From owner-cvs-all Tue Mar 20 10:58: 3 2001 Delivered-To: cvs-all@freebsd.org Received: from salmon.maths.tcd.ie (salmon.maths.tcd.ie [134.226.81.11]) by hub.freebsd.org (Postfix) with SMTP id D17A437B719; Tue, 20 Mar 2001 10:57:58 -0800 (PST) (envelope-from iedowse@maths.tcd.ie) Received: from walton.maths.tcd.ie by salmon.maths.tcd.ie with SMTP id ; 20 Mar 2001 18:57:57 +0000 (GMT) To: Robert Watson Cc: cvs-committers@FreeBSD.org, cvs-all@FreeBSD.org, iedowse@maths.tcd.ie Subject: Re: cvs commit: src/sbin/reboot reboot.c In-Reply-To: Your message of "Tue, 20 Mar 2001 12:59:34 EST." Date: Tue, 20 Mar 2001 18:57:57 +0000 From: Ian Dowse Message-ID: <200103201857.aa69023@salmon.maths.tcd.ie> Sender: owner-cvs-all@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG In message , Robe rt Watson writes: > >SIGTERM, then SIGKILL to take effect. Probably, it would be nice if these >values were consistent between various programs (shutdown, init, halt...). >Not sure what the "right" values are -- the ones there seem like >potentially decent approximations. Good point. It seems that there are just two implementations of the SIGTERM->SIGKILL delay; one in init(8) and one in reboot (halt is a hard link to reboot, and shutdown(8) signals init to do the killing). The reboot/halt code had the shorter delay of 5 seconds which is probably why I noticed the problem more frequently in that case. I could implement a similar dynamic delay in init I suppose. On my laptop I was seeing page-ins occurring for about 20 seconds after the SIGTERM, so 10 seconds isn't always enough. Ian To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe cvs-all" in the body of the message