From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-ua0-f199.google.com (mail-ua0-f199.google.com [209.85.217.199]) by kanga.kvack.org (Postfix) with ESMTP id 77FD6830F1 for ; Mon, 29 Aug 2016 23:39:19 -0400 (EDT) Received: by mail-ua0-f199.google.com with SMTP id 30so19861871uab.1 for ; Mon, 29 Aug 2016 20:39:19 -0700 (PDT) Received: from mail.linuxfoundation.org (mail.linuxfoundation.org. [140.211.169.12]) by mx.google.com with ESMTPS id g128si11354732ybg.234.2016.08.29.20.39.18 for (version=TLS1_2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128/128); Mon, 29 Aug 2016 20:39:18 -0700 (PDT) Date: Mon, 29 Aug 2016 20:39:16 -0700 From: Andrew Morton Subject: Re: [PATCH] thp: reduce usage of huge zero page's atomic counter Message-Id: <20160829203916.6a2b45845e8fb0c356cac17d@linux-foundation.org> In-Reply-To: <36b76a95-5025-ac64-0862-b98b2ebdeaf7@intel.com> References: <20160829155021.2a85910c3d6b16a7f75ffccd@linux-foundation.org> <36b76a95-5025-ac64-0862-b98b2ebdeaf7@intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-linux-mm@kvack.org List-ID: To: Aaron Lu Cc: Linux Memory Management List , "'Kirill A. Shutemov'" , Dave Hansen , Tim Chen , Huang Ying , Vlastimil Babka , Jerome Marchand , Andrea Arcangeli , Mel Gorman , Ebru Akagunduz , linux-kernel@vger.kernel.org On Tue, 30 Aug 2016 11:09:15 +0800 Aaron Lu wrote: > >> Case used for test on Haswell EP: > >> usemem -n 72 --readonly -j 0x200000 100G > >> Which spawns 72 processes and each will mmap 100G anonymous space and > >> then do read only access to that space sequentially with a step of 2MB. > >> > >> perf report for base commit: > >> 54.03% usemem [kernel.kallsyms] [k] get_huge_zero_page > >> perf report for this commit: > >> 0.11% usemem [kernel.kallsyms] [k] mm_get_huge_zero_page > > > > Does this mean that overall usemem runtime halved? > > Sorry for the confusion, the above line is extracted from perf report. > It shows the percent of CPU cycles executed in a specific function. > > The above two perf lines are used to show get_huge_zero_page doesn't > consume that much CPU cycles after applying the patch. > > > > > Do we have any numbers for something which is more real-wordly? > > Unfortunately, no real world numbers. > > We think the global atomic counter could be an issue for performance > so I'm trying to solve the problem. So, umm, we don't actually know if the patch is useful to anyone? Some more measurements would help things along, please. -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: email@kvack.org