From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from smtp.kernel.org (aws-us-west-2-korg-mail-alma10-1.taild15c8.ts.net [100.103.45.18]) (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 98BFD442397 for ; Fri, 16 Oct 2026 11:25:50 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=100.103.45.18 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1792149952; cv=none; b=uhrovgn5uuKym3VTNtLUgqNn2DCV1G5qSgrCwFY4JU4A2Dnlmw8nQdLR6dM8ltKmECoDwLf1Qb8BSkfekddlDTgMjuvmf6cBD+jO1NQrZgY8DYXZxa3308ma08XE5f9kRAwK3etUgjY33DLgXptILJxGwbxBd5xmblmlzIQnjLA= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1792149952; c=relaxed/simple; bh=5v9urK9HksbqZunJAsFacokdAo+ChxXhQSrnr1N194k=; h=Message-ID:Date:MIME-Version:Cc:Subject:To:References:From: In-Reply-To:Content-Type; b=JqOinoiy9RBpKSjC+Dm5Vo70ECGK8fIktHZNZdPfOAK9KxkMORpGRw7c2ddjJwxgxCadiFD7QUt7cYkEnPteetGehKyUCHBQWVfOVs7FxWIwKrgeHHgqRHQF8XvEuY6uVmLX22eDKabwaElvQHkuOEzKsZnsV5mMlsPYypgdCt8= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b=SKDOhBaJ; arc=none smtp.client-ip=100.103.45.18 Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b="SKDOhBaJ" Received: by smtp.kernel.org (Postfix) with ESMTPSA id AEA861F000C4; Fri, 16 Oct 2026 11:25:47 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=kernel.org; s=k20260515; t=1792149950; bh=bhTqRiskPttZ//Qh78NoZTiKwmkRVRPDuZMSTT8GKac=; h=Date:Cc:Subject:To:References:From:In-Reply-To; b=SKDOhBaJIdoNOFoY4lKUq3AnOCca4a44ZflyKjZC6Ec8mJ/+G/qiGkk5ShG6QkN5B tKOhPAcZzix8EiWi3oPgm4bNyseBNZVE54EXIkxdtbQFoRoTdVolePccbVN+5jigTL kgmFl3JIrbCPaSygqjZJQgz8ckokQ2S44b7gj1/95hvmEBr4q75h1VWZdiHgJUUd+H d2IWEFhrPvJlFtxFEfANWXp0xPqBTHk2E7bf0HdSH1Li6GUlM9cTX3aMyMAuUuei7O OyKPkKfPtk8zkViMwB4n3BpnwgSk5N6MFsnUT06FVk3LHkSmkr0tKEDTObIZjLz8gD yUan4gs1XWqmA== Message-ID: Date: Fri, 16 Oct 2026 19:25:46 +0800 Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 User-Agent: Mozilla Thunderbird Cc: chao@kernel.org, Barry Song , Nanzhe Zhao , Juan Yescas , Daeho Jeong , Dev Jain , David Hildenbrand , linux-kernel@vger.kernel.org, Bo Zhang , Kalesh Singh , Nanzhe Zhao , Pengfei Li , Ryan Roberts Subject: Re: [PATCH v5 08/16] f2fs: make GC migration large-folio aware To: Nanzhe Zhao , linux-f2fs-devel@lists.sourceforge.net, Jaegeuk Kim References: <20261014064445.2835813-1-zhaonanzhe@xiaomi.com> <20261014064445.2835813-9-zhaonanzhe@xiaomi.com> Content-Language: en-US From: Chao Yu In-Reply-To: <20261014064445.2835813-9-zhaonanzhe@xiaomi.com> Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 8bit On 10/14/26 14:44, Nanzhe Zhao wrote: > 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() 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 | 46 ++++++++++++++++++++++++++++++---------------- > fs/f2fs/f2fs.h | 1 + > fs/f2fs/gc.c | 36 ++++++++++++++++++++++++++++++++---- > fs/f2fs/segment.c | 2 +- > 4 files changed, 64 insertions(+), 21 deletions(-) > > diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c > index b5c048134e00..710213fea7a6 100644 > --- a/fs/f2fs/data.c > +++ b/fs/f2fs/data.c > @@ -1633,20 +1633,16 @@ 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 = f2fs_grab_cache_folio(mapping, index, for_write); > if (IS_ERR(folio)) > return folio; > > if (folio_test_large(folio)) { > - pgoff_t folio_index = mapping_align_index(mapping, index); > - unsigned long nr_pages = 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 = -ENOMEM; > + goto put_err; > + } > } > > if (f2fs_lookup_read_extent_cache_block(inode, index, > @@ -1682,7 +1678,7 @@ 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; > } > @@ -1695,9 +1691,11 @@ struct folio *f2fs_get_read_data_folio(struct inode *inode, pgoff_t index, > * f2fs_init_inode_metadata. > */ > if (dn.data_blkaddr == NEW_ADDR) { > - folio_zero_segment(folio, 0, folio_size(folio)); > - if (!folio_test_uptodate(folio)) > - folio_mark_uptodate(folio); > + size_t offset = 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; > } > @@ -1757,7 +1755,8 @@ struct folio *f2fs_get_lock_data_folio(struct inode *inode, pgoff_t index, > > /* wait for read completion */ > folio_lock(folio); > - if (unlikely(folio->mapping != mapping || !folio_test_uptodate(folio))) { > + if (unlikely(folio->mapping != mapping || > + !f2fs_ffs_test_blk_uptodate(folio, index))) { > f2fs_folio_put(folio, true); > return ERR_PTR(-EIO); > } > @@ -2906,6 +2905,22 @@ static void f2fs_ffs_mark_subrange_uptodate(struct folio *folio, size_t offset, > folio_mark_uptodate(folio); > } > > +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 = (struct f2fs_folio_state *)folio->private; > + nr_subpages = folio_nr_pages(folio); > + idx = index - folio->index; > + return test_bit(nr_subpages + idx, ffs->state); > +} > + > static unsigned int __ffs_mark_subrange_dirty(struct folio *folio, > struct f2fs_folio_state *ffs, size_t offset, size_t len) > { > @@ -2999,7 +3014,6 @@ bool f2fs_ffs_clear_subrange_dirty_and_test(struct folio *folio, size_t offset, > > return dirty; > } > - > static unsigned int ffs_next_dirty_subpage(struct f2fs_folio_state *ffs, > const struct folio *folio, unsigned int start, > unsigned int end) > @@ -3497,7 +3511,7 @@ bool f2fs_should_update_outplace(struct inode *inode, struct f2fs_io_info *fio) > return true; > > 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))) > diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h > index d976e6ad058b..36ec8266af69 100644 > --- a/fs/f2fs/f2fs.h > +++ b/fs/f2fs/f2fs.h > @@ -4493,6 +4493,7 @@ 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 offset, > size_t len); > diff --git a/fs/f2fs/gc.c b/fs/f2fs/gc.c > index d7fae1890835..5bb117e5e811 100644 > --- a/fs/f2fs/gc.c > +++ b/fs/f2fs/gc.c > @@ -1512,12 +1512,19 @@ static int move_data_page(struct inode *inode, block_t bidx, int gc_type, > unsigned int segno, int off) > { > struct folio *folio; > + size_t foff = 0; > + bool large = false; > int err = 0; > > folio = f2fs_get_lock_data_folio(inode, bidx, true); > if (IS_ERR(folio)) > return PTR_ERR(folio); > > + if (f2fs_folio_has_ffs(folio)) { > + large = true; > + foff = offset_in_folio(folio, (loff_t)bidx << PAGE_SHIFT); > + } > + > if (!check_valid_map(F2FS_I_SB(inode), segno, off)) { > err = -ENOENT; > goto out; > @@ -1532,6 +1539,8 @@ static int move_data_page(struct inode *inode, block_t bidx, int gc_type, > err = -EAGAIN; > goto out; > } > + if (large) > + f2fs_ffs_mark_subrange_dirty(folio, foff, PAGE_SIZE); > folio_mark_dirty(folio); > folio_set_f2fs_gcing(folio); > } else { > @@ -1544,32 +1553,51 @@ static int move_data_page(struct inode *inode, block_t bidx, int gc_type, > .op_flags = REQ_SYNC, > .old_blkaddr = NULL_ADDR, > .folio = folio, > + .folio_offset = bidx - folio->index, > + .folio_blkcnt = 1, > .encrypted_page = NULL, > .need_lock = LOCK_REQ, > .io_type = FS_GC_DATA_IO, > }; > - bool is_dirty = folio_test_dirty(folio); > + struct f2fs_folio_state *ffs = NULL; > + bool is_dirty = f2fs_ffs_test_blk_dirty(folio, bidx); > > retry: > f2fs_folio_wait_writeback(folio, true, true); > > + if (large) { > + ffs = (struct f2fs_folio_state *)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_block(inode); > - f2fs_remove_dirty_inode(inode); > + if (!large) { > + inode_dec_dirty_block(inode); > + f2fs_remove_dirty_inode(inode); > + } else if (f2fs_ffs_clear_subrange_dirty_and_test(folio, > + foff, PAGE_SIZE)) { > + folio_mark_dirty(folio); > + } > } > > + if (large) > + atomic_inc(&ffs->write_pages_pending); > folio_set_f2fs_gcing(folio); > > err = f2fs_do_write_data_page(&fio); Nanzhe, There is valid bug report from sashiko as below, I checked a bit w/ llm: *Normal fsync() Writeback Flow* 1. Buffered write dirties a folio: • folio_mark_dirty(folio) → filemap_dirty_folio() • Sets folio flag PG_dirty. • Sets XArray (radix-tree) mark PAGECACHE_TAG_DIRTY on the folio entry in mapping->i_pages. • (State: PG_dirty = 1, TAG_DIRTY = 1, TAG_TOWRITE = 0, TAG_WRITEBACK = 0) 2. fsync() starts and snapshots dirty folios (WB_SYNC_ALL): • f2fs_do_sync_file() → file_write_and_wait_range() → f2fs_write_cache_folios() → writeback_iter() • Calls tag_pages_for_writeback(), which scans mapping->i_pages for entries with PAGECACHE_TAG_DIRTY and tags them with PAGECACHE_TAG_TOWRITE. • (State: PG_dirty = 1, TAG_DIRTY = 1, TAG_TOWRITE = 1, TAG_WRITEBACK = 0) 3. writeback_iter() fetches the folio for writeout: • writeback_get_folio() looks up folios tagged with PAGECACHE_TAG_TOWRITE, locks the folio, and calls folio_clear_dirty_for_io(folio). • Clears PG_dirty on the folio, while keeping PAGECACHE_TAG_DIRTY and PAGECACHE_TAG_TOWRITE set in mapping->i_pages until writeback actually starts. • (State: PG_dirty = 0, TAG_DIRTY = 1, TAG_TOWRITE = 1, TAG_WRITEBACK = 0) 4. Filesystem starts writeback on the whole folio: • f2fs_write_cache_folios() calls folio_start_writeback(folio) → __folio_start_writeback(folio, false) and submits all dirty subpages in the folio. • Sets PG_writeback and PAGECACHE_TAG_WRITEBACK. • Because !folio_test_dirty(folio) is true, clears PAGECACHE_TAG_DIRTY. • Because keep_write == false, clears PAGECACHE_TAG_TOWRITE. • (State: PG_dirty = 0, TAG_DIRTY = 0, TAG_TOWRITE = 0, TAG_WRITEBACK = 1) 5. I/O completion & fsync() wait: • filemap_fdatawait_range() waits on PAGECACHE_TAG_WRITEBACK. • f2fs_write_end_bio() calls folio_end_writeback(folio), clearing PG_writeback and PAGECACHE_TAG_WRITEBACK. *Buggy Race Flow with move_data_page() on a Large Folio* Suppose a large folio has subpage 0 (clean, in a GC victim segment) and subpage 1 (dirty from a buffered write). Thread A (fsync) Thread B (FG_GC: move_data_page on subpage 0) -------------------------------------------- ----------------------------------------------------------- 1. Buffered write dirtied subpage 1: - ffs->state[dirty, 1] = 1 - folio_mark_dirty(folio) -> sets PG_dirty = 1 -> sets XArray PAGECACHE_TAG_DIRTY = 1 2. f2fs_do_sync_file() -> file_write_and_wait_range() -> f2fs_write_cache_folios() -> writeback_iter(): tag_pages_for_writeback() tags folio in XArray with PAGECACHE_TAG_TOWRITE = 1 [PG_dirty=1, TAG_DIRTY=1, TAG_TOWRITE=1] 3. move_data_page(inode, bidx=0, FG_GC) locks folio: - f2fs_ffs_mark_subrange_dirty(folio, subpage 0) - folio_mark_dirty(folio) - folio_clear_dirty_for_io(folio) -> clears PG_dirty = 0 - f2fs_ffs_clear_subrange_dirty_and_test(subpage 0) -> clears ffs->state[dirty, 0] = 0 -> sees ffs->state[dirty, 1] == 1 (still dirty!) -> calls folio_mark_dirty(folio): sets PG_dirty = 1, TAG_DIRTY = 1 4. Calls f2fs_do_write_data_page(&fio) for subpage 0 ONLY: - Calls folio_start_writeback(folio) -> __folio_start_writeback(folio, false): * sets PG_writeback = 1, TAG_WRITEBACK = 1 * folio_test_dirty(folio) is true -> keeps TAG_DIRTY = 1 * keep_write is false -> CLEARS PAGECACHE_TAG_TOWRITE = 0! - Submits bio for subpage 0 only; unlocks folio. [PG_dirty=1, TAG_DIRTY=1, TAG_TOWRITE=0, TAG_WRITEBACK=1] 5. writeback_get_folio() searches XArray for PAGECACHE_TAG_TOWRITE: -> Folio no longer has PAGECACHE_TAG_TOWRITE! -> writeback_iter() completely skips folio; subpage 1 is NEVER submitted for writeback. 6. filemap_fdatawait_range() waits for PAGECACHE_TAG_WRITEBACK: -> Subpage 0 bio finishes -> clears TAG_WRITEBACK. -> fsync() returns 0 (success) while subpage 1 is still dirty in RAM! *Potential Solution* We need to call __folio_start_writeback(folio, folio_test_dirty(folio)) to keep PAGECACHE_TAG_TOWRITE tag in folio xarry entry if there are still dirty page{,s} in large folio. --- a/fs/f2fs/data.c +++ b/fs/f2fs/data.c @@ -3761,7 +3761,7 @@ int f2fs_do_write_data_page(struct f2fs_io_info *fio) if (ipu_force || (__is_valid_data_blkaddr(fio->old_blkaddr) && need_inplace_update(fio))) { - folio_start_writeback(folio); + __folio_start_writeback(folio, folio_test_dirty(folio)); f2fs_put_dnode(&dn); if (fio->need_lock == LOCK_REQ) f2fs_unlock_op(fio->sbi, &lc); @@ -3788,7 +3788,7 @@ int f2fs_do_write_data_page(struct f2fs_io_info *fio) fio->version = ni.version; - folio_start_writeback(folio); + __folio_start_writeback(folio, folio_test_dirty(folio)); if (fio->compr_blocks && fio->old_blkaddr == COMPRESS_ADDR) f2fs_i_compr_blocks_update(inode, fio->compr_blocks - 1, false); Thanks, > if (err) { > folio_clear_f2fs_gcing(folio); > + if (large) > + atomic_dec(&ffs->write_pages_pending); > if (err == -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 96bfb7d63692..06fb92fc8479 100644 > --- a/fs/f2fs/segment.c > +++ b/fs/f2fs/segment.c > @@ -3863,7 +3863,7 @@ 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; > > - if (page_private_gcing(fio->page)) { > + if (folio_test_f2fs_gcing(fio->folio)) { > if (fio->sbi->am.atgc_enabled && > (fio->io_type == FS_DATA_IO) && > (fio->sbi->gc_mode != GC_URGENT_HIGH) &&