On Sun, Jun 15, 2003 at 08:50:49PM +0300, Jaakko Niemi wrote:
> I seem to be able to reproduce crash with 2.7.70-bk and .71.
> First, I tried getting dlink dwl-650 wlan card up on my thinkpad
> 570e, but orinoco_cs does not seem to want to even look at it.
> (any ideas what's the deal with that, btw?)
What happens if you plug in your cardbus card before the dlink wlan card?
If that doesn't work, please repeat your procedure to cause the oops.
In either case, could you send the output of lspci -vv at the
following points:
- directly after boot
- after you insert the cardbus card
- after you remove it
- after you re-insert (and get the oops)
Thanks.
--
Russell King ([email protected]) The developer of ARM Linux
http://www.arm.linux.org.uk/personal/aboutme.html
On Sun, Jun 15, 2003 at 11:00:00PM +0300, Jaakko Niemi wrote:
> Russell King <[email protected]> writes:
> > On Sun, Jun 15, 2003 at 08:50:49PM +0300, Jaakko Niemi wrote:
> >> I seem to be able to reproduce crash with 2.7.70-bk and .71.
> >> First, I tried getting dlink dwl-650 wlan card up on my thinkpad
> >> 570e, but orinoco_cs does not seem to want to even look at it.
> >> (any ideas what's the deal with that, btw?)
> >
> > What happens if you plug in your cardbus card before the dlink wlan card?
>
> Same thing.
Ok, I'm confused. I suspect that it may be something to do with two
PCI device structures appearing for the same device, however I don't
see that happening with the code which is in 2.5.7x.
Which kernel version first showed the problem?
Which modules are you loading?
Which version of cardmgr are you using?
Are you using any modules from the pcmcia-cs package?
--
Russell King ([email protected]) The developer of ARM Linux
http://www.arm.linux.org.uk/personal/aboutme.html
Russell King <[email protected]> writes:
> On Sun, Jun 15, 2003 at 11:00:00PM +0300, Jaakko Niemi wrote:
>> Russell King <[email protected]> writes:
>> > On Sun, Jun 15, 2003 at 08:50:49PM +0300, Jaakko Niemi wrote:
>> >> I seem to be able to reproduce crash with 2.7.70-bk and .71.
>> >> First, I tried getting dlink dwl-650 wlan card up on my thinkpad
>> >> 570e, but orinoco_cs does not seem to want to even look at it.
>> >> (any ideas what's the deal with that, btw?)
>> >
>> > What happens if you plug in your cardbus card before the dlink wlan card?
>>
>> Same thing.
>
> Ok, I'm confused. I suspect that it may be something to do with two
> PCI device structures appearing for the same device, however I don't
> see that happening with the code which is in 2.5.7x.
Hmm, that second instance was not there always. I'll try to get
a crash without it.
> Which kernel version first showed the problem?
2.5.71-bk13 was the first I managed to notice this with, iirc.
If you have some older version in mind, I can try that.
> Which modules are you loading?
Only those that come by default with this config:
snd, soundcore, usbkbd and usbcore
> Which version of cardmgr are you using?
3.2.2
> Are you using any modules from the pcmcia-cs package?
No.
--j
On Mon, Jun 16, 2003 at 02:46:00AM +0300, Jaakko Niemi wrote:
> > Which kernel version first showed the problem?
>
> 2.5.71-bk13 was the first I managed to notice this with, iirc.
> If you have some older version in mind, I can try that.
Which was the latest kernel version which didn't show the problem?
There doesn't seem to be any PCI, PCMCIA or driver model changes
from 2.5.70-bk12 to 2.5.70-bk13.
There are changes in:
-bk11 (pci)
-bk10 (pci)
-bk9 (driver model)
-bk4 (pci)
-bk2 (pcmcia)
--
Russell King ([email protected]) The developer of ARM Linux
http://www.arm.linux.org.uk/personal/aboutme.html
Russell King wrote:
>
> Which was the latest kernel version which didn't show the problem?
> There doesn't seem to be any PCI, PCMCIA or driver model changes
> from 2.5.70-bk12 to 2.5.70-bk13.
>
> There are changes in:
>
> -bk11 (pci)
> -bk10 (pci)
> -bk9 (driver model)
> -bk4 (pci)
> -bk2 (pcmcia)
I get the same problems: cardbus works only if I boot with the card
inserted. If I insert the card later I get two PCI device entries
for the device, and can't get any interrupts from the card (sometimes
the driver complains about the chip not responding (8139too)).
Tested with the following versions (exactly the same behaviour):
2.5.69-bk10
2.5.70
2.5.70-bk2
2.5.70-bk5
2.5.70-bk18
2.5.71
2.5.69-bk9 was OK.
/peter
Perhaps the changes to network hotplug are related. Did you see this?
>
> This patch causes an external script API change.
> Because network device go through the standard path, the action passed to the script
> is no longer register or unregister but is now "add" or "remove" like other devices.
> This is a good thing. When testing (at least on RHAT) just change /etc/hotplug/net.agent
> case statement:
>
> case $ACTION in
> add|register)
> # Don't do anything if the network is stopped
> if [ ! -f /var/lock/subsys/network ]; then
> exit 0
> fi
On Mon, Jun 16, 2003 at 09:15:06PM +0200, Peter Lundkvist wrote:
> Tested with the following versions (exactly the same behaviour):
> 2.5.69-bk10
>
> 2.5.69-bk9 was OK.
Great, this helps a lot. While I remove the bullet from my foot, could
you test this patch please?
--- linux/drivers/pcmcia/cs.c.old Mon Jun 16 21:17:45 2003
+++ linux/drivers/pcmcia/cs.c Mon Jun 16 21:24:23 2003
@@ -817,7 +817,8 @@
if ((skt->state & SOCKET_PRESENT) &&
!(status & SS_DETECT))
socket_shutdown(skt);
- if (status & SS_DETECT)
+ if (!(skt->state & SOCKET_PRESENT) &&
+ status & SS_DETECT)
socket_insert(skt);
}
if (events & SS_BATDEAD)
--
Russell King ([email protected]) The developer of ARM Linux
http://www.arm.linux.org.uk/personal/aboutme.html
Russell King wrote:
> Great, this helps a lot. While I remove the bullet from my foot, could
> you test this patch please?
>
> --- linux/drivers/pcmcia/cs.c.old Mon Jun 16 21:17:45 2003
> +++ linux/drivers/pcmcia/cs.c Mon Jun 16 21:24:23 2003
> @@ -817,7 +817,8 @@
> if ((skt->state & SOCKET_PRESENT) &&
> !(status & SS_DETECT))
> socket_shutdown(skt);
> - if (status & SS_DETECT)
> + if (!(skt->state & SOCKET_PRESENT) &&
> + status & SS_DETECT)
> socket_insert(skt);
> }
> if (events & SS_BATDEAD)
A quick test with this patch against 2.5.71: Works OK now!
Thank you
Peter
Peter Lundkvist <[email protected]> writes:
> Russell King wrote:
>> Great, this helps a lot. While I remove the bullet from my foot, could
>> you test this patch please?
>>
>> --- linux/drivers/pcmcia/cs.c.old Mon Jun 16 21:17:45 2003
>> +++ linux/drivers/pcmcia/cs.c Mon Jun 16 21:24:23 2003
>> @@ -817,7 +817,8 @@
>> if ((skt->state & SOCKET_PRESENT) &&
>> !(status & SS_DETECT))
>> socket_shutdown(skt);
>> - if (status & SS_DETECT)
>> + if (!(skt->state & SOCKET_PRESENT) &&
>> + status & SS_DETECT)
>> socket_insert(skt);
>> }
>> if (events & SS_BATDEAD)
>
> A quick test with this patch against 2.5.71: Works OK now!
Confirmation. Booted 2.5.72 with this patch and tried swapping and switching
cards about 50 times, and things look stable.
Next, how to get my d-link dwl-650 wlan card up and running. If I insert it,
link light on it lights up, and cardctl sees it in socket. However the drivers
do not find it, and there is no interface available. This happens at least
with 2.5.70 to .72. Anyone got suggestions where to start looking?
Thanks.
--j
On Wed, Jun 18, 2003 at 10:49:54PM +0300, Jaakko Niemi wrote:
> Next, how to get my d-link dwl-650 wlan card up and running. If I insert
> it, link light on it lights up, and cardctl sees it in socket. However
> the drivers do not find it, and there is no interface available. This
> happens at least with 2.5.70 to .72. Anyone got suggestions where to
> start looking?
Anything in /var/log/messages from cardmgr?
--
Russell King ([email protected]) The developer of ARM Linux
http://www.arm.linux.org.uk/personal/aboutme.html
Russell King <[email protected]> writes:
> On Wed, Jun 18, 2003 at 10:49:54PM +0300, Jaakko Niemi wrote:
>> Next, how to get my d-link dwl-650 wlan card up and running. If I insert
>> it, link light on it lights up, and cardctl sees it in socket. However
>> the drivers do not find it, and there is no interface available. This
>> happens at least with 2.5.70 to .72. Anyone got suggestions where to
>> start looking?
>
> Anything in /var/log/messages from cardmgr?
Nothing whatsoever, only thing I get when unplugging and re-inserting
the card is:
kernel: spurious 8259A interrupt: IRQ7.
--j
Russell King <[email protected]> writes:
> On Wed, Jun 18, 2003 at 10:49:54PM +0300, Jaakko Niemi wrote:
>> Next, how to get my d-link dwl-650 wlan card up and running. If I insert
>> it, link light on it lights up, and cardctl sees it in socket. However
>> the drivers do not find it, and there is no interface available. This
>> happens at least with 2.5.70 to .72. Anyone got suggestions where to
>> start looking?
>
> Anything in /var/log/messages from cardmgr?
Solved, the card was in monitor mode after playing with Kismet.
I took it back to normal mode and everything works great now.
It would be nice improvement to the driver to see the status.
Thanks.
--j