From mboxrd@z Thu Jan 1 00:00:00 1970 From: Andi Kleen Subject: Re: [RFC PATCH v2 0/7] x86/idle: add halt poll support Date: Tue, 29 Aug 2017 06:03:26 -0700 Message-ID: <87bmmyedb5.fsf@firstfloor.org> References: <1504007201-12904-1-git-send-email-yang.zhang.wz@gmail.com> Mime-Version: 1.0 Content-Type: text/plain Cc: linux-kernel@vger.kernel.org, kvm@vger.kernel.org, wanpeng.li@hotmail.com, mst@redhat.com, pbonzini@redhat.com, tglx@linutronix.de, rkrcmar@redhat.com, dmatlack@google.com, agraf@suse.de, peterz@infradead.org, linux-doc@vger.kernel.org To: Yang Zhang Return-path: In-Reply-To: <1504007201-12904-1-git-send-email-yang.zhang.wz@gmail.com> (Yang Zhang's message of "Tue, 29 Aug 2017 11:46:34 +0000") Sender: linux-doc-owner@vger.kernel.org List-Id: kvm.vger.kernel.org Yang Zhang writes: > Some latency-intensive workload will see obviously performance > drop when running inside VM. The main reason is that the overhead > is amplified when running inside VM. The most cost i have seen is > inside idle path. You could test with https://lkml.org/lkml/2017/7/9/204 which optimizes these idle paths with a fast path. If that works it would be a far better solution than explicit tunables and basically giving up by polling. -Andi