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 11BFAC47258 for ; Wed, 31 Jan 2024 13:17:00 +0000 (UTC) Received: by kanga.kvack.org (Postfix) id 9B0016B0082; Wed, 31 Jan 2024 08:16:59 -0500 (EST) Received: by kanga.kvack.org (Postfix, from userid 40) id 963096B0083; Wed, 31 Jan 2024 08:16:59 -0500 (EST) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id 801436B0085; Wed, 31 Jan 2024 08:16:59 -0500 (EST) X-Delivered-To: linux-mm@kvack.org Received: from relay.hostedemail.com (smtprelay0016.hostedemail.com [216.40.44.16]) by kanga.kvack.org (Postfix) with ESMTP id 6D6CB6B0082 for ; Wed, 31 Jan 2024 08:16:59 -0500 (EST) Received: from smtpin29.hostedemail.com (a10.router.float.18 [10.200.18.1]) by unirelay07.hostedemail.com (Postfix) with ESMTP id 162B6160D24 for ; Wed, 31 Jan 2024 13:16:59 +0000 (UTC) X-FDA: 81739656558.29.8E03497 Received: from foss.arm.com (foss.arm.com [217.140.110.172]) by imf21.hostedemail.com (Postfix) with ESMTP id CE27D1C001F for ; Wed, 31 Jan 2024 13:16:56 +0000 (UTC) Authentication-Results: imf21.hostedemail.com; dkim=none; spf=pass (imf21.hostedemail.com: domain of ryan.roberts@arm.com designates 217.140.110.172 as permitted sender) smtp.mailfrom=ryan.roberts@arm.com; dmarc=pass (policy=none) header.from=arm.com ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=hostedemail.com; s=arc-20220608; t=1706707017; 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; bh=qYx4UZW/zJMD/qtYwjNRYQvgnr/y2IQogBw7jLPz8qI=; b=MX/Sgw7gS0n14FV81VXzFQz99oYlEMBYMwYqMHD1np9wETyU25Hvde4Xa+AzZw2DhztJnT QQQgyvIBWQeoJBw2cKExeVFuzn+UlF+EOwUn1BWJX+xPkkhueJmYaDEMLsxlS9dghwsn5x wPujyLM4obnILOuHiDvtJcy0sH3cm80= ARC-Seal: i=1; s=arc-20220608; d=hostedemail.com; t=1706707017; a=rsa-sha256; cv=none; b=IrDSMk4RaNxcubZrnb0raFlvrAerZhMIrSCwaNDRIhNkNQyY0DAvqs3OFzOAaqO8hiMJKt VS2rsN6AwOwuGA4O3+T6A09NxjDaxOwZiUZENe+cl2FasgwSi7k+2ouvyzDbGVZy3CI6I0 42MydhKLA9qkIIbNiF7F9n97aF+VBEs= ARC-Authentication-Results: i=1; imf21.hostedemail.com; dkim=none; spf=pass (imf21.hostedemail.com: domain of ryan.roberts@arm.com designates 217.140.110.172 as permitted sender) smtp.mailfrom=ryan.roberts@arm.com; dmarc=pass (policy=none) header.from=arm.com Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.121.207.14]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id EC38DDA7; Wed, 31 Jan 2024 05:17:38 -0800 (PST) Received: from [10.57.79.60] (unknown [10.57.79.60]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPSA id 599E13F738; Wed, 31 Jan 2024 05:16:50 -0800 (PST) Message-ID: Date: Wed, 31 Jan 2024 13:16:48 +0000 MIME-Version: 1.0 User-Agent: Mozilla Thunderbird Subject: Re: [PATCH v3 00/15] mm/memory: optimize fork() with PTE-mapped THP Content-Language: en-GB To: David Hildenbrand , linux-kernel@vger.kernel.org Cc: linux-mm@kvack.org, Andrew Morton , Matthew Wilcox , Russell King , Catalin Marinas , Will Deacon , Dinh Nguyen , Michael Ellerman , Nicholas Piggin , Christophe Leroy , "Aneesh Kumar K.V" , "Naveen N. Rao" , Paul Walmsley , Palmer Dabbelt , Albert Ou , Alexander Gordeev , Gerald Schaefer , Heiko Carstens , Vasily Gorbik , Christian Borntraeger , Sven Schnelle , "David S. Miller" , linux-arm-kernel@lists.infradead.org, linuxppc-dev@lists.ozlabs.org, linux-riscv@lists.infradead.org, linux-s390@vger.kernel.org, sparclinux@vger.kernel.org References: <20240129124649.189745-1-david@redhat.com> <57eb82c7-4816-42a2-b5ab-cc221e289b21@arm.com> <714d0930-2202-48b6-9728-d248f820325e@arm.com> <30718fc8-15cf-41e4-922c-5cdbf00a0840@redhat.com> From: Ryan Roberts In-Reply-To: <30718fc8-15cf-41e4-922c-5cdbf00a0840@redhat.com> Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit X-Stat-Signature: pcoz3d4z6w6jdqgcobaejwp3pecd4bau X-Rspamd-Server: rspam10 X-Rspamd-Queue-Id: CE27D1C001F X-Rspam-User: X-HE-Tag: 1706707016-242931 X-HE-Meta: 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 TnwxLeXR 486t3fHtL2YkPInjoNHL4/8+LGWE0laD8SIqa6MMIyylcPqDKZMrskH/7S5TLqkPHbEo0YFIWsLGgUx8NaPuAX58cu8O9ESHAVMDQMSd8hl7UN4oBFSiC1xcW7j3d/0XGZ8Mn05KMyyULDGFndH1+tgJJTg== 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 31/01/2024 12:56, David Hildenbrand wrote: > On 31.01.24 13:37, Ryan Roberts wrote: >> On 31/01/2024 11:49, Ryan Roberts wrote: >>> On 31/01/2024 11:28, David Hildenbrand wrote: >>>> On 31.01.24 12:16, Ryan Roberts wrote: >>>>> On 31/01/2024 11:06, David Hildenbrand wrote: >>>>>> On 31.01.24 11:43, Ryan Roberts wrote: >>>>>>> On 29/01/2024 12:46, David Hildenbrand wrote: >>>>>>>> Now that the rmap overhaul[1] is upstream that provides a clean interface >>>>>>>> for rmap batching, let's implement PTE batching during fork when processing >>>>>>>> PTE-mapped THPs. >>>>>>>> >>>>>>>> This series is partially based on Ryan's previous work[2] to implement >>>>>>>> cont-pte support on arm64, but its a complete rewrite based on [1] to >>>>>>>> optimize all architectures independent of any such PTE bits, and to >>>>>>>> use the new rmap batching functions that simplify the code and prepare >>>>>>>> for further rmap accounting changes. >>>>>>>> >>>>>>>> We collect consecutive PTEs that map consecutive pages of the same large >>>>>>>> folio, making sure that the other PTE bits are compatible, and (a) adjust >>>>>>>> the refcount only once per batch, (b) call rmap handling functions only >>>>>>>> once per batch and (c) perform batch PTE setting/updates. >>>>>>>> >>>>>>>> While this series should be beneficial for adding cont-pte support on >>>>>>>> ARM64[2], it's one of the requirements for maintaining a total mapcount[3] >>>>>>>> for large folios with minimal added overhead and further changes[4] that >>>>>>>> build up on top of the total mapcount. >>>>>>>> >>>>>>>> Independent of all that, this series results in a speedup during fork with >>>>>>>> PTE-mapped THP, which is the default with THPs that are smaller than a PMD >>>>>>>> (for example, 16KiB to 1024KiB mTHPs for anonymous memory[5]). >>>>>>>> >>>>>>>> On an Intel Xeon Silver 4210R CPU, fork'ing with 1GiB of PTE-mapped folios >>>>>>>> of the same size (stddev < 1%) results in the following runtimes >>>>>>>> for fork() (shorter is better): >>>>>>>> >>>>>>>> Folio Size | v6.8-rc1 |      New | Change >>>>>>>> ------------------------------------------ >>>>>>>>          4KiB | 0.014328 | 0.014035 |   - 2% >>>>>>>>         16KiB | 0.014263 | 0.01196  |   -16% >>>>>>>>         32KiB | 0.014334 | 0.01094  |   -24% >>>>>>>>         64KiB | 0.014046 | 0.010444 |   -26% >>>>>>>>        128KiB | 0.014011 | 0.010063 |   -28% >>>>>>>>        256KiB | 0.013993 | 0.009938 |   -29% >>>>>>>>        512KiB | 0.013983 | 0.00985  |   -30% >>>>>>>>       1024KiB | 0.013986 | 0.00982  |   -30% >>>>>>>>       2048KiB | 0.014305 | 0.010076 |   -30% >>>>>>> >>>>>>> Just a heads up that I'm seeing some strange results on Apple M2. Fork for >>>>>>> order-0 is seemingly costing ~17% more. I'm using GCC 13.2 and was pretty >>>>>>> sure I >>>>>>> didn't see this problem with version 1; although that was on a different >>>>>>> baseline and I've thrown the numbers away so will rerun and try to debug >>>>>>> this. >>> >>> Numbers for v1 of the series, both on top of 6.8-rc1 and rebased to the same >>> mm-unstable base as v3 of the series (first 2 rows are from what I just posted >>> for context): >>> >>> | kernel             |   mean_rel |   std_rel | >>> |:-------------------|-----------:|----------:| >>> | mm-unstabe (base)  |       0.0% |      1.1% | >>> | mm-unstable + v3   |      16.7% |      0.8% | >>> | mm-unstable + v1   |      -2.5% |      1.7% | >>> | v6.8-rc1 + v1      |      -6.6% |      1.1% | >>> >>> So all looks good with v1. And seems to suggest mm-unstable has regressed by ~4% >>> vs v6.8-rc1. Is this really a useful benchmark? Does the raw performance of >>> fork() syscall really matter? Evidence suggests its moving all over the place - >>> breath on the code and it changes - not a great place to be when using the test >>> for gating purposes! >>> >>> Still with the old tests - I'll move to the new ones now. >>> >>> >>>>>>> >>>>>> >>>>>> So far, on my x86 tests (Intel, AMD EPYC), I was not able to observe this. >>>>>> fork() for order-0 was consistently effectively unchanged. Do you observe >>>>>> that >>>>>> on other ARM systems as well? >>>>> >>>>> Nope; running the exact same kernel binary and user space on Altra, I see >>>>> sensible numbers; >>>>> >>>>> fork order-0: -1.3% >>>>> fork order-9: -7.6% >>>>> dontneed order-0: -0.5% >>>>> dontneed order-9: 0.1% >>>>> munmap order-0: 0.0% >>>>> munmap order-9: -67.9% >>>>> >>>>> So I guess some pipelining issue that causes the M2 to stall more? >>>> >>>> With one effective added folio_test_large(), it could only be a code layout >>>> problem? Or the compiler does something stupid, but you say that you run the >>>> exact same kernel binary, so that doesn't make sense. >>> >>> Yup, same binary. We know this code is very sensitive - 1 cycle makes a big >>> difference. So could easily be code layout, branch prediction, etc... >>> >>>> >>>> I'm also surprised about the dontneed vs. munmap numbers. >>> >>> You mean the ones for Altra that I posted? (I didn't post any for M2). The altra >>> numbers look ok to me; dontneed has no change, and munmap has no change for >>> order-0 and is massively improved for order-9. >>> >>>   Doesn't make any sense >>>> (again, there was this VMA merging problem but it would still allow for >>>> batching >>>> within a single VMA that spans exactly one large folio). >>>> >>>> What are you using as baseline? Really just mm-unstable vs. >>>> mm-unstable+patches? >>> >>> yes. except for "v6.8-rc1 + v1" above. >>> >>>> >>>> Let's see if the new test changes the numbers you measure. >> >> Nope: looks the same. I've taken my test harness out of the picture and done >> everything manually from the ground up, with the old tests and the new. Headline >> is that I see similar numbers from both. > > I took me a while to get really reproducible numbers on Intel. Most importantly: > * Set a fixed CPU frequency, disabling any boost and avoiding any >   thermal throttling. > * Pin the test to CPUs and set a nice level. I'm already pinning the test to cpu 0. But for M2, at least, I'm running in a VM on top of macos, and I don't have a mechanism to pin the QEMU threads to the physical CPUs. Anyway, I don't think these are problems because for a given kernel build I can accurately repro numbers. > > Another thing is, to avoid systems where you can have NUMA effects within a > single socket. Otherwise, memory access latency is just random and depends on > what the buddy enjoys giving you. Yep; same. M2 is 1 NUMA node. On Altra, I'm disabling the second NUMA node to remove those effects. > > But you seem to get the same +17 even after reboots, so that indicates that the > CPU is not happy about the code for some reason. And the weird thing is, that > nothing significantly changed for order-0 folios between v1 and v3 that could > explain any of this. > > I'm not worried about 5% or so, nobody cares. But it would be good to have at > least an explanation why only that system shows +17%. Yep understood. > >> >> Some details: >>   - I'm running for 10 seconds then averaging the output > > Same here. > >>   - test is bimodal; first run (of 10 seconds) after boot is a bit faster on >>     average (up to 10%) than the rest; I could guess this is due to the memory >>     being allocated more contiguously the first few times through, so struct >>     pages have better locality, but that's a guess. > > I think it also has to do with the PCP lists, and the high-pcp auto tuning (I > played with disabling that). Running on a freshly booted system gave me > reproducible results. > > But yes: I was observing something similar on AMD EPYC, where you get > consecutive pages from the buddy, but once you allocate from the PCP it might no > longer be consecutive. > >>   - test is 5-10% slower when output is printed to terminal vs when redirected to >>     file. I've always effectively been redirecting. Not sure if this overhead >>     could start to dominate the regression and that's why you don't see it? > > That's weird, because we don't print while measuring? Anyhow, 5/10% variance on > some system is not the end of the world. I imagine its cache effects? More work to do to print the output could be evicting some code that's in the benchmark path? > >> >> I'm inclined to run this test for the last N kernel releases and if the number >> moves around significantly, conclude that these tests don't really matter. >> Otherwise its an exercise in randomly refactoring code until it works well, but >> that's just overfitting to the compiler and hw. What do you think? > > Personally, I wouldn't lose sleep if you see weird, unexplainable behavior on > some system (not even architecture!). Trying to optimize for that would indeed > be random refactorings. > > But I would not be so fast to say that "these tests don't really matter" and > then go wild and degrade them as much as you want. There are use cases that care > about fork performance especially with order-0 pages -- such as Redis. Indeed. But also remember that my fork baseline time is ~2.5ms, and I think you said yours was 14ms :) I'll continue to mess around with it until the end of the day. But I'm not making any headway, then I'll change tack; I'll just measure the performance of my contpte changes using your fork/zap stuff as the baseline and post based on that.