[PATCH net] net: systemport: Fix 64-bit stats deadlock

Subsystems: broadcom systemport ethernet driver, networking drivers, the rest

STALE3275d

6 messages, 3 authors, 2017-09-15 · open the first message on its own page

[PATCH net] net: systemport: Fix 64-bit stats deadlock

From: Florian Fainelli <f.fainelli@gmail.com>
Date: 2017-09-12 20:15:00

We can enter a deadlock situation because there is no sufficient protection
when ndo_get_stats64() runs in process context to guard against RX or TX NAPI
contexts running in softirq, this can lead to the following lockdep splat and
actual deadlock was experienced as well with an iperf session in the background
and a while loop doing ifconfig + ethtool.

[    5.780350] ================================
[    5.784679] WARNING: inconsistent lock state
[    5.789011] 4.13.0-rc7-02179-g32fae27c725d #70 Not tainted
[    5.794561] --------------------------------
[    5.798890] inconsistent {SOFTIRQ-ON-W} -> {IN-SOFTIRQ-W} usage.
[    5.804971] swapper/0/0 [HC0[0]:SC1[1]:HE0:SE0] takes:
[    5.810175]  (&syncp->seq#2){+.?...}, at: [<c0768a28>] bcm_sysport_tx_reclaim+0x30/0x54
[    5.818327] {SOFTIRQ-ON-W} state was registered at:
[    5.823278]   bcm_sysport_get_stats64+0x17c/0x258
[    5.828053]   dev_get_stats+0x38/0xac
[    5.831776]   rtnl_fill_stats+0x30/0x118
[    5.835761]   rtnl_fill_ifinfo+0x538/0xe24
[    5.839921]   rtmsg_ifinfo_build_skb+0x6c/0xd8
[    5.844430]   rtmsg_ifinfo_event.part.5+0x14/0x44
[    5.849201]   rtmsg_ifinfo+0x20/0x28
[    5.852837]   register_netdevice+0x628/0x6b8
[    5.857171]   register_netdev+0x14/0x24
[    5.861051]   bcm_sysport_probe+0x30c/0x438
[    5.865280]   platform_drv_probe+0x50/0xb0
[    5.869418]   driver_probe_device+0x2e8/0x450
[    5.873817]   __driver_attach+0x104/0x120
[    5.877871]   bus_for_each_dev+0x7c/0xc0
[    5.881834]   bus_add_driver+0x1b0/0x270
[    5.885797]   driver_register+0x78/0xf4
[    5.889675]   do_one_initcall+0x54/0x190
[    5.893646]   kernel_init_freeable+0x144/0x1d0
[    5.898135]   kernel_init+0x8/0x110
[    5.901665]   ret_from_fork+0x14/0x2c
[    5.905363] irq event stamp: 24263
[    5.908804] hardirqs last  enabled at (24262): [<c08eecf0>] net_rx_action+0xc4/0x4e4
[    5.916624] hardirqs last disabled at (24263): [<c0a7da00>] _raw_spin_lock_irqsave+0x1c/0x98
[    5.925143] softirqs last  enabled at (24258): [<c022a7fc>] irq_enter+0x84/0x98
[    5.932524] softirqs last disabled at (24259): [<c022a918>] irq_exit+0x108/0x16c
[    5.939985]
[    5.939985] other info that might help us debug this:
[    5.946576]  Possible unsafe locking scenario:
[    5.946576]
[    5.952556]        CPU0
[    5.955031]        ----
[    5.957506]   lock(&syncp->seq#2);
[    5.960955]   <Interrupt>
[    5.963604]     lock(&syncp->seq#2);
[    5.967227]
[    5.967227]  *** DEADLOCK ***
[    5.967227]
[    5.973222] 1 lock held by swapper/0/0:
[    5.977092]  #0:  (&(&ring->lock)->rlock){..-...}, at: [<c0768a18>] bcm_sysport_tx_reclaim+0x20/0x54

So just remove the u64_stats_update_begin()/end() pair in ndo_get_stats64()
since it does not appear to be useful for anything. No inconsistency was
observed with either ifconfig or ethtool, global TX counts equal the sum of
per-queue TX counts on a 32-bit architecture.

Fixes: 10377ba7673d ("net: systemport: Support 64bit statistics")
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
---
 drivers/net/ethernet/broadcom/bcmsysport.c | 3 ---
 1 file changed, 3 deletions(-)
diff --git a/drivers/net/ethernet/broadcom/bcmsysport.c b/drivers/net/ethernet/broadcom/bcmsysport.c
index a6572b51435a..c3c53f6cd9e6 100644
--- a/drivers/net/ethernet/broadcom/bcmsysport.c
+++ b/drivers/net/ethernet/broadcom/bcmsysport.c
@@ -1735,11 +1735,8 @@ static void bcm_sysport_get_stats64(struct net_device *dev,
 		stats->tx_packets += tx_packets;
 	}
 
-	/* lockless update tx_bytes and tx_packets */
-	u64_stats_update_begin(&priv->syncp);
 	stats64->tx_bytes = stats->tx_bytes;
 	stats64->tx_packets = stats->tx_packets;
-	u64_stats_update_end(&priv->syncp);
 
 	do {
 		start = u64_stats_fetch_begin_irq(&priv->syncp);
-- 
1.9.1

Re: [PATCH net] net: systemport: Fix 64-bit stats deadlock

From: Eric Dumazet <hidden>
Date: 2017-09-12 21:38:15

On Tue, 2017-09-12 at 13:14 -0700, Florian Fainelli wrote:
We can enter a deadlock situation because there is no sufficient protection
when ndo_get_stats64() runs in process context to guard against RX or TX NAPI
contexts running in softirq, this can lead to the following lockdep splat and
actual deadlock was experienced as well with an iperf session in the background
and a while loop doing ifconfig + ethtool.
quoted hunk
So just remove the u64_stats_update_begin()/end() pair in ndo_get_stats64()
since it does not appear to be useful for anything. No inconsistency was
observed with either ifconfig or ethtool, global TX counts equal the sum of
per-queue TX counts on a 32-bit architecture.

Fixes: 10377ba7673d ("net: systemport: Support 64bit statistics")
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
---
 drivers/net/ethernet/broadcom/bcmsysport.c | 3 ---
 1 file changed, 3 deletions(-)
diff --git a/drivers/net/ethernet/broadcom/bcmsysport.c b/drivers/net/ethernet/broadcom/bcmsysport.c
index a6572b51435a..c3c53f6cd9e6 100644
--- a/drivers/net/ethernet/broadcom/bcmsysport.c
+++ b/drivers/net/ethernet/broadcom/bcmsysport.c
@@ -1735,11 +1735,8 @@ static void bcm_sysport_get_stats64(struct net_device *dev,
 		stats->tx_packets += tx_packets;
 	}
 
-	/* lockless update tx_bytes and tx_packets */
-	u64_stats_update_begin(&priv->syncp);
Yes, this u64_stats_update_begin()/u64_stats_update_end() is bogus

But why do we even write on tx_bytes/tx_packets here ??? 

Seems very wrong anyway.

(ethtool -S does not call bcm_sysport_get_stats64() to refresh them )
 	stats64->tx_bytes = stats->tx_bytes;
 	stats64->tx_packets = stats->tx_packets;
-	u64_stats_update_end(&priv->syncp);
 
 	do {
 		start = u64_stats_fetch_begin_irq(&priv->syncp);

Re: [PATCH net] net: systemport: Fix 64-bit stats deadlock

From: Florian Fainelli <f.fainelli@gmail.com>
Date: 2017-09-12 21:48:31

On 09/12/2017 02:38 PM, Eric Dumazet wrote:
On Tue, 2017-09-12 at 13:14 -0700, Florian Fainelli wrote:
quoted
We can enter a deadlock situation because there is no sufficient protection
when ndo_get_stats64() runs in process context to guard against RX or TX NAPI
contexts running in softirq, this can lead to the following lockdep splat and
actual deadlock was experienced as well with an iperf session in the background
and a while loop doing ifconfig + ethtool.
quoted
So just remove the u64_stats_update_begin()/end() pair in ndo_get_stats64()
since it does not appear to be useful for anything. No inconsistency was
observed with either ifconfig or ethtool, global TX counts equal the sum of
per-queue TX counts on a 32-bit architecture.

Fixes: 10377ba7673d ("net: systemport: Support 64bit statistics")
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
---
 drivers/net/ethernet/broadcom/bcmsysport.c | 3 ---
 1 file changed, 3 deletions(-)
diff --git a/drivers/net/ethernet/broadcom/bcmsysport.c b/drivers/net/ethernet/broadcom/bcmsysport.c
index a6572b51435a..c3c53f6cd9e6 100644
--- a/drivers/net/ethernet/broadcom/bcmsysport.c
+++ b/drivers/net/ethernet/broadcom/bcmsysport.c
@@ -1735,11 +1735,8 @@ static void bcm_sysport_get_stats64(struct net_device *dev,
 		stats->tx_packets += tx_packets;
 	}
 
-	/* lockless update tx_bytes and tx_packets */
-	u64_stats_update_begin(&priv->syncp);
Yes, this u64_stats_update_begin()/u64_stats_update_end() is bogus

But why do we even write on tx_bytes/tx_packets here ??? 
That's for the ethtool -S netdev stats copy (that's on me, I added that
in the driver initial version), so yes, not very robust...
Seems very wrong anyway.

(ethtool -S does not call bcm_sysport_get_stats64() to refresh them )
Yes that might actually be the simplest way to get this fixed.
quoted
 	stats64->tx_bytes = stats->tx_bytes;
 	stats64->tx_packets = stats->tx_packets;
-	u64_stats_update_end(&priv->syncp);
 
 	do {
 		start = u64_stats_fetch_begin_irq(&priv->syncp);

-- 
Florian

Re: [PATCH net] net: systemport: Fix 64-bit stats deadlock

From: David Miller <davem@davemloft.net>
Date: 2017-09-15 21:25:12

From: Florian Fainelli <f.fainelli@gmail.com>
Date: Tue, 12 Sep 2017 13:14:26 -0700
We can enter a deadlock situation because there is no sufficient protection
when ndo_get_stats64() runs in process context to guard against RX or TX NAPI
contexts running in softirq, this can lead to the following lockdep splat and
actual deadlock was experienced as well with an iperf session in the background
and a while loop doing ifconfig + ethtool.
 ...
So just remove the u64_stats_update_begin()/end() pair in ndo_get_stats64()
since it does not appear to be useful for anything. No inconsistency was
observed with either ifconfig or ethtool, global TX counts equal the sum of
per-queue TX counts on a 32-bit architecture.

Fixes: 10377ba7673d ("net: systemport: Support 64bit statistics")
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Applied.

Re: [PATCH net] net: systemport: Fix 64-bit stats deadlock

From: Florian Fainelli <f.fainelli@gmail.com>
Date: 2017-09-15 21:28:45

On September 15, 2017 2:25:11 PM PDT, David Miller [off-list ref] wrote:
From: Florian Fainelli <f.fainelli@gmail.com>
Date: Tue, 12 Sep 2017 13:14:26 -0700
quoted
We can enter a deadlock situation because there is no sufficient
protection
quoted
when ndo_get_stats64() runs in process context to guard against RX or
TX NAPI
quoted
contexts running in softirq, this can lead to the following lockdep
splat and
quoted
actual deadlock was experienced as well with an iperf session in the
background
quoted
and a while loop doing ifconfig + ethtool.
...
quoted
So just remove the u64_stats_update_begin()/end() pair in
ndo_get_stats64()
quoted
since it does not appear to be useful for anything. No inconsistency
was
quoted
observed with either ifconfig or ethtool, global TX counts equal the
sum of
quoted
per-queue TX counts on a 32-bit architecture.

Fixes: 10377ba7673d ("net: systemport: Support 64bit statistics")
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Applied.
FYI, there is another patch needed to ensure consistency between ethtool reported stats and netdevice stats, will submit that after some more testing. Thanks!

-- 
Florian

Re: [PATCH net] net: systemport: Fix 64-bit stats deadlock

From: David Miller <davem@davemloft.net>
Date: 2017-09-15 21:53:02

From: Florian Fainelli <f.fainelli@gmail.com>
Date: Fri, 15 Sep 2017 14:28:40 -0700
FYI, there is another patch needed to ensure consistency between
ethtool reported stats and netdevice stats, will submit that after
some more testing. Thanks!
Yes, I saw that, th anks.
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help