Re: [patch] O(1) sys_exit(), threading, scalable-exit-2.5.31-A6

[Dave McCracken <dmccr@us.ibm.com>]

--On Monday, August 19, 2002 08:36:24 PM +0200 Ingo Molnar <mingo@elte.hu>
wrote:

> well, this means that we'd still have to iterate through both lists in
> wait4(), and we'd have to maintain the ptrace list(s) in all the relevant
> codepaths - does this buy us anything relative to -B4?

The lists would constitute the tasks that wait4() should consider, at
least.    And maintaining the list wouldn't be any more work than the
current reparenting. I do admit that I don't see a significant win,
codewise, other than aesthetics.

In looking at the code I was wondering something.  What happens to the real
parent of a ptraced task when it calls wait4()?  If that's its only child,
won't it return ECHILD?

Dave McCracken

======================================================================
Dave McCracken          IBM Linux Base Kernel Team      1-512-838-3059
dmccr@us.ibm.com                                        T/L   678-3059