netdev - [PATCH 3/5 v2] net/cpsw: don't rely only on netif_running() to check which device is active

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: <20130422083048.GA8162@linutronix.de>
Date:	Mon, 22 Apr 2013 10:30:48 +0200
From:	Sebastian Andrzej Siewior <bigeasy@...utronix.de>
To:	Mugunthan V N <mugunthanvnm@...com>
Cc:	netdev@...r.kernel.org, tglx@...utronix.de,
	"David S. Miller" <davem@...emloft.net>
Subject: [PATCH 3/5 v2] net/cpsw: don't rely only on netif_running() to
 check which device is active

netif_running() reports false before the ->ndo_stop() callback is
called. That means if one executes "ifconfig down" and the system
receives an interrupt before the interrupt source has been disabled we
hang for always for two reasons:
- we never disable the interrupt source because devices claim to be
  already inactive (or non-present) and don't feel responsible.
- since the ISR always reports IRQ_HANDLED the line is never deactivated
  because it looks like the ISR feels responsible.

This patch looks at both, netif_running() and IFF_UP to check if the
device is up. IFF_UP is set after ndo_open() completes and cleared after
ndo_close() completes and netif_running is set before ndo_open() is
called and cleared before ndo_close() completes. Using both will avoid
false positives during the bring down sequence.

Signed-off-by: Sebastian Andrzej Siewior <bigeasy@...utronix.de>
---
* Mugunthan V N | 2013-04-19 16:05:55 [+0530]:

>The same netif_running(). But why you are telling not to rely on this API?
>device state is updated immediately after successful ndo_open.

I tried to describe the reason for that in my patch description. Here 
it is again: netif_running() reports false before ndo_close() has been
called. That means an interrupt between the flag change and interrupt
disabling in ndo_close() will currently lockup the box (IRQ_NONE would
at least allow the core to disable the interrupt line).

Initialyly I decided against using IFF_UP as well but using would work
without the private active field. So here we have 

-v1…v2: replacing private active field with IFF_UP + netif_running()

How about this?

>Regards
>Mugunthan V N
 drivers/net/ethernet/ti/cpsw.c |   23 ++++++++++++++---------
 1 file changed, 14 insertions(+), 9 deletions(-)

diff --git a/drivers/net/ethernet/ti/cpsw.c b/drivers/net/ethernet/ti/cpsw.c
index 3b22a36..2705725 100644
--- a/drivers/net/ethernet/ti/cpsw.c
+++ b/drivers/net/ethernet/ti/cpsw.c
@@ -511,19 +511,24 @@ static irqreturn_t cpsw_interrupt(int irq, void *dev_id)
 {
 	struct cpsw_priv *priv = dev_id;
 
-	if (likely(netif_running(priv->ndev))) {
+	if (likely(netif_running(priv->ndev) || priv->ndev->flags & IFF_UP)) {
 		cpsw_intr_disable(priv);
 		cpsw_disable_irq(priv);
 		napi_schedule(&priv->napi);
-	} else {
-		priv = cpsw_get_slave_priv(priv, 1);
-		if (likely(priv) && likely(netif_running(priv->ndev))) {
-			cpsw_intr_disable(priv);
-			cpsw_disable_irq(priv);
-			napi_schedule(&priv->napi);
-		}
+		return IRQ_HANDLED;
+	}
+
+	priv = cpsw_get_slave_priv(priv, 1);
+	if (!priv)
+		return IRQ_NONE;
+
+	if (likely(netif_running(priv->ndev) || priv->ndev->flags & IFF_UP)) {
+		cpsw_intr_disable(priv);
+		cpsw_disable_irq(priv);
+		napi_schedule(&priv->napi);
+		return IRQ_HANDLED;
 	}
-	return IRQ_HANDLED;
+	return IRQ_NONE;
 }
 
 static int cpsw_poll(struct napi_struct *napi, int budget)
-- 
1.7.10.4

--
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