Skip to contentSkip to conversation

2.6.18-mm2 -- EIP: [<c11a962e>] klist_node_init+0x2b/0x3a SS:ESP 0068:f63a5f80

DISCUSSION Miles Lane 3 participants 9 messages Sep 29, 2006

lore
Share & tools
lore

Miles Lane

author
#0 lore raw

eth1: RealTek RTL8139 at 0xf9076800, 00:c0:9f:95:18:1b, IRQ 19
eth1: Identified 8139 chip type 'RTL-8100B/8139D'

BUG: unable to handle kernel NULL pointer dereference at virtual

address 000000d0
printing eip:
c11a962e
*pde = 00000000

Oops: 0000 [#1]
PREEMPT
last sysfs file: /class/firmware/0000:01:06.0/loading
Modules linked in: shpchp pci_hotplug intel_agp i2c_i801 agpgart

snd_intel8x0 i2c_core snd_intel8x0m snd_ac97_codec snd_ac97_bus snd_seq_dummy snd_seq_oss snd_seq_midi_event snd_seq snd_seq_device snd_pcm_oss snd_mixer_oss ata_generic ata_piix libata 8139too sdhci scsi_mod snd_pcm snd_timer psmouse snd soundcore snd_page_alloc 8139cp mii yenta_socket rsrc_nonstatic pcmcia_core ohci1394 serio_raw ipw2200 ieee1394 ide_cd cdrom rtc unix ehci_hcd ohci_hcd uhci_hcd usbcore ext3 jbd mbcache

CPU:    0
EIP:    0060:[<c11a962e>]    Not tainted VLI
EFLAGS: 00010296   (2.6.18-mm2 #8)

EIP is at klist_node_init+0x2b/0x3a
eax: dff16b08 ebx: 000000a0 ecx: c102ebd7 edx: f63a5f44
esi: dff16afc edi: f910e214 ebp: f63a5f88 esp: f63a5f80
ds: 007b es: 007b ss: 0068
Process probe-0000:01:0 (pid: 1697, ti=f63a4000 task=f639c030 task.ti=f63a4000)

Stack: 000000a0 dff16afc f63a5f98 c11a964f dff16a80 dff16afc f63a5fac c1124295
      00000000 dff16a80 f910e214 f63a5fc4 c1124338 f5c17e80 f5c17e80 f633bd90
      c11242f0 f63a5fe0 c102a834 ffffffff ffffffff c102a784 00000000 00000000
Call Trace:
[<c11a964f>] klist_add_tail+0x12/0x38
[<c1124295>] device_bind_driver+0x45/0xa0
[<c1124338>] really_probe+0x48/0xb3
[<c102a834>] kthread+0xb0/0xdc
[<c1003abb>] kernel_thread_helper+0x7/0x10
DWARF2 unwinder stuck at kernel_thread_helper+0x7/0x10

Leftover inexact backtrace:

[<c1003f02>] show_trace_log_lvl+0x12/0x25
[<c1003fa1>] show_stack_log_lvl+0x8c/0x97
[<c100412c>] show_registers+0x180/0x214
[<c1004355>] die+0x195/0x2b0
[<c10148a9>] do_page_fault+0x419/0x4e4
[<c11ac329>] error_code+0x39/0x40
[<c11a964f>] klist_add_tail+0x12/0x38
[<c1124295>] device_bind_driver+0x45/0xa0
[<c1124338>] really_probe+0x48/0xb3
[<c102a834>] kthread+0xb0/0xdc
[<c1003abb>] kernel_thread_helper+0x7/0x10
=======================
Code: 55 89 e5 56 53 89 c3 89 d6 8d 42 04 89 42 04 89 40 04 c7 42 10

00 00 00 00 8d 42 14 e8 f2 14 e8 ff 8d 46 0c e8 80 3d f1 ff 89 1e <8b> 53 30 85 d2 74 04 89 f0 ff d2 5b 5e 5d c3 55 89 e5 56 53 89
EIP: [<c11a962e>] klist_node_init+0x2b/0x3a SS:ESP 0068:f63a5f80 <6>ipw2200: Detected geography ZZM (11 802.11bg channels, 0 802.11a channels)

#1 lore raw
Quoted 10 lines · On Thu, 28 Sep 2006 19:13:23 -0700 "Miles Lane" <[email protected]> wrote:

eth1: RealTek RTL8139 at 0xf9076800, 00:c0:9f:95:18:1b, IRQ 19
eth1: Identified 8139 chip type 'RTL-8100B/8139D'

BUG: unable to handle kernel NULL pointer dereference at virtual

address 000000d0
printing eip:
c11a962e
*pde = 00000000

Oops: 0000 [#1]

PREEMPT
last sysfs file: /class/firmware/0000:01:06.0/loading

So you got to the stage of loading firmware.

Quoted 42 lines
Modules linked in: shpchp pci_hotplug intel_agp i2c_i801 agpgart

snd_intel8x0 i2c_core snd_intel8x0m snd_ac97_codec snd_ac97_bus snd_seq_dummy snd_seq_oss snd_seq_midi_event snd_seq snd_seq_device snd_pcm_oss snd_mixer_oss ata_generic ata_piix libata 8139too sdhci scsi_mod snd_pcm snd_timer psmouse snd soundcore snd_page_alloc 8139cp mii yenta_socket rsrc_nonstatic pcmcia_core ohci1394 serio_raw ipw2200 ieee1394 ide_cd cdrom rtc unix ehci_hcd ohci_hcd uhci_hcd usbcore ext3 jbd mbcache

CPU:    0
EIP:    0060:[<c11a962e>]    Not tainted VLI
EFLAGS: 00010296   (2.6.18-mm2 #8)

EIP is at klist_node_init+0x2b/0x3a
eax: dff16b08 ebx: 000000a0 ecx: c102ebd7 edx: f63a5f44
esi: dff16afc edi: f910e214 ebp: f63a5f88 esp: f63a5f80
ds: 007b es: 007b ss: 0068
Process probe-0000:01:0 (pid: 1697, ti=f63a4000 task=f639c030 task.ti=f63a4000)

Stack: 000000a0 dff16afc f63a5f98 c11a964f dff16a80 dff16afc f63a5fac c1124295
       00000000 dff16a80 f910e214 f63a5fc4 c1124338 f5c17e80 f5c17e80 f633bd90
       c11242f0 f63a5fe0 c102a834 ffffffff ffffffff c102a784 00000000 00000000
Call Trace:
 [<c11a964f>] klist_add_tail+0x12/0x38
 [<c1124295>] device_bind_driver+0x45/0xa0
 [<c1124338>] really_probe+0x48/0xb3
 [<c102a834>] kthread+0xb0/0xdc
 [<c1003abb>] kernel_thread_helper+0x7/0x10
DWARF2 unwinder stuck at kernel_thread_helper+0x7/0x10

Leftover inexact backtrace:

 [<c1003f02>] show_trace_log_lvl+0x12/0x25
 [<c1003fa1>] show_stack_log_lvl+0x8c/0x97
 [<c100412c>] show_registers+0x180/0x214
 [<c1004355>] die+0x195/0x2b0
 [<c10148a9>] do_page_fault+0x419/0x4e4
 [<c11ac329>] error_code+0x39/0x40
 [<c11a964f>] klist_add_tail+0x12/0x38
 [<c1124295>] device_bind_driver+0x45/0xa0
 [<c1124338>] really_probe+0x48/0xb3
 [<c102a834>] kthread+0xb0/0xdc
 [<c1003abb>] kernel_thread_helper+0x7/0x10
 =======================
Code: 55 89 e5 56 53 89 c3 89 d6 8d 42 04 89 42 04 89 40 04 c7 42 10
Quoted 4 lines

00 00 00 00 8d 42 14 e8 f2 14 e8 ff 8d 46 0c e8 80 3d f1 ff 89 1e <8b> 53 30 85 d2 74 04 89 f0 ff d2 5b 5e 5d c3 55 89 e5 56 53 89
EIP: [<c11a962e>] klist_node_init+0x2b/0x3a SS:ESP 0068:f63a5f80
<6>ipw2200: Detected geography ZZM (11 802.11bg channels, 0 802.11a channels)

Does setting CONFIG_PCI_MULTITHREAD_PROBE=n fix it?

author
#2 lore raw
Quoted 63 lines · Andrew Morton
Quoted 10 lines · On Thu, 28 Sep 2006 19:13:23 -0700 "Miles Lane" <[email protected]> wrote:

eth1: RealTek RTL8139 at 0xf9076800, 00:c0:9f:95:18:1b, IRQ 19
eth1: Identified 8139 chip type 'RTL-8100B/8139D'

BUG: unable to handle kernel NULL pointer dereference at virtual

address 000000d0
printing eip:
c11a962e
*pde = 00000000

Oops: 0000 [#1]

PREEMPT
last sysfs file: /class/firmware/0000:01:06.0/loading

So you got to the stage of loading firmware.

Quoted 42 lines
Modules linked in: shpchp pci_hotplug intel_agp i2c_i801 agpgart

snd_intel8x0 i2c_core snd_intel8x0m snd_ac97_codec snd_ac97_bus snd_seq_dummy snd_seq_oss snd_seq_midi_event snd_seq snd_seq_device snd_pcm_oss snd_mixer_oss ata_generic ata_piix libata 8139too sdhci scsi_mod snd_pcm snd_timer psmouse snd soundcore snd_page_alloc 8139cp mii yenta_socket rsrc_nonstatic pcmcia_core ohci1394 serio_raw ipw2200 ieee1394 ide_cd cdrom rtc unix ehci_hcd ohci_hcd uhci_hcd usbcore ext3 jbd mbcache

CPU:    0
EIP:    0060:[<c11a962e>]    Not tainted VLI
EFLAGS: 00010296   (2.6.18-mm2 #8)

EIP is at klist_node_init+0x2b/0x3a
eax: dff16b08 ebx: 000000a0 ecx: c102ebd7 edx: f63a5f44
esi: dff16afc edi: f910e214 ebp: f63a5f88 esp: f63a5f80
ds: 007b es: 007b ss: 0068
Process probe-0000:01:0 (pid: 1697, ti=f63a4000 task=f639c030 task.ti=f63a4000)

Stack: 000000a0 dff16afc f63a5f98 c11a964f dff16a80 dff16afc f63a5fac c1124295
       00000000 dff16a80 f910e214 f63a5fc4 c1124338 f5c17e80 f5c17e80 f633bd90
       c11242f0 f63a5fe0 c102a834 ffffffff ffffffff c102a784 00000000 00000000
Call Trace:
 [<c11a964f>] klist_add_tail+0x12/0x38
 [<c1124295>] device_bind_driver+0x45/0xa0
 [<c1124338>] really_probe+0x48/0xb3
 [<c102a834>] kthread+0xb0/0xdc
 [<c1003abb>] kernel_thread_helper+0x7/0x10
DWARF2 unwinder stuck at kernel_thread_helper+0x7/0x10

Leftover inexact backtrace:

 [<c1003f02>] show_trace_log_lvl+0x12/0x25
 [<c1003fa1>] show_stack_log_lvl+0x8c/0x97
 [<c100412c>] show_registers+0x180/0x214
 [<c1004355>] die+0x195/0x2b0
 [<c10148a9>] do_page_fault+0x419/0x4e4
 [<c11ac329>] error_code+0x39/0x40
 [<c11a964f>] klist_add_tail+0x12/0x38
 [<c1124295>] device_bind_driver+0x45/0xa0
 [<c1124338>] really_probe+0x48/0xb3
 [<c102a834>] kthread+0xb0/0xdc
 [<c1003abb>] kernel_thread_helper+0x7/0x10
 =======================
Code: 55 89 e5 56 53 89 c3 89 d6 8d 42 04 89 42 04 89 40 04 c7 42 10
Quoted 4 lines

00 00 00 00 8d 42 14 e8 f2 14 e8 ff 8d 46 0c e8 80 3d f1 ff 89 1e <8b> 53 30 85 d2 74 04 89 f0 ff d2 5b 5e 5d c3 55 89 e5 56 53 89
EIP: [<c11a962e>] klist_node_init+0x2b/0x3a SS:ESP 0068:f63a5f80
<6>ipw2200: Detected geography ZZM (11 802.11bg channels, 0 802.11a channels)

Quoted 1 line

Does setting CONFIG_PCI_MULTITHREAD_PROBE=n fix it?

Yes and no. The BUG no longer occurs, but either ipw2200 or NetworkManager is still not working. I am running Fedora development (rawhide). For some reason, NM still cannot establish a connection with this kernel build (this failed with 2.6.18-mm2 previously, but I thought it was related to the BUG). "iwlist scan" can corrently detect my access point. Also, I can use iwconfig and dhclient to establish a working connection. NM fails with this output in the message log (I'll try getting a more detailed debug log out of NM):

Sep 28 21:12:38 hogwarts NetworkManager: <information>  Activation
(eth0) started...
Sep 28 21:12:38 hogwarts NetworkManager: <information>  Activation
(eth0) Stage 1 of 5 (Device Prepare) scheduled...
Sep 28 21:12:38 hogwarts NetworkManager: <information>  Activation
(eth0) Stage 1 of 5 (Device Prepare) started...
Sep 28 21:12:38 hogwarts NetworkManager: <information>  Activation
(eth0) Stage 2 of 5 (Device Configure) scheduled...
Sep 28 21:12:38 hogwarts NetworkManager: <information>  Activation
(eth0) Stage 1 of 5 (Device Prepare) complete.
Sep 28 21:12:38 hogwarts NetworkManager: <information>  Activation
(eth0) Stage 2 of 5 (Device Configure) starting...
Sep 28 21:12:38 hogwarts NetworkManager: <information>  Activation
(eth0/wireless): access point 'linksys' is unencrypted, no key needed.
Sep 28 21:12:38 hogwarts NetworkManager: <information>  SUP: sending
command 'INTERFACE_ADD eth0                wext
/var/run/wpa_supplicant '
Sep 28 21:12:38 hogwarts NetworkManager: <information>  SUP: response was 'OK'
Sep 28 21:12:38 hogwarts NetworkManager: <information>  SUP: sending
command 'AP_SCAN 1'
Sep 28 21:12:38 hogwarts NetworkManager: <information>  SUP: response was 'OK'
Sep 28 21:12:38 hogwarts NetworkManager: <information>  SUP: sending
command 'ADD_NETWORK'
Sep 28 21:12:38 hogwarts NetworkManager: <information>  SUP: response was '0'
Sep 28 21:12:38 hogwarts NetworkManager: <information>  SUP: sending
command 'SET_NETWORK 0 ssid 6c696e6b737973'
Sep 28 21:12:38 hogwarts NetworkManager: <information>  SUP: response was 'OK'
Sep 28 21:12:38 hogwarts NetworkManager: <information>  SUP: sending
command 'SET_NETWORK 0 key_mgmt NONE'
Sep 28 21:12:38 hogwarts NetworkManager: <information>  SUP: response was 'OK'
Sep 28 21:12:38 hogwarts NetworkManager: <information>  SUP: sending
command 'ENABLE_NETWORK 0'
Sep 28 21:12:38 hogwarts NetworkManager: <information>  SUP: response was 'OK'
Sep 28 21:12:38 hogwarts NetworkManager: <information>  Activation
(eth0) Stage 2 of 5 (Device Configure) complete.
Sep 28 21:12:44 hogwarts NetworkManager: <information>
wpa_supplicant(3372): Trying to associate with 00:14:bf:1e:31:9b
(SSID='linksys' freq=0 ssage repeated 13 times
Sep 28 21:12:49 hogwarts NetworkManager: <information>
wpa_supplicant(3372): :bf:1e:31:9b (SSID='linksys' freq=0 MHz)
Sep 28 21:12:49 hogwarts NetworkManager: <information>
wpa_supplicant(3372): Trying to associate with 00:14:bf:1e:31:9b
(SSID='linksys' freq=0 MHz)
Sep 28 21:12:49 hogwarts last message repeated 12 times
Sep 28 21:12:54 hogwarts NetworkManager: <information>
wpa_supplicant(3372): req=0 MHz)
Sep 28 21:12:54 hogwarts NetworkManager: <information>
wpa_supplicant(3372): Trying to associate with 00:14:bf:1e:31:9b
(SSID='linksys' freq=0 MHz)
Sep 28 21:12:54 hogwarts last message repeated 13 times
Sep 28 21:12:58 hogwarts NetworkManager: <information>  Activation
(eth0/wireless): association took too long (>20s), failing activation.
Sep 28 21:12:58 hogwarts NetworkManager: <information>  Activation
(eth0) failure scheduled...
Sep 28 21:12:58 hogwarts NetworkManager: <information>  Activation

(eth0) failed for access point (linksys)

#3 lore raw
Quoted 64 lines · On Thu, 28 Sep 2006 21:31:19 -0700 "Miles Lane" <[email protected]> wrote:
Quoted 63 lines · Andrew Morton
Quoted 10 lines · On Thu, 28 Sep 2006 19:13:23 -0700 "Miles Lane" <[email protected]> wrote:

eth1: RealTek RTL8139 at 0xf9076800, 00:c0:9f:95:18:1b, IRQ 19
eth1: Identified 8139 chip type 'RTL-8100B/8139D'

BUG: unable to handle kernel NULL pointer dereference at virtual

address 000000d0
printing eip:
c11a962e
*pde = 00000000

Oops: 0000 [#1]

PREEMPT
last sysfs file: /class/firmware/0000:01:06.0/loading

So you got to the stage of loading firmware.

Quoted 46 lines
Modules linked in: shpchp pci_hotplug intel_agp i2c_i801 agpgart

snd_intel8x0 i2c_core snd_intel8x0m snd_ac97_codec snd_ac97_bus snd_seq_dummy snd_seq_oss snd_seq_midi_event snd_seq snd_seq_device snd_pcm_oss snd_mixer_oss ata_generic ata_piix libata 8139too sdhci scsi_mod snd_pcm snd_timer psmouse snd soundcore snd_page_alloc 8139cp mii yenta_socket rsrc_nonstatic pcmcia_core ohci1394 serio_raw ipw2200 ieee1394 ide_cd cdrom rtc unix ehci_hcd ohci_hcd uhci_hcd usbcore ext3 jbd mbcache

CPU:    0
EIP:    0060:[<c11a962e>]    Not tainted VLI
EFLAGS: 00010296   (2.6.18-mm2 #8)

EIP is at klist_node_init+0x2b/0x3a
eax: dff16b08 ebx: 000000a0 ecx: c102ebd7 edx: f63a5f44
esi: dff16afc edi: f910e214 ebp: f63a5f88 esp: f63a5f80
ds: 007b es: 007b ss: 0068
Process probe-0000:01:0 (pid: 1697, ti=f63a4000 task=f639c030 task.ti=f63a4000)

Stack: 000000a0 dff16afc f63a5f98 c11a964f dff16a80 dff16afc f63a5fac c1124295
       00000000 dff16a80 f910e214 f63a5fc4 c1124338 f5c17e80 f5c17e80 f633bd90
       c11242f0 f63a5fe0 c102a834 ffffffff ffffffff c102a784 00000000 00000000
Call Trace:
 [<c11a964f>] klist_add_tail+0x12/0x38
 [<c1124295>] device_bind_driver+0x45/0xa0
 [<c1124338>] really_probe+0x48/0xb3
 [<c102a834>] kthread+0xb0/0xdc
 [<c1003abb>] kernel_thread_helper+0x7/0x10
DWARF2 unwinder stuck at kernel_thread_helper+0x7/0x10

Leftover inexact backtrace:

 [<c1003f02>] show_trace_log_lvl+0x12/0x25
 [<c1003fa1>] show_stack_log_lvl+0x8c/0x97
 [<c100412c>] show_registers+0x180/0x214
 [<c1004355>] die+0x195/0x2b0
 [<c10148a9>] do_page_fault+0x419/0x4e4
 [<c11ac329>] error_code+0x39/0x40
 [<c11a964f>] klist_add_tail+0x12/0x38
 [<c1124295>] device_bind_driver+0x45/0xa0
 [<c1124338>] really_probe+0x48/0xb3
 [<c102a834>] kthread+0xb0/0xdc
 [<c1003abb>] kernel_thread_helper+0x7/0x10
 =======================
Code: 55 89 e5 56 53 89 c3 89 d6 8d 42 04 89 42 04 89 40 04 c7 42 10

00 00 00 00 8d 42 14 e8 f2 14 e8 ff 8d 46 0c e8 80 3d f1 ff 89 1e <8b> 53 30 85 d2 74 04 89 f0 ff d2 5b 5e 5d c3 55 89 e5 56 53 89
EIP: [<c11a962e>] klist_node_init+0x2b/0x3a SS:ESP 0068:f63a5f80
<6>ipw2200: Detected geography ZZM (11 802.11bg channels, 0 802.11a channels)

Quoted 3 lines
Quoted 1 line

Does setting CONFIG_PCI_MULTITHREAD_PROBE=n fix it?

Yes and no. The BUG no longer occurs,

OK, thanks. Note to Greg: CONFIG_PCI_MULTITHREAD_PROBE+ipw2200 = oops.

Quoted 8 lines

but either ipw2200 or
NetworkManager is still not working. I am running Fedora development (rawhide). For some reason, NM still cannot establish a connection with this kernel build (this failed with 2.6.18-mm2 previously, but I thought it was related to the BUG). "iwlist scan" can corrently detect my access point. Also, I can use iwconfig and dhclient to establish a working connection. NM fails with this output in the message log (I'll try getting a more detailed debug log out of NM):

Another reporter mentioned that he could only get his wireless to associate with his AP by using essid "any". But that was a different type of card.

ipw2200 works OK for me, fwiw.

There's a largeish ipw2200 update in mainline now. That'd be worth testing.

It looks like that ipw2200 update was in 2.6.18-mm1 though. Did that work?

author
#4 lore raw
Quoted 82 lines · Andrew Morton
Quoted 64 lines · On Thu, 28 Sep 2006 21:31:19 -0700 "Miles Lane" <[email protected]> wrote:
Quoted 63 lines · Andrew Morton
Quoted 10 lines · On Thu, 28 Sep 2006 19:13:23 -0700 "Miles Lane" <[email protected]> wrote:

eth1: RealTek RTL8139 at 0xf9076800, 00:c0:9f:95:18:1b, IRQ 19
eth1: Identified 8139 chip type 'RTL-8100B/8139D'

BUG: unable to handle kernel NULL pointer dereference at virtual

address 000000d0
printing eip:
c11a962e
*pde = 00000000

Oops: 0000 [#1]

PREEMPT
last sysfs file: /class/firmware/0000:01:06.0/loading

So you got to the stage of loading firmware.

Quoted 46 lines
Modules linked in: shpchp pci_hotplug intel_agp i2c_i801 agpgart

snd_intel8x0 i2c_core snd_intel8x0m snd_ac97_codec snd_ac97_bus snd_seq_dummy snd_seq_oss snd_seq_midi_event snd_seq snd_seq_device snd_pcm_oss snd_mixer_oss ata_generic ata_piix libata 8139too sdhci scsi_mod snd_pcm snd_timer psmouse snd soundcore snd_page_alloc 8139cp mii yenta_socket rsrc_nonstatic pcmcia_core ohci1394 serio_raw ipw2200 ieee1394 ide_cd cdrom rtc unix ehci_hcd ohci_hcd uhci_hcd usbcore ext3 jbd mbcache

CPU:    0
EIP:    0060:[<c11a962e>]    Not tainted VLI
EFLAGS: 00010296   (2.6.18-mm2 #8)

EIP is at klist_node_init+0x2b/0x3a
eax: dff16b08 ebx: 000000a0 ecx: c102ebd7 edx: f63a5f44
esi: dff16afc edi: f910e214 ebp: f63a5f88 esp: f63a5f80
ds: 007b es: 007b ss: 0068
Process probe-0000:01:0 (pid: 1697, ti=f63a4000 task=f639c030 task.ti=f63a4000)

Stack: 000000a0 dff16afc f63a5f98 c11a964f dff16a80 dff16afc f63a5fac c1124295
       00000000 dff16a80 f910e214 f63a5fc4 c1124338 f5c17e80 f5c17e80 f633bd90
       c11242f0 f63a5fe0 c102a834 ffffffff ffffffff c102a784 00000000 00000000
Call Trace:
 [<c11a964f>] klist_add_tail+0x12/0x38
 [<c1124295>] device_bind_driver+0x45/0xa0
 [<c1124338>] really_probe+0x48/0xb3
 [<c102a834>] kthread+0xb0/0xdc
 [<c1003abb>] kernel_thread_helper+0x7/0x10
DWARF2 unwinder stuck at kernel_thread_helper+0x7/0x10

Leftover inexact backtrace:

 [<c1003f02>] show_trace_log_lvl+0x12/0x25
 [<c1003fa1>] show_stack_log_lvl+0x8c/0x97
 [<c100412c>] show_registers+0x180/0x214
 [<c1004355>] die+0x195/0x2b0
 [<c10148a9>] do_page_fault+0x419/0x4e4
 [<c11ac329>] error_code+0x39/0x40
 [<c11a964f>] klist_add_tail+0x12/0x38
 [<c1124295>] device_bind_driver+0x45/0xa0
 [<c1124338>] really_probe+0x48/0xb3
 [<c102a834>] kthread+0xb0/0xdc
 [<c1003abb>] kernel_thread_helper+0x7/0x10
 =======================
Code: 55 89 e5 56 53 89 c3 89 d6 8d 42 04 89 42 04 89 40 04 c7 42 10

00 00 00 00 8d 42 14 e8 f2 14 e8 ff 8d 46 0c e8 80 3d f1 ff 89 1e <8b> 53 30 85 d2 74 04 89 f0 ff d2 5b 5e 5d c3 55 89 e5 56 53 89
EIP: [<c11a962e>] klist_node_init+0x2b/0x3a SS:ESP 0068:f63a5f80
<6>ipw2200: Detected geography ZZM (11 802.11bg channels, 0 802.11a channels)

Quoted 3 lines
Quoted 1 line

Does setting CONFIG_PCI_MULTITHREAD_PROBE=n fix it?

Yes and no. The BUG no longer occurs,

OK, thanks. Note to Greg: CONFIG_PCI_MULTITHREAD_PROBE+ipw2200 = oops.

Quoted 8 lines

but either ipw2200 or
NetworkManager is still not working. I am running Fedora development (rawhide). For some reason, NM still cannot establish a connection with this kernel build (this failed with 2.6.18-mm2 previously, but I thought it was related to the BUG). "iwlist scan" can corrently detect my access point. Also, I can use iwconfig and dhclient to establish a working connection. NM fails with this output in the message log (I'll try getting a more detailed debug log out of NM):

Quoted 4 lines

Another reporter mentioned that he could only get his wireless to associate with his AP by using essid "any". But that was a different type of card.

ipw2200 works OK for me, fwiw.

Shall I send you my .config?

Quoted 3 lines

There's a largeish ipw2200 update in mainline now. That'd be worth testing.

It looks like that ipw2200 update was in 2.6.18-mm1 though. Did that work?

Yes, it did. My main problem with mm1 was the broken CPUFreq, which is still busted in mm2. I am rebuilding with debugging enabled in the ipw2200 driver. I'll see if I can capture any useful information when I try to associate using NM.

Miles

#5 lore raw
Quoted 9 lines · On Thu, 28 Sep 2006 22:04:00 -0700 "Miles Lane" <[email protected]> wrote:
Quoted 9 lines
Quoted 8 lines

but either ipw2200 or
NetworkManager is still not working. I am running Fedora development (rawhide). For some reason, NM still cannot establish a connection with this kernel build (this failed with 2.6.18-mm2 previously, but I thought it was related to the BUG). "iwlist scan" can corrently detect my access point. Also, I can use iwconfig and dhclient to establish a working connection. NM fails with this output in the message log (I'll try getting a more detailed debug log out of NM):

Quoted 6 lines
Quoted 4 lines

Another reporter mentioned that he could only get his wireless to associate with his AP by using essid "any". But that was a different type of card.

ipw2200 works OK for me, fwiw.

Shall I send you my .config?

argh. I'm afraid my head pointer keeps on wrapping past my tail pointer recently.

But what the heck, yes please. I'm sure James will appreciate it ;)

Quoted 6 lines
Quoted 3 lines

There's a largeish ipw2200 update in mainline now. That'd be worth testing.

It looks like that ipw2200 update was in 2.6.18-mm1 though. Did that work?

Yes, it did. My main problem with mm1 was the broken CPUFreq, which is still busted in mm2.

Oh. Is anyone looking into that?

Quoted 3 lines

I am rebuilding with debugging enabled in the
ipw2200 driver. I'll see if I can capture any useful information when I try to associate using NM.

Testing mainline would be useful.

author
#6 lore raw
Quoted 15 lines · Andrew Morton
Quoted 12 lines · On Thu, 28 Sep 2006 22:04:00 -0700 "Miles Lane" <[email protected]> wrote:
Quoted 12 lines
Quoted 8 lines

but either ipw2200 or
NetworkManager is still not working. I am running Fedora development (rawhide). For some reason, NM still cannot establish a connection with this kernel build (this failed with 2.6.18-mm2 previously, but I thought it was related to the BUG). "iwlist scan" can corrently detect my access point. Also, I can use iwconfig and dhclient to establish a working connection. NM fails with this output in the message log (I'll try getting a more detailed debug log out of NM):

Another reporter mentioned that he could only get his wireless to associate with his AP by using essid "any". But that was a different type of card.

Quoted 8 lines
Quoted 3 lines
Quoted 1 line

ipw2200 works OK for me, fwiw.

Shall I send you my .config?

argh. I'm afraid my head pointer keeps on wrapping past my tail pointer recently.

But what the heck, yes please. I'm sure James will appreciate it ;)

Okay, I'll send it directly.

Quoted 8 lines
Quoted 6 lines
Quoted 3 lines

There's a largeish ipw2200 update in mainline now. That'd be worth testing.

It looks like that ipw2200 update was in 2.6.18-mm1 though. Did that work?

Yes, it did. My main problem with mm1 was the broken CPUFreq, which is still busted in mm2.

Oh. Is anyone looking into that?

I asked David Jones about it. He said he is in a push to get FC6 out the door and there are more pressing TODO items.

Quoted 5 lines
Quoted 3 lines

I am rebuilding with debugging enabled in the
ipw2200 driver. I'll see if I can capture any useful information when I try to associate using NM.

Testing mainline would be useful.

Will do.

Miles

author
#7 lore raw
In reply to Andrew Morton
Quoted 31 lines · Andrew Morton
Quoted 9 lines · On Thu, 28 Sep 2006 22:04:00 -0700 "Miles Lane" <[email protected]> wrote:
Quoted 9 lines
Quoted 8 lines

but either ipw2200 or
NetworkManager is still not working. I am running Fedora development (rawhide). For some reason, NM still cannot establish a connection with this kernel build (this failed with 2.6.18-mm2 previously, but I thought it was related to the BUG). "iwlist scan" can corrently detect my access point. Also, I can use iwconfig and dhclient to establish a working connection. NM fails with this output in the message log (I'll try getting a more detailed debug log out of NM):

Quoted 6 lines
Quoted 4 lines

Another reporter mentioned that he could only get his wireless to associate with his AP by using essid "any". But that was a different type of card.

ipw2200 works OK for me, fwiw.

Shall I send you my .config?

argh. I'm afraid my head pointer keeps on wrapping past my tail pointer recently.

But what the heck, yes please. I'm sure James will appreciate it ;)

Quoted 6 lines
Quoted 3 lines

There's a largeish ipw2200 update in mainline now. That'd be worth testing.

It looks like that ipw2200 update was in 2.6.18-mm1 though. Did that work?

Yes, it did. My main problem with mm1 was the broken CPUFreq, which is still busted in mm2.

Quoted 7 lines

Oh. Is anyone looking into that?

Quoted 3 lines

I am rebuilding with debugging enabled in the
ipw2200 driver. I'll see if I can capture any useful information when I try to associate using NM.

Testing mainline would be useful.

Bad news,
I tested 2.6.18-git11 tonight. Both Network Manager and CPUFreq (Speedstep) are busted here as well. I think I'll start trying changing some build options to try to narrow it down.

Miles

#8 lore raw
In reply to Andrew Morton
Quoted 7 lines · Andrew Morton
Quoted 3 lines · On Thu, 28 Sep 2006 21:31:19 -0700 "Miles Lane" <[email protected]> wrote:
Quoted 1 line

Does setting CONFIG_PCI_MULTITHREAD_PROBE=n fix it?

Yes and no. The BUG no longer occurs,

OK, thanks. Note to Greg: CONFIG_PCI_MULTITHREAD_PROBE+ipw2200 = oops.

Odd, it works for me here.

Hm, but that's not at boot time, I load the module at udev init time.

Miles, are you using this as a module or built into the kernel?

Hm, I'm also running it on a single proc machine, which really does not show any race conditions in the pci probe logic very well.

Anyone have a dual-core laptop they want to donate to the effort? :)

thanks,

greg k-h