On Thu, Mar 23, 2006 at 10:50:49PM +0100, Andi Kleen wrote:
> It happens without anybody ever using hangcheck for anything. It's just
> compiled in. You shouldn't assume everybody cares about that hang checking
> thing just because the driver is loaded.
Without disagreeing with you, why would someone compile in or
load hangcheck if they didn't want it? I mean the general
"distro-provided" case, etc. The random kernel hacker might try it, but
they can also learn why they don't want to.
> All the other watchdogs in Linux are only active when at least one person
> ever opened the device. That ought be right behaviour for hangcheck too.
The above said, I'm going to think about this issue. It will be
an additional pain for existing users (especially already-in-use
software that expects it to Just Work), but that doesn't mean it's a bad
idea.
The interface won't be open(2), though. This need not rely on a
process. Perhaps a sysfs knob.
Joel
--
"Not being known doesn't stop the truth from being true."
- Richard Bach
Joel Becker
Principal Software Developer
Oracle
E-mail: joel.becker@xxxxxxxxxx
Phone: (650) 506-8127
-
: send the line "unsubscribe linux-x86_64" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
