From patchwork Wed Mar 17 11:12:49 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Oscar Salvador X-Patchwork-Id: 12145557 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=-16.8 required=3.0 tests=BAYES_00, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER,INCLUDES_PATCH, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,USER_AGENT_GIT 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 E84DFC433E6 for ; Wed, 17 Mar 2021 11:13:11 +0000 (UTC) Received: from kanga.kvack.org (kanga.kvack.org [205.233.56.17]) by mail.kernel.org (Postfix) with ESMTP id 5F59464F69 for ; Wed, 17 Mar 2021 11:13:11 +0000 (UTC) DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org 5F59464F69 Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=suse.de Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=owner-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix) id 862198D0001; Wed, 17 Mar 2021 07:13:07 -0400 (EDT) Received: by kanga.kvack.org (Postfix, from userid 40) id 772E86B0074; Wed, 17 Mar 2021 07:13:07 -0400 (EDT) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id 550C58D0001; Wed, 17 Mar 2021 07:13:07 -0400 (EDT) X-Delivered-To: linux-mm@kvack.org Received: from forelay.hostedemail.com (smtprelay0153.hostedemail.com [216.40.44.153]) by kanga.kvack.org (Postfix) with ESMTP id 3058F6B0073 for ; Wed, 17 Mar 2021 07:13:07 -0400 (EDT) Received: from smtpin12.hostedemail.com (10.5.19.251.rfc1918.com [10.5.19.251]) by forelay02.hostedemail.com (Postfix) with ESMTP id DD88452D0 for ; Wed, 17 Mar 2021 11:13:06 +0000 (UTC) X-FDA: 77929104372.12.4165E0C Received: from mx2.suse.de (mx2.suse.de [195.135.220.15]) by imf21.hostedemail.com (Postfix) with ESMTP id 43529E0001AF for ; Wed, 17 Mar 2021 11:13:06 +0000 (UTC) X-Virus-Scanned: by amavisd-new at test-mx.suse.de Received: from relay2.suse.de (unknown [195.135.221.27]) by mx2.suse.de (Postfix) with ESMTP id 3D787AD74; Wed, 17 Mar 2021 11:13:05 +0000 (UTC) From: Oscar Salvador To: Andrew Morton Cc: Vlastimil Babka , David Hildenbrand , Michal Hocko , Muchun Song , Mike Kravetz , linux-mm@kvack.org, linux-kernel@vger.kernel.org, Oscar Salvador Subject: [PATCH v5 3/5] mm: Make alloc_contig_range handle free hugetlb pages Date: Wed, 17 Mar 2021 12:12:49 +0100 Message-Id: <20210317111251.17808-4-osalvador@suse.de> X-Mailer: git-send-email 2.13.7 In-Reply-To: <20210317111251.17808-1-osalvador@suse.de> References: <20210317111251.17808-1-osalvador@suse.de> X-Stat-Signature: hdc6a9sezi88brt3cpbbrffibf5cm44q X-Rspamd-Server: rspam02 X-Rspamd-Queue-Id: 43529E0001AF Received-SPF: none (suse.de>: No applicable sender policy available) receiver=imf21; identity=mailfrom; envelope-from=""; helo=mx2.suse.de; client-ip=195.135.220.15 X-HE-DKIM-Result: none/none X-HE-Tag: 1615979586-133274 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: alloc_contig_range will fail if it ever sees a HugeTLB page within the range we are trying to allocate, even when that page is free and can be easily reallocated. This has proved to be problematic for some users of alloc_contic_range, e.g: CMA and virtio-mem, where those would fail the call even when those pages lay in ZONE_MOVABLE and are free. We can do better by trying to replace such page. Free hugepages are tricky to handle so as to no userspace application notices disruption, we need to replace the current free hugepage with a new one. In order to do that, a new function called alloc_and_dissolve_huge_page is introduced. This function will first try to get a new fresh hugepage, and if it succeeds, it will replace the old one in the free hugepage pool. All operations are being handled under hugetlb_lock, so no races are possible. The only exception is when page's refcount is 0, but it still has not been flagged as PageHugeFreed. E.g, below scenario: CPU0 CPU1 __free_huge_page() isolate_or_dissolve_huge_page PageHuge() == T alloc_and_dissolve_huge_page alloc_fresh_huge_page() spin_lock(hugetlb_lock) // PageHuge() && !PageHugeFreed && // !PageCount() spin_unlock(hugetlb_lock) spin_lock(hugetlb_lock) 1) update_and_free_page PageHuge() == F __free_pages() 2) enqueue_huge_page SetPageHugeFreed() spin_unlock(&hugetlb_lock) spin_lock(hugetlb_lock) 1) PageHuge() == F (freed by case#1 from CPU0) 2) PageHuge() == T PageHugeFreed() == T - proceed with replacing the page In the case above we retry as the window race is quite small and we have high chances to succeed next time. With regard to the allocation, we restrict it to the node the page belongs to with __GFP_THISNODE, meaning we do not fallback on other node's zones. Note that gigantic hugetlb pages are fenced off since there is a cyclic dependency between them and alloc_contig_range. Signed-off-by: Oscar Salvador Reviewed-by: Mike Kravetz Acked-by: Michal Hocko --- include/linux/hugetlb.h | 6 +++ mm/compaction.c | 33 ++++++++++++++- mm/hugetlb.c | 109 +++++++++++++++++++++++++++++++++++++++++++++++- 3 files changed, 145 insertions(+), 3 deletions(-) diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h index cccd1aab69dd..bcff86ca616f 100644 --- a/include/linux/hugetlb.h +++ b/include/linux/hugetlb.h @@ -583,6 +583,7 @@ struct huge_bootmem_page { struct hstate *hstate; }; +int isolate_or_dissolve_huge_page(struct page *page); struct page *alloc_huge_page(struct vm_area_struct *vma, unsigned long addr, int avoid_reserve); struct page *alloc_huge_page_nodemask(struct hstate *h, int preferred_nid, @@ -865,6 +866,11 @@ static inline void huge_ptep_modify_prot_commit(struct vm_area_struct *vma, #else /* CONFIG_HUGETLB_PAGE */ struct hstate {}; +static inline int isolate_or_dissolve_huge_page(struct page *page) +{ + return -ENOMEM; +} + static inline struct page *alloc_huge_page(struct vm_area_struct *vma, unsigned long addr, int avoid_reserve) diff --git a/mm/compaction.c b/mm/compaction.c index 5769753a8f60..9f253fc3b4f9 100644 --- a/mm/compaction.c +++ b/mm/compaction.c @@ -810,6 +810,8 @@ isolate_migratepages_block(struct compact_control *cc, unsigned long low_pfn, bool skip_on_failure = false; unsigned long next_skip_pfn = 0; bool skip_updated = false; + bool fatal_error = false; + int ret = 0; cc->migrate_pfn = low_pfn; @@ -907,6 +909,32 @@ isolate_migratepages_block(struct compact_control *cc, unsigned long low_pfn, valid_page = page; } + if (PageHuge(page) && cc->alloc_contig) { + ret = isolate_or_dissolve_huge_page(page); + + /* + * Fail isolation in case isolate_or_dissolve_huge_page + * reports an error. In case of -ENOMEM, abort right away. + */ + if (ret < 0) { + /* + * Do not report -EBUSY down the chain. + */ + if (ret == -ENOMEM) + fatal_error = true; + else + ret = 0; + goto isolate_fail; + } + + /* + * Ok, the hugepage was dissolved. Now these pages are + * Buddy and cannot be re-allocated because they are + * isolated. Fall-through as the check below handles + * Buddy pages. + */ + } + /* * Skip if free. We read page order here without zone lock * which is generally unsafe, but the race window is small and @@ -1092,6 +1120,9 @@ isolate_migratepages_block(struct compact_control *cc, unsigned long low_pfn, */ next_skip_pfn += 1UL << cc->order; } + + if (fatal_error) + break; } /* @@ -1135,7 +1166,7 @@ isolate_migratepages_block(struct compact_control *cc, unsigned long low_pfn, cc->migrate_pfn = low_pfn; - return 0; + return ret; } /** diff --git a/mm/hugetlb.c b/mm/hugetlb.c index 5b1ab1f427c5..3194c1bd9e32 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -1035,13 +1035,18 @@ static bool vma_has_reserves(struct vm_area_struct *vma, long chg) return false; } +static void __enqueue_huge_page(struct list_head *list, struct page *page) +{ + list_move(&page->lru, list); + SetHPageFreed(page); +} + static void enqueue_huge_page(struct hstate *h, struct page *page) { int nid = page_to_nid(page); - list_move(&page->lru, &h->hugepage_freelists[nid]); + __enqueue_huge_page(&h->hugepage_freelists[nid], page); h->free_huge_pages++; h->free_huge_pages_node[nid]++; - SetHPageFreed(page); } static struct page *dequeue_huge_page_node_exact(struct hstate *h, int nid) @@ -2245,6 +2250,106 @@ static void restore_reserve_on_error(struct hstate *h, } } +/* + * alloc_and_dissolve_huge_page - Allocate a new page and dissolve the old one + * @h: struct hstate old page belongs to + * @old_page: Old page to dissolve + * Returns 0 on success, otherwise negated error. + */ + +static int alloc_and_dissolve_huge_page(struct hstate *h, struct page *old_page) +{ + gfp_t gfp_mask = htlb_alloc_mask(h) | __GFP_THISNODE; + int nid = page_to_nid(old_page); + struct page *new_page; + int ret = 0; + + /* + * Before dissolving the page, we need to allocate a new one, + * so the pool remains stable. + */ + new_page = alloc_fresh_huge_page(h, gfp_mask, nid, NULL, NULL); + if (!new_page) + return -ENOMEM; + + /* + * Pages got from Buddy are self-refcounted, but free hugepages + * need to have a refcount of 0. + */ + page_ref_dec(new_page); +retry: + spin_lock(&hugetlb_lock); + if (!PageHuge(old_page)) { + /* + * Freed from under us. Drop new_page too. + */ + update_and_free_page(h, new_page); + goto unlock; + } else if (page_count(old_page)) { + /* + * Someone has grabbed the page, fail for now. + */ + ret = -EBUSY; + update_and_free_page(h, new_page); + goto unlock; + } else if (!HPageFreed(old_page)) { + /* + * Page's refcount is 0 but it has not been enqueued in the + * freelist yet. Race window is small, so we can succed here if + * we retry. + */ + spin_unlock(&hugetlb_lock); + cond_resched(); + goto retry; + } else { + /* + * Ok, old_page is still a genuine free hugepage. Replace it + * with the new one. + */ + list_del(&old_page->lru); + update_and_free_page(h, old_page); + /* + * h->free_huge_pages{_node} counters do not need to be updated. + */ + __enqueue_huge_page(&h->hugepage_freelists[nid], new_page); + } +unlock: + spin_unlock(&hugetlb_lock); + + return ret; +} + +int isolate_or_dissolve_huge_page(struct page *page) +{ + struct hstate *h; + struct page *head; + + /* + * The page might have been dissolved from under our feet, so make sure + * to carefully check the state under the lock. + * Return success when racing as if we dissolved the page ourselves. + */ + spin_lock(&hugetlb_lock); + if (PageHuge(page)) { + head = compound_head(page); + h = page_hstate(head); + } else { + spin_unlock(&hugetlb_lock); + return 0; + } + spin_unlock(&hugetlb_lock); + + /* + * Fence off gigantic pages as there is a cyclic dependency between + * alloc_contig_range and them. Return -ENOME as this has the effect + * of bailing out right away without further retrying. + */ + if (hstate_is_gigantic(h)) + return -ENOMEM; + + return alloc_and_dissolve_huge_page(h, head); +} + struct page *alloc_huge_page(struct vm_area_struct *vma, unsigned long addr, int avoid_reserve) {