[Top] [All Lists]

Re: [PATCH 2.6.12-rc2] bonding: partially back out dev_set_mac_address

To: Jay Vosburgh <fubar@xxxxxxxxxx>
Subject: Re: [PATCH 2.6.12-rc2] bonding: partially back out dev_set_mac_address
From: Herbert Xu <herbert@xxxxxxxxxxxxxxxxxxx>
Date: Wed, 27 Apr 2005 12:15:11 +1000
Cc: "David S. Miller" <davem@xxxxxxxxxxxxx>, netdev@xxxxxxxxxxx, jgarzik@xxxxxxxxx
In-reply-to: <200504270209.j3R291ml007431@xxxxxxxxxxxxxxxxxxxxxx>
References: <20050426111845.GA8968@xxxxxxxxxxxxxxxxxxx> <200504270209.j3R291ml007431@xxxxxxxxxxxxxxxxxxxxxx>
Sender: netdev-bounce@xxxxxxxxxxx
User-agent: Mutt/1.5.6+20040907i
On Tue, Apr 26, 2005 at 07:09:01PM -0700, Jay Vosburgh wrote:
> >dev_ioctl => bond_enslave => bond_check_dev_link => slave_dev->ioctl
>       Why would it deadlock?  dev_ioctl holds RTNL, bonding grabs
> various bond locks, and the slave device ioctl handler may or may not
> get a lock of its own.  

I get it now.  I had thought that it was calling sys_ioctl but it's
going direct to slave_dev->ioctl which bypasses the rtnl.

So the bond_enslave path is safe but the timer path isn't as you
noted previously.

> >For the amount of work that these timers are doing, the overhead is
> >pretty small.  It is also gentler on the system when the CPU load
> >goes up.
>       Just so I'm clear: by "the overhead" do you mean the overhead of
> running everything in a work queue, or the overhead of calling out from
> a timer to a work queue for "special occasions"?

I'm referring to running everything out of a work queue.

Visit Openswan at
Email: Herbert Xu ~{PmV>HI~} <herbert@xxxxxxxxxxxxxxxxxxx>
Home Page:
PGP Key:

<Prev in Thread] Current Thread [Next in Thread>