang Yi

[permalink] [raw]
Subject: [PATCH 1/3] jbd2: protect buffers release with j_checkpoint_mutex

There is a race between jbd2_journal_try_to_free_buffers() and
jbd2_journal_destroy(), so the jbd2_log_do_checkpoint() may still
missing to detect the buffer write io error flag and lead to filesystem
inconsistency.

jbd2_journal_try_to_free_buffers() ext4_put_super()
jbd2_journal_destroy()
__jbd2_journal_remove_checkpoint()
detect buffer write error jbd2_log_do_checkpoint()
jbd2_cleanup_journal_tail()
<--- lead to inconsistency
jbd2_journal_abort()

Fix this issue by add j_checkpoint_mutex to protect journal buffer
release on jbd2_journal_try_to_free_buffers().

Signed-off-by: Zhang Yi <[email protected]>
---
fs/jbd2/transaction.c | 2 ++
1 file changed, 2 insertions(+)

diff --git a/fs/jbd2/transaction.c b/fs/jbd2/transaction.c
index 9396666b7314..b935b20cbae4 100644
--- a/fs/jbd2/transaction.c
+++ b/fs/jbd2/transaction.c
@@ -2123,6 +2123,7 @@ int jbd2_journal_try_to_free_buffers(journal_t *journal, struct page *page)

J_ASSERT(PageLocked(page));

+ mutex_lock(&journal->j_checkpoint_mutex);
head = page_buffers(page);
bh = head;
do {
@@ -2163,6 +2164,7 @@ int jbd2_journal_try_to_free_buffers(journal_t *journal, struct page *page)
if (has_write_io_error)
jbd2_journal_abort(journal, -EIO);

+ mutex_unlock(&journal->j_checkpoint_mutex);
return ret;
}

--
2.25.4

2021-04-08 13:46:34

by Jan Kara

[permalink] [raw]
Subject: Re: [PATCH 1/3] jbd2: protect buffers release with j_checkpoint_mutex

On Thu 08-04-21 19:36:16, Zhang Yi wrote:
> There is a race between jbd2_journal_try_to_free_buffers() and
> jbd2_journal_destroy(), so the jbd2_log_do_checkpoint() may still
> missing to detect the buffer write io error flag and lead to filesystem
> inconsistency.
>
> jbd2_journal_try_to_free_buffers() ext4_put_super()
> jbd2_journal_destroy()
> __jbd2_journal_remove_checkpoint()
> detect buffer write error jbd2_log_do_checkpoint()
> jbd2_cleanup_journal_tail()
> <--- lead to inconsistency
> jbd2_journal_abort()
>
> Fix this issue by add j_checkpoint_mutex to protect journal buffer
> release on jbd2_journal_try_to_free_buffers().
>
> Signed-off-by: Zhang Yi <[email protected]>

Thanks for the patch Zhang. I agree with your problem analysis but I don't
think the solution is correct:

> J_ASSERT(PageLocked(page));
>
> + mutex_lock(&journal->j_checkpoint_mutex);

We cannot grab j_checkpoint_mutex inside jbd2_journal_try_to_free_buffers()
(or even ext4_releasepage()) because that function is called withe a page
lock which ranks below the checkpoint mutex - generally page locks are
acquired within a transaction and thus all locks required to start a
transaction (and j_checkpoint_mutex is one of them) rank above the page
lock.

Also even if the lock ordering was OK, grabbing j_checkpoint_mutex for
every page from memory reclaim just to close this rare race seems like a
performance overkill.

What we seem to need is a quick way of marking the journal as "IO error
occured" in __journal_try_to_free_buffer() before actually removing the
buffer from the checkpoint list. Perhaps this marking could even happen
already in __jbd2_journal_remove_checkpoint() and we can reuse it in
jbd2_log_do_checkpoint() for IO error handling as well... And then once we
are in a safer context, we can do:

if (!is_journal_aborted(journal) && journal_io_error_happened(journal))
jbd2_journal_abort(...)

Honza

> head = page_buffers(page);
> bh = head;
> do {
> @@ -2163,6 +2164,7 @@ int jbd2_journal_try_to_free_buffers(journal_t *journal, struct page *page)
> if (has_write_io_error)
> jbd2_journal_abort(journal, -EIO);
>
> + mutex_unlock(&journal->j_checkpoint_mutex);
> return ret;
> }
>
> --
> 2.25.4
>
--
Jan Kara <[email protected]>
SUSE Labs, CR

2021-04-08 13:52:45

by kernel test robot

[permalink] [raw]
Subject: Re: [PATCH 3/3] ext4: add rcu to prevent use after free when umount filesystem

Hi Zhang,

Thank you for the patch! Yet something to improve:

[auto build test ERROR on ext4/dev]
[also build test ERROR on linus/master v5.12-rc6 next-20210408]
[cannot apply to tytso-fscrypt/master]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]

url: https://github.com/0day-ci/linux/commits/Zhang-Yi/ext4-fix-two-issue-about-bdev_try_to_free_page/20210408-193105
base: https://git.kernel.org/pub/scm/linux/kernel/git/tytso/ext4.git dev
config: alpha-randconfig-m031-20210408 (attached as .config)
compiler: alpha-linux-gcc (GCC) 9.3.0
reproduce (this is a W=1 build):
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
chmod +x ~/bin/make.cross
# https://github.com/0day-ci/linux/commit/1cd15af09e1887501090c23700b696d43ebf39ca
git remote add linux-review https://github.com/0day-ci/linux
git fetch --no-tags linux-review Zhang-Yi/ext4-fix-two-issue-about-bdev_try_to_free_page/20210408-193105
git checkout 1cd15af09e1887501090c23700b696d43ebf39ca
# save the attached .config to linux build tree
COMPILER_INSTALL_PATH=$HOME/0day COMPILER=gcc-9.3.0 make.cross ARCH=alpha

If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <[email protected]>

All errors (new ones prefixed by >>):

fs/ocfs2/journal.c: In function 'ocfs2_journal_shutdown':
>> fs/ocfs2/journal.c:1012:6: error: invalid use of void expression
1012 | if (!jbd2_journal_destroy(journal->j_journal) && !status) {
| ^


vim +1012 fs/ocfs2/journal.c

ccd979bdbce9fb Mark Fasheh 2005-12-15 955
ccd979bdbce9fb Mark Fasheh 2005-12-15 956 /*
ccd979bdbce9fb Mark Fasheh 2005-12-15 957 * If the journal has been kmalloc'd it needs to be freed after this
ccd979bdbce9fb Mark Fasheh 2005-12-15 958 * call.
ccd979bdbce9fb Mark Fasheh 2005-12-15 959 */
ccd979bdbce9fb Mark Fasheh 2005-12-15 960 void ocfs2_journal_shutdown(struct ocfs2_super *osb)
ccd979bdbce9fb Mark Fasheh 2005-12-15 961 {
ccd979bdbce9fb Mark Fasheh 2005-12-15 962 struct ocfs2_journal *journal = NULL;
ccd979bdbce9fb Mark Fasheh 2005-12-15 963 int status = 0;
ccd979bdbce9fb Mark Fasheh 2005-12-15 964 struct inode *inode = NULL;
ccd979bdbce9fb Mark Fasheh 2005-12-15 965 int num_running_trans = 0;
ccd979bdbce9fb Mark Fasheh 2005-12-15 966
ebdec83ba46c12 Eric Sesterhenn / snakebyte 2006-01-27 967 BUG_ON(!osb);
ccd979bdbce9fb Mark Fasheh 2005-12-15 968
ccd979bdbce9fb Mark Fasheh 2005-12-15 969 journal = osb->journal;
ccd979bdbce9fb Mark Fasheh 2005-12-15 970 if (!journal)
ccd979bdbce9fb Mark Fasheh 2005-12-15 971 goto done;
ccd979bdbce9fb Mark Fasheh 2005-12-15 972
ccd979bdbce9fb Mark Fasheh 2005-12-15 973 inode = journal->j_inode;
ccd979bdbce9fb Mark Fasheh 2005-12-15 974
ccd979bdbce9fb Mark Fasheh 2005-12-15 975 if (journal->j_state != OCFS2_JOURNAL_LOADED)
ccd979bdbce9fb Mark Fasheh 2005-12-15 976 goto done;
ccd979bdbce9fb Mark Fasheh 2005-12-15 977
2b4e30fbde4258 Joel Becker 2008-09-03 978 /* need to inc inode use count - jbd2_journal_destroy will iput. */
ccd979bdbce9fb Mark Fasheh 2005-12-15 979 if (!igrab(inode))
ccd979bdbce9fb Mark Fasheh 2005-12-15 980 BUG();
ccd979bdbce9fb Mark Fasheh 2005-12-15 981
ccd979bdbce9fb Mark Fasheh 2005-12-15 982 num_running_trans = atomic_read(&(osb->journal->j_num_trans));
b41079504c786e Tao Ma 2011-02-24 983 trace_ocfs2_journal_shutdown(num_running_trans);
ccd979bdbce9fb Mark Fasheh 2005-12-15 984
ccd979bdbce9fb Mark Fasheh 2005-12-15 985 /* Do a commit_cache here. It will flush our journal, *and*
ccd979bdbce9fb Mark Fasheh 2005-12-15 986 * release any locks that are still held.
ccd979bdbce9fb Mark Fasheh 2005-12-15 987 * set the SHUTDOWN flag and release the trans lock.
ccd979bdbce9fb Mark Fasheh 2005-12-15 988 * the commit thread will take the trans lock for us below. */
ccd979bdbce9fb Mark Fasheh 2005-12-15 989 journal->j_state = OCFS2_JOURNAL_IN_SHUTDOWN;
ccd979bdbce9fb Mark Fasheh 2005-12-15 990
ccd979bdbce9fb Mark Fasheh 2005-12-15 991 /* The OCFS2_JOURNAL_IN_SHUTDOWN will signal to commit_cache to not
ccd979bdbce9fb Mark Fasheh 2005-12-15 992 * drop the trans_lock (which we want to hold until we
ccd979bdbce9fb Mark Fasheh 2005-12-15 993 * completely destroy the journal. */
ccd979bdbce9fb Mark Fasheh 2005-12-15 994 if (osb->commit_task) {
ccd979bdbce9fb Mark Fasheh 2005-12-15 995 /* Wait for the commit thread */
b41079504c786e Tao Ma 2011-02-24 996 trace_ocfs2_journal_shutdown_wait(osb->commit_task);
ccd979bdbce9fb Mark Fasheh 2005-12-15 997 kthread_stop(osb->commit_task);
ccd979bdbce9fb Mark Fasheh 2005-12-15 998 osb->commit_task = NULL;
ccd979bdbce9fb Mark Fasheh 2005-12-15 999 }
ccd979bdbce9fb Mark Fasheh 2005-12-15 1000
ccd979bdbce9fb Mark Fasheh 2005-12-15 1001 BUG_ON(atomic_read(&(osb->journal->j_num_trans)) != 0);
ccd979bdbce9fb Mark Fasheh 2005-12-15 1002
c271c5c22b0a7c Sunil Mushran 2006-12-05 1003 if (ocfs2_mount_local(osb)) {
2b4e30fbde4258 Joel Becker 2008-09-03 1004 jbd2_journal_lock_updates(journal->j_journal);
2b4e30fbde4258 Joel Becker 2008-09-03 1005 status = jbd2_journal_flush(journal->j_journal);
2b4e30fbde4258 Joel Becker 2008-09-03 1006 jbd2_journal_unlock_updates(journal->j_journal);
c271c5c22b0a7c Sunil Mushran 2006-12-05 1007 if (status < 0)
c271c5c22b0a7c Sunil Mushran 2006-12-05 1008 mlog_errno(status);
c271c5c22b0a7c Sunil Mushran 2006-12-05 1009 }
c271c5c22b0a7c Sunil Mushran 2006-12-05 1010
d85400af790dba Junxiao Bi 2018-12-28 1011 /* Shutdown the kernel journal system */
d85400af790dba Junxiao Bi 2018-12-28 @1012 if (!jbd2_journal_destroy(journal->j_journal) && !status) {
c271c5c22b0a7c Sunil Mushran 2006-12-05 1013 /*
c271c5c22b0a7c Sunil Mushran 2006-12-05 1014 * Do not toggle if flush was unsuccessful otherwise
c271c5c22b0a7c Sunil Mushran 2006-12-05 1015 * will leave dirty metadata in a "clean" journal
c271c5c22b0a7c Sunil Mushran 2006-12-05 1016 */
539d8264093560 Sunil Mushran 2008-07-14 1017 status = ocfs2_journal_toggle_dirty(osb, 0, 0);
ccd979bdbce9fb Mark Fasheh 2005-12-15 1018 if (status < 0)
ccd979bdbce9fb Mark Fasheh 2005-12-15 1019 mlog_errno(status);
c271c5c22b0a7c Sunil Mushran 2006-12-05 1020 }
ae0dff683076b2 Sunil Mushran 2008-10-22 1021 journal->j_journal = NULL;
ccd979bdbce9fb Mark Fasheh 2005-12-15 1022
ccd979bdbce9fb Mark Fasheh 2005-12-15 1023 OCFS2_I(inode)->ip_open_count--;
ccd979bdbce9fb Mark Fasheh 2005-12-15 1024
ccd979bdbce9fb Mark Fasheh 2005-12-15 1025 /* unlock our journal */
e63aecb651ba73 Mark Fasheh 2007-10-18 1026 ocfs2_inode_unlock(inode, 1);
ccd979bdbce9fb Mark Fasheh 2005-12-15 1027
ccd979bdbce9fb Mark Fasheh 2005-12-15 1028 brelse(journal->j_bh);
ccd979bdbce9fb Mark Fasheh 2005-12-15 1029 journal->j_bh = NULL;
ccd979bdbce9fb Mark Fasheh 2005-12-15 1030
ccd979bdbce9fb Mark Fasheh 2005-12-15 1031 journal->j_state = OCFS2_JOURNAL_FREE;
ccd979bdbce9fb Mark Fasheh 2005-12-15 1032

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/[email protected]


Attachments:
(No filename) (8.57 kB)
.config.gz (26.80 kB)
Download all attachments

2021-04-08 13:56:47

by Jan Kara

[permalink] [raw]
Subject: Re: [PATCH 3/3] ext4: add rcu to prevent use after free when umount filesystem

On Thu 08-04-21 19:36:18, Zhang Yi wrote:
> There is a race between bdev_try_to_free_page() and
> jbd2_journal_destroy() that could end up triggering a use after free
> issue about journal.
>
> drop cache umount filesystem
> bdev_try_to_free_page()
> get journal
> jbd2_journal_try_to_free_buffers() ext4_put_super()
> kfree(journal)
> access journal <-- lead to UAF
>
> The above race also could happens between the bdev_try_to_free_page()
> and the error path of ext4_fill_super(). This patch avoid this race by
> add rcu protection around accessing sbi->s_journal in
> bdev_try_to_free_page() and destroy the journal after an rcu grace
> period.
>
> Signed-off-by: Zhang Yi <[email protected]>

OK, I see the problem. But cannot the use-after-free happen even for the
superblock itself (e.g., EXT4_SB(sb)->s_journal dereference)? I don't see
anything preventing that as blkdev_releasepage() just shamelessly does:

super = BDEV_I(page->mapping->host)->bdev.bd_super

without making sure the sb cannot go away the instant we load a pointer to
it. Or am I missing something Ted? If I'm right, we'd need some careful
sprinkling of RCU, READ_ONCE(), and careful superblock reference grabbing
to make bdev_try_to_free_page() safe against concurrent
kill_block_super()...

Honza

> ---
> fs/ext4/super.c | 33 ++++++++++++++++++++++++---------
> fs/jbd2/journal.c | 30 +++++++++++++++++++++++++++---
> include/linux/jbd2.h | 11 ++++++++++-
> 3 files changed, 61 insertions(+), 13 deletions(-)
>
> diff --git a/fs/ext4/super.c b/fs/ext4/super.c
> index 02ba47a5bc70..6bbaadc5357b 100644
> --- a/fs/ext4/super.c
> +++ b/fs/ext4/super.c
> @@ -1150,6 +1150,21 @@ static inline void ext4_quota_off_umount(struct super_block *sb)
> }
> #endif
>
> +static int ext4_journal_release(struct ext4_sb_info *sbi)
> +{
> + journal_t *journal = sbi->s_journal;
> + int ret;
> +
> + ret = jbd2_journal_release(journal);
> + sbi->s_journal = NULL;
> + /*
> + * Call rcu to prevent racing with bdev_try_to_free_page()
> + * accessing the journal at the same time.
> + */
> + call_rcu(&journal->j_rcu, jbd2_journal_release_rcu);
> + return ret;
> +}
> +
> static void ext4_put_super(struct super_block *sb)
> {
> struct ext4_sb_info *sbi = EXT4_SB(sb);
> @@ -1174,11 +1189,9 @@ static void ext4_put_super(struct super_block *sb)
>
> if (sbi->s_journal) {
> aborted = is_journal_aborted(sbi->s_journal);
> - err = jbd2_journal_destroy(sbi->s_journal);
> - sbi->s_journal = NULL;
> - if ((err < 0) && !aborted) {
> + err = ext4_journal_release(sbi);
> + if ((err < 0) && !aborted)
> ext4_abort(sb, -err, "Couldn't clean up the journal");
> - }
> }
>
> ext4_es_unregister_shrinker(sbi);
> @@ -1449,14 +1462,18 @@ static int ext4_nfs_commit_metadata(struct inode *inode)
> static int bdev_try_to_free_page(struct super_block *sb, struct page *page,
> gfp_t wait)
> {
> - journal_t *journal = EXT4_SB(sb)->s_journal;
> + journal_t *journal;
> int ret = 0;
>
> WARN_ON(PageChecked(page));
> if (!page_has_buffers(page))
> return 0;
> +
> + rcu_read_lock();
> + journal = READ_ONCE(EXT4_SB(sb)->s_journal);
> if (journal)
> ret = jbd2_journal_try_to_free_buffers(journal, page);
> + rcu_read_unlock();
> if (!ret)
> return try_to_free_buffers(page);
> return 0;
> @@ -5146,10 +5163,8 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
> ext4_xattr_destroy_cache(sbi->s_ea_block_cache);
> sbi->s_ea_block_cache = NULL;
>
> - if (sbi->s_journal) {
> - jbd2_journal_destroy(sbi->s_journal);
> - sbi->s_journal = NULL;
> - }
> + if (sbi->s_journal)
> + ext4_journal_release(sbi);
> failed_mount3a:
> ext4_es_unregister_shrinker(sbi);
> failed_mount3:
> diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c
> index 2dc944442802..071caaaa9de1 100644
> --- a/fs/jbd2/journal.c
> +++ b/fs/jbd2/journal.c
> @@ -76,6 +76,8 @@ EXPORT_SYMBOL(jbd2_journal_check_available_features);
> EXPORT_SYMBOL(jbd2_journal_set_features);
> EXPORT_SYMBOL(jbd2_journal_load);
> EXPORT_SYMBOL(jbd2_journal_destroy);
> +EXPORT_SYMBOL(jbd2_journal_release);
> +EXPORT_SYMBOL(jbd2_journal_release_rcu);
> EXPORT_SYMBOL(jbd2_journal_abort);
> EXPORT_SYMBOL(jbd2_journal_errno);
> EXPORT_SYMBOL(jbd2_journal_ack_err);
> @@ -1951,14 +1953,14 @@ int jbd2_journal_load(journal_t *journal)
> }
>
> /**
> - * jbd2_journal_destroy() - Release a journal_t structure.
> + * jbd2_journal_release() - Release a journal_t structure.
> * @journal: Journal to act on.
> *
> * Release a journal_t structure once it is no longer in use by the
> * journaled object.
> * Return <0 if we couldn't clean up the journal.
> */
> -int jbd2_journal_destroy(journal_t *journal)
> +int jbd2_journal_release(journal_t *journal)
> {
> int err = 0;
>
> @@ -2021,11 +2023,33 @@ int jbd2_journal_destroy(journal_t *journal)
> crypto_free_shash(journal->j_chksum_driver);
> kfree(journal->j_fc_wbuf);
> kfree(journal->j_wbuf);
> - kfree(journal);
>
> return err;
> }
>
> +/**
> + * jbd2_journal_release_rcu() - Free a journal_t structure.
> + * @rcu: rcu list node relate to the journal want to free.
> + *
> + * Freeing a journal_t structure after a rcu grace period.
> + */
> +void jbd2_journal_release_rcu(struct rcu_head *rcu)
> +{
> + kfree(container_of(rcu, journal_t, j_rcu));
> +}
> +
> +/**
> + * jbd2_journal_destroy() - Release and free a journal_t structure.
> + * @journal: Journal to act on.
> + *
> + * Release and free a journal_t structure once it is no longer in use
> + * by the journaled object.
> + */
> +void jbd2_journal_destroy(journal_t *journal)
> +{
> + jbd2_journal_release(journal);
> + kfree(journal);
> +}
>
> /**
> * jbd2_journal_check_used_features() - Check if features specified are used.
> diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h
> index 99d3cd051ac3..39a8d04596a2 100644
> --- a/include/linux/jbd2.h
> +++ b/include/linux/jbd2.h
> @@ -1238,6 +1238,13 @@ struct journal_s
> */
> __u32 j_csum_seed;
>
> + /**
> + * @j_rcu:
> + *
> + * Prevent racing between accessing and destroy at the same time.
> + */
> + struct rcu_head j_rcu;
> +
> #ifdef CONFIG_DEBUG_LOCK_ALLOC
> /**
> * @j_trans_commit_map:
> @@ -1509,7 +1516,9 @@ extern int jbd2_journal_set_features
> extern void jbd2_journal_clear_features
> (journal_t *, unsigned long, unsigned long, unsigned long);
> extern int jbd2_journal_load (journal_t *journal);
> -extern int jbd2_journal_destroy (journal_t *);
> +extern void jbd2_journal_destroy (journal_t *);
> +extern int jbd2_journal_release (journal_t *);
> +extern void jbd2_journal_release_rcu (struct rcu_head *rcu);
> extern int jbd2_journal_recover (journal_t *journal);
> extern int jbd2_journal_wipe (journal_t *, int);
> extern int jbd2_journal_skip_recovery (journal_t *);
> --
> 2.25.4
>
--
Jan Kara <[email protected]>
SUSE Labs, CR

2021-04-08 14:39:07

by Zhang Yi

[permalink] [raw]
Subject: Re: [PATCH 3/3] ext4: add rcu to prevent use after free when umount filesystem

On 2021/4/8 21:56, Jan Kara wrote:
> On Thu 08-04-21 19:36:18, Zhang Yi wrote:
>> There is a race between bdev_try_to_free_page() and
>> jbd2_journal_destroy() that could end up triggering a use after free
>> issue about journal.
>>
>> drop cache umount filesystem
>> bdev_try_to_free_page()
>> get journal
>> jbd2_journal_try_to_free_buffers() ext4_put_super()
>> kfree(journal)
>> access journal <-- lead to UAF
>>
>> The above race also could happens between the bdev_try_to_free_page()
>> and the error path of ext4_fill_super(). This patch avoid this race by
>> add rcu protection around accessing sbi->s_journal in
>> bdev_try_to_free_page() and destroy the journal after an rcu grace
>> period.
>>
>> Signed-off-by: Zhang Yi <[email protected]>
>
> OK, I see the problem. But cannot the use-after-free happen even for the
> superblock itself (e.g., EXT4_SB(sb)->s_journal dereference)? I don't see
> anything preventing that as blkdev_releasepage() just shamelessly does:
>
> super = BDEV_I(page->mapping->host)->bdev.bd_super
>
Hi, Jan.

I checked the superblock. In theory, the bdev_try_to_free_page() is invoked
with page locked, the umount process will wait the page unlock on
kill_block_super()->..->kill_bdev()->truncate_inode_pages_range() before free
superblock, so I guess the use-after-free problem couldn't happen in general.
But I think it's fragile and may invalidate if the bdev has more than one
operners(__blkdev_put() call kill_bdev only if bd_openers becomes zero)?
I will check it.

Thanks,
Yi.

> without making sure the sb cannot go away the instant we load a pointer to
> it. Or am I missing something Ted? If I'm right, we'd need some careful
> sprinkling of RCU, READ_ONCE(), and careful superblock reference grabbing
> to make bdev_try_to_free_page() safe against concurrent
> kill_block_super()...
>
> Honza
>
>> ---
>> fs/ext4/super.c | 33 ++++++++++++++++++++++++---------
>> fs/jbd2/journal.c | 30 +++++++++++++++++++++++++++---
>> include/linux/jbd2.h | 11 ++++++++++-
>> 3 files changed, 61 insertions(+), 13 deletions(-)
>>
>> diff --git a/fs/ext4/super.c b/fs/ext4/super.c
>> index 02ba47a5bc70..6bbaadc5357b 100644
>> --- a/fs/ext4/super.c
>> +++ b/fs/ext4/super.c
>> @@ -1150,6 +1150,21 @@ static inline void ext4_quota_off_umount(struct super_block *sb)
>> }
>> #endif
>>
>> +static int ext4_journal_release(struct ext4_sb_info *sbi)
>> +{
>> + journal_t *journal = sbi->s_journal;
>> + int ret;
>> +
>> + ret = jbd2_journal_release(journal);
>> + sbi->s_journal = NULL;
>> + /*
>> + * Call rcu to prevent racing with bdev_try_to_free_page()
>> + * accessing the journal at the same time.
>> + */
>> + call_rcu(&journal->j_rcu, jbd2_journal_release_rcu);
>> + return ret;
>> +}
>> +
>> static void ext4_put_super(struct super_block *sb)
>> {
>> struct ext4_sb_info *sbi = EXT4_SB(sb);
>> @@ -1174,11 +1189,9 @@ static void ext4_put_super(struct super_block *sb)
>>
>> if (sbi->s_journal) {
>> aborted = is_journal_aborted(sbi->s_journal);
>> - err = jbd2_journal_destroy(sbi->s_journal);
>> - sbi->s_journal = NULL;
>> - if ((err < 0) && !aborted) {
>> + err = ext4_journal_release(sbi);
>> + if ((err < 0) && !aborted)
>> ext4_abort(sb, -err, "Couldn't clean up the journal");
>> - }
>> }
>>
>> ext4_es_unregister_shrinker(sbi);
>> @@ -1449,14 +1462,18 @@ static int ext4_nfs_commit_metadata(struct inode *inode)
>> static int bdev_try_to_free_page(struct super_block *sb, struct page *page,
>> gfp_t wait)
>> {
>> - journal_t *journal = EXT4_SB(sb)->s_journal;
>> + journal_t *journal;
>> int ret = 0;
>>
>> WARN_ON(PageChecked(page));
>> if (!page_has_buffers(page))
>> return 0;
>> +
>> + rcu_read_lock();
>> + journal = READ_ONCE(EXT4_SB(sb)->s_journal);
>> if (journal)
>> ret = jbd2_journal_try_to_free_buffers(journal, page);
>> + rcu_read_unlock();
>> if (!ret)
>> return try_to_free_buffers(page);
>> return 0;
>> @@ -5146,10 +5163,8 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent)
>> ext4_xattr_destroy_cache(sbi->s_ea_block_cache);
>> sbi->s_ea_block_cache = NULL;
>>
>> - if (sbi->s_journal) {
>> - jbd2_journal_destroy(sbi->s_journal);
>> - sbi->s_journal = NULL;
>> - }
>> + if (sbi->s_journal)
>> + ext4_journal_release(sbi);
>> failed_mount3a:
>> ext4_es_unregister_shrinker(sbi);
>> failed_mount3:
>> diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c
>> index 2dc944442802..071caaaa9de1 100644
>> --- a/fs/jbd2/journal.c
>> +++ b/fs/jbd2/journal.c
>> @@ -76,6 +76,8 @@ EXPORT_SYMBOL(jbd2_journal_check_available_features);
>> EXPORT_SYMBOL(jbd2_journal_set_features);
>> EXPORT_SYMBOL(jbd2_journal_load);
>> EXPORT_SYMBOL(jbd2_journal_destroy);
>> +EXPORT_SYMBOL(jbd2_journal_release);
>> +EXPORT_SYMBOL(jbd2_journal_release_rcu);
>> EXPORT_SYMBOL(jbd2_journal_abort);
>> EXPORT_SYMBOL(jbd2_journal_errno);
>> EXPORT_SYMBOL(jbd2_journal_ack_err);
>> @@ -1951,14 +1953,14 @@ int jbd2_journal_load(journal_t *journal)
>> }
>>
>> /**
>> - * jbd2_journal_destroy() - Release a journal_t structure.
>> + * jbd2_journal_release() - Release a journal_t structure.
>> * @journal: Journal to act on.
>> *
>> * Release a journal_t structure once it is no longer in use by the
>> * journaled object.
>> * Return <0 if we couldn't clean up the journal.
>> */
>> -int jbd2_journal_destroy(journal_t *journal)
>> +int jbd2_journal_release(journal_t *journal)
>> {
>> int err = 0;
>>
>> @@ -2021,11 +2023,33 @@ int jbd2_journal_destroy(journal_t *journal)
>> crypto_free_shash(journal->j_chksum_driver);
>> kfree(journal->j_fc_wbuf);
>> kfree(journal->j_wbuf);
>> - kfree(journal);
>>
>> return err;
>> }
>>
>> +/**
>> + * jbd2_journal_release_rcu() - Free a journal_t structure.
>> + * @rcu: rcu list node relate to the journal want to free.
>> + *
>> + * Freeing a journal_t structure after a rcu grace period.
>> + */
>> +void jbd2_journal_release_rcu(struct rcu_head *rcu)
>> +{
>> + kfree(container_of(rcu, journal_t, j_rcu));
>> +}
>> +
>> +/**
>> + * jbd2_journal_destroy() - Release and free a journal_t structure.
>> + * @journal: Journal to act on.
>> + *
>> + * Release and free a journal_t structure once it is no longer in use
>> + * by the journaled object.
>> + */
>> +void jbd2_journal_destroy(journal_t *journal)
>> +{
>> + jbd2_journal_release(journal);
>> + kfree(journal);
>> +}
>>
>> /**
>> * jbd2_journal_check_used_features() - Check if features specified are used.
>> diff --git a/include/linux/jbd2.h b/include/linux/jbd2.h
>> index 99d3cd051ac3..39a8d04596a2 100644
>> --- a/include/linux/jbd2.h
>> +++ b/include/linux/jbd2.h
>> @@ -1238,6 +1238,13 @@ struct journal_s
>> */
>> __u32 j_csum_seed;
>>
>> + /**
>> + * @j_rcu:
>> + *
>> + * Prevent racing between accessing and destroy at the same time.
>> + */
>> + struct rcu_head j_rcu;
>> +
>> #ifdef CONFIG_DEBUG_LOCK_ALLOC
>> /**
>> * @j_trans_commit_map:
>> @@ -1509,7 +1516,9 @@ extern int jbd2_journal_set_features
>> extern void jbd2_journal_clear_features
>> (journal_t *, unsigned long, unsigned long, unsigned long);
>> extern int jbd2_journal_load (journal_t *journal);
>> -extern int jbd2_journal_destroy (journal_t *);
>> +extern void jbd2_journal_destroy (journal_t *);
>> +extern int jbd2_journal_release (journal_t *);
>> +extern void jbd2_journal_release_rcu (struct rcu_head *rcu);
>> extern int jbd2_journal_recover (journal_t *journal);
>> extern int jbd2_journal_wipe (journal_t *, int);
>> extern int jbd2_journal_skip_recovery (journal_t *);
>> --
>> 2.25.4
>>

2021-04-08 14:54:08

by Jan Kara

[permalink] [raw]
Subject: Re: [PATCH 3/3] ext4: add rcu to prevent use after free when umount filesystem

On Thu 08-04-21 22:38:08, Zhang Yi wrote:
> On 2021/4/8 21:56, Jan Kara wrote:
> > On Thu 08-04-21 19:36:18, Zhang Yi wrote:
> >> There is a race between bdev_try_to_free_page() and
> >> jbd2_journal_destroy() that could end up triggering a use after free
> >> issue about journal.
> >>
> >> drop cache umount filesystem
> >> bdev_try_to_free_page()
> >> get journal
> >> jbd2_journal_try_to_free_buffers() ext4_put_super()
> >> kfree(journal)
> >> access journal <-- lead to UAF
> >>
> >> The above race also could happens between the bdev_try_to_free_page()
> >> and the error path of ext4_fill_super(). This patch avoid this race by
> >> add rcu protection around accessing sbi->s_journal in
> >> bdev_try_to_free_page() and destroy the journal after an rcu grace
> >> period.
> >>
> >> Signed-off-by: Zhang Yi <[email protected]>
> >
> > OK, I see the problem. But cannot the use-after-free happen even for the
> > superblock itself (e.g., EXT4_SB(sb)->s_journal dereference)? I don't see
> > anything preventing that as blkdev_releasepage() just shamelessly does:
> >
> > super = BDEV_I(page->mapping->host)->bdev.bd_super
> >
> Hi, Jan.
>
> I checked the superblock. In theory, the bdev_try_to_free_page() is invoked
> with page locked, the umount process will wait the page unlock on
> kill_block_super()->..->kill_bdev()->truncate_inode_pages_range() before free
> superblock, so I guess the use-after-free problem couldn't happen in general.
> But I think it's fragile and may invalidate if the bdev has more than one
> operners(__blkdev_put() call kill_bdev only if bd_openers becomes zero)?

Yes, kill_bdev() is only called when bd_openers drops to 0 but there can be
other processes having the bdev open (non-exclusively).

Honza

--
Jan Kara <[email protected]>
SUSE Labs, CR