Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752620AbbHEIvp (ORCPT ); Wed, 5 Aug 2015 04:51:45 -0400 Received: from mail-wi0-f174.google.com ([209.85.212.174]:35588 "EHLO mail-wi0-f174.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751806AbbHEIvf (ORCPT ); Wed, 5 Aug 2015 04:51:35 -0400 Date: Wed, 5 Aug 2015 10:51:27 +0200 From: Ingo Molnar To: Brian Gerst Cc: the arch/x86 maintainers , Linux Kernel Mailing List , "H. Peter Anvin" , Denys Vlasenko , Andy Lutomirski , Linus Torvalds Subject: Re: [PATCH 5/8] x86/vm86: Add a separate config option for hardware IRQ handling Message-ID: <20150805085126.GB23169@gmail.com> References: <1438148483-11932-1-git-send-email-brgerst@gmail.com> <1438148483-11932-6-git-send-email-brgerst@gmail.com> <20150731085750.GA8289@gmail.com> <20150731135038.GA24619@gmail.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: User-Agent: Mutt/1.5.23 (2014-03-12) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1401 Lines: 38 * Brian Gerst wrote: > >> Disabling even less-used code that could have system stability impact. We've > >> discouraged user-mode drivers for a very long time. Ironically, other than > >> being configured through the vm86 syscall, there isn't really anything > >> vm86-specific about it. All it does is register an IRQ handler that sends a > >> signal to the task. > > > > So is this actually used by anything? Could we get away with disabling it, > > just to see whether anything cares? > > My best guess would be some very old X11 drivers that needed interrupts to run > the Video BIOS code. So let's keep it - but not complicate it with another layer of disabling logic. People that rely on legacies will enable vm86 as a single block - they won't necessarily know how deeply they rely on it. What _would_ be useful is to have a 3-mode vm86 sysctl: 1: enabled 0: disabled -1: disabled permanently (one-shot disabling after bootup) That way a distro can permanently disable vm86 for a particular bootup by setting it to -1 in /etc/sysctl.conf. The kernel should default that setting to '0'. Thanks, Ingo -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/