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 C27CDC77B75 for ; Fri, 5 May 2023 05:03:44 +0000 (UTC) Received: by kanga.kvack.org (Postfix) id D0B576B0075; Fri, 5 May 2023 01:03:43 -0400 (EDT) Received: by kanga.kvack.org (Postfix, from userid 40) id CBBCC6B0078; Fri, 5 May 2023 01:03:43 -0400 (EDT) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id BAA446B007B; Fri, 5 May 2023 01:03:43 -0400 (EDT) X-Delivered-To: linux-mm@kvack.org Received: from mga02.intel.com (mga02.intel.com [134.134.136.20]) by kanga.kvack.org (Postfix) with ESMTP id 71CBE6B0075 for ; Fri, 5 May 2023 01:03:43 -0400 (EDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=intel.com; i=@intel.com; q=dns/txt; s=Intel; t=1683263023; x=1714799023; h=from:to:cc:subject:in-reply-to:references:date: message-id:mime-version:content-transfer-encoding; bh=S4vyJou19RgYmzT65fd6fS70YFP8NSQ5x98Uj9rNLoc=; b=b7q2AbFlJucL4CBXx3QezDKJBqFMZTd8XzpeXilk4t/BPjzuWleUnA+G TZ1fbHw3d3wPQ0McerSqG9whNKvrKZT/qgyhta1s9yY29gdxEXoJEw6tv mTCEw5TY8FgYTdxSECTIxGaTQAZn3diDf3VYMoivAA274t40dUALed3DU gxydzl0U2gh34OqDeVPJs3Rg4S6SxjPqEtvGu/6odg28wGF9nKXqTBWvd OMTpwFfO1dzOW3jgkuVPqZ3NUZCg88qeNNTwR02hw8fG6Nvq78NmsuyBa yqa5dQFYjOhPLaJHlIeOubhYOKK8Vvwo2nmsEEzGKWyzb2ajIaOkoixXD g==; X-IronPort-AV: E=McAfee;i="6600,9927,10700"; a="338315035" X-IronPort-AV: E=Sophos;i="5.99,250,1677571200"; d="scan'208";a="338315035" Received: from fmsmga001.fm.intel.com ([10.253.24.23]) by orsmga101.jf.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 04 May 2023 22:03:41 -0700 X-ExtLoop1: 1 X-IronPort-AV: E=McAfee;i="6600,9927,10700"; a="841521484" X-IronPort-AV: E=Sophos;i="5.99,250,1677571200"; d="scan'208";a="841521484" Received: from yhuang6-desk2.sh.intel.com (HELO yhuang6-desk2.ccr.corp.intel.com) ([10.238.208.55]) by fmsmga001-auth.fm.intel.com with ESMTP/TLS/ECDHE-RSA-AES256-GCM-SHA384; 04 May 2023 22:03:36 -0700 From: "Huang, Ying" To: Yosry Ahmed Cc: Suren Baghdasaryan , Matthew Wilcox , akpm@linux-foundation.org, hannes@cmpxchg.org, mhocko@suse.com, josef@toxicpanda.com, jack@suse.cz, ldufour@linux.ibm.com, laurent.dufour@fr.ibm.com, michel@lespinasse.org, liam.howlett@oracle.com, jglisse@google.com, vbabka@suse.cz, minchan@google.com, dave@stgolabs.net, punit.agrawal@bytedance.com, lstoakes@gmail.com, hdanton@sina.com, apopple@nvidia.com, linux-mm@kvack.org, linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org, kernel-team@android.com, Ming Lei Subject: Re: [PATCH 1/3] mm: handle swap page faults under VMA lock if page is uncontended In-Reply-To: (Yosry Ahmed's message of "Wed, 3 May 2023 13:57:17 -0700") References: <20230501175025.36233-1-surenb@google.com> User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/27.1 (gnu/linux) Date: Fri, 05 May 2023 13:02:27 +0800 Message-ID: <87wn1nbcbg.fsf@yhuang6-desk2.ccr.corp.intel.com> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable 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: Yosry Ahmed writes: > On Wed, May 3, 2023 at 12:57=E2=80=AFPM Suren Baghdasaryan wrote: >> >> On Wed, May 3, 2023 at 1:34=E2=80=AFAM Yosry Ahmed wrote: >> > >> > On Tue, May 2, 2023 at 4:05=E2=80=AFPM Suren Baghdasaryan wrote: >> > > >> > > On Tue, May 2, 2023 at 3:31=E2=80=AFPM Matthew Wilcox wrote: >> > > > >> > > > On Tue, May 02, 2023 at 09:36:03AM -0700, Suren Baghdasaryan wrote: >> > > > > On Tue, May 2, 2023 at 8:03=E2=80=AFAM Matthew Wilcox wrote: >> > > > > > >> > > > > > On Mon, May 01, 2023 at 10:04:56PM -0700, Suren Baghdasaryan w= rote: >> > > > > > > On Mon, May 1, 2023 at 8:22=E2=80=AFPM Matthew Wilcox wrote: >> > > > > > > > >> > > > > > > > On Mon, May 01, 2023 at 07:30:13PM -0700, Suren Baghdasary= an wrote: >> > > > > > > > > On Mon, May 1, 2023 at 7:02=E2=80=AFPM Matthew Wilcox wrote: >> > > > > > > > > > >> > > > > > > > > > On Mon, May 01, 2023 at 10:50:23AM -0700, Suren Baghda= saryan wrote: >> > > > > > > > > > > +++ b/mm/memory.c >> > > > > > > > > > > @@ -3711,11 +3711,6 @@ vm_fault_t do_swap_page(struc= t vm_fault *vmf) >> > > > > > > > > > > if (!pte_unmap_same(vmf)) >> > > > > > > > > > > goto out; >> > > > > > > > > > > >> > > > > > > > > > > - if (vmf->flags & FAULT_FLAG_VMA_LOCK) { >> > > > > > > > > > > - ret =3D VM_FAULT_RETRY; >> > > > > > > > > > > - goto out; >> > > > > > > > > > > - } >> > > > > > > > > > > - >> > > > > > > > > > > entry =3D pte_to_swp_entry(vmf->orig_pte); >> > > > > > > > > > > if (unlikely(non_swap_entry(entry))) { >> > > > > > > > > > > if (is_migration_entry(entry)) { >> > > > > > > > > > >> > > > > > > > > > You're missing the necessary fallback in the (!folio) = case. >> > > > > > > > > > swap_readpage() is synchronous and will sleep. >> > > > > > > > > >> > > > > > > > > True, but is it unsafe to do that under VMA lock and has= to be done >> > > > > > > > > under mmap_lock? >> > > > > > > > >> > > > > > > > ... you were the one arguing that we didn't want to wait f= or I/O with >> > > > > > > > the VMA lock held? >> > > > > > > >> > > > > > > Well, that discussion was about waiting in folio_lock_or_ret= ry() with >> > > > > > > the lock being held. I argued against it because currently w= e drop >> > > > > > > mmap_lock lock before waiting, so if we don't drop VMA lock = we would >> > > > > > > be changing the current behavior which might introduce new >> > > > > > > regressions. In the case of swap_readpage and swapin_readahe= ad we >> > > > > > > already wait with mmap_lock held, so waiting with VMA lock h= eld does >> > > > > > > not introduce new problems (unless there is a need to hold m= map_lock). >> > > > > > > >> > > > > > > That said, you are absolutely correct that this situation ca= n be >> > > > > > > improved by dropping the lock in these cases too. I just did= n't want >> > > > > > > to attack everything at once. I believe after we agree on th= e approach >> > > > > > > implemented in https://lore.kernel.org/all/20230501175025.36= 233-3-surenb@google.com >> > > > > > > for dropping the VMA lock before waiting, these cases can be= added >> > > > > > > easier. Does that make sense? >> > > > > > >> > > > > > OK, I looked at this path some more, and I think we're fine. = This >> > > > > > patch is only called for SWP_SYNCHRONOUS_IO which is only set = for >> > > > > > QUEUE_FLAG_SYNCHRONOUS devices, which are brd, zram and nvdimms >> > > > > > (both btt and pmem). So the answer is that we don't sleep in = this >> > > > > > path, and there's no need to drop the lock. >> > > > > >> > > > > Yes but swapin_readahead does sleep, so I'll have to handle that= case >> > > > > too after this. >> > > > >> > > > Sleeping is OK, we do that in pXd_alloc()! Do we block on I/O any= where >> > > > in swapin_readahead()? It all looks like async I/O to me. >> > > >> > > Hmm. I thought that we have synchronous I/O in the following paths: >> > > swapin_readahead()->swap_cluster_readahead()->swap_readpage() >> > > swapin_readahead()->swap_vma_readahead()->swap_readpage() >> > > but just noticed that in both cases swap_readpage() is called with t= he >> > > synchronous parameter being false. So you are probably right here... >> > >> > In both swap_cluster_readahead() and swap_vma_readahead() it looks >> > like if the readahead window is 1 (aka we are not reading ahead), then >> > we jump to directly calling read_swap_cache_async() passing do_poll =3D >> > true, which means we may end up calling swap_readpage() passing >> > synchronous =3D true. >> > >> > I am not familiar with readahead heuristics, so I am not sure how >> > common this is, but it's something to think about. >> >> Uh, you are correct. If this branch is common, we could use the same >> "drop the lock and retry" pattern inside read_swap_cache_async(). That >> would be quite easy to implement. >> Thanks for checking on it! > > > I am honestly not sure how common this is. > > +Ying who might have a better idea. Checked the code and related history. It seems that we can just pass "synchronous =3D false" to swap_readpage() in read_swap_cache_async(). "synchronous =3D true" was introduced in commit 23955622ff8d ("swap: add block io poll in swapin path") to reduce swap read latency for block devices that can be polled. But in commit 9650b453a3d4 ("block: ignore RWF_HIPRI hint for sync dio"), the polling is deleted. So, we don't need to pass "synchronous =3D true" to swap_readpage() during swapin_readahead(), because we will wait the IO to complete in folio_lock_or_retry(). Best Regards, Huang, Ying >> >> >> > >> > > Does that mean swapin_readahead() might return a page which does not >> > > have its content swapped-in yet? >> > >