* [PATCH 3/3] bunch of drivers: avoid BUG at net/core/dev.c:1328
@ 2008-07-21 15:17 Anton Vorontsov
2008-07-21 15:32 ` David Miller
0 siblings, 1 reply; 2+ messages in thread
From: Anton Vorontsov @ 2008-07-21 15:17 UTC (permalink / raw)
To: Jeff Garzik, David Miller; +Cc: netdev
The drivers were touching net queue before it has been started, so without
this patch, the drivers will potentially BUG at net/core/dev.c:1328.
I don't have the hardware for the drivers below, so this patch is untested,
and thus should be carefully peer reviewed.
tc35815.c
au1000_eth.c
bfin_mac.c
macb.c
^ The four drivers are using phylib, they're calling netif_start_queue()
in open() callback. So trivially remove netif_tx_schedule_all().
Phylib will handle netif_carrier_*().
cpmac.c
fec_mpc52xx.c
fs_enet/fs_enet-main.c
sh_eth.c
^ The same as above, but these were also needlessly calling
netif_carrier_*() functions. So removed queue calls and also remove
carrier calls, since phylib will handle it. fs_enet-main.c also didn't
call netif_start_queue() at open(), this is fixed now.
Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com>
---
drivers/net/au1000_eth.c | 5 ++---
drivers/net/bfin_mac.c | 1 -
drivers/net/cpmac.c | 2 --
drivers/net/fec_mpc52xx.c | 5 -----
drivers/net/fs_enet/fs_enet-main.c | 7 ++-----
drivers/net/macb.c | 4 +---
drivers/net/sh_eth.c | 5 -----
drivers/net/tc35815.c | 1 -
8 files changed, 5 insertions(+), 25 deletions(-)
diff --git a/drivers/net/au1000_eth.c b/drivers/net/au1000_eth.c
index 3ab61e4..cb8be49 100644
--- a/drivers/net/au1000_eth.c
+++ b/drivers/net/au1000_eth.c
@@ -911,9 +911,8 @@ au1000_adjust_link(struct net_device *dev)
if(phydev->link != aup->old_link) {
// link state changed
- if (phydev->link) // link went up
- netif_tx_schedule_all(dev);
- else { // link went down
+ if (!phydev->link) {
+ /* link went down */
aup->old_speed = 0;
aup->old_duplex = -1;
}
diff --git a/drivers/net/bfin_mac.c b/drivers/net/bfin_mac.c
index a6a3da8..a8ec60e 100644
--- a/drivers/net/bfin_mac.c
+++ b/drivers/net/bfin_mac.c
@@ -357,7 +357,6 @@ static void bfin_mac_adjust_link(struct net_device *dev)
if (!lp->old_link) {
new_state = 1;
lp->old_link = 1;
- netif_tx_schedule_all(dev);
}
} else if (lp->old_link) {
new_state = 1;
diff --git a/drivers/net/cpmac.c b/drivers/net/cpmac.c
index fbd4280..a7800e5 100644
--- a/drivers/net/cpmac.c
+++ b/drivers/net/cpmac.c
@@ -945,10 +945,8 @@ static void cpmac_adjust_link(struct net_device *dev)
if (!priv->oldlink) {
new_state = 1;
priv->oldlink = 1;
- netif_tx_schedule_all(dev);
}
} else if (priv->oldlink) {
- netif_tx_stop_all_queues(dev);
new_state = 1;
priv->oldlink = 0;
priv->oldspeed = 0;
diff --git a/drivers/net/fec_mpc52xx.c b/drivers/net/fec_mpc52xx.c
index ae9ecb7..4e4f683 100644
--- a/drivers/net/fec_mpc52xx.c
+++ b/drivers/net/fec_mpc52xx.c
@@ -197,9 +197,6 @@ static void mpc52xx_fec_adjust_link(struct net_device *dev)
if (priv->link == PHY_DOWN) {
new_state = 1;
priv->link = phydev->link;
- netif_tx_schedule_all(dev);
- netif_carrier_on(dev);
- netif_start_queue(dev);
}
} else if (priv->link) {
@@ -207,8 +204,6 @@ static void mpc52xx_fec_adjust_link(struct net_device *dev)
priv->link = PHY_DOWN;
priv->speed = 0;
priv->duplex = -1;
- netif_stop_queue(dev);
- netif_carrier_off(dev);
}
if (new_state && netif_msg_link(priv))
diff --git a/drivers/net/fs_enet/fs_enet-main.c b/drivers/net/fs_enet/fs_enet-main.c
index 445763e..5291188 100644
--- a/drivers/net/fs_enet/fs_enet-main.c
+++ b/drivers/net/fs_enet/fs_enet-main.c
@@ -738,9 +738,6 @@ static void generic_adjust_link(struct net_device *dev)
if (!fep->oldlink) {
new_state = 1;
fep->oldlink = 1;
- netif_tx_schedule_all(dev);
- netif_carrier_on(dev);
- netif_start_queue(dev);
}
if (new_state)
@@ -750,8 +747,6 @@ static void generic_adjust_link(struct net_device *dev)
fep->oldlink = 0;
fep->oldspeed = 0;
fep->oldduplex = -1;
- netif_carrier_off(dev);
- netif_stop_queue(dev);
}
if (new_state && netif_msg_link(fep))
@@ -826,6 +821,8 @@ static int fs_enet_open(struct net_device *dev)
}
phy_start(fep->phydev);
+ netif_start_queue(dev);
+
return 0;
}
diff --git a/drivers/net/macb.c b/drivers/net/macb.c
index 0496d16..daba82b 100644
--- a/drivers/net/macb.c
+++ b/drivers/net/macb.c
@@ -164,9 +164,7 @@ static void macb_handle_link_change(struct net_device *dev)
}
if (phydev->link != bp->link) {
- if (phydev->link)
- netif_tx_schedule_all(dev);
- else {
+ if (!phydev->link) {
bp->speed = 0;
bp->duplex = -1;
}
diff --git a/drivers/net/sh_eth.c b/drivers/net/sh_eth.c
index a4bc812..c69ba13 100644
--- a/drivers/net/sh_eth.c
+++ b/drivers/net/sh_eth.c
@@ -642,17 +642,12 @@ static void sh_eth_adjust_link(struct net_device *ndev)
| ECMR_DM, ioaddr + ECMR);
new_state = 1;
mdp->link = phydev->link;
- netif_tx_schedule_all(ndev);
- netif_carrier_on(ndev);
- netif_start_queue(ndev);
}
} else if (mdp->link) {
new_state = 1;
mdp->link = PHY_DOWN;
mdp->speed = 0;
mdp->duplex = -1;
- netif_stop_queue(ndev);
- netif_carrier_off(ndev);
}
if (new_state)
diff --git a/drivers/net/tc35815.c b/drivers/net/tc35815.c
index 41d3ac4..a645e50 100644
--- a/drivers/net/tc35815.c
+++ b/drivers/net/tc35815.c
@@ -672,7 +672,6 @@ static void tc_handle_link_change(struct net_device *dev)
if (dev->flags & IFF_PROMISC)
tc35815_set_multicast_list(dev);
#endif
- netif_tx_schedule_all(dev);
} else {
lp->speed = 0;
lp->duplex = -1;
--
1.5.5.4
^ permalink raw reply related [flat|nested] 2+ messages in thread
* Re: [PATCH 3/3] bunch of drivers: avoid BUG at net/core/dev.c:1328
2008-07-21 15:17 [PATCH 3/3] bunch of drivers: avoid BUG at net/core/dev.c:1328 Anton Vorontsov
@ 2008-07-21 15:32 ` David Miller
0 siblings, 0 replies; 2+ messages in thread
From: David Miller @ 2008-07-21 15:32 UTC (permalink / raw)
To: avorontsov; +Cc: jeff, netdev
From: Anton Vorontsov <avorontsov@ru.mvista.com>
Date: Mon, 21 Jul 2008 19:17:58 +0400
> The drivers were touching net queue before it has been started, so without
> this patch, the drivers will potentially BUG at net/core/dev.c:1328.
>
> I don't have the hardware for the drivers below, so this patch is untested,
> and thus should be carefully peer reviewed.
>
> tc35815.c
> au1000_eth.c
> bfin_mac.c
> macb.c
> ^ The four drivers are using phylib, they're calling netif_start_queue()
> in open() callback. So trivially remove netif_tx_schedule_all().
> Phylib will handle netif_carrier_*().
>
> cpmac.c
> fec_mpc52xx.c
> fs_enet/fs_enet-main.c
> sh_eth.c
> ^ The same as above, but these were also needlessly calling
> netif_carrier_*() functions. So removed queue calls and also remove
> carrier calls, since phylib will handle it. fs_enet-main.c also didn't
> call netif_start_queue() at open(), this is fixed now.
>
> Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com>
Your analysis and patch both look accurate, but I'll let this
one sit for a day so that others can have a look and review.
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2008-07-21 15:32 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2008-07-21 15:17 [PATCH 3/3] bunch of drivers: avoid BUG at net/core/dev.c:1328 Anton Vorontsov
2008-07-21 15:32 ` David Miller
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).