* Re: Hotplug *seems* to hang during boot]
2004-10-04 18:52 Hotplug *seems* to hang during boot] Richard Wild
@ 2004-10-04 19:55 ` Dave Dodge
2004-10-04 20:44 ` Richard Wild
` (2 subsequent siblings)
3 siblings, 0 replies; 5+ messages in thread
From: Dave Dodge @ 2004-10-04 19:55 UTC (permalink / raw)
To: linux-hotplug
On Mon, Oct 04, 2004 at 07:14:07PM +0000, Richard Wild wrote:
> Some other information noticed: it seems (to me) quite likely that the
> message "cardmgr[953]: no sockets found!" is related to the problem,
> because when I boot the 2.4 kernel is says instead "cardmgr[80]: no
> pcmcia driver in /proc/devices".
cardmgr is a bit strange in how it accesses devices. Here's a quick
explanation in case it's relevant to the problem.
"no pcmcia driver in /proc/devices":
That means just what it says. cardmgr wants the major device number for
the pcmcia driver, so it reads through /proc/devices looking for it.
This error message indicates that it did not find a "pcmcia" line,
presumably because no driver was loaded.
"no sockets found!":
This indicates that it _did_ find "pcmcia" listed in /proc/devices,
and it tried to access the driver. Here's where things get a bit
unusual. After obtaining the major device number from /proc/devices,
cardmgr uses mknod(2) to create device files of its own. It starts
with minor device 0 and continues in a one-up loop until it fails.
For each minor number, it creates a device file with a name of the
form "cm-PID-COUNTER", opens the file, and then immediately unlinks
the file while holding it open. There are a number of directories
where it might try to do this, usually in this order:
/var/lib/pcmcia, /var/run, /dev, /tmp
The "no sockets found" message indicates that it managed to do the
mknod for minor device 0, but got an ENODEV when it then tried to open
the resulting file.
So what this seems to be saying is that:
- your 2.4 kernel has no PCMCIA stuff loaded at all.
- your 2.6 kernel has at least the "ds" driver, which
creates the "pcmcia" /proc/devices entry, but it has no
card sockets.
I don't know if this is part of your problem or not, but since cardmgr
does do some device file manipulation, it might be worth investigating.
-Dave Dodge
-------------------------------------------------------
This SF.net email is sponsored by: IT Product Guide on ITManagersJournal
Use IT products in your business? Tell us what you think of them. Give us
Your Opinions, Get Free ThinkGeek Gift Certificates! Click to find out more
http://productguide.itmanagersjournal.com/guidepromo.tmpl
_______________________________________________
Linux-hotplug-devel mailing list http://linux-hotplug.sourceforge.net
Linux-hotplug-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel
^ permalink raw reply [flat|nested] 5+ messages in thread* Re: Hotplug *seems* to hang during boot]
2004-10-04 18:52 Hotplug *seems* to hang during boot] Richard Wild
2004-10-04 19:55 ` Dave Dodge
@ 2004-10-04 20:44 ` Richard Wild
2004-10-04 21:16 ` Simone Gotti
2004-10-04 22:35 ` Richard Wild
3 siblings, 0 replies; 5+ messages in thread
From: Richard Wild @ 2004-10-04 20:44 UTC (permalink / raw)
To: linux-hotplug
Hi Dave,
thanks for your reply. Reading it made me think that the first thing to
try should be to recompile the 2.6.7 kernel with no PCMCIA support at
all. This is duly did, and when I attempted to boot the kernel cardmgr
reported that there was no pcmcia driver in /proc/devices as your
message suggested.
Unfortunately, hotplug still did not start. Whatever the problem is,
that wasn't it.
Interestingly, I tried booting the 2.6 kernel with "nohotplug" as an
option; this enabled it to boot as I said earlier (but with no NIC
function). I then decided to run "/etc/rc.d/rc.hotplug start" to see
what would happen. It did not merely block as I expected but completely
locked up the system: mouse, keyboard, no interrupts working at all as
far as I could see.
I hate to be a luser but I am quite stuck here. I'll do whatever
investigation is necessary but I could really use some advice on where
to start.
Thanks,
Rich.
Dave Dodge wrote:
>On Mon, Oct 04, 2004 at 07:14:07PM +0000, Richard Wild wrote:
>
>
>>Some other information noticed: it seems (to me) quite likely that the
>>message "cardmgr[953]: no sockets found!" is related to the problem,
>>because when I boot the 2.4 kernel is says instead "cardmgr[80]: no
>>pcmcia driver in /proc/devices".
>>
>>
>
>cardmgr is a bit strange in how it accesses devices. Here's a quick
>explanation in case it's relevant to the problem.
>
>"no pcmcia driver in /proc/devices":
>
>That means just what it says. cardmgr wants the major device number for
>the pcmcia driver, so it reads through /proc/devices looking for it.
>This error message indicates that it did not find a "pcmcia" line,
>presumably because no driver was loaded.
>
>"no sockets found!":
>
>This indicates that it _did_ find "pcmcia" listed in /proc/devices,
>and it tried to access the driver. Here's where things get a bit
>unusual. After obtaining the major device number from /proc/devices,
>cardmgr uses mknod(2) to create device files of its own. It starts
>with minor device 0 and continues in a one-up loop until it fails.
>For each minor number, it creates a device file with a name of the
>form "cm-PID-COUNTER", opens the file, and then immediately unlinks
>the file while holding it open. There are a number of directories
>where it might try to do this, usually in this order:
>
> /var/lib/pcmcia, /var/run, /dev, /tmp
>
>The "no sockets found" message indicates that it managed to do the
>mknod for minor device 0, but got an ENODEV when it then tried to open
>the resulting file.
>
>So what this seems to be saying is that:
>
> - your 2.4 kernel has no PCMCIA stuff loaded at all.
>
> - your 2.6 kernel has at least the "ds" driver, which
> creates the "pcmcia" /proc/devices entry, but it has no
> card sockets.
>
>I don't know if this is part of your problem or not, but since cardmgr
>does do some device file manipulation, it might be worth investigating.
>
> -Dave Dodge
>
>
>-------------------------------------------------------
>This SF.net email is sponsored by: IT Product Guide on ITManagersJournal
>Use IT products in your business? Tell us what you think of them. Give us
>Your Opinions, Get Free ThinkGeek Gift Certificates! Click to find out more
>http://productguide.itmanagersjournal.com/guidepromo.tmpl
>_______________________________________________
>Linux-hotplug-devel mailing list http://linux-hotplug.sourceforge.net
>Linux-hotplug-devel@lists.sourceforge.net
>https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel
>
>
>
-------------------------------------------------------
This SF.net email is sponsored by: IT Product Guide on ITManagersJournal
Use IT products in your business? Tell us what you think of them. Give us
Your Opinions, Get Free ThinkGeek Gift Certificates! Click to find out more
http://productguide.itmanagersjournal.com/guidepromo.tmpl
_______________________________________________
Linux-hotplug-devel mailing list http://linux-hotplug.sourceforge.net
Linux-hotplug-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: Hotplug *seems* to hang during boot]
2004-10-04 18:52 Hotplug *seems* to hang during boot] Richard Wild
2004-10-04 19:55 ` Dave Dodge
2004-10-04 20:44 ` Richard Wild
@ 2004-10-04 21:16 ` Simone Gotti
2004-10-04 22:35 ` Richard Wild
3 siblings, 0 replies; 5+ messages in thread
From: Simone Gotti @ 2004-10-04 21:16 UTC (permalink / raw)
To: linux-hotplug
[-- Attachment #1: Type: text/plain, Size: 1708 bytes --]
On Monday 04 October 2004 23:32, Richard Wild wrote:
> Hi Dave,
>
> thanks for your reply. Reading it made me think that the first thing to
> try should be to recompile the 2.6.7 kernel with no PCMCIA support at
> all. This is duly did, and when I attempted to boot the kernel cardmgr
> reported that there was no pcmcia driver in /proc/devices as your
> message suggested.
>
> Unfortunately, hotplug still did not start. Whatever the problem is,
> that wasn't it.
>
> Interestingly, I tried booting the 2.6 kernel with "nohotplug" as an
> option; this enabled it to boot as I said earlier (but with no NIC
> function). I then decided to run "/etc/rc.d/rc.hotplug start" to see
> what would happen. It did not merely block as I expected but completely
> locked up the system: mouse, keyboard, no interrupts working at all as
> far as I could see.
>
> I hate to be a luser but I am quite stuck here. I'll do whatever
> investigation is necessary but I could really use some advice on where
> to start.
>
> Thanks,
> Rich.
You can try editing the /etc/hotplug/hotplug.functions and uncomment the line
# DEBUG=yes; export DEBUG
so you can at least see in which part or on which module it's blocking. and if
it's a kernel problem.
P.S. I don't use slackware from a lot of time, but I thing that you can look
if you are using the latest hotplug (20040923) or install it.
If this don't work you can also try to apply the patches that you can find in
debian (I hope to have time to dig in them and send them splitted to this
list, so also the gentoo's and other distro users can use them).
Bye!
--
Simone Gotti
<simone.gotti@email.it>
http://kde-bluetooth.sf.net
[-- Attachment #2: Type: application/pgp-signature, Size: 190 bytes --]
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: Hotplug *seems* to hang during boot]
2004-10-04 18:52 Hotplug *seems* to hang during boot] Richard Wild
` (2 preceding siblings ...)
2004-10-04 21:16 ` Simone Gotti
@ 2004-10-04 22:35 ` Richard Wild
3 siblings, 0 replies; 5+ messages in thread
From: Richard Wild @ 2004-10-04 22:35 UTC (permalink / raw)
To: linux-hotplug
Hi,
I had that thought myself immediately after posting. I went through the
various scripts putting echo statements in and eventually arrived at
/etc/hotplug/pci.agent with the line:
load_drivers pci $MAP_CURRENT "$LABEL"
which expands to:
load_drivers pci /lib/modules/2.6.7/modules.pcimap PCI slot
That's where it seems to go off the trolley. I will post more if and
when I find it.
Thanks,
Rich.
Simone Gotti wrote:
> You can try editing the /etc/hotplug/hotplug.functions and uncomment
> the line
>
># DEBUG=yes; export DEBUG
>
>so you can at least see in which part or on which module it's blocking. and if
>it's a kernel problem.
>
>P.S. I don't use slackware from a lot of time, but I thing that you can look
>if you are using the latest hotplug (20040923) or install it.
>If this don't work you can also try to apply the patches that you can find in
>debian (I hope to have time to dig in them and send them splitted to this
>list, so also the gentoo's and other distro users can use them).
>
>Bye!
>
>
-------------------------------------------------------
This SF.net email is sponsored by: IT Product Guide on ITManagersJournal
Use IT products in your business? Tell us what you think of them. Give us
Your Opinions, Get Free ThinkGeek Gift Certificates! Click to find out more
http://productguide.itmanagersjournal.com/guidepromo.tmpl
_______________________________________________
Linux-hotplug-devel mailing list http://linux-hotplug.sourceforge.net
Linux-hotplug-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel
^ permalink raw reply [flat|nested] 5+ messages in thread