From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-pg0-f41.google.com ([74.125.83.41]:41292 "EHLO mail-pg0-f41.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751719AbeCPVcY (ORCPT ); Fri, 16 Mar 2018 17:32:24 -0400 Subject: Re: [RFC 2/2] page_frag_cache: Store metadata in struct page To: Matthew Wilcox , Alexander Duyck Cc: Alexander Duyck , linux-mm@vger.kernel.org, Netdev , Matthew Wilcox References: <20180315195329.7787-1-willy@infradead.org> <20180315195329.7787-3-willy@infradead.org> <20180316210500.GH27498@bombadil.infradead.org> From: Eric Dumazet Message-ID: <8e8c5cfa-ae76-e672-3da1-818b5df4448b@gmail.com> Date: Fri, 16 Mar 2018 14:32:22 -0700 MIME-Version: 1.0 In-Reply-To: <20180316210500.GH27498@bombadil.infradead.org> Content-Type: text/plain; charset=utf-8 Content-Language: en-US Content-Transfer-Encoding: 7bit Sender: netdev-owner@vger.kernel.org List-ID: On 03/16/2018 02:05 PM, Matthew Wilcox wrote: > > Obviously if the problem turns out to be the cacheline thrashing rather > than the call to page_to_virt, then this is pointless to test. > >> I won't be able to test the patches until next week, but I expect I >> will probably see a noticeable regression when performing a small >> packet routing test. > > I really appreciate you being willing to try this for me. I need to > get myself a dual-socket machine to test things like this. > It seems my prior mail/answer was lost. Issue is cacheline thrashing indeed, particularly on PowerPC (64 KB pages)