From owner-freebsd-hackers@FreeBSD.ORG Mon Oct 3 06:43:36 2011 Return-Path: Delivered-To: hackers@FreeBSD.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 75937106564A; Mon, 3 Oct 2011 06:43:36 +0000 (UTC) (envelope-from lev@FreeBSD.org) Received: from onlyone.friendlyhosting.spb.ru (onlyone.friendlyhosting.spb.ru [IPv6:2a01:4f8:131:60a2::2]) by mx1.freebsd.org (Postfix) with ESMTP id 123FC8FC0A; Mon, 3 Oct 2011 06:43:36 +0000 (UTC) Received: from lion.home.serebryakov.spb.ru (unknown [IPv6:2001:470:923f:1:906c:6af3:5301:18c6]) (Authenticated sender: lev@serebryakov.spb.ru) by onlyone.friendlyhosting.spb.ru (Postfix) with ESMTPA id 58DD74AC1C; Mon, 3 Oct 2011 10:43:34 +0400 (MSD) Date: Mon, 3 Oct 2011 10:43:27 +0400 From: Lev Serebryakov Organization: FreeBSD Project X-Priority: 3 (Normal) Message-ID: <535056585.20111003104327@serebryakov.spb.ru> To: Lawrence Stewart In-Reply-To: <4E89051D.30908@freebsd.org> References: <1258376930.20111002193223@serebryakov.spb.ru> <4E89051D.30908@freebsd.org> MIME-Version: 1.0 Content-Type: text/plain; charset=windows-1251 Content-Transfer-Encoding: quoted-printable Cc: hackers@FreeBSD.org Subject: Re: alq_open_flags() panics in _mtx_lock_flags() X-BeenThere: freebsd-hackers@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list Reply-To: lev@FreeBSD.org List-Id: Technical Discussions relating to FreeBSD List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Mon, 03 Oct 2011 06:43:36 -0000 Hello, Lawrence. You wrote 3 =EE=EA=F2=FF=E1=F0=FF 2011 =E3., 4:43:09: > I don't see anything obviously wrong with your call to alq_open_flags(), > other than that I think you want to set your buffer size to be a=20 > multiple of "sizeof(struct g_log_entry)". The small queue size wouldn't > cause the panic you're seeing though. > It's hard to draw any real conclusions without seeing your code. Are you > able to share it? As I wrote to current@, fs@ and geom@, it seems, that vnode API could not be called from geom threads (g_events), at least with topology lock held. So, it is not ALQ-specific problem, but some mis-interaction between GEOM and VFS levels. --=20 // Black Lion AKA Lev Serebryakov