Re: [PATCH net 3/3] bql: Avoid possible inconsistent calculation.

[Eric Dumazet <eric.dumazet@gmail.com>]

On Thu, 2012-05-31 at 07:25 +0900, Hiroaki SHIMODA wrote:
> dql->num_queued could change while processing dql_completed().
> To provide consistent calculation, added an on stack variable.
> 
> Signed-off-by: Hiroaki SHIMODA <shimoda.hiroaki@gmail.com>
> Cc: Tom Herbert <therbert@google.com>
> Cc: Eric Dumazet <eric.dumazet@gmail.com>
> Cc: Denys Fedoryshchenko <denys@visp.net.lb>
> ---
>  lib/dynamic_queue_limits.c |   12 +++++++-----
>  1 files changed, 7 insertions(+), 5 deletions(-)
> 
> diff --git a/lib/dynamic_queue_limits.c b/lib/dynamic_queue_limits.c
> index 0fafa77..0777c5a 100644
> --- a/lib/dynamic_queue_limits.c
> +++ b/lib/dynamic_queue_limits.c
> @@ -17,16 +17,18 @@
>  void dql_completed(struct dql *dql, unsigned int count)
>  {
>  	unsigned int inprogress, prev_inprogress, limit;
> -	unsigned int ovlimit, completed;
> +	unsigned int ovlimit, completed, num_queued;
>  	bool all_prev_completed;
>  
> +	num_queued = ACCESS_ONCE(dql->num_queued);
> +
>  	/* Can't complete more than what's in queue */
> -	BUG_ON(count > dql->num_queued - dql->num_completed);
> +	BUG_ON(count > num_queued - dql->num_completed);
>  
>  	completed = dql->num_completed + count;
>  	limit = dql->limit;
> -	ovlimit = POSDIFF(dql->num_queued - dql->num_completed, limit);
> -	inprogress = dql->num_queued - completed;
> +	ovlimit = POSDIFF(num_queued - dql->num_completed, limit);
> +	inprogress = num_queued - completed;
>  	prev_inprogress = dql->prev_num_queued - dql->num_completed;
>  	all_prev_completed = AFTER_EQ(completed, dql->prev_num_queued);
>  
> @@ -106,7 +108,7 @@ void dql_completed(struct dql *dql, unsigned int count)
>  	dql->prev_ovlimit = ovlimit;
>  	dql->prev_last_obj_cnt = dql->last_obj_cnt;
>  	dql->num_completed = completed;
> -	dql->prev_num_queued = dql->num_queued;
> +	dql->prev_num_queued = num_queued;
>  }
>  EXPORT_SYMBOL(dql_completed);
>  


Signed-off-by: Eric Dumazet <edumazet@google.com>

Thanks !

(I am wondering if using a sequence would be safer, so that we can
restart the computation in case a concurrent writer changes
'num_queued')