From mboxrd@z Thu Jan  1 00:00:00 1970
From: Andi Kleen <ak@suse.de>
Date: Mon, 22 Nov 2004 19:53:49 +0000
Subject: Re: [Lse-tech] Re: scalability of signal delivery for Posix Threads
Message-Id: <20041122195348.GB11097@wotan.suse.de>
List-Id: <linux-ia64.vger.kernel.org>
References: <41A20AF3.9030408@sgi.com> <20041122160705.GG25636@parcelfarce.linux.theplanet.co.uk> <41A242C1.10600@sgi.com>
In-Reply-To: <41A242C1.10600@sgi.com>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
To: Ray Bryant <raybry@sgi.com>
Cc: Matthew Wilcox <matthew@wil.cx>, Kernel Mailing List <linux-kernel@vger.kernel.org>, "linux-ia64@vger.kernel.org" <linux-ia64@vger.kernel.org>, lse-tech <lse-tech@lists.sourceforge.net>, holt@sgi.com, Dean Roe <roe@sgi.com>, Brian Sumner <bls@sgi.com>, John Hawkes <hawkes@tomahawk.engr.sgi.com>

> Well, the sighand->siglock is taken so many places in the kernel (>200 
> times)
> that RCUing its usage looks like a daunting change to make.

Agreed.  And having to wait for all CPUs in sigaction would also not
be nice.

> 
> In principle, I guess a seqlock could be made to work.  The idea would be 

seqlocks are reader only, but for signal delivery you need a writer to 
update state like the thread load balancing. We got all that gunk
from POSIX, before NPTL it would have been probably possible ;-)

-Andi