Re: Need new modutils aliases for 2.4.0? (Was)Re:Cannothotplugcardbus

Re: Need new modutils aliases for 2.4.0? (Was)Re:Cannothotplugcardbus

[David Brownell]

Yes, at the PCI level I think that's exactly right.

For "network" hotplugging, I'm not sure what the analagous policy
would be ... what comes to mind is getting the network interface
brought up exactly once. I saw some flakey behavior there a while
back, maybe it's gone now.  Aren't there distro portability issues,
like "ifup" not being universal?

- Dave


----- Original Message ----- 
From: David Hinds <dhinds@sonic.net>
To: David Brownell <david-b@pacbell.net>
Cc: <linux-hotplug-devel@lists.sourceforge.net>
Sent: Friday, January 26, 2001 10:10 AM
Subject: Re: Need new modutils aliases for 2.4.0? (Was)Re:Cannothotplugcardbus cards.


> On Fri, Jan 26, 2001 at 08:44:56AM -0800, David Brownell wrote:
> > 
> > That sounds good.  Can that also understand the network hotplugging?
> > Or is that going to need work yet?
> 
> I'm not sure what you mean?  When cardmgr sees a card that has a hot
> plug PCI driver, it will basically ignore the card completely and
> leave everything to /sbin/hotplug.  Isn't that what we want?
> 
> -- Dave


_______________________________________________
Linux-hotplug-devel mailing list  http://linux-hotplug.sourceforge.net
Linux-hotplug-devel@lists.sourceforge.net
http://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel

Re: Need new modutils aliases for 2.4.0? (Was) Re: Cannothotplugcardbus

[Miles Lane]

David Brownell wrote:
> 
> My thoughts exactly.  2.4 should be hotplugging the 2.4 kernel's
> Cardbus drivers ... unbundled pcmcia_cs drivers shouldn't be used,
> as a rule, the kernel drivers get more attention (and the expectation
> is that'll continue).
> 
> I think Miles still needs to get that "pci.agent" patch, which
> should help it read those variables correctly.  After we see more
> reports of that behaving, we can start to test "pcimodules".
> 
> (No, "/sbin/hotplug" has nothing to do with "cardmgr", they still
> need some sort of mutual-noninterference pact ...)

Cool.  PCI hotplugging is working for my Cardbus cards now.

I would like to explore the "mutual-noninterference pact" 
between Cardmgr and /sbin/hotplug.  It seems important to
get right.  Obviously, David Hinds will play a big role
in this.

	Miles

_______________________________________________
Linux-hotplug-devel mailing list  http://linux-hotplug.sourceforge.net
Linux-hotplug-devel@lists.sourceforge.net
http://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel

Re: Need new modutils aliases for 2.4.0? (Was) Re: Cannothotplugcardbus

[Miles Lane]

> David Hinds wrote:
>
> On Thu, Jan 25, 2001 at 01:43:18PM -0800, Miles Lane wrote:
> > David Brownell wrote:
> > >
> > > My thoughts exactly.  2.4 should be hotplugging the 2.4 kernel's
> > > Cardbus drivers ... unbundled pcmcia_cs drivers shouldn't be used,
> > > as a rule, the kernel drivers get more attention (and the expectation
> > > is that'll continue).
>
> Actually I think all the evidence suggests that at this point the
> unbundled pcmcia drivers still get considerably more attention than
> the kernel drivers, at least at the end user level.

Well, during the 2.5 time-frame, the kernel developers will
probably put more effort into improving the kernel drivers,
while you continue (?) to work on improving the pcmcia-cs
drivers.  I agree with you that the pcmcia-cs drivers are
getting *way* more use and testing.  I feel frustrated about
this.  We could be much further along if one or two kernel
developers devoted their focus to getting PCMCIA/Cardbus
development and testing into good shape.  Currently, people
who are willing to test the kernel drivers don't have a
"champion" to turn to.  You have done an admirable job of
gathering the community of users and serving them.

I'd be willing to do a lot of driver testing, but it requires
far more hardware than I have access to.

I do think that you, David, could help by evangelizing testing
of the kernel drivers.  Most of what people use PCMCIA/Cardbus
devices for is adjunct to mission-critical work.  That means
that if folks are willing to rebuild their kernel and do some
testing, they can boot back into their stable, comfy environments
pretty easily.  It would be a tremendous help if testers would
step up and simply tell us what's broken!

> > I would like to explore the "mutual-noninterference pact"
> > between Cardmgr and /sbin/hotplug.  It seems important to
> > get right.  Obviously, David Hinds will play a big role
> > in this.
>
> I have added code to cardmgr to parse modules.pcimap to figure out
> which cards should be handled by the hotplug stuff.  I still have a
> few things to clean up but I don't think it will be hard.

Wow!  That's great news!  Thanks for working on this.

        Miles

_______________________________________________
Linux-hotplug-devel mailing list  http://linux-hotplug.sourceforge.net
Linux-hotplug-devel@lists.sourceforge.net
http://lists.sourceforge.net/lists/listinfo/linux-hotplug-devel