From nobody Fri Sep 25 08:47:33 2026 Received: from m16.mail.126.com (m16.mail.126.com [220.197.31.9]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 3C46C332EBB for ; Tue, 15 Sep 2026 04:19:55 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=220.197.31.9 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1789445999; cv=none; b=XGrURTsdEGzbQjNg6AGTQEjic2Lb2p9sqKIYLaf116rnuCcejka59ljls6lIqokWpNgQ1UeHz+V9peP8rp+fcVe2OQtkXAw45SeDLzMJzm5hXRrWC09fyr/ik9ScbMba9+UeJBt7QrFBGOrJd4sgfYSEwRnnoZmmmzDWVsM95Ws= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1789445999; c=relaxed/simple; bh=CXMJf1VxpUJkNaOms1lHQUBt1ld2pxaY9LqsuqnoX1Q=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=ieqyhE8y1Nm+SlqGoOJ+65qeUxjvGbHW2FOdpBuVTzz+/9mdTzp/b/cUMfN8DsIVtrTeLsDU9VdrlnZHTj1z+p6lvKGhaZbEKTPSq1HRoKVNT6Uls3tOOIwFQ0UjFGL8PEUpwAJJfez98TtvTPEq1Oa6YxD+2rSck7rYOpYP93g= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=fail (p=quarantine dis=none) header.from=xiaomi.com; spf=pass smtp.mailfrom=126.com; dkim=pass (1024-bit key) header.d=126.com header.i=@126.com header.b=buvhhUul; arc=none smtp.client-ip=220.197.31.9 Authentication-Results: smtp.subspace.kernel.org; dmarc=fail (p=quarantine dis=none) header.from=xiaomi.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=126.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=126.com header.i=@126.com header.b="buvhhUul" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=126.com; s=s110527; h=From:To:Subject:Date:Message-ID:MIME-Version; bh=8q Z+qiQJIRwRVOe+x3CNCs894Wm+YjNv+O3pSDGzKS0=; b=buvhhUul9dq2PHLMfD 6W6GWSHmmNzeDsN38qq7p/SE6bc7QJCeVq2bX6FQb+0R7YtUgNAIRxZPjjk4VQjR hMVjHHG0iRh6BihxqAsdgVyiBP5Zmpz00g5CmthecJL8YM9qw8ZPelIiYw4X2dT+ FyAHjV9Tb4E7uKhDpr4nXOJlA= Received: from nzzhao-ThinkCentre-M760t.mioffice.cn (unknown []) by gzsmtp3 (Coremail) with SMTP id PikvCgDXreo9x6hq3jBvHw--.36918S3; Tue, 15 Sep 2026 12:19:11 +0800 (CST) From: Nanzhe Zhao To: linux-f2fs-devel@lists.sourceforge.net, Jaegeuk Kim , Chao Yu Cc: Barry Song , Juan Yescas , Dev Jain , linux-kernel@vger.kernel.org, David Hildenbrand , Bo Zhang , Kalesh Singh , Nanzhe Zhao , Pengfei Li , Ryan Roberts , Nanzhe Zhao Subject: [PATCH v2 01/14] f2fs: extend folio state for large folio write path Date: Tue, 15 Sep 2026 12:18:56 +0800 Message-ID: <20260915041909.2903887-2-zhaonanzhe@xiaomi.com> X-Mailer: git-send-email 2.43.0 In-Reply-To: <20260915041909.2903887-1-zhaonanzhe@xiaomi.com> References: <20260915041909.2903887-1-zhaonanzhe@xiaomi.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-CM-TRANSID: PikvCgDXreo9x6hq3jBvHw--.36918S3 X-Coremail-Antispam: 1Uf129KBjvJXoW3Ww4fKr1DGr1xWF17JryxZrb_yoWDJFWkpF WrGrn0yr48Jry7WF17JF1rZr1Skr92qa1UAFWxG34fA3Waqw1rWF1kJ34jqF43Jry8uFn3 AayYkr1rWF4YyF7anT9S1TB71UUUUU7qnTZGkaVYY2UrUUUUjbIjqfuFe4nvWSU5nxnvy2 9KBjDUYxBIdaVFxhVjvjDU0xZFpf9x07jjksgUUUUU= Sender: nzzhao@126.com X-CM-SenderInfo: xq22xtbr6rjloofrz/xtbBsADGlmqox0AXvwAA3N Content-Type: text/plain; charset="utf-8" Large folio write path needs a subpage status bitmap and write pages pending counter, while keeping compatible with f2fs private flags. Move struct f2fs_folio_state to f2fs.h, add private_flags and subpage state bitmap, and change PAGE_PRIVATE functions to be compatible with f2fs_folio_state. Allocate f2fs_folio_state via kzalloc instead of kmem_cache, since the state size depends on the folio order. Note: Now if a path wants to use f2fs_folio_state, it must call `folio_has_ffs` instead of `folio_test_large`` to make check. Signed-off-by: Nanzhe Zhao --- fs/f2fs/data.c | 72 ++++++++++++++++++++++++++++++++------------------ fs/f2fs/f2fs.h | 70 +++++++++++++++++++++++++++++++++++------------- 2 files changed, 98 insertions(+), 44 deletions(-) diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index a3b02c067e89..b3303109a696 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -32,15 +32,9 @@ =20 static struct kmem_cache *bio_post_read_ctx_cache; static struct kmem_cache *bio_entry_slab; -static struct kmem_cache *ffs_entry_slab; static mempool_t *bio_post_read_ctx_pool; static struct bio_set f2fs_bioset; =20 -struct f2fs_folio_state { - spinlock_t state_lock; - unsigned int read_pages_pending; -}; - struct f2fs_bio { struct work_struct work; struct bio bio; @@ -125,6 +119,9 @@ struct bio_post_read_ctx { block_t fs_blkaddr; }; =20 +static bool __ffs_mark_subrange_uptodate(struct folio *folio, + struct f2fs_folio_state *ffs, size_t offset, size_t len); + /* * Update and unlock a bio's pages, and free the bio. * @@ -139,6 +136,23 @@ struct bio_post_read_ctx { * called (i.e., I/O error or decryption error, but *not* verity error), a= nd * release the bio's reference to the decompress_io_ctx of the page's clus= ter. */ +/* + * Update read_pages_pending. + */ +static inline void f2fs_update_read_folio_pending(struct folio *folio, int= nr) +{ + struct f2fs_folio_state *ffs; + unsigned long flags; + + if (!f2fs_folio_has_ffs(folio)) + return; + + ffs =3D (struct f2fs_folio_state *)folio->private; + spin_lock_irqsave(&ffs->state_lock, flags); + ffs->read_pages_pending +=3D nr; + spin_unlock_irqrestore(&ffs->state_lock, flags); +} + static void f2fs_finish_read_bio(struct bio *bio, bool in_task) { struct folio_iter fi; @@ -147,7 +161,7 @@ static void f2fs_finish_read_bio(struct bio *bio, bool = in_task) =20 bio_for_each_folio_all(fi, bio) { struct folio *folio =3D fi.folio; - unsigned nr_pages =3D fi.length >> PAGE_SHIFT; + unsigned int nr_pages =3D fi.length >> PAGE_SHIFT; bool finished =3D true; =20 if (!folio_test_large(folio) && @@ -2487,17 +2501,31 @@ int f2fs_read_multi_pages(struct compress_ctx *cc, = struct bio **bio_ret, } #endif =20 -static struct f2fs_folio_state *ffs_find_or_alloc(struct folio *folio) +struct f2fs_folio_state *f2fs_ffs_find_or_alloc(struct folio *folio) { - struct f2fs_folio_state *ffs =3D folio->private; + struct f2fs_folio_state *ffs; + unsigned int nr_subpages =3D folio_nr_pages(folio); + unsigned long private =3D (unsigned long)folio->private; + + f2fs_bug_on(F2FS_F_SB(folio), !folio_test_large(folio)); + f2fs_bug_on(F2FS_F_SB(folio), + test_bit(PAGE_PRIVATE_NOT_POINTER, &private)); =20 - if (ffs) - return ffs; + if (f2fs_folio_has_ffs(folio)) + return (struct f2fs_folio_state *)folio->private; =20 - ffs =3D f2fs_kmem_cache_alloc(ffs_entry_slab, - GFP_NOIO | __GFP_ZERO, true, NULL); + ffs =3D f2fs_kmalloc(F2FS_F_SB(folio), + struct_size(ffs, state, BITS_TO_LONGS(2 * nr_subpages)), + GFP_NOFS | __GFP_ZERO); + if (!ffs) + return NULL; =20 spin_lock_init(&ffs->state_lock); + if (folio_test_uptodate(folio)) + bitmap_set(ffs->state, 0, nr_subpages); + if (folio_test_dirty(folio)) + bitmap_set(ffs->state, nr_subpages, nr_subpages); + folio_attach_private(folio, ffs); return ffs; } @@ -2506,7 +2534,7 @@ static void ffs_detach_free(struct folio *folio) { struct f2fs_folio_state *ffs; =20 - if (!folio_test_large(folio)) { + if (!f2fs_folio_has_ffs(folio)) { folio_detach_private(folio); return; } @@ -2516,7 +2544,8 @@ static void ffs_detach_free(struct folio *folio) return; =20 WARN_ON_ONCE(ffs->read_pages_pending !=3D 0); - kmem_cache_free(ffs_entry_slab, ffs); + WARN_ON_ONCE(atomic_read(&ffs->write_pages_pending)); + kfree(ffs); } =20 static int f2fs_read_data_large_folio(struct inode *inode, @@ -2529,7 +2558,7 @@ static int f2fs_read_data_large_folio(struct inode *i= node, pgoff_t index, offset, next_pgofs =3D 0; unsigned max_nr_pages =3D rac ? readahead_count(rac) : folio_nr_pages(folio); - unsigned nrpages; + unsigned int nrpages; struct f2fs_folio_state *ffs; int ret =3D 0; bool folio_in_bio =3D false; @@ -2605,7 +2634,7 @@ static int f2fs_read_data_large_folio(struct inode *i= node, * to prevent from premature folio_end_read() call on folio */ if (folio_test_large(folio)) { - ffs =3D ffs_find_or_alloc(folio); + ffs =3D f2fs_ffs_find_or_alloc(folio); =20 /* set the bitmap to wait */ spin_lock_irq(&ffs->state_lock); @@ -4522,21 +4551,12 @@ int __init f2fs_init_bio_entry_cache(void) if (!bio_entry_slab) return -ENOMEM; =20 - ffs_entry_slab =3D f2fs_kmem_cache_create("f2fs_ffs_slab", - sizeof(struct f2fs_folio_state)); - - if (!ffs_entry_slab) { - kmem_cache_destroy(bio_entry_slab); - return -ENOMEM; - } - return 0; } =20 void f2fs_destroy_bio_entry_cache(void) { kmem_cache_destroy(bio_entry_slab); - kmem_cache_destroy(ffs_entry_slab); } =20 static int f2fs_iomap_begin(struct inode *inode, loff_t offset, loff_t len= gth, diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index ebc621f302e1..9a2366a6c113 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -1618,6 +1618,18 @@ static inline void f2fs_clear_bit(unsigned int nr, c= har *addr); * Layout B: lowest bit should be 0 * page.private is a wrapped pointer. */ + +struct f2fs_folio_state { + spinlock_t state_lock; + unsigned int read_pages_pending; + atomic_t write_pages_pending; + unsigned long private_flags; + /* state[0..nr_subpages - 1] tracks uptodate subpages. + * state[nr_subpages..2 * nr_subpages - 1] tracks dirty subpages. + */ + unsigned long state[]; +}; + enum { PAGE_PRIVATE_NOT_POINTER, /* private contains non-pointer data */ PAGE_PRIVATE_ONGOING_MIGRATION, /* data page which is on-going migrating= */ @@ -1627,6 +1639,13 @@ enum { PAGE_PRIVATE_MAX }; =20 +static inline bool f2fs_folio_has_ffs(const struct folio *folio) +{ + unsigned long private =3D (unsigned long)folio->private; + + return folio_test_large(folio) && private; +} + /* For compression */ enum compress_algorithm_type { COMPRESS_LZO, @@ -2738,13 +2757,35 @@ static inline int inc_valid_block_count(struct f2fs= _sb_info *sbi, return -ENOSPC; } =20 +static inline unsigned long f2fs_folio_get_private_flags(const struct foli= o *folio) +{ + unsigned long private =3D (unsigned long)folio->private; + + if (f2fs_folio_has_ffs(folio)) { + struct f2fs_folio_state *ffs =3D (struct f2fs_folio_state *)private; + + return READ_ONCE(ffs->private_flags); + } + + if (test_bit(PAGE_PRIVATE_NOT_POINTER, &private)) + return private; + + return 0; +} + +static inline unsigned long *f2fs_folio_flags_addr(struct folio *folio) +{ + if (f2fs_folio_has_ffs(folio)) + return &((struct f2fs_folio_state *)folio->private)->private_flags; + + return (unsigned long *)&folio->private; +} + #define PAGE_PRIVATE_GET_FUNC(name, flagname) \ static inline bool folio_test_f2fs_##name(const struct folio *folio) \ { \ - unsigned long priv =3D (unsigned long)folio->private; \ - unsigned long v =3D (1UL << PAGE_PRIVATE_NOT_POINTER) | \ - (1UL << PAGE_PRIVATE_##flagname); \ - return (priv & v) =3D=3D v; \ + return f2fs_folio_get_private_flags(folio) & \ + (1UL << PAGE_PRIVATE_##flagname); \ } \ static inline bool page_private_##name(struct page *page) \ { \ @@ -2756,14 +2797,10 @@ static inline bool page_private_##name(struct page = *page) \ #define PAGE_PRIVATE_SET_FUNC(name, flagname) \ static inline void folio_set_f2fs_##name(struct folio *folio) \ { \ - unsigned long v =3D (1UL << PAGE_PRIVATE_NOT_POINTER) | \ - (1UL << PAGE_PRIVATE_##flagname); \ if (!folio->private) \ - folio_attach_private(folio, (void *)v); \ - else { \ - v |=3D (unsigned long)folio->private; \ - folio->private =3D (void *)v; \ - } \ + folio_attach_private(folio, \ + (void *)BIT(PAGE_PRIVATE_NOT_POINTER)); \ + set_bit(PAGE_PRIVATE_##flagname, f2fs_folio_flags_addr(folio)); \ } \ static inline void set_page_private_##name(struct page *page) \ { \ @@ -2776,13 +2813,10 @@ static inline void set_page_private_##name(struct p= age *page) \ #define PAGE_PRIVATE_CLEAR_FUNC(name, flagname) \ static inline void folio_clear_f2fs_##name(struct folio *folio) \ { \ - unsigned long v =3D (unsigned long)folio->private; \ - \ - v &=3D ~(1UL << PAGE_PRIVATE_##flagname); \ - if (v =3D=3D (1UL << PAGE_PRIVATE_NOT_POINTER)) \ + clear_bit(PAGE_PRIVATE_##flagname, \ + f2fs_folio_flags_addr(folio)); \ + if (folio->private =3D=3D (void *)BIT(PAGE_PRIVATE_NOT_POINTER)) \ folio_detach_private(folio); \ - else \ - folio->private =3D (void *)v; \ } \ static inline void clear_page_private_##name(struct page *page) \ { \ @@ -2808,7 +2842,7 @@ PAGE_PRIVATE_CLEAR_FUNC(atomic, ATOMIC_WRITE); =20 static inline unsigned long folio_get_f2fs_data(struct folio *folio) { - unsigned long data =3D (unsigned long)folio->private; + unsigned long data =3D f2fs_folio_get_private_flags(folio); =20 if (!test_bit(PAGE_PRIVATE_NOT_POINTER, &data)) return 0; --=20 2.43.0 From nobody Fri Sep 25 08:47:33 2026 Received: from m16.mail.126.com (m16.mail.126.com [220.197.31.6]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id AB95536829D for ; Tue, 15 Sep 2026 04:20:02 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=220.197.31.6 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1789446012; cv=none; b=YDpzdzpUnmiJeOzTpc2sQOG/fo7ire1Xp3W/okbWcabe69iHsUuQC2XgOa1MyAPMzkYZY/D4NIoO94qt9wQgjZBtQruLHBh+Ec+GXXMnuEtAk/Q1AqrpS9xm4fJ2YPptdbXpGI9mnoRoby3OJe2UeMkNU0Qtb8UAsD8ONyeUe5o= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1789446012; c=relaxed/simple; bh=N6WNVjAezz9llWorl4lKkATGSb+I46z5TOsht4sQuFc=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=Vf4bJU9rw4RL00VeYga8sjDN+oszEOCPFdL8FJLaagytRSBMrWI71ReP8csRkZKTVkRZAmwNV5+JiVS9nUyRwfCwDTDCNKkVscAb0JVZPhP2PzTHGWZtTvhtFOpfPM0tvRMwBNFCt1nYnAqQUVGdl2zT/8wLe7KjHSRlP6xAtiU= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=fail (p=quarantine dis=none) header.from=xiaomi.com; spf=pass smtp.mailfrom=126.com; dkim=pass (1024-bit key) header.d=126.com header.i=@126.com header.b=C+EtqRuB; arc=none smtp.client-ip=220.197.31.6 Authentication-Results: smtp.subspace.kernel.org; dmarc=fail (p=quarantine dis=none) header.from=xiaomi.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=126.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=126.com header.i=@126.com header.b="C+EtqRuB" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=126.com; s=s110527; h=From:To:Subject:Date:Message-ID:MIME-Version; bh=WC x93R3vjZNxUVy7wasf82kOWRkLEGf7p1dGYEpaSao=; b=C+EtqRuBu4vxquAIYH xWymUbTz6cOGfdzHPz4CA2jMCXVWqiUK00vpjim0DDtfsXQ3mtR0NqCNrsJhiY05 Teq3nF+IYNBNrwavHLFq0lxF59GVeiVuqs8+oiCPxjf+QOva13Ve0YvvtFd4dQjf gc/yUdJunQl9d7AntMxyZXtoU= Received: from nzzhao-ThinkCentre-M760t.mioffice.cn (unknown []) by gzsmtp3 (Coremail) with SMTP id PikvCgDXreo9x6hq3jBvHw--.36918S4; Tue, 15 Sep 2026 12:19:12 +0800 (CST) From: Nanzhe Zhao To: linux-f2fs-devel@lists.sourceforge.net, Jaegeuk Kim , Chao Yu Cc: Barry Song , Juan Yescas , Dev Jain , linux-kernel@vger.kernel.org, David Hildenbrand , Bo Zhang , Kalesh Singh , Nanzhe Zhao , Pengfei Li , Ryan Roberts , Nanzhe Zhao Subject: [PATCH v2 02/14] f2fs: carry subpage offset and count in write IO Date: Tue, 15 Sep 2026 12:18:57 +0800 Message-ID: <20260915041909.2903887-3-zhaonanzhe@xiaomi.com> X-Mailer: git-send-email 2.43.0 In-Reply-To: <20260915041909.2903887-1-zhaonanzhe@xiaomi.com> References: <20260915041909.2903887-1-zhaonanzhe@xiaomi.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-CM-TRANSID: PikvCgDXreo9x6hq3jBvHw--.36918S4 X-Coremail-Antispam: 1Uf129KBjvJXoW3tr4rZr1xKFy8XryxXr1UZFb_yoWkKF4DpF 15Ga4kGrs5GF17Wwn7tFn8JF1Sg343Wr48GFWfCa4Sqa10qwnag3WSkas8ZF15Kr4DCFyv qF4FyF1UWF1UJF7anT9S1TB71UUUUU7qnTZGkaVYY2UrUUUUjbIjqfuFe4nvWSU5nxnvy2 9KBjDUYxBIdaVFxhVjvjDU0xZFpf9x07j5Z23UUUUU= Sender: nzzhao@126.com X-CM-SenderInfo: xq22xtbr6rjloofrz/xtbBogDGlmqox0AMEwAA3o Content-Type: text/plain; charset="utf-8" Large folio write paths need to submit I/O for a range inside a folio instead of always submitting the whole folio from offset zero. Add folio_offset and folio_blkcnt to f2fs_io_info to describe the block offset inside the folio and the number of contiguous blocks covered by the I/O. Apply the new fields to the bio submit paths that need the subpage offset or contiguous block count. Signed-off-by: Nanzhe Zhao --- fs/f2fs/checkpoint.c | 2 ++ fs/f2fs/compress.c | 1 + fs/f2fs/data.c | 63 +++++++++++++++++++++++++++++++------------- fs/f2fs/f2fs.h | 11 ++++++++ fs/f2fs/gc.c | 2 ++ fs/f2fs/inline.c | 1 + fs/f2fs/node.c | 2 ++ fs/f2fs/segment.c | 4 ++- 8 files changed, 66 insertions(+), 20 deletions(-) diff --git a/fs/f2fs/checkpoint.c b/fs/f2fs/checkpoint.c index e95cd2118750..cd379183449e 100644 --- a/fs/f2fs/checkpoint.c +++ b/fs/f2fs/checkpoint.c @@ -264,6 +264,7 @@ static struct folio *__get_meta_folio(struct f2fs_sb_in= fo *sbi, pgoff_t index, .op_flags =3D REQ_META | REQ_PRIO, .old_blkaddr =3D index, .new_blkaddr =3D index, + .folio_blkcnt =3D 1, .encrypted_page =3D NULL, .is_por =3D !is_meta ? 1 : 0, }; @@ -456,6 +457,7 @@ int f2fs_ra_meta_pages(struct f2fs_sb_info *sbi, block_= t start, int nrpages, .type =3D META, .op =3D REQ_OP_READ, .op_flags =3D sync ? (REQ_META | REQ_PRIO) : REQ_RAHEAD, + .folio_blkcnt =3D 1, .encrypted_page =3D NULL, .in_list =3D 0, .is_por =3D (type =3D=3D META_POR) ? 1 : 0, diff --git a/fs/f2fs/compress.c b/fs/f2fs/compress.c index ce88092d9ce2..b130e43b4566 100644 --- a/fs/f2fs/compress.c +++ b/fs/f2fs/compress.c @@ -1282,6 +1282,7 @@ static int f2fs_write_compressed_pages(struct compres= s_ctx *cc, .op_flags =3D wbc_to_write_flags(wbc), .old_blkaddr =3D NEW_ADDR, .page =3D NULL, + .folio_blkcnt =3D 1, .encrypted_page =3D NULL, .compressed_page =3D NULL, .io_type =3D io_type, diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index b3303109a696..f14044e8a6f4 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -769,6 +769,9 @@ int f2fs_submit_page_bio(struct f2fs_io_info *fio) struct folio *fio_folio =3D fio->folio; struct folio *data_folio =3D fio->encrypted_page ? page_folio(fio->encrypted_page) : fio_folio; + pgoff_t fio_lblk =3D F2FS_FIO_LBLK(fio); + size_t bio_offset =3D F2FS_FIO_BIO_OFFSET(fio); + size_t bio_len =3D F2FS_FIO_BIO_SIZE(fio); =20 if (!f2fs_is_valid_blkaddr(fio->sbi, fio->new_blkaddr, fio->is_por ? META_POR : (__is_meta_io(fio) ? @@ -781,11 +784,11 @@ int f2fs_submit_page_bio(struct f2fs_io_info *fio) bio =3D __bio_alloc(fio, 1); =20 f2fs_set_bio_crypt_ctx(bio, fio_folio->mapping->host, - fio_folio->index, fio, GFP_NOIO); - bio_add_folio_nofail(bio, data_folio, folio_size(data_folio), 0); + fio_lblk, fio, GFP_NOIO); + bio_add_folio_nofail(bio, data_folio, bio_len, bio_offset); =20 if (fio->io_wbc && !is_read_io(fio->op)) - wbc_account_cgroup_owner(fio->io_wbc, fio_folio, PAGE_SIZE); + wbc_account_cgroup_owner(fio->io_wbc, fio_folio, bio_len); =20 inc_page_count(fio->sbi, is_read_io(fio->op) ? __read_io_type(data_folio) : WB_DATA_TYPE(fio->folio, false)); @@ -830,7 +833,8 @@ static bool io_is_mergeable(struct f2fs_sb_info *sbi, s= truct bio *bio, } =20 static void add_bio_entry(struct f2fs_sb_info *sbi, struct bio *bio, - struct folio *folio, enum temp_type temp) + struct folio *folio, size_t len, size_t offset, + enum temp_type temp) { struct f2fs_bio_info *io =3D sbi->write_io[DATA] + temp; struct bio_entry *be; @@ -839,7 +843,7 @@ static void add_bio_entry(struct f2fs_sb_info *sbi, str= uct bio *bio, be->bio =3D bio; bio_get(bio); =20 - bio_add_folio_nofail(bio, folio, folio_size(folio), 0); + bio_add_folio_nofail(bio, folio, len, offset); =20 f2fs_down_write(&io->bio_list_lock); list_add_tail(&be->list, &io->bio_list); @@ -856,6 +860,9 @@ static int add_ipu_page(struct f2fs_io_info *fio, struc= t bio **bio, struct folio *folio) { struct folio *fio_folio =3D fio->folio; + pgoff_t fio_lblk =3D F2FS_FIO_LBLK(fio); + size_t bio_offset =3D F2FS_FIO_BIO_OFFSET(fio); + size_t bio_len =3D F2FS_FIO_BIO_SIZE(fio); struct f2fs_sb_info *sbi =3D fio->sbi; enum temp_type temp; bool found =3D false; @@ -878,8 +885,8 @@ static int add_ipu_page(struct f2fs_io_info *fio, struc= t bio **bio, fio->new_blkaddr)); if (f2fs_crypt_mergeable_bio(*bio, fio_folio->mapping->host, - fio_folio->index, fio) && - bio_add_folio(*bio, folio, folio_size(folio), 0)) { + fio_lblk, fio) && + bio_add_folio(*bio, folio, bio_len, bio_offset)) { ret =3D 0; break; } @@ -993,6 +1000,10 @@ int f2fs_merge_page_bio(struct f2fs_io_info *fio) struct folio *data_folio =3D fio->encrypted_page ? page_folio(fio->encrypted_page) : fio->folio; struct folio *folio =3D fio->folio; + pgoff_t fio_lblk =3D F2FS_FIO_LBLK(fio); + unsigned int fio_cnt =3D F2FS_FIO_BLKCNT(fio); + size_t bio_offset =3D F2FS_FIO_BIO_OFFSET(fio); + size_t bio_len =3D F2FS_FIO_BIO_SIZE(fio); =20 if (!f2fs_is_valid_blkaddr(fio->sbi, fio->new_blkaddr, __is_meta_io(fio) ? META_GENERIC : DATA_GENERIC)) @@ -1007,9 +1018,10 @@ int f2fs_merge_page_bio(struct f2fs_io_info *fio) if (!bio) { bio =3D __bio_alloc(fio, BIO_MAX_VECS); f2fs_set_bio_crypt_ctx(bio, folio->mapping->host, - folio->index, fio, GFP_NOIO); + fio_lblk, fio, GFP_NOIO); =20 - add_bio_entry(fio->sbi, bio, data_folio, fio->temp); + add_bio_entry(fio->sbi, bio, data_folio, bio_len, + bio_offset, fio->temp); } else { if (add_ipu_page(fio, &bio, data_folio)) goto alloc_new; @@ -1020,7 +1032,7 @@ int f2fs_merge_page_bio(struct f2fs_io_info *fio) =20 inc_page_count(fio->sbi, WB_DATA_TYPE(folio, false)); =20 - *fio->last_block =3D fio->new_blkaddr; + *fio->last_block =3D fio->new_blkaddr + fio_cnt - 1; *fio->bio =3D bio; =20 return 0; @@ -1056,6 +1068,10 @@ void f2fs_submit_page_write(struct f2fs_io_info *fio) struct folio *bio_folio; struct f2fs_lock_context lc; enum count_type type; + pgoff_t fio_lblk; + unsigned int fio_cnt; + size_t bio_offset; + size_t bio_len; =20 f2fs_bug_on(sbi, is_read_io(fio->op)); =20 @@ -1094,33 +1110,39 @@ void f2fs_submit_page_write(struct f2fs_io_info *fi= o) /* set submitted =3D true as a return value */ fio->submitted =3D 1; =20 + fio_lblk =3D F2FS_FIO_LBLK(fio); + fio_cnt =3D F2FS_FIO_BLKCNT(fio); + type =3D WB_DATA_TYPE(bio_folio, fio->compressed_page); - inc_page_count(sbi, type); + for (unsigned int i =3D 0; i < fio_cnt; i++) + inc_page_count(sbi, type); =20 if (io->bio && (!io_is_mergeable(sbi, io->bio, io, fio, io->last_block_in_bio, fio->new_blkaddr) || !f2fs_crypt_mergeable_bio(io->bio, fio_inode(fio), - bio_folio->index, fio))) + fio_lblk, fio))) __submit_merged_bio(io); alloc_new: if (io->bio =3D=3D NULL) { io->bio =3D __bio_alloc(fio, BIO_MAX_VECS); f2fs_set_bio_crypt_ctx(io->bio, fio_inode(fio), - bio_folio->index, fio, GFP_NOIO); + fio_lblk, fio, GFP_NOIO); io->fio =3D *fio; } =20 - if (!bio_add_folio(io->bio, bio_folio, folio_size(bio_folio), 0)) { + bio_offset =3D F2FS_FIO_BIO_OFFSET(fio); + bio_len =3D F2FS_FIO_BIO_SIZE(fio); + + if (!bio_add_folio(io->bio, bio_folio, bio_len, bio_offset)) { __submit_merged_bio(io); goto alloc_new; } =20 if (fio->io_wbc) - wbc_account_cgroup_owner(fio->io_wbc, fio->folio, - folio_size(fio->folio)); + wbc_account_cgroup_owner(fio->io_wbc, fio->folio, bio_len); =20 - io->last_block_in_bio =3D fio->new_blkaddr; + io->last_block_in_bio =3D fio->new_blkaddr + fio_cnt - 1; =20 trace_f2fs_submit_folio_write(fio->folio, fio); #ifdef CONFIG_BLK_DEV_ZONED @@ -2999,7 +3021,8 @@ int f2fs_do_write_data_page(struct f2fs_io_info *fio) set_new_dnode(&dn, inode, NULL, NULL, 0); =20 if (need_inplace_update(fio) && - f2fs_lookup_read_extent_cache_block(inode, folio->index, + f2fs_lookup_read_extent_cache_block(inode, + F2FS_FIO_LBLK(fio), &fio->old_blkaddr)) { if (!f2fs_is_valid_blkaddr(fio->sbi, fio->old_blkaddr, DATA_GENERIC_ENHANCE)) @@ -3018,7 +3041,8 @@ int f2fs_do_write_data_page(struct f2fs_io_info *fio) if (fio->need_lock =3D=3D LOCK_REQ && !f2fs_trylock_op(fio->sbi, &lc)) return -EAGAIN; =20 - err =3D f2fs_get_dnode_of_data(&dn, folio->index, LOOKUP_NODE); + err =3D f2fs_get_dnode_of_data(&dn, F2FS_FIO_LBLK(fio), + LOOKUP_NODE); if (err) goto out; =20 @@ -3121,6 +3145,7 @@ int f2fs_write_single_data_page(struct folio *folio, = int *submitted, .op_flags =3D wbc_to_write_flags(wbc), .old_blkaddr =3D NULL_ADDR, .folio =3D folio, + .folio_blkcnt =3D 1, .encrypted_page =3D NULL, .submitted =3D 0, .compr_blocks =3D compr_blocks, diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 9a2366a6c113..326382ede1c3 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -1348,6 +1348,8 @@ struct f2fs_io_info { blk_opf_t op_flags; /* req_flag_bits */ block_t new_blkaddr; /* new block address to be written */ block_t old_blkaddr; /* old block address before Cow */ + pgoff_t folio_offset; /* offset in large folio */ + unsigned int folio_blkcnt; /* block count in large folio */ union { struct page *page; /* page to be written */ struct folio *folio; @@ -1368,6 +1370,15 @@ struct f2fs_io_info { sector_t *last_block; /* last block number in bio */ }; =20 +#define F2FS_FIO_LBLK(fio) \ + (((fio)->folio)->index + (fio)->folio_offset) +#define F2FS_FIO_BLKCNT(fio) \ + ((fio)->folio_blkcnt) +#define F2FS_FIO_BIO_OFFSET(fio) \ + ((fio)->folio_offset << PAGE_SHIFT) +#define F2FS_FIO_BIO_SIZE(fio) \ + F2FS_BLK_TO_BYTES((fio)->sbi, F2FS_FIO_BLKCNT(fio)) + struct bio_entry { struct bio *bio; struct list_head list; diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c index c70deac9f1a3..16fce62740d7 100644 --- a/fs/f2fs/gc.c +++ b/fs/f2fs/gc.c @@ -1227,6 +1227,7 @@ static int ra_data_block(struct inode *inode, pgoff_t= index) .temp =3D COLD, .op =3D REQ_OP_READ, .op_flags =3D 0, + .folio_blkcnt =3D 1, .encrypted_page =3D NULL, .in_list =3D 0, }; @@ -1337,6 +1338,7 @@ static int move_data_block(struct inode *inode, block= _t bidx, .temp =3D COLD, .op =3D REQ_OP_READ, .op_flags =3D 0, + .folio_blkcnt =3D 1, .encrypted_page =3D NULL, .in_list =3D 0, }; diff --git a/fs/f2fs/inline.c b/fs/f2fs/inline.c index 925bdaad013c..8db083b545bc 100644 --- a/fs/f2fs/inline.c +++ b/fs/f2fs/inline.c @@ -151,6 +151,7 @@ int f2fs_convert_inline_folio(struct dnode_of_data *dn,= struct folio *folio) .op =3D REQ_OP_WRITE, .op_flags =3D REQ_SYNC | REQ_PRIO, .folio =3D folio, + .folio_blkcnt =3D 1, .encrypted_page =3D NULL, .io_type =3D FS_DATA_IO, }; diff --git a/fs/f2fs/node.c b/fs/f2fs/node.c index f16144b75d11..a19abc34c14a 100644 --- a/fs/f2fs/node.c +++ b/fs/f2fs/node.c @@ -1475,6 +1475,7 @@ static int read_node_folio(struct folio *folio, blk_o= pf_t op_flags) .op =3D REQ_OP_READ, .op_flags =3D op_flags, .folio =3D folio, + .folio_blkcnt =3D 1, .encrypted_page =3D NULL, }; int err; @@ -1772,6 +1773,7 @@ static bool __write_node_folio(struct folio *folio, b= ool atomic, bool do_fsync, .op =3D REQ_OP_WRITE, .op_flags =3D wbc_to_write_flags(wbc), .folio =3D folio, + .folio_blkcnt =3D 1, .encrypted_page =3D NULL, .submitted =3D 0, .io_type =3D io_type, diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c index ffdd2c510b3c..644ddfb984c9 100644 --- a/fs/f2fs/segment.c +++ b/fs/f2fs/segment.c @@ -3873,7 +3873,8 @@ static int __get_segment_type_6(struct f2fs_io_info *= fio) if (file_is_cold(inode) || f2fs_need_compress_data(inode)) return CURSEG_COLD_DATA; =20 - type =3D __get_age_segment_type(inode, fio->folio->index); + type =3D __get_age_segment_type(inode, + F2FS_FIO_LBLK(fio)); if (type !=3D NO_CHECK_TYPE) return type; =20 @@ -4195,6 +4196,7 @@ void f2fs_do_write_meta_page(struct f2fs_sb_info *sbi= , struct folio *folio, .old_blkaddr =3D folio->index, .new_blkaddr =3D folio->index, .folio =3D folio, + .folio_blkcnt =3D 1, .encrypted_page =3D NULL, .in_list =3D 0, }; --=20 2.43.0 From nobody Fri Sep 25 08:47:33 2026 Received: from m16.mail.126.com (m16.mail.126.com [117.135.210.6]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 1B2B242E425 for ; Tue, 15 Sep 2026 04:19:59 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=117.135.210.6 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1789446003; cv=none; b=iV1luh1nXLBrHWipX1TXZ6FzT+0J4zdEUzSkflPyjmwKCGJIOwVpbF2++gwT/EQnMy+Rc4PIYuAMimPMqB/K9l63x4nzJiXgsNc6ajzjsAKZmtkep7t2MhJqNZ6Eoi7+QfNx1KrhLrudZs/JyE2Mct6AKkxOkww0uxMaExrm+1Q= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1789446003; c=relaxed/simple; bh=kPH71QfU3VqoCP939Pph5as7+QcO9n9Ygh2B6N8kdds=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=QSMD0xwUeH4IvPSZEiDd2YFcnvPfyFb1csrqKHPxE3LwPeN45Jgr0R++eZcxo54CPZJtnWwIn79MASC2rVV6APjCX6HWIMbmc553Eq9Laf8Uv4XpTzkvUP3gc9clwQZQfesKpVoG8liZ+Nof79HbZfP6WukWNOKCQ2UkrACthPo= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=fail (p=quarantine dis=none) header.from=xiaomi.com; spf=pass smtp.mailfrom=126.com; dkim=pass (1024-bit key) header.d=126.com header.i=@126.com header.b=SUz6a59P; arc=none smtp.client-ip=117.135.210.6 Authentication-Results: smtp.subspace.kernel.org; dmarc=fail (p=quarantine dis=none) header.from=xiaomi.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=126.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=126.com header.i=@126.com header.b="SUz6a59P" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=126.com; s=s110527; h=From:To:Subject:Date:Message-ID:MIME-Version; bh=p7 bex7KMzHy93Zk9zp68xSNwNNSpGZvyR0Q/fJUaVaU=; b=SUz6a59PeAy+F3KD+C mqdo0fVCJ/pUCTDl6QLI8JhcFtkMWQMrdqvz9THV0YBbi569ju6XZT4EOjrWHFUO vydQz5sMWXsjW+tbhc0OnAUc27IZJlv7e/jmGS2mxkoHAOGFBuf3ehnKWQTvYMtZ kRQUJlf28/ejslbLPw2gaLmpg= Received: from nzzhao-ThinkCentre-M760t.mioffice.cn (unknown []) by gzsmtp3 (Coremail) with SMTP id PikvCgDXreo9x6hq3jBvHw--.36918S5; Tue, 15 Sep 2026 12:19:13 +0800 (CST) From: Nanzhe Zhao To: linux-f2fs-devel@lists.sourceforge.net, Jaegeuk Kim , Chao Yu Cc: Barry Song , Juan Yescas , Dev Jain , linux-kernel@vger.kernel.org, David Hildenbrand , Bo Zhang , Kalesh Singh , Nanzhe Zhao , Pengfei Li , Ryan Roberts , Nanzhe Zhao Subject: [PATCH v2 03/14] f2fs: support regular file buffered writes on large folios Date: Tue, 15 Sep 2026 12:18:58 +0800 Message-ID: <20260915041909.2903887-4-zhaonanzhe@xiaomi.com> X-Mailer: git-send-email 2.43.0 In-Reply-To: <20260915041909.2903887-1-zhaonanzhe@xiaomi.com> References: <20260915041909.2903887-1-zhaonanzhe@xiaomi.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-CM-TRANSID: PikvCgDXreo9x6hq3jBvHw--.36918S5 X-Coremail-Antispam: 1Uf129KBjvAXoW3ZFWxZFWkAFW5CF4kAFWUJwb_yoW8Gryruo WfJws0qw4rXrnrZFWjkw17tFyUuan8C3s8JFs3ursxuFnrX3s09w47Kwn8Xayxur15GF13 CryxJa13GrW7Jr1fn29KB7ZKAUJUUUU8529EdanIXcx71UUUUU7v73VFW2AGmfu7bjvjm3 AaLaJ3UbIYCTnIWIevJa73UjIFyTuYvjxUcXo2UUUUU Sender: nzzhao@126.com X-CM-SenderInfo: xq22xtbr6rjloofrz/xtbBowHGlmqox0FoZwAA35 Content-Type: text/plain; charset="utf-8" To avoid the complexity of unlocking a large folio in write_begin, preallocate partial blocks for inodes that can use large folios. During write_begin, read only the partial head and tail 4K subpages that need read-before-write, and skip read I/O for the full middle subpages covered by the write. Signed-off-by: Nanzhe Zhao --- fs/f2fs/data.c | 300 +++++++++++++++++++++++++++++++++++++++++++++++-- fs/f2fs/f2fs.h | 1 + fs/f2fs/file.c | 17 ++- 3 files changed, 304 insertions(+), 14 deletions(-) diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index f14044e8a6f4..a490df4f50b3 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -1233,6 +1233,35 @@ static void f2fs_submit_page_read(struct inode *inod= e, struct fsverity_info *vi, f2fs_submit_read_bio(sbi, bio, DATA); } =20 +/* + * Synchronously read a single 4K subpage by reusing f2fs_submit_page_read= () + * so that iostat, trace, blk-crypto and post-read handling are all preser= ved. + * The caller must have already allocated ffs for the folio. + */ +static int f2fs_submit_page_read_sync(struct inode *inode, struct folio *f= olio, + pgoff_t index, block_t blkaddr) +{ + struct f2fs_folio_state *ffs =3D folio->private; + + /* Add bias so end_io does not call folio_end_read(). */ + f2fs_update_read_folio_pending(folio, 1); + + f2fs_submit_page_read(inode, NULL, folio, index, blkaddr, + REQ_OP_READ, false); + + /* Wait until all bios have completed (pending drops back to our bias). */ + while (READ_ONCE(ffs->read_pages_pending) !=3D 1) + f2fs_io_schedule_timeout(DEFAULT_SCHEDULE_TIMEOUT); + + /* Remove the bias. */ + f2fs_update_read_folio_pending(folio, -1); + + if (!f2fs_ffs_test_blk_uptodate(folio, index)) + return -EIO; + + return 0; +} + static void __set_data_blkaddr(struct dnode_of_data *dn, block_t blkaddr) { __le32 *addr =3D get_dnode_addr(dn->inode, dn->node_folio); @@ -2570,6 +2599,114 @@ static void ffs_detach_free(struct folio *folio) kfree(ffs); } =20 +bool f2fs_ffs_test_blk_uptodate(const struct folio *folio, pgoff_t index) +{ + struct f2fs_folio_state *ffs; + unsigned int idx; + + f2fs_bug_on(F2FS_F_SB(folio), !folio_contains(folio, index)); + + if (folio_test_uptodate(folio)) + return true; + + if (!f2fs_folio_has_ffs(folio)) + return false; + + ffs =3D (struct f2fs_folio_state *)folio->private; + idx =3D index - folio->index; + return test_bit(idx, ffs->state); +} + +static bool __ffs_mark_subrange_uptodate(struct folio *folio, + struct f2fs_folio_state *ffs, size_t offset, size_t len) +{ + unsigned int nr_subpages =3D folio_nr_pages(folio); + unsigned int start, end; + + start =3D offset >> PAGE_SHIFT; + end =3D (offset + len + PAGE_SIZE - 1) >> PAGE_SHIFT; + end =3D min(end, nr_subpages); + + bitmap_set(ffs->state, start, end - start); + return bitmap_full(ffs->state, nr_subpages); +} + +static void f2fs_ffs_mark_subrange_uptodate(struct folio *folio, size_t of= fset, + size_t len) +{ + struct f2fs_folio_state *ffs; + unsigned long flags; + bool mark_uptodate =3D false; + + f2fs_bug_on(F2FS_F_SB(folio), offset + len > folio_size(folio)); + + if (!f2fs_folio_has_ffs(folio)) { + folio_mark_uptodate(folio); + return; + } + + ffs =3D (struct f2fs_folio_state *)folio->private; + spin_lock_irqsave(&ffs->state_lock, flags); + mark_uptodate =3D __ffs_mark_subrange_uptodate(folio, ffs, offset, len) && + !ffs->read_pages_pending; + spin_unlock_irqrestore(&ffs->state_lock, flags); + if (mark_uptodate) + folio_mark_uptodate(folio); +} + +static void f2fs_ffs_mark_subrange_dirty(struct folio *folio, + size_t offset, size_t len) +{ + struct f2fs_folio_state *ffs; + unsigned int nr_subpages, start, end; + unsigned long flags; + + f2fs_bug_on(F2FS_F_SB(folio), offset + len > folio_size(folio)); + + if (!f2fs_folio_has_ffs(folio)) + return; + + ffs =3D (struct f2fs_folio_state *)folio->private; + nr_subpages =3D folio_nr_pages(folio); + start =3D offset >> PAGE_SHIFT; + end =3D (offset + len + PAGE_SIZE - 1) >> PAGE_SHIFT; + end =3D min(end, nr_subpages); + + spin_lock_irqsave(&ffs->state_lock, flags); + bitmap_set(ffs->state, nr_subpages + start, end - start); + spin_unlock_irqrestore(&ffs->state_lock, flags); +} + +static bool find_next_valid_block(const struct folio *folio, + size_t orig_off, size_t *need_off, + size_t len) +{ + size_t start =3D orig_off; + size_t end =3D start + len; + size_t head, tail; + pgoff_t index; + + if (start & (PAGE_SIZE - 1)) { + head =3D round_down(start, PAGE_SIZE); + index =3D folio->index + (head >> PAGE_SHIFT); + if (!f2fs_ffs_test_blk_uptodate(folio, index)) { + *need_off =3D head; + return true; + } + } + + if (end & (PAGE_SIZE - 1)) { + tail =3D round_down(end - 1, PAGE_SIZE); + index =3D folio->index + (tail >> PAGE_SHIFT); + if (!f2fs_ffs_test_blk_uptodate(folio, index)) { + *need_off =3D tail; + return true; + } + } + + return false; +} + static int f2fs_read_data_large_folio(struct inode *inode, struct fsverity_info *vi, struct readahead_control *rac, struct folio *folio) @@ -3927,6 +4064,110 @@ static int prepare_atomic_write_begin(struct f2fs_s= b_info *sbi, return 0; } =20 +static int prepare_large_folio_write_begin(struct inode *inode, + struct folio *folio, loff_t pos, + unsigned int len) +{ + struct f2fs_sb_info *sbi =3D F2FS_I_SB(inode); + struct f2fs_folio_state *ffs; + size_t ori_off =3D offset_in_folio(folio, pos); + size_t need_off =3D ori_off; + int err =3D 0; + + len =3D min_t(unsigned int, len, folio_size(folio) - ori_off); + + /* + * When folio minimum order is non-zero, the fsverity + * page_cache_write() path enters f2fs_write_begin() via + * aops->write_begin without going through f2fs_write_iter(), + * so preallocation from f2fs_write_iter() is skipped. In that + * case, if FI_PREALLOCATED_ALL is not set, we must preallocate + * the write blocks here. + */ + if (!is_inode_flag_set(inode, FI_PREALLOCATED_ALL)) { + struct f2fs_map_blocks map =3D {}; + + map.m_lblk =3D F2FS_BYTES_TO_BLK(sbi, pos); + map.m_len =3D F2FS_BLK_ALIGN(sbi, pos + len) - map.m_lblk; + + if (!IS_DEVICE_ALIASING(inode)) + map.m_may_create =3D true; + map.m_seg_type =3D NO_CHECK_TYPE; + + err =3D f2fs_map_blocks(inode, &map, F2FS_GET_BLOCK_PRE_AIO); + if (err) + return err; + } + + ffs =3D f2fs_ffs_find_or_alloc(folio); + if (!ffs) + return -ENOMEM; + + /* Skip read if the folio is already fully uptodate or the write + * covers the entire folio. + */ + if (folio_test_uptodate(folio) || len =3D=3D folio_size(folio)) + return 0; + + if (!f2fs_verity_in_progress(inode) && + !(pos & (PAGE_SIZE - 1)) && (pos + len) >=3D i_size_read(inode)) { + size_t zoff =3D ori_off + len; + + if (zoff < folio_size(folio)) + folio_zero_segment(folio, zoff, folio_size(folio)); + return 0; + } + + /* Inline data must have been converted before reaching here. */ + f2fs_bug_on(sbi, f2fs_has_inline_data(inode)); + + while (find_next_valid_block(folio, ori_off, + &need_off, len)) { + struct dnode_of_data dn; + pgoff_t index =3D folio->index + (need_off >> PAGE_SHIFT); + block_t blkaddr; + bool get_dn =3D false; + + if (!f2fs_lookup_read_extent_cache_block(inode, index, + &blkaddr)) { + if (IS_DEVICE_ALIASING(inode)) + return -ENODATA; + + set_new_dnode(&dn, inode, NULL, NULL, 0); + err =3D f2fs_get_dnode_of_data(&dn, index, LOOKUP_NODE); + if (err) + return err; + get_dn =3D true; + blkaddr =3D dn.data_blkaddr; + + if (blkaddr =3D=3D NEW_ADDR) { + size_t off =3D offset_in_folio(folio, + index << PAGE_SHIFT); + + folio_zero_segment(folio, off, off + PAGE_SIZE); + f2fs_ffs_mark_subrange_uptodate(folio, off, + PAGE_SIZE); + goto out; + } + + if (!f2fs_is_valid_blkaddr(sbi, blkaddr, + DATA_GENERIC_ENHANCE_READ)) { + err =3D -EFSCORRUPTED; + goto out; + } + } + + err =3D f2fs_submit_page_read_sync(inode, folio, index, blkaddr); +out: + if (get_dn) + f2fs_put_dnode(&dn); + if (err) + return err; + } + + return 0; +} + static int f2fs_write_begin(const struct kiocb *iocb, struct address_space *mapping, loff_t pos, unsigned len, struct folio **foliop, @@ -3938,6 +4179,7 @@ static int f2fs_write_begin(const struct kiocb *iocb, pgoff_t index =3D pos >> PAGE_SHIFT; bool need_balance =3D false; block_t blkaddr =3D NULL_ADDR; + fgf_t fgp =3D FGP_LOCK | FGP_WRITE | FGP_CREAT; int err =3D 0; =20 trace_f2fs_write_begin(inode, pos, len); @@ -3985,9 +4227,15 @@ static int f2fs_write_begin(const struct kiocb *iocb, * Do not use FGP_STABLE to avoid deadlock. * Will wait that below with our IO control. */ - folio =3D f2fs_filemap_get_folio(mapping, index, - FGP_LOCK | FGP_WRITE | FGP_CREAT, - mapping_gfp_mask(mapping)); + /* + * Inline data is carried in the first subpage. Keep the index #0 + * folio order-0 so that the existing inline read/convert paths + * (prepare_write_begin()) still apply; large folios are used from + * the second folio on. + */ + fgp |=3D fgf_set_order(f2fs_has_inline_data(inode) ? PAGE_SIZE : len); + folio =3D f2fs_filemap_get_folio(mapping, index, fgp, + mapping_gfp_mask(mapping)); if (IS_ERR(folio)) { err =3D PTR_ERR(folio); goto fail; @@ -4000,7 +4248,7 @@ static int f2fs_write_begin(const struct kiocb *iocb, if (f2fs_is_atomic_file(inode)) err =3D prepare_atomic_write_begin(sbi, folio, pos, len, &blkaddr, &need_balance); - else + else if (!folio_test_large(folio)) err =3D prepare_write_begin(sbi, folio, pos, len, &blkaddr, &need_balance); if (err) @@ -4021,6 +4269,14 @@ static int f2fs_write_begin(const struct kiocb *iocb, =20 f2fs_folio_wait_writeback(folio, DATA, false, true); =20 + if (folio_test_large(folio)) { + err =3D prepare_large_folio_write_begin(inode, + folio, pos, len); + if (!err) + return 0; + goto put_folio; + } + if (len =3D=3D folio_size(folio) || folio_test_uptodate(folio)) return 0; =20 @@ -4081,15 +4337,20 @@ static int f2fs_write_end(const struct kiocb *iocb, trace_f2fs_write_end(inode, pos, len, copied); =20 /* - * This should be come from len =3D=3D PAGE_SIZE, and we expect copied - * should be PAGE_SIZE. Otherwise, we treat it with zero copied and - * let generic_perform_write() try to copy data again through copied=3D0. + * If a short copy happens on a folio that isn't uptodate, we treat + * it with zero copied and let generic_perform_write() try to copy + * data again through copied=3D0. */ if (!folio_test_uptodate(folio)) { - if (unlikely(copied !=3D len)) + if (unlikely(copied !=3D len)) { copied =3D 0; - else + } else if (folio_test_large(folio)) { + f2fs_ffs_mark_subrange_uptodate(folio, + offset_in_folio(folio, pos), len); + } else { + /* This should be come from len =3D=3D PAGE_SIZE */ folio_mark_uptodate(folio); + } } =20 #ifdef CONFIG_F2FS_FS_COMPRESSION @@ -4108,6 +4369,9 @@ static int f2fs_write_end(const struct kiocb *iocb, if (!copied) goto unlock_out; =20 + if (folio_test_large(folio)) + f2fs_ffs_mark_subrange_dirty(folio, offset_in_folio(folio, pos), + copied); folio_mark_dirty(folio); =20 if (f2fs_is_atomic_file(inode)) @@ -4170,8 +4434,22 @@ static bool f2fs_dirty_data_folio(struct address_spa= ce *mapping, =20 trace_f2fs_set_page_dirty(folio, DATA); =20 - if (!folio_test_uptodate(folio)) - folio_mark_uptodate(folio); + if (!folio_test_uptodate(folio)) { + bool uptodate =3D true; + + if (f2fs_folio_has_ffs(folio)) { + struct f2fs_folio_state *ffs =3D + (struct f2fs_folio_state *)folio->private; + unsigned long flags; + + spin_lock_irqsave(&ffs->state_lock, flags); + uptodate =3D bitmap_full(ffs->state, folio_nr_pages(folio)) && + !ffs->read_pages_pending; + spin_unlock_irqrestore(&ffs->state_lock, flags); + } + if (uptodate) + folio_mark_uptodate(folio); + } BUG_ON(folio_test_swapcache(folio)); =20 if (filemap_dirty_folio(mapping, folio)) { diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 326382ede1c3..8eccca6e5d27 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -4392,6 +4392,7 @@ int f2fs_write_single_data_page(struct folio *folio, = int *submitted, struct writeback_control *wbc, enum iostat_type io_type, int compr_blocks, bool allow_balance); +bool f2fs_ffs_test_blk_uptodate(const struct folio *folio, pgoff_t index); void f2fs_write_failed(struct inode *inode, loff_t to); void f2fs_invalidate_folio(struct folio *folio, size_t offset, size_t leng= th); bool f2fs_release_folio(struct folio *folio, gfp_t wait); diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index b0bdc7a977b9..9071bd23e57b 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -5412,9 +5412,20 @@ static int f2fs_preallocate_blocks(struct kiocb *ioc= b, struct iov_iter *iter, return ret; } =20 - /* Do not preallocate blocks that will be written partially in 4KB. */ - map.m_lblk =3D F2FS_BLK_ALIGN(sbi, pos); - map.m_len =3D F2FS_BYTES_TO_BLK(sbi, pos + count); + if (!dio && mapping_large_folio_support(inode->i_mapping)) { + /* + * Preallocate all blocks touched by a large-folio buffered write so + * the regular write_begin path does not need to unlock the folio for + * f2fs_balance_fs(). Rechecking large-folio state after unlock is + * unreliable since partial truncation can split the folio. + */ + map.m_lblk =3D F2FS_BYTES_TO_BLK(sbi, pos); + map.m_len =3D F2FS_BLK_ALIGN(sbi, pos + count); + } else { + /* Do not preallocate blocks that will be written partially in 4KB. */ + map.m_lblk =3D F2FS_BLK_ALIGN(sbi, pos); + map.m_len =3D F2FS_BYTES_TO_BLK(sbi, pos + count); + } if (map.m_len > map.m_lblk) map.m_len -=3D map.m_lblk; else --=20 2.43.0 From nobody Fri Sep 25 08:47:33 2026 Received: from m16.mail.126.com (m16.mail.126.com [117.135.210.8]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 94A571F91F6 for ; Tue, 15 Sep 2026 04:19:54 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=117.135.210.8 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1789445997; cv=none; b=EqeOUHjDEfKW43ePY96yitJ0gDzJ89JsIDdtil2vqqFUFZyB3WoCpYJgMjcII5VOpRELsvfS3wDvhDuwXQKUkauNFgeg9OtOFUIqHPsXOrNgsmjjczYzyPpIR2cibnNL4mTYiRW2qu3cHfPGwQQ5T+Pjwh1U0APRS3Gb6VSN/ZI= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1789445997; c=relaxed/simple; bh=HgSsJRoK0J05KG0mwTFWrglOY9661dw2QVjfdEDKlJ8=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=nfCnmUoSETjZAKa4boGzcB5YIbtvftva0J3oKOF8AFyFS7QBjzyJ7D4FFKXQ2s7l54ZrmDkOfOw6rbdIGgmZuflUxt3NonRRoGBVw+hXSwplDPJ5P20ilC5kkJD8boKtJz/en7PVyffGi0TYCfGyDu3Uty7S4QP6iymGGyScnRA= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=fail (p=quarantine dis=none) header.from=xiaomi.com; spf=pass smtp.mailfrom=126.com; dkim=pass (1024-bit key) header.d=126.com header.i=@126.com header.b=NSBWDaZ0; arc=none smtp.client-ip=117.135.210.8 Authentication-Results: smtp.subspace.kernel.org; dmarc=fail (p=quarantine dis=none) header.from=xiaomi.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=126.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=126.com header.i=@126.com header.b="NSBWDaZ0" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=126.com; s=s110527; h=From:To:Subject:Date:Message-ID:MIME-Version; bh=f6 7LGm08S9tN97M8CGib1Eur96A/KAiCF+w+dAL+Kxc=; b=NSBWDaZ0X++7JFcmOc /RB1DYTeczlRi5JLi/bYgvbbYUWuXOLiYRDZ5EzdV2TSz60dIThNidhgrhcMaEZZ qMBxKpD4wvUNQcUO28aYBHqwKT15efp1G5eRY+x3LNe7lU2xREBo9dGejG0ffXe3 CBnXoSfVDR+OPLxnhGTnrw1iA= Received: from nzzhao-ThinkCentre-M760t.mioffice.cn (unknown []) by gzsmtp3 (Coremail) with SMTP id PikvCgDXreo9x6hq3jBvHw--.36918S6; Tue, 15 Sep 2026 12:19:14 +0800 (CST) From: Nanzhe Zhao To: linux-f2fs-devel@lists.sourceforge.net, Jaegeuk Kim , Chao Yu Cc: Barry Song , Juan Yescas , Dev Jain , linux-kernel@vger.kernel.org, David Hildenbrand , Bo Zhang , Kalesh Singh , Nanzhe Zhao , Pengfei Li , Ryan Roberts , Nanzhe Zhao Subject: [PATCH v2 04/14] f2fs: support atomic file large folios buffered write Date: Tue, 15 Sep 2026 12:18:59 +0800 Message-ID: <20260915041909.2903887-5-zhaonanzhe@xiaomi.com> X-Mailer: git-send-email 2.43.0 In-Reply-To: <20260915041909.2903887-1-zhaonanzhe@xiaomi.com> References: <20260915041909.2903887-1-zhaonanzhe@xiaomi.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-CM-TRANSID: PikvCgDXreo9x6hq3jBvHw--.36918S6 X-Coremail-Antispam: 1Uf129KBjvJXoWxAr4UWF45Wr1DJF13tF4DXFb_yoWruw4UpF WUKwn5KrZ3J343uwnavF4DZr1F9a4xXw47ZayfG3WfZ3W5Jw13KF48t343ZF15trWkA3W8 XF40yFyrW3WUJrJanT9S1TB71UUUUU7qnTZGkaVYY2UrUUUUjbIjqfuFe4nvWSU5nxnvy2 9KBjDUYxBIdaVFxhVjvjDU0xZFpf9x07jUNVkUUUUU= Sender: nzzhao@126.com X-CM-SenderInfo: xq22xtbr6rjloofrz/xtbBowLGlmqox0JodAAA3q Content-Type: text/plain; charset="utf-8" ioctl can convert an inode with large folio support into an atomic file. Support large folio buffered writes for atomic files as well. Add a large folio atomic write_begin helper that reserves COW mappings for the write range. For partial head and tail subpages, read the existing data from either the COW inode or the original inode before marking the subpage uptodate. Signed-off-by: Nanzhe Zhao --- fs/f2fs/data.c | 88 ++++++++++++++++++++++++++++++++++++++++++++++---- 1 file changed, 81 insertions(+), 7 deletions(-) diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index a490df4f50b3..eb5fb8e52851 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -4016,18 +4016,16 @@ static int __reserve_data_block(struct inode *inode= , pgoff_t index, } =20 static int prepare_atomic_write_begin(struct f2fs_sb_info *sbi, - struct folio *folio, loff_t pos, unsigned int len, + struct inode *inode, pgoff_t index, block_t *blk_addr, bool *node_changed) { - struct inode *inode =3D folio->mapping->host; struct inode *cow_inode =3D F2FS_I(inode)->cow_inode; - pgoff_t index =3D folio->index; int err =3D 0; block_t ori_blk_addr =3D NULL_ADDR; bool cow_has_reserved_block =3D false; =20 /* If pos is beyond the end of file, reserve a new block in COW inode */ - if ((pos & PAGE_MASK) >=3D i_size_read(inode)) + if ((index << PAGE_SHIFT) >=3D i_size_read(inode)) goto reserve_block; =20 /* Look for the block in COW inode first */ @@ -4168,6 +4166,74 @@ static int prepare_large_folio_write_begin(struct in= ode *inode, return 0; } =20 +static int prepare_large_folio_atomic_write_begin(struct inode *inode, + struct address_space *mapping, struct folio *folio, loff_t pos, + unsigned int len) +{ + struct f2fs_sb_info *sbi =3D F2FS_I_SB(inode); + size_t ori_off =3D offset_in_folio(folio, pos); + pgoff_t start, end, index; + unsigned int orig_order; + int err =3D 0; + + len =3D min_t(unsigned int, len, folio_size(folio) - ori_off); + + if (!f2fs_ffs_find_or_alloc(folio)) + return -ENOMEM; + + /* Inline data must have been converted before reaching here. */ + f2fs_bug_on(sbi, f2fs_has_inline_data(inode)); + + start =3D folio->index + (ori_off >> PAGE_SHIFT); + end =3D folio->index + ((ori_off + len - 1) >> PAGE_SHIFT); + + for (index =3D start; index <=3D end; index++) { + block_t blkaddr =3D NULL_ADDR; + bool node_changed =3D false; + size_t off =3D (index - folio->index) << PAGE_SHIFT; + + err =3D prepare_atomic_write_begin(sbi, inode, index, + &blkaddr, &node_changed); + if (err) + return err; + + if (f2fs_ffs_test_blk_uptodate(folio, index)) + goto balance; + + if (blkaddr =3D=3D NEW_ADDR) { + folio_zero_segment(folio, off, off + PAGE_SIZE); + f2fs_ffs_mark_subrange_uptodate(folio, off, PAGE_SIZE); + goto balance; + } + + if (!f2fs_is_valid_blkaddr(sbi, blkaddr, + DATA_GENERIC_ENHANCE_READ)) + return -EFSCORRUPTED; + + err =3D f2fs_submit_page_read_sync(inode, folio, index, + blkaddr); + if (err) + return err; +balance: + /* + * Expand the 4K-page balance decision per subpage: check + * right after each preallocated block. + */ + if (node_changed && !IS_NOQUOTA(inode) && + has_not_enough_free_secs(sbi, 0, 0)) { + orig_order =3D folio_order(folio); + folio_unlock(folio); + f2fs_balance_fs(sbi, true); + folio_lock(folio); + if (unlikely(folio->mapping !=3D mapping || + folio_order(folio) !=3D orig_order)) + return -EAGAIN; + } + } + + return 0; +} + static int f2fs_write_begin(const struct kiocb *iocb, struct address_space *mapping, loff_t pos, unsigned len, struct folio **foliop, @@ -4245,8 +4311,8 @@ static int f2fs_write_begin(const struct kiocb *iocb, =20 *foliop =3D folio; =20 - if (f2fs_is_atomic_file(inode)) - err =3D prepare_atomic_write_begin(sbi, folio, pos, len, + if (f2fs_is_atomic_file(inode) && !folio_test_large(folio)) + err =3D prepare_atomic_write_begin(sbi, inode, folio->index, &blkaddr, &need_balance); else if (!folio_test_large(folio)) err =3D prepare_write_begin(sbi, folio, pos, len, @@ -4270,10 +4336,18 @@ static int f2fs_write_begin(const struct kiocb *ioc= b, f2fs_folio_wait_writeback(folio, DATA, false, true); =20 if (folio_test_large(folio)) { - err =3D prepare_large_folio_write_begin(inode, + if (f2fs_is_atomic_file(inode)) + err =3D prepare_large_folio_atomic_write_begin(inode, + mapping, folio, pos, len); + else + err =3D prepare_large_folio_write_begin(inode, folio, pos, len); if (!err) return 0; + if (err =3D=3D -EAGAIN) { + f2fs_folio_put(folio, true); + goto repeat; + } goto put_folio; } =20 --=20 2.43.0 From nobody Fri Sep 25 08:47:33 2026 Received: from m16.mail.126.com (m16.mail.126.com [220.197.31.7]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 6305F35CB9C for ; Tue, 15 Sep 2026 04:20:02 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=220.197.31.7 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1789446020; cv=none; b=uiCiJ8RoWrO0IgVFGEAlTym1PM0S7fomah3/YVhvmk09hu4Msb6xTaLzt9gW1crNe+etQbaXTX3JMtBAgX2ErN9J/FBah9W7MqdcFtzZnIpa846nBAKS5qApe8YcVhBPlm4MGMGYOkIczjJBsdC9po1MCZ1NXbhK/Bdl29IESWg= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1789446020; c=relaxed/simple; bh=D2wUCbeeaWW7Hs0ZL8wFquQQjUuBxytrDJ9oua+0HMY=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=klAKvq1R6gGbhH5QSp8aaqCOnLdjneWiICPCybdZeMFtMSv67a2QlVS5sqdNBfSZXeR0SkdY1jNIhpxtKZH6OJdMrFyfmXCHa7glkS3e8iA9yZLVWPJNT16FOU/K9v/6Wooi7/8qeeUiyd5G36Z6w2HDGtPAW0VtT1ME08B6ODA= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=fail (p=quarantine dis=none) header.from=xiaomi.com; spf=pass smtp.mailfrom=126.com; dkim=pass (1024-bit key) header.d=126.com header.i=@126.com header.b=RPApR6zg; arc=none smtp.client-ip=220.197.31.7 Authentication-Results: smtp.subspace.kernel.org; dmarc=fail (p=quarantine dis=none) header.from=xiaomi.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=126.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=126.com header.i=@126.com header.b="RPApR6zg" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=126.com; s=s110527; h=From:To:Subject:Date:Message-ID:MIME-Version; bh=op ZbxStcInXcR1MH7uQFdPwVyP25So4wQhAPJUC05v4=; b=RPApR6zgS6CG3YRMGw sBuyDYsqbqFpxbP30vRWa0uSHtp9Ps6WaiDp5uMTvaTpZdn4cEy6XgNj3e/k2aos kc9LX2KF2BMEcUgZh5IWz3wa3MvSfDu9x/TLpLU/CD8pFml31uBYVL64dV+kct4f 9WNO7/6z+K4rMk6rY7utNBlCs= Received: from nzzhao-ThinkCentre-M760t.mioffice.cn (unknown []) by gzsmtp3 (Coremail) with SMTP id PikvCgDXreo9x6hq3jBvHw--.36918S7; Tue, 15 Sep 2026 12:19:15 +0800 (CST) From: Nanzhe Zhao To: linux-f2fs-devel@lists.sourceforge.net, Jaegeuk Kim , Chao Yu Cc: Barry Song , Juan Yescas , Dev Jain , linux-kernel@vger.kernel.org, David Hildenbrand , Bo Zhang , Kalesh Singh , Nanzhe Zhao , Pengfei Li , Ryan Roberts , Nanzhe Zhao Subject: [PATCH v2 05/14] f2fs: support large folio writeback Date: Tue, 15 Sep 2026 12:19:00 +0800 Message-ID: <20260915041909.2903887-6-zhaonanzhe@xiaomi.com> X-Mailer: git-send-email 2.43.0 In-Reply-To: <20260915041909.2903887-1-zhaonanzhe@xiaomi.com> References: <20260915041909.2903887-1-zhaonanzhe@xiaomi.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-CM-TRANSID: PikvCgDXreo9x6hq3jBvHw--.36918S7 X-Coremail-Antispam: 1Uf129KBjvAXoW3KFW5WF13WFW8AFykWFWUtwb_yoW8GryxCo WfJw4qqr18Kr17ArWj9w12qFyUuan8AF95CF4fur4DZ3ZFq3s09ws7Kwn8X3W7Xr1Ykr1x uryxK3ZxJrWxAFn3n29KB7ZKAUJUUUU8529EdanIXcx71UUUUU7v73VFW2AGmfu7bjvjm3 AaLaJ3UbIYCTnIWIevJa73UjIFyTuYvjxU7FksDUUUU Sender: nzzhao@126.com X-CM-SenderInfo: xq22xtbr6rjloofrz/xtbBsAPGlmqox0MX6AAA3a Content-Type: text/plain; charset="utf-8" Large folio can contain multiple dirty ranges. Add a folio-based writeback path for large-folio mapping files and keep the legacy f2fs_write_cache_pages() path unchanged for non large-folio mapping files. Signed-off-by: Nanzhe Zhao --- fs/f2fs/data.c | 412 ++++++++++++++++++++++++++++++++++++++++++++++++- 1 file changed, 407 insertions(+), 5 deletions(-) diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index eb5fb8e52851..1f9c99689f38 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -355,7 +355,9 @@ static void f2fs_write_end_bio(struct bio *bio) =20 bio_for_each_folio_all(fi, bio) { struct folio *folio =3D fi.folio; + unsigned int nr_pages =3D fi.length >> PAGE_SHIFT; enum count_type type; + bool finished =3D true; =20 #ifdef CONFIG_F2FS_FS_COMPRESSION if (f2fs_is_compressed_page(folio)) { @@ -379,10 +381,19 @@ static void f2fs_write_end_bio(struct bio *bio) folio->index, NODE_TYPE_REGULAR, true); f2fs_bug_on(sbi, folio->index !=3D nid_of_node(sbi, folio)); } + if (f2fs_folio_has_ffs(folio)) { + struct f2fs_folio_state *ffs =3D + (struct f2fs_folio_state *)folio->private; + + finished =3D atomic_sub_and_test(nr_pages, + &ffs->write_pages_pending); + } + if (f2fs_in_warm_node_list(sbi, folio)) f2fs_del_fsync_node_entry(sbi, folio); =20 - dec_page_count(sbi, type); + while (nr_pages--) + dec_page_count(sbi, type); =20 /* * we should access sbi before folio_end_writeback() to @@ -392,8 +403,10 @@ static void f2fs_write_end_bio(struct bio *bio) wq_has_sleeper(&sbi->cp_wait)) wake_up(&sbi->cp_wait); =20 - folio_clear_f2fs_gcing(folio); - folio_end_writeback(folio); + if (finished) { + folio_clear_f2fs_gcing(folio); + folio_end_writeback(folio); + } } =20 bio_put(bio); @@ -2654,7 +2667,7 @@ static void f2fs_ffs_mark_subrange_uptodate(struct fo= lio *folio, size_t offset, folio_mark_uptodate(folio); } =20 -static void f2fs_ffs_mark_subrange_dirty(struct folio *folio, +void f2fs_ffs_mark_subrange_dirty(struct folio *folio, size_t offset, size_t len) { struct f2fs_folio_state *ffs; @@ -2677,6 +2690,99 @@ static void f2fs_ffs_mark_subrange_dirty(struct foli= o *folio, spin_unlock_irqrestore(&ffs->state_lock, flags); } =20 +static bool __ffs_clear_subrange_dirty(struct folio *folio, + struct f2fs_folio_state *ffs, size_t offset, size_t len) +{ + unsigned int nr_subpages =3D folio_nr_pages(folio); + unsigned int start, end; + + start =3D offset >> PAGE_SHIFT; + end =3D (offset + len + PAGE_SIZE - 1) >> PAGE_SHIFT; + end =3D min(end, nr_subpages); + + bitmap_clear(ffs->state, nr_subpages + start, end - start); + return find_next_bit(ffs->state, 2 * nr_subpages, nr_subpages) < + 2 * nr_subpages; +} + +/** + * f2fs_ffs_clear_subrange_dirty_and_test - clear dirty bits and test rema= ining + * @folio: the large folio + * @offset: start byte offset within the folio + * @len: byte length of the subrange + * + * Clear the dirty bits of the 4K subpages covered by [offset, offset + le= n), + * and return whether the folio still has any dirty subpage left. + */ +bool f2fs_ffs_clear_subrange_dirty_and_test(struct folio *folio, size_t of= fset, + size_t len) +{ + struct f2fs_folio_state *ffs; + unsigned long flags; + bool dirty; + + if (!f2fs_folio_has_ffs(folio)) + return false; + + ffs =3D (struct f2fs_folio_state *)folio->private; + spin_lock_irqsave(&ffs->state_lock, flags); + dirty =3D __ffs_clear_subrange_dirty(folio, ffs, offset, len); + spin_unlock_irqrestore(&ffs->state_lock, flags); + + return dirty; +} + +static unsigned int ffs_next_dirty_subpage(struct f2fs_folio_state *ffs, + const struct folio *folio, unsigned int start, + unsigned int end) +{ + unsigned int nr_subpages =3D folio_nr_pages(folio); + + return find_next_bit(ffs->state, nr_subpages + end + 1, + nr_subpages + start) - nr_subpages; +} + +static unsigned int ffs_next_clean_subpage(struct f2fs_folio_state *ffs, + const struct folio *folio, unsigned int start, + unsigned int end) +{ + unsigned int nr_subpages =3D folio_nr_pages(folio); + + return find_next_zero_bit(ffs->state, nr_subpages + end + 1, + nr_subpages + start) - nr_subpages; +} + +static unsigned int ffs_find_dirty_range(struct folio *folio, + u64 *range_start, u64 range_end) +{ + struct f2fs_folio_state *ffs; + unsigned int start, end, nr_pages; + + if (*range_start >=3D range_end) + return 0; + + if (!f2fs_folio_has_ffs(folio)) + return range_end - *range_start; + + ffs =3D (struct f2fs_folio_state *)folio->private; + start =3D offset_in_folio(folio, *range_start) >> PAGE_SHIFT; + end =3D DIV_ROUND_UP(min_not_zero(offset_in_folio(folio, range_end), + folio_size(folio)), PAGE_SIZE) - 1; + + start =3D ffs_next_dirty_subpage(ffs, folio, start, end); + if (start > end) + return 0; + + if (start =3D=3D end) + nr_pages =3D 1; + else + nr_pages =3D ffs_next_clean_subpage(ffs, folio, + start + 1, end) - start; + + *range_start =3D folio_pos(folio) + ((u64)start << PAGE_SHIFT); + return (u64)nr_pages << PAGE_SHIFT; +} + static bool find_next_valid_block(const struct folio *folio, size_t orig_off, size_t *need_off, size_t len) @@ -3256,6 +3362,133 @@ int f2fs_do_write_data_page(struct f2fs_io_info *fi= o) return err; } =20 +static int f2fs_write_single_data_folio(struct folio *folio, int *submitte= d, + struct writeback_control *wbc, + enum iostat_type io_type, + u64 *pos, u64 len) +{ + struct inode *inode =3D folio->mapping->host; + struct f2fs_sb_info *sbi =3D F2FS_I_SB(inode); + bool atomic_commit =3D f2fs_is_atomic_file(inode) && + folio_test_f2fs_atomic(folio); + struct inode *dn_inode =3D atomic_commit ? + F2FS_I(inode)->cow_inode : inode; + struct f2fs_folio_state *ffs =3D NULL; + u64 folio_start =3D folio_pos(folio); + pgoff_t start_idx =3D (*pos - folio_start) >> PAGE_SHIFT; + pgoff_t end_idx =3D (*pos - folio_start + len - 1) >> PAGE_SHIFT; + int err =3D 0; + + for (pgoff_t i =3D start_idx; i <=3D end_idx && !err; i++) { + struct dnode_of_data dn; + struct node_info ni; + pgoff_t data_idx =3D folio->index + i; + bool ipu_force =3D false; + bool dn_held =3D false; + bool submitted_blk =3D false; + struct f2fs_io_info fio =3D { + .sbi =3D sbi, + .ino =3D inode->i_ino, + .type =3D DATA, + .op =3D REQ_OP_WRITE, + .op_flags =3D wbc_to_write_flags(wbc), + .old_blkaddr =3D NULL_ADDR, + .folio =3D folio, + .folio_offset =3D i, + .folio_blkcnt =3D 1, + .encrypted_page =3D NULL, + .submitted =3D 0, + .need_lock =3D LOCK_DONE, + .meta_gc =3D f2fs_meta_inode_gc_required(inode) ? 1 : 0, + .io_type =3D io_type, + .io_wbc =3D wbc, + }; + + if (f2fs_folio_has_ffs(folio)) { + ffs =3D (struct f2fs_folio_state *)folio->private; + atomic_inc(&ffs->write_pages_pending); + } + + set_new_dnode(&dn, dn_inode, NULL, NULL, 0); + + if (need_inplace_update(&fio) && + f2fs_lookup_read_extent_cache_block(inode, data_idx, + &fio.old_blkaddr)) { + if (!f2fs_is_valid_blkaddr(sbi, fio.old_blkaddr, + DATA_GENERIC_ENHANCE)) { + err =3D -EFSCORRUPTED; + goto block_done; + } + ipu_force =3D true; + goto got_it; + } + + err =3D f2fs_get_dnode_of_data(&dn, data_idx, LOOKUP_NODE); + if (err) + goto block_done; + dn_held =3D true; + + fio.old_blkaddr =3D dn.data_blkaddr; + +got_it: + if (__is_valid_data_blkaddr(fio.old_blkaddr) && + !f2fs_is_valid_blkaddr(sbi, fio.old_blkaddr, + DATA_GENERIC_ENHANCE)) { + err =3D -EFSCORRUPTED; + goto block_done; + } + + if (fio.meta_gc) + f2fs_wait_on_block_writeback(inode, fio.old_blkaddr); + + if (ipu_force || + (__is_valid_data_blkaddr(fio.old_blkaddr) && + need_inplace_update(&fio))) { + f2fs_put_dnode(&dn); + err =3D f2fs_inplace_write_data(&fio); + trace_f2fs_do_write_data_page(folio, IPU); + if (err) + goto block_done; + + if (submitted) + (*submitted)++; + submitted_blk =3D true; + set_inode_flag(inode, FI_UPDATE_WRITE); + goto block_done; + } + + err =3D f2fs_get_node_info(sbi, dn.nid, &ni, false); + if (err) + goto block_done; + + fio.version =3D ni.version; + + f2fs_outplace_write_data(&dn, &fio); + if (submitted) + (*submitted)++; + submitted_blk =3D true; + set_inode_flag(inode, FI_APPEND_WRITE); + trace_f2fs_do_write_data_page(folio, OPU); + goto block_done; +block_done: + if (dn_held) + f2fs_put_dnode(&dn); + if (!submitted_blk && ffs) + atomic_dec(&ffs->write_pages_pending); + if (!err) { + *pos =3D (data_idx + 1) << PAGE_SHIFT; + spin_lock(&F2FS_I(inode)->i_size_lock); + if (F2FS_I(inode)->last_disk_size < *pos) + F2FS_I(inode)->last_disk_size =3D *pos; + spin_unlock(&F2FS_I(inode)->i_size_lock); + } else { + file_set_keep_isize(inode); + } + } + + return err; +} + int f2fs_write_single_data_page(struct folio *folio, int *submitted, struct bio **bio, sector_t *last_block, @@ -3705,6 +3938,172 @@ static int f2fs_write_cache_pages(struct address_sp= ace *mapping, return ret; } =20 +static int f2fs_write_cache_folios(struct address_space *mapping, + struct writeback_control *wbc, + enum iostat_type io_type) +{ + struct folio *folio =3D NULL; + struct inode *inode =3D mapping->host; + struct f2fs_sb_info *sbi =3D F2FS_I_SB(inode); + struct f2fs_lock_context lc; + u64 pos =3D 0; + u64 end_pos =3D 0; + u32 r_len =3D 0; + int err =3D 0; + int submitted =3D 0; + int nwritten =3D 0; + bool op_locked =3D false; + bool next =3D false; + bool retry =3D false; + + if (get_dirty_pages(inode) <=3D SM_I(sbi)->min_hot_blocks) + set_inode_flag(inode, FI_HOT_DATA); + else + clear_inode_flag(inode, FI_HOT_DATA); + + while ((folio =3D writeback_iter(mapping, wbc, folio, &err))) { + struct f2fs_folio_state *ffs =3D NULL; + u64 isize; + size_t poff; + pgoff_t end_index; + bool verity_in_progress; + int folio_submitted =3D 0; + bool bias_added =3D false; + + submitted =3D 0; + next =3D true; + retry =3D false; + + if (atomic_read(&sbi->wb_sync_req[DATA]) && + wbc->sync_mode =3D=3D WB_SYNC_NONE) { + folio_redirty_for_writepage(wbc, folio); + next =3D false; + goto retry_out; + } +retry: + pos =3D folio_pos(folio); + end_pos =3D pos + folio_size(folio); + isize =3D i_size_read(inode); + verity_in_progress =3D f2fs_verity_in_progress(inode); + poff =3D 0; + end_index =3D 0; + + if (retry) { + + if (unlikely(folio->mapping !=3D mapping)) + goto retry_out; + + if (!folio_test_dirty(folio)) + goto retry_out; + + if (folio_test_writeback(folio)) { + if (wbc->sync_mode =3D=3D WB_SYNC_NONE) + goto retry_out; + f2fs_folio_wait_writeback(folio, DATA, true, true); + } + + if (!folio_clear_dirty_for_io(folio)) + goto retry_out; + } + + /* To avoid dealing with the complexity for one subrange is in bio + * while we trylock_op failed before writing another subrange. + * Try to lock_op before any subrange write for the folio. + */ + if (!op_locked) { + if (!f2fs_trylock_op(sbi, &lc)) { + folio_redirty_for_writepage(wbc, folio); + err =3D 0; + if (wbc->sync_mode !=3D WB_SYNC_ALL) + goto retry_out; + + retry =3D true; + folio_unlock(folio); + f2fs_io_schedule_timeout(DEFAULT_SCHEDULE_TIMEOUT); + folio_lock(folio); + goto retry; + } + op_locked =3D true; + } + + if (!verity_in_progress) { + poff =3D offset_in_folio(folio, isize); + end_index =3D isize >> PAGE_SHIFT; + + if (folio->index > end_index || + (folio->index =3D=3D end_index && poff =3D=3D 0)) + goto out; + + if (end_pos > isize) { + folio_zero_segment(folio, poff, folio_size(folio)); + end_pos =3D isize; + } + } + + folio_start_writeback(folio); + + if (!bias_added && f2fs_folio_has_ffs(folio)) { + ffs =3D (struct f2fs_folio_state *)folio->private; + WARN_ON_ONCE(atomic_read(&ffs->write_pages_pending) !=3D 0); + atomic_inc(&ffs->write_pages_pending); + bias_added =3D true; + } + + while ((r_len =3D ffs_find_dirty_range(folio, &pos, end_pos))) { + err =3D f2fs_write_single_data_folio(folio, &submitted, + wbc, io_type, &pos, r_len); + folio_submitted +=3D submitted; + if (err) + goto out; + + nwritten +=3D submitted; + } + + if (!err && folio_submitted && + f2fs_is_atomic_file(inode) && + folio_test_f2fs_atomic(folio)) + folio_clear_f2fs_atomic(folio); + +out: + /* + * Clear the dirty state of the subranges that were written back. + * If some subranges are still dirty (e.g. the folio was only + * partially written back), the folio must stay dirty so that + * the remaining subranges are written back later. + */ + if (f2fs_ffs_clear_subrange_dirty_and_test(folio, 0, + (err ? pos : end_pos) - folio_pos(folio))) + folio_redirty_for_writepage(wbc, folio); + else + inode_dec_dirty_pages(inode); + + if (bias_added) { + if (atomic_dec_and_test(&ffs->write_pages_pending)) + folio_end_writeback(folio); + } else if (!folio_submitted && folio_test_writeback(folio)) { + folio_end_writeback(folio); + } + +retry_out: + if (folio_test_locked(folio)) + folio_unlock(folio); + + if (op_locked) { + f2fs_unlock_op(sbi, &lc); + op_locked =3D false; + } + + if (err || !next) + break; + } + + if (nwritten) + f2fs_submit_merged_write_cond(F2FS_M_SB(mapping), mapping->host, + NULL, 0, DATA); + + return err; +} + static inline bool __should_serialize_io(struct inode *inode, struct writeback_control *wbc) { @@ -3799,7 +4198,10 @@ static int __f2fs_write_data_pages(struct address_sp= ace *mapping, account_writeback(inode, true); =20 blk_start_plug(&plug); - ret =3D f2fs_write_cache_pages(mapping, wbc, io_type); + if (mapping_large_folio_support(inode->i_mapping)) + ret =3D f2fs_write_cache_folios(mapping, wbc, io_type); + else + ret =3D f2fs_write_cache_pages(mapping, wbc, io_type); blk_finish_plug(&plug); =20 account_writeback(inode, false); --=20 2.43.0 From nobody Fri Sep 25 08:47:33 2026 Received: from m16.mail.126.com (m16.mail.126.com [117.135.210.7]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 9A035207A20 for ; Tue, 15 Sep 2026 04:19:54 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=117.135.210.7 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1789445997; cv=none; b=QxVyadk3vMoFteHukl/8fDffusIfcHqM5yAwG74U8TZH5gZ4bCz8dtAtEYnRncl21FX7myYlHsBQHLqpufXrGVskitFdO1SsNOzlEBrKgInYPTZlK6QYgxHd/58pVGdYZaCxHX0zJsdqTZnCQbdivUGJc9pZg2Zk4p0AYox21bk= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1789445997; c=relaxed/simple; bh=drA0gLUrUhgaz3/M7DhSYa0nPzXjcYsb/v12x18kVVg=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=dGtXsHpLDldy1nHyjoD162vFToYUnEJ5Pa4+0DVQj6tLlWa8/+gGmgjrdULzzlrnozeb5dwJ9kBKa7bDp+UnBYO7nWVThEEgTd9G6Ws5cxFgcWlXtBFehHMnDbilpTOaq0ZDQXnwDd7rmswm71j2Oox2Q/0tust+vqzX/7LU39Y= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=fail (p=quarantine dis=none) header.from=xiaomi.com; spf=pass smtp.mailfrom=126.com; dkim=pass (1024-bit key) header.d=126.com header.i=@126.com header.b=JOjYtbEt; arc=none smtp.client-ip=117.135.210.7 Authentication-Results: smtp.subspace.kernel.org; dmarc=fail (p=quarantine dis=none) header.from=xiaomi.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=126.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=126.com header.i=@126.com header.b="JOjYtbEt" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=126.com; s=s110527; h=From:To:Subject:Date:Message-ID:MIME-Version; bh=La K4nxrqbm6CTvTeDY9OaJp/SPb9ZvkcC8YMYFupLbI=; b=JOjYtbEtlusC7aG23t UCW6BlsKDENxEByxirdlCAc0BFRzYt8xf5mcEFb9LtIEpO2UJ01LZrgo8DYWEp9R Uak6gxAVWeyv1w6umYzz4sCXaiyCpKj4Oh+n+TZ/7GJ5ZQ2PQyqNCiHVBG8ASsph vQAtEr9LzNPjWjLITxy3PgIU4= Received: from nzzhao-ThinkCentre-M760t.mioffice.cn (unknown []) by gzsmtp3 (Coremail) with SMTP id PikvCgDXreo9x6hq3jBvHw--.36918S8; Tue, 15 Sep 2026 12:19:16 +0800 (CST) From: Nanzhe Zhao To: linux-f2fs-devel@lists.sourceforge.net, Jaegeuk Kim , Chao Yu Cc: Barry Song , Juan Yescas , Dev Jain , linux-kernel@vger.kernel.org, David Hildenbrand , Bo Zhang , Kalesh Singh , Nanzhe Zhao , Pengfei Li , Ryan Roberts , Nanzhe Zhao Subject: [PATCH v2 06/14] f2fs: prepare mmap write faults for large folios Date: Tue, 15 Sep 2026 12:19:01 +0800 Message-ID: <20260915041909.2903887-7-zhaonanzhe@xiaomi.com> X-Mailer: git-send-email 2.43.0 In-Reply-To: <20260915041909.2903887-1-zhaonanzhe@xiaomi.com> References: <20260915041909.2903887-1-zhaonanzhe@xiaomi.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-CM-TRANSID: PikvCgDXreo9x6hq3jBvHw--.36918S8 X-Coremail-Antispam: 1Uf129KBjvJXoW3Wr4kZF4kJr1rtw4fArW3GFg_yoWxWFy5pF W5G3s3JrZ3XFyUWrn7KF1qvw1rK34xurWUuFZ3G3WxZ3Wjqwn5KF10y340vF18trZ5Aan7 XF4jkFyUua4UGF7anT9S1TB71UUUUU7qnTZGkaVYY2UrUUUUjbIjqfuFe4nvWSU5nxnvy2 9KBjDUYxBIdaVFxhVjvjDU0xZFpf9x07jUManUUUUU= Sender: nzzhao@126.com X-CM-SenderInfo: xq22xtbr6rjloofrz/xtbBsQTHl2qox0QWcQAA3D Content-Type: text/plain; charset="utf-8" Now write protect `mmap` also need to support large folio, Change `f2fs_vm_page_mkwrite` to acheive that. Note it currently marks the whole large folio dirty to avoid data loss which causes write amplification. Further optimization is welcome. PG_mappedtodisk is useless in f2fs, so drop the folio_test_mappedtodisk() check and its goto out_sem shortcut in f2fs_vm_page_mkwrite(). We extend the folio_zero_segment() in mkwrite to zero the post-EOF part of the faulted folio for both order-0 and large folios, so the f2fs_zero_post_eof_page() call added to cover that shortcut is no longer needed. Signed-off-by: Nanzhe Zhao --- fs/f2fs/f2fs.h | 4 ++ fs/f2fs/file.c | 107 ++++++++++++++++++++++++++++++------------------- 2 files changed, 70 insertions(+), 41 deletions(-) diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 8eccca6e5d27..f48e2627d90a 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -4393,6 +4393,10 @@ int f2fs_write_single_data_page(struct folio *folio,= int *submitted, enum iostat_type io_type, int compr_blocks, bool allow_balance); bool f2fs_ffs_test_blk_uptodate(const struct folio *folio, pgoff_t index); +struct f2fs_folio_state *f2fs_ffs_find_or_alloc(struct folio *folio); +void f2fs_ffs_mark_subrange_dirty(struct folio *folio, size_t offset, size= _t len); +bool f2fs_ffs_clear_subrange_dirty_and_test(struct folio *folio, size_t of= fset, + size_t len); void f2fs_write_failed(struct inode *inode, loff_t to); void f2fs_invalidate_folio(struct folio *folio, size_t offset, size_t leng= th); bool f2fs_release_folio(struct folio *folio, gfp_t wait); diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index 9071bd23e57b..738a751c3903 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -118,13 +118,47 @@ static vm_fault_t f2fs_filemap_fault(struct vm_fault = *vmf) return ret; } =20 +static int f2fs_get_block_mkwrite(struct inode *inode, struct folio *folio, + pgoff_t index, bool need_alloc) +{ + struct dnode_of_data dn; + int err; + + set_new_dnode(&dn, inode, NULL, NULL, 0); + if (need_alloc) { + err =3D f2fs_get_block_locked(&dn, index); + } else { + err =3D f2fs_get_dnode_of_data(&dn, index, LOOKUP_NODE); + f2fs_put_dnode(&dn); + if (f2fs_is_pinned_file(inode) && + !__is_valid_data_blkaddr(dn.data_blkaddr)) + err =3D -EIO; + } + + if (err) + return err; + + f2fs_folio_wait_writeback(folio, DATA, false, true); + + /* wait for GCed page writeback via META_MAPPING */ + f2fs_wait_on_block_writeback(inode, dn.data_blkaddr); + + return 0; +} + static vm_fault_t f2fs_vm_page_mkwrite(struct vm_fault *vmf) { struct folio *folio =3D page_folio(vmf->page); struct inode *inode =3D file_inode(vmf->vma->vm_file); struct f2fs_sb_info *sbi =3D F2FS_I_SB(inode); - struct dnode_of_data dn; bool need_alloc =3D !f2fs_is_pinned_file(inode); + pgoff_t pidx =3D folio->index + folio_page_idx(folio, vmf->page); + loff_t pos =3D (loff_t)pidx << PAGE_SHIFT; + loff_t isize; + loff_t folio_start; + loff_t valid_end; + size_t dirty_len; + size_t subpage_off; int err =3D 0; vm_fault_t ret; =20 @@ -161,7 +195,7 @@ static vm_fault_t f2fs_vm_page_mkwrite(struct vm_fault = *vmf) =20 #ifdef CONFIG_F2FS_FS_COMPRESSION if (f2fs_compressed_file(inode)) { - int ret =3D f2fs_is_compressed_cluster(inode, folio->index); + int ret =3D f2fs_is_compressed_cluster(inode, pidx); =20 if (ret < 0) { err =3D ret; @@ -179,72 +213,63 @@ static vm_fault_t f2fs_vm_page_mkwrite(struct vm_faul= t *vmf) =20 f2fs_bug_on(sbi, f2fs_has_inline_data(inode)); =20 - err =3D f2fs_zero_post_eof_page(inode, - (folio->index + 1) << PAGE_SHIFT, true, false); - if (err) - goto out_pagefault; - file_update_time(vmf->vma->vm_file); filemap_invalidate_lock_shared(inode->i_mapping); =20 folio_lock(folio); + isize =3D i_size_read(inode); + folio_start =3D folio_pos(folio); + subpage_off =3D offset_in_folio(folio, pos); + valid_end =3D min_t(loff_t, folio_start + folio_size(folio), isize); + dirty_len =3D valid_end > folio_start ? valid_end - folio_start : 0; + if (unlikely(folio->mapping !=3D inode->i_mapping || - folio_pos(folio) > i_size_read(inode) || - !folio_test_uptodate(folio))) { + pos >=3D isize || + !f2fs_ffs_test_blk_uptodate(folio, + folio->index + (subpage_off >> PAGE_SHIFT)))) { folio_unlock(folio); err =3D -EFAULT; goto out_sem; } =20 - set_new_dnode(&dn, inode, NULL, NULL, 0); - if (need_alloc) { - /* block allocation */ - err =3D f2fs_get_block_locked(&dn, folio->index); - } else { - err =3D f2fs_get_dnode_of_data(&dn, folio->index, LOOKUP_NODE); - f2fs_put_dnode(&dn); - if (f2fs_is_pinned_file(inode) && - !__is_valid_data_blkaddr(dn.data_blkaddr)) - err =3D -EIO; - } + pgoff_t i, nr =3D DIV_ROUND_UP(dirty_len, PAGE_SIZE); =20 - if (err) { - folio_unlock(folio); - goto out_sem; + for (i =3D 0; i < nr; i++) { + err =3D f2fs_get_block_mkwrite(inode, folio, folio->index + i, + need_alloc); + if (err) { + folio_unlock(folio); + goto out_sem; + } } =20 - f2fs_folio_wait_writeback(folio, DATA, false, true); - - /* wait for GCed page writeback via META_MAPPING */ - f2fs_wait_on_block_writeback(inode, dn.data_blkaddr); - - /* - * check to see if the page is mapped already (no holes) - */ - if (folio_test_mappedtodisk(folio)) - goto out_sem; - /* page is wholly or partially inside EOF */ - if (((loff_t)(folio->index + 1) << PAGE_SHIFT) > - i_size_read(inode)) { - loff_t offset; + if (folio_start + folio_size(folio) > isize) { + size_t offset =3D offset_in_folio(folio, isize); =20 - offset =3D i_size_read(inode) & ~PAGE_MASK; folio_zero_segment(folio, offset, folio_size(folio)); } + + if (folio_test_large(folio)) { + if (!f2fs_ffs_find_or_alloc(folio)) { + folio_unlock(folio); + err =3D -ENOMEM; + goto out_sem; + } + f2fs_ffs_mark_subrange_dirty(folio, 0, dirty_len); + } folio_mark_dirty(folio); =20 - f2fs_update_iostat(sbi, inode, APP_MAPPED_IO, F2FS_BLKSIZE(sbi)); + f2fs_update_iostat(sbi, inode, APP_MAPPED_IO, dirty_len); f2fs_update_time(sbi, REQ_TIME); =20 out_sem: filemap_invalidate_unlock_shared(inode->i_mapping); -out_pagefault: sb_end_pagefault(inode->i_sb); out: ret =3D vmf_fs_error(err); =20 - trace_f2fs_vm_page_mkwrite(inode, folio->index, vmf->vma->vm_flags, ret); + trace_f2fs_vm_page_mkwrite(inode, pidx, vmf->vma->vm_flags, ret); return ret; } =20 --=20 2.43.0 From nobody Fri Sep 25 08:47:33 2026 Received: from m16.mail.126.com (m16.mail.126.com [220.197.31.9]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id C9B1A2E2665 for ; Tue, 15 Sep 2026 04:19:54 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=220.197.31.9 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1789445998; cv=none; b=IXgM6JpC9udG05v02aBKPxlr6gyNMl+TVSmJSFqJ9ZWVeccHucmcLL2WGZwbRJ2umwk9/e8qw45j3QlxrVE25ZomFfPZXL01nL/wEA+/pPAuClR5b9is0JA13+j1TZwMQMmR0kA39jLytmFk7lLqDhVmQWgYy3IHP8H+Wu6AO0o= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1789445998; c=relaxed/simple; bh=6Ce73nvdYYa65NgUDIFSu6Y4p67i74X+bPvVHyDmCuo=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=iLVmilgiuGOEqoh5axyMzNz9CBYi8z6R0oFNe92ECYc6Vvm1pjYzEtkpBMvpg2pobWrlIeBfh3BEe5n0wmo4ZDL0y9RhnIlFkZom9mzIS468MHY+q4K2t+FWCK5DkpGm03XdqbLimcjCGMHA45KEhaREg//zSkCFBa6C7nHb390= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=fail (p=quarantine dis=none) header.from=xiaomi.com; spf=pass smtp.mailfrom=126.com; dkim=pass (1024-bit key) header.d=126.com header.i=@126.com header.b=nDItdpZ2; arc=none smtp.client-ip=220.197.31.9 Authentication-Results: smtp.subspace.kernel.org; dmarc=fail (p=quarantine dis=none) header.from=xiaomi.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=126.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=126.com header.i=@126.com header.b="nDItdpZ2" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=126.com; s=s110527; h=From:To:Subject:Date:Message-ID:MIME-Version; bh=De OCd2h208uK1PYoGe91k/+tE0iG3xiMqrnXEBxLGoA=; b=nDItdpZ2gY3tkibnzL 8VOcQRG0rALZmtHCyTfD1Mn5/3dYCFD/3hEoP/GL27ekdMG8Ax3oH/qwedJ4Puwh kaETytQykHHN+NgjbBjjxbIGZ5BJnAq5XD7VbBG6Lbhm5B92NcBTJ3P1xenHaNqk kbL630GB+BGYjH2egyQDU/a4A= Received: from nzzhao-ThinkCentre-M760t.mioffice.cn (unknown []) by gzsmtp3 (Coremail) with SMTP id PikvCgDXreo9x6hq3jBvHw--.36918S9; Tue, 15 Sep 2026 12:19:16 +0800 (CST) From: Nanzhe Zhao To: linux-f2fs-devel@lists.sourceforge.net, Jaegeuk Kim , Chao Yu Cc: Barry Song , Juan Yescas , Dev Jain , linux-kernel@vger.kernel.org, David Hildenbrand , Bo Zhang , Kalesh Singh , Nanzhe Zhao , Pengfei Li , Ryan Roberts , Nanzhe Zhao Subject: [PATCH v2 07/14] f2fs: make GC migration large-folio aware Date: Tue, 15 Sep 2026 12:19:02 +0800 Message-ID: <20260915041909.2903887-8-zhaonanzhe@xiaomi.com> X-Mailer: git-send-email 2.43.0 In-Reply-To: <20260915041909.2903887-1-zhaonanzhe@xiaomi.com> References: <20260915041909.2903887-1-zhaonanzhe@xiaomi.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-CM-TRANSID: PikvCgDXreo9x6hq3jBvHw--.36918S9 X-Coremail-Antispam: 1Uf129KBjvJXoW3Kr4DCrWUAry3XF4UJF1xAFb_yoWDuw4xpF WUGF98KrWkJF1xWrn7tFn8Zr1Fk34fXr4UuFWxG34ft3WYqwn5KF18t34YyF1ftrW8A3Z2 qF4jka4UWF4UJF7anT9S1TB71UUUUU7qnTZGkaVYY2UrUUUUjbIjqfuFe4nvWSU5nxnvy2 9KBjDUYxBIdaVFxhVjvjDU0xZFpf9x07UShFsUUUUU= Sender: nzzhao@126.com X-CM-SenderInfo: xq22xtbr6rjloofrz/xtbBogXHl2qox0UMQwAA34 Content-Type: text/plain; charset="utf-8" GC can operate on a 4K block that is cached inside a large folio. The data lookup helpers therefore need to test and update uptodate state for the addressed subpage instead of rejecting large folios or treating the whole folio as the target block. Let f2fs_get_read_data_folio(), f2fs_find_data_folio(), and f2fs_get_lock_data_folio() to use subpage uptodate state. Submit single-block reads at the requested folio offset and zero only the addressed 4K range for NEW_ADDR. Also update `move_data_page` to mark, clear, and restore dirty state for the target subpage, and submit write I/O with the subpage offset recorded in f2fs_io_info. Signed-off-by: Nanzhe Zhao --- fs/f2fs/data.c | 75 ++++++++++++++++++++++++++++++----------------- fs/f2fs/f2fs.h | 1 + fs/f2fs/gc.c | 30 +++++++++++++++++-- fs/f2fs/segment.c | 2 +- 4 files changed, 78 insertions(+), 30 deletions(-) diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index 1f9c99689f38..287d83debf95 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -1226,19 +1226,26 @@ static struct bio *f2fs_grab_read_bio(struct inode = *inode, =20 /* This can handle encryption stuffs */ static void f2fs_submit_page_read(struct inode *inode, struct fsverity_inf= o *vi, - struct folio *folio, block_t blkaddr, - blk_opf_t op_flags, bool for_write) + struct folio *folio, pgoff_t index, + block_t blkaddr, blk_opf_t op_flags, + bool for_write) { struct f2fs_sb_info *sbi =3D F2FS_I_SB(inode); struct bio *bio; + size_t offset =3D 0; =20 - bio =3D f2fs_grab_read_bio(inode, vi, blkaddr, 1, op_flags, folio->index, + if (f2fs_folio_has_ffs(folio)) { + offset =3D offset_in_folio(folio, (loff_t)index << PAGE_SHIFT); + f2fs_update_read_folio_pending(folio, 1); + } + + bio =3D f2fs_grab_read_bio(inode, vi, blkaddr, 1, op_flags, index, for_write); =20 /* wait for GCed page writeback via META_MAPPING */ f2fs_wait_on_block_writeback(inode, blkaddr); =20 - if (!bio_add_folio(bio, folio, PAGE_SIZE, 0)) + if (!bio_add_folio(bio, folio, PAGE_SIZE, offset)) f2fs_bug_on(sbi, 1); =20 inc_page_count(sbi, F2FS_RD_DATA); @@ -1379,20 +1386,16 @@ struct folio *f2fs_get_read_data_folio(struct inode= *inode, pgoff_t index, struct dnode_of_data dn; struct folio *folio; int err; -retry: + folio =3D f2fs_grab_cache_folio(mapping, index, for_write); if (IS_ERR(folio)) return folio; =20 if (folio_test_large(folio)) { - pgoff_t folio_index =3D mapping_align_index(mapping, index); - unsigned long nr_pages =3D folio_nr_pages(folio); - - f2fs_folio_put(folio, true); - invalidate_inode_pages2_range(mapping, folio_index, - folio_index + nr_pages - 1); - f2fs_schedule_timeout(DEFAULT_SCHEDULE_TIMEOUT); - goto retry; + if (!f2fs_ffs_find_or_alloc(folio)) { + err =3D -ENOMEM; + goto put_err; + } } =20 if (f2fs_lookup_read_extent_cache_block(inode, index, @@ -1428,7 +1431,7 @@ struct folio *f2fs_get_read_data_folio(struct inode *= inode, pgoff_t index, goto put_err; } got_it: - if (folio_test_uptodate(folio)) { + if (f2fs_ffs_test_blk_uptodate(folio, index)) { folio_unlock(folio); return folio; } @@ -1441,15 +1444,17 @@ struct folio *f2fs_get_read_data_folio(struct inode= *inode, pgoff_t index, * f2fs_init_inode_metadata. */ if (dn.data_blkaddr =3D=3D NEW_ADDR) { - folio_zero_segment(folio, 0, folio_size(folio)); - if (!folio_test_uptodate(folio)) - folio_mark_uptodate(folio); + size_t offset =3D offset_in_folio(folio, + (loff_t)index << PAGE_SHIFT); + + folio_zero_segment(folio, offset, offset + PAGE_SIZE); + f2fs_ffs_mark_subrange_uptodate(folio, offset, PAGE_SIZE); folio_unlock(folio); return folio; } =20 - f2fs_submit_page_read(inode, f2fs_need_verity(inode, folio->index), - folio, dn.data_blkaddr, op_flags, for_write); + f2fs_submit_page_read(inode, f2fs_need_verity(inode, index), + folio, index, dn.data_blkaddr, op_flags, for_write); return folio; =20 put_err: @@ -1466,7 +1471,7 @@ struct folio *f2fs_find_data_folio(struct inode *inod= e, pgoff_t index, folio =3D f2fs_filemap_get_folio(mapping, index, FGP_ACCESSED, 0); if (IS_ERR(folio)) goto read; - if (folio_test_uptodate(folio)) + if (f2fs_ffs_test_blk_uptodate(folio, index)) return folio; f2fs_folio_put(folio, false); =20 @@ -1475,11 +1480,11 @@ struct folio *f2fs_find_data_folio(struct inode *in= ode, pgoff_t index, if (IS_ERR(folio)) return folio; =20 - if (folio_test_uptodate(folio)) + if (f2fs_ffs_test_blk_uptodate(folio, index)) return folio; =20 folio_wait_locked(folio); - if (unlikely(!folio_test_uptodate(folio))) { + if (unlikely(!f2fs_ffs_test_blk_uptodate(folio, index))) { f2fs_folio_put(folio, false); return ERR_PTR(-EIO); } @@ -1503,7 +1508,8 @@ struct folio *f2fs_get_lock_data_folio(struct inode *= inode, pgoff_t index, =20 /* wait for read completion */ folio_lock(folio); - if (unlikely(folio->mapping !=3D mapping || !folio_test_uptodate(folio)))= { + if (unlikely(folio->mapping !=3D mapping || + !f2fs_ffs_test_blk_uptodate(folio, index))) { f2fs_folio_put(folio, true); return ERR_PTR(-EIO); } @@ -2667,6 +2673,22 @@ static void f2fs_ffs_mark_subrange_uptodate(struct f= olio *folio, size_t offset, folio_mark_uptodate(folio); } =20 +bool f2fs_ffs_test_blk_dirty(const struct folio *folio, pgoff_t index) +{ + struct f2fs_folio_state *ffs; + unsigned int idx, nr_subpages; + + f2fs_bug_on(F2FS_F_SB(folio), !folio_contains(folio, index)); + + if (!f2fs_folio_has_ffs(folio)) + return folio_test_dirty(folio); + + ffs =3D folio->private; + nr_subpages =3D folio_nr_pages(folio); + idx =3D index - folio->index; + return test_bit(nr_subpages + idx, ffs->state); +} + void f2fs_ffs_mark_subrange_dirty(struct folio *folio, size_t offset, size_t len) { @@ -2724,14 +2746,13 @@ bool f2fs_ffs_clear_subrange_dirty_and_test(struct = folio *folio, size_t offset, if (!f2fs_folio_has_ffs(folio)) return false; =20 - ffs =3D (struct f2fs_folio_state *)folio->private; + ffs =3D folio->private; spin_lock_irqsave(&ffs->state_lock, flags); dirty =3D __ffs_clear_subrange_dirty(folio, ffs, offset, len); spin_unlock_irqrestore(&ffs->state_lock, flags); =20 return dirty; } - static unsigned int ffs_next_dirty_subpage(struct f2fs_folio_state *ffs, const struct folio *folio, unsigned int start, unsigned int end) @@ -3225,7 +3246,7 @@ bool f2fs_should_update_outplace(struct inode *inode,= struct f2fs_io_info *fio) return true; =20 if (fio) { - if (page_private_gcing(fio->page)) + if (folio_test_f2fs_gcing(fio->folio)) return true; if (unlikely(is_sbi_flag_set(sbi, SBI_CP_DISABLED) && f2fs_is_checkpointed_data(sbi, fio->old_blkaddr))) @@ -4781,7 +4802,7 @@ static int f2fs_write_begin(const struct kiocb *iocb, */ f2fs_submit_page_read(inode, NULL, /* can't write to fsverity files */ - folio, blkaddr, 0, true); + folio, index, blkaddr, 0, true); =20 folio_lock(folio); if (unlikely(folio->mapping !=3D mapping)) { diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index f48e2627d90a..6be08e45341c 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -4394,6 +4394,7 @@ int f2fs_write_single_data_page(struct folio *folio, = int *submitted, int compr_blocks, bool allow_balance); bool f2fs_ffs_test_blk_uptodate(const struct folio *folio, pgoff_t index); struct f2fs_folio_state *f2fs_ffs_find_or_alloc(struct folio *folio); +bool f2fs_ffs_test_blk_dirty(const struct folio *folio, pgoff_t index); void f2fs_ffs_mark_subrange_dirty(struct folio *folio, size_t offset, size= _t len); bool f2fs_ffs_clear_subrange_dirty_and_test(struct folio *folio, size_t of= fset, size_t len); diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c index 16fce62740d7..147836f1155f 100644 --- a/fs/f2fs/gc.c +++ b/fs/f2fs/gc.c @@ -1509,12 +1509,19 @@ static int move_data_page(struct inode *inode, bloc= k_t bidx, int gc_type, unsigned int segno, int off) { struct folio *folio; + size_t foff =3D 0; + bool large =3D false; int err =3D 0; =20 folio =3D f2fs_get_lock_data_folio(inode, bidx, true); if (IS_ERR(folio)) return PTR_ERR(folio); =20 + if (f2fs_folio_has_ffs(folio)) { + large =3D true; + foff =3D offset_in_folio(folio, (loff_t)bidx << PAGE_SHIFT); + } + if (!check_valid_map(F2FS_I_SB(inode), segno, off)) { err =3D -ENOENT; goto out; @@ -1529,6 +1536,8 @@ static int move_data_page(struct inode *inode, block_= t bidx, int gc_type, err =3D -EAGAIN; goto out; } + if (large) + f2fs_ffs_mark_subrange_dirty(folio, foff, PAGE_SIZE); folio_mark_dirty(folio); folio_set_f2fs_gcing(folio); } else { @@ -1541,32 +1550,49 @@ static int move_data_page(struct inode *inode, bloc= k_t bidx, int gc_type, .op_flags =3D REQ_SYNC, .old_blkaddr =3D NULL_ADDR, .folio =3D folio, + .folio_offset =3D bidx - folio->index, + .folio_blkcnt =3D 1, .encrypted_page =3D NULL, .need_lock =3D LOCK_REQ, .io_type =3D FS_GC_DATA_IO, }; - bool is_dirty =3D folio_test_dirty(folio); + struct f2fs_folio_state *ffs =3D NULL; + bool is_dirty =3D f2fs_ffs_test_blk_dirty(folio, bidx); =20 retry: f2fs_folio_wait_writeback(folio, DATA, true, true); =20 + if (large) { + ffs =3D folio->private; + f2fs_ffs_mark_subrange_dirty(folio, foff, PAGE_SIZE); + } folio_mark_dirty(folio); if (folio_clear_dirty_for_io(folio)) { inode_dec_dirty_pages(inode); f2fs_remove_dirty_inode(inode); + if (large && + f2fs_ffs_clear_subrange_dirty_and_test(folio, foff, PAGE_SIZE)) + folio_mark_dirty(folio); } =20 + if (large) + atomic_inc(&ffs->write_pages_pending); folio_set_f2fs_gcing(folio); =20 err =3D f2fs_do_write_data_page(&fio); if (err) { folio_clear_f2fs_gcing(folio); + if (large) + atomic_dec(&ffs->write_pages_pending); if (err =3D=3D -ENOMEM) { memalloc_retry_wait(GFP_NOFS); goto retry; } - if (is_dirty) + if (is_dirty) { + if (large) + f2fs_ffs_mark_subrange_dirty(folio, foff, PAGE_SIZE); folio_mark_dirty(folio); + } } } out: diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c index 644ddfb984c9..a81cc8c5c34a 100644 --- a/fs/f2fs/segment.c +++ b/fs/f2fs/segment.c @@ -3860,7 +3860,7 @@ static int __get_segment_type_6(struct f2fs_io_info *= fio) if (is_inode_flag_set(inode, FI_ALIGNED_WRITE)) return CURSEG_COLD_DATA_PINNED; =20 - if (page_private_gcing(fio->page)) { + if (folio_test_f2fs_gcing(fio->folio)) { if (fio->sbi->am.atgc_enabled && (fio->io_type =3D=3D FS_DATA_IO) && (fio->sbi->gc_mode !=3D GC_URGENT_HIGH) && --=20 2.43.0 From nobody Fri Sep 25 08:47:33 2026 Received: from m16.mail.126.com (m16.mail.126.com [117.135.210.9]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id C9B943EEAF8 for ; Tue, 15 Sep 2026 04:20:01 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=117.135.210.9 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1789446007; cv=none; b=HPhy6lRJMuLWOsCV3imhQwWUOSX4d8pQ8EOV/bVEnlxuwyUtvWvYXfcQHNQNSNRCoPzQ9ChhrwNi+9lgsyQt0KLmIg/EjMklg0M3k6yYC9IBXflMXKdMIq2W6BUoFqsabqSlzaq6vAKi+ZAi5CVAZOTpdgGB0r1fFsrr+PeYZCI= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1789446007; c=relaxed/simple; bh=S9F8vTTC1klBp+6TZHVJOCvZubzyU7a0EA3aGnGKR0g=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=uvGI8pJXB/gkw7owzZmpNJrpIWCUYYhLD8XQ59l8puW3ZkRb6DXBLWonf7NfKY3v1PuW0dBXitsXaRL30b9nShICht/D6uy0C12aQGiBkQrYtp+60ckiqPBdTJF4fhWjfSDu+y35M+zBbC/PMiM4iePC3xzEIVAt2e1CeyNoN1s= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=fail (p=quarantine dis=none) header.from=xiaomi.com; spf=pass smtp.mailfrom=126.com; dkim=pass (1024-bit key) header.d=126.com header.i=@126.com header.b=KbRSCNM9; arc=none smtp.client-ip=117.135.210.9 Authentication-Results: smtp.subspace.kernel.org; dmarc=fail (p=quarantine dis=none) header.from=xiaomi.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=126.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=126.com header.i=@126.com header.b="KbRSCNM9" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=126.com; s=s110527; h=From:To:Subject:Date:Message-ID:MIME-Version; bh=oi agiE9hFnek0JhwqrKr3HLWo3wKjcwb53js2X7yX1U=; b=KbRSCNM9QcphAVB1hB T/lNSKjzk5d330PDJuLy48zvyXqQa+3NV9yBrYRFPeHcbXY1Zm6qXxS0GLB8ONIo 27q/0xvK15Mm+46fJeDdzYLdvI7t/poahyuY3Dj8cd+yWuw96qJDsp4OfptbKpxe 8TvHd0p16p5VOku9uid5qn9/c= Received: from nzzhao-ThinkCentre-M760t.mioffice.cn (unknown []) by gzsmtp3 (Coremail) with SMTP id PikvCgDXreo9x6hq3jBvHw--.36918S10; Tue, 15 Sep 2026 12:19:17 +0800 (CST) From: Nanzhe Zhao To: linux-f2fs-devel@lists.sourceforge.net, Jaegeuk Kim , Chao Yu Cc: Barry Song , Juan Yescas , Dev Jain , linux-kernel@vger.kernel.org, David Hildenbrand , Bo Zhang , Kalesh Singh , Nanzhe Zhao , Pengfei Li , Ryan Roberts , Nanzhe Zhao Subject: [PATCH v2 08/14] f2fs: optimize small block size large folio read Date: Tue, 15 Sep 2026 12:19:03 +0800 Message-ID: <20260915041909.2903887-9-zhaonanzhe@xiaomi.com> X-Mailer: git-send-email 2.43.0 In-Reply-To: <20260915041909.2903887-1-zhaonanzhe@xiaomi.com> References: <20260915041909.2903887-1-zhaonanzhe@xiaomi.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-CM-TRANSID: PikvCgDXreo9x6hq3jBvHw--.36918S10 X-Coremail-Antispam: 1Uf129KBjvJXoWxJFW7urWUGF4UAryxXF18Zrb_yoWrKr1rpF W5Krn5KrZ5WFWxWr4IqF1DZr1ft348Wa1UGFZ3Gw1fX3Wjq34S9F1vy34YgF4rtrykCFna qF4rKryDWF1jqF7anT9S1TB71UUUUU7qnTZGkaVYY2UrUUUUjbIjqfuFe4nvWSU5nxnvy2 9KBjDUYxBIdaVFxhVjvjDU0xZFpf9x07jUZ2-UUUUU= Sender: nzzhao@126.com X-CM-SenderInfo: xq22xtbr6rjloofrz/xtbBowbHl2qox0ZomwAA3F Content-Type: text/plain; charset="utf-8" The original f2fs_read_data_large_folio() implementation has limited benefit with a 4KB block size, mainly because updating read_pages_pending greatly increases the number of spinlock operations. Use len_blks to batch read_pages_pending and iostat updates for contiguous mapped blocks. If the contiguous mapping covers the whole folio, skip f2fs_folio_state allocation for that folio. Signed-off-by: Nanzhe Zhao --- fs/f2fs/data.c | 65 ++++++++++++++++++++++++++++++++++---------------- 1 file changed, 45 insertions(+), 20 deletions(-) diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index 287d83debf95..09ad5cf0d9aa 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -173,8 +173,9 @@ static void f2fs_finish_read_bio(struct bio *bio, bool = in_task) continue; } =20 - if (folio_test_large(folio)) { - struct f2fs_folio_state *ffs =3D folio->private; + if (f2fs_folio_has_ffs(folio)) { + struct f2fs_folio_state *ffs =3D + (struct f2fs_folio_state *)folio->private; =20 spin_lock_irqsave(&ffs->state_lock, flags); ffs->read_pages_pending -=3D nr_pages; @@ -2844,8 +2845,7 @@ static int f2fs_read_data_large_folio(struct inode *i= node, pgoff_t index, offset, next_pgofs =3D 0; unsigned max_nr_pages =3D rac ? readahead_count(rac) : folio_nr_pages(folio); - unsigned int nrpages; - struct f2fs_folio_state *ffs; + unsigned int nrpages, len_blks; int ret =3D 0; bool folio_in_bio =3D false; =20 @@ -2868,11 +2868,17 @@ static int f2fs_read_data_large_folio(struct inode = *inode, folio_in_bio =3D false; index =3D folio->index; offset =3D 0; - ffs =3D NULL; nrpages =3D folio_nr_pages(folio); =20 - for (; nrpages; nrpages--, max_nr_pages--, index++, offset++) { + for (; nrpages; + nrpages -=3D len_blks, max_nr_pages -=3D len_blks, + index +=3D len_blks, offset +=3D len_blks) { sector_t block_nr; + bool whole_folio_in_bio; + unsigned int i; + + len_blks =3D 1; + /* * Map blocks using the previous result first. */ @@ -2901,13 +2907,31 @@ static int f2fs_read_data_large_folio(struct inode = *inode, got_it: if ((map.m_flags & F2FS_MAP_MAPPED)) { block_nr =3D map.m_pblk + index - map.m_lblk; - if (!f2fs_is_valid_blkaddr(F2FS_I_SB(inode), block_nr, + + len_blks =3D min_t(unsigned int, nrpages, max_nr_pages); + len_blks =3D min_t(unsigned int, len_blks, + (unsigned int)(map.m_lblk + map.m_len - index)); + + for (i =3D 0; i < len_blks; i++) { + if (!f2fs_is_valid_blkaddr(F2FS_I_SB(inode), + block_nr + i, DATA_GENERIC_ENHANCE_READ)) { - ret =3D -EFSCORRUPTED; - goto err_out; + ret =3D -EFSCORRUPTED; + goto err_out; + } } + + /* + * If an entire folio is added to one bio, + * folio_end_read() can complete the folio read status + * without relying on f2fs_folio_state. + */ + whole_folio_in_bio =3D offset =3D=3D 0 && + len_blks =3D=3D folio_nr_pages(folio); + } else { size_t page_offset =3D offset << PAGE_SHIFT; + folio_zero_range(folio, page_offset, PAGE_SIZE); if (vi && !fsverity_verify_blocks(vi, folio, PAGE_SIZE, page_offset)) { ret =3D -EIO; @@ -2917,15 +2941,13 @@ static int f2fs_read_data_large_folio(struct inode = *inode, } =20 /* We must increment read_pages_pending before possible BIOs submitting - * to prevent from premature folio_end_read() call on folio + * to prevent from premature folio_end_read() call on folio. */ - if (folio_test_large(folio)) { - ffs =3D f2fs_ffs_find_or_alloc(folio); + if (folio_test_large(folio) && !whole_folio_in_bio) { + f2fs_ffs_find_or_alloc(folio); =20 /* set the bitmap to wait */ - spin_lock_irq(&ffs->state_lock); - ffs->read_pages_pending++; - spin_unlock_irq(&ffs->state_lock); + f2fs_update_read_folio_pending(folio, len_blks); } =20 /* @@ -2949,17 +2971,20 @@ static int f2fs_read_data_large_folio(struct inode = *inode, * If the page is under writeback, we need to wait for * its completion to see the correct decrypted data. */ - f2fs_wait_on_block_writeback(inode, block_nr); + for (i =3D 0; i < len_blks; i++) + f2fs_wait_on_block_writeback(inode, block_nr + i); =20 - if (!bio_add_folio(bio, folio, F2FS_BLKSIZE(F2FS_I_SB(inode)), + if (!bio_add_folio(bio, folio, + len_blks * F2FS_BLKSIZE(F2FS_I_SB(inode)), offset << PAGE_SHIFT)) goto submit_and_realloc; =20 folio_in_bio =3D true; - inc_page_count(F2FS_I_SB(inode), F2FS_RD_DATA); + for (i =3D 0; i < len_blks; i++) + inc_page_count(F2FS_I_SB(inode), F2FS_RD_DATA); f2fs_update_iostat(F2FS_I_SB(inode), NULL, FS_DATA_READ_IO, - F2FS_BLKSIZE(F2FS_I_SB(inode))); - last_block_in_bio =3D block_nr; + len_blks * F2FS_BLKSIZE(F2FS_I_SB(inode))); + last_block_in_bio =3D block_nr + len_blks - 1; } trace_f2fs_read_folio(folio, DATA); err_out: --=20 2.43.0 From nobody Fri Sep 25 08:47:33 2026 Received: from m16.mail.126.com (m16.mail.126.com [117.135.210.6]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id AC2313D6488 for ; Tue, 15 Sep 2026 04:20:01 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=117.135.210.6 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1789446007; cv=none; b=XS+Rh1mA9jYppK+nzu6sjKSC63D64JkaJVlt05NZeAw1X5Gm4hlkyVytwnedTC+fOBtBUylAEPnN4zLUrpk2N3bre/ybavDhIiT4n4ln/V2FcPndJitlttg0iWQDcgfvsl7A+BD6G6AcyQsvYgmDTjXD4iBAhZSZl/j5l33YchQ= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1789446007; c=relaxed/simple; bh=5GllvYFHKB3oxiWSOxPy75yfEIXgbGcS5SfBwQ9U3zY=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=VnfD1HxbCZAqsM/UfEEQBCtmZEEVdZ1pn93b5MjjxD2R0b2gH+iBYc/oqJL65/CpHIYDoHRbW++++a2A/sx/0JwZzmr3Aur5TQ07uN5NBOr5a6pYA1mmTDGmcEjzelT6tcMvWT74bw9AIt09BMjejO2DklJrg6lrCNksf+NWo8Y= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=fail (p=quarantine dis=none) header.from=xiaomi.com; spf=pass smtp.mailfrom=126.com; dkim=pass (1024-bit key) header.d=126.com header.i=@126.com header.b=BN4gCwp+; arc=none smtp.client-ip=117.135.210.6 Authentication-Results: smtp.subspace.kernel.org; dmarc=fail (p=quarantine dis=none) header.from=xiaomi.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=126.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=126.com header.i=@126.com header.b="BN4gCwp+" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=126.com; s=s110527; h=From:To:Subject:Date:Message-ID:MIME-Version; bh=84 hc+WBF4d/CffpQuiXGcMHzVZBCeWFgiS9xB1Izx4k=; b=BN4gCwp+Y1RqRC7mlX 0e+l0Uxl3Q9xPUAGzb70xLlWnRJRbOaRBIkGkRfYJy87eu2jz2M4cttgrJQLGSY7 1lfaeZzaNBgc4IkDLFtlmBkw005h66aUHakp78JXiFx/QykhUTkFYF3rQzsSG3KB 6v4PgqBPs3+KBXR6c+YfLh+1g= Received: from nzzhao-ThinkCentre-M760t.mioffice.cn (unknown []) by gzsmtp3 (Coremail) with SMTP id PikvCgDXreo9x6hq3jBvHw--.36918S11; Tue, 15 Sep 2026 12:19:18 +0800 (CST) From: Nanzhe Zhao To: linux-f2fs-devel@lists.sourceforge.net, Jaegeuk Kim , Chao Yu Cc: Barry Song , Juan Yescas , Dev Jain , linux-kernel@vger.kernel.org, David Hildenbrand , Bo Zhang , Kalesh Singh , Nanzhe Zhao , Pengfei Li , Ryan Roberts , Nanzhe Zhao Subject: [PATCH v2 09/14] f2fs: support partial uptodate large folio read Date: Tue, 15 Sep 2026 12:19:04 +0800 Message-ID: <20260915041909.2903887-10-zhaonanzhe@xiaomi.com> X-Mailer: git-send-email 2.43.0 In-Reply-To: <20260915041909.2903887-1-zhaonanzhe@xiaomi.com> References: <20260915041909.2903887-1-zhaonanzhe@xiaomi.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-CM-TRANSID: PikvCgDXreo9x6hq3jBvHw--.36918S11 X-Coremail-Antispam: 1Uf129KBjvJXoWxuF4DGF1kCr17tF1UZrWfAFb_yoWruFy7pF W8K398KrW5JrW3WF1SqFs3Zr1SqryIqay7GFZ3G34fJ3WYqa4SgFykt34YkF4rtrykCF1f Za15GryUWF4jyFJanT9S1TB71UUUUU7qnTZGkaVYY2UrUUUUjbIjqfuFe4nvWSU5nxnvy2 9KBjDUYxBIdaVFxhVjvjDU0xZFpf9x07jUZ2-UUUUU= Sender: nzzhao@126.com X-CM-SenderInfo: xq22xtbr6rjloofrz/xtbBowbHl2qox0ZopAAA36 Content-Type: text/plain; charset="utf-8" Buffered write can have write bytes smaller than folio size for cases when folio minimum order is not zero. This can left partially uptodate folio in page cache. So we skip uptodate subpage read in read_data_large_folio. Also mark hole subpage uptodate in uptodate bitmap. Signed-off-by: Nanzhe Zhao --- fs/f2fs/data.c | 71 +++++++++++++++++++++++++++++++++++++++++++++++++- 1 file changed, 70 insertions(+), 1 deletion(-) diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index 09ad5cf0d9aa..d9e3a5c370d7 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -163,6 +163,7 @@ static void f2fs_finish_read_bio(struct bio *bio, bool = in_task) struct folio *folio =3D fi.folio; unsigned int nr_pages =3D fi.length >> PAGE_SHIFT; bool finished =3D true; + bool uptodate =3D bio->bi_status =3D=3D BLK_STS_OK; =20 if (!folio_test_large(folio) && f2fs_is_compressed_page(folio)) { @@ -178,6 +179,9 @@ static void f2fs_finish_read_bio(struct bio *bio, bool = in_task) (struct f2fs_folio_state *)folio->private; =20 spin_lock_irqsave(&ffs->state_lock, flags); + if (bio->bi_status =3D=3D BLK_STS_OK) + uptodate =3D __ffs_mark_subrange_uptodate(folio, ffs, + fi.offset, fi.length); ffs->read_pages_pending -=3D nr_pages; finished =3D !ffs->read_pages_pending; spin_unlock_irqrestore(&ffs->state_lock, flags); @@ -194,7 +198,8 @@ static void f2fs_finish_read_bio(struct bio *bio, bool = in_task) bio->bi_status =3D BLK_STS_IOERR; =20 if (finished) - folio_end_read(folio, bio->bi_status =3D=3D BLK_STS_OK); + folio_end_read(folio, + bio->bi_status =3D=3D BLK_STS_OK && uptodate); } =20 if (ctx) @@ -2835,6 +2840,56 @@ static bool find_next_valid_block(const struct folio= *folio, return false; } =20 +static unsigned int ffs_next_uptodate_subpage(struct f2fs_folio_state *ffs, + unsigned int start, unsigned int end) +{ + return find_next_bit(ffs->state, end + 1, start); +} + +static unsigned int ffs_next_nonuptodate_subpage(struct f2fs_folio_state *= ffs, + unsigned int start, unsigned int end) +{ + return find_next_zero_bit(ffs->state, end + 1, start); +} + +static void f2fs_skip_fully_uptodate_front(struct folio *folio, + pgoff_t *index, pgoff_t *offset, unsigned int *nrpages, + unsigned int *max_nr_pages) +{ + struct f2fs_folio_state *ffs; + unsigned int next, skipped; + + if (!f2fs_folio_has_ffs(folio) || !*nrpages) + return; + + ffs =3D folio->private; + next =3D ffs_next_nonuptodate_subpage(ffs, *offset, + *offset + *nrpages - 1); + skipped =3D next - *offset; + if (!skipped) + return; + + *index +=3D skipped; + *offset +=3D skipped; + *nrpages -=3D skipped; + *max_nr_pages -=3D skipped; +} + +static void f2fs_truncate_read_extent(struct folio *folio, pgoff_t offset, + unsigned int *len_blks) +{ + struct f2fs_folio_state *ffs; + unsigned int next, end; + + if (!f2fs_folio_has_ffs(folio) || *len_blks <=3D 1) + return; + + ffs =3D folio->private; + end =3D offset + *len_blks - 1; + next =3D ffs_next_uptodate_subpage(ffs, offset + 1, end); + if (next <=3D end) + *len_blks =3D next - offset; +} static int f2fs_read_data_large_folio(struct inode *inode, struct fsverity_info *vi, struct readahead_control *rac, struct folio *folio) @@ -2846,6 +2901,7 @@ static int f2fs_read_data_large_folio(struct inode *i= node, unsigned max_nr_pages =3D rac ? readahead_count(rac) : folio_nr_pages(folio); unsigned int nrpages, len_blks; + struct f2fs_folio_state *ffs; int ret =3D 0; bool folio_in_bio =3D false; =20 @@ -2879,6 +2935,11 @@ static int f2fs_read_data_large_folio(struct inode *= inode, =20 len_blks =3D 1; =20 + f2fs_skip_fully_uptodate_front(folio, &index, &offset, + &nrpages, &max_nr_pages); + if (!nrpages) + break; + /* * Map blocks using the previous result first. */ @@ -2911,6 +2972,7 @@ static int f2fs_read_data_large_folio(struct inode *i= node, len_blks =3D min_t(unsigned int, nrpages, max_nr_pages); len_blks =3D min_t(unsigned int, len_blks, (unsigned int)(map.m_lblk + map.m_len - index)); + f2fs_truncate_read_extent(folio, offset, &len_blks); =20 for (i =3D 0; i < len_blks; i++) { if (!f2fs_is_valid_blkaddr(F2FS_I_SB(inode), @@ -2937,6 +2999,13 @@ static int f2fs_read_data_large_folio(struct inode *= inode, ret =3D -EIO; goto err_out; } + if (folio_test_large(folio)) { + ffs =3D f2fs_ffs_find_or_alloc(folio); + spin_lock_irq(&ffs->state_lock); + __ffs_mark_subrange_uptodate(folio, ffs, + page_offset, PAGE_SIZE); + spin_unlock_irq(&ffs->state_lock); + } continue; } =20 --=20 2.43.0 From nobody Fri Sep 25 08:47:33 2026 Received: from m16.mail.126.com (m16.mail.126.com [117.135.210.6]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 35CF9336883 for ; Tue, 15 Sep 2026 04:20:36 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=117.135.210.6 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1789446039; cv=none; b=qibYyKy2kSD5drCeFDm2JPVEwSTLfZxzNiSVyE/jOzFwY26ztMF6qc7LN8O/reZ/M6h8j6ofDH08HID3KMZR+nHBbv7OzGct3HGmWwFlbFLRc1TEBLn1Pb9r5dtFvLyUURnMD8Fed82032hcOP00/dMRSvld/QlDKe1lmAB8yyg= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1789446039; c=relaxed/simple; bh=Z9DN6YJ/BsepswIpAuuiNFz21ILM9xadLwZgJpGmrs8=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=L9gGLHumHzffiYPT+9MD973i/7nOtCOa17JxJ3u3Od/eHBXOaEDaDYMWi0uKFD0lnPSPeBJBzFlLJaSjx/2h3W5MsXvFuGu60AtYfoD92HrdNIU7IaqZcN9I7U/gY4w+In1CwcGdGityTGlelOgzSOmzbM7trtIMmvqrU68dJQo= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=fail (p=quarantine dis=none) header.from=xiaomi.com; spf=pass smtp.mailfrom=126.com; dkim=pass (1024-bit key) header.d=126.com header.i=@126.com header.b=cv+Vlzm+; arc=none smtp.client-ip=117.135.210.6 Authentication-Results: smtp.subspace.kernel.org; dmarc=fail (p=quarantine dis=none) header.from=xiaomi.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=126.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=126.com header.i=@126.com header.b="cv+Vlzm+" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=126.com; s=s110527; h=From:To:Subject:Date:Message-ID:MIME-Version; bh=x9 ZlzP7KpSO+gqVQvXv2VHlI3hj+Jnhl/T+yZmtX9aA=; b=cv+Vlzm+6vFH8TIKpU +jCaq6/IFUgwG6Tbq7jIdyH98NikcwLI7CkLITFjHHWvXJ7qydLpLNo/ide14Hfp nXxG+XDRVBCEBXo8oVTJ8psGGryxtLk6cH+KVj2/ZuLUBaf5KploI9IstmBi0pEr h+z/lZ0nqtQjqKQcqPtplD4Tk= Received: from nzzhao-ThinkCentre-M760t.mioffice.cn (unknown []) by gzsmtp3 (Coremail) with SMTP id PikvCgDXreo9x6hq3jBvHw--.36918S12; Tue, 15 Sep 2026 12:19:19 +0800 (CST) From: Nanzhe Zhao To: linux-f2fs-devel@lists.sourceforge.net, Jaegeuk Kim , Chao Yu Cc: Barry Song , Juan Yescas , Dev Jain , linux-kernel@vger.kernel.org, David Hildenbrand , Bo Zhang , Kalesh Singh , Nanzhe Zhao , Pengfei Li , Ryan Roberts , Nanzhe Zhao Subject: [PATCH v2 10/14] f2fs: handle partial truncate of large folio dirty subpages Date: Tue, 15 Sep 2026 12:19:05 +0800 Message-ID: <20260915041909.2903887-11-zhaonanzhe@xiaomi.com> X-Mailer: git-send-email 2.43.0 In-Reply-To: <20260915041909.2903887-1-zhaonanzhe@xiaomi.com> References: <20260915041909.2903887-1-zhaonanzhe@xiaomi.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-CM-TRANSID: PikvCgDXreo9x6hq3jBvHw--.36918S12 X-Coremail-Antispam: 1Uf129KBjvJXoWxAr4fuFyUXr4Utw15Xw4Dtwb_yoW5KF1fpF WDCr98JFs8J3Zruw1fXFn8Xr1fK3s7Wr4xCa97Gwn3J3WYq3sxKF1kt34FgFs8Jry8AF1f Xr1UtFW5WFyUKF7anT9S1TB71UUUUU7qnTZGkaVYY2UrUUUUjbIjqfuFe4nvWSU5nxnvy2 9KBjDUYxBIdaVFxhVjvjDU0xZFpf9x07jjbyZUUUUU= Sender: nzzhao@126.com X-CM-SenderInfo: xq22xtbr6rjloofrz/xtbBsQfHl2qox0cWjwAA39 Content-Type: text/plain; charset="utf-8" A large folio can be partial truncated and stays in folio mapping, we need to clear the subrange dirty bits and uptodate bits that the partial truncate covers. If this partial truncate happens to clear the last subrange dirty bits, then cancel the whole folio dirty state. Also add a guard in f2fs_write_single_data_folio() so a large folio subpage whose disk block was already truncated (NULL_ADDR) is skipped Signed-off-by: Nanzhe Zhao --- fs/f2fs/data.c | 66 +++++++++++++++++++++++++++++++++++++++++++++++++- 1 file changed, 65 insertions(+), 1 deletion(-) diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index d9e3a5c370d7..bf9b9c9b722f 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -2679,6 +2679,34 @@ static void f2fs_ffs_mark_subrange_uptodate(struct f= olio *folio, size_t offset, folio_mark_uptodate(folio); } =20 +static void ffs_clear_subrange_uptodate(struct folio *folio, + size_t offset, size_t len) +{ + struct f2fs_folio_state *ffs; + unsigned int nr_subpages, start, end; + unsigned long flags; + + f2fs_bug_on(F2FS_F_SB(folio), offset + len > folio_size(folio)); + + if (!f2fs_folio_has_ffs(folio)) { + if (folio_test_uptodate(folio)) + folio_clear_uptodate(folio); + return; + } + + ffs =3D (struct f2fs_folio_state *)folio->private; + nr_subpages =3D folio_nr_pages(folio); + start =3D offset >> PAGE_SHIFT; + end =3D (offset + len + PAGE_SIZE - 1) >> PAGE_SHIFT; + end =3D min(end, nr_subpages); + + spin_lock_irqsave(&ffs->state_lock, flags); + bitmap_clear(ffs->state, start, end - start); + spin_unlock_irqrestore(&ffs->state_lock, flags); + if (folio_test_uptodate(folio)) + folio_clear_uptodate(folio); +} + bool f2fs_ffs_test_blk_dirty(const struct folio *folio, pgoff_t index) { struct f2fs_folio_state *ffs; @@ -3545,6 +3573,14 @@ static int f2fs_write_single_data_folio(struct folio= *folio, int *submitted, =20 fio.old_blkaddr =3D dn.data_blkaddr; =20 + /* This page is already truncated */ + if (fio.old_blkaddr =3D=3D NULL_ADDR) { + ffs_clear_subrange_uptodate(folio, + i << PAGE_SHIFT, PAGE_SIZE); + folio_clear_f2fs_gcing(folio); + goto block_done; + } + got_it: if (__is_valid_data_blkaddr(fio.old_blkaddr) && !f2fs_is_valid_blkaddr(sbi, fio.old_blkaddr, @@ -4987,8 +5023,36 @@ void f2fs_invalidate_folio(struct folio *folio, size= _t offset, size_t length) struct f2fs_sb_info *sbi =3D F2FS_I_SB(inode); =20 if (inode->i_ino >=3D F2FS_ROOT_INO(sbi) && - (offset || length !=3D folio_size(folio))) + (offset || length !=3D folio_size(folio))) { + size_t clear_start =3D round_up(offset, PAGE_SIZE); + size_t clear_end =3D round_down(offset + length, PAGE_SIZE); + size_t clear_length; + bool dirty; + + /* + * If the truncated range falls within a single subpage, no + * subpage state needs to be cleared. + */ + if (clear_start >=3D clear_end || !f2fs_folio_has_ffs(folio)) + return; + + clear_length =3D clear_end - clear_start; + dirty =3D f2fs_ffs_clear_subrange_dirty_and_test(folio, + clear_start, clear_length); + ffs_clear_subrange_uptodate(folio, clear_start, clear_length); + + /* + * If the truncated subrange happens to clear the remaining + * dirty bitmap of the whole folio, cancel the folio-level + * dirty state. + */ + if (!dirty && folio_test_dirty(folio)) { + inode_dec_dirty_pages(inode); + f2fs_remove_dirty_inode(inode); + folio_cancel_dirty(folio); + } return; + } =20 if (folio_test_dirty(folio)) { if (inode->i_ino =3D=3D F2FS_META_INO(sbi)) { --=20 2.43.0 From nobody Fri Sep 25 08:47:33 2026 Received: from m16.mail.126.com (m16.mail.126.com [117.135.210.7]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 7B0B826FD9B for ; Tue, 15 Sep 2026 04:25:38 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=117.135.210.7 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1789446341; cv=none; b=W4EajOz2T6nIVfx9Ug7VuEGnEfuHWcqXko3BdBlCx/nYIFUbVGH9Sydt6MMnhK89hyLApKoLh5FClpOyJEC/JczK4WR1y2BYlhkUpgrvB7BukSVU7JFbJ2FxJrr2ze5YUJLEVJM6fLRDsEOL6XccVIM38fNyYPx4wx/Mjm5vu0c= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1789446341; c=relaxed/simple; bh=BiCsoswgpdC7GMVj5HeviDLEvntohEuSPlAB07BVrG0=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=blnr1kY0q1RiNwY6Sw08M30NzUxSpi86Eis+b79KfywuZGu9Usemd342g12qYcWZG7e0eoxXiI4/fizmAP31CNRI4b6/um+NbD+JR45ave+Y0hgucofIzr1nDeEf63B4vH7wlp14ggf+E6Jam5SJmZzzSHQw4Qg0UQLaeqqP92w= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=fail (p=quarantine dis=none) header.from=xiaomi.com; spf=pass smtp.mailfrom=126.com; dkim=pass (1024-bit key) header.d=126.com header.i=@126.com header.b=mkHh/Axv; arc=none smtp.client-ip=117.135.210.7 Authentication-Results: smtp.subspace.kernel.org; dmarc=fail (p=quarantine dis=none) header.from=xiaomi.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=126.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=126.com header.i=@126.com header.b="mkHh/Axv" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=126.com; s=s110527; h=From:To:Subject:Date:Message-ID:MIME-Version; bh=4u 7cyU8gIpacYtDPlo3aSPUr2tv6Jzc7cfjlIxvbPhA=; b=mkHh/AxvOxnRAB60ws KpxfpaWEsf9u7h67Biq4UdRqA3eWVgLGuc7f47BReIyR0riVbod6PFgSQoPHNQs7 jr6DTBMVAxqH6VSWQgSTiPRaQHMw4KB7l+54XiwgiJxdazsaq4UM6Ju7S12AZ+vJ /MPWysbJeYmOW5vHYaxm1MHPE= Received: from nzzhao-ThinkCentre-M760t.mioffice.cn (unknown []) by gzga-smtp-mtada-g0-4 (Coremail) with SMTP id _____wDXG+KfyKhqTDskBQ--.23380S2; Tue, 15 Sep 2026 12:25:04 +0800 (CST) From: Nanzhe Zhao To: linux-f2fs-devel@lists.sourceforge.net, Jaegeuk Kim , Chao Yu Cc: Barry Song , Juan Yescas , Dev Jain , linux-kernel@vger.kernel.org, David Hildenbrand , Bo Zhang , Kalesh Singh , Nanzhe Zhao , Pengfei Li , Ryan Roberts , Nanzhe Zhao Subject: [PATCH v2 11/14] f2fs: fix zeroing paths for large folios Date: Tue, 15 Sep 2026 12:25:00 +0800 Message-ID: <20260915042503.2912616-1-zhaonanzhe@xiaomi.com> X-Mailer: git-send-email 2.43.0 In-Reply-To: <20260915041909.2903887-1-zhaonanzhe@xiaomi.com> References: <20260915041909.2903887-1-zhaonanzhe@xiaomi.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-CM-TRANSID: _____wDXG+KfyKhqTDskBQ--.23380S2 X-Coremail-Antispam: 1Uf129KBjvJXoWxWF4DKF1xAFy7Gr15Kr17Wrg_yoW7JF1rpr yUGasrGrs5GrW8ur1IqF1DZw1S93sxWF1UCFWxGasaya4Yq3s5KF18t34j9F15trW8Arn7 X3WrKFW5WF4DJF7anT9S1TB71UUUUU7qnTZGkaVYY2UrUUUUjbIjqfuFe4nvWSU5nxnvy2 9KBjDUYxBIdaVFxhVjvjDU0xZFpf9x07UA3ktUUUUU= Sender: nzzhao@126.com X-CM-SenderInfo: xq22xtbr6rjloofrz/xtbBogAf7mqoyKAZKQAA3J Content-Type: text/plain; charset="utf-8" Several f2fs zeroing paths still use PAGE_SIZE based offsets after a file mapping can contain large folios. This is fine for order-0 folios, but it zeros the wrong range once the target block is not at offset 0 in a large folio. Use offset_in_folio() to translate the file block index to the folio offset before zeroing data in truncate_partial_data_page(), fill_zero(), and f2fs_get_new_data_folio(). Signed-off-by: Nanzhe Zhao --- fs/f2fs/data.c | 16 ++++++++++++---- fs/f2fs/f2fs.h | 1 + fs/f2fs/file.c | 21 ++++++++++++++++----- 3 files changed, 29 insertions(+), 9 deletions(-) diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index bf9b9c9b722f..295424f2ade5 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -1549,6 +1549,12 @@ struct folio *f2fs_get_new_data_folio(struct inode *= inode, return ERR_PTR(-ENOMEM); } =20 + if (folio_test_large(folio) && !f2fs_ffs_find_or_alloc(folio)) { + f2fs_folio_put(folio, true); + f2fs_folio_put(ifolio, true); + return ERR_PTR(-ENOMEM); + } + set_new_dnode(&dn, inode, ifolio, NULL, 0); err =3D f2fs_reserve_block(&dn, index); if (err) { @@ -1562,9 +1568,11 @@ struct folio *f2fs_get_new_data_folio(struct inode *= inode, goto got_it; =20 if (dn.data_blkaddr =3D=3D NEW_ADDR) { - folio_zero_segment(folio, 0, folio_size(folio)); - if (!folio_test_uptodate(folio)) - folio_mark_uptodate(folio); + size_t off =3D offset_in_folio(folio, + (loff_t)index << PAGE_SHIFT); + + folio_zero_segment(folio, off, off + PAGE_SIZE); + f2fs_ffs_mark_subrange_uptodate(folio, off, PAGE_SIZE); } else { f2fs_folio_put(folio, true); =20 @@ -2656,7 +2664,7 @@ static bool __ffs_mark_subrange_uptodate(struct folio= *folio, return bitmap_full(ffs->state, nr_subpages); } =20 -static void f2fs_ffs_mark_subrange_uptodate(struct folio *folio, size_t of= fset, +void f2fs_ffs_mark_subrange_uptodate(struct folio *folio, size_t offset, size_t len) { struct f2fs_folio_state *ffs; diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index 6be08e45341c..d99a812810d3 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -4381,6 +4381,7 @@ struct folio *f2fs_get_lock_data_folio(struct inode *= inode, pgoff_t index, bool for_write); struct folio *f2fs_get_new_data_folio(struct inode *inode, struct folio *ifolio, pgoff_t index, bool new_i_size); +void f2fs_ffs_mark_subrange_uptodate(struct folio *folio, size_t offset, s= ize_t len); int f2fs_do_write_data_page(struct f2fs_io_info *fio); int f2fs_map_blocks(struct inode *inode, struct f2fs_map_blocks *map, int = flag); int f2fs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index 738a751c3903..54aa045b549e 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -825,12 +825,13 @@ void f2fs_truncate_data_blocks_range(struct dnode_of_= data *dn, int count) static int truncate_partial_data_page(struct inode *inode, u64 from, bool cache_only) { - loff_t offset =3D from & (PAGE_SIZE - 1); pgoff_t index =3D from >> PAGE_SHIFT; struct address_space *mapping =3D inode->i_mapping; struct folio *folio; + size_t folio_off; + size_t blk_end; =20 - if (!offset && !cache_only) + if (!(from & (PAGE_SIZE - 1)) && !cache_only) return 0; =20 if (cache_only) { @@ -848,12 +849,18 @@ static int truncate_partial_data_page(struct inode *i= node, u64 from, return PTR_ERR(folio) =3D=3D -ENOENT ? 0 : PTR_ERR(folio); truncate_out: f2fs_folio_wait_writeback(folio, DATA, true, true); - folio_zero_segment(folio, offset, folio_size(folio)); + folio_off =3D offset_in_folio(folio, from); + blk_end =3D min_t(size_t, round_up(folio_off, PAGE_SIZE), + folio_size(folio)); + folio_zero_segment(folio, folio_off, blk_end); =20 /* An encrypted inode should have a key and truncate the last page. */ f2fs_bug_on(F2FS_I_SB(inode), cache_only && IS_ENCRYPTED(inode)); - if (!cache_only) + if (!cache_only) { + f2fs_ffs_mark_subrange_dirty(folio, folio_off, + blk_end - folio_off); folio_mark_dirty(folio); + } f2fs_folio_put(folio, true); return 0; } @@ -1321,6 +1328,7 @@ static int fill_zero(struct inode *inode, pgoff_t ind= ex, struct f2fs_sb_info *sbi =3D F2FS_I_SB(inode); struct folio *folio; struct f2fs_lock_context lc; + size_t folio_off; =20 if (!len) return 0; @@ -1335,7 +1343,10 @@ static int fill_zero(struct inode *inode, pgoff_t in= dex, return PTR_ERR(folio); =20 f2fs_folio_wait_writeback(folio, DATA, true, true); - folio_zero_range(folio, start, len); + folio_off =3D offset_in_folio(folio, + (loff_t)index << PAGE_SHIFT) + start; + folio_zero_range(folio, folio_off, len); + f2fs_ffs_mark_subrange_dirty(folio, folio_off, len); folio_mark_dirty(folio); f2fs_folio_put(folio, true); return 0; --=20 2.43.0 From nobody Fri Sep 25 08:47:33 2026 Received: from m16.mail.126.com (m16.mail.126.com [117.135.210.7]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id A84972EA754 for ; Tue, 15 Sep 2026 04:25:38 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=117.135.210.7 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1789446341; cv=none; b=kKIlBCE0kGpq98ym0JlYN1o87GyaWke0huAYf+kVBuvOIVtyJjsNQ+CYMMgzqWe0YwOsA5M/hAEKaM286JVUtPiDr4rB/E8y6sRAbVocUcvL+O1IHrg9hlb8eIIp55lQBFuDyX0jN3cOvuyEppMqs8JI/x9tIER4pejg+FtHPeI= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1789446341; c=relaxed/simple; bh=98V6SAFLSemsKaAedK+2Ue7kL3AHriW/BUrBwixC3Lw=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=PDq4o63Swf3yN5/NefuWbQPgckk13K0/MgR3Imlamg7Mk/raePCjwIuzCwnDY7sg9Y1Ky9TalD/SNY2VQwYfaxM/BZNDzYeUUgNXgwimK/JvQtSnqoXObnINGJijR98z7FduTCili/tBxY5MmVm7Ccc2++n73ZUR1IqvQqKFDGU= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=fail (p=quarantine dis=none) header.from=xiaomi.com; spf=pass smtp.mailfrom=126.com; dkim=pass (1024-bit key) header.d=126.com header.i=@126.com header.b=SvTRYz8O; arc=none smtp.client-ip=117.135.210.7 Authentication-Results: smtp.subspace.kernel.org; dmarc=fail (p=quarantine dis=none) header.from=xiaomi.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=126.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=126.com header.i=@126.com header.b="SvTRYz8O" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=126.com; s=s110527; h=From:To:Subject:Date:Message-ID:MIME-Version; bh=Kj RYvMkldZCdlNU79CyV1S07+8U1NHukSrBQclKwm7w=; b=SvTRYz8OkWqX2QkZGV 7DlKSFboiiQ1+EC90Ee2qdUmqHOaSYNKUsAw1lyROj/UsDb1bpo8lggXFPxERsTm k0RVgRNs1+B97bBw8AnJaltUAT9eDmPcpDN9FnFlSCokk9vpotGT+2NJKoF/kzGj lfSkqETzNcS3as6Bh5E4XbTuU= Received: from nzzhao-ThinkCentre-M760t.mioffice.cn (unknown []) by gzga-smtp-mtada-g0-4 (Coremail) with SMTP id _____wDXG+KfyKhqTDskBQ--.23380S3; Tue, 15 Sep 2026 12:25:05 +0800 (CST) From: Nanzhe Zhao To: linux-f2fs-devel@lists.sourceforge.net, Jaegeuk Kim , Chao Yu Cc: Barry Song , Juan Yescas , Dev Jain , linux-kernel@vger.kernel.org, David Hildenbrand , Bo Zhang , Kalesh Singh , Nanzhe Zhao , Pengfei Li , Ryan Roberts , Nanzhe Zhao Subject: [PATCH v2 12/14] f2fs: handle block cloning within the same large folio Date: Tue, 15 Sep 2026 12:25:01 +0800 Message-ID: <20260915042503.2912616-2-zhaonanzhe@xiaomi.com> X-Mailer: git-send-email 2.43.0 In-Reply-To: <20260915041909.2903887-1-zhaonanzhe@xiaomi.com> References: <20260915041909.2903887-1-zhaonanzhe@xiaomi.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-CM-TRANSID: _____wDXG+KfyKhqTDskBQ--.23380S3 X-Coremail-Antispam: 1Uf129KBjvJXoWxXF4UWFyUAry7CF1fKr4kWFg_yoW5Cr1rpF y5Grs5Z34kCF1xWryftF17ZryFgr97Ww4UJFs7uw1fX3W5X3WfKr1Iv34UuF1FqrWDAw48 XryUGFy8uF4UXFJanT9S1TB71UUUUU7qnTZGkaVYY2UrUUUUjbIjqfuFe4nvWSU5nxnvy2 9KBjDUYxBIdaVFxhVjvjDU0xZFpf9x07jjUDJUUUUU= Sender: nzzhao@126.com X-CM-SenderInfo: xq22xtbr6rjloofrz/xtbBsAEf7mqoyKEkqQAA3m Content-Type: text/plain; charset="utf-8" With large folios, the source and destination block in __clone_blkaddrs() can belong to the same folio during same-inode operations such as collapse range and insert range. In that case, locking the source folio and then looking up the destination folio can try to lock the same folio again. It also copies from and to offset 0, which is only correct for order-0 folios. Detect the same-inode, same-folio case before looking up the destination folio. Then in this case,we reserve the destination block explicitly. Signed-off-by: Nanzhe Zhao --- fs/f2fs/file.c | 61 +++++++++++++++++++++++++++++++++++++++++++------- 1 file changed, 53 insertions(+), 8 deletions(-) diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index 54aa045b549e..a36d243ca305 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -1586,25 +1586,70 @@ static int __clone_blkaddrs(struct inode *src_inode= , struct inode *dst_inode, f2fs_put_dnode(&dn); } else { struct folio *fsrc, *fdst; + size_t src_off, dst_off; + bool same_folio; =20 fsrc =3D f2fs_get_lock_data_folio(src_inode, src + i, true); if (IS_ERR(fsrc)) return PTR_ERR(fsrc); - fdst =3D f2fs_get_new_data_folio(dst_inode, NULL, dst + i, - true); - if (IS_ERR(fdst)) { - f2fs_folio_put(fsrc, true); - return PTR_ERR(fdst); + + src_off =3D offset_in_folio(fsrc, + (loff_t)(src + i) << PAGE_SHIFT); + + /* + * For same-inode operations (collapse/insert), src and + * dst may fall within the same large folio. Detect this + * to avoid self-deadlock on folio lock. + */ + same_folio =3D src_inode =3D=3D dst_inode && + folio_contains(fsrc, dst + i); + if (same_folio) { + struct dnode_of_data dn; + + /* Reserve block for dst before copying */ + set_new_dnode(&dn, dst_inode, NULL, NULL, 0); + ret =3D f2fs_reserve_block(&dn, dst + i); + if (ret) { + f2fs_folio_put(fsrc, true); + return ret; + } + fdst =3D fsrc; + } else { + fdst =3D f2fs_get_new_data_folio(dst_inode, NULL, + dst + i, true); + if (IS_ERR(fdst)) { + f2fs_folio_put(fsrc, true); + return PTR_ERR(fdst); + } } =20 + dst_off =3D offset_in_folio(fdst, + (loff_t)(dst + i) << PAGE_SHIFT); f2fs_folio_wait_writeback(fdst, DATA, true, true); - - memcpy_folio(fdst, 0, fsrc, 0, PAGE_SIZE); + memcpy_folio(fdst, dst_off, fsrc, src_off, + PAGE_SIZE); + if (folio_test_large(fdst)) { + if (!f2fs_ffs_find_or_alloc(fdst)) { + f2fs_folio_put(fdst, true); + if (!same_folio) + f2fs_folio_put(fsrc, true); + return -ENOMEM; + } + f2fs_ffs_mark_subrange_uptodate(fdst, + dst_off, PAGE_SIZE); + f2fs_ffs_mark_subrange_dirty(fdst, dst_off, + PAGE_SIZE); + } folio_mark_dirty(fdst); + if (same_folio && i_size_read(dst_inode) < + ((loff_t)(dst + i + 1) << PAGE_SHIFT)) + f2fs_i_size_write(dst_inode, + ((loff_t)(dst + i + 1) << PAGE_SHIFT)); folio_set_f2fs_gcing(fdst); f2fs_folio_put(fdst, true); - f2fs_folio_put(fsrc, true); + if (!same_folio) + f2fs_folio_put(fsrc, true); =20 ret =3D f2fs_truncate_hole(src_inode, src + i, src + i + 1); --=20 2.43.0 From nobody Fri Sep 25 08:47:33 2026 Received: from m16.mail.126.com (m16.mail.126.com [220.197.31.9]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 0F9C535C684 for ; Tue, 15 Sep 2026 04:25:38 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=220.197.31.9 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1789446342; cv=none; b=mM2y43q5zHuj7n/0z3cvMpJGiMno22HwnOCfdus30k11p/h6MiTMqvSZ0K8JYk5yygYclE9VCn0NBkNOWUm8foL+coOQiJHshV6P0Jik66M2HuO4WH7pDKiZkqh7V6jUkFDeaZ6DyvRQR0H38G7uclQAIxuJpcOQGL+Y5ToN740= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1789446342; c=relaxed/simple; bh=JUjHJbeg+4rAQjkWQ6Gs24yHJOFAAqjZ8ww/Etfa6RE=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=pHoRjXAP7qF7+MIieRcJf0/IsUzVD0H5c3goKYwTHFiWZFyX8W0SO2kik2pGa4nmyGOiI4ORJNN+6544IYzue5HdEbHSh1PInuVJuY3EBTAMlGpLyS/gHhpv7196Ei8nW6GM/ZN9MlVohnPA3zXvjXaefXE0NPqFXqWzdSn4Ixk= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=fail (p=quarantine dis=none) header.from=xiaomi.com; spf=pass smtp.mailfrom=126.com; dkim=pass (1024-bit key) header.d=126.com header.i=@126.com header.b=BaLb4DMD; arc=none smtp.client-ip=220.197.31.9 Authentication-Results: smtp.subspace.kernel.org; dmarc=fail (p=quarantine dis=none) header.from=xiaomi.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=126.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=126.com header.i=@126.com header.b="BaLb4DMD" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=126.com; s=s110527; h=From:To:Subject:Date:Message-ID:MIME-Version; bh=vG +0jXI9nLTrSkQdUgMu6cidCTbemoJ3PlzdQyVyDPI=; b=BaLb4DMDOSn872XucB /U+MNGS/TwdLQOnSmeOYX7GNBqBw4D+0T2DD6opq8X/kH5uuwk4MevA3UdGNVGeV RNKeMqkQy+oaBd1h8NNXfvtSjAm9hGx1oSIJ5vfed9XURqNs7bkKryCHKHlY6hEA LqwxPGLJgzoNRISdPq9i0Wii8= Received: from nzzhao-ThinkCentre-M760t.mioffice.cn (unknown []) by gzga-smtp-mtada-g0-4 (Coremail) with SMTP id _____wDXG+KfyKhqTDskBQ--.23380S4; Tue, 15 Sep 2026 12:25:06 +0800 (CST) From: Nanzhe Zhao To: linux-f2fs-devel@lists.sourceforge.net, Jaegeuk Kim , Chao Yu Cc: Barry Song , Juan Yescas , Dev Jain , linux-kernel@vger.kernel.org, David Hildenbrand , Bo Zhang , Kalesh Singh , Nanzhe Zhao , Pengfei Li , Ryan Roberts , Nanzhe Zhao Subject: [PATCH v2 13/14] f2fs: allow large folio support to writeable files Date: Tue, 15 Sep 2026 12:25:02 +0800 Message-ID: <20260915042503.2912616-3-zhaonanzhe@xiaomi.com> X-Mailer: git-send-email 2.43.0 In-Reply-To: <20260915041909.2903887-1-zhaonanzhe@xiaomi.com> References: <20260915041909.2903887-1-zhaonanzhe@xiaomi.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-CM-TRANSID: _____wDXG+KfyKhqTDskBQ--.23380S4 X-Coremail-Antispam: 1Uf129KBjvJXoWxKFWfXr1DKw1xWw1rXr13Arb_yoW7XFWDpF Z8GF1rGr4DuFy5Wr1ktFn0vr1Fqa4xG3yUAa97Cw4kA3WDX34SgF1xKa45uFWUtrWkJrWI qF4jkryUWr15GFJanT9S1TB71UUUUU7qnTZGkaVYY2UrUUUUjbIjqfuFe4nvWSU5nxnvy2 9KBjDUYxBIdaVFxhVjvjDU0xZFpf9x07j5cTPUUUUU= Sender: nzzhao@126.com X-CM-SenderInfo: xq22xtbr6rjloofrz/xtbBsAIf7mqoyKIksQAA3+ Content-Type: text/plain; charset="utf-8" Now we make all write path support large folios, so we open permission to let writeable file set large folio mapping. Keep fs-layer encrypted files excluded unless inline encryption is enabled, since f2fs_encrypt_one_page() still encrypts one PAGE_SIZE page and cannot handle large folios. Inline files are allowed as well; their index #0 folio stays order-0 (see the large folio write path), so the existing inline read/convert paths are not affected and large folios are used from the second folio on. Signed-off-by: Nanzhe Zhao --- fs/f2fs/data.c | 2 +- fs/f2fs/f2fs.h | 17 +++++++++++++++++ fs/f2fs/file.c | 16 ---------------- fs/f2fs/inline.c | 6 ++++++ fs/f2fs/inode.c | 4 +--- fs/f2fs/namei.c | 1 + 6 files changed, 26 insertions(+), 20 deletions(-) diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index 295424f2ade5..c9dba8d0ad3d 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -2941,7 +2941,7 @@ static int f2fs_read_data_large_folio(struct inode *i= node, int ret =3D 0; bool folio_in_bio =3D false; =20 - if (!IS_IMMUTABLE(inode) || f2fs_compressed_file(inode)) { + if (f2fs_compressed_file(inode)) { if (folio) folio_unlock(folio); return -EOPNOTSUPP; diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h index d99a812810d3..d608376b1456 100644 --- a/fs/f2fs/f2fs.h +++ b/fs/f2fs/f2fs.h @@ -5251,6 +5251,23 @@ static inline bool f2fs_quota_file(struct f2fs_sb_in= fo *sbi, nid_t ino) return false; } =20 +static inline void f2fs_mapping_set_large_folio(struct inode *inode) +{ + if (!S_ISREG(inode->i_mode) || + f2fs_compressed_file(inode) || + f2fs_quota_file(F2FS_I_SB(inode), inode->i_ino)) + return; + + /* + * Keep the folio order range at 0 as a transitional step; large + * folios will be enabled once the whole write path is audited. + */ + if (IS_IMMUTABLE(inode)) + mapping_set_folio_min_order(inode->i_mapping, 0); + else + mapping_set_folio_order_range(inode->i_mapping, 0, 0); +} + static inline bool f2fs_block_unit_discard(struct f2fs_sb_info *sbi) { return F2FS_OPTION(sbi).discard_unit =3D=3D DISCARD_UNIT_BLOCK; diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index a36d243ca305..db71cadefabd 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -162,17 +162,6 @@ static vm_fault_t f2fs_vm_page_mkwrite(struct vm_fault= *vmf) int err =3D 0; vm_fault_t ret; =20 - /* - * We only support large folio on the read case. - * Don't make any dirty pages. - */ - if (unlikely(IS_IMMUTABLE(inode)) || - mapping_large_folio_support(inode->i_mapping)) { - f2fs_err(sbi, "Not expected: immutable: %d large_folio: %d", - IS_IMMUTABLE(inode), - mapping_large_folio_support(inode->i_mapping)); - return VM_FAULT_SIGBUS; - } =20 if (is_inode_flag_set(inode, FI_COMPRESS_RELEASED)) { err =3D -EIO; @@ -712,9 +701,6 @@ static int f2fs_file_open(struct inode *inode, struct f= ile *filp) if (!f2fs_is_compress_backend_ready(inode)) return -EOPNOTSUPP; =20 - if (mapping_large_folio_support(inode->i_mapping) && - filp->f_mode & FMODE_WRITE) - return -EOPNOTSUPP; =20 err =3D fsverity_file_open(inode, filp); if (err) @@ -1186,8 +1172,6 @@ int f2fs_setattr(struct mnt_idmap *idmap, struct dent= ry *dentry, return -EPERM; =20 if ((attr->ia_valid & ATTR_SIZE)) { - if (mapping_large_folio_support(inode->i_mapping)) - return -EOPNOTSUPP; if (IS_DEVICE_ALIASING(inode)) return -EPERM; if (!f2fs_is_compress_backend_ready(inode)) diff --git a/fs/f2fs/inline.c b/fs/f2fs/inline.c index 8db083b545bc..51cc2e6c0284 100644 --- a/fs/f2fs/inline.c +++ b/fs/f2fs/inline.c @@ -238,6 +238,12 @@ int f2fs_convert_inline_inode(struct inode *inode) if (IS_ERR(folio)) return PTR_ERR(folio); =20 + /* + * Inline data is only supported in an order-0 index #0 folio; the + * inline write and mmap fault paths never request a larger order. + */ + f2fs_bug_on(sbi, folio_test_large(folio)); + f2fs_lock_op(sbi, &lc); =20 ifolio =3D f2fs_get_inode_folio(sbi, inode->i_ino); diff --git a/fs/f2fs/inode.c b/fs/f2fs/inode.c index 24205a35be04..c8a4c62c9927 100644 --- a/fs/f2fs/inode.c +++ b/fs/f2fs/inode.c @@ -643,9 +643,7 @@ struct inode *f2fs_iget(struct super_block *sb, unsigne= d long ino) inode->i_op =3D &f2fs_file_inode_operations; inode->i_fop =3D &f2fs_file_operations; inode->i_mapping->a_ops =3D &f2fs_dblock_aops; - if (IS_IMMUTABLE(inode) && !f2fs_compressed_file(inode) && - !f2fs_quota_file(sbi, inode->i_ino)) - mapping_set_folio_min_order(inode->i_mapping, 0); + f2fs_mapping_set_large_folio(inode); } else if (S_ISDIR(inode->i_mode)) { inode->i_op =3D &f2fs_dir_inode_operations; inode->i_fop =3D &f2fs_dir_operations; diff --git a/fs/f2fs/namei.c b/fs/f2fs/namei.c index ff86ee07290d..9d0cec72aacc 100644 --- a/fs/f2fs/namei.c +++ b/fs/f2fs/namei.c @@ -341,6 +341,7 @@ static struct inode *f2fs_new_inode(struct mnt_idmap *i= dmap, f2fs_set_inode_flags(inode); =20 f2fs_init_extent_tree(inode); + f2fs_mapping_set_large_folio(inode); =20 trace_f2fs_new_inode(inode, 0); return inode; --=20 2.43.0 From nobody Fri Sep 25 08:47:33 2026 Received: from m16.mail.126.com (m16.mail.126.com [117.135.210.9]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 33B7B336883 for ; Tue, 15 Sep 2026 04:25:36 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=117.135.210.9 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1789446340; cv=none; b=bjtaBwha0JyET3+yin6xnkBKO89TRWTyZcsCpkYihJ/ibaIjPhPHtV3gI177Tb8ieBLZ8izD/vE2ML3f96l9V6JZoAbknnRXzlmupn4BCa4Qvtpv/956A0hHX6zYSJbH+5fl6UuXAXD/qHeMVNrqWcJFCt92jg+bNV4J0L1CDIA= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1789446340; c=relaxed/simple; bh=zu2r8fKg37WUcJm24sbqJsnSwlTxVzemQnIW9xyeMKE=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=eKF6UoZIN/Z0UzflcI5J9S8OTfPCFLUz/SZ5stPRZIvzhdp0toiB5I+GtEQhlI32jAUlaq1KZZiveKVxlZD3clSVG+c6/HTb9kljAcCphGyxEbZHdLVN+XTYOAh7++P3A41OEKLBqGvN9vXBjzyqLfn27M6sAxX3WjcoOAWz4gk= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=fail (p=quarantine dis=none) header.from=xiaomi.com; spf=pass smtp.mailfrom=126.com; dkim=pass (1024-bit key) header.d=126.com header.i=@126.com header.b=fDc69c98; arc=none smtp.client-ip=117.135.210.9 Authentication-Results: smtp.subspace.kernel.org; dmarc=fail (p=quarantine dis=none) header.from=xiaomi.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=126.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=126.com header.i=@126.com header.b="fDc69c98" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=126.com; s=s110527; h=From:To:Subject:Date:Message-ID:MIME-Version; bh=Ll 7cLNOzxNOjZdNKXLvuqbWxGXfC8y1PJVzt99BmzAw=; b=fDc69c98ey1yoFI5Jd hmFi+sStw1vUqj+sadIU0NywfwAx5Xts/Ab/6mUTJ2uTmxgNt2PMtX+BGjHOglhr cbmdFvVzaHB/hEaeXpfptLLJFdiXEdqpupCn+qSfhhYpdViVjV07N0VkBh6tndEK l4OU8F+ICRla9hczG5Ufvd0h8= Received: from nzzhao-ThinkCentre-M760t.mioffice.cn (unknown []) by gzga-smtp-mtada-g0-4 (Coremail) with SMTP id _____wDXG+KfyKhqTDskBQ--.23380S5; Tue, 15 Sep 2026 12:25:07 +0800 (CST) From: Nanzhe Zhao To: linux-f2fs-devel@lists.sourceforge.net, Jaegeuk Kim , Chao Yu Cc: Barry Song , Juan Yescas , Dev Jain , linux-kernel@vger.kernel.org, David Hildenbrand , Bo Zhang , Kalesh Singh , Nanzhe Zhao , Pengfei Li , Ryan Roberts , Nanzhe Zhao Subject: [PATCH v2 14/14] f2fs: make compressed files compatible with large folio Date: Tue, 15 Sep 2026 12:25:03 +0800 Message-ID: <20260915042503.2912616-4-zhaonanzhe@xiaomi.com> X-Mailer: git-send-email 2.43.0 In-Reply-To: <20260915041909.2903887-1-zhaonanzhe@xiaomi.com> References: <20260915041909.2903887-1-zhaonanzhe@xiaomi.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-CM-TRANSID: _____wDXG+KfyKhqTDskBQ--.23380S5 X-Coremail-Antispam: 1Uf129KBjvJXoWxGw45XFWUJFyrJw1fGr48Zwb_yoWrJr4fpr W5GryrKrWvqryUuwn7Aa1Ygr1ftry8G3yUZFWxKw18AF15XrnY9FWxt3Z8WFWUJry5ZF4v va1jkryUX3W5taUanT9S1TB71UUUUU7qnTZGkaVYY2UrUUUUjbIjqfuFe4nvWSU5nxnvy2 9KBjDUYxBIdaVFxhVjvjDU0xZFpf9x07jOdbbUUUUU= Sender: nzzhao@126.com X-CM-SenderInfo: xq22xtbr6rjloofrz/xtbBowMf7mqoyKN04wAA3v Content-Type: text/plain; charset="utf-8" The compression flag is the hint indicates that the inode can be compressed, when the inode is using large folio, we expected it keeps using the large folio read/write paths and its data stays uncompressed on disk until the inode is evicted and re-read Let f2fs_write_begin() skip the compression overwrite preparation for such inodes and remove the compressed-file gate in f2fs_read_data_large_folio() so the data is simply read/written as regular blocks. For the same reason, reject F2FS_IOC_COMPRESS_FILE with -EOPNOTSUPP once the inode mapping is switched to large folios. Signed-off-by: Nanzhe Zhao --- Documentation/filesystems/f2fs.rst | 10 ++++++++++ fs/f2fs/compress.c | 2 ++ fs/f2fs/data.c | 9 ++------- fs/f2fs/file.c | 7 +++++++ 4 files changed, 21 insertions(+), 7 deletions(-) diff --git a/Documentation/filesystems/f2fs.rst b/Documentation/filesystems= /f2fs.rst index 771216f45207..b4f436ffd9ac 100644 --- a/Documentation/filesystems/f2fs.rst +++ b/Documentation/filesystems/f2fs.rst @@ -941,6 +941,16 @@ Compression implementation reserved via ioctl(F2FS_IOC_RESERVE_COMPRESS_BLOCKS) or the file size is truncated to zero. =20 +- Compression and large folios are not effective at the same time on a fil= e: + a compressed inode does not use large folios, while an inode which is + using large folios keeps its data uncompressed on disk. If the compress= ion + flag is set on an inode that is already using large folios, the flag wor= ks + as a hint until the inode is evicted: the inode keeps using the large fo= lio + read/write paths, f2fs_write_begin() skips the compression overwrite + preparation, and ioctl(F2FS_IOC_COMPRESS_FILE) fails with -EOPNOTSUPP. + Once the inode is evicted and read back, it uses order-0 folios again and + compression is applied as usual. + Compress metadata layout:: =20 [Dnode Structure] diff --git a/fs/f2fs/compress.c b/fs/f2fs/compress.c index b130e43b4566..fa0f8449a12c 100644 --- a/fs/f2fs/compress.c +++ b/fs/f2fs/compress.c @@ -78,6 +78,8 @@ bool f2fs_is_compressed_page(struct folio *folio) return false; if (folio_test_f2fs_nonpointer(folio)) return false; + if (f2fs_folio_has_ffs(folio)) + return false; =20 f2fs_bug_on(F2FS_F_SB(folio), *((u32 *)folio->private) !=3D F2FS_COMPRESSED_PAGE_MAGIC); diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c index c9dba8d0ad3d..f5421334ecf3 100644 --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -2941,12 +2941,6 @@ static int f2fs_read_data_large_folio(struct inode *= inode, int ret =3D 0; bool folio_in_bio =3D false; =20 - if (f2fs_compressed_file(inode)) { - if (folio) - folio_unlock(folio); - return -EOPNOTSUPP; - } - map.m_seg_type =3D NO_CHECK_TYPE; =20 if (rac) @@ -4828,7 +4822,8 @@ static int f2fs_write_begin(const struct kiocb *iocb, } =20 #ifdef CONFIG_F2FS_FS_COMPRESSION - if (f2fs_compressed_file(inode)) { + if (f2fs_compressed_file(inode) && + !mapping_large_folio_support(inode->i_mapping)) { int ret; struct page *page; =20 diff --git a/fs/f2fs/file.c b/fs/f2fs/file.c index db71cadefabd..c5cb0b0384b6 100644 --- a/fs/f2fs/file.c +++ b/fs/f2fs/file.c @@ -5002,6 +5002,13 @@ static int f2fs_ioc_compress_file(struct file *filp) if (!(filp->f_mode & FMODE_WRITE)) return -EBADF; =20 + /* + * The mapping is already using large folios, where the data is kept + * uncompressed, so refuse to start compressing the file. + */ + if (mapping_large_folio_support(inode->i_mapping)) + return -EOPNOTSUPP; + f2fs_balance_fs(sbi, true); =20 ret =3D mnt_want_write_file(filp); --=20 2.43.0