netdev - Re: d80211 deadlock with wpa

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] [day] [month] [year] [list]

Message-ID: <20070714152414.0d9d00ea@logostar.upir.cz>
Date:	Sat, 14 Jul 2007 15:24:14 +0200
From:	Jiri Benc <jbenc@...e.cz>
To:	Johannes Berg <johannes@...solutions.net>
Cc:	seventh guardian <seventhguardian@...il.com>,
	netdev@...r.kernel.org, linux-wireless@...r.kernel.org,
	"John W. Linville" <linville@...driver.com>
Subject: Re: d80211 deadlock with wpa_supplicant and rmmod

On Fri, 13 Jul 2007 23:45:16 +0200, Johannes Berg wrote:
> On Fri, 2007-07-13 at 13:05 +0100, seventh guardian wrote:
> 
> > Hello, sorry for the delay but I only managed to try it out today..
> > I've applied the patch and tortured the system as far as I could, and
> > no hang at all :)
> > 
> > I've read the thread and it apparently replaces one bad problem for a
> > lesser one, so I understand if it's not applied.. But at least there
> > is a place in the code to point as the culprit :)
> 
> It definitely should be applied anyway :) I guess John just hasn't
> caught up with all threads yet.

That's my fault, I missed that patch. If everybody put [PATCH] to subject
of mails containing a patch... (sigh)

> John: This is about the patch in the "mac80211/bcm43xx deadlock" thread
> from Michael, it kills some rtnl_lock()/unlock() that causes this
> deadlock.

Yes, please apply (the patch is below). Of course, we'll need a proper fix
but until it is ready this should be applied to prevent deadlocks.

Thanks,

 Jiri

---

Subject: [PATCH] mac80211: remove rtnl locking in ieee80211_sta.c
From: Michael Wu <flamingice@...rmilk.net>

The rtnl is held in ieee80211_sta.c to prevent some potential races with
userspace. Unfortunately, it also has the potential for deadlocks on
interface down. This patch removes the rtnl locking to eliminate the
deadlocks.

Signed-off-by: Michael Wu <flamingice@...rmilk.net>
---

 net/mac80211/ieee80211_sta.c |    6 ------
 1 files changed, 0 insertions(+), 6 deletions(-)

diff --git a/net/mac80211/ieee80211_sta.c b/net/mac80211/ieee80211_sta.c
index 91b545c..4862797 100644
--- a/net/mac80211/ieee80211_sta.c
+++ b/net/mac80211/ieee80211_sta.c
@@ -25,7 +25,6 @@
 #include <linux/wireless.h>
 #include <linux/random.h>
 #include <linux/etherdevice.h>
-#include <linux/rtnetlink.h>
 #include <net/iw_handler.h>
 #include <asm/types.h>
 
@@ -2073,12 +2072,9 @@ static int ieee80211_sta_config_auth(struct net_device *dev,
 	struct ieee80211_sta_bss *bss, *selected = NULL;
 	int top_rssi = 0, freq;
 
-	rtnl_lock();
-
 	if (!ifsta->auto_channel_sel && !ifsta->auto_bssid_sel &&
 	    !ifsta->auto_ssid_sel) {
 		ifsta->state = IEEE80211_AUTHENTICATE;
-		rtnl_unlock();
 		ieee80211_sta_reset_auth(dev, ifsta);
 		return 0;
 	}
@@ -2121,7 +2117,6 @@ static int ieee80211_sta_config_auth(struct net_device *dev,
 		ieee80211_sta_set_bssid(dev, selected->bssid);
 		ieee80211_rx_bss_put(dev, selected);
 		ifsta->state = IEEE80211_AUTHENTICATE;
-		rtnl_unlock();
 		ieee80211_sta_reset_auth(dev, ifsta);
 		return 0;
 	} else {
@@ -2132,7 +2127,6 @@ static int ieee80211_sta_config_auth(struct net_device *dev,
 		} else
 			ifsta->state = IEEE80211_DISABLED;
 	}
-	rtnl_unlock();
 	return -1;
 }
 

-- 
Jiri Benc
SUSE Labs
-
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html