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=-12.7 required=3.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,FREEMAIL_FORGED_FROMDOMAIN,FREEMAIL_FROM, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED autolearn=ham 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 1B965C433DB for ; Fri, 29 Jan 2021 20:46:51 +0000 (UTC) Received: from kanga.kvack.org (kanga.kvack.org [205.233.56.17]) by mail.kernel.org (Postfix) with ESMTP id 907A564DE7 for ; Fri, 29 Jan 2021 20:46:50 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 907A564DE7 Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=gmail.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=owner-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix) id BC63D8D0005; Fri, 29 Jan 2021 15:46:49 -0500 (EST) Received: by kanga.kvack.org (Postfix, from userid 40) id B77348D0001; Fri, 29 Jan 2021 15:46:49 -0500 (EST) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id AB3DC8D0005; Fri, 29 Jan 2021 15:46:49 -0500 (EST) X-Delivered-To: linux-mm@kvack.org Received: from forelay.hostedemail.com (smtprelay0068.hostedemail.com [216.40.44.68]) by kanga.kvack.org (Postfix) with ESMTP id 95D438D0001 for ; Fri, 29 Jan 2021 15:46:49 -0500 (EST) Received: from smtpin14.hostedemail.com (10.5.19.251.rfc1918.com [10.5.19.251]) by forelay01.hostedemail.com (Postfix) with ESMTP id 5FD3B180AD837 for ; Fri, 29 Jan 2021 20:46:49 +0000 (UTC) X-FDA: 77759996538.14.pie27_0e08c69275ab Received: from filter.hostedemail.com (10.5.16.251.rfc1918.com [10.5.16.251]) by smtpin14.hostedemail.com (Postfix) with ESMTP id 3D89E18229835 for ; Fri, 29 Jan 2021 20:46:49 +0000 (UTC) X-HE-Tag: pie27_0e08c69275ab X-Filterd-Recvd-Size: 12970 Received: from mail-ed1-f41.google.com (mail-ed1-f41.google.com [209.85.208.41]) by imf47.hostedemail.com (Postfix) with ESMTP for ; Fri, 29 Jan 2021 20:46:48 +0000 (UTC) Received: by mail-ed1-f41.google.com with SMTP id a14so12108530edu.7 for ; Fri, 29 Jan 2021 12:46:48 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc; bh=0u/SrW/ayXXMayejHeDskpnMWtid3p2YEcmD4tncPvE=; b=EAF9RO7KqcSoomD4vFgYqUcAZnypKz7YNj6H28ryadwzS2HNdmpNsiYzy5aGwmC6ks 0qMFoq+29QcJoCyic7z4sL9pU8WEnKm5kPRMSIU6xBW0CsaNTZ8uwNGlzLrjG4lGS2mb rkDKOUJPWbw8mYTO4WUSUxZRgclU1DputL0x4umkO3OcfUsGT07NZb6Y79JRrG9zLzii zp9Hro9c8fnzsMlbbEGGMlctQbWaHye3qCK8y3cMKr3JXbl/Ib5R+5xNbNn7lCTWAN8K MhqvFu+4nxp/BGRmRbfCVh8TdHnXSZUgcinIAcJpVUvG/AszEy/KoU1MxKcopAjeE9F+ CO+w== 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=0u/SrW/ayXXMayejHeDskpnMWtid3p2YEcmD4tncPvE=; b=Hjxt5Zh3igzyeAWBftMOBTOCuOM4va9wqbUo6co6Jb0BnV0B/A+XZCYKt5GL9sSLDf IACg+5RcB7NoOcRGjPw2OsWg0k0oUv+ffu/AQ202p3i5HzgevWD+0OlMZ0fPjxa/oSlD 7aCcGFJza471XAbGomGBzFwHeZ9hsDHRbH7RTOy15NSlGoIGdc58psNnkIXXU0Z4zN8v FEEWY45ucoAJXQGFMIMM440O9LIRKseVAk16jUxOPi662BRJpJEMNgoHWEqCKqdu5+Jv cOzSOxnV5UgcEL40dn5UVbz88SYW2q/UNNDZOH8bbrTU/CVbD+Hywhx327LwbMZqpyNZ o8Sg== X-Gm-Message-State: AOAM530fMKydZYHTtExuceEybxchzovHrZgOw0AfJY/vFlKRd46vQ5nF MW4FbE6J9SRtMN4wMGT3/MCyeCTIvQ7McKGvsGg= X-Google-Smtp-Source: ABdhPJy+9IgT4LWH6dgrcWboPQSMX0J+kwBdGsfBIzjz6HAtvFG2yH/K63MaJtOA2IZpGdd7mf0FDhluQkDKrYyj1uM= X-Received: by 2002:a05:6402:1bde:: with SMTP id ch30mr7292820edb.151.1611953207600; Fri, 29 Jan 2021 12:46:47 -0800 (PST) MIME-Version: 1.0 References: <20210126003411.2AC51464@viggo.jf.intel.com> <20210126003421.45897BF4@viggo.jf.intel.com> In-Reply-To: <20210126003421.45897BF4@viggo.jf.intel.com> From: Yang Shi Date: Fri, 29 Jan 2021 12:46:35 -0800 Message-ID: Subject: Re: [RFC][PATCH 05/13] mm/numa: automatically generate node migration order To: Dave Hansen Cc: Linux Kernel Mailing List , Linux MM , Yang Shi , David Rientjes , Huang Ying , Dan Williams , David Hildenbrand , Oscar Salvador Content-Type: text/plain; charset="UTF-8" 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 Mon, Jan 25, 2021 at 4:41 PM Dave Hansen wrote: > > > From: Dave Hansen > > When memory fills up on a node, memory contents can be > automatically migrated to another node. The biggest problems are > knowing when to migrate and to where the migration should be > targeted. > > The most straightforward way to generate the "to where" list > would be to follow the page allocator fallback lists. Those > lists already tell us if memory is full where to look next. It > would also be logical to move memory in that order. > > But, the allocator fallback lists have a fatal flaw: most nodes > appear in all the lists. This would potentially lead to > migration cycles (A->B, B->A, A->B, ...). > > Instead of using the allocator fallback lists directly, keep a > separate node migration ordering. But, reuse the same data used > to generate page allocator fallback in the first place: > find_next_best_node(). > > This means that the firmware data used to populate node distances > essentially dictates the ordering for now. It should also be > architecture-neutral since all NUMA architectures have a working > find_next_best_node(). > > The protocol for node_demotion[] access and writing is not > standard. It has no specific locking and is intended to be read > locklessly. Readers must take care to avoid observing changes > that appear incoherent. This was done so that node_demotion[] > locking has no chance of becoming a bottleneck on large systems > with lots of CPUs in direct reclaim. > > This code is unused for now. It will be called later in the > series. > > Signed-off-by: Dave Hansen > Cc: Yang Shi > Cc: David Rientjes > Cc: Huang Ying > Cc: Dan Williams > Cc: David Hildenbrand > Cc: osalvador > --- > > b/mm/internal.h | 5 + > b/mm/migrate.c | 137 +++++++++++++++++++++++++++++++++++++++++++++++++++++- > b/mm/page_alloc.c | 2 > 3 files changed, 142 insertions(+), 2 deletions(-) > > diff -puN mm/internal.h~auto-setup-default-migration-path-from-firmware mm/internal.h > --- a/mm/internal.h~auto-setup-default-migration-path-from-firmware 2021-01-25 16:23:10.607866706 -0800 > +++ b/mm/internal.h 2021-01-25 16:23:10.616866706 -0800 > @@ -515,12 +515,17 @@ static inline void mminit_validate_memmo > > #ifdef CONFIG_NUMA > extern int node_reclaim(struct pglist_data *, gfp_t, unsigned int); > +extern int find_next_best_node(int node, nodemask_t *used_node_mask); > #else > static inline int node_reclaim(struct pglist_data *pgdat, gfp_t mask, > unsigned int order) > { > return NODE_RECLAIM_NOSCAN; > } > +static inline int find_next_best_node(int node, nodemask_t *used_node_mask) > +{ > + return NUMA_NO_NODE; > +} > #endif > > extern int hwpoison_filter(struct page *p); > diff -puN mm/migrate.c~auto-setup-default-migration-path-from-firmware mm/migrate.c > --- a/mm/migrate.c~auto-setup-default-migration-path-from-firmware 2021-01-25 16:23:10.609866706 -0800 > +++ b/mm/migrate.c 2021-01-25 16:23:10.617866706 -0800 > @@ -1161,6 +1161,10 @@ out: > return rc; > } > > +/* > + * Writes to this array occur without locking. READ_ONCE() > + * is recommended for readers to ensure consistent reads. > + */ > static int node_demotion[MAX_NUMNODES] = {[0 ... MAX_NUMNODES - 1] = NUMA_NO_NODE}; > > /** > @@ -1174,7 +1178,13 @@ static int node_demotion[MAX_NUMNODES] = > */ > int next_demotion_node(int node) > { > - return node_demotion[node]; > + /* > + * node_demotion[] is updated without excluding > + * this function from running. READ_ONCE() avoids > + * reading multiple, inconsistent 'node' values > + * during an update. > + */ Don't we need a smp_rmb() here? The single write barrier might be not enough in migration target set. Typically a write barrier should be used in pairs with a read barrier. > + return READ_ONCE(node_demotion[node]); Why not consolidate the patch #4 in this patch? The patch #4 just add the definition of node_demotion and the function, then the function is changed in this patch, and the function is not used by anyone between the adding and changing. > } > > /* > @@ -3124,3 +3134,128 @@ void migrate_vma_finalize(struct migrate > } > EXPORT_SYMBOL(migrate_vma_finalize); > #endif /* CONFIG_DEVICE_PRIVATE */ > + > +/* Disable reclaim-based migration. */ > +static void disable_all_migrate_targets(void) > +{ > + int node; > + > + for_each_online_node(node) > + node_demotion[node] = NUMA_NO_NODE; > +} > + > +/* > + * Find an automatic demotion target for 'node'. > + * Failing here is OK. It might just indicate > + * being at the end of a chain. > + */ > +static int establish_migrate_target(int node, nodemask_t *used) > +{ > + int migration_target; > + > + /* > + * Can not set a migration target on a > + * node with it already set. > + * > + * No need for READ_ONCE() here since this > + * in the write path for node_demotion[]. > + * This should be the only thread writing. > + */ > + if (node_demotion[node] != NUMA_NO_NODE) > + return NUMA_NO_NODE; > + > + migration_target = find_next_best_node(node, used); > + if (migration_target == NUMA_NO_NODE) > + return NUMA_NO_NODE; > + > + node_demotion[node] = migration_target; > + > + return migration_target; > +} > + > +/* > + * When memory fills up on a node, memory contents can be > + * automatically migrated to another node instead of > + * discarded at reclaim. > + * > + * Establish a "migration path" which will start at nodes > + * with CPUs and will follow the priorities used to build the > + * page allocator zonelists. > + * > + * The difference here is that cycles must be avoided. If > + * node0 migrates to node1, then neither node1, nor anything > + * node1 migrates to can migrate to node0. > + * > + * This function can run simultaneously with readers of > + * node_demotion[]. However, it can not run simultaneously > + * with itself. Exclusion is provided by memory hotplug events > + * being single-threaded. Maybe an example diagram for the physical topology and how the migration target is generated in the comment seems helpful to understand the code. > + */ > +void __set_migration_target_nodes(void) > +{ > + nodemask_t next_pass = NODE_MASK_NONE; > + nodemask_t this_pass = NODE_MASK_NONE; > + nodemask_t used_targets = NODE_MASK_NONE; > + int node; > + > + /* > + * Avoid any oddities like cycles that could occur > + * from changes in the topology. This will leave > + * a momentary gap when migration is disabled. > + */ > + disable_all_migrate_targets(); > + > + /* > + * Ensure that the "disable" is visible across the system. > + * Readers will see either a combination of before+disable > + * state or disable+after. They will never see before and > + * after state together. > + * > + * The before+after state together might have cycles and > + * could cause readers to do things like loop until this > + * function finishes. This ensures they can only see a > + * single "bad" read and would, for instance, only loop > + * once. > + */ > + smp_wmb(); > + > + /* > + * Allocations go close to CPUs, first. Assume that > + * the migration path starts at the nodes with CPUs. > + */ > + next_pass = node_states[N_CPU]; > +again: > + this_pass = next_pass; > + next_pass = NODE_MASK_NONE; > + /* > + * To avoid cycles in the migration "graph", ensure > + * that migration sources are not future targets by > + * setting them in 'used_targets'. Do this only > + * once per pass so that multiple source nodes can > + * share a target node. > + * > + * 'used_targets' will become unavailable in future > + * passes. This limits some opportunities for > + * multiple source nodes to share a desintation. s/desination/destination > + */ > + nodes_or(used_targets, used_targets, this_pass); > + for_each_node_mask(node, this_pass) { > + int target_node = establish_migrate_target(node, &used_targets); > + > + if (target_node == NUMA_NO_NODE) > + continue; > + > + /* Visit targets from this pass in the next pass: */ > + node_set(target_node, next_pass); > + } > + /* Is another pass necessary? */ > + if (!nodes_empty(next_pass)) > + goto again; > +} > + > +void set_migration_target_nodes(void) It seems this function is not called outside migrate.c, so it should be static. > +{ > + get_online_mems(); > + __set_migration_target_nodes(); > + put_online_mems(); > +} > diff -puN mm/page_alloc.c~auto-setup-default-migration-path-from-firmware mm/page_alloc.c > --- a/mm/page_alloc.c~auto-setup-default-migration-path-from-firmware 2021-01-25 16:23:10.612866706 -0800 > +++ b/mm/page_alloc.c 2021-01-25 16:23:10.619866706 -0800 > @@ -5704,7 +5704,7 @@ static int node_load[MAX_NUMNODES]; > * > * Return: node id of the found node or %NUMA_NO_NODE if no node is found. > */ > -static int find_next_best_node(int node, nodemask_t *used_node_mask) > +int find_next_best_node(int node, nodemask_t *used_node_mask) > { > int n, val; > int min_val = INT_MAX; > _ >