From mboxrd@z Thu Jan 1 00:00:00 1970 From: Florian Weimer Subject: Re: Supporting core-specific instruction sets (e.g. big.LITTLE) with restartable sequences Date: Fri, 02 Nov 2018 20:18:12 +0100 Message-ID: <877ehvb85n.fsf@oldenburg.str.redhat.com> References: <313542172.8.1541171544337.JavaMail.zimbra@efficios.com> Mime-Version: 1.0 Content-Type: text/plain Return-path: In-Reply-To: <313542172.8.1541171544337.JavaMail.zimbra@efficios.com> (Mathieu Desnoyers's message of "Fri, 2 Nov 2018 11:12:24 -0400 (EDT)") Sender: linux-kernel-owner@vger.kernel.org To: Mathieu Desnoyers Cc: Richard Henderson , Will Deacon , linux-kernel , libc-alpha , Carlos O'Donell , Joseph Myers , Szabolcs Nagy , Thomas Gleixner , Ben Maurer , Peter Zijlstra , "Paul E. McKenney" , Boqun Feng , Dave Watson , Paul Turner , linux-api List-Id: linux-api@vger.kernel.org * Mathieu Desnoyers: > Basically, the use-cases targeted are those where some cores on the > system support a larger instruction set than others. So for instance, > some cores could use a faster atomic add instruction than others, > which should rely on a slower fallback. This is also the same story > for reading the performance monitoring unit counters from user-space: > it depends on the feature-set supported by the CPU on which the > instruction is issued. Same applies to cores having different > cache-line sizes. The kernel needs to present a consistent view to userspace, the common denominator. I don't think there is any other way. The situation is not new at all, by the way. It also arises with VM and process migration. In glibc, we do not re-run CPU feature selection upon resume (and how could we? function pointers would have to change), and we have no plans to implement anything differently. Thanks, Florian