Subject: Re: pr/35143 and layer_node_find()
To: None <email@example.com, gnats-bugs@NetBSD.org>
From: Chuck Silvers <firstname.lastname@example.org>
Date: 11/30/2006 08:27:30
On Wed, Nov 29, 2006 at 11:20:19AM -0800, Bill Studenmund wrote:
> But we can't return a valid vp in this case. To get into this corner case,
> we have to have another thread in the kernel actively in the process of
> cleaning said vnode. To be honest, I'm not 100% sure how a big-lock kernel
> got into this case, but it did...
... and now I realize that my proposal doesn't actually fix the problem
in this PR. as you say, we don't understand how it could have happened yet.
I think we need to figure that out before we decide on a fix.