Received-SPF: pass (google.com: domain of linux-ext4-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Content-Type: text/plain;
        charset=us-ascii
Mime-Version: 1.0 (Mac OS X Mail 12.4 \(3445.104.15\))
Subject: Re: [PATCH 2/4] ext4: skip non-loaded groups at cr=0/1 when scanning
 for good groups
From:   =?utf-8?B?0JHQu9Cw0LPQvtC00LDRgNC10L3QutC+INCQ0YDRgtGR0Lw=?= 
        <artem.blagodarenko@gmail.com>
In-Reply-To: <20200717155352.1053040-3-tytso@mit.edu>
Date:   Fri, 24 Jul 2020 14:27:48 +0300
Cc:     Ext4 Developers List <linux-ext4@vger.kernel.org>,
        Alex Zhuravlev <bzzz@whamcloud.com>,
        Alex Zhuravlev <azhuravlev@whamcloud.com>,
        Andreas Dilger <adilger@whamcloud.com>
Content-Transfer-Encoding: quoted-printable
Message-Id: <81448013-525F-4C51-A9F3-88969CB49B70@gmail.com>
References: <20200717155352.1053040-1-tytso@mit.edu>
 <20200717155352.1053040-3-tytso@mit.edu>
To:     Theodore Ts'o <tytso@mit.edu>
Sender: linux-ext4-owner@vger.kernel.org
Precedence: bulk

Looks good.

Reviewed-by: Artem Blagodarenko <artem.blagodarenko@gmail.com>

> On 17 Jul 2020, at 18:53, Theodore Ts'o <tytso@mit.edu> wrote:
>=20
> From: Alex Zhuravlev <azhuravlev@whamcloud.com>
>=20
> cr=3D0 is supposed to be an optimization to save CPU cycles, but if
> buddy data (in memory) is not initialized then all this makes no sense
> as we have to do sync IO taking a lot of cycles.  also, at cr=3D0
> mballoc doesn't store any avaibale chunk. cr=3D1 also skips groups =
using
> heuristic based on avg. fragment size. it's more useful to skip such
> groups and switch to cr=3D2 where groups will be scanned for available
> chunks.
>=20
> using sparse image and dm-slow virtual device of 120TB was
> simulated. then the image was formatted and filled using debugfs to
> mark ~85% of available space as busy.  mount process w/o the patch
> couldn't complete in half an hour (according to vmstat it would take
> ~10-11 hours).  With the patch applied mount took ~20 seconds.
>=20
> Lustre-bug-id: https://jira.whamcloud.com/browse/LU-12988
> Signed-off-by: Alex Zhuravlev <bzzz@whamcloud.com>
> Reviewed-by: Andreas Dilger <adilger@whamcloud.com>
> ---
> fs/ext4/mballoc.c | 13 ++++++++++++-
> 1 file changed, 12 insertions(+), 1 deletion(-)
>=20
> diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c
> index 8a1e6e03c088..172994349bf6 100644
> --- a/fs/ext4/mballoc.c
> +++ b/fs/ext4/mballoc.c
> @@ -2195,7 +2195,18 @@ static int ext4_mb_good_group_nolock(struct =
ext4_allocation_context *ac,
>=20
> 	/* We only do this if the grp has never been initialized */
> 	if (unlikely(EXT4_MB_GRP_NEED_INIT(grp))) {
> -		ret =3D ext4_mb_init_group(ac->ac_sb, group, GFP_NOFS);
> +		struct ext4_group_desc *gdp =3D ext4_get_group_desc(sb, =
group,
> +								  NULL);
> +		int ret;
> +
> +		/* cr=3D0/1 is a very optimistic search to find large
> +		 * good chunks almost for free. if buddy data is
> +		 * not ready, then this optimization makes no sense */
> +		if (cr < 2 &&
> +		    !(ext4_has_group_desc_csum(sb) &&
> +		      (gdp->bg_flags & =
cpu_to_le16(EXT4_BG_BLOCK_UNINIT))))
> +			return 0;
> +		ret =3D ext4_mb_init_group(sb, group, GFP_NOFS);
> 		if (ret)
> 			return ret;
> 	}
> --=20
> 2.24.1
>=20