Can't kill hung remote CP copy
Joe Zeff
joe at zeff.us
Sat Apr 14 19:42:13 UTC 2012
On 04/14/2012 12:25 PM, Michael Hennebry wrote:
> Under what circumstance would killing a waiting process
> be worse than a process that should have waited,
> but terminated instead?
I think that the idea is that if the process is in an uninterruptable
sleep, it's for a good reason and the kernel is expected to assume that
the programmer knew what he was doing and why. Most of the time, that's
a good assumption; in fact, almost always. Alas, there's currently no
way to override that on the very rare occasions that it's wrong.
More information about the users
mailing list