Re: biodone() and splbio ?

[Andrew Doran <ad%NetBSD.org@localhost>]

On Wed, Jan 20, 2010 at 11:14:38AM +0900, Masao Uebayashi wrote:
> > Too bad we don't have a geenric SPLASSERT() for this kind of stuff ...
> 
> It's surprising to me we've not had SPLASSERT() now.  I think SPLASSERT() is
> possible if we don't make it generic; make it a nop on platforms which can't
> compare IPL.

The plan is that spl calls become ~extinct, only to be made by low level areas
of the kernel. Once you convert a bit of code over to locks, LOCKDEBUG
does what you want already, in a machine-independent way.

This is why I at least find SPLASSERT uninteresting and never spent any
time on it.