[PATCH] usbnet: Fix using smp_processor_id() in preemptible code warnings

[PATCH] usbnet: Fix using smp_processor_id() in preemptible code warnings

[Zqiang]

Syzbot reported the following warning:

BUG: using smp_processor_id() in preemptible [00000000] code: dhcpcd/2879
caller is usbnet_skb_return+0x74/0x490 drivers/net/usb/usbnet.c:331
CPU: 1 UID: 0 PID: 2879 Comm: dhcpcd Not tainted 6.15.0-rc4-syzkaller-00098-g615dca38c2ea #0 PREEMPT(voluntary)
Call Trace:
 <TASK>
 __dump_stack lib/dump_stack.c:94 [inline]
 dump_stack_lvl+0x16c/0x1f0 lib/dump_stack.c:120
 check_preemption_disabled+0xd0/0xe0 lib/smp_processor_id.c:49
 usbnet_skb_return+0x74/0x490 drivers/net/usb/usbnet.c:331
 usbnet_resume_rx+0x4b/0x170 drivers/net/usb/usbnet.c:708
 usbnet_change_mtu+0x1be/0x220 drivers/net/usb/usbnet.c:417
 __dev_set_mtu net/core/dev.c:9443 [inline]
 netif_set_mtu_ext+0x369/0x5c0 net/core/dev.c:9496
 netif_set_mtu+0xb0/0x160 net/core/dev.c:9520
 dev_set_mtu+0xae/0x170 net/core/dev_api.c:247
 dev_ifsioc+0xa31/0x18d0 net/core/dev_ioctl.c:572
 dev_ioctl+0x223/0x10e0 net/core/dev_ioctl.c:821
 sock_do_ioctl+0x19d/0x280 net/socket.c:1204
 sock_ioctl+0x42f/0x6a0 net/socket.c:1311
 vfs_ioctl fs/ioctl.c:51 [inline]
 __do_sys_ioctl fs/ioctl.c:906 [inline]
 __se_sys_ioctl fs/ioctl.c:892 [inline]
 __x64_sys_ioctl+0x190/0x200 fs/ioctl.c:892
 do_syscall_x64 arch/x86/entry/syscall_64.c:63 [inline]
 do_syscall_64+0xcd/0x260 arch/x86/entry/syscall_64.c:94
 entry_SYSCALL_64_after_hwframe+0x77/0x7f

The usbnet_skb_return() can be invoked in preemptible task-context,
this commit therefore use get_cpu_ptr/put_cpu_ptr() instead of
this_cpu_ptr() to get stats64 pointer.

Fixes: 43daa96b166c ("usbnet: Stop RX Q on MTU change")
Link: https://lore.kernel.org/all/681607f0.a70a0220.254cdc.001a.GAE@google.com/T/
Signed-off-by: Zqiang <qiang.zhang@linux.dev>
---
 drivers/net/usb/usbnet.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/net/usb/usbnet.c b/drivers/net/usb/usbnet.c
index 511c4154cf74..89f79a0bdc43 100644
--- a/drivers/net/usb/usbnet.c
+++ b/drivers/net/usb/usbnet.c
@@ -328,7 +328,7 @@ static void __usbnet_status_stop_force(struct usbnet *dev)
  */
 void usbnet_skb_return (struct usbnet *dev, struct sk_buff *skb)
 {
-	struct pcpu_sw_netstats *stats64 = this_cpu_ptr(dev->net->tstats);
+	struct pcpu_sw_netstats *stats64;
 	unsigned long flags;
 	int	status;
 
@@ -341,10 +341,12 @@ void usbnet_skb_return (struct usbnet *dev, struct sk_buff *skb)
 	if (skb->protocol == 0)
 		skb->protocol = eth_type_trans (skb, dev->net);
 
+	stats64 = get_cpu_ptr(dev->net->tstats);
 	flags = u64_stats_update_begin_irqsave(&stats64->syncp);
 	u64_stats_inc(&stats64->rx_packets);
 	u64_stats_add(&stats64->rx_bytes, skb->len);
 	u64_stats_update_end_irqrestore(&stats64->syncp, flags);
+	put_cpu_ptr(stats64);
 
 	netif_dbg(dev, rx_status, dev->net, "< rx, len %zu, type 0x%x\n",
 		  skb->len + sizeof (struct ethhdr), skb->protocol);
-- 
2.17.1

Re: [PATCH] usbnet: Fix using smp_processor_id() in preemptible code warnings

[Jakub Kicinski]

On Tue, 30 Sep 2025 16:46:36 +0800 Zqiang wrote:
> The usbnet_skb_return() can be invoked in preemptible task-context,
> this commit therefore use get_cpu_ptr/put_cpu_ptr() instead of
> this_cpu_ptr() to get stats64 pointer.

We also call netif_rx() which historically wanted to be in bh context.
I think that disabling bh in usbnet_resume_rx() may be a better idea.
-- 
pw-bot: cr

Re: [PATCH] usbnet: Fix using smp_processor_id() in preemptible code warnings

[Zqiang]

> 
> On Tue, 30 Sep 2025 16:46:36 +0800 Zqiang wrote:
> 
> > 
> > The usbnet_skb_return() can be invoked in preemptible task-context,
> >  this commit therefore use get_cpu_ptr/put_cpu_ptr() instead of
> >  this_cpu_ptr() to get stats64 pointer.
> > 
> We also call netif_rx() which historically wanted to be in bh context.
> I think that disabling bh in usbnet_resume_rx() may be a better idea.


Actually, the netif_rx() will disable bh internally when call from
process context.
but the skb_defer_rx_timestamp() is usually called in the bh or
interrupt context.

Anyway, I will disable bh in usbnet_resume_rx() and resend.


Thanks
Zqiang


> -- 
> pw-bot: cr
>

Re: [PATCH] usbnet: Fix using smp_processor_id() in preemptible code warnings

[Jakub Kicinski]

On Wed, 01 Oct 2025 00:39:18 +0000 Zqiang wrote:
> > We also call netif_rx() which historically wanted to be in bh context.
> > I think that disabling bh in usbnet_resume_rx() may be a better idea.  
> 
> Actually, the netif_rx() will disable bh internally when call from
> process context.

Sigh. I said "historically". Commit under Fixes is very old
so the fix needs to be backported. Look thru the git history 
you'll see that netif_rx() did not disable BH when the bad commit 
was written.

> but the skb_defer_rx_timestamp() is usually called in the bh or
> interrupt context.
> 
> Anyway, I will disable bh in usbnet_resume_rx() and resend.

If you think you have better solution, please do not hesitate 
to propose.

Also make sure you read
https://www.kernel.org/doc/html/next/process/maintainer-netdev.html
if you haven't already.

Re: [PATCH] usbnet: Fix using smp_processor_id() in preemptible code warnings

[Zqiang]

> 
> On Wed, 01 Oct 2025 00:39:18 +0000 Zqiang wrote:
> 
> > 
> > We also call netif_rx() which historically wanted to be in bh context.
> >  I think that disabling bh in usbnet_resume_rx() may be a better idea. 
> >  
> >  Actually, the netif_rx() will disable bh internally when call from
> >  process context.
> > 
> Sigh. I said "historically". Commit under Fixes is very old
> so the fix needs to be backported. Look thru the git history 
> you'll see that netif_rx() did not disable BH when the bad commit 
> was written.
> 
> > 
> > but the skb_defer_rx_timestamp() is usually called in the bh or
> >  interrupt context.
> >  
> >  Anyway, I will disable bh in usbnet_resume_rx() and resend.
> > 
> If you think you have better solution, please do not hesitate 
> to propose.

Maybe we can use netif_receive_skb() instead of skb_defer_rx_timestamp()
and netif_rcv() in the usbnet_skb_return() and disbale bh in usbnet_resume_rx().

Any thoughts ？

Thanks
Zqiang

> 
> Also make sure you read
> https://www.kernel.org/doc/html/next/process/maintainer-netdev.html
> if you haven't already.
>