Access counts.

[Poul-Henning Kamp]

In message <20040917094948.GP30151 at darkness.comp.waw.pl>, Pawel Jakub Dawidek w
rites:

>+> >	taste thread			gmirror thread
>+> >	------------			--------------
>+> >	g_access(da0s1a, 1, 0, 0)
>+> >	GEOM:g_access(da0s1, 1, 0, 1)
>+> >					g_access(da0s1, 0, 1, 1) -> EPERM
>+> >
>+> >This is not the same consumer in taste thread and gmirror thread.
>+>=20
>+> Why would you want to open with exclusive in taste ?  If you find that
>+> you need that once you've looked for metadata you can upgrade at that
>+> point.
>
>I don't want to! Look a bit closer. I'm opening da0s1a with r1w0e0, but
>da0s1a opens da0s1 with r1w0e1. I don't touch da0s1 in taste anymore,
>I'm only opening da0s1a there.

So what you're saying that you are using two different consumers
for da0s1 inside the same instance of g_mirror ?   One for trafic
(the one you open in your access) and one for metadata (in gmirror
thread)  ?

If so Why ?

If they're not different I don't see the problem...

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.