From owner-freebsd-current@FreeBSD.ORG Tue Sep 29 16:40:08 2009 Return-Path: Delivered-To: freebsd-current@freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:4f8:fff6::34]) by hub.freebsd.org (Postfix) with ESMTP id 373491065672; Tue, 29 Sep 2009 16:40:08 +0000 (UTC) (envelope-from peter@wingless.org) Received: from mail.pfak.org (satan.pfak.org [IPv6:2001:470:b:14::d813:b29a]) by mx1.freebsd.org (Postfix) with ESMTP id 165BB8FC35; Tue, 29 Sep 2009 16:40:08 +0000 (UTC) Received: by mail.pfak.org (Postfix, from userid 65534) id C263340DF; Tue, 29 Sep 2009 09:40:07 -0700 (PDT) X-Spam-Checker-Version: SpamAssassin 3.2.5 (2008-06-10) on mail.pfak.org X-Spam-Level: X-Spam-Status: No, score=-4.4 required=5.0 tests=ALL_TRUSTED,BAYES_00 autolearn=no version=3.2.5 Received: from [192.168.1.123] (S0106001c105b9c2c.vf.shawcable.net [96.48.72.201]) (using TLSv1 with cipher DHE-RSA-CAMELLIA256-SHA (256/256 bits)) (No client certificate requested) (Authenticated sender: peter@kieser.ca) by mail.pfak.org (Postfix) with ESMTPSA id 4164840C8; Tue, 29 Sep 2009 09:40:07 -0700 (PDT) Message-ID: <4AC23869.2050601@wingless.org> Date: Tue, 29 Sep 2009 09:40:09 -0700 From: Peter Kieser User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.2; en-US; rv:1.9.1.4pre) Gecko/20090915 Thunderbird/3.0b4 MIME-Version: 1.0 To: Alan Cox References: <4ABB200E.3020706@wingless.org> <4ABC585D.6090609@wingless.org> <4ABC5CA7.10702@cs.rice.edu> In-Reply-To: <4ABC5CA7.10702@cs.rice.edu> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Cc: alc@freebsd.org, freebsd-current@freebsd.org Subject: Re: Superpages and VMware ESXi 3.5u4 crashes X-BeenThere: freebsd-current@freebsd.org X-Mailman-Version: 2.1.5 Precedence: list List-Id: Discussions about the use of FreeBSD-current List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 29 Sep 2009 16:40:08 -0000 On 9/24/2009 11:01 PM, Alan Cox wrote: > Peter Kieser wrote: >> On 9/24/2009 10:23 PM, Alan Cox wrote: >>> I would agree that this seems like a bug in ESX. Nonetheless, I'm >>> curious to know something about our your system. What type of processor >>> are you using, and in particular is ESX using hardware supported nested >>> paging? >>> >>> Regards, >>> Alan >>> >> >> I'm using an AMD Phenom 9950. How would I tell if ESXi is using >> hardware supported nested paging? Is this something I can tell in >> FreeBSD? > > Ok. Your processor does support nested paging, although you may have to > explicitly enable it in your BIOS. FreeBSD can't tell whether you're > using nested paging. Support for nested paging is implemented in the > hypervisor and is invisible to the guess, so only ESX can tell you if > it's being used. Unfortunately, I'm not familiar enough with ESX to tell > you how to determine this. > > If you're not currently using nested paging, it's plausible that turning > it on could make this problem go away. Likewise, if you are currently > using nested paging, turning it off may help. If nothing else, this > could help narrow down where the problem lies. > > Alan I do not have the option in my BIOS to enable nested pages. However upgrading to ESXi 4.0 made the problem go away. Seems like a pretty large bug in VMware. Unfortunately I'm using "unsupported" hardware so VMware is unwilling to even look into the bug. -Peter