Re: Several Misc SCSI updates...

[Doug Ledford <dledford@redhat.com>]

On Sun, Nov 17, 2002 at 12:29:25PM -0800, Linus Torvalds wrote:
> 
> On Sun, 17 Nov 2002, Doug Ledford wrote:
> >
> > These bring the scsi subsys up to the new module loader semantics.  There 
> > is more work to be done on inter-module locking here, but we need to solve 
> > the whole module->live is 0 during init problem first or else it's a waste 
> > of time.
> 
> Hey, just remove the "live" test, I think it's over-eager and likely to 
> just cause extra code to work around it rather than fix anything.

Won't work.  module->live is what Rusty uses to indicate that the module 
is in the process of unloading, which is when we *do* want the attempt to 
module_get() to fail.  I think the process out to basically be:

load module into mem
set module->live = 1
call module_init
export module syms
done loading module

on module exit:
unexport module syms
set module->live 0
call module_exit
free module memory
done unloading.

That *should* solve all the races Rusty is trying to solve without the 
problems we've had so far, but this is only after a few minutes of 
thinking....

-- 
  Doug Ledford <dledford@redhat.com>     919-754-3700 x44233
         Red Hat, Inc. 
         1801 Varsity Dr.
         Raleigh, NC 27606