[PATCH] block: drbd: remove a stay unlock in __drbd_send_protocol()

[PATCH] block: drbd: remove a stay unlock in __drbd_send_protocol()

[Dan Carpenter]

There are two callers of this function and they both unlock the mutex so
this ends up being a double unlock.

Fixes: 44ed167da748 ("drbd: rcu_read_lock() and rcu_dereference() for tconn->net_conf")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
---
Static analisys.  Not tested.  There is a comment about the lock next to
the caller in drbd_nl.c that I didn't understand:

drivers/block/drbd/drbd_nl.c
  2509          crypto_free_shash(connection->integrity_tfm);
  2510          connection->integrity_tfm = crypto.integrity_tfm;
  2511          if (connection->cstate >= C_WF_REPORT_PARAMS && connection->agreed_pro_version >= 100)
  2512                  /* Do this without trying to take connection->data.mutex again.  */
                           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
What does this mean?  We're already holding that lock.  We took it near
the start of the function.

  2513                  __drbd_send_protocol(connection, P_PROTOCOL_UPDATE);
  2514  
  2515          crypto_free_shash(connection->cram_hmac_tfm);
  2516          connection->cram_hmac_tfm = crypto.cram_hmac_tfm;
  2517  
  2518          mutex_unlock(&connection->resource->conf_update);
  2519          mutex_unlock(&connection->data.mutex);
                ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
Unlocked here.

 drivers/block/drbd/drbd_main.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/drivers/block/drbd/drbd_main.c b/drivers/block/drbd/drbd_main.c
index 5b248763a672..a18155cdce41 100644
--- a/drivers/block/drbd/drbd_main.c
+++ b/drivers/block/drbd/drbd_main.c
@@ -786,7 +786,6 @@ int __drbd_send_protocol(struct drbd_connection *connection, enum drbd_packet cm
 
 	if (nc->tentative && connection->agreed_pro_version < 92) {
 		rcu_read_unlock();
-		mutex_unlock(&sock->mutex);
 		drbd_err(connection, "--dry-run is not supported by peer");
 		return -EOPNOTSUPP;
 	}
-- 
2.20.1

Re: [PATCH] block: drbd: remove a stay unlock in __drbd_send_protocol()

[Philipp Reisner]

Hi Dan,

yes, your patch it obviously correct. The comment you are
referring to is badly worded. We will remove it.

Jens,

are you taking this patch as it is?

best regards,
 Phil

Am Donnerstag, 7. November 2019, 08:48:47 CET schrieb Dan Carpenter:
> There are two callers of this function and they both unlock the mutex so
> this ends up being a double unlock.
> 
> Fixes: 44ed167da748 ("drbd: rcu_read_lock() and rcu_dereference() for
> tconn->net_conf") Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
> ---
> Static analisys.  Not tested.  There is a comment about the lock next to
> the caller in drbd_nl.c that I didn't understand:
> 
> drivers/block/drbd/drbd_nl.c
>   2509          crypto_free_shash(connection->integrity_tfm);
>   2510          connection->integrity_tfm = crypto.integrity_tfm;
>   2511          if (connection->cstate >= C_WF_REPORT_PARAMS &&
> connection->agreed_pro_version >= 100) 2512                  /* Do this
> without trying to take connection->data.mutex again.  */
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ What does this
> mean?  We're already holding that lock.  We took it near the start of the
> function.
> 
>   2513                  __drbd_send_protocol(connection, P_PROTOCOL_UPDATE);
> 2514
>   2515          crypto_free_shash(connection->cram_hmac_tfm);
>   2516          connection->cram_hmac_tfm = crypto.cram_hmac_tfm;
>   2517
>   2518          mutex_unlock(&connection->resource->conf_update);
>   2519          mutex_unlock(&connection->data.mutex);
>                 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> Unlocked here.
> 
>  drivers/block/drbd/drbd_main.c | 1 -
>  1 file changed, 1 deletion(-)
> 
> diff --git a/drivers/block/drbd/drbd_main.c b/drivers/block/drbd/drbd_main.c
> index 5b248763a672..a18155cdce41 100644
> --- a/drivers/block/drbd/drbd_main.c
> +++ b/drivers/block/drbd/drbd_main.c
> @@ -786,7 +786,6 @@ int __drbd_send_protocol(struct drbd_connection
> *connection, enum drbd_packet cm
> 
>  	if (nc->tentative && connection->agreed_pro_version < 92) {
>  		rcu_read_unlock();
> -		mutex_unlock(&sock->mutex);
>  		drbd_err(connection, "--dry-run is not supported by peer");
>  		return -EOPNOTSUPP;
>  	}


-- 
LINBIT | Keeping The Digital World Running

DRBD® and LINBIT® are registered trademarks of LINBIT, Austria.

Re: [PATCH] block: drbd: remove a stay unlock in __drbd_send_protocol()

[Jens Axboe]

On 11/8/19 3:46 AM, Philipp Reisner wrote:
> Hi Dan,
> 
> yes, your patch it obviously correct. The comment you are
> referring to is badly worded. We will remove it.
> 
> Jens,
> 
> are you taking this patch as it is?

Yep, I'll queue it up.

-- 
Jens Axboe

Re: [PATCH] block: drbd: remove a stay unlock in __drbd_send_protocol()

[Bart Van Assche]

On 11/6/19 11:48 PM, Dan Carpenter wrote:
> There are two callers of this function and they both unlock the mutex so
> this ends up being a double unlock.

Is there a typo in the patch subject (stay -> stray)?

Thanks,

Bart.

Re: [PATCH] block: drbd: remove a stay unlock in __drbd_send_protocol()

[Jens Axboe]

On 11/8/19 9:26 AM, Bart Van Assche wrote:
> On 11/6/19 11:48 PM, Dan Carpenter wrote:
>> There are two callers of this function and they both unlock the mutex so
>> this ends up being a double unlock.
> 
> Is there a typo in the patch subject (stay -> stray)?

I fixed that up while applying, fwiw.

-- 
Jens Axboe