From owner-freebsd-current Sat Mar 25 2:12:44 2000 Delivered-To: freebsd-current@freebsd.org Received: from apollo.backplane.com (apollo.backplane.com [216.240.41.2]) by hub.freebsd.org (Postfix) with ESMTP id AF0E037B520 for ; Sat, 25 Mar 2000 02:12:35 -0800 (PST) (envelope-from dillon@apollo.backplane.com) Received: (from dillon@localhost) by apollo.backplane.com (8.9.3/8.9.1) id CAA19991; Sat, 25 Mar 2000 02:12:07 -0800 (PST) (envelope-from dillon) Date: Sat, 25 Mar 2000 02:12:07 -0800 (PST) From: Matthew Dillon Message-Id: <200003251012.CAA19991@apollo.backplane.com> To: Bruce Evans Cc: freebsd-current@FreeBSD.ORG Subject: Re: Anyone know why the syscall interface is using the doreti mechanism? References: Sender: owner-freebsd-current@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG Ah, excellent summary Bruce! Now I know what to look for and test re: syscall returns. I'm confident I can at least test for the cases without needing the MP lock, which is all we really need to be able to optimize the critical path. I have also successfully removed all the nasty *CPL_LOCK stuff. It was only half-implemented anyway and virtually none of it would have been useful for interrupt threads. It's kinda nice having the SMP spl*() call overhead back down to the UP case. I'll post the new patch tomorrow. It's getting clean enough that you can almost understand the interrupt code :-) -Matt To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message