Hi,
On Tue, Jun 23, 2020 at 09:57:50PM +0200, Greg Kroah-Hartman wrote:
> From: Bob Peterson <rpeterso@xxxxxxxxxx>
>
> [ Upstream commit 83d060ca8d90fa1e3feac227f995c013100862d3 ]
>
> Before this patch, transactions could be merged into the system
> transaction by function gfs2_merge_trans(), but the transaction ail
> lists were never merged. Because the ail flushing mechanism can run
> separately, bd elements can be attached to the transaction's buffer
> list during the transaction (trans_add_meta, etc) but quickly moved
> to its ail lists. Later, in function gfs2_trans_end, the transaction
> can be freed (by gfs2_trans_end) while it still has bd elements
> queued to its ail lists, which can cause it to either lose track of
> the bd elements altogether (memory leak) or worse, reference the bd
> elements after the parent transaction has been freed.
>
> Although I've not seen any serious consequences, the problem becomes
> apparent with the previous patch's addition of:
>
> gfs2_assert_warn(sdp, list_empty(&tr->tr_ail1_list));
>
> to function gfs2_trans_free().
>
> This patch adds logic into gfs2_merge_trans() to move the merged
> transaction's ail lists to the sdp transaction. This prevents the
> use-after-free. To do this properly, we need to hold the ail lock,
> so we pass sdp into the function instead of the transaction itself.
>
> Signed-off-by: Bob Peterson <rpeterso@xxxxxxxxxx>
> Signed-off-by: Andreas Gruenbacher <agruenba@xxxxxxxxxx>
> Signed-off-by: Sasha Levin <sashal@xxxxxxxxxx>
> ---
> fs/gfs2/log.c | 11 +++++++++--
> 1 file changed, 9 insertions(+), 2 deletions(-)
>
> diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c
> index d3f0612e33471..06752db213d21 100644
> --- a/fs/gfs2/log.c
> +++ b/fs/gfs2/log.c
> @@ -877,8 +877,10 @@ void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl, u32 flags)
> * @new: New transaction to be merged
> */
>
> -static void gfs2_merge_trans(struct gfs2_trans *old, struct gfs2_trans *new)
> +static void gfs2_merge_trans(struct gfs2_sbd *sdp, struct gfs2_trans *new)
> {
> + struct gfs2_trans *old = sdp->sd_log_tr;
> +
> WARN_ON_ONCE(!test_bit(TR_ATTACHED, &old->tr_flags));
>
> old->tr_num_buf_new += new->tr_num_buf_new;
> @@ -890,6 +892,11 @@ static void gfs2_merge_trans(struct gfs2_trans *old, struct gfs2_trans *new)
>
> list_splice_tail_init(&new->tr_databuf, &old->tr_databuf);
> list_splice_tail_init(&new->tr_buf, &old->tr_buf);
> +
> + spin_lock(&sdp->sd_ail_lock);
> + list_splice_tail_init(&new->tr_ail1_list, &old->tr_ail1_list);
> + list_splice_tail_init(&new->tr_ail2_list, &old->tr_ail2_list);
> + spin_unlock(&sdp->sd_ail_lock);
> }
>
> static void log_refund(struct gfs2_sbd *sdp, struct gfs2_trans *tr)
> @@ -901,7 +908,7 @@ static void log_refund(struct gfs2_sbd *sdp, struct gfs2_trans *tr)
> gfs2_log_lock(sdp);
>
> if (sdp->sd_log_tr) {
> - gfs2_merge_trans(sdp->sd_log_tr, tr);
> + gfs2_merge_trans(sdp, tr);
> } else if (tr->tr_num_buf_new || tr->tr_num_databuf_new) {
> gfs2_assert_withdraw(sdp, test_bit(TR_ALLOCED, &tr->tr_flags));
> sdp->sd_log_tr = tr;
> --
> 2.25.1
In Debian two user confirmed issues on writing on a GFS2 partition
with this commit applied. The initial Debian report is at
https://bugs.debian.org/968567 and Daniel Craig reported it into
Bugzilla at https://bugzilla.kernel.org/show_bug.cgi?id=209217 .
Writing to a gfs2 filesystem fails and results in a soft lookup of the
machine for kernels with that commit applied. I cannot reporduce the
issue myself due not having a respective setup available, but Daniel
described a minimal serieos of steps to reproduce the issue.
This might affect as well other stable series where this commit was
applied, as there was a similar report for someone running 5.4.58 in
https://www.redhat.com/archives/linux-cluster/2020-August/msg00000.html
.
Regards,
Salvatore
