* [PATCH -next] jbd2: Fix null-ptr-deref when process reserved list in jbd2_journal_commit_transaction
@ 2022-03-17 1:27 Ye Bin
2022-03-17 10:00 ` Jan Kara
0 siblings, 1 reply; 3+ messages in thread
From: Ye Bin @ 2022-03-17 1:27 UTC (permalink / raw)
To: tytso, adilger.kernel, linux-ext4; +Cc: linux-kernel, jack, lczerner, Ye Bin
we got issue as follows:
[ 72.796117] EXT4-fs error (device sda): ext4_journal_check_start:83: comm fallocate: Detected aborted journal
[ 72.826847] EXT4-fs (sda): Remounting filesystem read-only
fallocate: fallocate failed: Read-only file system
[ 74.791830] jbd2_journal_commit_transaction: jh=0xffff9cfefe725d90 bh=0x0000000000000000 end delay
[ 74.793597] ------------[ cut here ]------------
[ 74.794203] kernel BUG at fs/jbd2/transaction.c:2063!
[ 74.794886] invalid opcode: 0000 [#1] PREEMPT SMP PTI
[ 74.795533] CPU: 4 PID: 2260 Comm: jbd2/sda-8 Not tainted 5.17.0-rc8-next-20220315-dirty #150
[ 74.798327] RIP: 0010:__jbd2_journal_unfile_buffer+0x3e/0x60
[ 74.801971] RSP: 0018:ffffa828c24a3cb8 EFLAGS: 00010202
[ 74.802694] RAX: 0000000000000000 RBX: 0000000000000000 RCX: 0000000000000000
[ 74.803601] RDX: 0000000000000001 RSI: ffff9cfefe725d90 RDI: ffff9cfefe725d90
[ 74.804554] RBP: ffff9cfefe725d90 R08: 0000000000000000 R09: ffffa828c24a3b20
[ 74.805471] R10: 0000000000000001 R11: 0000000000000001 R12: ffff9cfefe725d90
[ 74.806385] R13: ffff9cfefe725d98 R14: 0000000000000000 R15: ffff9cfe833a4d00
[ 74.807301] FS: 0000000000000000(0000) GS:ffff9d01afb00000(0000) knlGS:0000000000000000
[ 74.808338] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 74.809084] CR2: 00007f2b81bf4000 CR3: 0000000100056000 CR4: 00000000000006e0
[ 74.810047] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[ 74.810981] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[ 74.811897] Call Trace:
[ 74.812241] <TASK>
[ 74.812566] __jbd2_journal_refile_buffer+0x12f/0x180
[ 74.813246] jbd2_journal_refile_buffer+0x4c/0xa0
[ 74.813869] jbd2_journal_commit_transaction.cold+0xa1/0x148
[ 74.817550] kjournald2+0xf8/0x3e0
[ 74.819056] kthread+0x153/0x1c0
[ 74.819963] ret_from_fork+0x22/0x30
Above issue may happen as follows:
write truncate kjournald2
generic_perform_write
ext4_write_begin
ext4_walk_page_buffers
do_journal_get_write_access ->add BJ_Reserved list
ext4_journalled_write_end
ext4_walk_page_buffers
write_end_fn
ext4_handle_dirty_metadata
***************JBD2 ABORT**************
jbd2_journal_dirty_metadata
-> return -EROFS, jh in reserved_list
jbd2_journal_commit_transaction
while (commit_transaction->t_reserved_list)
jh = commit_transaction->t_reserved_list;
truncate_pagecache_range
do_invalidatepage
ext4_journalled_invalidatepage
jbd2_journal_invalidatepage
journal_unmap_buffer
__dispose_buffer
__jbd2_journal_unfile_buffer
jbd2_journal_put_journal_head ->put last ref_count
__journal_remove_journal_head
bh->b_private = NULL;
jh->b_bh = NULL;
jbd2_journal_refile_buffer(journal, jh);
bh = jh2bh(jh);
->bh is NULL, later will trigger null-ptr-deref
journal_free_journal_head(jh);
As after 96f1e0974575 commit, handle reserved list will not hold "journal->j_state_lock"
when kjournald2 commit transaction. So journal_unmap_buffer maybe free
journal_head when handle reserved list. And lead to null-ptr-deref or some
strange errors.
As reserved list almost time is empty. Use "journal->j_state_lock" to protect
handle reserved list can simply solve above issue.
Fixes: 96f1e0974575("jbd2: avoid long hold times of j_state_lock while committing a transaction")
Signed-off-by: Ye Bin <yebin10@huawei.com>
---
fs/jbd2/commit.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/fs/jbd2/commit.c b/fs/jbd2/commit.c
index 5b9408e3b370..2b737b928d26 100644
--- a/fs/jbd2/commit.c
+++ b/fs/jbd2/commit.c
@@ -488,7 +488,6 @@ void jbd2_journal_commit_transaction(journal_t *journal)
jbd2_journal_wait_updates(journal);
commit_transaction->t_state = T_SWITCH;
- write_unlock(&journal->j_state_lock);
J_ASSERT (atomic_read(&commit_transaction->t_outstanding_credits) <=
journal->j_max_transaction_buffers);
@@ -527,6 +526,7 @@ void jbd2_journal_commit_transaction(journal_t *journal)
jbd2_journal_refile_buffer(journal, jh);
}
+ write_unlock(&journal->j_state_lock);
/*
* Now try to drop any written-back buffers from the journal's
* checkpoint lists. We do this *before* commit because it potentially
--
2.31.1
^ permalink raw reply related [flat|nested] 3+ messages in thread
* Re: [PATCH -next] jbd2: Fix null-ptr-deref when process reserved list in jbd2_journal_commit_transaction
2022-03-17 1:27 [PATCH -next] jbd2: Fix null-ptr-deref when process reserved list in jbd2_journal_commit_transaction Ye Bin
@ 2022-03-17 10:00 ` Jan Kara
[not found] ` <62331D5C.2010808@huawei.com>
0 siblings, 1 reply; 3+ messages in thread
From: Jan Kara @ 2022-03-17 10:00 UTC (permalink / raw)
To: Ye Bin; +Cc: tytso, adilger.kernel, linux-ext4, linux-kernel, jack, lczerner
On Thu 17-03-22 09:27:55, Ye Bin wrote:
> we got issue as follows:
> [ 72.796117] EXT4-fs error (device sda): ext4_journal_check_start:83: comm fallocate: Detected aborted journal
> [ 72.826847] EXT4-fs (sda): Remounting filesystem read-only
> fallocate: fallocate failed: Read-only file system
> [ 74.791830] jbd2_journal_commit_transaction: jh=0xffff9cfefe725d90 bh=0x0000000000000000 end delay
> [ 74.793597] ------------[ cut here ]------------
> [ 74.794203] kernel BUG at fs/jbd2/transaction.c:2063!
> [ 74.794886] invalid opcode: 0000 [#1] PREEMPT SMP PTI
> [ 74.795533] CPU: 4 PID: 2260 Comm: jbd2/sda-8 Not tainted 5.17.0-rc8-next-20220315-dirty #150
> [ 74.798327] RIP: 0010:__jbd2_journal_unfile_buffer+0x3e/0x60
> [ 74.801971] RSP: 0018:ffffa828c24a3cb8 EFLAGS: 00010202
> [ 74.802694] RAX: 0000000000000000 RBX: 0000000000000000 RCX: 0000000000000000
> [ 74.803601] RDX: 0000000000000001 RSI: ffff9cfefe725d90 RDI: ffff9cfefe725d90
> [ 74.804554] RBP: ffff9cfefe725d90 R08: 0000000000000000 R09: ffffa828c24a3b20
> [ 74.805471] R10: 0000000000000001 R11: 0000000000000001 R12: ffff9cfefe725d90
> [ 74.806385] R13: ffff9cfefe725d98 R14: 0000000000000000 R15: ffff9cfe833a4d00
> [ 74.807301] FS: 0000000000000000(0000) GS:ffff9d01afb00000(0000) knlGS:0000000000000000
> [ 74.808338] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> [ 74.809084] CR2: 00007f2b81bf4000 CR3: 0000000100056000 CR4: 00000000000006e0
> [ 74.810047] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
> [ 74.810981] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
> [ 74.811897] Call Trace:
> [ 74.812241] <TASK>
> [ 74.812566] __jbd2_journal_refile_buffer+0x12f/0x180
> [ 74.813246] jbd2_journal_refile_buffer+0x4c/0xa0
> [ 74.813869] jbd2_journal_commit_transaction.cold+0xa1/0x148
> [ 74.817550] kjournald2+0xf8/0x3e0
> [ 74.819056] kthread+0x153/0x1c0
> [ 74.819963] ret_from_fork+0x22/0x30
>
> Above issue may happen as follows:
> write truncate kjournald2
> generic_perform_write
> ext4_write_begin
> ext4_walk_page_buffers
> do_journal_get_write_access ->add BJ_Reserved list
> ext4_journalled_write_end
> ext4_walk_page_buffers
> write_end_fn
> ext4_handle_dirty_metadata
> ***************JBD2 ABORT**************
> jbd2_journal_dirty_metadata
> -> return -EROFS, jh in reserved_list
> jbd2_journal_commit_transaction
> while (commit_transaction->t_reserved_list)
> jh = commit_transaction->t_reserved_list;
> truncate_pagecache_range
> do_invalidatepage
> ext4_journalled_invalidatepage
> jbd2_journal_invalidatepage
> journal_unmap_buffer
> __dispose_buffer
> __jbd2_journal_unfile_buffer
> jbd2_journal_put_journal_head ->put last ref_count
> __journal_remove_journal_head
> bh->b_private = NULL;
> jh->b_bh = NULL;
> jbd2_journal_refile_buffer(journal, jh);
> bh = jh2bh(jh);
> ->bh is NULL, later will trigger null-ptr-deref
> journal_free_journal_head(jh);
>
> As after 96f1e0974575 commit, handle reserved list will not hold "journal->j_state_lock"
> when kjournald2 commit transaction. So journal_unmap_buffer maybe free
> journal_head when handle reserved list. And lead to null-ptr-deref or some
> strange errors.
> As reserved list almost time is empty. Use "journal->j_state_lock" to protect
> handle reserved list can simply solve above issue.
>
> Fixes: 96f1e0974575("jbd2: avoid long hold times of j_state_lock while committing a transaction")
> Signed-off-by: Ye Bin <yebin10@huawei.com>
Good spotting! Thanks for the analysis and the patch!
> diff --git a/fs/jbd2/commit.c b/fs/jbd2/commit.c
> index 5b9408e3b370..2b737b928d26 100644
> --- a/fs/jbd2/commit.c
> +++ b/fs/jbd2/commit.c
> @@ -488,7 +488,6 @@ void jbd2_journal_commit_transaction(journal_t *journal)
> jbd2_journal_wait_updates(journal);
>
> commit_transaction->t_state = T_SWITCH;
> - write_unlock(&journal->j_state_lock);
>
> J_ASSERT (atomic_read(&commit_transaction->t_outstanding_credits) <=
> journal->j_max_transaction_buffers);
> @@ -527,6 +526,7 @@ void jbd2_journal_commit_transaction(journal_t *journal)
> jbd2_journal_refile_buffer(journal, jh);
> }
>
> + write_unlock(&journal->j_state_lock);
> /*
> * Now try to drop any written-back buffers from the journal's
> * checkpoint lists. We do this *before* commit because it potentially
Honestly, using j_state_lock seems like unnecessarily big hammer for this.
I'd rather use journal->j_list_lock for this which is much more natural for
list processing. And we grab it anyway from jbd2_journal_refile_buffer() so
it is not like there will be any additional overhead anyway.
Honza
--
Jan Kara <jack@suse.com>
SUSE Labs, CR
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: [PATCH -next] jbd2: Fix null-ptr-deref when process reserved list in jbd2_journal_commit_transaction
[not found] ` <62331D5C.2010808@huawei.com>
@ 2022-03-17 13:30 ` Jan Kara
0 siblings, 0 replies; 3+ messages in thread
From: Jan Kara @ 2022-03-17 13:30 UTC (permalink / raw)
To: yebin; +Cc: Jan Kara, tytso, adilger.kernel, linux-ext4, linux-kernel, lczerner
On Thu 17-03-22 19:37:00, yebin wrote:
>
>
> On 2022/3/17 18:00, Jan Kara wrote:
> > On Thu 17-03-22 09:27:55, Ye Bin wrote:
> > > we got issue as follows:
> > > [ 72.796117] EXT4-fs error (device sda): ext4_journal_check_start:83: comm fallocate: Detected aborted journal
> > > [ 72.826847] EXT4-fs (sda): Remounting filesystem read-only
> > > fallocate: fallocate failed: Read-only file system
> > > [ 74.791830] jbd2_journal_commit_transaction: jh=0xffff9cfefe725d90 bh=0x0000000000000000 end delay
> > > [ 74.793597] ------------[ cut here ]------------
> > > [ 74.794203] kernel BUG at fs/jbd2/transaction.c:2063!
> > > [ 74.794886] invalid opcode: 0000 [#1] PREEMPT SMP PTI
> > > [ 74.795533] CPU: 4 PID: 2260 Comm: jbd2/sda-8 Not tainted 5.17.0-rc8-next-20220315-dirty #150
> > > [ 74.798327] RIP: 0010:__jbd2_journal_unfile_buffer+0x3e/0x60
> > > [ 74.801971] RSP: 0018:ffffa828c24a3cb8 EFLAGS: 00010202
> > > [ 74.802694] RAX: 0000000000000000 RBX: 0000000000000000 RCX: 0000000000000000
> > > [ 74.803601] RDX: 0000000000000001 RSI: ffff9cfefe725d90 RDI: ffff9cfefe725d90
> > > [ 74.804554] RBP: ffff9cfefe725d90 R08: 0000000000000000 R09: ffffa828c24a3b20
> > > [ 74.805471] R10: 0000000000000001 R11: 0000000000000001 R12: ffff9cfefe725d90
> > > [ 74.806385] R13: ffff9cfefe725d98 R14: 0000000000000000 R15: ffff9cfe833a4d00
> > > [ 74.807301] FS: 0000000000000000(0000) GS:ffff9d01afb00000(0000) knlGS:0000000000000000
> > > [ 74.808338] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> > > [ 74.809084] CR2: 00007f2b81bf4000 CR3: 0000000100056000 CR4: 00000000000006e0
> > > [ 74.810047] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
> > > [ 74.810981] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
> > > [ 74.811897] Call Trace:
> > > [ 74.812241] <TASK>
> > > [ 74.812566] __jbd2_journal_refile_buffer+0x12f/0x180
> > > [ 74.813246] jbd2_journal_refile_buffer+0x4c/0xa0
> > > [ 74.813869] jbd2_journal_commit_transaction.cold+0xa1/0x148
> > > [ 74.817550] kjournald2+0xf8/0x3e0
> > > [ 74.819056] kthread+0x153/0x1c0
> > > [ 74.819963] ret_from_fork+0x22/0x30
> > >
> > > Above issue may happen as follows:
> > > write truncate kjournald2
> > > generic_perform_write
> > > ext4_write_begin
> > > ext4_walk_page_buffers
> > > do_journal_get_write_access ->add BJ_Reserved list
> > > ext4_journalled_write_end
> > > ext4_walk_page_buffers
> > > write_end_fn
> > > ext4_handle_dirty_metadata
> > > ***************JBD2 ABORT**************
> > > jbd2_journal_dirty_metadata
> > > -> return -EROFS, jh in reserved_list
> > > jbd2_journal_commit_transaction
> > > while (commit_transaction->t_reserved_list)
> > > jh = commit_transaction->t_reserved_list;
> > > truncate_pagecache_range
> > > do_invalidatepage
> > > ext4_journalled_invalidatepage
> > > jbd2_journal_invalidatepage
> > > journal_unmap_buffer
> > > __dispose_buffer
> > > __jbd2_journal_unfile_buffer
> > > jbd2_journal_put_journal_head ->put last ref_count
> > > __journal_remove_journal_head
> > > bh->b_private = NULL;
> > > jh->b_bh = NULL;
> > > jbd2_journal_refile_buffer(journal, jh);
> > > bh = jh2bh(jh);
> > > ->bh is NULL, later will trigger null-ptr-deref
> > > journal_free_journal_head(jh);
> > >
> > > As after 96f1e0974575 commit, handle reserved list will not hold "journal->j_state_lock"
> > > when kjournald2 commit transaction. So journal_unmap_buffer maybe free
> > > journal_head when handle reserved list. And lead to null-ptr-deref or some
> > > strange errors.
> > > As reserved list almost time is empty. Use "journal->j_state_lock" to protect
> > > handle reserved list can simply solve above issue.
> > >
> > > Fixes: 96f1e0974575("jbd2: avoid long hold times of j_state_lock while committing a transaction")
> > > Signed-off-by: Ye Bin <yebin10@huawei.com>
> > Good spotting! Thanks for the analysis and the patch!
> >
> > > diff --git a/fs/jbd2/commit.c b/fs/jbd2/commit.c
> > > index 5b9408e3b370..2b737b928d26 100644
> > > --- a/fs/jbd2/commit.c
> > > +++ b/fs/jbd2/commit.c
> > > @@ -488,7 +488,6 @@ void jbd2_journal_commit_transaction(journal_t *journal)
> > > jbd2_journal_wait_updates(journal);
> > > commit_transaction->t_state = T_SWITCH;
> > > - write_unlock(&journal->j_state_lock);
> > > J_ASSERT (atomic_read(&commit_transaction->t_outstanding_credits) <=
> > > journal->j_max_transaction_buffers);
> > > @@ -527,6 +526,7 @@ void jbd2_journal_commit_transaction(journal_t *journal)
> > > jbd2_journal_refile_buffer(journal, jh);
> > > }
> > > + write_unlock(&journal->j_state_lock);
> > > /*
> > > * Now try to drop any written-back buffers from the journal's
> > > * checkpoint lists. We do this *before* commit because it potentially
> > Honestly, using j_state_lock seems like unnecessarily big hammer for this.
> > I'd rather use journal->j_list_lock for this which is much more natural for
> > list processing. And we grab it anyway from jbd2_journal_refile_buffer() so
> > it is not like there will be any additional overhead anyway.
> >
> > Honza
> If use journal->j_list_lock for reserved list, we must change order of
> jh->b_state_lock and
> journal->j_list_lock.
> <../cgi-bin/global.cgi?pattern=j_list_lock&type=symbol>
> I have another idea as follows. Do you think it is feasible?
Oh right, I forgot the lock ordering is not very convenient for these two
locks. Ok, then I probably prefer your original solution to the one below.
But please add a comment like:
We use journal->j_state_lock here to serialize processing of
t_reserved_list with eviction of buffers from journal_unmap_buffer().
Thanks!
Honza
> diff --git a/fs/jbd2/transaction.c b/fs/jbd2/transaction.c
> index 73ed02f061e1..00c51a92a5cc 100644
> --- a/fs/jbd2/transaction.c
> +++ b/fs/jbd2/transaction.c
> @@ -2282,6 +2282,7 @@ static int journal_unmap_buffer(journal_t *journal,
> struct buffer_head *bh,
> if (!jh)
> goto zap_buffer_unlocked;
>
> +repeat:
> /* OK, we have data buffer in journaled mode */
> write_lock(&journal->j_state_lock);
> spin_lock(&jh->b_state_lock);
> @@ -2357,6 +2358,11 @@ static int journal_unmap_buffer(journal_t *journal,
> struct buffer_head *bh,
> goto zap_buffer;
> }
> }
> + } else if (transaction->t_state == T_SWITCH) {
> + spin_unlock(&journal->j_list_lock);
> + spin_unlock(&jh->b_state_lock);
> + wait_transaction_switching(journal);
> + goto repeat;
> } else if (transaction == journal->j_committing_transaction) {
> JBUFFER_TRACE(jh, "on committing transaction");
> /*
> --
>
>
--
Jan Kara <jack@suse.com>
SUSE Labs, CR
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2022-03-17 13:30 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2022-03-17 1:27 [PATCH -next] jbd2: Fix null-ptr-deref when process reserved list in jbd2_journal_commit_transaction Ye Bin
2022-03-17 10:00 ` Jan Kara
[not found] ` <62331D5C.2010808@huawei.com>
2022-03-17 13:30 ` Jan Kara
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).