All of lore.kernel.org
 help / color / mirror / Atom feed
* [Bug 209407] New: skb_over_panic in cdc_mbim_tx_fixup
@ 2020-09-27 15:34 bugzilla-daemon
  2020-12-31  6:05 ` [Bug 209407] " bugzilla-daemon
                   ` (6 more replies)
  0 siblings, 7 replies; 8+ messages in thread
From: bugzilla-daemon @ 2020-09-27 15:34 UTC (permalink / raw)
  To: linux-usb

https://bugzilla.kernel.org/show_bug.cgi?id=209407

            Bug ID: 209407
           Summary: skb_over_panic in cdc_mbim_tx_fixup
           Product: Drivers
           Version: 2.5
    Kernel Version: 5.8.12
          Hardware: x86-64
                OS: Linux
              Tree: Mainline
            Status: NEW
          Severity: normal
          Priority: P1
         Component: USB
          Assignee: drivers_usb@kernel-bugs.kernel.org
          Reporter: jks@iki.fi
        Regression: No

Created attachment 292681
  --> https://bugzilla.kernel.org/attachment.cgi?id=292681&action=edit
Debug information including the dmesg dump

I'm running a PCEngines APU3C4 system as a router, with a Huawei
909s-120 LTE module for the upstream connection. It has been crashing
once every few days. I found a way to cause the crash (or at least one
case of it) quite reliably by transferring a large file from a host to
AWS S3 via the router. This happened initially on Debian's kernel
4.19.132-1 but I can reproduce it on an unmodified 5.8.12. The
following dumps are all from 5.8.12.

Some lines from the dmesg dump (the full dump in the attachment along
with other debug information):

> [ 2297.135441] skbuff: skb_over_panic: text:ffffffffc09bf27a len:122 put:122
> head:0000000000000000 data:0000000000000000 tail:0x7a end:0x0 dev:<NULL>
> [ 2297.137499] BUG: unable to handle page fault for address: ffff90b01631c000
> [ 2297.167354] Call Trace:
> [ 2297.167369]  <IRQ>
> [ 2297.167409]  cdc_ncm_fill_tx_frame+0x83a/0x970 [cdc_ncm]
> [ 2297.306125]  cdc_mbim_tx_fixup+0x1d9/0x240 [cdc_mbim]
> [ 2297.311578]  usbnet_start_xmit+0x5d/0x720 [usbnet]
> [ 2297.316682]  dev_hard_start_xmit+0x93/0x1f0

cdc_ncm_fill_tx_frame+0x83a is in cdc_ncm_fill_tx_frame
(./include/linux/skbuff.h:2252):

> 2246  static inline void *skb_put_zero(struct sk_buff *skb, unsigned int len)
> 2247  {
> 2248          void *tmp = skb_put(skb, len);
> 2249
> 2250          memset(tmp, 0, len);
> 2251
> 2252          return tmp;
> 2253  }

cdc_mbim_tx_fixup+0x1d9 is in cdc_mbim_tx_fixup
(drivers/net/usb/cdc_mbim.c:292):

> 288                   }
> 289           }
> 290
> 291           spin_lock_bh(&ctx->mtx);
> 292           skb_out = cdc_ncm_fill_tx_frame(dev, skb, sign);
> 293           spin_unlock_bh(&ctx->mtx);
> 294           return skb_out;


My corresponding Debian bug report is at
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=970736

-- 
You are receiving this mail because:
You are watching the assignee of the bug.

^ permalink raw reply	[flat|nested] 8+ messages in thread
end of thread, other threads:[~2021-01-22 12:22 UTC | newest]

Thread overview: 8+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-09-27 15:34 [Bug 209407] New: skb_over_panic in cdc_mbim_tx_fixup bugzilla-daemon
2020-12-31  6:05 ` [Bug 209407] " bugzilla-daemon
2020-12-31 12:20 ` bugzilla-daemon
2021-01-01  8:29 ` bugzilla-daemon
2021-01-01  9:52 ` bugzilla-daemon
2021-01-01 16:25 ` [Bug 209407] cdc_ncm_fill_tx_frame calls memset with huge length when called from cdc_mbim_tx_fixup bugzilla-daemon
2021-01-03 15:02 ` bugzilla-daemon
2021-01-22 12:20 ` bugzilla-daemon

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.