From mboxrd@z Thu Jan 1 00:00:00 1970 From: Chao Yu Subject: RE: [f2fs-dev] [PATCH V2] f2fs: check all ones or zeros bitmap with bitops for better mount performance Date: Mon, 28 Oct 2013 15:07:11 +0800 Message-ID: <000201ced3ac$799274c0$6cb75e40$@samsung.com> References: <000401ced092$28087a90$78196fb0$@samsung.com> <003501ced384$31fb7530$95f25f90$@samsung.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: In-reply-to: <003501ced384$31fb7530$95f25f90$@samsung.com> Content-language: zh-cn Sender: linux-fsdevel-owner@vger.kernel.org To: =?utf-8?B?J+ydtOywveunjCc=?= , jaegeuk.kim@samsung.com Cc: linux-fsdevel@vger.kernel.org, =?utf-8?B?J+iwreWnnSc=?= , linux-kernel@vger.kernel.org, linux-f2fs-devel@lists.sourceforge.net List-Id: linux-f2fs-devel.lists.sourceforge.net Hi Lee, > -----Original Message----- > From: =EC=9D=B4=EC=B0=BD=EB=A7=8C [mailto:cm224.lee@samsung.com] > Sent: Monday, October 28, 2013 10:20 AM > To: 'Chao Yu'; jaegeuk.kim@samsung.com > Cc: linux-fsdevel@vger.kernel.org; '=E8=B0=AD=E5=A7=9D'; linux-kernel= @vger.kernel.org; > linux-f2fs-devel@lists.sourceforge.net > Subject: RE: [f2fs-dev] [PATCH V2] f2fs: check all ones or zeros bitm= ap with > bitops for better mount performance >=20 > To check whether bitmap are all zeros or all ones, I think memcmp is = more > neat. > But I don't know exactly performance gap between memcmp and > find_next_bit. With the result of my test, find_next_bit takes less time than memcmp. If we could use {test, set, clear}_bit_le intead of f2fs_{test, set, cl= ear}_bit, the following patch could be used for better performance. This one is better than the V2 patch for performance with mixed bitmap = in my test. diff --git a/fs/f2fs/segment.h b/fs/f2fs/segment.h index 7f94d78..32153eb --- a/fs/f2fs/segment.h +++ b/fs/f2fs/segment.h @@ -544,7 +544,8 @@ static inline void check_block_count(struct f2fs_sb= _info *sbi, struct f2fs_sm_info *sm_info =3D SM_I(sbi); unsigned int end_segno =3D sm_info->segment_count - 1; int valid_blocks =3D 0; - int i; + int i, cur_pos =3D 0, next_pos; + bool is_valid; =20 /* check segment usage */ BUG_ON(GET_SIT_VBLOCKS(raw_sit) > sbi->blocks_per_seg); @@ -553,9 +554,21 @@ static inline void check_block_count(struct f2fs_s= b_info *sbi, BUG_ON(segno > end_segno); =20 /* check bitmap with valid block count */ - for (i =3D 0; i < sbi->blocks_per_seg; i++) - if (f2fs_test_bit(i, raw_sit->valid_map)) - valid_blocks++; + cur_pos =3D 0; + is_valid =3D test_bit_le(0, raw_sit->valid_map) ? true : false; + do { + if (is_valid) { + next_pos =3D find_next_zero_bit_le(&raw_sit->va= lid_map, + sbi->blocks_per_seg, + cur_pos); + valid_blocks +=3D next_pos - cur_pos; + } else + next_pos =3D find_next_bit_le(&raw_sit->valid_m= ap, + sbi->blocks_per_seg, + cur_pos); + cur_pos =3D next_pos; + is_valid =3D !is_valid; + } while (cur_pos < sbi->blocks_per_seg); BUG_ON(GET_SIT_VBLOCKS(raw_sit) !=3D valid_blocks); } >=20 >=20 > -----Original Message----- > From: Chao Yu [mailto:chao2.yu@samsung.com] > Sent: Thursday, October 24, 2013 5:21 PM > To: jaegeuk.kim@samsung.com > Cc: linux-fsdevel@vger.kernel.org; '=E8=B0=AD=E5=A7=9D'; linux-kernel= @vger.kernel.org; > linux-f2fs-devel@lists.sourceforge.net > Subject: [f2fs-dev] [PATCH V2] f2fs: check all ones or zeros bitmap w= ith bitops > for better mount performance >=20 > Previously, check_block_count check valid_map with bit data type in c= ommon > scenario that sit has all ones or zeros bitmap, it makes low mount > performance. > So let's check the special bitmap with integer data type instead of t= he bit one. >=20 > v2: > use find_next_bit_le/find_next_zero_bit_le for better performance= and > readable as Jaegeuk suggested. >=20 > Suggested-by: Jaegeuk Kim > Signed-off-by: Tan Shu > Signed-off-by: Yu Chao > --- > fs/f2fs/segment.h | 17 +++++++++++++++++ > 1 file changed, 17 insertions(+) >=20 > diff --git a/fs/f2fs/segment.h b/fs/f2fs/segment.h index 7f94d78..d25= b6af > 100644 > --- a/fs/f2fs/segment.h > +++ b/fs/f2fs/segment.h > @@ -552,6 +552,23 @@ static inline void check_block_count(struct > f2fs_sb_info *sbi, > /* check boundary of a given segment number */ > BUG_ON(segno > end_segno); >=20 > + /* check all ones or zeros valid_map */ > + if (GET_SIT_VBLOCKS(raw_sit) =3D=3D 0) { > + int pos =3D find_next_bit_le(&raw_sit->valid_map, > + sbi->blocks_per_seg, > + 0); > + if (pos !=3D sbi->blocks_per_seg) > + BUG(); > + return; > + } else if (GET_SIT_VBLOCKS(raw_sit) =3D=3D sbi->blocks_per_seg) { > + int pos =3D find_next_zero_bit_le(&raw_sit->valid_map, > + sbi->blocks_per_seg, > + 0); > + if (pos !=3D sbi->blocks_per_seg) > + BUG(); > + return; > + } > + > /* check bitmap with valid block count */ > for (i =3D 0; i < sbi->blocks_per_seg; i++) > if (f2fs_test_bit(i, raw_sit->valid_map)) > -- > 1.7.9.5 >=20 >=20 > ---------------------------------------------------------------------= --------- > October Webinars: Code for Performance > Free Intel webinars can help you accelerate application performance. > Explore tips for MPI, OpenMP, advanced profiling, and more. Get the m= ost from > the latest Intel processors and coprocessors. See abstracts and regis= ter > > http://pubads.g.doubleclick.net/gampad/clk?id=3D60135991&iu=3D/4140/o= stg.clktr > k > _______________________________________________ > Linux-f2fs-devel mailing list > Linux-f2fs-devel@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/linux-f2fs-devel -- 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