All of lore.kernel.org
 help / color / mirror / Atom feed
From: Marek Lindner <mareklindner@neomailbox.ch>
To: b.a.t.m.a.n@lists.open-mesh.org
Cc: Simon Wunderlich <simon@open-mesh.com>
Subject: Re: [B.A.T.M.A.N.] [PATCHv2] batman-adv: remove obsolete deleted attribute for gateway node
Date: Tue, 14 Jul 2015 00:05:22 +0800	[thread overview]
Message-ID: <3846505.HTcyTpK9ms@voltaire> (raw)
In-Reply-To: <1435317085-31015-1-git-send-email-sw@simonwunderlich.de>

[-- Attachment #1: Type: text/plain, Size: 2651 bytes --]

On Friday, June 26, 2015 13:11:25 Simon Wunderlich wrote:
> @@ -537,14 +530,26 @@ void batadv_gw_node_update(struct batadv_priv
> *bat_priv, /* Note: We don't need a NULL check here, since curr_gw never
>  		 * gets dereferenced.
>  		 */
> +		spin_lock_bh(&bat_priv->gw.list_lock);
> +		hlist_for_each_entry_safe(gw_node_tmp, node_tmp,
> +					  &bat_priv->gw.list, list) {
> +			if (gw_node_tmp != gw_node)
> +				continue;
> +
> +			hlist_del_rcu(&gw_node->list);
> +			batadv_gw_node_free_ref(gw_node);
> +		}
> +		spin_unlock_bh(&bat_priv->gw.list_lock);

An entire hlist_for_each_entry_safe() loop isn't necessary. 
hlist_del_init_rcu() should suffice.


>  		curr_gw = batadv_gw_get_selected_gw_node(bat_priv);
>  		if (gw_node == curr_gw)
>  			batadv_gw_reselect(bat_priv);
> +
> +		if (curr_gw)
> +			batadv_gw_node_free_ref(curr_gw);
>  	}
> 
>  out:
> -	if (curr_gw)
> -		batadv_gw_node_free_ref(curr_gw);
>  	if (gw_node)
>  		batadv_gw_node_free_ref(gw_node);
>  }

After the batadv_gw_node_free_ref() 'bat_priv->gw.curr_gw' points to random 
memory ...



> @@ -562,37 +567,21 @@ void batadv_gw_node_delete(struct batadv_priv
> *bat_priv,
> 
>  void batadv_gw_node_purge(struct batadv_priv *bat_priv)
>  {
> -	struct batadv_gw_node *gw_node, *curr_gw;
> +	struct batadv_gw_node *gw_node;
>  	struct hlist_node *node_tmp;
> -	unsigned long timeout = msecs_to_jiffies(2 * BATADV_PURGE_TIMEOUT);
> -	int do_reselect = 0;
> 
> -	curr_gw = batadv_gw_get_selected_gw_node(bat_priv);
> +	if (atomic_read(&bat_priv->mesh_state) == BATADV_MESH_ACTIVE)
> +		return;

Why checking for BATADV_MESH_ACTIVE if the function only ever gets called from 
batadv_mesh_free() which sets BATADV_MESH_DEACTIVATING ?

If you change the meaning of the function you could also rename it to 
batadv_gw_node_free() to be consistent with the other cleanup functions we 
have.



>  	spin_lock_bh(&bat_priv->gw.list_lock);
> -
>  	hlist_for_each_entry_safe(gw_node, node_tmp,
>  				  &bat_priv->gw.list, list) {
> -		if (((!gw_node->deleted) ||
> -		     (time_before(jiffies, gw_node->deleted + timeout))) &&
> -		    atomic_read(&bat_priv->mesh_state) == BATADV_MESH_ACTIVE)
> -			continue;
> -
> -		if (curr_gw == gw_node)
> -			do_reselect = 1;
> 
>  		hlist_del_rcu(&gw_node->list);
>  		batadv_gw_node_free_ref(gw_node);
>  	}
> 
>  	spin_unlock_bh(&bat_priv->gw.list_lock);
> -
> -	/* gw_reselect() needs to acquire the gw_list_lock */
> -	if (do_reselect)
> -		batadv_gw_reselect(bat_priv);
> -
> -	if (curr_gw)
> -		batadv_gw_node_free_ref(curr_gw);
>  }

At this point 'bat_priv->gw.curr_gw' points to random memory if I am not 
mistaken.

Cheers,
Marek

[-- Attachment #2: This is a digitally signed message part. --]
[-- Type: application/pgp-signature, Size: 473 bytes --]

  reply	other threads:[~2015-07-13 16:05 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-06-26 11:11 [B.A.T.M.A.N.] [PATCHv2] batman-adv: remove obsolete deleted attribute for gateway node Simon Wunderlich
2015-07-13 16:05 ` Marek Lindner [this message]
2015-07-13 16:50   ` Marek Lindner

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=3846505.HTcyTpK9ms@voltaire \
    --to=mareklindner@neomailbox.ch \
    --cc=b.a.t.m.a.n@lists.open-mesh.org \
    --cc=simon@open-mesh.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.