From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:37750) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1b5Ib1-0004Xu-KN for qemu-devel@nongnu.org; Tue, 24 May 2016 16:06:32 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1b5Iax-0004ea-0h for qemu-devel@nongnu.org; Tue, 24 May 2016 16:06:30 -0400 Received: from out2-smtp.messagingengine.com ([66.111.4.26]:44256) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1b5Iau-0004ck-Nh for qemu-devel@nongnu.org; Tue, 24 May 2016 16:06:26 -0400 From: "Emilio G. Cota" Date: Tue, 24 May 2016 16:06:11 -0400 Message-Id: <1464120374-8950-1-git-send-email-cota@braap.org> Subject: [Qemu-devel] [PATCH v2 0/3] atomics: fix RCU perf. regression + update documentation List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: QEMU Developers , MTTCG Devel Cc: =?UTF-8?q?Alex=20Benn=C3=A9e?= , Paolo Bonzini , Richard Henderson , Sergey Fedorov v1: https://lists.gnu.org/archive/html/qemu-devel/2016-05/msg03661.html Patch 1 hasn't changed from v1 (where it was patch 2 though). Patches 2 and 3 fix a not-so-small-after-all RCU performance regression we introduced when transitioning to __atomic primitives. I got an arm64 machine to test today and a workload that issues a lot of atomic_read_rcu's, such as a 100%-lookup qht-bench test, can gain ~12% in performance. [ in v1's 0/2 message I mentioned rcutorture; It turns out it's not as dependent on atomic_read_rcu as I thought, so it's not a good benchmark to measure this effect. ] Thanks, Emilio