linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH V2] bcache: fix potential deadlock problem in btree_gc_coalesce
@ 2020-04-26  8:06 Zhiqiang Liu
  2020-04-26  9:36 ` Coly Li
  0 siblings, 1 reply; 4+ messages in thread
From: Zhiqiang Liu @ 2020-04-26  8:06 UTC (permalink / raw)
  To: Coly Li, kmo, linux-bcache, linux-kernel
  Cc: wubo (T), Mingfangsen, Yanxiaodan, linfeilong, renxudong

From: Zhiqiang Liu <liuzhiqiang26@huawei.com>

coccicheck reports:
  drivers/md//bcache/btree.c:1538:1-7: preceding lock on line 1417

btree_gc_coalesce func is designed to coalesce two adjacent nodes in
new_nodes[GC_MERGE_NODES] and finally release one node. All nodes`write_lock,
new_nodes[i]->write_lock, are holded before coalescing adjacent nodes,
and them will be released after coalescing successfully.

However, if the coalescing process fails, such as no enough space of new_nodes[1]
to fit all of the remaining keys in new_nodes[0] and realloc keylist failed, we
will goto to out_nocoalesce tag directly without releasing new_nodes[i]->write_lock.
Then, a deadlock will occur after calling btree_node_free to free new_nodes[i],
which also try to acquire new_nodes[i]->write_lock.

Here, we add a new tag 'out_unlock_nocoalesce' before out_nocoalesce tag to release
new_nodes[i]->write_lock when coalescing process fails.

--
V1->V2: rewrite commit log (suggested by Coly Li) and rename the patch

Fixes: 2a285686c1 ("bcache: btree locking rework")
Signed-off-by: Zhiqiang Liu <liuzhiqiang26@huawei.com>
---
 drivers/md/bcache/btree.c | 8 ++++++--
 1 file changed, 6 insertions(+), 2 deletions(-)

diff --git a/drivers/md/bcache/btree.c b/drivers/md/bcache/btree.c
index fa872df4e770..cad8b0b97e33 100644
--- a/drivers/md/bcache/btree.c
+++ b/drivers/md/bcache/btree.c
@@ -1447,7 +1447,7 @@ static int btree_gc_coalesce(struct btree *b, struct btree_op *op,
 			if (__set_blocks(n1, n1->keys + n2->keys,
 					 block_bytes(b->c)) >
 			    btree_blocks(new_nodes[i]))
-				goto out_nocoalesce;
+				goto out_unlock_nocoalesce;

 			keys = n2->keys;
 			/* Take the key of the node we're getting rid of */
@@ -1476,7 +1476,7 @@ static int btree_gc_coalesce(struct btree *b, struct btree_op *op,

 		if (__bch_keylist_realloc(&keylist,
 					  bkey_u64s(&new_nodes[i]->key)))
-			goto out_nocoalesce;
+			goto out_unlock_nocoalesce;

 		bch_btree_node_write(new_nodes[i], &cl);
 		bch_keylist_add(&keylist, &new_nodes[i]->key);
@@ -1522,6 +1522,10 @@ static int btree_gc_coalesce(struct btree *b, struct btree_op *op,
 	/* Invalidated our iterator */
 	return -EINTR;

+out_unlock_nocoalesce:
+	for (i = 0; i < nodes; i++)
+		mutex_unlock(&new_nodes[i]->write_lock);
+
 out_nocoalesce:
 	closure_sync(&cl);

-- 
2.19.1



^ permalink raw reply related	[flat|nested] 4+ messages in thread

* Re: [PATCH V2] bcache: fix potential deadlock problem in btree_gc_coalesce
  2020-04-26  8:06 [PATCH V2] bcache: fix potential deadlock problem in btree_gc_coalesce Zhiqiang Liu
@ 2020-04-26  9:36 ` Coly Li
  0 siblings, 0 replies; 4+ messages in thread
From: Coly Li @ 2020-04-26  9:36 UTC (permalink / raw)
  To: Zhiqiang Liu, kmo, linux-bcache, linux-kernel
  Cc: wubo (T), Mingfangsen, Yanxiaodan, linfeilong, renxudong

On 2020/4/26 16:06, Zhiqiang Liu wrote:
> From: Zhiqiang Liu <liuzhiqiang26@huawei.com>
> 
> coccicheck reports:
>   drivers/md//bcache/btree.c:1538:1-7: preceding lock on line 1417
> 
> btree_gc_coalesce func is designed to coalesce two adjacent nodes in
> new_nodes[GC_MERGE_NODES] and finally release one node. All nodes`write_lock,
> new_nodes[i]->write_lock, are holded before coalescing adjacent nodes,
> and them will be released after coalescing successfully.
> 
> However, if the coalescing process fails, such as no enough space of new_nodes[1]
> to fit all of the remaining keys in new_nodes[0] and realloc keylist failed, we
> will goto to out_nocoalesce tag directly without releasing new_nodes[i]->write_lock.
> Then, a deadlock will occur after calling btree_node_free to free new_nodes[i],
> which also try to acquire new_nodes[i]->write_lock.
> 
> Here, we add a new tag 'out_unlock_nocoalesce' before out_nocoalesce tag to release
> new_nodes[i]->write_lock when coalescing process fails.
> 
> --
> V1->V2: rewrite commit log (suggested by Coly Li) and rename the patch
> 
> Fixes: 2a285686c1 ("bcache: btree locking rework")
> Signed-off-by: Zhiqiang Liu <liuzhiqiang26@huawei.com>

OK, I will add it to my for-test queue.

Thanks.

Coly Li


> ---
>  drivers/md/bcache/btree.c | 8 ++++++--
>  1 file changed, 6 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/md/bcache/btree.c b/drivers/md/bcache/btree.c
> index fa872df4e770..cad8b0b97e33 100644
> --- a/drivers/md/bcache/btree.c
> +++ b/drivers/md/bcache/btree.c
> @@ -1447,7 +1447,7 @@ static int btree_gc_coalesce(struct btree *b, struct btree_op *op,
>  			if (__set_blocks(n1, n1->keys + n2->keys,
>  					 block_bytes(b->c)) >
>  			    btree_blocks(new_nodes[i]))
> -				goto out_nocoalesce;
> +				goto out_unlock_nocoalesce;
> 
>  			keys = n2->keys;
>  			/* Take the key of the node we're getting rid of */
> @@ -1476,7 +1476,7 @@ static int btree_gc_coalesce(struct btree *b, struct btree_op *op,
> 
>  		if (__bch_keylist_realloc(&keylist,
>  					  bkey_u64s(&new_nodes[i]->key)))
> -			goto out_nocoalesce;
> +			goto out_unlock_nocoalesce;
> 
>  		bch_btree_node_write(new_nodes[i], &cl);
>  		bch_keylist_add(&keylist, &new_nodes[i]->key);
> @@ -1522,6 +1522,10 @@ static int btree_gc_coalesce(struct btree *b, struct btree_op *op,
>  	/* Invalidated our iterator */
>  	return -EINTR;
> 
> +out_unlock_nocoalesce:
> +	for (i = 0; i < nodes; i++)
> +		mutex_unlock(&new_nodes[i]->write_lock);
> +
>  out_nocoalesce:
>  	closure_sync(&cl);
> 


^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [PATCH V2] bcache: fix potential deadlock problem in btree_gc_coalesce
  2020-04-26 18:16 Markus Elfring
@ 2020-06-12  1:01 ` Zhiqiang Liu
  0 siblings, 0 replies; 4+ messages in thread
From: Zhiqiang Liu @ 2020-06-12  1:01 UTC (permalink / raw)
  To: Markus Elfring, linux-bcache
  Cc: linux-kernel, kernel-janitors, Coly Li, Feilong Lin,
	Kent Overstreet, mingfangsen, renxudong1, Wu Bo, yanxiaodan

Sorry for late reply. I will improve the commit log as your suggestion in v3 patch.
Thanks for your suggestion.

On 2020/4/27 2:16, Markus Elfring wrote:
>> --
>> V1->V2: rewrite commit log (suggested by Coly Li) and rename the patch
> 
> * The patch version description should be placed behind triple dashes.
> 
> * I would find a shorter version identification (without the arrow)
>   also sufficient.
> 
> 
>> Fixes: 2a285686c1 ("bcache: btree locking rework")>
> 
> Will a longer commit identifier be safer for the final change description?
> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/Documentation/process/submitting-patches.rst?id=b2768df24ec400dd4f7fa79542f797e904812053#n183
> 
> Regards,
> Markus
> 


^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [PATCH V2] bcache: fix potential deadlock problem in btree_gc_coalesce
@ 2020-04-26 18:16 Markus Elfring
  2020-06-12  1:01 ` Zhiqiang Liu
  0 siblings, 1 reply; 4+ messages in thread
From: Markus Elfring @ 2020-04-26 18:16 UTC (permalink / raw)
  To: Zhiqiang Liu, linux-bcache
  Cc: linux-kernel, kernel-janitors, Coly Li, Feilong Lin,
	Kent Overstreet, mingfangsen, renxudong1, Wu Bo, yanxiaodan

> --
> V1->V2: rewrite commit log (suggested by Coly Li) and rename the patch

* The patch version description should be placed behind triple dashes.

* I would find a shorter version identification (without the arrow)
  also sufficient.


> Fixes: 2a285686c1 ("bcache: btree locking rework")>

Will a longer commit identifier be safer for the final change description?
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/Documentation/process/submitting-patches.rst?id=b2768df24ec400dd4f7fa79542f797e904812053#n183

Regards,
Markus

^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2020-06-12  1:01 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-04-26  8:06 [PATCH V2] bcache: fix potential deadlock problem in btree_gc_coalesce Zhiqiang Liu
2020-04-26  9:36 ` Coly Li
2020-04-26 18:16 Markus Elfring
2020-06-12  1:01 ` Zhiqiang Liu

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).