On Wed, Sep 28, 2011 at 03:07:52AM +0000, Adrian Chadd wrote:
> Author: adrian
> Date: Wed Sep 28 03:07:51 2011
> New Revision: 225821
> URL: http://svn.freebsd.org/changeset/base/225821
> 
> Log:
>   Fix lock order to be correcter.
>   
>   Nothing else locks these two queues (cabq, avp mcastq), but it should
>   be consistent and correct.
> 
> Modified:
>   head/sys/dev/ath/if_ath.c
> 
> Modified: head/sys/dev/ath/if_ath.c
> ==============================================================================
> --- head/sys/dev/ath/if_ath.c Wed Sep 28 03:05:04 2011        (r225820)
> +++ head/sys/dev/ath/if_ath.c Wed Sep 28 03:07:51 2011        (r225821)
> @@ -2620,8 +2620,8 @@ ath_beacon_generate(struct ath_softc *sc
>               }
>               /* NB: gated by beacon so safe to start here */
>               ath_hal_txstart(ah, cabq->axq_qnum);
> -             ATH_TXQ_UNLOCK(cabq);
>               ATH_TXQ_UNLOCK(&avp->av_mcastq);
> +             ATH_TXQ_UNLOCK(cabq);
>       }
>       return bf;
>  }
Does the unlock order matter ? Unlock cannot block.

I do not object against the change, rather, I am nit-picking against the
commit message.

Attachment: pgpuuYLtFeh95.pgp
Description: PGP signature

Reply via email to