Has there been any progress in the Intel536 one-connection-only problem? There
was some activity about IRQ conflicts and possibly this same issue, but
looking back through my recent linmodems emails I don't see it anymore.
Issue:
Since 2.6.30, the 536 driver only allows one connect. If it dies, or the user
closes it, another cannot be made. Attempting to start pppd, the only message
on failure is:
2009-07-22T15:57:49.853127-04:00 atr2 pppd[28736]: Failed to open
/dev/536ep: Bad address
Removing the module appears to work, as does re-insertion, but then on
subsequent calls to pppd:
# pppd call isp
BUG: unable to handle kernel paging request at ebffffff
IP: [<c0159906>] cdev_get+0xb/0x33
*pde = 00000000
Oops: 0000 [#1]
last sysfs file: /sys/module/Intel536/refcnt
Modules linked in: Intel536 wlan_scan_ap ath_rate_minstrel ath_pci wlan
ath_hal sha256_generic cbc dm_round_robin dm_multipath dm_snapshot dm_zero
dm_crypt dm_mirror dm_region_hash dm_log dm_mod ipt_MASQUERADE iptable_nat
nf_nat xt_comment xt_iprange xt_multiport ipt_REJECT ipt_LOG xt_limit
nf_conntrack_ipv4 nf_defrag_ipv4 xt_state nf_conntrack iptable_filter
ppp_deflate bsd_comp ppp_async ppp_generic slhc xfs exportfs twofish_i586
twofish_common aes_i586 aes_generic msr cpuid microcode camellia cryptd
des_generic serpent sha1_generic sha512_generic blowfish lp hangcheck_timer
i915 i2c_i801 pcspkr parport_pc parport rtc_cmos uhci_hcd ehci_hcd iTCO_wdt
configfs tcp_highspeed ipv6 snd_pcm_oss snd_mixer_oss snd_seq_dummy
snd_seq_oss snd_seq_midi_event snd_seq snd_seq_device snd_dummy snd_intel8x0
snd_intel8x0m snd_ac97_codec snd_pcm snd_timer snd snd_page_alloc ac97_bus
soundcore [last
unloaded: Intel536]
Pid: 28756, comm: pppd Not tainted (2.6.30.1 #1) D300
EIP: 0060:[<c0159906>] EFLAGS: 00010246 CPU: 0
EIP is at cdev_get+0xb/0x33
EAX: e4582708 EBX: ebffffff ECX: c03767b8 EDX: d7063400
ESI: 00000000 EDI: d7063400 EBP: c0159a34 ESP: dc7b0e88
DS: 007b ES: 007b FS: 0000 GS: 0033 SS: 0068
Process pppd (pid: 28756, ti=dc7b0000 task=dd42ca50 task.ti=dc7b0000)
Stack:
dcc925c0 e4582708 c0159ac5 c0167081 dc7b0f0c df0ce500 d7063400 dcc925c0
c0159a34 c01563f6 00000000 dcccfd80 dcccfd80 d7063400 00000803 dc7b0f78
c0156574 d7063400 00000000 d7063d80 dc7b0f0c 00000000 c015f32a c1866000
Call Trace:
[<c0159ac5>] ? chrdev_open+0x91/0xfc
[<c0167081>] ? mntput_no_expire+0x16/0x9b
[<c0159a34>] ? chrdev_open+0x0/0xfc
[<c01563f6>] ? __dentry_open+0x10c/0x1f5
[<c0156574>] ? nameidata_to_filp+0x29/0x3a
[<c015f32a>] ? do_filp_open+0x38b/0x623
[<c0149852>] ? do_wp_page+0x1b6/0x3dc
[<c016613a>] ? alloc_fd+0x46/0xa5
[<c01561fd>] ? do_sys_open+0x45/0xd3
[<c01562cf>] ? sys_open+0x1e/0x23
[<c01026c8>] ? sysenter_do_call+0x12/0x26
Code: 8a ec 00 00 00 89 8a e8 00 00 00 c7 82 f0 00 00 00 00 00 00 00 8b 48
2c 39 f9 75 cc 5b 5e 5f c3 56 53 8b 58 24 85 db 74 13 31 f6 <83> 3b 02 74 1e
8d 93 28 01 00 00 ff 83 28 01 00 00 e8 da 43 06
EIP: [<c0159906>] cdev_get+0xb/0x33 SS:ESP 0068:dc7b0e88
CR2: 00000000ebffffff
---[ end trace ad2097a3f9102d48 ]---
The end result is, if the link goes down, you have to reboot the entire system
to get back online!
Kern: 2.6.30.1 (but same issue as with 2.6.30)
01:01.0 Communication controller: Intel Corporation 536EP Data Fax Modem
Src package: intel-536-537_2009_06_22.tar.bz2
