Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
From:   Naohiro Aota <Naohiro.Aota@wdc.com>
To:     Anand Jain <anand.jain@oracle.com>,
        "linux-btrfs@vger.kernel.org" <linux-btrfs@vger.kernel.org>,
        David Sterba <dsterba@suse.com>
CC:     Chris Mason <clm@fb.com>, Josef Bacik <josef@toxicpanda.com>,
        Qu Wenruo <wqu@suse.com>, Nikolay Borisov <nborisov@suse.com>,
        "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
        Hannes Reinecke <hare@suse.com>,
        "linux-fsdevel@vger.kernel.org" <linux-fsdevel@vger.kernel.org>,
        Damien Le Moal <Damien.LeMoal@wdc.com>,
        =?iso-8859-1?Q?Matias_Bj=F8rling?= <mb@lightnvm.io>,
        Johannes Thumshirn <jthumshirn@suse.de>,
        Bart Van Assche <bvanassche@acm.org>
Subject: Re: [PATCH 09/19] btrfs: limit super block locations in HMZONED mode
Thread-Topic: [PATCH 09/19] btrfs: limit super block locations in HMZONED mode
Thread-Index: AQHVHTKI41mlZbldhUiKC20uoMUKKg==
Date:   Fri, 28 Jun 2019 06:39:54 +0000
Message-ID: <SN6PR04MB52315AB9E229D0BAAB930D5B8CFC0@SN6PR04MB5231.namprd04.prod.outlook.com>
References: <20190607131025.31996-1-naohiro.aota@wdc.com>
 <20190607131025.31996-10-naohiro.aota@wdc.com>
 <0ca3c475-fe10-4135-ddc9-7a82cc966d9a@oracle.com>
Accept-Language: en-US
Content-Language: en-US
wdcipoutbound: EOP-TRUE
Content-Type: text/plain; charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0
X-MS-Exchange-CrossTenant-Network-Message-Id: ea6f06c6-7936-476d-7c3d-08d6fb937130
X-MS-Exchange-CrossTenant-originalarrivaltime: 28 Jun 2019 06:39:54.8658
 (UTC)
X-MS-Exchange-CrossTenant-fromentityheader: Hosted
X-MS-Exchange-CrossTenant-id: b61c8803-16f3-4c35-9b17-6f65f441df86
X-MS-Exchange-CrossTenant-mailboxtype: HOSTED
X-MS-Exchange-CrossTenant-userprincipalname: Naohiro.Aota1@wdc.com
X-MS-Exchange-Transport-CrossTenantHeadersStamped: SN6PR04MB5119
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org

On 2019/06/28 12:56, Anand Jain wrote:=0A=
> On 7/6/19 9:10 PM, Naohiro Aota wrote:=0A=
>> When in HMZONED mode, make sure that device super blocks are located in=
=0A=
>> randomly writable zones of zoned block devices. That is, do not write su=
per=0A=
>> blocks in sequential write required zones of host-managed zoned block=0A=
>> devices as update would not be possible.=0A=
> =0A=
>    By design all copies of SB must be updated at each transaction,=0A=
>    as they are redundant copies they must match at the end of=0A=
>    each transaction.=0A=
> =0A=
>    Instead of skipping the sb updates, why not alter number of=0A=
>    copies at the time of mkfs.btrfs?=0A=
> =0A=
> Thanks, Anand=0A=
=0A=
That is exactly what the patched code does. It updates all the SB=0A=
copies, but it just avoids writing a copy to sequential writing=0A=
required zones. Mkfs.btrfs do the same. So, all the available SB=0A=
copies always match after a transaction. At the SB location in a=0A=
sequential write required zone, you will see zeroed region (in the=0A=
next version of the patch series), but that is easy to ignore: it=0A=
lacks even BTRFS_MAGIC.=0A=
=0A=
The number of SB copy available on HMZONED device will vary=0A=
by its zone size and its zone layout.=0A=
=0A=
Thanks,=0A=
=0A=
> =0A=
>> Signed-off-by: Damien Le Moal <damien.lemoal@wdc.com>=0A=
>> Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>=0A=
>> ---=0A=
>>    fs/btrfs/disk-io.c     | 11 +++++++++++=0A=
>>    fs/btrfs/disk-io.h     |  1 +=0A=
>>    fs/btrfs/extent-tree.c |  4 ++++=0A=
>>    fs/btrfs/scrub.c       |  2 ++=0A=
>>    4 files changed, 18 insertions(+)=0A=
>>=0A=
>> diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c=0A=
>> index 7c1404c76768..ddbb02906042 100644=0A=
>> --- a/fs/btrfs/disk-io.c=0A=
>> +++ b/fs/btrfs/disk-io.c=0A=
>> @@ -3466,6 +3466,13 @@ struct buffer_head *btrfs_read_dev_super(struct b=
lock_device *bdev)=0A=
>>    	return latest;=0A=
>>    }=0A=
>>    =0A=
>> +int btrfs_check_super_location(struct btrfs_device *device, u64 pos)=0A=
>> +{=0A=
>> +	/* any address is good on a regular (zone_size =3D=3D 0) device */=0A=
>> +	/* non-SEQUENTIAL WRITE REQUIRED zones are capable on a zoned device *=
/=0A=
>> +	return device->zone_size =3D=3D 0 || !btrfs_dev_is_sequential(device, =
pos);=0A=
>> +}=0A=
>> +=0A=
>>    /*=0A=
>>     * Write superblock @sb to the @device. Do not wait for completion, a=
ll the=0A=
>>     * buffer heads we write are pinned.=0A=
>> @@ -3495,6 +3502,8 @@ static int write_dev_supers(struct btrfs_device *d=
evice,=0A=
>>    		if (bytenr + BTRFS_SUPER_INFO_SIZE >=3D=0A=
>>    		    device->commit_total_bytes)=0A=
>>    			break;=0A=
>> +		if (!btrfs_check_super_location(device, bytenr))=0A=
>> +			continue;=0A=
>>    =0A=
>>    		btrfs_set_super_bytenr(sb, bytenr);=0A=
>>    =0A=
>> @@ -3561,6 +3570,8 @@ static int wait_dev_supers(struct btrfs_device *de=
vice, int max_mirrors)=0A=
>>    		if (bytenr + BTRFS_SUPER_INFO_SIZE >=3D=0A=
>>    		    device->commit_total_bytes)=0A=
>>    			break;=0A=
>> +		if (!btrfs_check_super_location(device, bytenr))=0A=
>> +			continue;=0A=
>>    =0A=
>>    		bh =3D __find_get_block(device->bdev,=0A=
>>    				      bytenr / BTRFS_BDEV_BLOCKSIZE,=0A=
>> diff --git a/fs/btrfs/disk-io.h b/fs/btrfs/disk-io.h=0A=
>> index a0161aa1ea0b..70e97cd6fa76 100644=0A=
>> --- a/fs/btrfs/disk-io.h=0A=
>> +++ b/fs/btrfs/disk-io.h=0A=
>> @@ -141,6 +141,7 @@ struct extent_map *btree_get_extent(struct btrfs_ino=
de *inode,=0A=
>>    		struct page *page, size_t pg_offset, u64 start, u64 len,=0A=
>>    		int create);=0A=
>>    int btrfs_get_num_tolerated_disk_barrier_failures(u64 flags);=0A=
>> +int btrfs_check_super_location(struct btrfs_device *device, u64 pos);=
=0A=
>>    int __init btrfs_end_io_wq_init(void);=0A=
>>    void __cold btrfs_end_io_wq_exit(void);=0A=
>>    =0A=
>> diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c=0A=
>> index 3d41d840fe5c..ae2c895d08c4 100644=0A=
>> --- a/fs/btrfs/extent-tree.c=0A=
>> +++ b/fs/btrfs/extent-tree.c=0A=
>> @@ -267,6 +267,10 @@ static int exclude_super_stripes(struct btrfs_block=
_group_cache *cache)=0A=
>>    			return ret;=0A=
>>    	}=0A=
>>    =0A=
>> +	/* we won't have super stripes in sequential zones */=0A=
>> +	if (cache->alloc_type =3D=3D BTRFS_ALLOC_SEQ)=0A=
>> +		return 0;=0A=
>> +=0A=
>>    	for (i =3D 0; i < BTRFS_SUPER_MIRROR_MAX; i++) {=0A=
>>    		bytenr =3D btrfs_sb_offset(i);=0A=
>>    		ret =3D btrfs_rmap_block(fs_info, cache->key.objectid,=0A=
>> diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c=0A=
>> index f7b29f9db5e2..36ad4fad7eaf 100644=0A=
>> --- a/fs/btrfs/scrub.c=0A=
>> +++ b/fs/btrfs/scrub.c=0A=
>> @@ -3720,6 +3720,8 @@ static noinline_for_stack int scrub_supers(struct =
scrub_ctx *sctx,=0A=
>>    		if (bytenr + BTRFS_SUPER_INFO_SIZE >=0A=
>>    		    scrub_dev->commit_total_bytes)=0A=
>>    			break;=0A=
>> +		if (!btrfs_check_super_location(scrub_dev, bytenr))=0A=
>> +			continue;=0A=
>>    =0A=
>>    		ret =3D scrub_pages(sctx, bytenr, BTRFS_SUPER_INFO_SIZE, bytenr,=0A=
>>    				  scrub_dev, BTRFS_EXTENT_FLAG_SUPER, gen, i,=0A=
>>=0A=
> =0A=
> =0A=
=0A=