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 aws-us-west-2-korg-lkml-1.web.codeaurora.org (localhost.localdomain [127.0.0.1]) by smtp.lore.kernel.org (Postfix) with ESMTP id EB999C433EF for ; Fri, 21 Jan 2022 13:42:17 +0000 (UTC) Received: from relay5-d.mail.gandi.net (relay5-d.mail.gandi.net [217.70.183.197]) by mx.groups.io with SMTP id smtpd.web10.12008.1642772536838288413 for ; Fri, 21 Jan 2022 05:42:17 -0800 Authentication-Results: mx.groups.io; dkim=missing; spf=pass (domain: bootlin.com, ip: 217.70.183.197, mailfrom: michael.opdenacker@bootlin.com) Received: (Authenticated sender: michael.opdenacker@bootlin.com) by mail.gandi.net (Postfix) with ESMTPSA id B5E2F1C0003; Fri, 21 Jan 2022 13:42:14 +0000 (UTC) Cc: Ross Burton , OE-core Subject: Re: [OE-core] [PATCH] runqemu: automatically turn on "kvm" on x86 CPUs with VT To: Alexander Kanavin , Richard Purdie References: <20220120152146.1912559-1-michael.opdenacker@bootlin.com> <27cd215a-abb7-a873-40a6-86df335691b1@bootlin.com> <56724a11b917e32b196f8aeb7c4678c4e0903d49.camel@linuxfoundation.org> From: Michael Opdenacker Organization: Bootlin Message-ID: Date: Fri, 21 Jan 2022 14:42:14 +0100 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:78.0) Gecko/20100101 Thunderbird/78.14.0 MIME-Version: 1.0 In-Reply-To: Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 7bit Content-Language: en-US List-Id: X-Webhook-Received: from li982-79.members.linode.com [45.33.32.79] by aws-us-west-2-korg-lkml-1.web.codeaurora.org with HTTPS for ; Fri, 21 Jan 2022 13:42:17 -0000 X-Groupsio-URL: https://lists.openembedded.org/g/openembedded-core/message/160828 On 1/21/22 12:43 PM, Alexander Kanavin wrote: > On Thu, 20 Jan 2022 at 21:46, Richard Purdie > > wrote: > > > I think the patch as it stands will actually regress things for > many users. You > must be already in the right groups to be able to access kvm on > your system but > I don't think that is the default for many distros. On those > distros you would > see a "permission denied" message from qemu which users find very > confusing and > I think this is why the code is in the form it is in. > > We might be able to improve the test further with the open > permissions Alexander > mentions but that is probably necessary to make the patch mergable. > > > I guess what we can do for a start is to check if host and target > arches match (e.g. x86 on x86), and then if kvm is not asked for, > gently suggest to the users that they do, for a much better > experience, with a link to the kvm on yocto guide. Then we can look > into auto-attempting it carefully, so that such attempts don't make > the experience worse. Thanks for your ideas. If you think this idea of automatically enabling KVM is risky, it sounds like a good idea indeed to make this suggestion when applicable. I'll also update the documentation to remind people. I guess that mainly beginner users are impacted, and those are the ones who will read the our docs. Thanks again Michael. -- Michael Opdenacker, Bootlin Embedded Linux and Kernel engineering https://bootlin.com