On Wed, Jun 13, 2012 at 02:52:26PM -0700, Greg Kroah-Hartman wrote:
> On Tue, May 22, 2012 at 02:04:37PM +0200, Jan Safrata wrote:
> > Running following code when 'option' driver compatible usb modem is used
> >
> > int fd = open("/dev/ttyUSB3", O_RDWR);
> > if (fd < 0)
> > exit(1)
> > while (1) {
> > fd = open("/dev/ttyUSB3", O_RDWR);
> > if (fd >= 0)
> > close(fd);
> > usleep(200 * 1000);
> > }
> >
> > and disconnecting the modem results with following oops:
> >
> > option1 ttyUSB3: GSM modem (1-port) converter now disconnected from ttyUSB3
> > option 1-2.6:1.2: device disconnected
> > BUG: unable to handle kernel paging request at 2f406578
> > IP: [<f84fd4a4>] 0xf84fd4a3
> > *pdpt = 00000000366f0001 *pde = 0000000000000000
> > Oops: 0000 [#1] SMP
> > Modules linked in: eetis_ts option usb_wwan emgd(O) backlight cfbcopyarea drm_kms_helper drm cfbimgblt cfbfillrect usbserial cdc_acm snd_hda_codec_cirrus snd_hda_intel snd_hda_codec snd_pcm snd_timer snd snd_page_alloc
> >
> > Pid: 2298, comm: modemtest Tainted: G O 3.3.6-acs4+ #6 SECO 0866
> > EIP: 0060:[<f84fd4a4>] EFLAGS: 00010246 CPU: 0
> > EIP is at 0xf84fd4a4
> > EAX: 2f406576 EBX: f561d9c0 ECX: f41455a0 EDX: f41455a0
> > ESI: f401b400 EDI: 00000000 EBP: f55e7ccc ESP: f55e7ca4
> > DS: 007b ES: 007b FS: 00d8 GS: 0033 SS: 0068
> > Process modemtest (pid: 2298, ti=f55e6000 task=f6c96420 task.ti=f55e6000)
> > Stack:
> > 00000001 00000003 c152b9a0 c152b99c 00000003 00000286 f55e7cdc f401b400
> > f561d9c0 00000000 f55e7cf8 f85e7c60 00000286 f401b40c f55e7cf0 f40a7600
> > f41455a0 f561d9f8 f401b404 00000001 f401b40c f55e7d00 f85a7068 f55e7d20
> > Call Trace:
> > [<f85e7c60>] usb_wwan_dtr_rts+0x60/0xa0 [usb_wwan]
> > [<f85a7068>] 0xf85a7067
> > [<c1210593>] tty_port_close_start+0xe3/0x1c0
> > [<c10b74bc>] ? fasync_remove_entry+0x1c/0x90
> > [<c1210d46>] tty_port_close+0x16/0x50
> > [<f85a76da>] 0xf85a76d9
> > [<c120817d>] ? __tty_fasync+0x5d/0x140
> > [<c120946e>] tty_release+0x10e/0x4c0
> > [<c1210b12>] ? tty_port_open+0xa2/0xc0
> > [<f85a774a>] ? 0xf85a7749
> > [<c1209b7a>] tty_open+0x35a/0x560
> > [<c10acaf8>] chrdev_open+0x98/0x180
> > [<c10a7652>] __dentry_open.isra.16+0x1b2/0x2a0
> > [<c10a8514>] nameidata_to_filp+0x54/0x70
> > [<c10aca60>] ? cdev_put+0x20/0x20
> > [<c10b5677>] do_last.isra.47+0x267/0x730
> > [<c10b5d04>] path_openat+0x1c4/0x310
> > [<c10b5f2d>] do_filp_open+0x2d/0x80
> > [<c10c0174>] ? alloc_fd+0x34/0xe0
> > [<c10a8614>] do_sys_open+0xe4/0x1b0
> > [<c1040000>] ? posix_timer_fn+0xc0/0xd0
> > [<c10a870e>] sys_open+0x2e/0x40
> > [<c13f2c10>] sysenter_do_call+0x12/0x26
> > Code: ba f4 ff ff ff eb bc 55 89 e5 8d 64 24 d8 80 3d 5c 0b 50 f8 00 89 75 f8 89 5d f4 89 7d fc 8b 18 89 c6 8b 43 08 8b 53 4c 8b 40 04 <0f> b6 78 02 0f 85 9f 00 00 00 8b 4a 0c 89 f8 ba 01 00 00 00 e8
> > EIP: [<f84fd4a4>] 0xf84fd4a4 SS:ESP 0068:f55e7ca4
> > CR2: 000000002f406578
> > ---[ end trace 3b8de3e8ffdf7ff2 ]---
> >
> > Signed-off-by: Jan Safrata <jan.nikitenko@xxxxxxxxx>
> > ---
> > drivers/usb/serial/option.c | 6 +++++-
> > 1 files changed, 5 insertions(+), 1 deletions(-)
>
> What was the result of the discussion about this patch? I'm thinking it
> wasn't complete, right?
Yes, you're right -- this patch should not be applied.
We've identified what appears to be a bug in tty (tty_port_lower_dtr_rts
being called after tty_vhangup as well as tty_unregister_device) and are
waiting for Alan Cox to provide some input.
Either this needs to be fixed in tty or (perhaps in the mean-time) we
need to fix this for all usb-serial drivers by checking for disconnect
in usb-serial core (I posted an RFC-patch in the thread which does
this).
Thanks,
Johan
--
To unsubscribe from this list: send the line "unsubscribe linux-usb" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
