From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-3.8 required=3.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,HEADER_FROM_DIFFERENT_DOMAINS,MAILING_LIST_MULTI,SPF_HELO_NONE, SPF_PASS autolearn=no autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 2CCE4C433DB for ; Sat, 20 Feb 2021 01:19:02 +0000 (UTC) Received: from kanga.kvack.org (kanga.kvack.org [205.233.56.17]) by mail.kernel.org (Postfix) with ESMTP id B35A464E57 for ; Sat, 20 Feb 2021 01:19:01 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org B35A464E57 Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=intel.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=owner-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix) id 2669E6B0005; Fri, 19 Feb 2021 20:19:01 -0500 (EST) Received: by kanga.kvack.org (Postfix, from userid 40) id 1F2586B006C; Fri, 19 Feb 2021 20:19:01 -0500 (EST) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id 0E0236B006E; Fri, 19 Feb 2021 20:19:01 -0500 (EST) X-Delivered-To: linux-mm@kvack.org Received: from forelay.hostedemail.com (smtprelay0193.hostedemail.com [216.40.44.193]) by kanga.kvack.org (Postfix) with ESMTP id E26646B0005 for ; Fri, 19 Feb 2021 20:19:00 -0500 (EST) Received: from smtpin23.hostedemail.com (10.5.19.251.rfc1918.com [10.5.19.251]) by forelay05.hostedemail.com (Postfix) with ESMTP id 9DAC4180AC177 for ; Sat, 20 Feb 2021 01:19:00 +0000 (UTC) X-FDA: 77836887240.23.4E17DF8 Received: from mail-ed1-f52.google.com (mail-ed1-f52.google.com [209.85.208.52]) by imf27.hostedemail.com (Postfix) with ESMTP id E033980192C6 for ; Sat, 20 Feb 2021 01:18:55 +0000 (UTC) Received: by mail-ed1-f52.google.com with SMTP id o3so12928516edv.4 for ; Fri, 19 Feb 2021 17:18:59 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=intel-com.20150623.gappssmtp.com; s=20150623; h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc; bh=8I/++zqgWe2vZKzB2I3EDGwNh2FcyG9ignZ4r6L3DZY=; b=w4aQQRje/jerTeZglVuAtAD8RzQVEbuKDYLK7xJfV681Dqh3LGBt/VgLQ0AIvOdX/O Q3RGmg6wctRTBw2/2RqK8xyi32T4YzA/IYjuJDALNl2Kd6+KVqOiFhE46UDZVJTTEd6t uZDswIbehTiCRBllrGM3L8S/XNMokHkIfTi+4PtkdTi3CM9bUjnZMUlWxALfm4/uvhTY ewdBHWR2oDbEGBRaO7E4tIReG4brzSwgqP2R7Sn/A5h6YdU2Sqq02aEQUceYNQe1NrNJ X3xtH8rHRDzz8/KIixO4PiCwgj5i7OFR1jEwhZgMNE/QDaqquna64D+1T0PW2PSgr5bJ y5yQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=8I/++zqgWe2vZKzB2I3EDGwNh2FcyG9ignZ4r6L3DZY=; b=j4B6jySvO0ZhuTC4kahX2BedVOTnNkdsWvUtJ5+vIVPzwjb5DhMiLJ+1BhpHpgIbD/ rIXVcAlbY/VD0lNdzn4FsvRH+UoBt6N83vmzsJ5KKcF/QJFhXxbTg0HIrABAQPVZGi2Q B2Q5miX/MaPvfA3gzs4yxGXIcXwjoPmYj4y9ZTqfYm9pZ3mM7vYiTnrWuBGtTpmyldFv 87jNAX6a+LSYJRay7oqztmSDVhhy65r/ldG2Jiv1WWsBEN5666G511bi1h3ioyZ7X01y UnYMCFFGNP3O1QFQKGc2RRo1LGVxznylEGib4HElbb7IUihBiurEW2F5JeJty1CWccuG iXIg== X-Gm-Message-State: AOAM532ygh2czFM7VaIKY5fE//CU1UDHIz1DDUUotSXqpasbNlHGLifV SkWQ7SD12q0Y3jr9hF3SRicYOwwmudYrbH8gNRHmyA== X-Google-Smtp-Source: ABdhPJwGcgVw8LZGjLQwH8vQSWkjNN7/CtuprkImbJhEapBBfAjZp4GCVOHgmxwHw9iipUama/sr+PVi/PiU+Ak2KEA= X-Received: by 2002:aa7:d315:: with SMTP id p21mr10307428edq.300.1613783938776; Fri, 19 Feb 2021 17:18:58 -0800 (PST) MIME-Version: 1.0 References: <20201208172901.17384-1-joao.m.martins@oracle.com> In-Reply-To: <20201208172901.17384-1-joao.m.martins@oracle.com> From: Dan Williams Date: Fri, 19 Feb 2021 17:18:51 -0800 Message-ID: Subject: Re: [PATCH RFC 0/9] mm, sparse-vmemmap: Introduce compound pagemaps To: Joao Martins Cc: Linux MM , Ira Weiny , linux-nvdimm , Matthew Wilcox , Jason Gunthorpe , Jane Chu , Muchun Song , Mike Kravetz , Andrew Morton Content-Type: text/plain; charset="UTF-8" X-Stat-Signature: j63bqdy768h3c3db6ruuq6s1pzehpgi1 X-Rspamd-Server: rspam02 X-Rspamd-Queue-Id: E033980192C6 Received-SPF: none (intel.com>: No applicable sender policy available) receiver=imf27; identity=mailfrom; envelope-from=""; helo=mail-ed1-f52.google.com; client-ip=209.85.208.52 X-HE-DKIM-Result: pass/pass X-HE-Tag: 1613783935-780389 X-Bogosity: Ham, tests=bogofilter, spamicity=0.000000, version=1.2.4 Sender: owner-linux-mm@kvack.org Precedence: bulk X-Loop: owner-majordomo@kvack.org List-ID: On Tue, Dec 8, 2020 at 9:32 AM Joao Martins wrote: > > Hey, > > This small series, attempts at minimizing 'struct page' overhead by > pursuing a similar approach as Muchun Song series "Free some vmemmap > pages of hugetlb page"[0] but applied to devmap/ZONE_DEVICE. > > [0] https://lore.kernel.org/linux-mm/20201130151838.11208-1-songmuchun@bytedance.com/ Clever! > > The link above describes it quite nicely, but the idea is to reuse tail > page vmemmap areas, particular the area which only describes tail pages. > So a vmemmap page describes 64 struct pages, and the first page for a given > ZONE_DEVICE vmemmap would contain the head page and 63 tail pages. The second > vmemmap page would contain only tail pages, and that's what gets reused across > the rest of the subsection/section. The bigger the page size, the bigger the > savings (2M hpage -> save 6 vmemmap pages; 1G hpage -> save 4094 vmemmap pages). > > In terms of savings, per 1Tb of memory, the struct page cost would go down > with compound pagemap: > > * with 2M pages we lose 4G instead of 16G (0.39% instead of 1.5% of total memory) > * with 1G pages we lose 8MB instead of 16G (0.0007% instead of 1.5% of total memory) Nice! > > Along the way I've extended it past 'struct page' overhead *trying* to address a > few performance issues we knew about for pmem, specifically on the > {pin,get}_user_pages* function family with device-dax vmas which are really > slow even of the fast variants. THP is great on -fast variants but all except > hugetlbfs perform rather poorly on non-fast gup. > > So to summarize what the series does: > > Patches 1-5: Much like Muchun series, we reuse tail page areas across a given > page size (namely @align was referred by remaining memremap/dax code) and > enabling of memremap to initialize the ZONE_DEVICE pages as compound pages or a > given @align order. The main difference though, is that contrary to the hugetlbfs > series, there's no vmemmap for the area, because we are onlining it. IOW no > freeing of pages of already initialized vmemmap like the case for hugetlbfs, > which simplifies the logic (besides not being arch-specific). After these, > there's quite visible region bootstrap of pmem memmap given that we would > initialize fewer struct pages depending on the page size. > > NVDIMM namespace bootstrap improves from ~750ms to ~190ms/<=1ms on emulated NVDIMMs > with 2M and 1G respectivally. The net gain in improvement is similarly observed > in proportion when running on actual NVDIMMs. I > > Patch 6 - 8: Optimize grabbing/release a page refcount changes given that we > are working with compound pages i.e. we do 1 increment/decrement to the head > page for a given set of N subpages compared as opposed to N individual writes. > {get,pin}_user_pages_fast() for zone_device with compound pagemap consequently > improves considerably, and unpin_user_pages() improves as well when passed a > set of consecutive pages: > > before after > (get_user_pages_fast 1G;2M page size) ~75k us -> ~3.2k ; ~5.2k us > (pin_user_pages_fast 1G;2M page size) ~125k us -> ~3.4k ; ~5.5k us Compelling! > > The RDMA patch (patch 8/9) is to demonstrate the improvement for an existing > user. For unpin_user_pages() we have an additional test to demonstrate the > improvement. The test performs MR reg/unreg continuously and measuring its > rate for a given period. So essentially ib_mem_get and ib_mem_release being > stress tested which at the end of day means: pin_user_pages_longterm() and > unpin_user_pages() for a scatterlist: > > Before: > 159 rounds in 5.027 sec: 31617.923 usec / round (device-dax) > 466 rounds in 5.009 sec: 10748.456 usec / round (hugetlbfs) > > After: > 305 rounds in 5.010 sec: 16426.047 usec / round (device-dax) > 1073 rounds in 5.004 sec: 4663.622 usec / round (hugetlbfs) Why does hugetlbfs get faster for a ZONE_DEVICE change? Might answer that question myself when I get to patch 8. > > Patch 9: Improves {pin,get}_user_pages() and its longterm counterpart. It > is very experimental, and I imported most of follow_hugetlb_page(), except > that we do the same trick as gup-fast. In doing the patch I feel this batching > should live in follow_page_mask() and having that being changed to return a set > of pages/something-else when walking over PMD/PUDs for THP / devmap pages. This > patch then brings the previous test of mr reg/unreg (above) on parity > between device-dax and hugetlbfs. > > Some of the patches are a little fresh/WIP (specially patch 3 and 9) and we are > still running tests. Hence the RFC, asking for comments and general direction > of the work before continuing. Will go look at the code, but I don't see anything scary conceptually here. The fact that pfn_to_page() does not need to change is among the most compelling features of this approach.