Re: mutexes, locks and so on...

[Johnny Billquist <bqt%softjar.se@localhost>]

On 11/17/10 13:55, Thor Lancelot Simon wrote:
> On Wed, Nov 17, 2010 at 01:31:51PM +0100, Johnny Billquist wrote:
> > But you could perhaps argue that this is a bug in itself, and
> > mutex_vector_enter should not take a spin mutex on its own when a
> > mutex_spin_enter function exist.
>
> I don't think that's right.  I don't believe spin mutexes are _required_
> to be taken with mutex_spin_enter() -- mutex_enter() should work fine.
> It has to, because it is supposed to be invisible to the programmer
> whether mutexes are spin or adaptive, except for the initial IPL used
> to allocate them.
>
> At least, that is how I understand it.  Which may be wrong.

You misunderstood me. :-)
If code would have called mutex_enter, I can handle it correctly.
It is that mutex_vector_enter can also take a spin mutex, and it will 
not take it the same way as mutex_spin_enter (or mutex_enter) will do 
it. mutex_vector_enter has it's own spin mutex enter code, which if 
invoked, means that mutex_enter and mutex_spin_enter cannot actually 
have their own implementation, but must in fact be the same as the one 
in mutex_vector_enter.

        Johnny