So the use count doesn't include the module dependancies? If sr_mod depends
on cdrom and cdrom is marked as [sr_mod] should that not count as 1 in the
use count.
> As to why sr_mod and isofs don't get unloaded, it is a feature of the new
> module system that is trying to prevent a race condition in kerneld.
> What would happen is, on some IDE systems, it would take some time (more
> than a minute) for the CD to come on line and begin reading data. In the
> meantime, the isofs module got unloaded and so the mount failed. This
> would happen over and over.
>
> So The Plan is that kerneld won't automatically unload a module that has
> never been used, so that we don't run into that situation.
>
> I'm not sure how far along the chain you have to go in the scsi subsystem
> before the bits trip to say that the module is in use, but at this point
> I can only presume it wasn't far enough.
Cool idea. Maybe we could have an option in the config file which could
specify a different wait time for different modules so people with slow
devices could set the timeout on those modules to 2 minutes, for instance.
-- [======================================================================] [ Kevin Lentin Email: K.Lentin@cs.monash.edu.au ] [ finger kevinl@fangorn.cs.monash.edu.au for PGP public key block. ] [ KeyId: 06808EED FingerPrint: 6024308DE1F84314 811B511DBA6FD596 ] [======================================================================]