From owner-freebsd-questions Tue Feb 20 10:19:51 2001 Delivered-To: freebsd-questions@freebsd.org Received: from ns1.hutchtel.net (ns1.hutchtel.net [206.9.112.100]) by hub.freebsd.org (Postfix) with ESMTP id 09AB837B503; Tue, 20 Feb 2001 10:19:42 -0800 (PST) (envelope-from jpaetzel@hutchtel.net) Received: from mark9 (hutch-409.hutchtel.net [209.105.45.137]) by ns1.hutchtel.net (8.9.1/8.9.0) with SMTP id BAA15627; Sat, 17 Feb 2001 01:20:06 -0600 (CST) Message-ID: <00e801c098b2$089a3b60$6100000a@vladsempire.net> From: "Josh Paetzel" To: "Robert Watson" , "Anish A Patankar" Cc: References: Subject: Re: doubt about init process.. Date: Sat, 17 Feb 2001 01:19:27 -0600 MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.50.4133.2400 X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4133.2400 Sender: owner-freebsd-questions@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG ----- Original Message ----- From: "Robert Watson" To: "Anish A Patankar" Cc: Sent: Friday, February 16, 2001 9:06 PM Subject: Re: doubt about init process.. > > Generally speaking, sending a source code question to two mailing lists > and to the entire core team individually addressed is considered to be in > somewhat poor taste. We usually encourage systems programming questions > to be addressed to the freebsd-hackers mailing list. > > In answer to your question, I'm not sure off-hand. My guess would be that > init runs with a signal handling set inherited from proc0, and proc0 is > initialized to ignore all signals by default. The signal handling is > probably reset for children processes by execsig() (or maybe it's > sigexec()) in kern_sig.c, meaning that children processes will get a > normal default signal set. init probably also initializes handlers for a > set of basic signals to handle runlevel stuff. This is all just a guess > -- the best way to find out for sure would be to read the source code, or > if you're into more experimental tactics, you could set a breakpoint using > the kernel debugger in one of the init_main() routines associated with > starting up init, and check the signal masks there to see whether they're > reset or initialized during the fork performed on proc0. > > Robert N M Watson FreeBSD Core Team, TrustedBSD Project > robert@fledge.watson.org NAI Labs, Safeport Network Services > > On Fri, 16 Feb 2001, Anish A Patankar wrote: > > > > > The init process on FreeBSD is a user-level process (as opposed to > > a kernel-level process) that, other than having been started by the > > kernel > > itself as part of system startup, reacts just like any other process on > > the system. However if you send init the SIGKILL signal, even as root, > > init does not die? Can you please tell me why this is so?? > > Thanks > > Anish. Killing init does essentially kills the system. Unlike a traditional Unix system, FreeBSD does not allow anyone to kill init, as there are more graceful ways to shut the system down. It's a feature, not a bug. Josh To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message