Re: [PATCH] fcoe: Don't hold rtnl_mutex in fcoe_update_src_mac

[Bart Van Assche <bvanassche@acm.org>]

On 03/13/12 22:52, Robert Love wrote:
> The rtnl_mutex was held to protect calls to dev_uc_add
> and dev_uc_del. Holding rtnl is not required as those
> functions make use of the netif_addr_lock* API to
> protect the MAC changing.
>
> This change fixes the following regression by removing
> the rtnl usage when fcoe_update_src_mac is called.
>
> https://bugzilla.kernel.org/show_bug.cgi?id=42918
>
> the existing dependency chain (in reverse order) is:
>
> -> #1 (&fip->ctlr_mutex){+.+...}:
>        [<c1091f70>] lock_acquire+0x80/0x1b0
>        [<c147655d>] mutex_lock_nested+0x6d/0x340
>        [<f8970c32>] fcoe_ctlr_link_up+0x22/0x180 [libfcoe]
>        [<f894620e>] fcoe_create+0x47e/0x6e0 [fcoe]
>        [<f8973dd3>] fcoe_transport_create+0x143/0x250 [libfcoe]
>        [<c10527e0>] param_attr_store+0x30/0x60
>        [<c1052696>] module_attr_store+0x26/0x40
>        [<c11a201e>] sysfs_write_file+0xae/0x100
>        [<c11449df>] vfs_write+0x8f/0x160
>        [<c1144cbd>] sys_write+0x3d/0x70
>        [<c147a0c4>] syscall_call+0x7/0xb
>
> -> #0 (rtnl_mutex){+.+.+.}:
>        [<c109164b>] __lock_acquire+0x140b/0x1720
>        [<c1091f70>] lock_acquire+0x80/0x1b0
>        [<c147655d>] mutex_lock_nested+0x6d/0x340
>        [<c13a10c4>] rtnl_lock+0x14/0x20
>        [<f89445ac>] fcoe_update_src_mac+0x2c/0xb0 [fcoe]
>        [<f8971712>] fcoe_ctlr_timer_work+0x712/0xb60 [libfcoe]
>        [<c104fb69>] process_one_work+0x179/0x5d0
>        [<c10502f1>] worker_thread+0x121/0x2d0
>        [<c10550ed>] kthread+0x7d/0x90
>        [<c1481a82>] kernel_thread_helper+0x6/0x10
>
> other info that might help us debug this:
>
>  Possible unsafe locking scenario:
>
>        CPU0                    CPU1
>        ----                    ----
>   lock(&fip->ctlr_mutex);
>                                lock(rtnl_mutex);
>                                lock(&fip->ctlr_mutex);
>   lock(rtnl_mutex);
>
>  *** DEADLOCK ***
>
> Signed-off-by: Robert Love <robert.w.love@intel.com>
> ---
>  drivers/scsi/fcoe/fcoe.c |    2 --
>  1 files changed, 0 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/scsi/fcoe/fcoe.c b/drivers/scsi/fcoe/fcoe.c
> index e959960..85b8203 100644
> --- a/drivers/scsi/fcoe/fcoe.c
> +++ b/drivers/scsi/fcoe/fcoe.c
> @@ -539,13 +539,11 @@ static void fcoe_update_src_mac(struct fc_lport *lport, u8 *addr)
>  	struct fcoe_port *port = lport_priv(lport);
>  	struct fcoe_interface *fcoe = port->priv;
>  
> -	rtnl_lock();
>  	if (!is_zero_ether_addr(port->data_src_addr))
>  		dev_uc_del(fcoe->netdev, port->data_src_addr);
>  	if (!is_zero_ether_addr(addr))
>  		dev_uc_add(fcoe->netdev, addr);
>  	memcpy(port->data_src_addr, addr, ETH_ALEN);
> -	rtnl_unlock();
>  }
>  
>  /**

Seems to be sufficient here to avoid the lockdep complaint.

Tested-by: Bart Van Assche <bvanassche@acm.org>