netdev - Re: unregister_netdevice: waiting for batadv_slave

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <1731a084-79fb-4bc6-9e0b-9b17f3345c4b@I-love.SAKURA.ne.jp>
Date: Tue, 23 Sep 2025 23:45:48 +0900
From: Tetsuo Handa <penguin-kernel@...ove.SAKURA.ne.jp>
To: Marek Lindner <marek.lindner@...lbox.org>,
        Simon Wunderlich <sw@...onwunderlich.de>,
        Antonio Quartulli <antonio@...delbit.com>,
        Sven Eckelmann <sven@...fation.org>,
        "David S. Miller"
 <davem@...emloft.net>,
        Eric Dumazet <edumazet@...gle.com>, Jakub Kicinski <kuba@...nel.org>,
        Paolo Abeni <pabeni@...hat.com>, Simon Horman <horms@...nel.org>,
        b.a.t.m.a.n@...ts.open-mesh.org,
        Network Development <netdev@...r.kernel.org>,
        Linus Lüssing <linus.luessing@...3.blue>
Subject: Re: unregister_netdevice: waiting for batadv_slave_0 to become free.
 Usage count = 2

On 2025/09/22 23:09, Tetsuo Handa wrote:
> I suspect that batadv_hard_if_event_meshif() has something to do upon
> NETDEV_UNREGISTER event because batadv_hard_if_event_meshif() receives
> NETDEV_POST_INIT / NETDEV_REGISTER / NETDEV_UNREGISTER / NETDEV_PRE_UNINIT
> events when this reproducer is executed, but I don't know what to do...

With a change show bottom, the reproducer no longer triggers this problem.
But is this change correct?

Commit 9e6b5648bbc4 ("batman-adv: Fix duplicated OGMs on NETDEV_UP")
replaced batadv_iv_iface_activate() (which is called via iface.activate()
 from batadv_hardif_activate_interface()) with batadv_iv_iface_enabled()
(which is called via iface.enabled() from batadv_hardif_enable_interface()).
But that commit missed that batadv_hardif_activate_interface() is called from
both batadv_hardif_enable_interface() and batadv_hard_if_event().

Since batadv_iv_ogm_schedule_buff() updates if_status to BATADV_IF_ACTIVE
only when if_status was BATADV_IF_TO_BE_ACTIVATED, we need to call
batadv_iv_ogm_schedule_buff() from batadv_iv_ogm_schedule() from
batadv_iv_iface_enabled() via iface.enabled() with
if_status == BATADV_IF_TO_BE_ACTIVATED if we want iface.enabled() from
batadv_hardif_enable_interface() to update if_status to BATADV_IF_ACTIVE.

But when IFF_UP is not set upon creation, batadv_hardif_enable_interface()
does not call batadv_hardif_activate_interface(), which means that
if_status remains BATADV_IF_INACTIVE despite
batadv_iv_ogm_schedule_buff() is called via iface.enabled().

And when IFF_UP is set after creation, batadv_hard_if_event() calls
batadv_hardif_activate_interface(). But despite "Interface activated: %s\n"
message being printed, if_status remains BATADV_IF_TO_BE_ACTIVATED because
iface.activate == NULL due to above-mentioned commit.

Since we need to call iface.enabled() instead of iface.activate() so that
batadv_iv_ogm_schedule_buff() will update if_status to BATADV_IF_ACTIVE,
move iface.enabled() from batadv_hardif_enable_interface() to
batadv_hardif_activate_interface().

diff --git a/net/batman-adv/hard-interface.c b/net/batman-adv/hard-interface.c
index bace57e4f9a5..403785f649ff 100644
--- a/net/batman-adv/hard-interface.c
+++ b/net/batman-adv/hard-interface.c
@@ -673,6 +673,8 @@ batadv_hardif_activate_interface(struct batadv_hard_iface *hard_iface)

 	if (bat_priv->algo_ops->iface.activate)
 		bat_priv->algo_ops->iface.activate(hard_iface);
+	if (bat_priv->algo_ops->iface.enabled)
+		bat_priv->algo_ops->iface.enabled(hard_iface);

 out:
 	batadv_hardif_put(primary_if);
@@ -770,9 +772,6 @@ int batadv_hardif_enable_interface(struct batadv_hard_iface *hard_iface,

 	batadv_hardif_recalc_extra_skbroom(mesh_iface);

-	if (bat_priv->algo_ops->iface.enabled)
-		bat_priv->algo_ops->iface.enabled(hard_iface);
-
 out:
 	return 0;