Re: [B.A.T.M.A.N.] [RFC 2/6] batman-adv: Detect coding nodes and remove these after timeout

[Antonio Quartulli <ordex@autistici.org>]

[-- Attachment #1: Type: text/plain, Size: 1805 bytes --]

On Mon, Nov 19, 2012 at 02:22:00PM +0100, Martin Hundebøll wrote:
[...]

> diff --git a/network-coding.c b/network-coding.c
> index 8d930e7..29f7744 100644
> --- a/network-coding.c
> +++ b/network-coding.c
> @@ -17,8 +17,12 @@
>   * 02110-1301, USA
>   */
>  
> +#include <linux/debugfs.h>
> +

Ok, I don't want to make you feel a cavy :) but it is better to correct this
stuff when we can: kernel-includes (<..h>) should follow project-includes
("..h")

>  #include "main.h"
>  #include "network-coding.h"
> +#include "originator.h"
> +#include "hard-interface.h"
>  
>  /**
> @@ -68,11 +218,197 @@ static void batadv_nc_worker(struct work_struct *work)
>  	priv_nc = container_of(delayed_work, struct batadv_priv_nc, work);
>  	bat_priv = container_of(priv_nc, struct batadv_priv, nc);
>  
> +	batadv_nc_purge_orig_hash(bat_priv);
> +

ah, now this function makes sense (with respect to what I said in reply to 1/6).
Would it be better to add the worker here instead of adding it in 1/6?

>  	/* Schedule a new check */
>  	batadv_nc_start_timer(bat_priv);
>  }

[...]
> @@ -268,9 +274,12 @@ struct batadv_priv_dat {
>  /**
>   * struct batadv_priv_nc - per mesh interface network coding private data
>   * @work: work queue callback item for cleanup
> + * @min_tq: only consider neighbors for encoding if neigh_tq > min_tq
>   */
>  struct batadv_priv_nc {
>  	struct delayed_work work;
> +	struct dentry *debug_dir;
> +	u8 min_tq;

If I am not mistaken min_tq is never modified. Why do we need to store it in a
bat_priv field? And being a constant is it worth a define? In this way it an be
easily found and tuned like the other constant params.

Cheers,

-- 
Antonio Quartulli

..each of us alone is worth nothing..
Ernesto "Che" Guevara

[-- Attachment #2: Type: application/pgp-signature, Size: 198 bytes --]