From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail.linutronix.de (146.0.238.70:993) by crypto-ml.lab.linutronix.de with IMAP4-SSL for ; 09 Jan 2019 17:31:27 -0000 Received: from mga14.intel.com ([192.55.52.115]) by Galois.linutronix.de with esmtps (TLS1.2:DHE_RSA_AES_256_CBC_SHA256:256) (Exim 4.80) (envelope-from ) id 1ghHhN-00061u-Oc for speck@linutronix.de; Wed, 09 Jan 2019 18:31:26 +0100 Date: Wed, 9 Jan 2019 09:31:22 -0800 From: Andi Kleen Subject: [MODERATED] Re: [PATCH v3 00/32] MDSv3 12 Message-ID: <20190109173122.GP6118@tassilo.jf.intel.com> References: MIME-Version: 1.0 In-Reply-To: Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 7bit To: speck@linutronix.de List-ID: On Wed, Jan 09, 2019 at 09:09:24AM -0800, speck for Linus Torvalds wrote: > > > > I kept the support for software sequences because from what I'm hearing > > some CPUs might need them. If that's not the case they can be still > > removed. > > I'm not taking the crazy code that even Intel says is not needed. I don't know what parts of Intel you're talking with, but the parts I talk with say that it's likely some CPUs won't be able to do the microcode update for VERW. > > Gus, you need to realize that we still have taste and we still care > about maintainability. "Theoretical attack on older CPU's" is not an > excuse to throw that out the window. Are you saying we shouldn't support a MDS fix on those CPUs? What do you suggest to do instead? BTW it may be possible to remove one of the sequences i have today (the haswell+ version) if those CPUs get covered, but I expect to need to add a new version of the software sequence (for KNL) too, so it would need two at least. Also the software sequence stuff is ugly -- I fully agree on that -- but it's all in a single file and doesn't have much impact elsewhere. -Andi