From: Josef Bacik <josef@toxicpanda.com>
To: Naohiro Aota <naohiro.aota@wdc.com>,
linux-btrfs@vger.kernel.org, dsterba@suse.com
Cc: hare@suse.com, linux-fsdevel@vger.kernel.org
Subject: Re: [PATCH v9 14/41] btrfs: load zone's alloction offset
Date: Mon, 2 Nov 2020 15:29:36 -0500 [thread overview]
Message-ID: <2201bd4a-03ee-b68b-683b-0ba079b071dc@toxicpanda.com> (raw)
In-Reply-To: <1730f278-39d5-cd82-7cd5-a48d826df2ef@toxicpanda.com>
On 11/2/20 3:25 PM, Josef Bacik wrote:
> On 10/30/20 9:51 AM, Naohiro Aota wrote:
>> Zoned btrfs must allocate blocks at the zones' write pointer. The device's
>> write pointer position can be mapped to a logical address within a block
>> group. This commit adds "alloc_offset" to track the logical address.
>>
>> This logical address is populated in btrfs_load_block-group_zone_info()
>
> btrfs_load_block_group_zone_info()
>
>> from write pointers of corresponding zones.
>>
>> For now, zoned btrfs only support the SINGLE profile. Supporting non-SINGLE
>> profile with zone append writing is not trivial. For example, in the DUP
>> profile, we send a zone append writing IO to two zones on a device. The
>> device reply with written LBAs for the IOs. If the offsets of the returned
>> addresses from the beginning of the zone are different, then it results in
>> different logical addresses.
>>
>> We need fine-grained logical to physical mapping to support such separated
>> physical address issue. Since it should require additional metadata type,
>> disable non-SINGLE profiles for now.
>>
>> Signed-off-by: Naohiro Aota <naohiro.aota@wdc.com>
>> ---
>> fs/btrfs/block-group.c | 15 ++++
>> fs/btrfs/block-group.h | 6 ++
>> fs/btrfs/zoned.c | 153 +++++++++++++++++++++++++++++++++++++++++
>> fs/btrfs/zoned.h | 6 ++
>> 4 files changed, 180 insertions(+)
>>
>> diff --git a/fs/btrfs/block-group.c b/fs/btrfs/block-group.c
>> index e989c66aa764..920b2708c7f2 100644
>> --- a/fs/btrfs/block-group.c
>> +++ b/fs/btrfs/block-group.c
>> @@ -15,6 +15,7 @@
>> #include "delalloc-space.h"
>> #include "discard.h"
>> #include "raid56.h"
>> +#include "zoned.h"
>> /*
>> * Return target flags in extended format or 0 if restripe for this chunk_type
>> @@ -1935,6 +1936,13 @@ static int read_one_block_group(struct btrfs_fs_info
>> *info,
>> goto error;
>> }
>> + ret = btrfs_load_block_group_zone_info(cache);
>> + if (ret) {
>> + btrfs_err(info, "failed to load zone info of bg %llu",
>> + cache->start);
>> + goto error;
>> + }
>> +
>> /*
>> * We need to exclude the super stripes now so that the space info has
>> * super bytes accounted for, otherwise we'll think we have more space
>> @@ -2161,6 +2169,13 @@ int btrfs_make_block_group(struct btrfs_trans_handle
>> *trans, u64 bytes_used,
>> cache->last_byte_to_unpin = (u64)-1;
>> cache->cached = BTRFS_CACHE_FINISHED;
>> cache->needs_free_space = 1;
>> +
>> + ret = btrfs_load_block_group_zone_info(cache);
>> + if (ret) {
>> + btrfs_put_block_group(cache);
>> + return ret;
>> + }
>> +
>> ret = exclude_super_stripes(cache);
>> if (ret) {
>> /* We may have excluded something, so call this just in case */
>> diff --git a/fs/btrfs/block-group.h b/fs/btrfs/block-group.h
>> index adfd7583a17b..14e3043c9ce7 100644
>> --- a/fs/btrfs/block-group.h
>> +++ b/fs/btrfs/block-group.h
>> @@ -183,6 +183,12 @@ struct btrfs_block_group {
>> /* Record locked full stripes for RAID5/6 block group */
>> struct btrfs_full_stripe_locks_tree full_stripe_locks_root;
>> +
>> + /*
>> + * Allocation offset for the block group to implement sequential
>> + * allocation. This is used only with ZONED mode enabled.
>> + */
>> + u64 alloc_offset;
>> };
>> static inline u64 btrfs_block_group_end(struct btrfs_block_group *block_group)
>> diff --git a/fs/btrfs/zoned.c b/fs/btrfs/zoned.c
>> index 4411d786597a..0aa821893a51 100644
>> --- a/fs/btrfs/zoned.c
>> +++ b/fs/btrfs/zoned.c
>> @@ -3,14 +3,20 @@
>> #include <linux/bitops.h>
>> #include <linux/slab.h>
>> #include <linux/blkdev.h>
>> +#include <linux/sched/mm.h>
>> #include "ctree.h"
>> #include "volumes.h"
>> #include "zoned.h"
>> #include "rcu-string.h"
>> #include "disk-io.h"
>> +#include "block-group.h"
>> /* Maximum number of zones to report per blkdev_report_zones() call */
>> #define BTRFS_REPORT_NR_ZONES 4096
>> +/* Invalid allocation pointer value for missing devices */
>> +#define WP_MISSING_DEV ((u64)-1)
>> +/* Pseudo write pointer value for conventional zone */
>> +#define WP_CONVENTIONAL ((u64)-2)
>> static int copy_zone_info_cb(struct blk_zone *zone, unsigned int idx,
>> void *data)
>> @@ -733,3 +739,150 @@ int btrfs_ensure_empty_zones(struct btrfs_device
>> *device, u64 start, u64 size)
>> return 0;
>> }
>> +
>> +int btrfs_load_block_group_zone_info(struct btrfs_block_group *cache)
>> +{
>> + struct btrfs_fs_info *fs_info = cache->fs_info;
>> + struct extent_map_tree *em_tree = &fs_info->mapping_tree;
>> + struct extent_map *em;
>> + struct map_lookup *map;
>> + struct btrfs_device *device;
>> + u64 logical = cache->start;
>> + u64 length = cache->length;
>> + u64 physical = 0;
>> + int ret;
>> + int i;
>> + unsigned int nofs_flag;
>> + u64 *alloc_offsets = NULL;
>> + u32 num_sequential = 0, num_conventional = 0;
>> +
>> + if (!btrfs_is_zoned(fs_info))
>> + return 0;
>> +
>> + /* Sanity check */
>> + if (!IS_ALIGNED(length, fs_info->zone_size)) {
>> + btrfs_err(fs_info, "unaligned block group at %llu + %llu",
>> + logical, length);
>> + return -EIO;
>> + }
>> +
>> + /* Get the chunk mapping */
>> + read_lock(&em_tree->lock);
>> + em = lookup_extent_mapping(em_tree, logical, length);
>> + read_unlock(&em_tree->lock);
>> +
>> + if (!em)
>> + return -EINVAL;
>> +
>> + map = em->map_lookup;
>> +
>> + /*
>> + * Get the zone type: if the group is mapped to a non-sequential zone,
>> + * there is no need for the allocation offset (fit allocation is OK).
>> + */
>> + alloc_offsets = kcalloc(map->num_stripes, sizeof(*alloc_offsets),
>> + GFP_NOFS);
>> + if (!alloc_offsets) {
>> + free_extent_map(em);
>> + return -ENOMEM;
>> + }
>> +
>> + for (i = 0; i < map->num_stripes; i++) {
>> + bool is_sequential;
>> + struct blk_zone zone;
>> +
>> + device = map->stripes[i].dev;
>> + physical = map->stripes[i].physical;
>> +
>> + if (device->bdev == NULL) {
>> + alloc_offsets[i] = WP_MISSING_DEV;
>> + continue;
>> + }
>> +
>> + is_sequential = btrfs_dev_is_sequential(device, physical);
>> + if (is_sequential)
>> + num_sequential++;
>> + else
>> + num_conventional++;
>> +
>> + if (!is_sequential) {
>> + alloc_offsets[i] = WP_CONVENTIONAL;
>> + continue;
>> + }
>> +
>> + /*
>> + * This zone will be used for allocation, so mark this
>> + * zone non-empty.
>> + */
>> + btrfs_dev_clear_zone_empty(device, physical);
>> +
>> + /*
>> + * The group is mapped to a sequential zone. Get the zone write
>> + * pointer to determine the allocation offset within the zone.
>> + */
>> + WARN_ON(!IS_ALIGNED(physical, fs_info->zone_size));
>> + nofs_flag = memalloc_nofs_save();
>> + ret = btrfs_get_dev_zone(device, physical, &zone);
>> + memalloc_nofs_restore(nofs_flag);
>> + if (ret == -EIO || ret == -EOPNOTSUPP) {
>> + ret = 0;
>> + alloc_offsets[i] = WP_MISSING_DEV;
>> + continue;
>> + } else if (ret) {
>> + goto out;
>> + }
>> +
>> + switch (zone.cond) {
>> + case BLK_ZONE_COND_OFFLINE:
>> + case BLK_ZONE_COND_READONLY:
>> + btrfs_err(fs_info, "Offline/readonly zone %llu",
>> + physical >> device->zone_info->zone_size_shift);
>> + alloc_offsets[i] = WP_MISSING_DEV;
>> + break;
>> + case BLK_ZONE_COND_EMPTY:
>> + alloc_offsets[i] = 0;
>> + break;
>> + case BLK_ZONE_COND_FULL:
>> + alloc_offsets[i] = fs_info->zone_size;
>> + break;
>> + default:
>> + /* Partially used zone */
>> + alloc_offsets[i] =
>> + ((zone.wp - zone.start) << SECTOR_SHIFT);
>> + break;
>> + }
>> + }
>> +
>> + if (num_conventional > 0) {
>> + /*
>> + * Since conventional zones does not have write pointer, we
>> + * cannot determine alloc_offset from the pointer
>> + */
>> + ret = -EINVAL;
>> + goto out;
>> + }
>
> Does this mean we can't have zoned with a device that has conventional and
> sequential zones? I thought such things existed currently? Thanks,
I see this changes in a follow up patch, ignore me, you can add
Reviewed-by: Josef Bacik <josef@toxicpanda.com>
Thanks,
Josef
next prev parent reply other threads:[~2020-11-02 20:29 UTC|newest]
Thread overview: 119+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-10-30 13:51 [PATCH v9 00/41] btrfs: zoned block device support Naohiro Aota
2020-10-30 13:51 ` [PATCH v9 01/41] block: add bio_add_zone_append_page Naohiro Aota
2020-10-30 13:51 ` [PATCH v9 02/41] iomap: support REQ_OP_ZONE_APPEND Naohiro Aota
2020-11-02 5:34 ` Naohiro Aota
2020-11-02 16:55 ` Darrick J. Wong
2020-11-02 17:39 ` Johannes Thumshirn
2020-10-30 13:51 ` [PATCH v9 03/41] btrfs: introduce ZONED feature flag Naohiro Aota
2020-10-30 13:51 ` [PATCH v9 04/41] btrfs: Get zone information of zoned block devices Naohiro Aota
2020-11-02 16:53 ` Josef Bacik
2020-11-02 16:58 ` Johannes Thumshirn
2020-11-02 21:07 ` Naohiro Aota
2020-11-03 12:02 ` David Sterba
2020-10-30 13:51 ` [PATCH v9 05/41] btrfs: Check and enable ZONED mode Naohiro Aota
2020-11-03 12:13 ` David Sterba
2020-11-06 9:36 ` Naohiro Aota
2020-10-30 13:51 ` [PATCH v9 06/41] btrfs: introduce max_zone_append_size Naohiro Aota
2020-11-02 16:57 ` Josef Bacik
2020-11-03 12:16 ` David Sterba
2020-10-30 13:51 ` [PATCH v9 07/41] btrfs: disallow space_cache in ZONED mode Naohiro Aota
2020-11-02 17:02 ` Josef Bacik
2020-11-02 17:37 ` Johannes Thumshirn
2020-11-03 12:48 ` David Sterba
2020-11-10 10:14 ` Naohiro Aota
2020-10-30 13:51 ` [PATCH v9 08/41] btrfs: disallow NODATACOW " Naohiro Aota
2020-11-02 17:05 ` Josef Bacik
2020-11-03 12:57 ` David Sterba
2020-10-30 13:51 ` [PATCH v9 09/41] btrfs: disable fallocate " Naohiro Aota
2020-11-03 13:00 ` David Sterba
2020-10-30 13:51 ` [PATCH v9 10/41] btrfs: disallow mixed-bg " Naohiro Aota
2020-11-03 13:01 ` David Sterba
2020-10-30 13:51 ` [PATCH v9 11/41] btrfs: implement log-structured superblock for " Naohiro Aota
2020-11-02 18:22 ` Josef Bacik
2020-11-02 18:53 ` Johannes Thumshirn
2020-11-02 19:01 ` Josef Bacik
2020-11-02 19:31 ` Johannes Thumshirn
2020-11-03 8:21 ` Naohiro Aota
2020-11-02 18:54 ` Josef Bacik
2020-11-03 3:31 ` Naohiro Aota
2020-11-03 13:15 ` David Sterba
2020-11-03 14:10 ` David Sterba
2020-11-06 10:37 ` Naohiro Aota
2020-10-30 13:51 ` [PATCH v9 12/41] btrfs: implement zoned chunk allocator Naohiro Aota
2020-11-02 20:09 ` Josef Bacik
2020-11-02 22:21 ` Naohiro Aota
2020-11-03 13:23 ` David Sterba
2020-10-30 13:51 ` [PATCH v9 13/41] btrfs: verify device extent is aligned to zone Naohiro Aota
2020-11-02 20:14 ` Josef Bacik
2020-10-30 13:51 ` [PATCH v9 14/41] btrfs: load zone's alloction offset Naohiro Aota
2020-11-02 20:25 ` Josef Bacik
2020-11-02 20:29 ` Josef Bacik [this message]
2020-11-02 22:43 ` Naohiro Aota
2020-11-02 22:40 ` Naohiro Aota
2020-11-03 13:28 ` David Sterba
2020-10-30 13:51 ` [PATCH v9 15/41] btrfs: emulate write pointer for conventional zones Naohiro Aota
2020-11-02 20:37 ` Josef Bacik
2020-11-03 1:25 ` Naohiro Aota
2020-11-03 13:32 ` David Sterba
2020-10-30 13:51 ` [PATCH v9 16/41] btrfs: track unusable bytes for zones Naohiro Aota
2020-11-03 14:25 ` Josef Bacik
2020-10-30 13:51 ` [PATCH v9 17/41] btrfs: do sequential extent allocation in ZONED mode Naohiro Aota
2020-11-03 14:28 ` Josef Bacik
2020-10-30 13:51 ` [PATCH v9 18/41] btrfs: reset zones of unused block groups Naohiro Aota
2020-11-03 14:34 ` Josef Bacik
2020-11-10 10:40 ` Naohiro Aota
2020-10-30 13:51 ` [PATCH v9 19/41] btrfs: redirty released extent buffers in ZONED mode Naohiro Aota
2020-11-03 14:41 ` Josef Bacik
2020-11-06 9:11 ` Johannes Thumshirn
2020-11-06 15:01 ` Josef Bacik
2020-10-30 13:51 ` [PATCH v9 20/41] btrfs: extract page adding function Naohiro Aota
2020-11-03 14:45 ` Josef Bacik
2020-10-30 13:51 ` [PATCH v9 21/41] btrfs: use bio_add_zone_append_page for zoned btrfs Naohiro Aota
2020-11-03 14:55 ` Josef Bacik
2020-11-10 10:42 ` Naohiro Aota
2020-10-30 13:51 ` [PATCH v9 22/41] btrfs: handle REQ_OP_ZONE_APPEND as writing Naohiro Aota
2020-11-03 14:57 ` Josef Bacik
2020-10-30 13:51 ` [PATCH v9 23/41] btrfs: split ordered extent when bio is sent Naohiro Aota
2020-11-03 15:29 ` Josef Bacik
2020-10-30 13:51 ` [PATCH v9 24/41] btrfs: extend btrfs_rmap_block for specifying a device Naohiro Aota
2020-11-03 15:32 ` Josef Bacik
2020-11-06 10:52 ` Johannes Thumshirn
2020-10-30 13:51 ` [PATCH v9 25/41] btrfs: use ZONE_APPEND write for ZONED btrfs Naohiro Aota
2020-11-03 15:55 ` Josef Bacik
2020-10-30 13:51 ` [PATCH v9 26/41] btrfs: enable zone append writing for direct IO Naohiro Aota
2020-11-03 15:56 ` Josef Bacik
2020-10-30 13:51 ` [PATCH v9 27/41] btrfs: introduce dedicated data write path for ZONED mode Naohiro Aota
2020-11-03 15:57 ` Josef Bacik
2020-10-30 13:51 ` [PATCH v9 28/41] btrfs: serialize meta IOs on " Naohiro Aota
2020-11-03 16:04 ` Josef Bacik
2020-10-30 13:51 ` [PATCH v9 29/41] btrfs: wait existing extents before truncating Naohiro Aota
2020-10-30 13:51 ` [PATCH v9 30/41] btrfs: avoid async metadata checksum on ZONED mode Naohiro Aota
2020-11-03 16:05 ` Josef Bacik
2020-10-30 13:51 ` [PATCH v9 31/41] btrfs: mark block groups to copy for device-replace Naohiro Aota
2020-11-03 17:09 ` Josef Bacik
2020-10-30 13:51 ` [PATCH v9 32/41] btrfs: implement cloning for ZONED device-replace Naohiro Aota
2020-11-03 17:15 ` Josef Bacik
2020-10-30 13:51 ` [PATCH v9 33/41] btrfs: implement copying " Naohiro Aota
2020-11-03 17:19 ` Josef Bacik
2020-11-10 11:09 ` Naohiro Aota
2020-10-30 13:51 ` [PATCH v9 34/41] btrfs: support dev-replace in ZONED mode Naohiro Aota
2020-11-03 20:34 ` Josef Bacik
2020-10-30 13:51 ` [PATCH v9 35/41] btrfs: enable relocation " Naohiro Aota
2020-11-03 20:39 ` Josef Bacik
2020-10-30 13:51 ` [PATCH v9 36/41] btrfs: relocate block group to repair IO failure in ZONED Naohiro Aota
2020-10-30 13:51 ` [PATCH v9 37/41] btrfs: split alloc_log_tree() Naohiro Aota
2020-11-03 20:42 ` Josef Bacik
2020-11-03 22:10 ` Amy Parker
2020-11-10 11:12 ` Naohiro Aota
2020-10-30 13:51 ` [PATCH v9 38/41] btrfs: extend zoned allocator to use dedicated tree-log block group Naohiro Aota
2020-11-03 20:47 ` Josef Bacik
2020-11-10 6:37 ` Naohiro Aota
2020-10-30 13:51 ` [PATCH v9 39/41] btrfs: serialize log transaction on ZONED mode Naohiro Aota
2020-11-03 20:49 ` Josef Bacik
2020-10-30 13:51 ` [PATCH v9 40/41] btrfs: reorder log node allocation Naohiro Aota
2020-11-03 20:49 ` Josef Bacik
2020-10-30 13:51 ` [PATCH v9 41/41] btrfs: enable to mount ZONED incompat flag Naohiro Aota
2020-10-31 3:40 ` [PATCH v9 01/41] block: add bio_add_zone_append_page Jens Axboe
2020-11-02 5:15 ` Naohiro Aota
2020-11-02 8:24 ` Johannes Thumshirn
2020-11-03 11:54 ` [PATCH v9 00/41] btrfs: zoned block device support David Sterba
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=2201bd4a-03ee-b68b-683b-0ba079b071dc@toxicpanda.com \
--to=josef@toxicpanda.com \
--cc=dsterba@suse.com \
--cc=hare@suse.com \
--cc=linux-btrfs@vger.kernel.org \
--cc=linux-fsdevel@vger.kernel.org \
--cc=naohiro.aota@wdc.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).