From owner-freebsd-current@FreeBSD.ORG  Thu Nov 23 16:58:58 2006
Return-Path: <owner-freebsd-current@FreeBSD.ORG>
X-Original-To: freebsd-current@freebsd.org
Delivered-To: freebsd-current@freebsd.org
Received: from mx1.FreeBSD.org (mx1.freebsd.org [69.147.83.52])
	by hub.freebsd.org (Postfix) with ESMTP id 7909416A403;
	Thu, 23 Nov 2006 16:58:58 +0000 (UTC)
	(envelope-from bakul@bitblocks.com)
Received: from mail.bitblocks.com (bitblocks.com [209.204.185.216])
	by mx1.FreeBSD.org (Postfix) with ESMTP id 1F63D43D69;
	Thu, 23 Nov 2006 16:58:21 +0000 (GMT)
	(envelope-from bakul@bitblocks.com)
Received: from bitblocks.com (localhost.bitblocks.com [127.0.0.1])
	by mail.bitblocks.com (Postfix) with ESMTP id 5931F5B3E;
	Thu, 23 Nov 2006 08:58:57 -0800 (PST)
To: Hajimu UMEMOTO <ume@freebsd.org>
In-reply-to: Your message of "Thu, 23 Nov 2006 21:07:36 +0900."
	<ygeu00qs6wn.wl%ume@mahoroba.org> 
Date: Thu, 23 Nov 2006 08:58:57 -0800
From: Bakul Shah <bakul@bitblocks.com>
Message-Id: <20061123165857.5931F5B3E@mail.bitblocks.com>
Cc: freebsd-current@freebsd.org
Subject: Re: unkillable qemu: sleeping in devdrn state 
X-BeenThere: freebsd-current@freebsd.org
X-Mailman-Version: 2.1.5
Precedence: list
List-Id: Discussions about the use of FreeBSD-current
	<freebsd-current.freebsd.org>
List-Unsubscribe: <http://lists.freebsd.org/mailman/listinfo/freebsd-current>, 
	<mailto:freebsd-current-request@freebsd.org?subject=unsubscribe>
List-Archive: <http://lists.freebsd.org/pipermail/freebsd-current>
List-Post: <mailto:freebsd-current@freebsd.org>
List-Help: <mailto:freebsd-current-request@freebsd.org?subject=help>
List-Subscribe: <http://lists.freebsd.org/mailman/listinfo/freebsd-current>,
	<mailto:freebsd-current-request@freebsd.org?subject=subscribe>
X-List-Received-Date: Thu, 23 Nov 2006 16:58:58 -0000

> bakul> With  a fairly recent -current kernel qemu sleeps in "devdrn"
> bakul> state and is unkillable.  /proc doesn't have qemu's pid entry
> bakul> but it does show up in a ps listing!  This bug is not 100%
> bakul> repeatable but close enough.  AFAIK this problem appears only
> bakul> when kqemu is used with qemu.
> 
> Though I'm not using qemu, I met similar problem with portupgrade.
> The portupgrade sleeps in "devdrm" state.
> Don't you set kern.pts.enable to 1?  It seems the problem doesn't
> occur when kern.pts.enable=0.

I tried that but the problem remained.  The following "fixed"
it but I don't trust it is the right fix.  IMHO destroy_dev()
needs to do dev_relthread() or we need to figure out why
si_threadcount is > 0 and fix that so that we don't have to
go fix all the drivers.

--- kqemu-freebsd.c-old Thu Nov 19 08:36:25 2006
+++ kqemu-freebsd.c   Mon Nov 20 01:05:56 2006
@@ -321,6 +321,7 @@
 #if __FreeBSD_version >= 500000
     dev->si_drv1 = NULL;
     TAILQ_REMOVE(&kqemuhead, ks, kqemu_ent);
+    dev_relthread(dev); 
     destroy_dev(dev);
 #endif
     free(ks, M_KQEMU);