Skip to contentSkip to conversation

oops with USB serial irda adapter

DISCUSSION Chuck Ebbert 2 participants 3 messages Jun 20, 2007 – Jun 21, 2007

lore
Share & tools
lore
#0 lore raw

[https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=241598]

Oops happens here in kernel 2.6.20.11:
drivers/usb/serial/ir-usb.c, line 557:

                /* Notify the tty driver that the termios have changed. */
                port->tty->ldisc.set_termios(port->tty, NULL);

(ldisc.set_termios is NULL)

BUG: unable to handle kernel NULL pointer dereference at virtual address 00000000
 printing eip:
00000000
*pde = 00000000
Oops: 0000 [#1]
SMP 
last sysfs file: /class/usb_device/usbdev2.2/dev
Modules linked in: irda_usb ir_usb usbserial irnet ppp_generic slhc irtty_sir sir_dev ircomm_tty ircomm irda crc_ccitt fuse radeon drm ipv6 autofs4 hidp rfcomm l2cap bluetooth sunrpc nf_conntrack_ftp nf_conntrack_netbios_ns ipt_REJECT nf_conntrack_ipv4 xt_state nf_conntrack nfnetlink xt_tcpudp iptable_filter ip_tables x_tables dm_multipath video sbs i2c_ec dock button battery ac lp floppy snd_emu10k1_synth snd_emux_synth snd_seq_virmidi snd_seq_midi_emul snd_emu10k1 snd_rawmidi snd_ac97_codec ac97_bus snd_seq_dummy snd_seq_oss snd_seq_midi_event snd_seq tulip parport_pc snd_pcm_oss snd_mixer_oss parport snd_pcm pcspkr bt878 bttv snd_seq_device video_buf ir_common iTCO_wdt snd_timer compat_ioctl32 i2c_algo_bit snd_page_alloc btcx_risc iTCO_vendor_support snd_util_mem tveeprom videodev snd_hwdep v4l2_common snd v4l1_compat emu10k1_gp i2c_i801 ide_cd gameport serio_raw i2c_core soundcore cdrom dm_snapshot dm_zero dm_mirror dm_mod ata_piix libata sd_mod scsi_mod ext3 jbd ehci_hcd
 ohci_hcd uhci_hcd
CPU:    0
EIP:    0060:[<00000000>]    Not tainted VLI
EFLAGS: 00210246   (2.6.20-1.2948.fc6 #1)

EIP is at 0x0
eax: d1016800 ebx: 0000000d ecx: d1016800 edx: 00000000
esi: d1016800 edi: c0805ed8 ebp: e2512800 esp: c6197e04
ds: 007b es: 007b ss: 0068
Process irattach (pid: 6007, ti=c6197000 task=d1d053b0 task.ti=c6197000)

Stack: f0f2bca1 00200286 00200286 d101680c c05357f5 00000005 0000000b e2512800 
       c6197e50 c0805ed8 d1016800 f0f0f249 7f1c0300 01000415 00000003 00000000 
       00000000 d1016800 f0d394b0 00000005 00000000 000008bd 00000000 7f1c030b 
Call Trace:
 [<f0f2bca1>] ir_set_termios+0x1e4/0x2a3 [ir_usb]
 [<c05357f5>] set_termios+0x2fc/0x312
 [<f0f0f249>] serial_set_termios+0x83/0x89 [usbserial]
 [<f0d394b0>] irtty_open+0x5f/0x113 [irtty_sir]
 [<c0532f38>] tty_ioctl+0x86f/0xae7
 [<c0533d11>] n_tty_open+0x0/0x8a
 [<c0533cec>] n_tty_close+0x0/0x25
 [<c0533bff>] n_tty_flush_buffer+0x0/0x43
 [<c0533b04>] n_tty_chars_in_buffer+0x0/0x57
 [<c0534dfd>] read_chan+0x0/0x543
 [<c0533d9b>] write_chan+0x0/0x28f
 [<c053580b>] n_tty_ioctl+0x0/0x411
 [<c05338e7>] n_tty_set_termios+0x0/0x1c3
 [<c0534cf0>] normal_poll+0x0/0x10d
 [<c053402a>] n_tty_receive_buf+0x0/0xcc6
 [<c0533aaa>] n_tty_write_wakeup+0x0/0x27
 [<c045d846>] __handle_mm_fault+0x898/0x8ba
 [<c045f26f>] unmap_region+0xd8/0xf7
 [<c0478445>] do_ioctl+0x4d/0x63
 [<c047869f>] vfs_ioctl+0x244/0x256
 [<c04786fd>] sys_ioctl+0x4c/0x67
 [<c0403fe0>] syscall_call+0x7/0xb
 =======================
Code:  Bad EIP value.

EIP: [<00000000>] 0x0 SS:ESP 0068:c6197e04

#1 lore raw
Quoted 9 lines · On Wed, 20 Jun 2007 18:16:47 -0400 Chuck Ebbert <[email protected]> wrote:

[https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=241598]

Oops happens here in kernel 2.6.20.11:
drivers/usb/serial/ir-usb.c, line 557:

                /* Notify the tty driver that the termios have changed. */
                port->tty->ldisc.set_termios(port->tty, NULL);

(ldisc.set_termios is NULL)

Yeah the driver is full of bugs - its also broken for the way its set_termios tries to play 'guess it the properties have changed' and gets speed checks wrong, how it snoops the baud bits and some other stuff it does.

If you make it look like a real serial driver it ought to work fine. If you've got hardware I can send you test patches ?

#2 lore raw
Quoted 13 lines · Alan Cox
Quoted 9 lines · On Wed, 20 Jun 2007 18:16:47 -0400 Chuck Ebbert <[email protected]> wrote:

[https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=241598]

Oops happens here in kernel 2.6.20.11:
drivers/usb/serial/ir-usb.c, line 557:

                /* Notify the tty driver that the termios have changed. */
                port->tty->ldisc.set_termios(port->tty, NULL);

(ldisc.set_termios is NULL)

Quoted 7 lines

Yeah the driver is full of bugs - its also broken for the way its set_termios tries to play 'guess it the properties have changed' and gets speed checks wrong, how it snoops the baud bits and some other stuff it does.

If you make it look like a real serial driver it ought to work fine. If you've got hardware I can send you test patches ?

I think I can get the bug reporter to test, so go ahead and send patches.