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 ; 29 Jun 2018 17:01:42 -0000 Received: from mga01.intel.com ([192.55.52.88]) by Galois.linutronix.de with esmtps (TLS1.2:DHE_RSA_AES_256_CBC_SHA256:256) (Exim 4.80) (envelope-from ) id 1fYwmC-0003BQ-BO for speck@linutronix.de; Fri, 29 Jun 2018 19:01:40 +0200 Date: Fri, 29 Jun 2018 10:01:38 -0700 From: "Luck, Tony" Subject: [MODERATED] Re: Re: [patch V4 13/13] x86/apic: Ignore secondary threads if nosmt=force Message-ID: <20180629170137.GA18334@agluck-desk> References: <20180620201907.304694346@linutronix.de> <20180620201933.588840902@linutronix.de> <9a4cea33-05a7-b6d9-7f49-692603bd047f@linux.intel.com> <20180629082617.GA3752@zn.tnic> MIME-Version: 1.0 In-Reply-To: <20180629082617.GA3752@zn.tnic> Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 7bit To: speck@linutronix.de List-ID: On Fri, Jun 29, 2018 at 10:26:17AM +0200, speck for Borislav Petkov wrote: > Ok, question: what exactly does the hw/microcode do on an offlined > logical CPU: does it test CR4.MCE first and assert that IERR# or does it > look at MCG_CTL/MCi_CTL too so that we can try to disable MCE reporting > on those offlined cores? MCi_CTL may or may not save you. Banks are shared across cores or the whole package. So if you bring any CPU online that shares the bank that reports the error, then the offline CPUs that share that bank will have the machine check enabled. For the hyper-thread disabled case you are guaranteed to be screwed. -Tony