diff mbox series

nfc: pn533: initialize struct pn533_out_arg properly

Message ID 20230306214838.237801-1-pchelkin@ispras.ru
State New
Headers
Series nfc: pn533: initialize struct pn533_out_arg properly |

Commit Message

Fedor Pchelkin March 6, 2023, 9:48 p.m. UTC 
  struct pn533_out_arg used as a temporary context for out_urb is not
initialized properly. Its uninitialized 'phy' field can be dereferenced in
error cases inside pn533_out_complete() callback function. It causes the
following failure:

general protection fault, probably for non-canonical address 0xdffffc0000000000: 0000 [#1] PREEMPT SMP KASAN
KASAN: null-ptr-deref in range [0x0000000000000000-0x0000000000000007]
CPU: 1 PID: 0 Comm: swapper/1 Not tainted 6.2.0-rc3-next-20230110-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 10/26/2022
RIP: 0010:pn533_out_complete.cold+0x15/0x44 drivers/nfc/pn533/usb.c:441
Call Trace:
 <IRQ>
 __usb_hcd_giveback_urb+0x2b6/0x5c0 drivers/usb/core/hcd.c:1671
 usb_hcd_giveback_urb+0x384/0x430 drivers/usb/core/hcd.c:1754
 dummy_timer+0x1203/0x32d0 drivers/usb/gadget/udc/dummy_hcd.c:1988
 call_timer_fn+0x1da/0x800 kernel/time/timer.c:1700
 expire_timers+0x234/0x330 kernel/time/timer.c:1751
 __run_timers kernel/time/timer.c:2022 [inline]
 __run_timers kernel/time/timer.c:1995 [inline]
 run_timer_softirq+0x326/0x910 kernel/time/timer.c:2035
 __do_softirq+0x1fb/0xaf6 kernel/softirq.c:571
 invoke_softirq kernel/softirq.c:445 [inline]
 __irq_exit_rcu+0x123/0x180 kernel/softirq.c:650
 irq_exit_rcu+0x9/0x20 kernel/softirq.c:662
 sysvec_apic_timer_interrupt+0x97/0xc0 arch/x86/kernel/apic/apic.c:1107

Initialize the field with the pn533_usb_phy currently used.

Fixes: 9dab880d675b ("nfc: pn533: Wait for out_urb's completion in pn533_usb_send_frame()")
Reported-by: syzbot+1e608ba4217c96d1952f@syzkaller.appspotmail.com
Signed-off-by: Fedor Pchelkin <pchelkin@ispras.ru>
---
 drivers/nfc/pn533/usb.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Comments

Simon Horman March 7, 2023, 3:45 p.m. UTC | #1 
On Tue, Mar 07, 2023 at 12:48:38AM +0300, Fedor Pchelkin wrote:
> struct pn533_out_arg used as a temporary context for out_urb is not
> initialized properly. Its uninitialized 'phy' field can be dereferenced in
> error cases inside pn533_out_complete() callback function. It causes the
> following failure:
> 
> general protection fault, probably for non-canonical address 0xdffffc0000000000: 0000 [#1] PREEMPT SMP KASAN
> KASAN: null-ptr-deref in range [0x0000000000000000-0x0000000000000007]
> CPU: 1 PID: 0 Comm: swapper/1 Not tainted 6.2.0-rc3-next-20230110-syzkaller #0
> Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 10/26/2022
> RIP: 0010:pn533_out_complete.cold+0x15/0x44 drivers/nfc/pn533/usb.c:441
> Call Trace:
>  <IRQ>
>  __usb_hcd_giveback_urb+0x2b6/0x5c0 drivers/usb/core/hcd.c:1671
>  usb_hcd_giveback_urb+0x384/0x430 drivers/usb/core/hcd.c:1754
>  dummy_timer+0x1203/0x32d0 drivers/usb/gadget/udc/dummy_hcd.c:1988
>  call_timer_fn+0x1da/0x800 kernel/time/timer.c:1700
>  expire_timers+0x234/0x330 kernel/time/timer.c:1751
>  __run_timers kernel/time/timer.c:2022 [inline]
>  __run_timers kernel/time/timer.c:1995 [inline]
>  run_timer_softirq+0x326/0x910 kernel/time/timer.c:2035
>  __do_softirq+0x1fb/0xaf6 kernel/softirq.c:571
>  invoke_softirq kernel/softirq.c:445 [inline]
>  __irq_exit_rcu+0x123/0x180 kernel/softirq.c:650
>  irq_exit_rcu+0x9/0x20 kernel/softirq.c:662
>  sysvec_apic_timer_interrupt+0x97/0xc0 arch/x86/kernel/apic/apic.c:1107
> 
> Initialize the field with the pn533_usb_phy currently used.
> 
> Fixes: 9dab880d675b ("nfc: pn533: Wait for out_urb's completion in pn533_usb_send_frame()")
> Reported-by: syzbot+1e608ba4217c96d1952f@syzkaller.appspotmail.com
> Signed-off-by: Fedor Pchelkin <pchelkin@ispras.ru>
> ---
>  drivers/nfc/pn533/usb.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/drivers/nfc/pn533/usb.c b/drivers/nfc/pn533/usb.c
> index ed9c5e2cf3ad..159b331b3bf6 100644
> --- a/drivers/nfc/pn533/usb.c
> +++ b/drivers/nfc/pn533/usb.c
> @@ -162,7 +162,7 @@ static int pn533_usb_send_frame(struct pn533 *dev,
>  				struct sk_buff *out)
>  {
>  	struct pn533_usb_phy *phy = dev->phy;
> -	struct pn533_out_arg arg;
> +	struct pn533_out_arg arg = {.phy = phy};

nit: This doesn't follow reverse xmas tree ordering - longest to shortest line.
     It's probably not worth respinning, but I expect the preferred
     approach is (*completely untested!*)

	...
	struct pn533_out_arg arg;
	...

	arg.phy = phy;

>  	void *cntx;
>  	int rc;
>  
> -- 
> 2.34.1
>
Krzysztof Kozlowski March 8, 2023, 10:22 a.m. UTC | #2 
On 06/03/2023 22:48, Fedor Pchelkin wrote:
> struct pn533_out_arg used as a temporary context for out_urb is not
> initialized properly. Its uninitialized 'phy' field can be dereferenced in
> error cases inside pn533_out_complete() callback function. It causes the
> following failure:
> 


Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>

Best regards,
Krzysztof
Jakub Kicinski March 9, 2023, 2:30 a.m. UTC | #3 
On Tue, 7 Mar 2023 16:45:30 +0100 Simon Horman wrote:
> nit: This doesn't follow reverse xmas tree ordering - longest to shortest line.
>      It's probably not worth respinning, but I expect the preferred
>      approach is (*completely untested!*)
> 
> 	...
> 	struct pn533_out_arg arg;
> 	...
> 
> 	arg.phy = phy;

Let's do it this way.

And please make sure that the patch is based on top of the netdev/net
tree from git.kernel.org, looks like this version doesn't apply cleanly.
Fedor Pchelkin March 9, 2023, 4:10 p.m. UTC | #4 
On Tue, Mar 07, 2023 at 04:45:30PM +0100, Simon Horman wrote:
> 
> nit: This doesn't follow reverse xmas tree ordering - longest to shortest line.
>      It's probably not worth respinning, but I expect the preferred
>      approach is (*completely untested!*)
> 
> 	...
> 	struct pn533_out_arg arg;
> 	...
> 
> 	arg.phy = phy;
> 

That is much prettier, thanks for advice!

BTW, is reverse xmas tree ordering considered to be a general
recommendation for all new kernel patches?
Simon Horman March 9, 2023, 4:41 p.m. UTC | #5 
On Thu, Mar 09, 2023 at 07:10:25PM +0300, Fedor Pchelkin wrote:
> On Tue, Mar 07, 2023 at 04:45:30PM +0100, Simon Horman wrote:
> > 
> > nit: This doesn't follow reverse xmas tree ordering - longest to shortest line.
> >      It's probably not worth respinning, but I expect the preferred
> >      approach is (*completely untested!*)
> > 
> > 	...
> > 	struct pn533_out_arg arg;
> > 	...
> > 
> > 	arg.phy = phy;
> > 
> 
> That is much prettier, thanks for advice!
> 
> BTW, is reverse xmas tree ordering considered to be a general
> recommendation for all new kernel patches?

I think that the situation is that some subsystem maintainers require it,
while other's do not. So no.
diff mbox series

Patch

diff --git a/drivers/nfc/pn533/usb.c b/drivers/nfc/pn533/usb.c
index ed9c5e2cf3ad..159b331b3bf6 100644
--- a/drivers/nfc/pn533/usb.c
+++ b/drivers/nfc/pn533/usb.c
@@ -162,7 +162,7 @@  static int pn533_usb_send_frame(struct pn533 *dev,
 				struct sk_buff *out)
 {
 	struct pn533_usb_phy *phy = dev->phy;
-	struct pn533_out_arg arg;
+	struct pn533_out_arg arg = {.phy = phy};
 	void *cntx;
 	int rc;