From owner-freebsd-current Thu Sep 17 13:29:13 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id NAA20499 for freebsd-current-outgoing; Thu, 17 Sep 1998 13:29:13 -0700 (PDT) (envelope-from owner-freebsd-current@FreeBSD.ORG) Received: from godzilla.zeta.org.au (godzilla.zeta.org.au [203.15.68.22]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id NAA20445 for ; Thu, 17 Sep 1998 13:28:52 -0700 (PDT) (envelope-from bde@godzilla.zeta.org.au) Received: (from bde@localhost) by godzilla.zeta.org.au (8.8.7/8.8.7) id GAA00254; Fri, 18 Sep 1998 06:28:26 +1000 Date: Fri, 18 Sep 1998 06:28:26 +1000 From: Bruce Evans Message-Id: <199809172028.GAA00254@godzilla.zeta.org.au> To: bde@zeta.org.au, jhay@mikom.csir.co.za Subject: Re: Death by SIGXCPU (problems with our clock code) Cc: current@FreeBSD.ORG Sender: owner-freebsd-current@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG >Is it possible to lower the priority of the sio interrupts? To maybe Not without weakening them, especially for timestamping. >not configure them as fast interrupts? On the time servers the serial Not configuring them as fast interrupts works, except it is hard-coded (`something |= RI_FAST') and it weakens them. BTW, I'm still trying to figure out how to configure fast interrupts for the pci cy driver. Configuring them either way just works provided cy's pci interrupt is not shared. It seems to be fundamentally impossible to share an interrupt between fast and slow interrupt handlers. >Will disable_intr()/enable_intr() work on the SMP machines? Maybe. I think the giant lock prevents other CPUs entering the kernel while one is processing clkintr(), so even plain disable_intr()/ enable_intr() probably works (disable_intr() is a macro in clock.c in the SMP case). Bruce To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message