Re: [RFC][PATCH 0/3] fork: Add the ability to create tasks with given pids

From: Tejun Heo <tj@kernel.org>
To: Pavel Emelyanov <xemul@parallels.com>
Cc: Oleg Nesterov <oleg@redhat.com>,
	Linus Torvalds <torvalds@linux-foundation.org>,
	Andrew Morton <akpm@linux-foundation.org>,
	Alan Cox <alan@linux.intel.com>,
	Roland McGrath <roland@hack.frob.com>,
	Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
	Cyrill Gorcunov <gorcunov@openvz.org>,
	James Bottomley <jbottomley@parallels.com>
Subject: Re: [RFC][PATCH 0/3] fork: Add the ability to create tasks with given pids
Date: Tue, 22 Nov 2011 07:29:39 -0800	[thread overview]
Message-ID: <20111122152939.GC322@google.com> (raw)
In-Reply-To: <20111122152312.GB322@google.com>

On Tue, Nov 22, 2011 at 07:23:12AM -0800, Tejun Heo wrote:
> I think the question should usually be the other way around.  Why
> deviate when the deviation doesn't earn any tangible benefit?  If you
> think setting it explicitly is justified, explain why in the comment
> of the setter and places where those explicit settings are.

Hmmm... I think I trimmed a bit too much.  Let me add back a bit.

So, either just use set_last_pid() or explain in the comment of
set_last_pid() that there are other places which set last_pid but it
won't race because they're synchronized through outer lock and similar
comment where the explicit setting of last_pid is too.  As it
currently stands, it really is difficult to find out who else would be
changing last_pid - it's buried in an argument to a proc parse
function.  IMHO just using set_last_pid() would be better here.

Thanks.

-- 
tejun