From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from smtp.kernel.org (aws-us-west-2-korg-mail-1.web.codeaurora.org [10.30.226.201]) (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 3868F2D8DA3 for ; Mon, 13 Apr 2026 11:35:46 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=10.30.226.201 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1776080146; cv=none; b=lsonEhVJpYT3NkXO6VQFCTpNzLW7wEPJIawzrM9nBE54bV5g8pi5SRBdnJk1Y82XyjalhCas9g0IVc1LEFEjX2vugXQFxOygO3glmazwxaxLio042w5RJBewAsIyyZSaeeYNV/+OFYD96FJpxY5JF2P3qfHz38czjQLzE0+fMjM= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1776080146; c=relaxed/simple; bh=gZpu/c5zzguUdsoAwuODrkDtYotcRT8T5XjV2Ca7kC8=; h=Message-ID:Date:MIME-Version:Cc:Subject:To:References:From: In-Reply-To:Content-Type; b=LulXo51BWCdK/2qGdZza0tazJShuDjFZmA+m6MYDhQ6DfVzqLjTxmosoepLDeWXbtsFh7rryfcqBd4Grxg5ZO1+rYLvHcxtRJ1kGB6nLNLuGiQ+kur+l2yL+N6MlsJYFFbkazMKyPG/2HGMA1OBYuwUw1fhrg9ErHD4sqJHpxaI= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b=G1uYNsXm; arc=none smtp.client-ip=10.30.226.201 Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b="G1uYNsXm" Received: by smtp.kernel.org (Postfix) with ESMTPSA id A8660C116C6; Mon, 13 Apr 2026 11:35:44 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=k20201202; t=1776080146; bh=gZpu/c5zzguUdsoAwuODrkDtYotcRT8T5XjV2Ca7kC8=; h=Date:Cc:Subject:To:References:From:In-Reply-To:From; b=G1uYNsXmaq5U6HI7jYvlX5siiWGOODHi4/m2LEib2vQcKVDjshwspmWNx3yO1TZZ/ 0lqp3Xqr9g33+P3IG2e6GciNg6X4lIDDf6YLSg3j5skIe9SIje4I46artHCeFSchKP 7VdNViDxJbLbIOAiwHHZ65KPjp+q+UJx9PJ5X9LFi2rgLlNGkbnXAf9vQ8o2ttTUbK En6aevWPHeDCujpMP/TFHahNziIGgZyRZDDoSUjEMsMh/3gO0IAKtOMnDg1odWSlq5 qiTt6ZQSiRJ1JCxWQfLW+XDzT8GQdnnRwTLyl/b2kuIJF3FoQdPqwcYD6ofq+aQ3fk qe/ri0mezop3A== Message-ID: <2ad55daf-5a1c-4777-8a06-689c554f4f5b@kernel.org> Date: Mon, 13 Apr 2026 19:35:44 +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, linux-f2fs-devel@lists.sourceforge.net, linux-kernel@vger.kernel.org Subject: Re: [PATCH] f2fs: add page-order information for large folio reads in iostat To: Daniel Lee , Jaegeuk Kim References: <20260408060840.316332-1-chullee@google.com> Content-Language: en-US From: Chao Yu In-Reply-To: <20260408060840.316332-1-chullee@google.com> Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit On 4/8/2026 2:08 PM, Daniel Lee wrote: > Track read folio counts by order in F2FS iostat sysfs and tracepoints. > > Signed-off-by: Daniel Lee > --- > fs/f2fs/data.c | 4 ++++ > fs/f2fs/f2fs.h | 3 +++ > fs/f2fs/iostat.c | 38 ++++++++++++++++++++++++++++++++++++- > fs/f2fs/iostat.h | 4 ++++ > include/trace/events/f2fs.h | 21 ++++++++++++++++---- > 5 files changed, 65 insertions(+), 5 deletions(-) > > diff --git a/fs/f2fs/data.c b/fs/f2fs/data.c > index a210a7a627c6..965d4e6443c6 100644 > --- a/fs/f2fs/data.c > +++ b/fs/f2fs/data.c > @@ -2508,6 +2508,8 @@ static int f2fs_read_data_large_folio(struct inode *inode, > if (!folio) > goto out; > > + f2fs_update_read_folio_count(F2FS_I_SB(inode), folio); > + > folio_in_bio = false; > index = folio->index; > offset = 0; > @@ -2682,6 +2684,8 @@ static int f2fs_mpage_readpages(struct inode *inode, struct fsverity_info *vi, > prefetchw(&folio->flags); > } > > + f2fs_update_read_folio_count(F2FS_I_SB(inode), folio); > + > #ifdef CONFIG_F2FS_FS_COMPRESSION > index = folio->index; > > diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h > index 56c4af4b1737..e40b6b2784ee 100644 > --- a/fs/f2fs/f2fs.h > +++ b/fs/f2fs/f2fs.h > @@ -10,6 +10,7 @@ > > #include > #include > +#include > #include > #include > #include > @@ -2034,6 +2035,8 @@ struct f2fs_sb_info { > unsigned long long iostat_count[NR_IO_TYPE]; > unsigned long long iostat_bytes[NR_IO_TYPE]; > unsigned long long prev_iostat_bytes[NR_IO_TYPE]; > + unsigned long long iostat_read_folio_count[NR_PAGE_ORDERS]; > + unsigned long long prev_iostat_read_folio_count[NR_PAGE_ORDERS]; > bool iostat_enable; > unsigned long iostat_next_period; > unsigned int iostat_period_ms; > diff --git a/fs/f2fs/iostat.c b/fs/f2fs/iostat.c > index f8703038e1d8..ae265e3e9b2c 100644 > --- a/fs/f2fs/iostat.c > +++ b/fs/f2fs/iostat.c > @@ -34,6 +34,7 @@ int __maybe_unused iostat_info_seq_show(struct seq_file *seq, void *offset) > { > struct super_block *sb = seq->private; > struct f2fs_sb_info *sbi = F2FS_SB(sb); > + int i; > > if (!sbi->iostat_enable) > return 0; > @@ -76,6 +77,12 @@ int __maybe_unused iostat_info_seq_show(struct seq_file *seq, void *offset) > IOSTAT_INFO_SHOW("fs node", FS_NODE_READ_IO); > IOSTAT_INFO_SHOW("fs meta", FS_META_READ_IO); > > + /* print read folio order stats */ > + seq_printf(seq, "%-23s", "fs read folio order:"); > + for (i = 0; i < NR_PAGE_ORDERS; i++) > + seq_printf(seq, " %llu", sbi->iostat_read_folio_count[i]); > + seq_putc(seq, '\n'); > + > /* print other IOs */ > seq_puts(seq, "[OTHER]\n"); > IOSTAT_INFO_SHOW("fs discard", FS_DISCARD_IO); > @@ -113,6 +120,7 @@ static inline void __record_iostat_latency(struct f2fs_sb_info *sbi) > static inline void f2fs_record_iostat(struct f2fs_sb_info *sbi) > { > unsigned long long iostat_diff[NR_IO_TYPE]; > + unsigned long long read_folio_count_diff[NR_PAGE_ORDERS]; > int i; > unsigned long flags; > > @@ -133,9 +141,15 @@ static inline void f2fs_record_iostat(struct f2fs_sb_info *sbi) > sbi->prev_iostat_bytes[i]; > sbi->prev_iostat_bytes[i] = sbi->iostat_bytes[i]; > } > + > + for (i = 0; i < NR_PAGE_ORDERS; i++) { > + read_folio_count_diff[i] = sbi->iostat_read_folio_count[i] - > + sbi->prev_iostat_read_folio_count[i]; > + sbi->prev_iostat_read_folio_count[i] = sbi->iostat_read_folio_count[i]; > + } > spin_unlock_irqrestore(&sbi->iostat_lock, flags); > > - trace_f2fs_iostat(sbi, iostat_diff); > + trace_f2fs_iostat(sbi, iostat_diff, read_folio_count_diff); > > __record_iostat_latency(sbi); > } > @@ -151,6 +165,10 @@ void f2fs_reset_iostat(struct f2fs_sb_info *sbi) > sbi->iostat_bytes[i] = 0; > sbi->prev_iostat_bytes[i] = 0; > } > + for (i = 0; i < NR_PAGE_ORDERS; i++) { > + sbi->iostat_read_folio_count[i] = 0; > + sbi->prev_iostat_read_folio_count[i] = 0; > + } > spin_unlock_irq(&sbi->iostat_lock); > > spin_lock_irq(&sbi->iostat_lat_lock); > @@ -165,6 +183,24 @@ static inline void __f2fs_update_iostat(struct f2fs_sb_info *sbi, > sbi->iostat_count[type]++; > } > > +void f2fs_update_read_folio_count(struct f2fs_sb_info *sbi, struct folio *folio) > +{ > + unsigned int order = folio_order(folio); > + unsigned long flags; > + > + if (!sbi->iostat_enable) > + return; > + > + if (order >= NR_PAGE_ORDERS) > + order = NR_PAGE_ORDERS - 1; > + > + spin_lock_irqsave(&sbi->iostat_lock, flags); > + sbi->iostat_read_folio_count[order]++; > + spin_unlock_irqrestore(&sbi->iostat_lock, flags); > + > + f2fs_record_iostat(sbi); > +} > + > void f2fs_update_iostat(struct f2fs_sb_info *sbi, struct inode *inode, > enum iostat_type type, unsigned long long io_bytes) > { > diff --git a/fs/f2fs/iostat.h b/fs/f2fs/iostat.h > index eb99d05cf272..2025225b5bed 100644 > --- a/fs/f2fs/iostat.h > +++ b/fs/f2fs/iostat.h > @@ -34,6 +34,8 @@ extern int __maybe_unused iostat_info_seq_show(struct seq_file *seq, > extern void f2fs_reset_iostat(struct f2fs_sb_info *sbi); > extern void f2fs_update_iostat(struct f2fs_sb_info *sbi, struct inode *inode, > enum iostat_type type, unsigned long long io_bytes); > +extern void f2fs_update_read_folio_count(struct f2fs_sb_info *sbi, > + struct folio *folio); > > struct bio_iostat_ctx { > struct f2fs_sb_info *sbi; > @@ -68,6 +70,8 @@ extern void f2fs_destroy_iostat(struct f2fs_sb_info *sbi); > #else > static inline void f2fs_update_iostat(struct f2fs_sb_info *sbi, struct inode *inode, > enum iostat_type type, unsigned long long io_bytes) {} > +static inline void f2fs_update_read_folio_count(struct f2fs_sb_info *sbi, > + struct folio *folio) {} > static inline void iostat_update_and_unbind_ctx(struct bio *bio) {} > static inline void iostat_alloc_and_bind_ctx(struct f2fs_sb_info *sbi, > struct bio *bio, struct bio_post_read_ctx *ctx) {} > diff --git a/include/trace/events/f2fs.h b/include/trace/events/f2fs.h > index 9364e6775562..49c6cbb6f989 100644 > --- a/include/trace/events/f2fs.h > +++ b/include/trace/events/f2fs.h > @@ -2116,9 +2116,10 @@ DEFINE_EVENT(f2fs_zip_end, f2fs_decompress_pages_end, > #ifdef CONFIG_F2FS_IOSTAT > TRACE_EVENT(f2fs_iostat, > > - TP_PROTO(struct f2fs_sb_info *sbi, unsigned long long *iostat), > + TP_PROTO(struct f2fs_sb_info *sbi, unsigned long long *iostat, > + unsigned long long *read_folio_count), > > - TP_ARGS(sbi, iostat), > + TP_ARGS(sbi, iostat, read_folio_count), > > TP_STRUCT__entry( > __field(dev_t, dev) > @@ -2150,6 +2151,7 @@ TRACE_EVENT(f2fs_iostat, > __field(unsigned long long, fs_mrio) > __field(unsigned long long, fs_discard) > __field(unsigned long long, fs_reset_zone) > + __array(unsigned long long, read_folio_count, 16) We only print from read_folio_count[0] to read_folio_count[10], so is __array(unsigned long long, read_folio_count, 11) enough? Thanks, > ), > > TP_fast_assign( > @@ -2182,6 +2184,9 @@ TRACE_EVENT(f2fs_iostat, > __entry->fs_mrio = iostat[FS_META_READ_IO]; > __entry->fs_discard = iostat[FS_DISCARD_IO]; > __entry->fs_reset_zone = iostat[FS_ZONE_RESET_IO]; > + memset(__entry->read_folio_count, 0, sizeof(__entry->read_folio_count)); > + memcpy(__entry->read_folio_count, read_folio_count, > + sizeof(unsigned long long) * min_t(int, NR_PAGE_ORDERS, 16)); > ), > > TP_printk("dev = (%d,%d), " > @@ -2194,7 +2199,9 @@ TRACE_EVENT(f2fs_iostat, > "app [read=%llu (direct=%llu, buffered=%llu), mapped=%llu], " > "compr(buffered=%llu, mapped=%llu)], " > "fs [data=%llu, (gc_data=%llu, cdata=%llu), " > - "node=%llu, meta=%llu]", > + "node=%llu, meta=%llu], " > + "read_folio_count [0=%llu, 1=%llu, 2=%llu, 3=%llu, 4=%llu, " > + "5=%llu, 6=%llu, 7=%llu, 8=%llu, 9=%llu, 10=%llu]", > show_dev(__entry->dev), __entry->app_wio, __entry->app_dio, > __entry->app_bio, __entry->app_mio, __entry->app_bcdio, > __entry->app_mcdio, __entry->fs_dio, __entry->fs_cdio, > @@ -2205,7 +2212,13 @@ TRACE_EVENT(f2fs_iostat, > __entry->app_rio, __entry->app_drio, __entry->app_brio, > __entry->app_mrio, __entry->app_bcrio, __entry->app_mcrio, > __entry->fs_drio, __entry->fs_gdrio, > - __entry->fs_cdrio, __entry->fs_nrio, __entry->fs_mrio) > + __entry->fs_cdrio, __entry->fs_nrio, __entry->fs_mrio, > + __entry->read_folio_count[0], __entry->read_folio_count[1], > + __entry->read_folio_count[2], __entry->read_folio_count[3], > + __entry->read_folio_count[4], __entry->read_folio_count[5], > + __entry->read_folio_count[6], __entry->read_folio_count[7], > + __entry->read_folio_count[8], __entry->read_folio_count[9], > + __entry->read_folio_count[10]) > ); > > #ifndef __F2FS_IOSTAT_LATENCY_TYPE