From nobody Mon Sep 15 14:20:10 2025 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 vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 582C1C54EBC for ; Thu, 12 Jan 2023 12:40:29 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S229698AbjALMk0 (ORCPT ); Thu, 12 Jan 2023 07:40:26 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:32818 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S230422AbjALMjz (ORCPT ); Thu, 12 Jan 2023 07:39:55 -0500 Received: from mail-wm1-x32a.google.com (mail-wm1-x32a.google.com [IPv6:2a00:1450:4864:20::32a]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 04A9949163 for ; Thu, 12 Jan 2023 04:39:53 -0800 (PST) Received: by mail-wm1-x32a.google.com with SMTP id g19-20020a05600c4ed300b003d9eb1dbc0aso11961418wmq.3 for ; Thu, 12 Jan 2023 04:39:53 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20210112; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=KnX9+IwlYoEli7Cxl1hBSElWY2G0wjT9ONp5xH+mJuY=; b=BmEtIm5KxLpYcH2+rFudi2b2EqZYY95/hg9kFAqv1fHXnNg7vHYznS1zI6POhv0Kax URYQLw5zaT9kX1QS0uUF7Z3n4Ua6qaTVab7o3Eu9+7+Hdavh11vvRBZbtBYaID/5e7/9 JMIdRQikFGst75Ynmb2K7bhIkBcwxrAdXIOhoJQl6b+2N50a6vZ+9moOM24ojGx6aTgo XnJR47AlRVA3oGJLNctsYa75KpHOE7y9BWUrFQqzB2L/3ZWfuwPpaUjAAQXivm+h5/Y8 0iWAObRtk71FHnLv57ynfXAdiWdrjnrmG6SMOghTCTZcYt+e+WOIIgn7FeVVFfzygAN/ CtcA== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=KnX9+IwlYoEli7Cxl1hBSElWY2G0wjT9ONp5xH+mJuY=; b=oClqBHtIYq9MQ2FWfFL4qu7PBqo7H1DAyjbbsp+wcMDIp39hX36J0bAkL/WFe6Kkgb BgBaUAHu4EvnOUWCWtp7kRMVrU+4nOkYWw0XGhAGKotqR+ER9PUeFCkndboOit+OvSrl kTN8RjapWjvUG2FrtPyQushstFhu88ob7wyYlqWKJheuNJbvuvHBKra5xA3Yd1magXfF C6DF4Xj8rxe7IVdCjzfiPzeUnLLAa+nzcSQO32PYLnAxKMBXi8wr8JiA209J/fQuReGI VG2cHgkm6ZEe+B2xR27asebH2Hyg9IBhIKxvJXHEG+84Ivtmw7+5Zo/SXuqAVnNkxpwf UI/w== X-Gm-Message-State: AFqh2krJ5ewqFKv78614dXrQ0WtWabvbyh4STQ/ffFdrM2io60kHfChi iuNfE0Vfkd/117lfd0MPVIs= X-Google-Smtp-Source: AMrXdXstJtcljsZ7fhP+s06VKguzL0WQXuBmkK97SmkLfKKb1mqVoxRmi7HSxR4XE27QhAcApIWdew== X-Received: by 2002:a05:600c:348b:b0:3d1:f16b:30e6 with SMTP id a11-20020a05600c348b00b003d1f16b30e6mr55625635wmq.28.1673527192234; Thu, 12 Jan 2023 04:39:52 -0800 (PST) Received: from lucifer.home (host86-164-169-89.range86-164.btcentralplus.com. [86.164.169.89]) by smtp.googlemail.com with ESMTPSA id q1-20020a1ce901000000b003b3307fb98fsm20890797wmc.24.2023.01.12.04.39.50 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Thu, 12 Jan 2023 04:39:51 -0800 (PST) From: Lorenzo Stoakes To: linux-mm@kvack.org, Andrew Morton , linux-kernel@vger.kernel.org Cc: Matthew Wilcox , Hugh Dickins , Vlastimil Babka , Liam Howlett , William Kucharski , Christian Brauner , Jonathan Corbet , Mike Rapoport , Joel Fernandes , Geert Uytterhoeven , Lorenzo Stoakes Subject: [PATCH v4 2/5] mm: mlock: use folios and a folio batch internally Date: Thu, 12 Jan 2023 12:39:29 +0000 Message-Id: <9f894d54d568773f4ed3cb0eef5f8932f62c95f4.1673526881.git.lstoakes@gmail.com> X-Mailer: git-send-email 2.39.0 In-Reply-To: References: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Type: text/plain; charset="utf-8" This brings mlock in line with the folio batches declared in mm/swap.c and makes the code more consistent across the two. The existing mechanism for identifying which operation each folio in the batch is undergoing is maintained, i.e. using the lower 2 bits of the struct folio address (previously struct page address). This should continue to function correctly as folios remain at least system word-aligned. All invoctions of mlock() pass either a non-compound page or the head of a THP-compound page and no tail pages need updating so this functionality works with struct folios being used internally rather than struct pages. In this patch the external interface is kept identical to before in order to maintain separation between patches in the series, using a rather awkward conversion from struct page to struct folio in relevant functions. However, this maintenance of the existing interface is intended to be temporary - the next patch in the series will update the interfaces to accept folios directly. Signed-off-by: Lorenzo Stoakes Acked-by: Vlastimil Babka --- mm/mlock.c | 246 +++++++++++++++++++++++++++-------------------------- 1 file changed, 124 insertions(+), 122 deletions(-) diff --git a/mm/mlock.c b/mm/mlock.c index 7032f6dd0ce1..f8e8d30ab08a 100644 --- a/mm/mlock.c +++ b/mm/mlock.c @@ -28,12 +28,12 @@ =20 #include "internal.h" =20 -struct mlock_pvec { +struct mlock_fbatch { local_lock_t lock; - struct pagevec vec; + struct folio_batch fbatch; }; =20 -static DEFINE_PER_CPU(struct mlock_pvec, mlock_pvec) =3D { +static DEFINE_PER_CPU(struct mlock_fbatch, mlock_fbatch) =3D { .lock =3D INIT_LOCAL_LOCK(lock), }; =20 @@ -48,192 +48,192 @@ bool can_do_mlock(void) EXPORT_SYMBOL(can_do_mlock); =20 /* - * Mlocked pages are marked with PageMlocked() flag for efficient testing + * Mlocked folios are marked with the PG_mlocked flag for efficient testing * in vmscan and, possibly, the fault path; and to support semi-accurate * statistics. * - * An mlocked page [PageMlocked(page)] is unevictable. As such, it will - * be placed on the LRU "unevictable" list, rather than the [in]active lis= ts. - * The unevictable list is an LRU sibling list to the [in]active lists. - * PageUnevictable is set to indicate the unevictable state. + * An mlocked folio [folio_test_mlocked(folio)] is unevictable. As such, = it + * will be ostensibly placed on the LRU "unevictable" list (actually no su= ch + * list exists), rather than the [in]active lists. PG_unevictable is set to + * indicate the unevictable state. */ =20 -static struct lruvec *__mlock_page(struct page *page, struct lruvec *lruve= c) +static struct lruvec *__mlock_folio(struct folio *folio, struct lruvec *lr= uvec) { /* There is nothing more we can do while it's off LRU */ - if (!TestClearPageLRU(page)) + if (!folio_test_clear_lru(folio)) return lruvec; =20 - lruvec =3D folio_lruvec_relock_irq(page_folio(page), lruvec); + lruvec =3D folio_lruvec_relock_irq(folio, lruvec); =20 - if (unlikely(page_evictable(page))) { + if (unlikely(folio_evictable(folio))) { /* - * This is a little surprising, but quite possible: - * PageMlocked must have got cleared already by another CPU. - * Could this page be on the Unevictable LRU? I'm not sure, - * but move it now if so. + * This is a little surprising, but quite possible: PG_mlocked + * must have got cleared already by another CPU. Could this + * folio be unevictable? I'm not sure, but move it now if so. */ - if (PageUnevictable(page)) { - del_page_from_lru_list(page, lruvec); - ClearPageUnevictable(page); - add_page_to_lru_list(page, lruvec); + if (folio_test_unevictable(folio)) { + lruvec_del_folio(lruvec, folio); + folio_clear_unevictable(folio); + lruvec_add_folio(lruvec, folio); + __count_vm_events(UNEVICTABLE_PGRESCUED, - thp_nr_pages(page)); + folio_nr_pages(folio)); } goto out; } =20 - if (PageUnevictable(page)) { - if (PageMlocked(page)) - page->mlock_count++; + if (folio_test_unevictable(folio)) { + if (folio_test_mlocked(folio)) + folio->mlock_count++; goto out; } =20 - del_page_from_lru_list(page, lruvec); - ClearPageActive(page); - SetPageUnevictable(page); - page->mlock_count =3D !!PageMlocked(page); - add_page_to_lru_list(page, lruvec); - __count_vm_events(UNEVICTABLE_PGCULLED, thp_nr_pages(page)); + lruvec_del_folio(lruvec, folio); + folio_clear_active(folio); + folio_set_unevictable(folio); + folio->mlock_count =3D !!folio_test_mlocked(folio); + lruvec_add_folio(lruvec, folio); + __count_vm_events(UNEVICTABLE_PGCULLED, folio_nr_pages(folio)); out: - SetPageLRU(page); + folio_set_lru(folio); return lruvec; } =20 -static struct lruvec *__mlock_new_page(struct page *page, struct lruvec *l= ruvec) +static struct lruvec *__mlock_new_folio(struct folio *folio, struct lruvec= *lruvec) { - VM_BUG_ON_PAGE(PageLRU(page), page); + VM_BUG_ON_FOLIO(folio_test_lru(folio), folio); =20 - lruvec =3D folio_lruvec_relock_irq(page_folio(page), lruvec); + lruvec =3D folio_lruvec_relock_irq(folio, lruvec); =20 /* As above, this is a little surprising, but possible */ - if (unlikely(page_evictable(page))) + if (unlikely(folio_evictable(folio))) goto out; =20 - SetPageUnevictable(page); - page->mlock_count =3D !!PageMlocked(page); - __count_vm_events(UNEVICTABLE_PGCULLED, thp_nr_pages(page)); + folio_set_unevictable(folio); + folio->mlock_count =3D !!folio_test_mlocked(folio); + __count_vm_events(UNEVICTABLE_PGCULLED, folio_nr_pages(folio)); out: - add_page_to_lru_list(page, lruvec); - SetPageLRU(page); + lruvec_add_folio(lruvec, folio); + folio_set_lru(folio); return lruvec; } =20 -static struct lruvec *__munlock_page(struct page *page, struct lruvec *lru= vec) +static struct lruvec *__munlock_folio(struct folio *folio, struct lruvec *= lruvec) { - int nr_pages =3D thp_nr_pages(page); + int nr_pages =3D folio_nr_pages(folio); bool isolated =3D false; =20 - if (!TestClearPageLRU(page)) + if (!folio_test_clear_lru(folio)) goto munlock; =20 isolated =3D true; - lruvec =3D folio_lruvec_relock_irq(page_folio(page), lruvec); + lruvec =3D folio_lruvec_relock_irq(folio, lruvec); =20 - if (PageUnevictable(page)) { + if (folio_test_unevictable(folio)) { /* Then mlock_count is maintained, but might undercount */ - if (page->mlock_count) - page->mlock_count--; - if (page->mlock_count) + if (folio->mlock_count) + folio->mlock_count--; + if (folio->mlock_count) goto out; } /* else assume that was the last mlock: reclaim will fix it if not */ =20 munlock: - if (TestClearPageMlocked(page)) { - __mod_zone_page_state(page_zone(page), NR_MLOCK, -nr_pages); - if (isolated || !PageUnevictable(page)) + if (folio_test_clear_mlocked(folio)) { + __zone_stat_mod_folio(folio, NR_MLOCK, -nr_pages); + if (isolated || !folio_test_unevictable(folio)) __count_vm_events(UNEVICTABLE_PGMUNLOCKED, nr_pages); else __count_vm_events(UNEVICTABLE_PGSTRANDED, nr_pages); } =20 - /* page_evictable() has to be checked *after* clearing Mlocked */ - if (isolated && PageUnevictable(page) && page_evictable(page)) { - del_page_from_lru_list(page, lruvec); - ClearPageUnevictable(page); - add_page_to_lru_list(page, lruvec); + /* folio_evictable() has to be checked *after* clearing Mlocked */ + if (isolated && folio_test_unevictable(folio) && folio_evictable(folio)) { + lruvec_del_folio(lruvec, folio); + folio_clear_unevictable(folio); + lruvec_add_folio(lruvec, folio); __count_vm_events(UNEVICTABLE_PGRESCUED, nr_pages); } out: if (isolated) - SetPageLRU(page); + folio_set_lru(folio); return lruvec; } =20 /* - * Flags held in the low bits of a struct page pointer on the mlock_pvec. + * Flags held in the low bits of a struct folio pointer on the mlock_fbatc= h. */ -#define LRU_PAGE 0x1 -#define NEW_PAGE 0x2 -static inline struct page *mlock_lru(struct page *page) +#define LRU_FOLIO 0x1 +#define NEW_FOLIO 0x2 +static inline struct folio *mlock_lru(struct folio *folio) { - return (struct page *)((unsigned long)page + LRU_PAGE); + return (struct folio *)((unsigned long)folio + LRU_FOLIO); } =20 -static inline struct page *mlock_new(struct page *page) +static inline struct folio *mlock_new(struct folio *folio) { - return (struct page *)((unsigned long)page + NEW_PAGE); + return (struct folio *)((unsigned long)folio + NEW_FOLIO); } =20 /* - * mlock_pagevec() is derived from pagevec_lru_move_fn(): - * perhaps that can make use of such page pointer flags in future, - * but for now just keep it for mlock. We could use three separate - * pagevecs instead, but one feels better (munlocking a full pagevec - * does not need to drain mlocking pagevecs first). + * mlock_folio_batch() is derived from folio_batch_move_lru(): perhaps tha= t can + * make use of such folio pointer flags in future, but for now just keep i= t for + * mlock. We could use three separate folio batches instead, but one feels + * better (munlocking a full folio batch does not need to drain mlocking f= olio + * batches first). */ -static void mlock_pagevec(struct pagevec *pvec) +static void mlock_folio_batch(struct folio_batch *fbatch) { struct lruvec *lruvec =3D NULL; unsigned long mlock; - struct page *page; + struct folio *folio; int i; =20 - for (i =3D 0; i < pagevec_count(pvec); i++) { - page =3D pvec->pages[i]; - mlock =3D (unsigned long)page & (LRU_PAGE | NEW_PAGE); - page =3D (struct page *)((unsigned long)page - mlock); - pvec->pages[i] =3D page; + for (i =3D 0; i < folio_batch_count(fbatch); i++) { + folio =3D fbatch->folios[i]; + mlock =3D (unsigned long)folio & (LRU_FOLIO | NEW_FOLIO); + folio =3D (struct folio *)((unsigned long)folio - mlock); + fbatch->folios[i] =3D folio; =20 - if (mlock & LRU_PAGE) - lruvec =3D __mlock_page(page, lruvec); - else if (mlock & NEW_PAGE) - lruvec =3D __mlock_new_page(page, lruvec); + if (mlock & LRU_FOLIO) + lruvec =3D __mlock_folio(folio, lruvec); + else if (mlock & NEW_FOLIO) + lruvec =3D __mlock_new_folio(folio, lruvec); else - lruvec =3D __munlock_page(page, lruvec); + lruvec =3D __munlock_folio(folio, lruvec); } =20 if (lruvec) unlock_page_lruvec_irq(lruvec); - release_pages(pvec->pages, pvec->nr); - pagevec_reinit(pvec); + release_pages(fbatch->folios, fbatch->nr); + folio_batch_reinit(fbatch); } =20 void mlock_page_drain_local(void) { - struct pagevec *pvec; + struct folio_batch *fbatch; =20 - local_lock(&mlock_pvec.lock); - pvec =3D this_cpu_ptr(&mlock_pvec.vec); - if (pagevec_count(pvec)) - mlock_pagevec(pvec); - local_unlock(&mlock_pvec.lock); + local_lock(&mlock_fbatch.lock); + fbatch =3D this_cpu_ptr(&mlock_fbatch.fbatch); + if (folio_batch_count(fbatch)) + mlock_folio_batch(fbatch); + local_unlock(&mlock_fbatch.lock); } =20 void mlock_page_drain_remote(int cpu) { - struct pagevec *pvec; + struct folio_batch *fbatch; =20 WARN_ON_ONCE(cpu_online(cpu)); - pvec =3D &per_cpu(mlock_pvec.vec, cpu); - if (pagevec_count(pvec)) - mlock_pagevec(pvec); + fbatch =3D &per_cpu(mlock_fbatch.fbatch, cpu); + if (folio_batch_count(fbatch)) + mlock_folio_batch(fbatch); } =20 bool need_mlock_page_drain(int cpu) { - return pagevec_count(&per_cpu(mlock_pvec.vec, cpu)); + return folio_batch_count(&per_cpu(mlock_fbatch.fbatch, cpu)); } =20 /** @@ -242,10 +242,10 @@ bool need_mlock_page_drain(int cpu) */ void mlock_folio(struct folio *folio) { - struct pagevec *pvec; + struct folio_batch *fbatch; =20 - local_lock(&mlock_pvec.lock); - pvec =3D this_cpu_ptr(&mlock_pvec.vec); + local_lock(&mlock_fbatch.lock); + fbatch =3D this_cpu_ptr(&mlock_fbatch.fbatch); =20 if (!folio_test_set_mlocked(folio)) { int nr_pages =3D folio_nr_pages(folio); @@ -255,10 +255,10 @@ void mlock_folio(struct folio *folio) } =20 folio_get(folio); - if (!pagevec_add(pvec, mlock_lru(&folio->page)) || + if (!folio_batch_add(fbatch, mlock_lru(folio)) || folio_test_large(folio) || lru_cache_disabled()) - mlock_pagevec(pvec); - local_unlock(&mlock_pvec.lock); + mlock_folio_batch(fbatch); + local_unlock(&mlock_fbatch.lock); } =20 /** @@ -267,20 +267,22 @@ void mlock_folio(struct folio *folio) */ void mlock_new_page(struct page *page) { - struct pagevec *pvec; - int nr_pages =3D thp_nr_pages(page); + struct folio_batch *fbatch; + struct folio *folio =3D page_folio(page); + int nr_pages =3D folio_nr_pages(folio); =20 - local_lock(&mlock_pvec.lock); - pvec =3D this_cpu_ptr(&mlock_pvec.vec); - SetPageMlocked(page); - mod_zone_page_state(page_zone(page), NR_MLOCK, nr_pages); + local_lock(&mlock_fbatch.lock); + fbatch =3D this_cpu_ptr(&mlock_fbatch.fbatch); + folio_set_mlocked(folio); + + zone_stat_mod_folio(folio, NR_MLOCK, nr_pages); __count_vm_events(UNEVICTABLE_PGMLOCKED, nr_pages); =20 - get_page(page); - if (!pagevec_add(pvec, mlock_new(page)) || - PageHead(page) || lru_cache_disabled()) - mlock_pagevec(pvec); - local_unlock(&mlock_pvec.lock); + folio_get(folio); + if (!folio_batch_add(fbatch, mlock_new(folio)) || + folio_test_large(folio) || lru_cache_disabled()) + mlock_folio_batch(fbatch); + local_unlock(&mlock_fbatch.lock); } =20 /** @@ -289,20 +291,20 @@ void mlock_new_page(struct page *page) */ void munlock_page(struct page *page) { - struct pagevec *pvec; + struct folio_batch *fbatch; + struct folio *folio =3D page_folio(page); =20 - local_lock(&mlock_pvec.lock); - pvec =3D this_cpu_ptr(&mlock_pvec.vec); + local_lock(&mlock_fbatch.lock); + fbatch =3D this_cpu_ptr(&mlock_fbatch.fbatch); /* - * TestClearPageMlocked(page) must be left to __munlock_page(), - * which will check whether the page is multiply mlocked. + * folio_test_clear_mlocked(folio) must be left to __munlock_folio(), + * which will check whether the folio is multiply mlocked. */ - - get_page(page); - if (!pagevec_add(pvec, page) || - PageHead(page) || lru_cache_disabled()) - mlock_pagevec(pvec); - local_unlock(&mlock_pvec.lock); + folio_get(folio); + if (!folio_batch_add(fbatch, folio) || + folio_test_large(folio) || lru_cache_disabled()) + mlock_folio_batch(fbatch); + local_unlock(&mlock_fbatch.lock); } =20 static int mlock_pte_range(pmd_t *pmd, unsigned long addr, --=20 2.39.0