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 Received: from kanga.kvack.org (kanga.kvack.org [205.233.56.17]) by smtp.lore.kernel.org (Postfix) with ESMTP id A2435C4707B for ; Wed, 10 Jan 2024 11:38:59 +0000 (UTC) Received: by kanga.kvack.org (Postfix) id 1A8868D0003; Wed, 10 Jan 2024 06:38:59 -0500 (EST) Received: by kanga.kvack.org (Postfix, from userid 40) id 1582C8D0001; Wed, 10 Jan 2024 06:38:59 -0500 (EST) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id 020D78D0003; Wed, 10 Jan 2024 06:38:58 -0500 (EST) X-Delivered-To: linux-mm@kvack.org Received: from relay.hostedemail.com (smtprelay0010.hostedemail.com [216.40.44.10]) by kanga.kvack.org (Postfix) with ESMTP id E35618D0001 for ; Wed, 10 Jan 2024 06:38:58 -0500 (EST) Received: from smtpin09.hostedemail.com (a10.router.float.18 [10.200.18.1]) by unirelay08.hostedemail.com (Postfix) with ESMTP id B4362140473 for ; Wed, 10 Jan 2024 11:38:58 +0000 (UTC) X-FDA: 81663204756.09.4FBC729 Received: from mail-vk1-f180.google.com (mail-vk1-f180.google.com [209.85.221.180]) by imf02.hostedemail.com (Postfix) with ESMTP id 003AA80005 for ; Wed, 10 Jan 2024 11:38:56 +0000 (UTC) Authentication-Results: imf02.hostedemail.com; dkim=pass header.d=gmail.com header.s=20230601 header.b=eNvIi0x7; spf=pass (imf02.hostedemail.com: domain of 21cnbao@gmail.com designates 209.85.221.180 as permitted sender) smtp.mailfrom=21cnbao@gmail.com; dmarc=pass (policy=none) header.from=gmail.com ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=hostedemail.com; s=arc-20220608; t=1704886737; h=from:from:sender:reply-to:subject:subject:date:date: message-id:message-id:to:to:cc:cc:mime-version:mime-version: content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references:dkim-signature; bh=CmA0k4RnSX44/LB+bsZGRqOWuXutbPeONNXGQBUs9c0=; b=LUsC5dT9IUpcsfkCWkxyNugTRSstDLu7qIjCRCMWbUdBPWD0ZQp9VAIeIY41poP0hw21MS m9cOeHwulOGBiCTDYR16NLOMEl1DNFsYEbDaF+GEqe7/LWehuYn8d0Fd9f8NijHiQ0ctKt ibnwcY1rkyeVbN8j3LeLleVouZjFmu8= ARC-Seal: i=1; s=arc-20220608; d=hostedemail.com; t=1704886737; a=rsa-sha256; cv=none; b=UJQI0skKWALcna+NqzOKpIXXsygau3pNZWTyVEmqgB4CTYWb6IcoYUlKZJVdAojqVmpor4 p7swjA9c4pxX8NDSlpM0ZIg3tI4+3v+QljYBpeTYWTHvNkL4JUQih2NsoPID062OFB0WkK UN18Z2jroDzOB64SI3wLc/uz9O0sLzw= ARC-Authentication-Results: i=1; imf02.hostedemail.com; dkim=pass header.d=gmail.com header.s=20230601 header.b=eNvIi0x7; spf=pass (imf02.hostedemail.com: domain of 21cnbao@gmail.com designates 209.85.221.180 as permitted sender) smtp.mailfrom=21cnbao@gmail.com; dmarc=pass (policy=none) header.from=gmail.com Received: by mail-vk1-f180.google.com with SMTP id 71dfb90a1353d-4b731a13d76so737784e0c.1 for ; Wed, 10 Jan 2024 03:38:56 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20230601; t=1704886736; x=1705491536; darn=kvack.org; h=content-transfer-encoding:cc:to:subject:message-id:date:from :in-reply-to:references:mime-version:from:to:cc:subject:date :message-id:reply-to; bh=CmA0k4RnSX44/LB+bsZGRqOWuXutbPeONNXGQBUs9c0=; b=eNvIi0x7NePAa8FXKsN1pO0apaGvYkvGNW3N/xMyTUGyKe9zVXDajCs4UP/2ibveCx tEZmDzDHwfhb/wGq7bPFc3pqnHF40dGeBkgKqhBD1GRKpqszJVdAleYObwTrhGf2EeM1 ix8smBsw9p99JajIC0dNwqv7h1N0sm+pWaPfBycAcZSIYgzZogzgj0eiDpzU0QTTp9Ue c2kVvywarjwLilvijvQ6xUrtZXD4l8fUqoKAFWGdevoAC29xAaYXo3waCtAk/9WGtlSW zQnVuFWcxK8xaGb0fWX93pRwhaN1m46y2i5PqJM5F86ikcltfz71Lp1eat13IugXHtQe 5wLg== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1704886736; x=1705491536; h=content-transfer-encoding:cc:to:subject:message-id:date:from :in-reply-to:references:mime-version:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=CmA0k4RnSX44/LB+bsZGRqOWuXutbPeONNXGQBUs9c0=; b=IFhW0tD1oElcJWVnzdfN6AANw2jWXLJNfSgenVSahcYoJaOlgmOIZqKfU2lsVvFieu V0kcD3TjABqDM9GrVLhGdohhPbVTsys3NAxRdk+Yg8aDEN9OkdDzH492XPHMib3eK0zg 8uYWiXB1TfFVop6ZORTiiN1D69ZBr+vGovhxB6Weu+im5mrlwSojwrS/qhlvDCxtLmpy jYzrwaJ19Zfhtz6t91nWPqg35ypPGastjB4dYCmJO8IKFLsiDCEEKdVfFsco40MfZpK1 ziNEEJdzzcVhp+p2wgiV8LIQCJLAuX8mI5Uh1X7KsJSo0POI/6//fH9HDJdccW0uj8tA 3vBg== X-Gm-Message-State: AOJu0Yy1tjdV9tji9kAS3H1VhATOaw4tKimeBQmWFcFSwnyqr+lVFIr8 ogcn1fgPgHYwsxnunThX3+09z+rWryYi9Zbcko0= X-Google-Smtp-Source: AGHT+IGScEEQtbsVEfA+HzKk8qz2CBRMVwQwVQwuOesBD6rdu67k+jC41meiJ7WDgkl1dsId/GRzFZWfiriYiUxLOp0= X-Received: by 2002:a05:6122:1b0d:b0:4b7:18b3:b151 with SMTP id er13-20020a0561221b0d00b004b718b3b151mr417013vkb.24.1704886735946; Wed, 10 Jan 2024 03:38:55 -0800 (PST) MIME-Version: 1.0 References: <20240102153828.1002295-1-ryan.roberts@arm.com> <4e7445a0-acc9-487f-999f-a2b6d03d265e@nvidia.com> <3bd5e4a3-9f67-4483-9a0e-9abb5eb783cd@arm.com> <94ebe62b-5f55-4be9-b464-4105b4692496@arm.com> <68d5ce7e-6587-47c6-bd0f-988adf5d92a4@arm.com> <974a2670-7fa9-425e-921e-8d54a596e6cf@arm.com> <6c77f143-9c2c-4d17-9a2a-d69d9adf2eea@arm.com> In-Reply-To: <6c77f143-9c2c-4d17-9a2a-d69d9adf2eea@arm.com> From: Barry Song <21cnbao@gmail.com> Date: Wed, 10 Jan 2024 19:38:44 +0800 Message-ID: Subject: Re: [RFC PATCH v1] tools/mm: Add thpmaps script to dump THP usage info To: Ryan Roberts Cc: David Hildenbrand , John Hubbard , Andrew Morton , Zenghui Yu , Matthew Wilcox , Kefeng Wang , Zi Yan , Alistair Popple , linux-mm@kvack.org Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable X-Rspamd-Queue-Id: 003AA80005 X-Rspam-User: X-Stat-Signature: as3i7z4z1uqbkx3cfd5xzr5wzt7ba7q3 X-Rspamd-Server: rspam03 X-HE-Tag: 1704886736-599998 X-HE-Meta: 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 /roUlcsf Efj2pYGdJiC8FSJT959C0tf+nlSNoU2gGW0orHGuUVpuYpYpg9ISsM9+wv/IFQ7xduVzhbur3FueiM8tDuCFojVkTjC0utE5Hl475ETz+kqIE0OBjJdwzp9SA5qmpZUEX9Tz8aC93MHTQPKwdBi1kKubttkL0EmCRy/6nKwkSxZkAMIwrxcJ6ZaFBydordKzia2Yhrkh0yUvYrGSe/jcruQj/exhcDkSDlWLbGJ1pid6fVChqgn38e4qBlDDevBvTY2XcwnjPTSHu3g2c36muwy6rAxtuh2J2lu2WjUbPFgy0DIaT1oUOmDbsJcN7yE9kPs1P7LO2lxy5Id8FdNWa3AsMoTPsWMQqdDNPIvburkOQQf1wXZpX26yNCS2+BSdni+xzxl7ZGHr4h9i9mP2oQwc97XVFbaV4FLVam58ChdV5ByCJYl5BFvamrA== 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: List-Subscribe: List-Unsubscribe: On Wed, Jan 10, 2024 at 7:21=E2=80=AFPM Ryan Roberts = wrote: > > On 10/01/2024 11:00, David Hildenbrand wrote: > > On 10.01.24 11:55, Ryan Roberts wrote: > >> On 10/01/2024 10:42, David Hildenbrand wrote: > >>> On 10.01.24 11:38, Ryan Roberts wrote: > >>>> On 10/01/2024 10:30, Barry Song wrote: > >>>>> On Wed, Jan 10, 2024 at 6:23=E2=80=AFPM Ryan Roberts wrote: > >>>>>> > >>>>>> On 10/01/2024 09:09, Barry Song wrote: > >>>>>>> On Wed, Jan 10, 2024 at 4:58=E2=80=AFPM Ryan Roberts wrote: > >>>>>>>> > >>>>>>>> On 10/01/2024 08:02, Barry Song wrote: > >>>>>>>>> On Wed, Jan 10, 2024 at 12:16=E2=80=AFPM John Hubbard wrote: > >>>>>>>>>> > >>>>>>>>>> On 1/9/24 19:51, Barry Song wrote: > >>>>>>>>>>> On Wed, Jan 10, 2024 at 11:35=E2=80=AFAM John Hubbard > >>>>>>>>>>> wrote: > >>>>>>>>>> ... > >>>>>>>>>>>> Hi Ryan, > >>>>>>>>>>>> > >>>>>>>>>>>> One thing that immediately came up during some recent testin= g of mTHP > >>>>>>>>>>>> on arm64: the pid requirement is sometimes a little awkward.= I'm > >>>>>>>>>>>> running > >>>>>>>>>>>> tests on a machine at a time for now, inside various contain= ers and > >>>>>>>>>>>> such, and it would be nice if there were an easy way to get = some > >>>>>>>>>>>> numbers > >>>>>>>>>>>> for the mTHPs across the whole machine. > >>>>>>>> > >>>>>>>> Just to confirm, you're expecting these "global" stats be truely= global > >>>>>>>> and not > >>>>>>>> per-container? (asking because you exploicitly mentioned being i= n a > >>>>>>>> container). > >>>>>>>> If you want per-container, then you can probably just create the= container > >>>>>>>> in a > >>>>>>>> cgroup? > >>>>>>>> > >>>>>>>>>>>> > >>>>>>>>>>>> I'm not sure if that changes anything about thpmaps here. Pr= obably > >>>>>>>>>>>> this is fine as-is. But I wanted to give some initial reacti= ons from > >>>>>>>>>>>> just some quick runs: the global state would be convenient. > >>>>>>>> > >>>>>>>> Thanks for taking this for a spin! Appreciate the feedback. > >>>>>>>> > >>>>>>>>>>> > >>>>>>>>>>> +1. but this seems to be impossible by scanning pagemap? > >>>>>>>>>>> so may we add this statistics information in kernel just like > >>>>>>>>>>> /proc/meminfo or a separate /proc/mthp_info? > >>>>>>>>>>> > >>>>>>>>>> > >>>>>>>>>> Yes. From my perspective, it looks like the global stats are m= ore useful > >>>>>>>>>> initially, and the more detailed per-pid or per-cgroup stats a= re the > >>>>>>>>>> next level of investigation. So feels odd to start with the mo= re > >>>>>>>>>> detailed stats. > >>>>>>>>>> > >>>>>>>>> > >>>>>>>>> probably because this can be done without the modification of t= he kernel. > >>>>>>>> > >>>>>>>> Yes indeed, as John said in an earlier thread, my previous attem= pts to add > >>>>>>>> stats > >>>>>>>> directly in the kernel got pushback; DavidH was concerned that w= e don't > >>>>>>>> really > >>>>>>>> know exectly how to account mTHPs yet > >>>>>>>> (whole/partial/aligned/unaligned/per-size/etc) so didn't want to= end up > >>>>>>>> adding > >>>>>>>> the wrong ABI and having to maintain it forever. There has also = been some > >>>>>>>> pushback regarding adding more values to multi-value files in sy= sfs, so > >>>>>>>> David > >>>>>>>> was suggesting coming up with a whole new scheme at some point (= I know > >>>>>>>> /proc/meminfo isn't sysfs, but the equivalent files for NUMA nod= es and > >>>>>>>> cgroups > >>>>>>>> do live in sysfs). > >>>>>>>> > >>>>>>>> Anyway, this script was my attempt to 1) provide a short term so= lution > >>>>>>>> to the > >>>>>>>> "we need some stats" request and 2) provide a context in which t= o explore > >>>>>>>> what > >>>>>>>> the right stats are - this script can evolve without the ABI pro= blem. > >>>>>>>> > >>>>>>>>> The detailed per-pid or per-cgroup is still quite useful to my = case in > >>>>>>>>> which > >>>>>>>>> we set mTHP enabled/disabled and allowed sizes according to vma= types, > >>>>>>>>> eg. libc_malloc, java heaps etc. > >>>>>>>>> > >>>>>>>>> Different vma types can have different anon_name. So I can use = the > >>>>>>>>> detailed > >>>>>>>>> info to find out if specific VMAs have gotten mTHP properly and= how many > >>>>>>>>> they have gotten. > >>>>>>>>> > >>>>>>>>>> However, Ryan did clearly say, above, "In future we may wish t= o > >>>>>>>>>> introduce stats directly into the kernel (e.g. smaps or simila= r)". And > >>>>>>>>>> earlier he ran into some pushback on trying to set up /proc or= /sys > >>>>>>>>>> values because this is still such an early feature. > >>>>>>>>>> > >>>>>>>>>> I wonder if we could put the global stats in debugfs for now? = That's > >>>>>>>>>> specifically supposed to be a "we promise *not* to keep this A= BI stable" > >>>>>>>>>> location. > >>>>>>>> > >>>>>>>> Now that I think about it, I wonder if we can add a --global mod= e to the > >>>>>>>> script > >>>>>>>> (or just infer global when neither --pid nor --cgroup are provid= ed). I > >>>>>>>> think I > >>>>>>>> should be able to determine all the physical memory ranges from > >>>>>>>> /proc/iomem, > >>>>>>>> then grab all the info we need from /proc/kpageflags. We should = then be > >>>>>>>> able to > >>>>>>>> process it all in much the same way as for --pid/--cgroup and pr= ovide the > >>>>>>>> same > >>>>>>>> stats, but it will apply globally. What do you think? > >>>>>> > >>>>>> Having now thought about this for a few mins (in the shower, if an= yone wants > >>>>>> the > >>>>>> complete picture :) ), this won't quite work. This approach doesn'= t have the > >>>>>> virtual mapping information so the best it can do is tell us "how = many of > >>>>>> each > >>>>>> size of THP are allocated?" - it doesn't tell us anything about wh= ether they > >>>>>> are > >>>>>> fully or partially mapped or what their alignment is (all necessar= y if we > >>>>>> want > >>>>>> to know if they are contpte-mapped). So I don't think this approac= h is > >>>>>> going to > >>>>>> be particularly useful. > >>>>>> > >>>>>> And this is also the big problem if we want to gather stats inside= the > >>>>>> kernel; > >>>>>> if we want something equivalant to /proc/meminfo's > >>>>>> AnonHugePages/ShmemPmdMapped/FilePmdMapped, we need to consider no= t just the > >>>>>> allocation of the THP but also whether it is mapped. That's easy f= or > >>>>>> PMD-mappings, because there is only one entry to consider - when y= ou set it, > >>>>>> you > >>>>>> increment the number of PMD-mapped THPs, when you clear it, you de= crement. > >>>>>> But > >>>>>> for PTE-mappings it's harder; you know the size when you are mappi= ng so its > >>>>>> easy > >>>>>> to increment, but you can do a partial unmap, so you would need to= scan the > >>>>>> PTEs > >>>>>> to figure out if we are unmapping the first page of a previously > >>>>>> fully-PTE-mapped THP, which is expensive. We would need a cheap me= chanism to > >>>>>> determine "is this folio fully and contiguously mapped in at least= one > >>>>>> process?". > >>>>> > >>>>> as OPPO's approach I shared to you before is maintaining two mapcou= nt > >>>>> 1. entire map > >>>>> 2. subpage's map > >>>>> 3. if 1 and 2 both exist, it is DoubleMapped. > >>>>> > >>>>> This isn't a problem for us. and everytime if we do a partial unmap= , > >>>>> we have an explicit > >>>>> cont_pte split which will decrease the entire map and increase the > >>>>> subpage's mapcount. > >>>>> > >>>>> but its downside is that we expose this info to mm-core. > >>>> > >>>> OK, but I think we have a slightly more generic situation going on w= ith the > >>>> upstream; If I've understood correctly, you are using the PTE_CONT b= it in the > >>>> PTE to determne if its fully mapped? That works for your case where = you only > >>>> have 1 size of THP that you care about (contpte-size). But for the u= pstream, we > >>>> have multi-size THP so we can't use the PTE_CONT bit to determine if= its fully > >>>> mapped because we can only use that bit if the THP is at least 64K a= nd aligned, > >>>> and only on arm64. We would need a SW bit for this purpose, and the = mm would > >>>> need to update that SW bit for every PTE one the full -> partial map > >>>> transition. > >>> > >>> Oh no. Let's not make everything more complicated for the purpose of = some stats. > >>> > >> > >> Indeed, I was intending to argue *against* doing it this way. Fundamen= tally, if > >> we want to know what's fully mapped and what's not, then I don't see a= ny way > >> other than by scanning the page tables and we might as well do that in= user > >> space with this script. > >> > >> Although, I expect you will shortly make a proposal that is simple to = implement > >> and prove me wrong ;-) > > > > Unlikely :) As you said, once you have multiple folio sizes, it stops r= eally > > making sense. > > > > Assume you have a 128 kiB pageache folio, and half of that is mapped. Y= ou can > > set cont-pte bits on that half and all is fine. Or AMD can benefit from= it's > > optimizations without the cont-pte bit and everything is fine. > > Yes, but for debug and optimization, its useful to know when THPs are > fully/partially mapped, when they are unaligned etc. Anyway, the script d= oes > that for us, and I think we are tending towards agreement that there are > unlikely to be any cost benefits by moving it into the kernel. frequent partial unmap can defeat all purpose for us to use large folios. just imagine a large folio can soon be splitted after it is formed. we lose the performance gain and might get regression instead. and this can be very frequent, for example, one userspace heap management is releasing memory page by page. In our real product deployment, we might not care about the second partial unmapped, we do care about the first partial unmapped as we can use this to know if split has ever happened on this large folios. an partial unmappe= d subpage can be unlikely re-mapped back. so i guess 1st unmap is probably enough, at least for my product. I mean we care about if partial unmap has ever happened on a large folio more than ho= w they are exactly partially unmapped :-) > > > > > We want simple stats that tell us which folio sizes are actually alloca= ted. For > > everything else, just scan the process to figure out what exactly is go= ing on. > > > > Certainly that's much easier to do. But is it valuable? It might be if we= also > keep stats for the number of failures to allocate the various sizes - the= n we > can see what percentage of high order allocation attempts are successful,= which > is probably useful. > Thanks Barry