From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jaegeuk Kim Subject: Re: [f2fs-dev] [PATCH] f2fs: split sbi->write_mutex for DATA/NODE/META to avoid unnecessary race Date: Tue, 19 Nov 2013 12:18:16 +0900 Message-ID: <1384831096.26319.25.camel@kjgkr> References: <001201cee4c8$de450360$9acf0a20$@samsung.com> Reply-To: jaegeuk.kim@samsung.com Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org, linux-f2fs-devel@lists.sourceforge.net, =?UTF-8?Q?=E8=B0=AD=E5=A7=9D?= To: Chao Yu Return-path: Received: from mailout3.samsung.com ([203.254.224.33]:9898 "EHLO mailout3.samsung.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751503Ab3KSDTI convert rfc822-to-8bit (ORCPT ); Mon, 18 Nov 2013 22:19:08 -0500 In-reply-to: <001201cee4c8$de450360$9acf0a20$@samsung.com> Sender: linux-fsdevel-owner@vger.kernel.org List-ID: Hi, 2013-11-19 (=ED=99=94), 09:43 +0800, Chao Yu: > All DATA/NODE/META bio buffers in superblock is protected by=20 > 'sbi->write_mutex', but each bio buffer area is independent, So we=20 > should split write_mutex to three for DATA/NODE/META. Agreed, one comment below though. Anyway I'll send v2. Thanks, :) >=20 > Signed-off-by: Chao Yu > --- > fs/f2fs/f2fs.h | 2 +- > fs/f2fs/segment.c | 8 ++++---- > fs/f2fs/super.c | 4 +++- > 3 files changed, 8 insertions(+), 6 deletions(-) >=20 > diff --git a/fs/f2fs/f2fs.h b/fs/f2fs/f2fs.h > index 2df1e61..1c67521 100644 > --- a/fs/f2fs/f2fs.h > +++ b/fs/f2fs/f2fs.h > @@ -362,7 +362,7 @@ struct f2fs_sb_info { > struct f2fs_sm_info *sm_info; /* segment manager */ > struct bio *bio[NR_PAGE_TYPE]; /* bios to merge */ > sector_t last_block_in_bio[NR_PAGE_TYPE]; /* last block number */ > - struct mutex write_mutex; /* mutex for writing IOs */ > + struct mutex write_mutex[NR_PAGE_TYPE]; /* mutex for writing IOs */ > =20 > /* for checkpoint */ > struct f2fs_checkpoint *ckpt; /* raw checkpoint pointer */ > diff --git a/fs/f2fs/segment.c b/fs/f2fs/segment.c > index 556965f..a36713ae 100644 > --- a/fs/f2fs/segment.c > +++ b/fs/f2fs/segment.c > @@ -661,9 +661,9 @@ void f2fs_submit_bio(struct f2fs_sb_info *sbi, en= um > page_type type, bool sync) > if (!sbi->bio[btype]) > return; > =20 > - mutex_lock(&sbi->write_mutex); > + mutex_lock(&sbi->write_mutex[type]); btype instead of type. > do_submit_bio(sbi, type, sync); > - mutex_unlock(&sbi->write_mutex); > + mutex_unlock(&sbi->write_mutex[type]); > } > =20 > static void submit_write_page(struct f2fs_sb_info *sbi, struct page = *page, > @@ -674,7 +674,7 @@ static void submit_write_page(struct f2fs_sb_info= *sbi, > struct page *page, > =20 > verify_block_addr(sbi, blk_addr); > =20 > - mutex_lock(&sbi->write_mutex); > + mutex_lock(&sbi->write_mutex[type]); > =20 > inc_page_count(sbi, F2FS_WRITEBACK); > =20 > @@ -709,7 +709,7 @@ retry: > =20 > sbi->last_block_in_bio[type] =3D blk_addr; > =20 > - mutex_unlock(&sbi->write_mutex); > + mutex_unlock(&sbi->write_mutex[type]); > trace_f2fs_submit_write_page(page, blk_addr, type); > } > =20 > diff --git a/fs/f2fs/super.c b/fs/f2fs/super.c > index f56941c..6928c0a 100644 > --- a/fs/f2fs/super.c > +++ b/fs/f2fs/super.c > @@ -874,7 +874,9 @@ static int f2fs_fill_super(struct super_block *sb= , void > *data, int silent) > mutex_init(&sbi->node_write); > sbi->por_doing =3D false; > spin_lock_init(&sbi->stat_lock); > - mutex_init(&sbi->write_mutex); > + mutex_init(&sbi->write_mutex[DATA]); > + mutex_init(&sbi->write_mutex[NODE]); > + mutex_init(&sbi->write_mutex[META]); > init_rwsem(&sbi->cp_rwsem); > init_waitqueue_head(&sbi->cp_wait); > init_sb_info(sbi); --=20 Jaegeuk Kim Samsung -- To unsubscribe from this list: send the line "unsubscribe linux-fsdevel= " in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html