From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 86B30C433FE for ; Wed, 16 Nov 2022 23:52:13 +0000 (UTC) Received: from list by lists.xenproject.org with outflank-mailman.444426.699620 (Exim 4.92) (envelope-from ) id 1ovSCF-0000f0-2X; Wed, 16 Nov 2022 23:51:59 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 444426.699620; Wed, 16 Nov 2022 23:51:59 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ovSCE-0000et-W9; Wed, 16 Nov 2022 23:51:58 +0000 Received: by outflank-mailman (input) for mailman id 444426; Wed, 16 Nov 2022 23:51:58 +0000 Received: from mail.xenproject.org ([104.130.215.37]) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ovSCE-0000en-6b for xen-devel@lists.xenproject.org; Wed, 16 Nov 2022 23:51:58 +0000 Received: from xenbits.xenproject.org ([104.239.192.120]) by mail.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1ovSC4-0001n3-1n; Wed, 16 Nov 2022 23:51:48 +0000 Received: from gw1.octic.net ([88.97.20.152] helo=[10.0.1.102]) by xenbits.xenproject.org with esmtpsa (TLS1.3:ECDHE_RSA_AES_128_GCM_SHA256:128) (Exim 4.92) (envelope-from ) id 1ovSC3-0002nk-RU; Wed, 16 Nov 2022 23:51:47 +0000 X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=xen.org; s=20200302mail; h=Content-Transfer-Encoding:Content-Type:In-Reply-To:From: References:Cc:To:Subject:MIME-Version:Date:Message-ID; bh=Vck3wzYrPUmcOloZJeqWop495rMArVD8XXqEHlY6zGM=; b=kQLO0M/EvutGbgwPeot677xb/M roE+ZwNNdMtWD2a/RWgKFuoAYf2Ro0+UVdnT5oCXszPwsQ/lGcuY/0YY6ZjUx0maHtkaivtMYFSeu OO+7PzhYOt6jFuegYx7AVpoFs/27Bn0f0KhRkpPn5lcAE13C/uFeQ7emRSmyqejaMUEA=; Message-ID: Date: Wed, 16 Nov 2022 23:51:44 +0000 MIME-Version: 1.0 User-Agent: Mozilla/5.0 (Macintosh; Intel Mac OS X 10.15; rv:102.0) Gecko/20100101 Thunderbird/102.4.2 Subject: Re: [PATCH 3/4] xen/arm, libxl: Revert XEN_DOMCTL_shadow_op; use p2m mempool hypercalls To: Stefano Stabellini , Andrew Cooper Cc: Jan Beulich , Xen-devel , Xen Security Team , Roger Pau Monne , Wei Liu , Volodymyr Babchuk , Bertrand Marquis , Henry Wang , Anthony Perard References: <20221026102018.4144-1-andrew.cooper3@citrix.com> <20221026102018.4144-4-andrew.cooper3@citrix.com> <269f4e70-07c2-5cf6-f42e-6f98e2ce70dd@citrix.com> From: Julien Grall In-Reply-To: Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 8bit Hi, On 16/11/2022 23:44, Stefano Stabellini wrote: > On Wed, 16 Nov 2022, Andrew Cooper wrote: >> On 16/11/2022 08:30, Jan Beulich wrote: >>> On 16.11.2022 03:00, Stefano Stabellini wrote: >>>> On Wed, 16 Nov 2022, Andrew Cooper wrote: >>>>> On 16/11/2022 01:37, Stefano Stabellini wrote: >>>>>> On Wed, 26 Oct 2022, Andrew Cooper wrote: >>>>>>> This reverts most of commit cf2a68d2ffbc3ce95e01449d46180bddb10d24a0, and bits >>>>>>> of cbea5a1149ca7fd4b7cdbfa3ec2e4f109b601ff7. >>>>>>> >>>>>>> First of all, with ARM borrowing x86's implementation, the logic to set the >>>>>>> pool size should have been common, not duplicated. Introduce >>>>>>> libxl__domain_set_p2m_pool_size() as a shared implementation, and use it from >>>>>>> the ARM and x86 paths. It is left as an exercise to the reader to judge how >>>>>>> libxl/xl can reasonably function without the ability to query the pool size... >>>>>>> >>>>>>> Remove ARM's p2m_domctl() infrastructure now the functioanlity has been >>>>>>> replaced with a working and unit tested interface. >>>>>>> >>>>>>> This is part of XSA-409 / CVE-2022-33747. >>>>>> Genuine question: I can see this patch removes the implementation of >>>>>> XEN_DOMCTL_SHADOW_OP_SET_ALLOCATION on ARM. It also switches libxl (both >>>>>> ARM and x86) to the new hypercall. >>>>>> >>>>>> Why keep the old hypercall (XEN_DOMCTL_shadow_op and >>>>>> XEN_DOMCTL_SHADOW_OP_SET_ALLOCATION) implementation on x86 (not on ARM)? >>>>>> >>>>>> Is that because it was only recently implemented? And not actually >>>>>> present in any past Xen release? >>>>>> >>>>>> If so, please add a note about this in the commit message. Also, if that >>>>>> is the case, I think this patch series should go in 4.17. If it is too >>>>>> late to get it in before the release, then we should backport it to 4.17 >>>>>> as soon as possible. That's because ideally we want to keep the >>>>>> hypercall interface changes down to a minimum. >>>>> On ARM, the hypercall has existed for a little over 4 weeks, and isn't >>>>> in any released version of Xen (yet). >>>>> >>>>> On x86, the hypercall has existed for more than a decade, and has known >>>>> out-of-tree users.  It needs to be deprecated properly, which in this >>>>> case means "phased out in the 4.18 cycle once known callers have been >>>>> adapted to the new hypercall". >>>> Understoon. Then I am in favor of getting all 4 patches in 4.17, either >>>> before the release or via backports. >>> Removing something from the domctl interface generally requires bumping >>> the interface version, so some extra care may need applying if such an >>> interface change was to be backported to any stable branch. >> >> To be clear, I have no plans to remove the x86 "older" interface in this >> patch series.  It will definitely break out of tree users. >> >> In the 4.18 timeframe, we can see about retiring the older hypercalls, >> but as a non-backportable change. > > For ARM, given that XEN_DOMCTL_shadow_op has not been enabled for long, > maybe we can get away without bumping the interface version? IMHO how long it was out doesn't matter. Once we do a release, we should avoid changing the interface in minor version. This is because a user may start to rely on it and we don't want to break them for minor releases. Cheers, -- Julien Grall