Skip to contentSkip to conversation

btrfs: zoned block device support

PATCH Superseded Naohiro Aota 11 participants 121 messages Nov 10, 2020 – Dec 12, 2020

Subsystems (4)

v10 of 15 Superseded Latest is v15 →

lore
Share & tools
lore
#0 lore raw

v10 cover btrfs: zoned block device support

This series adds zoned block device support to btrfs.

This series is also available on github.
Kernel https://github.com/naota/linux/tree/btrfs-zoned-v10
Userland https://github.com/naota/btrfs-progs/tree/btrfs-zoned
xfstests https://github.com/naota/fstests/tree/btrfs-zoned

Userland tool depends on patched util-linux (libblkid and wipefs) to handle log-structured superblock. To ease the testing, pre-compiled static linked userland tools are available here:
https://wdc.app.box.com/s/fnhqsb3otrvgkstq66o6bvdw6tk525kp

This v10 still leaves the following issues left for later fix. But, the first part of the series should be good shape to be merged.
- Bio submission path & splitting an ordered extent
- Redirtying freed tree blocks
- Switch to keeping it dirty
- Not working correctly for now
- Dedicated tree-log block group
- We need tree-log for zoned device
- Dbench (32 clients) is 85% slower with "-o notreelog"
- Need to separate tree-log block group from other metadata space_info
- Relocation
- Use normal write command for relocation
- Relocated device extents must be reset
- It should be discarded on regular btrfs too though

Changes from v9:
- Extract iomap_dio_bio_opflags() to set the proper bi_opf flag
- write pointer emulation
- Rewrite using btrfs_previous_extent_item()
- Convert ASSERT() to runtime check
- Exclude regular superblock positions
- Fix an error on writing to conventional zones
- Take the transaction lock in mark_block_group_to_copy()
- Rename 'hmzoned_devices' to 'zoned_devices' in btrfs_check_zoned_mode()
- Add do_discard_extent() helper
- Move zoned check into fetch_cluster_info()
- Drop setting bdev to bio in btrfs_bio_add_page() (will fix later once we support multiple devices)
- Subtract bytes_zone_unusable properly when removing a block group
- Add "struct block_device *bdev" directly to btrfs_rmap_block()
- Rename btrfs_zone_align to btrfs_align_offset_to_zone
- Add comment to use pr_info in place of btrfs_info
- Add comment for superblock log zones
- Fix coding style
- Fix typos

btrfs-progs and xfstests series will follow.

This version of ZONED btrfs switched from normal write command to zone append write command. You do not need to specify LBA (at the write pointer) to write for zone append write command. Instead, you only select a zone to write with its start LBA. Then the device (NVMe ZNS), or the emulation of zone append command in the sd driver in the case of SAS or SATA HDDs, automatically writes the data at the write pointer position and return the written LBA as a command reply.

The benefit of using the zone append write command is that write command issuing order does not matter. So, we can eliminate block group lock and utilize asynchronous checksum, which can reorder the IOs.

Eliminating the lock improves performance. In particular, on a workload with massive competing to the same zone [1], we observed 36% performance improvement compared to normal write.

[1] Fio running 16 jobs with 4KB random writes for 5 minutes

However, there are some limitations. We cannot use the non-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.

For the same reason, we cannot issue multiple IOs for one ordered extent. Thus, the size of an ordered extent is limited under max_zone_append_size. This limitation will cause fragmentation and increased usage of metadata. In the future, we can add optimization to merge ordered extents after end_bio.

* Patch series description

A zoned block device consists of a number of zones. Zones are either conventional and accepting random writes or sequential and requiring that writes be issued in LBA order from each zone write pointer position. This patch series ensures that the sequential write constraint of sequential zones is respected while fundamentally not changing BtrFS block and I/O management for block stored in conventional zones.

To achieve this, the default chunk size of btrfs is changed on zoned block devices so that chunks are always aligned to a zone. Allocation of blocks within a chunk is changed so that the allocation is always sequential from the beginning of the chunks. To do so, an allocation pointer is added to block groups and used as the allocation hint. The allocation changes also ensure that blocks freed below the allocation pointer are ignored, resulting in sequential block allocation regardless of the chunk usage.

The zone of a chunk is reset to allow reuse of the zone only when the block group is being freed, that is, when all the chunks of the block group are unused.

For btrfs volumes composed of multiple zoned disks, a restriction is added to ensure that all disks have the same zone size. This restriction matches the existing constraint that all chunks in a block group must have the same size.

* Enabling tree-log

The tree-log feature does not work on ZONED mode as is. Blocks for a tree-log tree are allocated mixed with other metadata blocks, and btrfs writes and syncs the tree-log blocks to devices at the time of fsync(), which is different timing than a global transaction commit. As a result, both writing tree-log blocks and writing other metadata blocks become non-sequential writes which ZONED mode must avoid.

This series introduces a dedicated block group for tree-log blocks to create two metadata writing streams, one for tree-log blocks and the other for metadata blocks. As a result, each write stream can now be written to devices separately and sequentially.

* Log-structured superblock

Superblock (and its copies) is the only data structure in btrfs which has a fixed location on a device. Since we cannot overwrite in a sequential write required zone, we cannot place superblock in the zone.

This series implements superblock log writing. It uses two zones as a circular buffer to write updated superblocks. Once the first zone is filled up, start writing into the second zone. The first zone will be reset once both zones are filled. We can determine the postion of the latest superblock by reading the write pointer information from a device.

* Patch series organization

Patches 1 and 2 are preparing patches for block and iomap layer.

Patch 3 introduces the ZONED incompatible feature flag to indicate that the btrfs volume was formatted for use on zoned block devices.

Patches 4 to 6 implement functions to gather information on the zones of the device (zones type, write pointer position, and max_zone_append_size).

Patches 7 to 10 disable features which are not compatible with the sequential write constraints of zoned block devices. These includes space_cache, NODATACOW, fallocate, and MIXED_BG.

Patch 11 implements the log-structured superblock writing.

Patches 12 and 13 tweak the device extent allocation for ZONED mode and add verification to check if a device extent is properly aligned to zones.

Patches 14 to 17 implements sequential block allocator for ZONED mode.

Patch 18 implement a zone reset for unused block groups.

Patches 19 to 30 implement the writing path for several types of IO (non-compressed data, direct IO, and metadata). These include re-dirtying once-freed metadata blocks to prevent write holes.

Patches 31 to 40 tweak some btrfs features work with ZONED mode. These include device-replace, relocation, repairing IO error, and tree-log.

Finally, patch 41 adds the ZONED feature to the list of supported features.

* Patch testing note

** Zone-aware util-linux

Since the log-structured superblock feature changed the location of superblock magic, the current util-linux (libblkid) cannot detect ZONED btrfs anymore. You need to apply a to-be posted patch to util-linux to make it "zone aware".

** Testing device

You need devices with zone append writing command support to run ZONED btrfs.

Other than real devices, null_blk supports zone append write command. You can use memory backed null_blk to run the test on it. Following script creates 12800 MB /dev/nullb0.

    sysfs=/sys/kernel/config/nullb/nullb0
    size=12800 # MB
    
    # drop nullb0
    if [[ -d $sysfs ]]; then
            echo 0 > "${sysfs}"/power
            rmdir $sysfs
    fi
    lsmod | grep -q null_blk && rmmod null_blk
    modprobe null_blk nr_devices=0
    
    mkdir "${sysfs}"
    
    echo "${size}" > "${sysfs}"/size
    echo 1 > "${sysfs}"/zoned
    echo 0 > "${sysfs}"/zone_nr_conv
    echo 1 > "${sysfs}"/memory_backed
    
    echo 1 > "${sysfs}"/power
    udevadm settle

Zoned SCSI devices such as SMR HDDs or scsi_debug also support the zone append command as an emulated command within the SCSI sd driver. This emulation is completely transparent to the user and provides the same semantic as a NVMe ZNS native drive support.

Also, there is a qemu patch available to enable NVMe ZNS device.

** xfstests

We ran xfstests on ZONED btrfs, and, if we omit some cases that are known to fail currently, all test cases pass.

Cases that can be ignored:
1) failing also with the regular btrfs on regular devices,
2) trying to test fallocate feature without testing with "_require_xfs_io_command "falloc"",
3) trying to test incompatible features for ZONED btrfs (e.g. RAID5/6)
4) trying to use incompatible setup for ZONED btrfs (e.g. dm-linear not aligned to zone boundary, swap)
5) trying to create a file system with too small size, (we require at least 9 zones to initiate a ZONED btrfs)
6) dropping original MKFS_OPTIONS ("-O zoned"), so it cannot create ZONED btrfs (btrfs/003)
7) having ENOSPC which incurred by larger metadata block group size

I will send a patch series for xfstests to handle these cases (2-6) properly.

Patched xfstests is available here:

https://github.com/naota/fstests/tree/btrfs-zoned

Also, you need to apply the following patch if you run xfstests with tcmu devices. xfstests btrfs/003 failed to "_devmgt_add" after "_devmgt_remove" without this patch.

https://marc.info/?l=linux-scsi&m=156498625421698&w=2

v9 https://lore.kernel.org/linux-btrfs/[email protected]/
v8 https://lore.kernel.org/linux-btrfs/[email protected]/
v7 https://lore.kernel.org/linux-btrfs/[email protected]/
v6 https://lore.kernel.org/linux-btrfs/[email protected]/
v5 https://lore.kernel.org/linux-btrfs/[email protected]/
v4 https://lwn.net/Articles/797061/
v3 https://lore.kernel.org/linux-btrfs/[email protected]/
v2 https://lore.kernel.org/linux-btrfs/[email protected]/
v1 https://lore.kernel.org/linux-btrfs/[email protected]/

Changelog
v9
- Direct-IO path now follow several hardware restrictions (other than max_zone_append_size) by using ZONE_APPEND support of iomap
- introduces union of fs_info->zone_size and fs_info->zoned [Johannes]
- and use btrfs_is_zoned(fs_info) in place of btrfs_fs_incompat(fs_info, ZONED)
- print if zoned is enabled or not when printing module info [Johannes]
- drop patch of disabling inode_cache on ZONED
- moved for_teelog flag to a proper location [Johannes]
- Code style fixes [Johannes]
- Add comment about adding physical layer things to ordered extent structure
- Pass file_offset explicitly to extract_ordered_extent() instead of determining it from bio
- Bug fixes
- write out fsync region so that the logical address of ordered extents and checksums are properly finalized
- free zone_info at umount time
- fix superblock log handling when entering zones[1] in the first time
- fixes double free of log-tree roots [Johannes]
- Drop erroneous ASSERT in do_allocation_zoned()
v8
- Use bio_add_hw_page() to build up bio to honor hardware restrictions
- add bio_add_zone_append_page() as a wrapper of the function
- Split file extent on submitting bio
- If bio_add_zone_append_page() fails, split the file extent and send out bio
- so, we can ensure one bio == one file extent
- Fix build bot issues
- Rebased on misc-next
v7:
- Use zone append write command instead of normal write command
- Bio issuing order does not matter
- No need to use lock anymore
- Can use asynchronous checksum
- Removed RAID support for now
- Rename HMZONED to ZONED
- Split some patches
- Rebased on kdave/for-5.9-rc3 + iomap direct IO
v6:
- Use bitmap helpers (Johannes)
- Code cleanup (Johannes)
- Rebased on kdave/for-5.5
- Enable the tree-log feature.
- Treat conventional zones as sequential zones, so we can now allow mixed allocation of conventional zone and sequential write required zone to construct a block group.
- Implement log-structured superblock
- No need for one conventional zone at the beginning of a device.
- Fix deadlock of direct IO writing
- Fix building with !CONFIG_BLK_DEV_ZONED (Johannes)
- Fix leak of zone_info (Johannes)
v5:
- Rebased on kdave/for-5.5
- Enable the tree-log feature.
- Treat conventional zones as sequential zones, so we can now allow mixed allocation of conventional zone and sequential write required zone to construct a block group.
- Implement log-structured superblock
- No need for one conventional zone at the beginning of a device.
- Fix deadlock of direct IO writing
- Fix building with !CONFIG_BLK_DEV_ZONED (Johannes)
- Fix leak of zone_info (Johannes)
v4:
- Move memory allcation of zone informattion out of btrfs_get_dev_zones() (Anand)
- Add disabled features table in commit log (Anand)
- Ensure "max_chunk_size >= devs_min * data_stripes * zone_size"
v3:
- Serialize allocation and submit_bio instead of bio buffering in btrfs_map_bio().
-- Disable async checksum/submit in HMZONED mode
- Introduce helper functions and hmzoned.c/h (Josef, David)
- Add support for repairing IO failure
- Add support for NOCOW direct IO write (Josef)
- Disable preallocation entirely
-- Disable INODE_MAP_CACHE
-- relocation is reworked not to rely on preallocation in HMZONED mode
- Disable NODATACOW
-Disable MIXED_BG
- Device extent that cover super block position is banned (David)
v2:
- Add support for dev-replace
-- To support dev-replace, moved submit_buffer one layer up. It now
handles bio instead of btrfs_bio.
-- Mark unmirrored Block Group readonly only when there are writable
mirrored BGs. Necessary to handle degraded RAID.
- Expire worker use vanilla delayed_work instead of btrfs's async-thread
- Device extent allocator now ensure that region is on the same zone type.
- Add delayed allocation shrinking.
- Rename btrfs_drop_dev_zonetypes() to btrfs_destroy_dev_zonetypes
- Fix
-- Use SECTOR_SHIFT (Nikolay)
-- Use btrfs_err (Nikolay)

Johannes Thumshirn (1):
block: add bio_add_zone_append_page

Naohiro Aota (40):
iomap: support REQ_OP_ZONE_APPEND
btrfs: introduce ZONED feature flag
btrfs: get zone information of zoned block devices
btrfs: check and enable ZONED mode
btrfs: introduce max_zone_append_size
btrfs: disallow space_cache in ZONED mode
btrfs: disallow NODATACOW in ZONED mode
btrfs: disable fallocate in ZONED mode
btrfs: disallow mixed-bg in ZONED mode
btrfs: implement log-structured superblock for ZONED mode
btrfs: implement zoned chunk allocator
btrfs: verify device extent is aligned to zone
btrfs: load zone's alloction offset
btrfs: emulate write pointer for conventional zones
btrfs: track unusable bytes for zones
btrfs: do sequential extent allocation in ZONED mode
btrfs: reset zones of unused block groups
btrfs: redirty released extent buffers in ZONED mode
btrfs: extract page adding function
btrfs: use bio_add_zone_append_page for zoned btrfs
btrfs: handle REQ_OP_ZONE_APPEND as writing
btrfs: split ordered extent when bio is sent
btrfs: extend btrfs_rmap_block for specifying a device
btrfs: use ZONE_APPEND write for ZONED btrfs
btrfs: enable zone append writing for direct IO
btrfs: introduce dedicated data write path for ZONED mode
btrfs: serialize meta IOs on ZONED mode
btrfs: wait existing extents before truncating
btrfs: avoid async metadata checksum on ZONED mode
btrfs: mark block groups to copy for device-replace
btrfs: implement cloning for ZONED device-replace
btrfs: implement copying for ZONED device-replace
btrfs: support dev-replace in ZONED mode
btrfs: enable relocation in ZONED mode
btrfs: relocate block group to repair IO failure in ZONED
btrfs: split alloc_log_tree()
btrfs: extend zoned allocator to use dedicated tree-log block group
btrfs: serialize log transaction on ZONED mode
btrfs: reorder log node allocation
btrfs: enable to mount ZONED incompat flag

 block/bio.c                       |   38 +
 fs/btrfs/Makefile                 |    1 +
 fs/btrfs/block-group.c            |   84 +-
 fs/btrfs/block-group.h            |   18 +-
 fs/btrfs/ctree.h                  |   20 +-
 fs/btrfs/dev-replace.c            |  195 +++++
 fs/btrfs/dev-replace.h            |    3 +
 fs/btrfs/disk-io.c                |   93 ++-
 fs/btrfs/disk-io.h                |    2 +
 fs/btrfs/extent-tree.c            |  218 ++++-
 fs/btrfs/extent_io.c              |  130 ++-
 fs/btrfs/extent_io.h              |    2 +
 fs/btrfs/file.c                   |    6 +-
 fs/btrfs/free-space-cache.c       |   58 ++
 fs/btrfs/free-space-cache.h       |    2 +
 fs/btrfs/inode.c                  |  164 +++-
 fs/btrfs/ioctl.c                  |   13 +
 fs/btrfs/ordered-data.c           |   79 ++
 fs/btrfs/ordered-data.h           |   10 +
 fs/btrfs/relocation.c             |   35 +-
 fs/btrfs/scrub.c                  |  145 ++++
 fs/btrfs/space-info.c             |   13 +-
 fs/btrfs/space-info.h             |    4 +-
 fs/btrfs/super.c                  |   19 +-
 fs/btrfs/sysfs.c                  |    4 +
 fs/btrfs/tests/extent-map-tests.c |    2 +-
 fs/btrfs/transaction.c            |   10 +
 fs/btrfs/transaction.h            |    3 +
 fs/btrfs/tree-log.c               |   52 +-
 fs/btrfs/volumes.c                |  322 +++++++-
 fs/btrfs/volumes.h                |    7 +
 fs/btrfs/zoned.c                  | 1272 +++++++++++++++++++++++++++++
 fs/btrfs/zoned.h                  |  295 +++++++
 fs/iomap/direct-io.c              |   41 +-
 include/linux/bio.h               |    2 +
 include/linux/iomap.h             |    1 +
 include/uapi/linux/btrfs.h        |    1 +
 37 files changed, 3246 insertions(+), 118 deletions(-)
 create mode 100644 fs/btrfs/zoned.c
 create mode 100644 fs/btrfs/zoned.h
Signature · 1 line
2.27.0
#1 lore raw

PATCH v10 1/41 block: add bio_add_zone_append_page

From: Johannes Thumshirn <[email protected]>

Add bio_add_zone_append_page(), a wrapper around bio_add_hw_page() which is intended to be used by file systems that directly add pages to a bio instead of using bio_iov_iter_get_pages().

---
 block/bio.c         | 38 ++++++++++++++++++++++++++++++++++++++
 include/linux/bio.h |  2 ++
 2 files changed, 40 insertions(+)
block/bio.c +38 -0
index 58d765400226..c8943201c26c 100644
@@ -853,6 +853,44 @@ int bio_add_pc_page(struct request_queue *q, struct bio *bio,
 }
 EXPORT_SYMBOL(bio_add_pc_page);
 
+/**
+ * bio_add_zone_append_page - attempt to add page to zone-append bio
+ * @bio: destination bio
+ * @page: page to add
+ * @len: vec entry length
+ * @offset: vec entry offset
+ *
+ * Attempt to add a page to the bio_vec maplist of a bio that will be submitted
+ * for a zone-append request. This can fail for a number of reasons, such as the
+ * bio being full or the target block device is not a zoned block device or
+ * other limitations of the target block device. The target block device must
+ * allow bio's up to PAGE_SIZE, so it is always possible to add a single page
+ * to an empty bio.
+ *
+ * Returns: number of bytes added to the bio, or 0 in case of a failure.
+ */
+int bio_add_zone_append_page(struct bio *bio, struct page *page,
+			     unsigned int len, unsigned int offset)
+{
+	struct request_queue *q;
+	bool same_page = false;
+
+	if (WARN_ON_ONCE(bio_op(bio) != REQ_OP_ZONE_APPEND))
+		return 0;
+
+	if (WARN_ON_ONCE(!bio->bi_disk))
+		return 0;
+
+	q = bio->bi_disk->queue;
+
+	if (WARN_ON_ONCE(!blk_queue_is_zoned(q)))
+		return 0;
+
+	return bio_add_hw_page(q, bio, page, len, offset,
+			       queue_max_zone_append_sectors(q), &same_page);
+}
+EXPORT_SYMBOL_GPL(bio_add_zone_append_page);
+
 /**
  * __bio_try_merge_page - try appending data to an existing bvec.
  * @bio: destination bio
include/linux/bio.h +2 -0
index c6d765382926..7ef300cb4e9a 100644
@@ -442,6 +442,8 @@ void bio_chain(struct bio *, struct bio *);
 extern int bio_add_page(struct bio *, struct page *, unsigned int,unsigned int);
 extern int bio_add_pc_page(struct request_queue *, struct bio *, struct page *,
 			   unsigned int, unsigned int);
+int bio_add_zone_append_page(struct bio *bio, struct page *page,
+			     unsigned int len, unsigned int offset);
 bool __bio_try_merge_page(struct bio *bio, struct page *page,
 		unsigned int len, unsigned int off, bool *same_page);
 void __bio_add_page(struct bio *bio, struct page *page,
Signature · 1 line
2.27.0
Reviewed-by on v10 1/41
#2 lore raw

Re: [PATCH v10 01/41] block: add bio_add_zone_append_page

block/bio.c int bio_add_pc_page(struct request_queue *q, struct bio *bio, · line 882

Quoted 11 lines
block/bio.c +11 -0
+int bio_add_zone_append_page(struct bio *bio, struct page *page,
+			     unsigned int len, unsigned int offset)
+{
+	struct request_queue *q;
+	bool same_page = false;
+
+	if (WARN_ON_ONCE(bio_op(bio) != REQ_OP_ZONE_APPEND))
+		return 0;
+
+	if (WARN_ON_ONCE(!bio->bi_disk))
+		return 0;

Do we need this check? I'd rather just initialize q at declaration time and let the NULL pointer deref be the obvious sign for a grave programming error..

Except for that the patch looks good to me:

#3 lore raw

Re: [PATCH v10 01/41] block: add bio_add_zone_append_page

Quoted 3 lines · Christoph Hellwig

Do we need this check? I'd rather just initialize q at declaration time and let the NULL pointer deref be the obvious sign for a grave programming error..

I wanted to be a bit more gentle, but I don't mind the other way around as well.

#4 lore raw

PATCH v10 2/41 iomap: support REQ_OP_ZONE_APPEND

In reply to Naohiro Aota

A ZONE_APPEND bio must follow hardware restrictions (e.g. not exceeding max_zone_append_sectors) not to be split. bio_iov_iter_get_pages builds such restricted bio using __bio_iov_append_get_pages if bio_op(bio) == REQ_OP_ZONE_APPEND.

To utilize it, we need to set the bio_op before calling bio_iov_iter_get_pages(). This commit introduces IOMAP_F_ZONE_APPEND, so that iomap user can set the flag to indicate they want REQ_OP_ZONE_APPEND and restricted bio.

---
 fs/iomap/direct-io.c  | 41 +++++++++++++++++++++++++++++++++++------
 include/linux/iomap.h |  1 +
 2 files changed, 36 insertions(+), 6 deletions(-)
fs/iomap/direct-io.c +35 -6
index c1aafb2ab990..f04572a55a09 100644
@@ -200,6 +200,34 @@ iomap_dio_zero(struct iomap_dio *dio, struct iomap *iomap, loff_t pos,
 	iomap_dio_submit_bio(dio, iomap, bio, pos);
 }
 
+/*
+ * Figure out the bio's operation flags from the dio request, the
+ * mapping, and whether or not we want FUA.  Note that we can end up
+ * clearing the WRITE_FUA flag in the dio request.
+ */
+static inline unsigned int
+iomap_dio_bio_opflags(struct iomap_dio *dio, struct iomap *iomap, bool use_fua)
+{
+	unsigned int opflags = REQ_SYNC | REQ_IDLE;
+
+	if (!(dio->flags & IOMAP_DIO_WRITE)) {
+		WARN_ON_ONCE(iomap->flags & IOMAP_F_ZONE_APPEND);
+		return REQ_OP_READ;
+	}
+
+	if (iomap->flags & IOMAP_F_ZONE_APPEND)
+		opflags |= REQ_OP_ZONE_APPEND;
+	else
+		opflags |= REQ_OP_WRITE;
+
+	if (use_fua)
+		opflags |= REQ_FUA;
+	else
+		dio->flags &= ~IOMAP_DIO_WRITE_FUA;
+
+	return opflags;
+}
+
 static loff_t
 iomap_dio_bio_actor(struct inode *inode, loff_t pos, loff_t length,
 		struct iomap_dio *dio, struct iomap *iomap)
@@ -278,6 +306,13 @@ iomap_dio_bio_actor(struct inode *inode, loff_t pos, loff_t length,
 		bio->bi_private = dio;
 		bio->bi_end_io = iomap_dio_bio_end_io;
 
+		/*
+		 * Set the operation flags early so that bio_iov_iter_get_pages
+		 * can set up the page vector appropriately for a ZONE_APPEND
+		 * operation.
+		 */
+		bio->bi_opf = iomap_dio_bio_opflags(dio, iomap, use_fua);
+
 		ret = bio_iov_iter_get_pages(bio, dio->submit.iter);
 		if (unlikely(ret)) {
 			/*
@@ -292,14 +327,8 @@ iomap_dio_bio_actor(struct inode *inode, loff_t pos, loff_t length,
 
 		n = bio->bi_iter.bi_size;
 		if (dio->flags & IOMAP_DIO_WRITE) {
-			bio->bi_opf = REQ_OP_WRITE | REQ_SYNC | REQ_IDLE;
-			if (use_fua)
-				bio->bi_opf |= REQ_FUA;
-			else
-				dio->flags &= ~IOMAP_DIO_WRITE_FUA;
 			task_io_account_write(n);
 		} else {
-			bio->bi_opf = REQ_OP_READ;
 			if (dio->flags & IOMAP_DIO_DIRTY)
 				bio_set_pages_dirty(bio);
 		}
include/linux/iomap.h +1 -0
index 4d1d3c3469e9..1bccd1880d0d 100644
@@ -54,6 +54,7 @@ struct vm_fault;
 #define IOMAP_F_SHARED		0x04
 #define IOMAP_F_MERGED		0x08
 #define IOMAP_F_BUFFER_HEAD	0x10
+#define IOMAP_F_ZONE_APPEND	0x20
 
 /*
  * Flags set by the core iomap code during operations:
Signature · 1 line
2.27.0
#5 lore raw

Re: [PATCH v10 02/41] iomap: support REQ_OP_ZONE_APPEND

fs/iomap/direct-io.c iomap_dio_bio_actor(struct inode *inode, loff_t pos, loff_t length, · line 314

Quoted 11 lines
fs/iomap/direct-io.c +6 -0
 		struct iomap_dio *dio, struct iomap *iomap)
@@ -278,6 +306,13 @@ iomap_dio_bio_actor(struct inode *inode, loff_t pos, loff_t length,
 		bio->bi_private = dio;
 		bio->bi_end_io = iomap_dio_bio_end_io;
 
+		/*
+		 * Set the operation flags early so that bio_iov_iter_get_pages
+		 * can set up the page vector appropriately for a ZONE_APPEND
+		 * operation.
+		 */
+		bio->bi_opf = iomap_dio_bio_opflags(dio, iomap, use_fua);

We could just calculate the flags once before the loop if we touch this anyway.

But otherwise this looks ok to me.

#6 lore raw

Re: [PATCH v10 02/41] iomap: support REQ_OP_ZONE_APPEND

In reply to Naohiro Aota
Quoted 21 lines · Naohiro Aota

A ZONE_APPEND bio must follow hardware restrictions (e.g. not exceeding max_zone_append_sectors) not to be split. bio_iov_iter_get_pages builds such restricted bio using __bio_iov_append_get_pages if bio_op(bio) == REQ_OP_ZONE_APPEND.

To utilize it, we need to set the bio_op before calling bio_iov_iter_get_pages(). This commit introduces IOMAP_F_ZONE_APPEND, so that iomap user can set the flag to indicate they want REQ_OP_ZONE_APPEND and restricted bio.

---
 fs/iomap/direct-io.c  | 41 +++++++++++++++++++++++++++++++++++------
 include/linux/iomap.h |  1 +
 2 files changed, 36 insertions(+), 6 deletions(-)
fs/iomap/direct-io.c +0 -0
index c1aafb2ab990..f04572a55a09 100644

fs/iomap/direct-io.c iomap_dio_zero(struct iomap_dio *dio, struct iomap *iomap, loff_t pos, · line 209

Quoted 10 lines
fs/iomap/direct-io.c +7 -0
@@ -200,6 +200,34 @@ iomap_dio_zero(struct iomap_dio *dio, struct iomap *iomap, loff_t pos,
 	iomap_dio_submit_bio(dio, iomap, bio, pos);
 }
 
+/*
+ * Figure out the bio's operation flags from the dio request, the
+ * mapping, and whether or not we want FUA.  Note that we can end up
+ * clearing the WRITE_FUA flag in the dio request.
+ */
+static inline unsigned int
+iomap_dio_bio_opflags(struct iomap_dio *dio, struct iomap *iomap, bool use_fua)

Hmm, just to check my understanding of what iomap has to do to support all this:

When we're wanting to use a ZONE_APPEND command, the @iomap structure has to have IOMAP_F_ZONE_APPEND set in iomap->flags, iomap->type is set to IOMAP_MAPPED, but what should iomap->addr be set to?

I gather from what I see in zonefs and the relevant NVME proposal that iomap->addr should be set to the (byte) address of the zone we want to append to? And if we do that, then bio->bi_iter.bi_sector will be set to sector address of iomap->addr, right?

(I got lost trying to figure out how btrfs sets ->addr for appends.)

Then when the IO completes, the block layer sets bio->bi_iter.bi_sector to wherever the drive told it that it actually wrote the bio, right?

If that's true, then that implies that need_zeroout must always be false for an append operation, right? Does that also mean that the directio request has to be aligned to an fs block and not just the sector size?

Can userspace send a directio append that crosses a zone boundary? If so, what happens if a direct append to a lower address fails but a direct append to a higher address succeeds?

Quoted 25 lines
fs/iomap/direct-io.c +21 -0
+{
+	unsigned int opflags = REQ_SYNC | REQ_IDLE;
+
+	if (!(dio->flags & IOMAP_DIO_WRITE)) {
+		WARN_ON_ONCE(iomap->flags & IOMAP_F_ZONE_APPEND);
+		return REQ_OP_READ;
+	}
+
+	if (iomap->flags & IOMAP_F_ZONE_APPEND)
+		opflags |= REQ_OP_ZONE_APPEND;
+	else
+		opflags |= REQ_OP_WRITE;
+
+	if (use_fua)
+		opflags |= REQ_FUA;
+	else
+		dio->flags &= ~IOMAP_DIO_WRITE_FUA;
+
+	return opflags;
+}
+
 static loff_t
 iomap_dio_bio_actor(struct inode *inode, loff_t pos, loff_t length,
 		struct iomap_dio *dio, struct iomap *iomap)

fs/iomap/direct-io.c iomap_dio_bio_actor(struct inode *inode, loff_t pos, loff_t length, · line 314

Quoted 9 lines
fs/iomap/direct-io.c +6 -0
@@ -278,6 +306,13 @@ iomap_dio_bio_actor(struct inode *inode, loff_t pos, loff_t length,
 		bio->bi_private = dio;
 		bio->bi_end_io = iomap_dio_bio_end_io;
 
+		/*
+		 * Set the operation flags early so that bio_iov_iter_get_pages
+		 * can set up the page vector appropriately for a ZONE_APPEND
+		 * operation.
+		 */
+		bio->bi_opf = iomap_dio_bio_opflags(dio, iomap, use_fua);

I'm also vaguely wondering how to communicate the write location back to the filesystem when the bio completes? btrfs handles the bio completion completely so it doesn't have a problem, but for other filesystems (cough future xfs cough) either we'd have to add a new callback for append operations; or I guess everyone could hook the bio endio.

Admittedly that's not really your problem, and for all I know hch is already working on this.

--D

Quoted 34 lines
fs/iomap/direct-io.c +1 -6
+
 		ret = bio_iov_iter_get_pages(bio, dio->submit.iter);
 		if (unlikely(ret)) {
 			/*
@@ -292,14 +327,8 @@ iomap_dio_bio_actor(struct inode *inode, loff_t pos, loff_t length,
 
 		n = bio->bi_iter.bi_size;
 		if (dio->flags & IOMAP_DIO_WRITE) {
-			bio->bi_opf = REQ_OP_WRITE | REQ_SYNC | REQ_IDLE;
-			if (use_fua)
-				bio->bi_opf |= REQ_FUA;
-			else
-				dio->flags &= ~IOMAP_DIO_WRITE_FUA;
 			task_io_account_write(n);
 		} else {
-			bio->bi_opf = REQ_OP_READ;
 			if (dio->flags & IOMAP_DIO_DIRTY)
 				bio_set_pages_dirty(bio);
 		}
include/linux/iomap.h +1 -0
index 4d1d3c3469e9..1bccd1880d0d 100644
@@ -54,6 +54,7 @@ struct vm_fault;
 #define IOMAP_F_SHARED		0x04
 #define IOMAP_F_MERGED		0x08
 #define IOMAP_F_BUFFER_HEAD	0x10
+#define IOMAP_F_ZONE_APPEND	0x20
 
 /*
  * Flags set by the core iomap code during operations:
Signature · 1 line
2.27.0
#7 lore raw

Re: [PATCH v10 02/41] iomap: support REQ_OP_ZONE_APPEND

Quoted 50 lines · Darrick J. Wong
Quoted 21 lines · Naohiro Aota

A ZONE_APPEND bio must follow hardware restrictions (e.g. not exceeding max_zone_append_sectors) not to be split. bio_iov_iter_get_pages builds such restricted bio using __bio_iov_append_get_pages if bio_op(bio) == REQ_OP_ZONE_APPEND.

To utilize it, we need to set the bio_op before calling bio_iov_iter_get_pages(). This commit introduces IOMAP_F_ZONE_APPEND, so that iomap user can set the flag to indicate they want REQ_OP_ZONE_APPEND and restricted bio.

---
 fs/iomap/direct-io.c  | 41 +++++++++++++++++++++++++++++++++++------
 include/linux/iomap.h |  1 +
 2 files changed, 36 insertions(+), 6 deletions(-)
fs/iomap/direct-io.c +0 -0
index c1aafb2ab990..f04572a55a09 100644

fs/iomap/direct-io.c iomap_dio_zero(struct iomap_dio *dio, struct iomap *iomap, loff_t pos, · line 209

Quoted 10 lines
fs/iomap/direct-io.c +7 -0
@@ -200,6 +200,34 @@ iomap_dio_zero(struct iomap_dio *dio, struct iomap *iomap, loff_t pos,
 	iomap_dio_submit_bio(dio, iomap, bio, pos);
 }
 
+/*
+ * Figure out the bio's operation flags from the dio request, the
+ * mapping, and whether or not we want FUA.  Note that we can end up
+ * clearing the WRITE_FUA flag in the dio request.
+ */
+static inline unsigned int
+iomap_dio_bio_opflags(struct iomap_dio *dio, struct iomap *iomap, bool use_fua)

Hmm, just to check my understanding of what iomap has to do to support all this:

When we're wanting to use a ZONE_APPEND command, the @iomap structure has to have IOMAP_F_ZONE_APPEND set in iomap->flags, iomap->type is set to IOMAP_MAPPED, but what should iomap->addr be set to?

I gather from what I see in zonefs and the relevant NVME proposal that iomap->addr should be set to the (byte) address of the zone we want to append to? And if we do that, then bio->bi_iter.bi_sector will be set to sector address of iomap->addr, right?

(I got lost trying to figure out how btrfs sets ->addr for appends.)

Then when the IO completes, the block layer sets bio->bi_iter.bi_sector to wherever the drive told it that it actually wrote the bio, right?

Quoted 7 lines

If that's true, then that implies that need_zeroout must always be false for an append operation, right? Does that also mean that the directio request has to be aligned to an fs block and not just the sector size?

Can userspace send a directio append that crosses a zone boundary? If so, what happens if a direct append to a lower address fails but a direct append to a higher address succeeds?

Bleh, vim tomfoolery == missing sentence. Change the above paragraph to read:

Can userspace send a directio append that crosses a zone boundary? Can we issue multiple bios for a single append write? What happens if a direct append to a lower address fails but a direct append to a higher address succeeds?

--D

Quoted 80 lines
Quoted 25 lines
fs/iomap/direct-io.c +21 -0
+{
+	unsigned int opflags = REQ_SYNC | REQ_IDLE;
+
+	if (!(dio->flags & IOMAP_DIO_WRITE)) {
+		WARN_ON_ONCE(iomap->flags & IOMAP_F_ZONE_APPEND);
+		return REQ_OP_READ;
+	}
+
+	if (iomap->flags & IOMAP_F_ZONE_APPEND)
+		opflags |= REQ_OP_ZONE_APPEND;
+	else
+		opflags |= REQ_OP_WRITE;
+
+	if (use_fua)
+		opflags |= REQ_FUA;
+	else
+		dio->flags &= ~IOMAP_DIO_WRITE_FUA;
+
+	return opflags;
+}
+
 static loff_t
 iomap_dio_bio_actor(struct inode *inode, loff_t pos, loff_t length,
 		struct iomap_dio *dio, struct iomap *iomap)

fs/iomap/direct-io.c iomap_dio_bio_actor(struct inode *inode, loff_t pos, loff_t length, · line 314

Quoted 9 lines
fs/iomap/direct-io.c +6 -0
@@ -278,6 +306,13 @@ iomap_dio_bio_actor(struct inode *inode, loff_t pos, loff_t length,
 		bio->bi_private = dio;
 		bio->bi_end_io = iomap_dio_bio_end_io;
 
+		/*
+		 * Set the operation flags early so that bio_iov_iter_get_pages
+		 * can set up the page vector appropriately for a ZONE_APPEND
+		 * operation.
+		 */
+		bio->bi_opf = iomap_dio_bio_opflags(dio, iomap, use_fua);

I'm also vaguely wondering how to communicate the write location back to the filesystem when the bio completes? btrfs handles the bio completion completely so it doesn't have a problem, but for other filesystems (cough future xfs cough) either we'd have to add a new callback for append operations; or I guess everyone could hook the bio endio.

Admittedly that's not really your problem, and for all I know hch is already working on this.

--D

Quoted 34 lines
fs/iomap/direct-io.c +1 -6
+
 		ret = bio_iov_iter_get_pages(bio, dio->submit.iter);
 		if (unlikely(ret)) {
 			/*
@@ -292,14 +327,8 @@ iomap_dio_bio_actor(struct inode *inode, loff_t pos, loff_t length,
 
 		n = bio->bi_iter.bi_size;
 		if (dio->flags & IOMAP_DIO_WRITE) {
-			bio->bi_opf = REQ_OP_WRITE | REQ_SYNC | REQ_IDLE;
-			if (use_fua)
-				bio->bi_opf |= REQ_FUA;
-			else
-				dio->flags &= ~IOMAP_DIO_WRITE_FUA;
 			task_io_account_write(n);
 		} else {
-			bio->bi_opf = REQ_OP_READ;
 			if (dio->flags & IOMAP_DIO_DIRTY)
 				bio_set_pages_dirty(bio);
 		}
include/linux/iomap.h +1 -0
index 4d1d3c3469e9..1bccd1880d0d 100644
@@ -54,6 +54,7 @@ struct vm_fault;
 #define IOMAP_F_SHARED		0x04
 #define IOMAP_F_MERGED		0x08
 #define IOMAP_F_BUFFER_HEAD	0x10
+#define IOMAP_F_ZONE_APPEND	0x20
 
 /*
  * Flags set by the core iomap code during operations:
Signature · 1 line
2.27.0
#8 lore raw

Re: [PATCH v10 02/41] iomap: support REQ_OP_ZONE_APPEND

In reply to Darrick J. Wong
Quoted 8 lines · Darrick J. Wong

When we're wanting to use a ZONE_APPEND command, the @iomap structure has to have IOMAP_F_ZONE_APPEND set in iomap->flags, iomap->type is set to IOMAP_MAPPED, but what should iomap->addr be set to?

I gather from what I see in zonefs and the relevant NVME proposal that iomap->addr should be set to the (byte) address of the zone we want to append to? And if we do that, then bio->bi_iter.bi_sector will be set to sector address of iomap->addr, right?

Yes.

Quoted 2 lines

Then when the IO completes, the block layer sets bio->bi_iter.bi_sector to wherever the drive told it that it actually wrote the bio, right?

Yes.

Quoted 3 lines

If that's true, then that implies that need_zeroout must always be false for an append operation, right? Does that also mean that the directio request has to be aligned to an fs block and not just the sector size?

I think so, yes.

Quoted 3 lines

Can userspace send a directio append that crosses a zone boundary? If so, what happens if a direct append to a lower address fails but a direct append to a higher address succeeds?

Userspace doesn't know about zone boundaries. It can send I/O larger than a zone, but the file system has to split it into multiple I/Os just like when it has to cross and AG boundary in XFS.

Quoted 8 lines

I'm also vaguely wondering how to communicate the write location back to the filesystem when the bio completes? btrfs handles the bio completion completely so it doesn't have a problem, but for other filesystems (cough future xfs cough) either we'd have to add a new callback for append operations; or I guess everyone could hook the bio endio.

Admittedly that's not really your problem, and for all I know hch is already working on this.

I think any non-trivial file system needs to override the bio completion handler for writes anyway, so this seems reasonable. It might be worth documenting, though.

Reviewed-by on v10 2/41
#9 lore raw

Re: [PATCH v10 02/41] iomap: support REQ_OP_ZONE_APPEND

In reply to Naohiro Aota
Quoted 11 lines · Naohiro Aota

A ZONE_APPEND bio must follow hardware restrictions (e.g. not exceeding max_zone_append_sectors) not to be split. bio_iov_iter_get_pages builds such restricted bio using __bio_iov_append_get_pages if bio_op(bio) == REQ_OP_ZONE_APPEND.

To utilize it, we need to set the bio_op before calling bio_iov_iter_get_pages(). This commit introduces IOMAP_F_ZONE_APPEND, so that iomap user can set the flag to indicate they want REQ_OP_ZONE_APPEND and restricted bio.

Christoph's answers seem reasonable to me, so

Er... do you want me to take this one via the iomap tree?

--D

Quoted 88 lines
---
 fs/iomap/direct-io.c  | 41 +++++++++++++++++++++++++++++++++++------
 include/linux/iomap.h |  1 +
 2 files changed, 36 insertions(+), 6 deletions(-)
fs/iomap/direct-io.c +35 -6
index c1aafb2ab990..f04572a55a09 100644
@@ -200,6 +200,34 @@ iomap_dio_zero(struct iomap_dio *dio, struct iomap *iomap, loff_t pos,
 	iomap_dio_submit_bio(dio, iomap, bio, pos);
 }
 
+/*
+ * Figure out the bio's operation flags from the dio request, the
+ * mapping, and whether or not we want FUA.  Note that we can end up
+ * clearing the WRITE_FUA flag in the dio request.
+ */
+static inline unsigned int
+iomap_dio_bio_opflags(struct iomap_dio *dio, struct iomap *iomap, bool use_fua)
+{
+	unsigned int opflags = REQ_SYNC | REQ_IDLE;
+
+	if (!(dio->flags & IOMAP_DIO_WRITE)) {
+		WARN_ON_ONCE(iomap->flags & IOMAP_F_ZONE_APPEND);
+		return REQ_OP_READ;
+	}
+
+	if (iomap->flags & IOMAP_F_ZONE_APPEND)
+		opflags |= REQ_OP_ZONE_APPEND;
+	else
+		opflags |= REQ_OP_WRITE;
+
+	if (use_fua)
+		opflags |= REQ_FUA;
+	else
+		dio->flags &= ~IOMAP_DIO_WRITE_FUA;
+
+	return opflags;
+}
+
 static loff_t
 iomap_dio_bio_actor(struct inode *inode, loff_t pos, loff_t length,
 		struct iomap_dio *dio, struct iomap *iomap)
@@ -278,6 +306,13 @@ iomap_dio_bio_actor(struct inode *inode, loff_t pos, loff_t length,
 		bio->bi_private = dio;
 		bio->bi_end_io = iomap_dio_bio_end_io;
 
+		/*
+		 * Set the operation flags early so that bio_iov_iter_get_pages
+		 * can set up the page vector appropriately for a ZONE_APPEND
+		 * operation.
+		 */
+		bio->bi_opf = iomap_dio_bio_opflags(dio, iomap, use_fua);
+
 		ret = bio_iov_iter_get_pages(bio, dio->submit.iter);
 		if (unlikely(ret)) {
 			/*
@@ -292,14 +327,8 @@ iomap_dio_bio_actor(struct inode *inode, loff_t pos, loff_t length,
 
 		n = bio->bi_iter.bi_size;
 		if (dio->flags & IOMAP_DIO_WRITE) {
-			bio->bi_opf = REQ_OP_WRITE | REQ_SYNC | REQ_IDLE;
-			if (use_fua)
-				bio->bi_opf |= REQ_FUA;
-			else
-				dio->flags &= ~IOMAP_DIO_WRITE_FUA;
 			task_io_account_write(n);
 		} else {
-			bio->bi_opf = REQ_OP_READ;
 			if (dio->flags & IOMAP_DIO_DIRTY)
 				bio_set_pages_dirty(bio);
 		}
include/linux/iomap.h +1 -0
index 4d1d3c3469e9..1bccd1880d0d 100644
@@ -54,6 +54,7 @@ struct vm_fault;
 #define IOMAP_F_SHARED		0x04
 #define IOMAP_F_MERGED		0x08
 #define IOMAP_F_BUFFER_HEAD	0x10
+#define IOMAP_F_ZONE_APPEND	0x20
 
 /*
  * Flags set by the core iomap code during operations:
Signature · 1 line
2.27.0
#10 lore raw

Re: [PATCH v10 02/41] iomap: support REQ_OP_ZONE_APPEND

Quoted 11 lines · Darrick J. Wong
Quoted 10 lines · Naohiro Aota

A ZONE_APPEND bio must follow hardware restrictions (e.g. not exceeding max_zone_append_sectors) not to be split. bio_iov_iter_get_pages builds such restricted bio using __bio_iov_append_get_pages if bio_op(bio) == REQ_OP_ZONE_APPEND.

To utilize it, we need to set the bio_op before calling bio_iov_iter_get_pages(). This commit introduces IOMAP_F_ZONE_APPEND, so that iomap user can set the flag to indicate they want REQ_OP_ZONE_APPEND and restricted bio.

Quoted 7 lines
Quoted 1 line

Christoph's answers seem reasonable to me, so

Er... do you want me to take this one via the iomap tree?

Ahm probably yes, but we have an update addressing Christoph's coming with the next version series.

#11 lore raw

Re: [PATCH v10 02/41] iomap: support REQ_OP_ZONE_APPEND

In reply to Naohiro Aota

Btw, another thing I noticed:

when using io_uring to submit a write to btrfs that ends up using Zone Append we'll hit the

	if (WARN_ON_ONCE(is_bvec))
		return -EINVAL;

case in bio_iov_iter_get_pages with the changes in this series.

#12 lore raw

Re: [PATCH v10 02/41] iomap: support REQ_OP_ZONE_APPEND

Quoted 9 lines · Christoph Hellwig

Btw, another thing I noticed:

when using io_uring to submit a write to btrfs that ends up using Zone Append we'll hit the

	if (WARN_ON_ONCE(is_bvec))
		return -EINVAL;

case in bio_iov_iter_get_pages with the changes in this series.

Yes this warning is totally bogus. It was in there from the beginning of the zone-append series and I have no idea why I didn't kill it.

IIRC Chaitanya had a patch in his nvmet zoned series removing it.

#13 lore raw

Re: [PATCH v10 02/41] iomap: support REQ_OP_ZONE_APPEND

Quoted 9 lines · Johannes Thumshirn
Quoted 8 lines · Christoph Hellwig

Btw, another thing I noticed:

when using io_uring to submit a write to btrfs that ends up using Zone Append we'll hit the

	if (WARN_ON_ONCE(is_bvec))
		return -EINVAL;
Quoted 6 lines
Quoted 1 line

case in bio_iov_iter_get_pages with the changes in this series.

Yes this warning is totally bogus. It was in there from the beginning of the zone-append series and I have no idea why I didn't kill it.

IIRC Chaitanya had a patch in his nvmet zoned series removing it.

Yes, but it is wrong. What we need is a version of __bio_iov_bvec_add_pages that takes the hardware limits into account.

#14 lore raw

Re: [PATCH v10 02/41] iomap: support REQ_OP_ZONE_APPEND

Quoted 12 lines · On 09/12/2020 11:10, [email protected] wrote:
Quoted 11 lines · Johannes Thumshirn
Quoted 9 lines · Christoph Hellwig

Btw, another thing I noticed:

when using io_uring to submit a write to btrfs that ends up using Zone Append we'll hit the

	if (WARN_ON_ONCE(is_bvec))
		return -EINVAL;

case in bio_iov_iter_get_pages with the changes in this series.

Quoted 8 lines
Quoted 4 lines

Yes this warning is totally bogus. It was in there from the beginning of the zone-append series and I have no idea why I didn't kill it.

IIRC Chaitanya had a patch in his nvmet zoned series removing it.

Yes, but it is wrong. What we need is a version of __bio_iov_bvec_add_pages that takes the hardware limits into account.

Ah now I understand the situation, I'm on it.

#15 lore raw

Re: [PATCH v10 02/41] iomap: support REQ_OP_ZONE_APPEND

Quoted 16 lines · Johannes Thumshirn
Quoted 15 lines · On 09/12/2020 11:10, [email protected] wrote:
Quoted 14 lines · Johannes Thumshirn
Quoted 9 lines · Christoph Hellwig

Btw, another thing I noticed:

when using io_uring to submit a write to btrfs that ends up using Zone Append we'll hit the

	if (WARN_ON_ONCE(is_bvec))
		return -EINVAL;

case in bio_iov_iter_get_pages with the changes in this series.

Yes this warning is totally bogus. It was in there from the beginning of the zone-append series and I have no idea why I didn't kill it.

Quoted 8 lines
Quoted 5 lines
Quoted 1 line

IIRC Chaitanya had a patch in his nvmet zoned series removing it.

Yes, but it is wrong. What we need is a version of __bio_iov_bvec_add_pages that takes the hardware limits into account.

Ah now I understand the situation, I'm on it.

OK got something, just need to test it.

#16 lore raw

Re: [PATCH v10 02/41] iomap: support REQ_OP_ZONE_APPEND

Quoted 19 lines · Johannes Thumshirn
Quoted 18 lines · Johannes Thumshirn
Quoted 17 lines · On 09/12/2020 11:10, [email protected] wrote:
Quoted 15 lines · Johannes Thumshirn
Quoted 9 lines · Christoph Hellwig

Btw, another thing I noticed:

when using io_uring to submit a write to btrfs that ends up using Zone Append we'll hit the

	if (WARN_ON_ONCE(is_bvec))
		return -EINVAL;

case in bio_iov_iter_get_pages with the changes in this series.

Yes this warning is totally bogus. It was in there from the beginning of the zone-append series and I have no idea why I didn't kill it.

IIRC Chaitanya had a patch in his nvmet zoned series removing it.

Quoted 9 lines
Quoted 6 lines
Quoted 3 lines

Yes, but it is wrong. What we need is a version of __bio_iov_bvec_add_pages that takes the hardware limits into account.

Ah now I understand the situation, I'm on it.

OK got something, just need to test it.

I just ran tests with my solution and to verify it worked as expected I ran the test without it. Interestingly the WARN_ON() didn't trigger for me. Here's the fio command line I've used:

fio --ioengine io_uring --rw readwrite --bs 1M --size 1G --time_based   \
    --runtime 1m --filename /mnt/test/io_uring --name io_uring-test     \
    --direct 1 --numjobs $NPROC

I did verify it's using zone append though.

What did you use to trigger the warning?

#17 lore raw

Re: [PATCH v10 02/41] iomap: support REQ_OP_ZONE_APPEND

Johhanes/Christoph,

Quoted 10 lines · Johannes Thumshirn
Quoted 9 lines · Johannes Thumshirn
Quoted 8 lines · Johannes Thumshirn
Quoted 7 lines · On 09/12/2020 11:10, [email protected] wrote:
Quoted 6 lines · Johannes Thumshirn
Quoted 5 lines · Christoph Hellwig

Btw, another thing I noticed:

when using io_uring to submit a write to btrfs that ends up using Zone Append we'll hit the

Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 4 lines
	if (WARN_ON_ONCE(is_bvec))
		return -EINVAL;

case in bio_iov_iter_get_pages with the changes in this series.

Yes this warning is totally bogus. It was in there from the beginning of the zone-append series and I have no idea why I didn't kill it.

IIRC Chaitanya had a patch in his nvmet zoned series removing it.

Even though that patch is not needed I've tested that with the NVMeOF backend to build bios from bvecs with bio_iov_iter_get_pages(), I can still send that patch, please confirm.

#18 lore raw

Re: [PATCH v10 02/41] iomap: support REQ_OP_ZONE_APPEND

Quoted 17 lines · Chaitanya Kulkarni

Johhanes/Christoph,

Quoted 14 lines · Johannes Thumshirn
Quoted 13 lines · Johannes Thumshirn
Quoted 12 lines · Johannes Thumshirn
Quoted 11 lines · On 09/12/2020 11:10, [email protected] wrote:
Quoted 10 lines · Johannes Thumshirn
Quoted 9 lines · Christoph Hellwig

Btw, another thing I noticed:

when using io_uring to submit a write to btrfs that ends up using Zone Append we'll hit the

	if (WARN_ON_ONCE(is_bvec))
		return -EINVAL;

case in bio_iov_iter_get_pages with the changes in this series.

Quoted 9 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines

Yes this warning is totally bogus. It was in there from the beginning of the zone-append series and I have no idea why I didn't kill it.

IIRC Chaitanya had a patch in his nvmet zoned series removing it.

Even though that patch is not needed I've tested that with the NVMeOF backend to build bios from bvecs with bio_iov_iter_get_pages(), I can still send that patch, please confirm.

I have the following locally, but I fail to call it in my tests:

commit ea93c91a70204a23ebf9e22b19fbf8add644e12e

Author: Johannes Thumshirn <[email protected]>
Date:   Wed Dec 9 23:26:38 2020 +0900

block: provide __bio_iov_bvec_add_append_pages

block/bio.c +29 -3
index 5c6982902330..dc8178ca796f 100644
@@ -992,6 +992,31 @@ void bio_release_pages(struct bio *bio, bool mark_dirty)
 }
 EXPORT_SYMBOL_GPL(bio_release_pages);
 
+static int __bio_iov_bvec_add_append_pages(struct bio *bio,
+                                          struct iov_iter *iter)
+{
+       const struct bio_vec *bv = iter->bvec;
+       struct request_queue *q = bio->bi_disk->queue;
+       unsigned int max_append_sectors = queue_max_zone_append_sectors(q);
+       unsigned int len;
+       size_t size;
+
+       if (WARN_ON_ONCE(!max_append_sectors))
+               return -EINVAL;
+
+       if (WARN_ON_ONCE(iter->iov_offset > bv->bv_len))
+               return -EINVAL;
+
+       len = min_t(size_t, bv->bv_len - iter->iov_offset, iter->count);
+       size = bio_add_hw_page(q, bio, bv->bv_page, len,
+                              bv->bv_offset + iter->iov_offset,
+                              max_append_sectors, false);
+       if (unlikely(size != len))
+               return -EINVAL;
+       iov_iter_advance(iter, size);
+       return 0;
+}
+
 static int __bio_iov_bvec_add_pages(struct bio *bio, struct iov_iter *iter)
 {
        const struct bio_vec *bv = iter->bvec;
@@ -1142,9 +1167,10 @@ int bio_iov_iter_get_pages(struct bio *bio, struct iov_iter *iter)
 
        do {
                if (bio_op(bio) == REQ_OP_ZONE_APPEND) {
-                       if (WARN_ON_ONCE(is_bvec))
-                               return -EINVAL;
-                       ret = __bio_iov_append_get_pages(bio, iter);
+                       if (is_bvec)
+                               ret = __bio_iov_bvec_add_append_pages(bio, iter);
+                       else
+                               ret = __bio_iov_append_get_pages(bio, iter);
                } else {
                        if (is_bvec)
                                ret = __bio_iov_bvec_add_pages(bio, iter);

It's basically __bio_iov_bvec_add_pages() respecting the max_zone_sectors queue limit.

#19 lore raw

PATCH v10 3/41 btrfs: introduce ZONED feature flag

In reply to Naohiro Aota

This patch introduces the ZONED incompat flag. The flag indicates that the volume management will satisfy the constraints imposed by host-managed zoned block devices.

---
 fs/btrfs/sysfs.c           | 2 ++
 include/uapi/linux/btrfs.h | 1 +
 2 files changed, 3 insertions(+)
fs/btrfs/sysfs.c +2 -0
index 279d9262b676..828006020bbd 100644
@@ -263,6 +263,7 @@ BTRFS_FEAT_ATTR_INCOMPAT(no_holes, NO_HOLES);
 BTRFS_FEAT_ATTR_INCOMPAT(metadata_uuid, METADATA_UUID);
 BTRFS_FEAT_ATTR_COMPAT_RO(free_space_tree, FREE_SPACE_TREE);
 BTRFS_FEAT_ATTR_INCOMPAT(raid1c34, RAID1C34);
+BTRFS_FEAT_ATTR_INCOMPAT(zoned, ZONED);
 
 static struct attribute *btrfs_supported_feature_attrs[] = {
 	BTRFS_FEAT_ATTR_PTR(mixed_backref),
@@ -278,6 +279,7 @@ static struct attribute *btrfs_supported_feature_attrs[] = {
 	BTRFS_FEAT_ATTR_PTR(metadata_uuid),
 	BTRFS_FEAT_ATTR_PTR(free_space_tree),
 	BTRFS_FEAT_ATTR_PTR(raid1c34),
+	BTRFS_FEAT_ATTR_PTR(zoned),
 	NULL
 };
 
include/uapi/linux/btrfs.h +1 -0
index 2c39d15a2beb..5df73001aad4 100644
@@ -307,6 +307,7 @@ struct btrfs_ioctl_fs_info_args {
 #define BTRFS_FEATURE_INCOMPAT_NO_HOLES		(1ULL << 9)
 #define BTRFS_FEATURE_INCOMPAT_METADATA_UUID	(1ULL << 10)
 #define BTRFS_FEATURE_INCOMPAT_RAID1C34		(1ULL << 11)
+#define BTRFS_FEATURE_INCOMPAT_ZONED		(1ULL << 12)
 
 struct btrfs_ioctl_feature_flags {
 	__u64 compat_flags;
Signature · 1 line
2.27.0
#20 lore raw

Re: [PATCH v10 03/41] btrfs: introduce ZONED feature flag

Quoted 18 lines · Naohiro Aota

This patch introduces the ZONED incompat flag. The flag indicates that the volume management will satisfy the constraints imposed by host-managed zoned block devices.

---
 fs/btrfs/sysfs.c           | 2 ++
 include/uapi/linux/btrfs.h | 1 +
 2 files changed, 3 insertions(+)
fs/btrfs/sysfs.c +0 -0
index 279d9262b676..828006020bbd 100644

fs/btrfs/sysfs.c BTRFS_FEAT_ATTR_INCOMPAT(no_holes, NO_HOLES); · line 268

Quoted 6 lines
fs/btrfs/sysfs.c +2 -0
@@ -263,6 +263,7 @@ BTRFS_FEAT_ATTR_INCOMPAT(no_holes, NO_HOLES);
 BTRFS_FEAT_ATTR_INCOMPAT(metadata_uuid, METADATA_UUID);
 BTRFS_FEAT_ATTR_COMPAT_RO(free_space_tree, FREE_SPACE_TREE);
 BTRFS_FEAT_ATTR_INCOMPAT(raid1c34, RAID1C34);
+BTRFS_FEAT_ATTR_INCOMPAT(zoned, ZONED);
 
+	BTRFS_FEAT_ATTR_PTR(zoned),

As we're going to add zoned support incrementally, we can't advertise the support in sysfs until it's feature complete. Until then it's going to be under CONFIG_BTRFS_DEBUG. This has been folded to this patch and changelog updated.

#21 lore raw

PATCH v10 4/41 btrfs: get zone information of zoned block devices

In reply to Naohiro Aota

If a zoned block device is found, get its zone information (number of zones and zone size) using the new helper function btrfs_get_dev_zone_info(). To avoid costly run-time zone report commands to test the device zones type during block allocation, attach the seq_zones bitmap to the device structure to indicate if a zone is sequential or accept random writes. Also it attaches the empty_zones bitmap to indicate if a zone is empty or not.

This patch also introduces the helper function btrfs_dev_is_sequential() to test if the zone storing a block is a sequential write required zone and btrfs_dev_is_empty_zone() to test if the zone is a empty zone.

---
 fs/btrfs/Makefile      |   1 +
 fs/btrfs/dev-replace.c |   5 ++
 fs/btrfs/super.c       |   5 ++
 fs/btrfs/volumes.c     |  19 ++++-
 fs/btrfs/volumes.h     |   4 +
 fs/btrfs/zoned.c       | 182 +++++++++++++++++++++++++++++++++++++++++
 fs/btrfs/zoned.h       |  91 +++++++++++++++++++++
 7 files changed, 305 insertions(+), 2 deletions(-)
 create mode 100644 fs/btrfs/zoned.c
 create mode 100644 fs/btrfs/zoned.h
fs/btrfs/Makefile +1 -0
index e738f6206ea5..0497fdc37f90 100644
@@ -16,6 +16,7 @@ btrfs-y += super.o ctree.o extent-tree.o print-tree.o root-tree.o dir-item.o \
 btrfs-$(CONFIG_BTRFS_FS_POSIX_ACL) += acl.o
 btrfs-$(CONFIG_BTRFS_FS_CHECK_INTEGRITY) += check-integrity.o
 btrfs-$(CONFIG_BTRFS_FS_REF_VERIFY) += ref-verify.o
+btrfs-$(CONFIG_BLK_DEV_ZONED) += zoned.o
 
 btrfs-$(CONFIG_BTRFS_FS_RUN_SANITY_TESTS) += tests/free-space-tests.o \
 	tests/extent-buffer-tests.o tests/btrfs-tests.o \
fs/btrfs/dev-replace.c +5 -0
index 20ce1970015f..6f6d77224c2b 100644
@@ -21,6 +21,7 @@
 #include "rcu-string.h"
 #include "dev-replace.h"
 #include "sysfs.h"
+#include "zoned.h"
 
 /*
  * Device replace overview
@@ -291,6 +292,10 @@ static int btrfs_init_dev_replace_tgtdev(struct btrfs_fs_info *fs_info,
 	set_blocksize(device->bdev, BTRFS_BDEV_BLOCKSIZE);
 	device->fs_devices = fs_info->fs_devices;
 
+	ret = btrfs_get_dev_zone_info(device);
+	if (ret)
+		goto error;
+
 	mutex_lock(&fs_info->fs_devices->device_list_mutex);
 	list_add(&device->dev_list, &fs_info->fs_devices->devices);
 	fs_info->fs_devices->num_devices++;
fs/btrfs/super.c +5 -0
index 8840a4fa81eb..ed55014fd1bd 100644
@@ -2462,6 +2462,11 @@ static void __init btrfs_print_mod_info(void)
 #endif
 #ifdef CONFIG_BTRFS_FS_REF_VERIFY
 			", ref-verify=on"
+#endif
+#ifdef CONFIG_BLK_DEV_ZONED
+			", zoned=yes"
+#else
+			", zoned=no"
 #endif
 			;
 	pr_info("Btrfs loaded, crc32c=%s%s\n", crc32c_impl(), options);
fs/btrfs/volumes.c +17 -2
index 58b9c419a2b6..e787bf89f761 100644
@@ -31,6 +31,7 @@
 #include "space-info.h"
 #include "block-group.h"
 #include "discard.h"
+#include "zoned.h"
 
 const struct btrfs_raid_attr btrfs_raid_array[BTRFS_NR_RAID_TYPES] = {
 	[BTRFS_RAID_RAID10] = {
@@ -374,6 +375,7 @@ void btrfs_free_device(struct btrfs_device *device)
 	rcu_string_free(device->name);
 	extent_io_tree_release(&device->alloc_state);
 	bio_put(device->flush_bio);
+	btrfs_destroy_dev_zone_info(device);
 	kfree(device);
 }
 
@@ -667,6 +669,11 @@ static int btrfs_open_one_device(struct btrfs_fs_devices *fs_devices,
 	clear_bit(BTRFS_DEV_STATE_IN_FS_METADATA, &device->dev_state);
 	device->mode = flags;
 
+	/* Get zone type information of zoned block devices */
+	ret = btrfs_get_dev_zone_info(device);
+	if (ret != 0)
+		goto error_free_page;
+
 	fs_devices->open_devices++;
 	if (test_bit(BTRFS_DEV_STATE_WRITEABLE, &device->dev_state) &&
 	    device->devid != BTRFS_DEV_REPLACE_DEVID) {
@@ -1143,6 +1150,7 @@ static void btrfs_close_one_device(struct btrfs_device *device)
 		device->bdev = NULL;
 	}
 	clear_bit(BTRFS_DEV_STATE_WRITEABLE, &device->dev_state);
+	btrfs_destroy_dev_zone_info(device);
 
 	device->fs_info = NULL;
 	atomic_set(&device->dev_stats_ccnt, 0);
@@ -2543,6 +2551,14 @@ int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path
 	}
 	rcu_assign_pointer(device->name, name);
 
+	device->fs_info = fs_info;
+	device->bdev = bdev;
+
+	/* Get zone type information of zoned block devices */
+	ret = btrfs_get_dev_zone_info(device);
+	if (ret)
+		goto error_free_device;
+
 	trans = btrfs_start_transaction(root, 0);
 	if (IS_ERR(trans)) {
 		ret = PTR_ERR(trans);
@@ -2559,8 +2575,6 @@ int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path
 					 fs_info->sectorsize);
 	device->disk_total_bytes = device->total_bytes;
 	device->commit_total_bytes = device->total_bytes;
-	device->fs_info = fs_info;
-	device->bdev = bdev;
 	set_bit(BTRFS_DEV_STATE_IN_FS_METADATA, &device->dev_state);
 	clear_bit(BTRFS_DEV_STATE_REPLACE_TGT, &device->dev_state);
 	device->mode = FMODE_EXCL;
@@ -2707,6 +2721,7 @@ int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path
 		sb->s_flags |= SB_RDONLY;
 	if (trans)
 		btrfs_end_transaction(trans);
+	btrfs_destroy_dev_zone_info(device);
 error_free_device:
 	btrfs_free_device(device);
 error:
fs/btrfs/volumes.h +4 -0
index bf27ac07d315..9c07b97a2260 100644
@@ -51,6 +51,8 @@ struct btrfs_io_geometry {
 #define BTRFS_DEV_STATE_REPLACE_TGT	(3)
 #define BTRFS_DEV_STATE_FLUSH_SENT	(4)
 
+struct btrfs_zoned_device_info;
+
 struct btrfs_device {
 	struct list_head dev_list; /* device_list_mutex */
 	struct list_head dev_alloc_list; /* chunk mutex */
@@ -64,6 +66,8 @@ struct btrfs_device {
 
 	struct block_device *bdev;
 
+	struct btrfs_zoned_device_info *zone_info;
+
 	/* the mode sent to blkdev_get */
 	fmode_t mode;
 
fs/btrfs/zoned.c new file +182 -0

182 lines not on this page · raw message

fs/btrfs/zoned.h new file +91 -0

91 lines not on this page · raw message

Signature · 1 line
2.27.0
#22 lore raw

Re: [PATCH v10 04/41] btrfs: get zone information of zoned block devices

Quoted 7 lines
fs/btrfs/super.c +0 -0
index 8840a4fa81eb..ed55014fd1bd 100644

fs/btrfs/super.c static void __init btrfs_print_mod_info(void) · line 2470

Quoted 9 lines
fs/btrfs/super.c +5 -0
@@ -2462,6 +2462,11 @@ static void __init btrfs_print_mod_info(void)
  #endif
  #ifdef CONFIG_BTRFS_FS_REF_VERIFY
  			", ref-verify=on"
+#endif
+#ifdef CONFIG_BLK_DEV_ZONED
+			", zoned=yes"
+#else
+			", zoned=no"
  #endif

IMO, we don't need this, as most of the generic kernel will be compiled with the CONFIG_BLK_DEV_ZONED defined.
For review purpose we may want to know if the mounted device is a zoned device. So log of zone device and its type may be useful when we have verified the zoned devices in the open_ctree().

fs/btrfs/super.c void btrfs_free_device(struct btrfs_device *device) · line 379

Quoted 6 lines
fs/btrfs/super.c +1 -0
@@ -374,6 +375,7 @@ void btrfs_free_device(struct btrfs_device *device)
  	rcu_string_free(device->name);
  	extent_io_tree_release(&device->alloc_state);
  	bio_put(device->flush_bio);
 
+	btrfs_destroy_dev_zone_info(device);

Free of btrfs_device::zone_info is already happening in the path..

btrfs_close_one_device()
btrfs_destroy_dev_zone_info()

We don't need this..

btrfs_free_device()
btrfs_destroy_dev_zone_info()

Quoted 5 lines
fs/btrfs/super.c +1 -0
@@ -2543,6 +2551,14 @@ int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path
  	}
  	rcu_assign_pointer(device->name, name);
 
+	device->fs_info = fs_info;

fs/btrfs/super.c int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path · line 2564

Quoted 10 lines
fs/btrfs/super.c +7 -0
+	device->bdev = bdev;
+
+	/* Get zone type information of zoned block devices */
+	ret = btrfs_get_dev_zone_info(device);
+	if (ret)
+		goto error_free_device;
+
  	trans = btrfs_start_transaction(root, 0);
  	if (IS_ERR(trans)) {
  		ret = PTR_ERR(trans);

It should be something like goto error_free_zone from here.

fs/btrfs/super.c int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path · line 2723

Quoted 4 lines
fs/btrfs/super.c +0 -0
@@ -2707,6 +2721,7 @@ int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path
  		sb->s_flags |= SB_RDONLY;
  	if (trans)
  		btrfs_end_transaction(trans);

error_free_zone:

fs/btrfs/super.c int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path · line 2727

Quoted 4 lines
fs/btrfs/super.c +1 -0
+	btrfs_destroy_dev_zone_info(device);
  error_free_device:
  	btrfs_free_device(device);
  error:

As mentioned we don't need btrfs_destroy_dev_zone_info() again in btrfs_free_device(). Otherwise we end up calling btrfs_destroy_dev_zone_info twice here.

Thanks, Anand

#23 lore raw

Re: [PATCH v10 04/41] btrfs: get zone information of zoned block devices

Quoted 11 lines · Anand Jain
Quoted 11 lines
fs/btrfs/super.c +3 -0
index 8840a4fa81eb..ed55014fd1bd 100644
@@ -2462,6 +2462,11 @@ static void __init btrfs_print_mod_info(void)
  #endif
  #ifdef CONFIG_BTRFS_FS_REF_VERIFY
  			", ref-verify=on"
+#endif
+#ifdef CONFIG_BLK_DEV_ZONED
+			", zoned=yes"
Quoted 9 lines

fs/btrfs/super.c static void __init btrfs_print_mod_info(void) · line 2470

Quoted 3 lines
fs/btrfs/super.c +2 -0
+#else
+			", zoned=no"
  #endif

IMO, we don't need this, as most of the generic kernel will be compiled with the CONFIG_BLK_DEV_ZONED defined.
For review purpose we may want to know if the mounted device is a zoned device. So log of zone device and its type may be useful when we have verified the zoned devices in the open_ctree().

David explicitly asked for this in [1] so we included it.

[1] https://lore.kernel.org/linux-btrfs/[email protected]

#24 lore raw

Re: [PATCH v10 04/41] btrfs: get zone information of zoned block devices

Quoted 22 lines · Johannes Thumshirn
Quoted 11 lines · Anand Jain
Quoted 11 lines
fs/btrfs/super.c +3 -0
index 8840a4fa81eb..ed55014fd1bd 100644
@@ -2462,6 +2462,11 @@ static void __init btrfs_print_mod_info(void)
  #endif
  #ifdef CONFIG_BTRFS_FS_REF_VERIFY
  			", ref-verify=on"
+#endif
+#ifdef CONFIG_BLK_DEV_ZONED
+			", zoned=yes"
Quoted 9 lines

fs/btrfs/super.c static void __init btrfs_print_mod_info(void) · line 2470

Quoted 3 lines
fs/btrfs/super.c +2 -0
+#else
+			", zoned=no"
  #endif

IMO, we don't need this, as most of the generic kernel will be compiled with the CONFIG_BLK_DEV_ZONED defined.
For review purpose we may want to know if the mounted device is a zoned device. So log of zone device and its type may be useful when we have verified the zoned devices in the open_ctree().

Quoted 4 lines

David explicitly asked for this in [1] so we included it.

[1] https://lore.kernel.org/linux-btrfs/[email protected]

And as of now, not all generic kernels are compiled with CONFIG_BLK_DEV_ZONED. E.g. RHEL and CentOS. That may change in the future, but it should not be assumed that CONFIG_BLK_DEV_ZONED is always enabled.

Signature · 2 lines
Damien Le Moal
Western Digital Research
#25 lore raw

Re: [PATCH v10 04/41] btrfs: get zone information of zoned block devices

Quoted 23 lines · Damien Le Moal
Quoted 22 lines · Johannes Thumshirn
Quoted 20 lines · Anand Jain
Quoted 5 lines
fs/btrfs/super.c +0 -0
index 8840a4fa81eb..ed55014fd1bd 100644

fs/btrfs/super.c static void __init btrfs_print_mod_info(void) · line 2470

Quoted 9 lines
fs/btrfs/super.c +5 -0
@@ -2462,6 +2462,11 @@ static void __init btrfs_print_mod_info(void)
   #endif
   #ifdef CONFIG_BTRFS_FS_REF_VERIFY
   			", ref-verify=on"
+#endif
+#ifdef CONFIG_BLK_DEV_ZONED
+			", zoned=yes"
+#else
+			", zoned=no"
   #endif

IMO, we don't need this, as most of the generic kernel will be compiled with the CONFIG_BLK_DEV_ZONED defined.
For review purpose we may want to know if the mounted device is a zoned device. So log of zone device and its type may be useful when we have verified the zoned devices in the open_ctree().

Quoted 9 lines
Quoted 4 lines

David explicitly asked for this in [1] so we included it.

[1] https://lore.kernel.org/linux-btrfs/[email protected]

And as of now, not all generic kernels are compiled with CONFIG_BLK_DEV_ZONED. E.g. RHEL and CentOS. That may change in the future, but it should not be assumed that CONFIG_BLK_DEV_ZONED is always enabled.

Ok. My comment was from the long term perspective. I am fine if you want to keep it.

#26 lore raw

Re: [PATCH v10 04/41] btrfs: get zone information of zoned block devices

Quoted 29 lines · Anand Jain
Quoted 28 lines · Damien Le Moal
Quoted 26 lines · Johannes Thumshirn
Quoted 11 lines · Anand Jain
Quoted 11 lines
fs/btrfs/super.c +3 -0
index 8840a4fa81eb..ed55014fd1bd 100644
@@ -2462,6 +2462,11 @@ static void __init btrfs_print_mod_info(void)
   #endif
   #ifdef CONFIG_BTRFS_FS_REF_VERIFY
   			", ref-verify=on"
+#endif
+#ifdef CONFIG_BLK_DEV_ZONED
+			", zoned=yes"
Quoted 9 lines

fs/btrfs/super.c static void __init btrfs_print_mod_info(void) · line 2470

Quoted 3 lines
fs/btrfs/super.c +2 -0
+#else
+			", zoned=no"
   #endif

IMO, we don't need this, as most of the generic kernel will be compiled with the CONFIG_BLK_DEV_ZONED defined.
For review purpose we may want to know if the mounted device is a zoned device. So log of zone device and its type may be useful when we have verified the zoned devices in the open_ctree().

David explicitly asked for this in [1] so we included it.

[1] https://lore.kernel.org/linux-btrfs/[email protected]

Quoted 6 lines
Quoted 3 lines

And as of now, not all generic kernels are compiled with CONFIG_BLK_DEV_ZONED. E.g. RHEL and CentOS. That may change in the future, but it should not be assumed that CONFIG_BLK_DEV_ZONED is always enabled.

Ok. My comment was from the long term perspective. I am fine if you want to keep it.

The idea is to let the module announce which conditionally built features are there according to fs/btrfs/Makefile and Kconfig. Besides ACLs that should be always on and self-tests that run right after module load, all other are there and we should keep the list up to date.

#27 lore raw

Re: [PATCH v10 04/41] btrfs: get zone information of zoned block devices

In reply to Anand Jain
Quoted 67 lines · Anand Jain
Quoted 7 lines
fs/btrfs/super.c +0 -0
index 8840a4fa81eb..ed55014fd1bd 100644

fs/btrfs/super.c static void __init btrfs_print_mod_info(void) · line 2470

Quoted 9 lines
fs/btrfs/super.c +5 -0
@@ -2462,6 +2462,11 @@ static void __init btrfs_print_mod_info(void)
  #endif
  #ifdef CONFIG_BTRFS_FS_REF_VERIFY
  			", ref-verify=on"
+#endif
+#ifdef CONFIG_BLK_DEV_ZONED
+			", zoned=yes"
+#else
+			", zoned=no"
  #endif

IMO, we don't need this, as most of the generic kernel will be compiled with the CONFIG_BLK_DEV_ZONED defined.
For review purpose we may want to know if the mounted device is a zoned device. So log of zone device and its type may be useful when we have verified the zoned devices in the open_ctree().

fs/btrfs/super.c void btrfs_free_device(struct btrfs_device *device) · line 379

Quoted 6 lines
fs/btrfs/super.c +1 -0
@@ -374,6 +375,7 @@ void btrfs_free_device(struct btrfs_device *device)
  	rcu_string_free(device->name);
  	extent_io_tree_release(&device->alloc_state);
  	bio_put(device->flush_bio);
 
+	btrfs_destroy_dev_zone_info(device);

Free of btrfs_device::zone_info is already happening in the path..

btrfs_close_one_device()
btrfs_destroy_dev_zone_info()

We don't need this..

btrfs_free_device()
btrfs_destroy_dev_zone_info()

Quoted 5 lines
fs/btrfs/super.c +1 -0
@@ -2543,6 +2551,14 @@ int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path
  	}
  	rcu_assign_pointer(device->name, name);
 
+	device->fs_info = fs_info;

fs/btrfs/super.c int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path · line 2564

Quoted 10 lines
fs/btrfs/super.c +7 -0
+	device->bdev = bdev;
+
+	/* Get zone type information of zoned block devices */
+	ret = btrfs_get_dev_zone_info(device);
+	if (ret)
+		goto error_free_device;
+
  	trans = btrfs_start_transaction(root, 0);
  	if (IS_ERR(trans)) {
  		ret = PTR_ERR(trans);

It should be something like goto error_free_zone from here.

fs/btrfs/super.c int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path · line 2723

Quoted 4 lines
fs/btrfs/super.c +0 -0
@@ -2707,6 +2721,7 @@ int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path
  		sb->s_flags |= SB_RDONLY;
  	if (trans)
  		btrfs_end_transaction(trans);

error_free_zone:

Quoted 8 lines

fs/btrfs/super.c int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path · line 2727

Quoted 4 lines
fs/btrfs/super.c +1 -0
+	btrfs_destroy_dev_zone_info(device);
  error_free_device:
  	btrfs_free_device(device);
  error:

As mentioned we don't need btrfs_destroy_dev_zone_info() again in btrfs_free_device(). Otherwise we end up calling btrfs_destroy_dev_zone_info twice here.

Which doesn't do any harm as:
void btrfs_destroy_dev_zone_info(struct btrfs_device *device)

{
        struct btrfs_zoned_device_info *zone_info = device->zone_info;

        if (!zone_info)
                return;

	/* ... */
        device->zone_info = NULL;
}

Not sure what would be the preferred style here

#28 lore raw

Re: [PATCH v10 04/41] btrfs: get zone information of zoned block devices

In reply to Anand Jain
Quoted 32 lines · Anand Jain
Quoted 7 lines
fs/btrfs/super.c +0 -0
index 8840a4fa81eb..ed55014fd1bd 100644

fs/btrfs/super.c static void __init btrfs_print_mod_info(void) · line 2470

Quoted 9 lines
fs/btrfs/super.c +5 -0
@@ -2462,6 +2462,11 @@ static void __init btrfs_print_mod_info(void)
 #endif
 #ifdef CONFIG_BTRFS_FS_REF_VERIFY
 			", ref-verify=on"
+#endif
+#ifdef CONFIG_BLK_DEV_ZONED
+			", zoned=yes"
+#else
+			", zoned=no"
 #endif

IMO, we don't need this, as most of the generic kernel will be compiled with the CONFIG_BLK_DEV_ZONED defined.
For review purpose we may want to know if the mounted device is a zoned device. So log of zone device and its type may be useful when we have verified the zoned devices in the open_ctree().

fs/btrfs/super.c void btrfs_free_device(struct btrfs_device *device) · line 379

Quoted 6 lines
fs/btrfs/super.c +1 -0
@@ -374,6 +375,7 @@ void btrfs_free_device(struct btrfs_device *device)
 	rcu_string_free(device->name);
 	extent_io_tree_release(&device->alloc_state);
 	bio_put(device->flush_bio);
 
+	btrfs_destroy_dev_zone_info(device);

Free of btrfs_device::zone_info is already happening in the path..

Quoted 7 lines

btrfs_close_one_device()
btrfs_destroy_dev_zone_info()

We don't need this..

btrfs_free_device()
btrfs_destroy_dev_zone_info()

Ah, yes, I once had it only in btrfs_free_device() and noticed that it does not free the device zone info on umount. So, I added one in btrfs_close_one_device() and forgot to remove the other one. I'll drop it from btrfs_free_device().

Quoted 20 lines
Quoted 6 lines
fs/btrfs/super.c +1 -0
@@ -2543,6 +2551,14 @@ int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path
 	}
 	rcu_assign_pointer(device->name, name);
+	device->fs_info = fs_info;

fs/btrfs/super.c int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path · line 2563

Quoted 10 lines
fs/btrfs/super.c +7 -0
+	device->bdev = bdev;
+
+	/* Get zone type information of zoned block devices */
+	ret = btrfs_get_dev_zone_info(device);
+	if (ret)
+		goto error_free_device;
+
 	trans = btrfs_start_transaction(root, 0);
 	if (IS_ERR(trans)) {
 		ret = PTR_ERR(trans);

It should be something like goto error_free_zone from here.

Quoted 7 lines

fs/btrfs/super.c int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path · line 2723

Quoted 4 lines
fs/btrfs/super.c +0 -0
@@ -2707,6 +2721,7 @@ int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path
 		sb->s_flags |= SB_RDONLY;
 	if (trans)
 		btrfs_end_transaction(trans);

error_free_zone:

And, I'll do something like this.

Quoted 11 lines

fs/btrfs/super.c int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path · line 2727

Quoted 4 lines
fs/btrfs/super.c +1 -0
+	btrfs_destroy_dev_zone_info(device);
 error_free_device:
 	btrfs_free_device(device);
 error:

As mentioned we don't need btrfs_destroy_dev_zone_info() again in btrfs_free_device(). Otherwise we end up calling btrfs_destroy_dev_zone_info twice here.

Thanks, Anand

#29 lore raw

Re: [PATCH v10 04/41] btrfs: get zone information of zoned block devices

Also, %device->fs_info is not protected. It is better to avoid using fs_info when we are still at open_fs_devices(). Yeah, the unknown part can be better. We need to fix it as a whole. For now, you can use something like...

-------------------------

fs/btrfs/zoned.c +5 -13
index 1223d5b0e411..e857bb304d28 100644
@@ -130,19 +130,11 @@ int btrfs_get_dev_zone_info(struct btrfs_device *device)
         * (device <unknown>) ..."
         */
 
-       rcu_read_lock();
-       if (device->fs_info)
-               btrfs_info(device->fs_info,
-                       "host-%s zoned block device %s, %u zones of %llu bytes",
-                       bdev_zoned_model(bdev) == BLK_ZONED_HM ? "managed" : "aware",
-                       rcu_str_deref(device->name), zone_info->nr_zones,
-                       zone_info->zone_size);
-       else
-               pr_info("BTRFS info: host-%s zoned block device %s, %u zones of %llu bytes",
-                       bdev_zoned_model(bdev) == BLK_ZONED_HM ? "managed" : "aware",
-                       rcu_str_deref(device->name), zone_info->nr_zones,
-                       zone_info->zone_size);
-       rcu_read_unlock();
+       btrfs_info_in_rcu(NULL,
+               "host-%s zoned block device %s, %u zones of %llu bytes",
+               bdev_zoned_model(bdev) == BLK_ZONED_HM ? "managed" : "aware",
+               rcu_str_deref(device->name), zone_info->nr_zones,
+               zone_info->zone_size);
 
        return 0;
 ---------------------------

Thanks, Anand

Quoted 89 lines · Naohiro Aota
Quoted 32 lines · Anand Jain
Quoted 15 lines
fs/btrfs/super.c +0 -0
index 8840a4fa81eb..ed55014fd1bd 100644
@@ -2462,6 +2462,11 @@ static void __init btrfs_print_mod_info(void)
  #endif
  #ifdef CONFIG_BTRFS_FS_REF_VERIFY

             ", ref-verify=on"

fs/btrfs/super.c +5 -0
+#endif
+#ifdef CONFIG_BLK_DEV_ZONED
+            ", zoned=yes"
+#else
+            ", zoned=no"
Quoted 1 line

 #endif

IMO, we don't need this, as most of the generic kernel will be compiled with the CONFIG_BLK_DEV_ZONED defined.
For review purpose we may want to know if the mounted device is a zoned device. So log of zone device and its type may be useful when we have verified the zoned devices in the open_ctree().

fs/btrfs/super.c void btrfs_free_device(struct btrfs_device *device) · line 379

Quoted 6 lines
fs/btrfs/super.c +1 -0
@@ -374,6 +375,7 @@ void btrfs_free_device(struct btrfs_device *device)
      rcu_string_free(device->name);
      extent_io_tree_release(&device->alloc_state);
      bio_put(device->flush_bio);
 
+    btrfs_destroy_dev_zone_info(device);

Free of btrfs_device::zone_info is already happening in the path..

Quoted 7 lines

btrfs_close_one_device()
  btrfs_destroy_dev_zone_info()

We don't need this..

btrfs_free_device()
 btrfs_destroy_dev_zone_info()

Ah, yes, I once had it only in btrfs_free_device() and noticed that it does not free the device zone info on umount. So, I added one in btrfs_close_one_device() and forgot to remove the other one. I'll drop it from btrfs_free_device().

Quoted 21 lines
Quoted 7 lines
fs/btrfs/super.c +1 -0
@@ -2543,6 +2551,14 @@ int btrfs_init_new_device(struct btrfs_fs_info 
 *fs_info, const char *device_path
      }
      rcu_assign_pointer(device->name, name);
+    device->fs_info = fs_info;

fs/btrfs/super.c int btrfs_init_new_device(struct btrfs_fs_info · line 2564

Quoted 10 lines
fs/btrfs/super.c +7 -0
+    device->bdev = bdev;
+
+    /* Get zone type information of zoned block devices */
+    ret = btrfs_get_dev_zone_info(device);
+    if (ret)
+        goto error_free_device;
+
      trans = btrfs_start_transaction(root, 0);
      if (IS_ERR(trans)) {
          ret = PTR_ERR(trans);

It should be something like goto error_free_zone from here.

Quoted 8 lines

fs/btrfs/super.c int btrfs_init_new_device(struct btrfs_fs_info · line 2724

Quoted 5 lines
fs/btrfs/super.c +0 -0
@@ -2707,6 +2721,7 @@ int btrfs_init_new_device(struct btrfs_fs_info 
 *fs_info, const char *device_path
          sb->s_flags |= SB_RDONLY;
      if (trans)
          btrfs_end_transaction(trans);

error_free_zone:

And, I'll do something like this.

Quoted 11 lines
Quoted 4 lines
fs/btrfs/super.c +1 -0
+    btrfs_destroy_dev_zone_info(device);

 error_free_device:
     btrfs_free_device(device);
 error:

As mentioned we don't need btrfs_destroy_dev_zone_info() again in  btrfs_free_device(). Otherwise we end up calling btrfs_destroy_dev_zone_info twice here.

Thanks, Anand

Reviewed-by
#30 lore raw

Re: [PATCH v10 04/41] btrfs: get zone information of zoned block devices

Below two comments are fixed in the misc-next.

Thanks.

Quoted 71 lines · Anand Jain

Also, %device->fs_info is not protected. It is better to avoid using fs_info when we are still at open_fs_devices(). Yeah, the unknown part can be better. We need to fix it as a whole. For now, you can use something like...

-------------------------

fs/btrfs/zoned.c +3 -13
index 1223d5b0e411..e857bb304d28 100644
@@ -130,19 +130,11 @@ int btrfs_get_dev_zone_info(struct btrfs_device 
 *device)
          * (device <unknown>) ..."
          */
 
-       rcu_read_lock();
-       if (device->fs_info)
-               btrfs_info(device->fs_info,
-                       "host-%s zoned block device %s, %u zones of %llu 
 bytes",
-                       bdev_zoned_model(bdev) == BLK_ZONED_HM ? 
 "managed" : "aware",
-                       rcu_str_deref(device->name), zone_info->nr_zones,
-                       zone_info->zone_size);
-       else
-               pr_info("BTRFS info: host-%s zoned block device %s, %u 
 zones of %llu bytes",
-                       bdev_zoned_model(bdev) == BLK_ZONED_HM ? 
 "managed" : "aware",
-                       rcu_str_deref(device->name), zone_info->nr_zones,
-                       zone_info->zone_size);
-       rcu_read_unlock();
+       btrfs_info_in_rcu(NULL,
+               "host-%s zoned block device %s, %u zones of %llu bytes",
+               bdev_zoned_model(bdev) == BLK_ZONED_HM ? "managed" : 

"aware",
+               rcu_str_deref(device->name), zone_info->nr_zones, +               zone_info->zone_size);

        return 0;
 ---------------------------

Thanks, Anand

Quoted 21 lines
Quoted 9 lines

Review void btrfs_free_device(struct btrfs_device *device) · line 379

Quoted 6 lines
(quoted diff) +1 -0
@@ -374,6 +375,7 @@ void btrfs_free_device(struct btrfs_device *device)
      rcu_string_free(device->name);
      extent_io_tree_release(&device->alloc_state);
      bio_put(device->flush_bio);
 
+    btrfs_destroy_dev_zone_info(device);

Free of btrfs_device::zone_info is already happening in the path..

Quoted 7 lines

btrfs_close_one_device()
  btrfs_destroy_dev_zone_info()

We don't need this..

btrfs_free_device()
 btrfs_destroy_dev_zone_info()

Ah, yes, I once had it only in btrfs_free_device() and noticed that it does not free the device zone info on umount. So, I added one in btrfs_close_one_device() and forgot to remove the other one. I'll drop it from btrfs_free_device().

#31 lore raw

Re: [PATCH v10 04/41] btrfs: get zone information of zoned block devices

In reply to Naohiro Aota

fs/btrfs/zoned.c line 52

Quoted 6 lines · Naohiro Aota
fs/btrfs/zoned.c new file +6 -0
+int btrfs_get_dev_zone_info(struct btrfs_device *device)
+{
+	struct btrfs_zoned_device_info *zone_info = NULL;
+	struct block_device *bdev = device->bdev;
+	sector_t nr_sectors = bdev->bd_part->nr_sects;
+	sector_t sector = 0;

I'd rather replace the sector_t types with u64. The type is unsigned long and does not have the same width on 32/64 bit. The typecasts must be used and if not, bugs happen (and happened).

Quoted 7 lines
fs/btrfs/zoned.c new file +7 -0
+	struct blk_zone *zones = NULL;
+	unsigned int i, nreported = 0, nr_zones;
+	unsigned int zone_sectors;
+	int ret;
+
+	if (!bdev_is_zoned(bdev))
+		return 0;

fs/btrfs/zoned.c line 69

Quoted 10 lines
fs/btrfs/zoned.c new file +10 -0
+
+	if (device->zone_info)
+		return 0;
+
+	zone_info = kzalloc(sizeof(*zone_info), GFP_KERNEL);
+	if (!zone_info)
+		return -ENOMEM;
+
+	zone_sectors = bdev_zone_sectors(bdev);
+	ASSERT(is_power_of_2(zone_sectors));

As is_power_of_2 works only on longs, this needs to be opencoded as there's no unsigned long long version.

#32 lore raw

Re: [PATCH v10 04/41] btrfs: get zone information of zoned block devices

Quoted 11 lines · David Sterba

fs/btrfs/zoned.c line 52

Quoted 6 lines · Naohiro Aota
fs/btrfs/zoned.c new file +6 -0
+int btrfs_get_dev_zone_info(struct btrfs_device *device)
+{
+	struct btrfs_zoned_device_info *zone_info = NULL;
+	struct block_device *bdev = device->bdev;
+	sector_t nr_sectors = bdev->bd_part->nr_sects;
+	sector_t sector = 0;

I'd rather replace the sector_t types with u64. The type is unsigned long and does not have the same width on 32/64 bit. The typecasts must be used and if not, bugs happen (and happened).

Like in the same function a few lines below

   95         /* Get zones type */
   96         while (sector < nr_sectors) {
   97                 nr_zones = BTRFS_REPORT_NR_ZONES;
   98                 ret = btrfs_get_dev_zones(device, sector << SECTOR_SHIFT, zones,
   99                                           &nr_zones);

sector without a type cast to u64

#33 lore raw

Re: [PATCH v10 04/41] btrfs: get zone information of zoned block devices

In reply to David Sterba

Hi David,

Quoted 11 lines · David Sterba

fs/btrfs/zoned.c line 52

Quoted 6 lines · Naohiro Aota
fs/btrfs/zoned.c new file +6 -0
+int btrfs_get_dev_zone_info(struct btrfs_device *device)
+{
+	struct btrfs_zoned_device_info *zone_info = NULL;
+	struct block_device *bdev = device->bdev;
+	sector_t nr_sectors = bdev->bd_part->nr_sects;
+	sector_t sector = 0;

I'd rather replace the sector_t types with u64. The type is unsigned long and does not have the same width on 32/64 bit. The typecasts must be used and if not, bugs happen (and happened).

Since kernel 5.2, sector_t is unconditionally defined as u64 in linux/type.h:

typedef u64 sector_t;

CONFIG_LBDAF does not exist anymore.

I am not against using u64 at all, but using sector_t makes it clear what the unit is for the values at hand.

Quoted 21 lines
Quoted 8 lines
fs/btrfs/zoned.c new file +7 -0
+	struct blk_zone *zones = NULL;
+	unsigned int i, nreported = 0, nr_zones;
+	unsigned int zone_sectors;
+	int ret;
+
+	if (!bdev_is_zoned(bdev))
+		return 0;

fs/btrfs/zoned.c line 69

Quoted 10 lines
fs/btrfs/zoned.c new file +10 -0
+
+	if (device->zone_info)
+		return 0;
+
+	zone_info = kzalloc(sizeof(*zone_info), GFP_KERNEL);
+	if (!zone_info)
+		return -ENOMEM;
+
+	zone_sectors = bdev_zone_sectors(bdev);
+	ASSERT(is_power_of_2(zone_sectors));

As is_power_of_2 works only on longs, this needs to be opencoded as there's no unsigned long long version.

Signature · 2 lines
Damien Le Moal
Western Digital
#34 lore raw

Re: [PATCH v10 04/41] btrfs: get zone information of zoned block devices

Quoted 15 lines · Damien Le Moal

Hi David,

Quoted 11 lines · David Sterba

fs/btrfs/zoned.c line 52

Quoted 6 lines · Naohiro Aota
fs/btrfs/zoned.c new file +6 -0
+int btrfs_get_dev_zone_info(struct btrfs_device *device)
+{
+	struct btrfs_zoned_device_info *zone_info = NULL;
+	struct block_device *bdev = device->bdev;
+	sector_t nr_sectors = bdev->bd_part->nr_sects;
+	sector_t sector = 0;

I'd rather replace the sector_t types with u64. The type is unsigned long and does not have the same width on 32/64 bit. The typecasts must be used and if not, bugs happen (and happened).

Quoted 5 lines

Since kernel 5.2, sector_t is unconditionally defined as u64 in linux/type.h:

typedef u64 sector_t;

CONFIG_LBDAF does not exist anymore.

That's great, I was not aware of that.

Quoted 2 lines

I am not against using u64 at all, but using sector_t makes it clear what the unit is for the values at hand.

Yeah agreed, I'll switch it back.

#35 lore raw

Re: [PATCH v10 04/41] btrfs: get zone information of zoned block devices

In reply to Naohiro Aota

fs/btrfs/zoned.c line 107

Quoted 13 lines · Naohiro Aota
fs/btrfs/zoned.c new file +13 -0
+	/* Get zones type */
+	while (sector < nr_sectors) {
+		nr_zones = BTRFS_REPORT_NR_ZONES;
+		ret = btrfs_get_dev_zones(device, sector << SECTOR_SHIFT, zones,
+					  &nr_zones);
+		if (ret)
+			goto out;
+
+		for (i = 0; i < nr_zones; i++) {
+			if (zones[i].type == BLK_ZONE_TYPE_SEQWRITE_REQ)
+				set_bit(nreported, zone_info->seq_zones);
+			if (zones[i].cond == BLK_ZONE_COND_EMPTY)
+				set_bit(nreported, zone_info->empty_zones);

set_bit is atomic and it's not needed as nothing else could be touching the bitmap, so I'll switch it to plain __set_bit.

#36 lore raw

PATCH v10 5/41 btrfs: check and enable ZONED mode

In reply to Naohiro Aota

This commit introduces the function btrfs_check_zoned_mode() to check if ZONED flag is enabled on the file system and if the file system consists of zoned devices with equal zone size.

---
 fs/btrfs/ctree.h       | 11 ++++++
 fs/btrfs/dev-replace.c |  7 ++++
 fs/btrfs/disk-io.c     | 11 ++++++
 fs/btrfs/super.c       |  1 +
 fs/btrfs/volumes.c     |  5 +++
 fs/btrfs/zoned.c       | 81 ++++++++++++++++++++++++++++++++++++++++++
 fs/btrfs/zoned.h       | 26 ++++++++++++++
 7 files changed, 142 insertions(+)
fs/btrfs/ctree.h +11 -0
index aac3d6f4e35b..453f41ca024e 100644
@@ -948,6 +948,12 @@ struct btrfs_fs_info {
 	/* Type of exclusive operation running */
 	unsigned long exclusive_operation;
 
+	/* Zone size when in ZONED mode */
+	union {
+		u64 zone_size;
+		u64 zoned;
+	};
+
 #ifdef CONFIG_BTRFS_FS_REF_VERIFY
 	spinlock_t ref_verify_lock;
 	struct rb_root block_tree;
@@ -3595,4 +3601,9 @@ static inline int btrfs_is_testing(struct btrfs_fs_info *fs_info)
 }
 #endif
 
+static inline bool btrfs_is_zoned(struct btrfs_fs_info *fs_info)
+{
+	return fs_info->zoned != 0;
+}
+
 #endif
fs/btrfs/dev-replace.c +7 -0
index 6f6d77224c2b..db87f1aa604b 100644
@@ -238,6 +238,13 @@ static int btrfs_init_dev_replace_tgtdev(struct btrfs_fs_info *fs_info,
 		return PTR_ERR(bdev);
 	}
 
+	if (!btrfs_check_device_zone_type(fs_info, bdev)) {
+		btrfs_err(fs_info,
+			  "dev-replace: zoned type of target device mismatch with filesystem");
+		ret = -EINVAL;
+		goto error;
+	}
+
 	sync_blockdev(bdev);
 
 	list_for_each_entry(device, &fs_info->fs_devices->devices, dev_list) {
fs/btrfs/disk-io.c +11 -0
index 764001609a15..e76ac4da208d 100644
@@ -42,6 +42,7 @@
 #include "block-group.h"
 #include "discard.h"
 #include "space-info.h"
+#include "zoned.h"
 
 #define BTRFS_SUPER_FLAG_SUPP	(BTRFS_HEADER_FLAG_WRITTEN |\
 				 BTRFS_HEADER_FLAG_RELOC |\
@@ -2976,6 +2977,8 @@ int __cold open_ctree(struct super_block *sb, struct btrfs_fs_devices *fs_device
 	if (features & BTRFS_FEATURE_INCOMPAT_SKINNY_METADATA)
 		btrfs_info(fs_info, "has skinny extents");
 
+	fs_info->zoned = features & BTRFS_FEATURE_INCOMPAT_ZONED;
+
 	/*
 	 * flag our filesystem as having big metadata blocks if
 	 * they are bigger than the page size
@@ -3130,7 +3133,15 @@ int __cold open_ctree(struct super_block *sb, struct btrfs_fs_devices *fs_device
 
 	btrfs_free_extra_devids(fs_devices, 1);
 
+	ret = btrfs_check_zoned_mode(fs_info);
+	if (ret) {
+		btrfs_err(fs_info, "failed to inititialize zoned mode: %d",
+			  ret);
+		goto fail_block_groups;
+	}
+
 	ret = btrfs_sysfs_add_fsid(fs_devices);
+
 	if (ret) {
 		btrfs_err(fs_info, "failed to init sysfs fsid interface: %d",
 				ret);
fs/btrfs/super.c +1 -0
index ed55014fd1bd..3312fe08168f 100644
@@ -44,6 +44,7 @@
 #include "backref.h"
 #include "space-info.h"
 #include "sysfs.h"
+#include "zoned.h"
 #include "tests/btrfs-tests.h"
 #include "block-group.h"
 #include "discard.h"
fs/btrfs/volumes.c +5 -0
index e787bf89f761..10827892c086 100644
@@ -2518,6 +2518,11 @@ int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path
 	if (IS_ERR(bdev))
 		return PTR_ERR(bdev);
 
+	if (!btrfs_check_device_zone_type(fs_info, bdev)) {
+		ret = -EINVAL;
+		goto error;
+	}
+
 	if (fs_devices->seeding) {
 		seeding_dev = 1;
 		down_write(&sb->s_umount);
fs/btrfs/zoned.c +81 -0

84 lines not on this page · raw message

fs/btrfs/zoned.h +26 -0
index c9e69ff87ab9..bcb1cb99a4f3 100644
@@ -4,6 +4,7 @@
 #define BTRFS_ZONED_H
 
 #include <linux/types.h>
+#include <linux/blkdev.h>
 
 struct btrfs_zoned_device_info {
 	/*
@@ -22,6 +23,7 @@ int btrfs_get_dev_zone(struct btrfs_device *device, u64 pos,
 		       struct blk_zone *zone);
 int btrfs_get_dev_zone_info(struct btrfs_device *device);
 void btrfs_destroy_dev_zone_info(struct btrfs_device *device);
+int btrfs_check_zoned_mode(struct btrfs_fs_info *fs_info);
 #else /* CONFIG_BLK_DEV_ZONED */
 static inline int btrfs_get_dev_zone(struct btrfs_device *device, u64 pos,
 				     struct blk_zone *zone)
@@ -36,6 +38,15 @@ static inline int btrfs_get_dev_zone_info(struct btrfs_device *device)
 
 static inline void btrfs_destroy_dev_zone_info(struct btrfs_device *device) { }
 
+static inline int btrfs_check_zoned_mode(struct btrfs_fs_info *fs_info)
+{
+	if (!btrfs_is_zoned(fs_info))
+		return 0;
+
+	btrfs_err(fs_info, "Zoned block devices support is not enabled");
+	return -EOPNOTSUPP;
+}
+
 #endif
 
 static inline bool btrfs_dev_is_sequential(struct btrfs_device *device, u64 pos)
@@ -88,4 +99,19 @@ static inline void btrfs_dev_clear_zone_empty(struct btrfs_device *device,
 	btrfs_dev_set_empty_zone_bit(device, pos, false);
 }
 
+static inline bool btrfs_check_device_zone_type(struct btrfs_fs_info *fs_info,
+						struct block_device *bdev)
+{
+	u64 zone_size;
+
+	if (btrfs_is_zoned(fs_info)) {
+		zone_size = (u64)bdev_zone_sectors(bdev) << SECTOR_SHIFT;
+		/* Do not allow non-zoned device */
+		return bdev_is_zoned(bdev) && fs_info->zone_size == zone_size;
+	}
+
+	/* Do not allow Host Manged zoned device */
+	return bdev_zoned_model(bdev) != BLK_ZONED_HM;
+}
+
 #endif
Signature · 1 line
2.27.0
#37 lore raw

Re: [PATCH v10 05/41] btrfs: check and enable ZONED mode

Quoted 53 lines · Naohiro Aota

This commit introduces the function btrfs_check_zoned_mode() to check if ZONED flag is enabled on the file system and if the file system consists of zoned devices with equal zone size.

---
  fs/btrfs/ctree.h       | 11 ++++++
  fs/btrfs/dev-replace.c |  7 ++++
  fs/btrfs/disk-io.c     | 11 ++++++
  fs/btrfs/super.c       |  1 +
  fs/btrfs/volumes.c     |  5 +++
  fs/btrfs/zoned.c       | 81 ++++++++++++++++++++++++++++++++++++++++++
  fs/btrfs/zoned.h       | 26 ++++++++++++++
  7 files changed, 142 insertions(+)
fs/btrfs/ctree.h +11 -0
index aac3d6f4e35b..453f41ca024e 100644
@@ -948,6 +948,12 @@ struct btrfs_fs_info {
  	/* Type of exclusive operation running */
  	unsigned long exclusive_operation;
 
+	/* Zone size when in ZONED mode */
+	union {
+		u64 zone_size;
+		u64 zoned;
+	};
+
  #ifdef CONFIG_BTRFS_FS_REF_VERIFY
  	spinlock_t ref_verify_lock;
  	struct rb_root block_tree;
@@ -3595,4 +3601,9 @@ static inline int btrfs_is_testing(struct btrfs_fs_info *fs_info)
  }
  #endif
 
+static inline bool btrfs_is_zoned(struct btrfs_fs_info *fs_info)
+{
+	return fs_info->zoned != 0;
+}
+
  #endif
fs/btrfs/dev-replace.c +0 -0
index 6f6d77224c2b..db87f1aa604b 100644
@@ -238,6 +238,13 @@ static int btrfs_init_dev_replace_tgtdev(struct btrfs_fs_info *fs_info,
  		return PTR_ERR(bdev);
  	}
 

fs/btrfs/dev-replace.c static int btrfs_init_dev_replace_tgtdev(struct btrfs_fs_info *fs_info, · line 250

Quoted 10 lines
fs/btrfs/dev-replace.c +7 -0
+	if (!btrfs_check_device_zone_type(fs_info, bdev)) {
+		btrfs_err(fs_info,
+			  "dev-replace: zoned type of target device mismatch with filesystem");
+		ret = -EINVAL;
+		goto error;
+	}
+
  	sync_blockdev(bdev);
 
  	list_for_each_entry(device, &fs_info->fs_devices->devices, dev_list) {

I am not sure if it is done in some other patch. But we still have to check for

(model == BLK_ZONED_HA && incompat_zoned))

right? What if in a non-zoned FS, a zoned device is added through the replace. No?

Quoted 55 lines
fs/btrfs/disk-io.c +11 -0
index 764001609a15..e76ac4da208d 100644
@@ -42,6 +42,7 @@
  #include "block-group.h"
  #include "discard.h"
  #include "space-info.h"
+#include "zoned.h"
 
  #define BTRFS_SUPER_FLAG_SUPP	(BTRFS_HEADER_FLAG_WRITTEN |\
  				 BTRFS_HEADER_FLAG_RELOC |\
@@ -2976,6 +2977,8 @@ int __cold open_ctree(struct super_block *sb, struct btrfs_fs_devices *fs_device
  	if (features & BTRFS_FEATURE_INCOMPAT_SKINNY_METADATA)
  		btrfs_info(fs_info, "has skinny extents");
 
+	fs_info->zoned = features & BTRFS_FEATURE_INCOMPAT_ZONED;
+
  	/*
  	 * flag our filesystem as having big metadata blocks if
  	 * they are bigger than the page size
@@ -3130,7 +3133,15 @@ int __cold open_ctree(struct super_block *sb, struct btrfs_fs_devices *fs_device
 
  	btrfs_free_extra_devids(fs_devices, 1);
 
+	ret = btrfs_check_zoned_mode(fs_info);
+	if (ret) {
+		btrfs_err(fs_info, "failed to inititialize zoned mode: %d",
+			  ret);
+		goto fail_block_groups;
+	}
+
  	ret = btrfs_sysfs_add_fsid(fs_devices);
+
  	if (ret) {
  		btrfs_err(fs_info, "failed to init sysfs fsid interface: %d",
  				ret);
fs/btrfs/super.c +1 -0
index ed55014fd1bd..3312fe08168f 100644
@@ -44,6 +44,7 @@
  #include "backref.h"
  #include "space-info.h"
  #include "sysfs.h"
+#include "zoned.h"
  #include "tests/btrfs-tests.h"
  #include "block-group.h"
  #include "discard.h"
fs/btrfs/volumes.c +0 -0
index e787bf89f761..10827892c086 100644
@@ -2518,6 +2518,11 @@ int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path
  	if (IS_ERR(bdev))

fs/btrfs/volumes.c int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path · line 2528

Quoted 10 lines
fs/btrfs/volumes.c +5 -0
  		return PTR_ERR(bdev);
 
+	if (!btrfs_check_device_zone_type(fs_info, bdev)) {
+		ret = -EINVAL;
+		goto error;
+	}
+
  	if (fs_devices->seeding) {
  		seeding_dev = 1;
  		down_write(&sb->s_umount);

Same here too. It can also happen that a zone device is added to a non zoned fs.

Thanks, Anand

Quoted 146 lines
fs/btrfs/zoned.c +81 -0
index b7ffe6670d3a..1223d5b0e411 100644
@@ -180,3 +180,84 @@ int btrfs_get_dev_zone(struct btrfs_device *device, u64 pos,
 
  	return 0;
  }
+
+int btrfs_check_zoned_mode(struct btrfs_fs_info *fs_info)
+{
+	struct btrfs_fs_devices *fs_devices = fs_info->fs_devices;
+	struct btrfs_device *device;
+	u64 zoned_devices = 0;
+	u64 nr_devices = 0;
+	u64 zone_size = 0;
+	const bool incompat_zoned = btrfs_is_zoned(fs_info);
+	int ret = 0;
+
+	/* Count zoned devices */
+	list_for_each_entry(device, &fs_devices->devices, dev_list) {
+		enum blk_zoned_model model;
+
+		if (!device->bdev)
+			continue;
+
+		model = bdev_zoned_model(device->bdev);
+		if (model == BLK_ZONED_HM ||
+		    (model == BLK_ZONED_HA && incompat_zoned)) {
+			zoned_devices++;
+			if (!zone_size) {
+				zone_size = device->zone_info->zone_size;
+			} else if (device->zone_info->zone_size != zone_size) {
+				btrfs_err(fs_info,
+					  "zoned: unequal block device zone sizes: have %llu found %llu",
+					  device->zone_info->zone_size,
+					  zone_size);
+				ret = -EINVAL;
+				goto out;
+			}
+		}
+		nr_devices++;
+	}
+
+	if (!zoned_devices && !incompat_zoned)
+		goto out;
+
+	if (!zoned_devices && incompat_zoned) {
+		/* No zoned block device found on ZONED FS */
+		btrfs_err(fs_info,
+			  "zoned: no zoned devices found on a zoned filesystem");
+		ret = -EINVAL;
+		goto out;
+	}
+
+	if (zoned_devices && !incompat_zoned) {
+		btrfs_err(fs_info,
+			  "zoned: mode not enabled but zoned device found");
+		ret = -EINVAL;
+		goto out;
+	}
+
+	if (zoned_devices != nr_devices) {
+		btrfs_err(fs_info,
+			  "zoned: cannot mix zoned and regular devices");
+		ret = -EINVAL;
+		goto out;
+	}
+
+	/*
+	 * stripe_size is always aligned to BTRFS_STRIPE_LEN in
+	 * __btrfs_alloc_chunk(). Since we want stripe_len == zone_size,
+	 * check the alignment here.
+	 */
+	if (!IS_ALIGNED(zone_size, BTRFS_STRIPE_LEN)) {
+		btrfs_err(fs_info,
+			  "zoned: zone size not aligned to stripe %u",
+			  BTRFS_STRIPE_LEN);
+		ret = -EINVAL;
+		goto out;
+	}
+
+	fs_info->zone_size = zone_size;
+
+	btrfs_info(fs_info, "zoned mode enabled with zone size %llu",
+		   fs_info->zone_size);
+out:
+	return ret;
+}
fs/btrfs/zoned.h +26 -0
index c9e69ff87ab9..bcb1cb99a4f3 100644
@@ -4,6 +4,7 @@
  #define BTRFS_ZONED_H
 
  #include <linux/types.h>
+#include <linux/blkdev.h>
 
  struct btrfs_zoned_device_info {
  	/*
@@ -22,6 +23,7 @@ int btrfs_get_dev_zone(struct btrfs_device *device, u64 pos,
  		       struct blk_zone *zone);
  int btrfs_get_dev_zone_info(struct btrfs_device *device);
  void btrfs_destroy_dev_zone_info(struct btrfs_device *device);
+int btrfs_check_zoned_mode(struct btrfs_fs_info *fs_info);
  #else /* CONFIG_BLK_DEV_ZONED */
  static inline int btrfs_get_dev_zone(struct btrfs_device *device, u64 pos,
  				     struct blk_zone *zone)
@@ -36,6 +38,15 @@ static inline int btrfs_get_dev_zone_info(struct btrfs_device *device)
 
  static inline void btrfs_destroy_dev_zone_info(struct btrfs_device *device) { }
 
+static inline int btrfs_check_zoned_mode(struct btrfs_fs_info *fs_info)
+{
+	if (!btrfs_is_zoned(fs_info))
+		return 0;
+
+	btrfs_err(fs_info, "Zoned block devices support is not enabled");
+	return -EOPNOTSUPP;
+}
+
  #endif
 
  static inline bool btrfs_dev_is_sequential(struct btrfs_device *device, u64 pos)
@@ -88,4 +99,19 @@ static inline void btrfs_dev_clear_zone_empty(struct btrfs_device *device,
  	btrfs_dev_set_empty_zone_bit(device, pos, false);
  }
 
+static inline bool btrfs_check_device_zone_type(struct btrfs_fs_info *fs_info,
+						struct block_device *bdev)
+{
+	u64 zone_size;
+
+	if (btrfs_is_zoned(fs_info)) {
+		zone_size = (u64)bdev_zone_sectors(bdev) << SECTOR_SHIFT;
+		/* Do not allow non-zoned device */
+		return bdev_is_zoned(bdev) && fs_info->zone_size == zone_size;
+	}
+
+	/* Do not allow Host Manged zoned device */
+	return bdev_zoned_model(bdev) != BLK_ZONED_HM;
+}
+
  #endif
#38 lore raw

Re: [PATCH v10 05/41] btrfs: check and enable ZONED mode

Quoted 63 lines · Anand Jain
Quoted 62 lines · Naohiro Aota

This commit introduces the function btrfs_check_zoned_mode() to check if ZONED flag is enabled on the file system and if the file system consists of zoned devices with equal zone size.

---
  fs/btrfs/ctree.h       | 11 ++++++
  fs/btrfs/dev-replace.c |  7 ++++
  fs/btrfs/disk-io.c     | 11 ++++++
  fs/btrfs/super.c       |  1 +
  fs/btrfs/volumes.c     |  5 +++
  fs/btrfs/zoned.c       | 81 ++++++++++++++++++++++++++++++++++++++++++
  fs/btrfs/zoned.h       | 26 ++++++++++++++
  7 files changed, 142 insertions(+)
fs/btrfs/ctree.h +11 -0
index aac3d6f4e35b..453f41ca024e 100644
@@ -948,6 +948,12 @@ struct btrfs_fs_info {
  	/* Type of exclusive operation running */
  	unsigned long exclusive_operation;
 
+	/* Zone size when in ZONED mode */
+	union {
+		u64 zone_size;
+		u64 zoned;
+	};
+
  #ifdef CONFIG_BTRFS_FS_REF_VERIFY
  	spinlock_t ref_verify_lock;
  	struct rb_root block_tree;
@@ -3595,4 +3601,9 @@ static inline int btrfs_is_testing(struct btrfs_fs_info *fs_info)
  }
  #endif
 
+static inline bool btrfs_is_zoned(struct btrfs_fs_info *fs_info)
+{
+	return fs_info->zoned != 0;
+}
+
  #endif
fs/btrfs/dev-replace.c +7 -0
index 6f6d77224c2b..db87f1aa604b 100644
@@ -238,6 +238,13 @@ static int btrfs_init_dev_replace_tgtdev(struct btrfs_fs_info *fs_info,
  		return PTR_ERR(bdev);
  	}
 
+	if (!btrfs_check_device_zone_type(fs_info, bdev)) {
+		btrfs_err(fs_info,
+			  "dev-replace: zoned type of target device mismatch with filesystem");
+		ret = -EINVAL;
+		goto error;
+	}
+
  	sync_blockdev(bdev);
 
Quoted 6 lines

fs/btrfs/dev-replace.c static int btrfs_init_dev_replace_tgtdev(struct btrfs_fs_info *fs_info, · line 250

Quoted 1 line
fs/btrfs/dev-replace.c +0 -0
  	list_for_each_entry(device, &fs_info->fs_devices->devices, dev_list) {

I am not sure if it is done in some other patch. But we still have to check for

  (model == BLK_ZONED_HA && incompat_zoned))

Do you really mean BLK_ZONED_HA, ie. host-aware (HA)? btrfs_check_device_zone_type checks for _HM.

Quoted 2 lines

right? What if in a non-zoned FS, a zoned device is added through the replace. No?

The types of devices cannot mix, yeah. So I'd like to know the answer as well.

Quoted 17 lines
Quoted 4 lines
fs/btrfs/volumes.c +0 -0
@@ -2518,6 +2518,11 @@ int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path
  	if (IS_ERR(bdev))

fs/btrfs/volumes.c int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path · line 2528

Quoted 10 lines
fs/btrfs/volumes.c +5 -0
  		return PTR_ERR(bdev);
 
+	if (!btrfs_check_device_zone_type(fs_info, bdev)) {
+		ret = -EINVAL;
+		goto error;
+	}
+
  	if (fs_devices->seeding) {
  		seeding_dev = 1;
  		down_write(&sb->s_umount);

Same here too. It can also happen that a zone device is added to a non zoned fs.

#39 lore raw

Re: [PATCH v10 05/41] btrfs: check and enable ZONED mode

Quoted 69 lines · David Sterba
Quoted 68 lines · Anand Jain
Quoted 53 lines · Naohiro Aota

This commit introduces the function btrfs_check_zoned_mode() to check if ZONED flag is enabled on the file system and if the file system consists of zoned devices with equal zone size.

---
   fs/btrfs/ctree.h       | 11 ++++++
   fs/btrfs/dev-replace.c |  7 ++++
   fs/btrfs/disk-io.c     | 11 ++++++
   fs/btrfs/super.c       |  1 +
   fs/btrfs/volumes.c     |  5 +++
   fs/btrfs/zoned.c       | 81 ++++++++++++++++++++++++++++++++++++++++++
   fs/btrfs/zoned.h       | 26 ++++++++++++++
   7 files changed, 142 insertions(+)
fs/btrfs/ctree.h +11 -0
index aac3d6f4e35b..453f41ca024e 100644
@@ -948,6 +948,12 @@ struct btrfs_fs_info {
   	/* Type of exclusive operation running */
   	unsigned long exclusive_operation;
 
+	/* Zone size when in ZONED mode */
+	union {
+		u64 zone_size;
+		u64 zoned;
+	};
+
   #ifdef CONFIG_BTRFS_FS_REF_VERIFY
   	spinlock_t ref_verify_lock;
   	struct rb_root block_tree;
@@ -3595,4 +3601,9 @@ static inline int btrfs_is_testing(struct btrfs_fs_info *fs_info)
   }
   #endif
 
+static inline bool btrfs_is_zoned(struct btrfs_fs_info *fs_info)
+{
+	return fs_info->zoned != 0;
+}
+
   #endif
fs/btrfs/dev-replace.c +0 -0
index 6f6d77224c2b..db87f1aa604b 100644
@@ -238,6 +238,13 @@ static int btrfs_init_dev_replace_tgtdev(struct btrfs_fs_info *fs_info,
   		return PTR_ERR(bdev);
   	}
 

fs/btrfs/dev-replace.c static int btrfs_init_dev_replace_tgtdev(struct btrfs_fs_info *fs_info, · line 250

Quoted 10 lines
fs/btrfs/dev-replace.c +7 -0
+	if (!btrfs_check_device_zone_type(fs_info, bdev)) {
+		btrfs_err(fs_info,
+			  "dev-replace: zoned type of target device mismatch with filesystem");
+		ret = -EINVAL;
+		goto error;
+	}
+
   	sync_blockdev(bdev);
 
   	list_for_each_entry(device, &fs_info->fs_devices->devices, dev_list) {

I am not sure if it is done in some other patch. But we still have to check for

Quoted 4 lines
Quoted 1 line
   (model == BLK_ZONED_HA && incompat_zoned))

Do you really mean BLK_ZONED_HA, ie. host-aware (HA)? btrfs_check_device_zone_type checks for _HM.

Still confusing to me. The below function, which is part of this patch, says we don't support BLK_ZONED_HM. So does it mean we allow BLK_ZONED_HA only?

+static inline bool btrfs_check_device_zone_type(struct btrfs_fs_info *fs_info,
+ struct block_device *bdev)
+{
+ u64 zone_size;
+
+ if (btrfs_is_zoned(fs_info)) {
+ zone_size = (u64)bdev_zone_sectors(bdev) << SECTOR_SHIFT;
+ /* Do not allow non-zoned device */
+ return bdev_is_zoned(bdev) && fs_info->zone_size == zone_size;
+ }
+
+ /* Do not allow Host Manged zoned device */
+ return bdev_zoned_model(bdev) != BLK_ZONED_HM;
+}

Also, if there is a new type of zoned device in the future, the older kernel should be able to reject the newer zone device types.

And, if possible could you rename above function to btrfs_zone_type_is_valid(). Or better.

Quoted 24 lines
Quoted 2 lines

right? What if in a non-zoned FS, a zoned device is added through the replace. No?

The types of devices cannot mix, yeah. So I'd like to know the answer as well.

Quoted 17 lines
Quoted 4 lines
fs/btrfs/volumes.c +0 -0
@@ -2518,6 +2518,11 @@ int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path
   	if (IS_ERR(bdev))

fs/btrfs/volumes.c int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path · line 2528

Quoted 10 lines
fs/btrfs/volumes.c +5 -0
   		return PTR_ERR(bdev);
 
+	if (!btrfs_check_device_zone_type(fs_info, bdev)) {
+		ret = -EINVAL;
+		goto error;
+	}
+
   	if (fs_devices->seeding) {
   		seeding_dev = 1;
   		down_write(&sb->s_umount);

Same here too. It can also happen that a zone device is added to a non zoned fs.

Thanks.

#40 lore raw

Re: [PATCH v10 05/41] btrfs: check and enable ZONED mode

Quoted 82 lines · Anand Jain
Quoted 69 lines · David Sterba
Quoted 68 lines · Anand Jain
Quoted 53 lines · Naohiro Aota

This commit introduces the function btrfs_check_zoned_mode() to check if ZONED flag is enabled on the file system and if the file system consists of zoned devices with equal zone size.

---
   fs/btrfs/ctree.h       | 11 ++++++
   fs/btrfs/dev-replace.c |  7 ++++
   fs/btrfs/disk-io.c     | 11 ++++++
   fs/btrfs/super.c       |  1 +
   fs/btrfs/volumes.c     |  5 +++
   fs/btrfs/zoned.c       | 81 ++++++++++++++++++++++++++++++++++++++++++
   fs/btrfs/zoned.h       | 26 ++++++++++++++
   7 files changed, 142 insertions(+)
fs/btrfs/ctree.h +11 -0
index aac3d6f4e35b..453f41ca024e 100644
@@ -948,6 +948,12 @@ struct btrfs_fs_info {
   	/* Type of exclusive operation running */
   	unsigned long exclusive_operation;
 
+	/* Zone size when in ZONED mode */
+	union {
+		u64 zone_size;
+		u64 zoned;
+	};
+
   #ifdef CONFIG_BTRFS_FS_REF_VERIFY
   	spinlock_t ref_verify_lock;
   	struct rb_root block_tree;
@@ -3595,4 +3601,9 @@ static inline int btrfs_is_testing(struct btrfs_fs_info *fs_info)
   }
   #endif
 
+static inline bool btrfs_is_zoned(struct btrfs_fs_info *fs_info)
+{
+	return fs_info->zoned != 0;
+}
+
   #endif
fs/btrfs/dev-replace.c +0 -0
index 6f6d77224c2b..db87f1aa604b 100644
@@ -238,6 +238,13 @@ static int btrfs_init_dev_replace_tgtdev(struct btrfs_fs_info *fs_info,
   		return PTR_ERR(bdev);
   	}
 

fs/btrfs/dev-replace.c static int btrfs_init_dev_replace_tgtdev(struct btrfs_fs_info *fs_info, · line 250

Quoted 10 lines
fs/btrfs/dev-replace.c +7 -0
+	if (!btrfs_check_device_zone_type(fs_info, bdev)) {
+		btrfs_err(fs_info,
+			  "dev-replace: zoned type of target device mismatch with filesystem");
+		ret = -EINVAL;
+		goto error;
+	}
+
   	sync_blockdev(bdev);
 
   	list_for_each_entry(device, &fs_info->fs_devices->devices, dev_list) {

I am not sure if it is done in some other patch. But we still have to check for

Quoted 4 lines
Quoted 1 line
   (model == BLK_ZONED_HA && incompat_zoned))

Do you really mean BLK_ZONED_HA, ie. host-aware (HA)? btrfs_check_device_zone_type checks for _HM.

Still confusing to me. The below function, which is part of this patch, says we don't support BLK_ZONED_HM. So does it mean we allow BLK_ZONED_HA only?

+static inline bool btrfs_check_device_zone_type(struct btrfs_fs_info *fs_info,

fs/btrfs/zoned.h static inline void btrfs_dev_clear_zone_empty(struct btrfs_device *device, · line 109

Quoted 7 lines
fs/btrfs/zoned.h +7 -0
+						struct block_device *bdev)
+{
+	u64 zone_size;
+
+	if (btrfs_is_zoned(fs_info)) {
+		zone_size = (u64)bdev_zone_sectors(bdev) << SECTOR_SHIFT;
+		/* Do not allow non-zoned device */

This comment does not make sense. It should be:

		/* Only allow zoned devices with the same zone size */

fs/btrfs/zoned.h static inline void btrfs_dev_clear_zone_empty(struct btrfs_device *device, · line 114

Quoted 5 lines
fs/btrfs/zoned.h +5 -0
+		return bdev_is_zoned(bdev) && fs_info->zone_size == zone_size;
+	}
+
+	/* Do not allow Host Manged zoned device */
+	return bdev_zoned_model(bdev) != BLK_ZONED_HM;

The comment is also wrong. It should read:

	/* Allow only host managed zoned devices */

This is because we decided to treat host aware devices in the same way as regular block devices, since HA drives are backward compatible with regular block devices.

Quoted 38 lines

+}

Also, if there is a new type of zoned device in the future, the older kernel should be able to reject the newer zone device types.

And, if possible could you rename above function to btrfs_zone_type_is_valid(). Or better.

Quoted 24 lines
Quoted 2 lines

right? What if in a non-zoned FS, a zoned device is added through the replace. No?

The types of devices cannot mix, yeah. So I'd like to know the answer as well.

Quoted 17 lines
Quoted 4 lines
fs/btrfs/volumes.c +0 -0
@@ -2518,6 +2518,11 @@ int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path
   	if (IS_ERR(bdev))

fs/btrfs/volumes.c int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path · line 2528

Quoted 10 lines
fs/btrfs/volumes.c +5 -0
   		return PTR_ERR(bdev);
 
+	if (!btrfs_check_device_zone_type(fs_info, bdev)) {
+		ret = -EINVAL;
+		goto error;
+	}
+
   	if (fs_devices->seeding) {
   		seeding_dev = 1;
   		down_write(&sb->s_umount);

Same here too. It can also happen that a zone device is added to a non zoned fs.

Thanks.

Signature · 2 lines
Damien Le Moal
Western Digital Research
#41 lore raw

Re: [PATCH v10 05/41] btrfs: check and enable ZONED mode

Quoted 103 lines · Damien Le Moal
Quoted 82 lines · Anand Jain
Quoted 69 lines · David Sterba
Quoted 68 lines · Anand Jain
Quoted 53 lines · Naohiro Aota

This commit introduces the function btrfs_check_zoned_mode() to check if ZONED flag is enabled on the file system and if the file system consists of zoned devices with equal zone size.

---
    fs/btrfs/ctree.h       | 11 ++++++
    fs/btrfs/dev-replace.c |  7 ++++
    fs/btrfs/disk-io.c     | 11 ++++++
    fs/btrfs/super.c       |  1 +
    fs/btrfs/volumes.c     |  5 +++
    fs/btrfs/zoned.c       | 81 ++++++++++++++++++++++++++++++++++++++++++
    fs/btrfs/zoned.h       | 26 ++++++++++++++
    7 files changed, 142 insertions(+)
fs/btrfs/ctree.h +11 -0
index aac3d6f4e35b..453f41ca024e 100644
@@ -948,6 +948,12 @@ struct btrfs_fs_info {
    	/* Type of exclusive operation running */
    	unsigned long exclusive_operation;
 
+	/* Zone size when in ZONED mode */
+	union {
+		u64 zone_size;
+		u64 zoned;
+	};
+
    #ifdef CONFIG_BTRFS_FS_REF_VERIFY
    	spinlock_t ref_verify_lock;
    	struct rb_root block_tree;
@@ -3595,4 +3601,9 @@ static inline int btrfs_is_testing(struct btrfs_fs_info *fs_info)
    }
    #endif
 
+static inline bool btrfs_is_zoned(struct btrfs_fs_info *fs_info)
+{
+	return fs_info->zoned != 0;
+}
+
    #endif
fs/btrfs/dev-replace.c +0 -0
index 6f6d77224c2b..db87f1aa604b 100644
@@ -238,6 +238,13 @@ static int btrfs_init_dev_replace_tgtdev(struct btrfs_fs_info *fs_info,
    		return PTR_ERR(bdev);
    	}
 

fs/btrfs/dev-replace.c static int btrfs_init_dev_replace_tgtdev(struct btrfs_fs_info *fs_info, · line 250

Quoted 10 lines
fs/btrfs/dev-replace.c +7 -0
+	if (!btrfs_check_device_zone_type(fs_info, bdev)) {
+		btrfs_err(fs_info,
+			  "dev-replace: zoned type of target device mismatch with filesystem");
+		ret = -EINVAL;
+		goto error;
+	}
+
    	sync_blockdev(bdev);
 
    	list_for_each_entry(device, &fs_info->fs_devices->devices, dev_list) {

I am not sure if it is done in some other patch. But we still have to check for

Quoted 4 lines
Quoted 1 line
    (model == BLK_ZONED_HA && incompat_zoned))

Do you really mean BLK_ZONED_HA, ie. host-aware (HA)? btrfs_check_device_zone_type checks for _HM.

Still confusing to me. The below function, which is part of this patch, says we don't support BLK_ZONED_HM. So does it mean we allow BLK_ZONED_HA only?

+static inline bool btrfs_check_device_zone_type(struct btrfs_fs_info *fs_info,

fs/btrfs/zoned.h static inline void btrfs_dev_clear_zone_empty(struct btrfs_device *device, · line 109

Quoted 7 lines
fs/btrfs/zoned.h +7 -0
+						struct block_device *bdev)
+{
+	u64 zone_size;
+
+	if (btrfs_is_zoned(fs_info)) {
+		zone_size = (u64)bdev_zone_sectors(bdev) << SECTOR_SHIFT;
+		/* Do not allow non-zoned device */

This comment does not make sense. It should be:

		/* Only allow zoned devices with the same zone size */

fs/btrfs/zoned.h static inline void btrfs_dev_clear_zone_empty(struct btrfs_device *device, · line 114

Quoted 5 lines
fs/btrfs/zoned.h +5 -0
+		return bdev_is_zoned(bdev) && fs_info->zone_size == zone_size;
+	}
+
+	/* Do not allow Host Manged zoned device */
+	return bdev_zoned_model(bdev) != BLK_ZONED_HM;

The comment is also wrong. It should read:

Quoted 5 lines
	/* Allow only host managed zoned devices */

This is because we decided to treat host aware devices in the same way as regular block devices, since HA drives are backward compatible with regular block devices.

Yeah, I read about them, but I have questions like do an FS work on top of a BLK_ZONED_HA without modification?
Are we ok to replace an HM device with a HA device? Or add a HA device
to a btrfs on an HM device.

Thanks.

Quoted 41 lines
Quoted 39 lines

+}

Also, if there is a new type of zoned device in the future, the older kernel should be able to reject the newer zone device types.

And, if possible could you rename above function to btrfs_zone_type_is_valid(). Or better.

Quoted 24 lines
Quoted 2 lines

right? What if in a non-zoned FS, a zoned device is added through the replace. No?

The types of devices cannot mix, yeah. So I'd like to know the answer as well.

Quoted 17 lines
Quoted 4 lines
fs/btrfs/volumes.c +0 -0
@@ -2518,6 +2518,11 @@ int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path
    	if (IS_ERR(bdev))

fs/btrfs/volumes.c int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path · line 2528

Quoted 10 lines
fs/btrfs/volumes.c +5 -0
    		return PTR_ERR(bdev);
 
+	if (!btrfs_check_device_zone_type(fs_info, bdev)) {
+		ret = -EINVAL;
+		goto error;
+	}
+
    	if (fs_devices->seeding) {
    		seeding_dev = 1;
    		down_write(&sb->s_umount);

Same here too. It can also happen that a zone device is added to a non zoned fs.

Thanks.

#42 lore raw

Re: [PATCH v10 05/41] btrfs: check and enable ZONED mode

Quoted 106 lines · Anand Jain
Quoted 105 lines · Damien Le Moal
Quoted 82 lines · Anand Jain
Quoted 69 lines · David Sterba
Quoted 68 lines · Anand Jain
Quoted 53 lines · Naohiro Aota

This commit introduces the function btrfs_check_zoned_mode() to check if ZONED flag is enabled on the file system and if the file system consists of zoned devices with equal zone size.

---
    fs/btrfs/ctree.h       | 11 ++++++
    fs/btrfs/dev-replace.c |  7 ++++
    fs/btrfs/disk-io.c     | 11 ++++++
    fs/btrfs/super.c       |  1 +
    fs/btrfs/volumes.c     |  5 +++
    fs/btrfs/zoned.c       | 81 ++++++++++++++++++++++++++++++++++++++++++
    fs/btrfs/zoned.h       | 26 ++++++++++++++
    7 files changed, 142 insertions(+)
fs/btrfs/ctree.h +11 -0
index aac3d6f4e35b..453f41ca024e 100644
@@ -948,6 +948,12 @@ struct btrfs_fs_info {
    	/* Type of exclusive operation running */
    	unsigned long exclusive_operation;
 
+	/* Zone size when in ZONED mode */
+	union {
+		u64 zone_size;
+		u64 zoned;
+	};
+
    #ifdef CONFIG_BTRFS_FS_REF_VERIFY
    	spinlock_t ref_verify_lock;
    	struct rb_root block_tree;
@@ -3595,4 +3601,9 @@ static inline int btrfs_is_testing(struct btrfs_fs_info *fs_info)
    }
    #endif
 
+static inline bool btrfs_is_zoned(struct btrfs_fs_info *fs_info)
+{
+	return fs_info->zoned != 0;
+}
+
    #endif
fs/btrfs/dev-replace.c +0 -0
index 6f6d77224c2b..db87f1aa604b 100644
@@ -238,6 +238,13 @@ static int btrfs_init_dev_replace_tgtdev(struct btrfs_fs_info *fs_info,
    		return PTR_ERR(bdev);
    	}
 

fs/btrfs/dev-replace.c static int btrfs_init_dev_replace_tgtdev(struct btrfs_fs_info *fs_info, · line 250

Quoted 10 lines
fs/btrfs/dev-replace.c +7 -0
+	if (!btrfs_check_device_zone_type(fs_info, bdev)) {
+		btrfs_err(fs_info,
+			  "dev-replace: zoned type of target device mismatch with filesystem");
+		ret = -EINVAL;
+		goto error;
+	}
+
    	sync_blockdev(bdev);
 
    	list_for_each_entry(device, &fs_info->fs_devices->devices, dev_list) {

I am not sure if it is done in some other patch. But we still have to check for

Quoted 4 lines
Quoted 1 line
    (model == BLK_ZONED_HA && incompat_zoned))

Do you really mean BLK_ZONED_HA, ie. host-aware (HA)? btrfs_check_device_zone_type checks for _HM.

Still confusing to me. The below function, which is part of this patch, says we don't support BLK_ZONED_HM. So does it mean we allow BLK_ZONED_HA only?

+static inline bool btrfs_check_device_zone_type(struct btrfs_fs_info *fs_info,

fs/btrfs/zoned.h static inline void btrfs_dev_clear_zone_empty(struct btrfs_device *device, · line 109

Quoted 7 lines
fs/btrfs/zoned.h +7 -0
+						struct block_device *bdev)
+{
+	u64 zone_size;
+
+	if (btrfs_is_zoned(fs_info)) {
+		zone_size = (u64)bdev_zone_sectors(bdev) << SECTOR_SHIFT;
+		/* Do not allow non-zoned device */

This comment does not make sense. It should be:

		/* Only allow zoned devices with the same zone size */

fs/btrfs/zoned.h static inline void btrfs_dev_clear_zone_empty(struct btrfs_device *device, · line 114

Quoted 5 lines
fs/btrfs/zoned.h +5 -0
+		return bdev_is_zoned(bdev) && fs_info->zone_size == zone_size;
+	}
+
+	/* Do not allow Host Manged zoned device */
+	return bdev_zoned_model(bdev) != BLK_ZONED_HM;

The comment is also wrong. It should read:

	/* Allow only host managed zoned devices */
Quoted 7 lines
Quoted 3 lines

This is because we decided to treat host aware devices in the same way as regular block devices, since HA drives are backward compatible with regular block devices.

Yeah, I read about them, but I have questions like do an FS work on top of a BLK_ZONED_HA without modification?

Yes. These drives are fully backward compatible and accept random writes anywhere. Performance however is potentially a different story as the drive will eventually need to do internal garbage collection of some sort, exactly like an SSD, but definitely not at SSD speeds :)

Quoted 2 lines

Are we ok to replace an HM device with a HA device? Or add a HA device
to a btrfs on an HM device.

We have a choice here: we can treat HA drives as regular devices or treat them as HM devices. Anything in between does not make sense. I am fine either way, the main reason being that there are no HA drive on the market today that I know of (this model did not have a lot of success due to the potentially very unpredictable performance depending on the use case).

So the simplest thing to do is, in my opinion, to ignore their "zone" characteristics and treat them as regular disks. But treating them as HM drives is a simple to do too.

Of note is that a host-aware drive will be reported by the block layer as BLK_ZONED_HA only as long as the drive does not have any partition. If it does, then the block layer will treat the drive as a regular disk.

Quoted 46 lines

Thanks.

Quoted 41 lines
Quoted 39 lines

+}

Also, if there is a new type of zoned device in the future, the older kernel should be able to reject the newer zone device types.

And, if possible could you rename above function to btrfs_zone_type_is_valid(). Or better.

Quoted 24 lines
Quoted 2 lines

right? What if in a non-zoned FS, a zoned device is added through the replace. No?

The types of devices cannot mix, yeah. So I'd like to know the answer as well.

Quoted 17 lines
Quoted 4 lines
fs/btrfs/volumes.c +0 -0
@@ -2518,6 +2518,11 @@ int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path
    	if (IS_ERR(bdev))

fs/btrfs/volumes.c int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path · line 2528

Quoted 10 lines
fs/btrfs/volumes.c +5 -0
    		return PTR_ERR(bdev);
 
+	if (!btrfs_check_device_zone_type(fs_info, bdev)) {
+		ret = -EINVAL;
+		goto error;
+	}
+
    	if (fs_devices->seeding) {
    		seeding_dev = 1;
    		down_write(&sb->s_umount);

Same here too. It can also happen that a zone device is added to a non zoned fs.

Thanks.

Signature · 2 lines
Damien Le Moal
Western Digital Research
#43 lore raw

Re: [PATCH v10 05/41] btrfs: check and enable ZONED mode

Quoted 129 lines · Damien Le Moal
Quoted 106 lines · Anand Jain
Quoted 105 lines · Damien Le Moal
Quoted 82 lines · Anand Jain
Quoted 69 lines · David Sterba
Quoted 68 lines · Anand Jain
Quoted 53 lines · Naohiro Aota

This commit introduces the function btrfs_check_zoned_mode() to check if ZONED flag is enabled on the file system and if the file system consists of zoned devices with equal zone size.

---
     fs/btrfs/ctree.h       | 11 ++++++
     fs/btrfs/dev-replace.c |  7 ++++
     fs/btrfs/disk-io.c     | 11 ++++++
     fs/btrfs/super.c       |  1 +
     fs/btrfs/volumes.c     |  5 +++
     fs/btrfs/zoned.c       | 81 ++++++++++++++++++++++++++++++++++++++++++
     fs/btrfs/zoned.h       | 26 ++++++++++++++
     7 files changed, 142 insertions(+)
fs/btrfs/ctree.h +11 -0
index aac3d6f4e35b..453f41ca024e 100644
@@ -948,6 +948,12 @@ struct btrfs_fs_info {
     	/* Type of exclusive operation running */
     	unsigned long exclusive_operation;
 
+	/* Zone size when in ZONED mode */
+	union {
+		u64 zone_size;
+		u64 zoned;
+	};
+
     #ifdef CONFIG_BTRFS_FS_REF_VERIFY
     	spinlock_t ref_verify_lock;
     	struct rb_root block_tree;
@@ -3595,4 +3601,9 @@ static inline int btrfs_is_testing(struct btrfs_fs_info *fs_info)
     }
     #endif
 
+static inline bool btrfs_is_zoned(struct btrfs_fs_info *fs_info)
+{
+	return fs_info->zoned != 0;
+}
+
     #endif
fs/btrfs/dev-replace.c +0 -0
index 6f6d77224c2b..db87f1aa604b 100644
@@ -238,6 +238,13 @@ static int btrfs_init_dev_replace_tgtdev(struct btrfs_fs_info *fs_info,
     		return PTR_ERR(bdev);
     	}
 

fs/btrfs/dev-replace.c static int btrfs_init_dev_replace_tgtdev(struct btrfs_fs_info *fs_info, · line 250

Quoted 10 lines
fs/btrfs/dev-replace.c +7 -0
+	if (!btrfs_check_device_zone_type(fs_info, bdev)) {
+		btrfs_err(fs_info,
+			  "dev-replace: zoned type of target device mismatch with filesystem");
+		ret = -EINVAL;
+		goto error;
+	}
+
     	sync_blockdev(bdev);
 
     	list_for_each_entry(device, &fs_info->fs_devices->devices, dev_list) {

I am not sure if it is done in some other patch. But we still have to check for

Quoted 4 lines
Quoted 1 line
     (model == BLK_ZONED_HA && incompat_zoned))

Do you really mean BLK_ZONED_HA, ie. host-aware (HA)? btrfs_check_device_zone_type checks for _HM.

Still confusing to me. The below function, which is part of this patch, says we don't support BLK_ZONED_HM. So does it mean we allow BLK_ZONED_HA only?

+static inline bool btrfs_check_device_zone_type(struct btrfs_fs_info *fs_info,

Review

Quoted 7 lines
(quoted diff) +7 -0
+						struct block_device *bdev)
+{
+	u64 zone_size;
+
+	if (btrfs_is_zoned(fs_info)) {
+		zone_size = (u64)bdev_zone_sectors(bdev) << SECTOR_SHIFT;
+		/* Do not allow non-zoned device */

This comment does not make sense. It should be:

		/* Only allow zoned devices with the same zone size */

Review

Quoted 5 lines
(quoted diff) +5 -0
+		return bdev_is_zoned(bdev) && fs_info->zone_size == zone_size;
+	}
+
+	/* Do not allow Host Manged zoned device */
+	return bdev_zoned_model(bdev) != BLK_ZONED_HM;

The comment is also wrong. It should read:

	/* Allow only host managed zoned devices */
Quoted 7 lines
Quoted 3 lines

This is because we decided to treat host aware devices in the same way as regular block devices, since HA drives are backward compatible with regular block devices.

Yeah, I read about them, but I have questions like do an FS work on top of a BLK_ZONED_HA without modification?

Yes. These drives are fully backward compatible and accept random writes anywhere. Performance however is potentially a different story as the drive will eventually need to do internal garbage collection of some sort, exactly like an SSD, but definitely not at SSD speeds :)

Quoted 2 lines

Are we ok to replace an HM device with a HA device? Or add a HA device
to a btrfs on an HM device.

We have a choice here: we can treat HA drives as regular devices or treat them as HM devices. Anything in between does not make sense. I am fine either way, the main reason being that there are no HA drive on the market today that I know of (this model did not have a lot of success due to the potentially very unpredictable performance depending on the use case).

Quoted 6 lines

So the simplest thing to do is, in my opinion, to ignore their "zone" characteristics and treat them as regular disks. But treating them as HM drives is a simple to do too.

Quoted 1 line

Of note is that a host-aware drive will be reported by the block layer as

BLK_ZONED_HA only as long as the drive does not have any partition. If it does, then the block layer will treat the drive as a regular disk.

IMO. For now, it is better to check for the BLK_ZONED_HA explicitly in a non-zoned-btrfs. And check for BLK_ZONED_HM explicitly in a zoned-btrfs. This way, if there is another type of BLK_ZONED_xx in the future, we have the opportunity to review to support it. As below [1]...

[1]
bool btrfs_check_device_type()

{
	if (bdev_is_zoned()) {
		if (btrfs_is_zoned())
			if (bdev_zoned_model == BLK_ZONED_HM)
			/* also check the zone_size. */
				return true;
		else
			if (bdev_zoned_model == BLK_ZONED_HA)
			/* a regular device and FS, no zone_size to check I think? */
				return true;
	} else {
		if (!btrfs_is_zoned())
			return true
	}

	return false;
}

Thanks.

Quoted 42 lines
Quoted 42 lines

Thanks.

Quoted 38 lines
Quoted 38 lines

+}

Also, if there is a new type of zoned device in the future, the older kernel should be able to reject the newer zone device types.

And, if possible could you rename above function to btrfs_zone_type_is_valid(). Or better.

Quoted 24 lines
Quoted 2 lines

right? What if in a non-zoned FS, a zoned device is added through the replace. No?

The types of devices cannot mix, yeah. So I'd like to know the answer as well.

Quoted 17 lines
Quoted 4 lines
fs/btrfs/volumes.c +0 -0
@@ -2518,6 +2518,11 @@ int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path
     	if (IS_ERR(bdev))

fs/btrfs/volumes.c int btrfs_init_new_device(struct btrfs_fs_info *fs_info, const char *device_path · line 2528

Quoted 10 lines
fs/btrfs/volumes.c +5 -0
     		return PTR_ERR(bdev);
 
+	if (!btrfs_check_device_zone_type(fs_info, bdev)) {
+		ret = -EINVAL;
+		goto error;
+	}
+
     	if (fs_devices->seeding) {
     		seeding_dev = 1;
     		down_write(&sb->s_umount);

Same here too. It can also happen that a zone device is added to a non zoned fs.

Thanks.

#44 lore raw

Re: [PATCH v10 05/41] btrfs: check and enable ZONED mode

Quoted 133 lines · Anand Jain
Quoted 132 lines · Damien Le Moal
Quoted 106 lines · Anand Jain
Quoted 105 lines · Damien Le Moal
Quoted 82 lines · Anand Jain
Quoted 69 lines · David Sterba
Quoted 68 lines · Anand Jain
Quoted 53 lines · Naohiro Aota

This commit introduces the function btrfs_check_zoned_mode() to check if ZONED flag is enabled on the file system and if the file system consists of zoned devices with equal zone size.

---
     fs/btrfs/ctree.h       | 11 ++++++
     fs/btrfs/dev-replace.c |  7 ++++
     fs/btrfs/disk-io.c     | 11 ++++++
     fs/btrfs/super.c       |  1 +
     fs/btrfs/volumes.c     |  5 +++
     fs/btrfs/zoned.c       | 81 ++++++++++++++++++++++++++++++++++++++++++
     fs/btrfs/zoned.h       | 26 ++++++++++++++
     7 files changed, 142 insertions(+)
fs/btrfs/ctree.h +11 -0
index aac3d6f4e35b..453f41ca024e 100644
@@ -948,6 +948,12 @@ struct btrfs_fs_info {
     	/* Type of exclusive operation running */
     	unsigned long exclusive_operation;
 
+	/* Zone size when in ZONED mode */
+	union {
+		u64 zone_size;
+		u64 zoned;
+	};
+
     #ifdef CONFIG_BTRFS_FS_REF_VERIFY
     	spinlock_t ref_verify_lock;
     	struct rb_root block_tree;
@@ -3595,4 +3601,9 @@ static inline int btrfs_is_testing(struct btrfs_fs_info *fs_info)
     }
     #endif
 
+static inline bool btrfs_is_zoned(struct btrfs_fs_info *fs_info)
+{
+	return fs_info->zoned != 0;
+}
+
     #endif
fs/btrfs/dev-replace.c +0 -0
index 6f6d77224c2b..db87f1aa604b 100644
@@ -238,6 +238,13 @@ static int btrfs_init_dev_replace_tgtdev(struct btrfs_fs_info *fs_info,
     		return PTR_ERR(bdev);
     	}
 

fs/btrfs/dev-replace.c static int btrfs_init_dev_replace_tgtdev(struct btrfs_fs_info *fs_info, · line 250

Quoted 10 lines
fs/btrfs/dev-replace.c +7 -0
+	if (!btrfs_check_device_zone_type(fs_info, bdev)) {
+		btrfs_err(fs_info,
+			  "dev-replace: zoned type of target device mismatch with filesystem");
+		ret = -EINVAL;
+		goto error;
+	}
+
     	sync_blockdev(bdev);
 
     	list_for_each_entry(device, &fs_info->fs_devices->devices, dev_list) {

I am not sure if it is done in some other patch. But we still have to check for

Quoted 4 lines
Quoted 1 line
     (model == BLK_ZONED_HA && incompat_zoned))

Do you really mean BLK_ZONED_HA, ie. host-aware (HA)? btrfs_check_device_zone_type checks for _HM.

Still confusing to me. The below function, which is part of this patch, says we don't support BLK_ZONED_HM. So does it mean we allow BLK_ZONED_HA only?

+static inline bool btrfs_check_device_zone_type(struct btrfs_fs_info *fs_info,

Review

Quoted 7 lines
(quoted diff) +7 -0
+						struct block_device *bdev)
+{
+	u64 zone_size;
+
+	if (btrfs_is_zoned(fs_info)) {
+		zone_size = (u64)bdev_zone_sectors(bdev) << SECTOR_SHIFT;
+		/* Do not allow non-zoned device */

This comment does not make sense. It should be:

		/* Only allow zoned devices with the same zone size */

Review

Quoted 5 lines
(quoted diff) +5 -0
+		return bdev_is_zoned(bdev) && fs_info->zone_size == zone_size;
+	}
+
+	/* Do not allow Host Manged zoned device */
+	return bdev_zoned_model(bdev) != BLK_ZONED_HM;

The comment is also wrong. It should read:

	/* Allow only host managed zoned devices */
Quoted 7 lines
Quoted 3 lines

This is because we decided to treat host aware devices in the same way as regular block devices, since HA drives are backward compatible with regular block devices.

Yeah, I read about them, but I have questions like do an FS work on top of a BLK_ZONED_HA without modification?

Yes. These drives are fully backward compatible and accept random writes anywhere. Performance however is potentially a different story as the drive will eventually need to do internal garbage collection of some sort, exactly like an SSD, but definitely not at SSD speeds :)

Quoted 2 lines

Are we ok to replace an HM device with a HA device? Or add a HA device
to a btrfs on an HM device.

We have a choice here: we can treat HA drives as regular devices or treat them as HM devices. Anything in between does not make sense. I am fine either way, the main reason being that there are no HA drive on the market today that I know of (this model did not have a lot of success due to the potentially very unpredictable performance depending on the use case).

So the simplest thing to do is, in my opinion, to ignore their "zone" characteristics and treat them as regular disks. But treating them as HM drives is a simple to do too.

Quoted 6 lines
Quoted 3 lines
Quoted 1 line

Of note is that a host-aware drive will be reported by the block layer as

BLK_ZONED_HA only as long as the drive does not have any partition. If it does, then the block layer will treat the drive as a regular disk.

IMO. For now, it is better to check for the BLK_ZONED_HA explicitly in a non-zoned-btrfs. And check for BLK_ZONED_HM explicitly in a zoned-btrfs.

Sure, we can. But since HA drives are backward compatible, not sure the HA check for non-zoned make sense. As long as the zoned flag is not set, the drive can be used like a regular disk. If the user really want to use it as a zoned drive, then it can format with force selecting the zoned flag in btrfs super. Then the HA drive will be used as a zoned disk, exactly like HM disks.

Quoted 2 lines

This way, if there is another type of BLK_ZONED_xx in the future, we have the opportunity to review to support it. As below [1]...

It is very unlikely that we will see any other zone model. ZNS adopted the HM model in purpose, to avoid multiplying the possible models, making the ecosystem effort a nightmare.

Quoted 21 lines

[1]
bool btrfs_check_device_type()

{
	if (bdev_is_zoned()) {
		if (btrfs_is_zoned())
			if (bdev_zoned_model == BLK_ZONED_HM)
			/* also check the zone_size. */
				return true;
		else
			if (bdev_zoned_model == BLK_ZONED_HA)
			/* a regular device and FS, no zone_size to check I think? */
				return true;
	} else {
		if (!btrfs_is_zoned())
			return true
	}

	return false;
}
Quoted 1 line

Thanks.

Works for me. May be reverse the conditions to make things easier to read and understand:

bool btrfs_check_device_type()

{
	if (btrfs_is_zoned()) {
		if (bdev_is_zoned()) {
			/* also check the zone_size. */
			return true;
		}

		/*
		 * Regular device: emulate zones with zone size equal
		 * to device extent size.
		 */
		return true;
	}

	if (bdev_zoned_model == BLK_ZONED_HM) {
		/* Zoned HM device require zoned btrfs */
		return false;
	}

	/* Regular device or zoned HA device used as a regular device */
	return true;
}
Signature · 2 lines
Damien Le Moal
Western Digital Research
#45 lore raw

Re: [PATCH v10 05/41] btrfs: check and enable ZONED mode

Quoted 195 lines · Damien Le Moal
Quoted 133 lines · Anand Jain
Quoted 132 lines · Damien Le Moal
Quoted 106 lines · Anand Jain
Quoted 105 lines · Damien Le Moal
Quoted 82 lines · Anand Jain
Quoted 69 lines · David Sterba
Quoted 68 lines · Anand Jain
Quoted 53 lines · Naohiro Aota

This commit introduces the function btrfs_check_zoned_mode() to check if ZONED flag is enabled on the file system and if the file system consists of zoned devices with equal zone size.

---
      fs/btrfs/ctree.h       | 11 ++++++
      fs/btrfs/dev-replace.c |  7 ++++
      fs/btrfs/disk-io.c     | 11 ++++++
      fs/btrfs/super.c       |  1 +
      fs/btrfs/volumes.c     |  5 +++
      fs/btrfs/zoned.c       | 81 ++++++++++++++++++++++++++++++++++++++++++
      fs/btrfs/zoned.h       | 26 ++++++++++++++
      7 files changed, 142 insertions(+)
fs/btrfs/ctree.h +11 -0
index aac3d6f4e35b..453f41ca024e 100644
@@ -948,6 +948,12 @@ struct btrfs_fs_info {
      	/* Type of exclusive operation running */
      	unsigned long exclusive_operation;
 
+	/* Zone size when in ZONED mode */
+	union {
+		u64 zone_size;
+		u64 zoned;
+	};
+
      #ifdef CONFIG_BTRFS_FS_REF_VERIFY
      	spinlock_t ref_verify_lock;
      	struct rb_root block_tree;
@@ -3595,4 +3601,9 @@ static inline int btrfs_is_testing(struct btrfs_fs_info *fs_info)
      }
      #endif
 
+static inline bool btrfs_is_zoned(struct btrfs_fs_info *fs_info)
+{
+	return fs_info->zoned != 0;
+}
+
      #endif
fs/btrfs/dev-replace.c +0 -0
index 6f6d77224c2b..db87f1aa604b 100644
@@ -238,6 +238,13 @@ static int btrfs_init_dev_replace_tgtdev(struct btrfs_fs_info *fs_info,
      		return PTR_ERR(bdev);
      	}
 

fs/btrfs/dev-replace.c static int btrfs_init_dev_replace_tgtdev(struct btrfs_fs_info *fs_info, · line 250

Quoted 10 lines
fs/btrfs/dev-replace.c +7 -0
+	if (!btrfs_check_device_zone_type(fs_info, bdev)) {
+		btrfs_err(fs_info,
+			  "dev-replace: zoned type of target device mismatch with filesystem");
+		ret = -EINVAL;
+		goto error;
+	}
+
      	sync_blockdev(bdev);
 
      	list_for_each_entry(device, &fs_info->fs_devices->devices, dev_list) {

I am not sure if it is done in some other patch. But we still have to check for

Quoted 4 lines
Quoted 1 line
      (model == BLK_ZONED_HA && incompat_zoned))

Do you really mean BLK_ZONED_HA, ie. host-aware (HA)? btrfs_check_device_zone_type checks for _HM.

Still confusing to me. The below function, which is part of this patch, says we don't support BLK_ZONED_HM. So does it mean we allow BLK_ZONED_HA only?

+static inline bool btrfs_check_device_zone_type(struct btrfs_fs_info *fs_info,

Review

Quoted 7 lines
(quoted diff) +7 -0
+						struct block_device *bdev)
+{
+	u64 zone_size;
+
+	if (btrfs_is_zoned(fs_info)) {
+		zone_size = (u64)bdev_zone_sectors(bdev) << SECTOR_SHIFT;
+		/* Do not allow non-zoned device */

This comment does not make sense. It should be:

		/* Only allow zoned devices with the same zone size */

Review

Quoted 5 lines
(quoted diff) +5 -0
+		return bdev_is_zoned(bdev) && fs_info->zone_size == zone_size;
+	}
+
+	/* Do not allow Host Manged zoned device */
+	return bdev_zoned_model(bdev) != BLK_ZONED_HM;

The comment is also wrong. It should read:

	/* Allow only host managed zoned devices */
Quoted 7 lines
Quoted 3 lines

This is because we decided to treat host aware devices in the same way as regular block devices, since HA drives are backward compatible with regular block devices.

Yeah, I read about them, but I have questions like do an FS work on top of a BLK_ZONED_HA without modification?

Yes. These drives are fully backward compatible and accept random writes anywhere. Performance however is potentially a different story as the drive will eventually need to do internal garbage collection of some sort, exactly like an SSD, but definitely not at SSD speeds :)

Quoted 2 lines

Are we ok to replace an HM device with a HA device? Or add a HA device
to a btrfs on an HM device.

We have a choice here: we can treat HA drives as regular devices or treat them as HM devices. Anything in between does not make sense. I am fine either way, the main reason being that there are no HA drive on the market today that I know of (this model did not have a lot of success due to the potentially very unpredictable performance depending on the use case).

So the simplest thing to do is, in my opinion, to ignore their "zone" characteristics and treat them as regular disks. But treating them as HM drives is a simple to do too.

Quoted 6 lines
Quoted 3 lines
Quoted 1 line

Of note is that a host-aware drive will be reported by the block layer as

BLK_ZONED_HA only as long as the drive does not have any partition. If it does, then the block layer will treat the drive as a regular disk.

IMO. For now, it is better to check for the BLK_ZONED_HA explicitly in a non-zoned-btrfs. And check for BLK_ZONED_HM explicitly in a zoned-btrfs.

Sure, we can. But since HA drives are backward compatible, not sure the HA check for non-zoned make sense. As long as the zoned flag is not set, the drive can be used like a regular disk. If the user really want to use it as a zoned drive, then it can format with force selecting the zoned flag in btrfs super. Then the HA drive will be used as a zoned disk, exactly like HM disks.

Quoted 2 lines

This way, if there is another type of BLK_ZONED_xx in the future, we have the opportunity to review to support it. As below [1]...

It is very unlikely that we will see any other zone model. ZNS adopted the HM model in purpose, to avoid multiplying the possible models, making the ecosystem effort a nightmare.

Quoted 21 lines

[1]
bool btrfs_check_device_type()

{
	if (bdev_is_zoned()) {
		if (btrfs_is_zoned())
			if (bdev_zoned_model == BLK_ZONED_HM)
			/* also check the zone_size. */
				return true;
		else
			if (bdev_zoned_model == BLK_ZONED_HA)
			/* a regular device and FS, no zone_size to check I think? */
				return true;
	} else {
		if (!btrfs_is_zoned())
			return true
	}

	return false;
}
Quoted 1 line

Thanks.

Works for me. May be reverse the conditions to make things easier to read and understand:

bool btrfs_check_device_type()

{
	if (btrfs_is_zoned()) {
		if (bdev_is_zoned()) {
			/* also check the zone_size. */
			return true;
		}

		/*
		 * Regular device: emulate zones with zone size equal
		 * to device extent size.
		 */
		return true;
	}
Quoted 10 lines
	if (bdev_zoned_model == BLK_ZONED_HM) {
		/* Zoned HM device require zoned btrfs */
		return false;
	}

	/* Regular device or zoned HA device used as a regular device */
	return true;
}

Yeah. Makes sense.

Thanks.

#46 lore raw

Re: [PATCH v10 05/41] btrfs: check and enable ZONED mode

In reply to Damien Le Moal
Quoted 5 lines · Damien Le Moal

Yes. These drives are fully backward compatible and accept random writes anywhere. Performance however is potentially a different story as the drive will eventually need to do internal garbage collection of some sort, exactly like an SSD, but definitely not at SSD speeds :)

Quoted 8 lines
Quoted 2 lines

Are we ok to replace an HM device with a HA device? Or add a HA device
to a btrfs on an HM device.

We have a choice here: we can treat HA drives as regular devices or treat them as HM devices. Anything in between does not make sense. I am fine either way, the main reason being that there are no HA drive on the market today that I know of (this model did not have a lot of success due to the potentially very unpredictable performance depending on the use case).

So there will be no testing against HA drives? And no btrfs developers will have one? And they have very different timing and possibly failure modes from "normal" disks when they do GC?

I think there is no option but to disallow them. If HA drives start to appear in significant numbers then that would be easy enough to change, after suitable testing.

Quoted 3 lines

Of note is that a host-aware drive will be reported by the block layer as BLK_ZONED_HA only as long as the drive does not have any partition. If it does, then the block layer will treat the drive as a regular disk.

That is a bit of a shame. With that unfortunate decision in the block layer, system managers need to realise that partitioning an HA disk means they may be entering territory untested by their filesystem.

#47 lore raw

Re: [PATCH v10 05/41] btrfs: check and enable ZONED mode

Quoted 15 lines · Graham Cobb
Quoted 5 lines · Damien Le Moal

Yes. These drives are fully backward compatible and accept random writes anywhere. Performance however is potentially a different story as the drive will eventually need to do internal garbage collection of some sort, exactly like an SSD, but definitely not at SSD speeds :)

Quoted 8 lines
Quoted 2 lines

Are we ok to replace an HM device with a HA device? Or add a HA device
to a btrfs on an HM device.

We have a choice here: we can treat HA drives as regular devices or treat them as HM devices. Anything in between does not make sense. I am fine either way, the main reason being that there are no HA drive on the market today that I know of (this model did not have a lot of success due to the potentially very unpredictable performance depending on the use case).

Quoted 7 lines

So there will be no testing against HA drives? And no btrfs developers will have one? And they have very different timing and possibly failure modes from "normal" disks when they do GC?

I think there is no option but to disallow them. If HA drives start to appear in significant numbers then that would be easy enough to change, after suitable testing.

Works for me. Even simpler :)

Quoted 8 lines
Quoted 4 lines

Of note is that a host-aware drive will be reported by the block layer as BLK_ZONED_HA only as long as the drive does not have any partition. If it does, then the block layer will treat the drive as a regular disk.

That is a bit of a shame. With that unfortunate decision in the block layer, system managers need to realise that partitioning an HA disk means they may be entering territory untested by their filesystem.

Well, not really. HA drives, per specifications, are backward compatible. If they are partitioned, the block layer will force a regular drive mode use, hiding their zoned interface, which is completely optional to use in the first place.

If by "untested territory" you mean the possibility of hitting drive FW bugs coming from the added complexity of internal GC, then I would argue that this is a common territory for any FS on any drive, especially SSDs: device FW bugs do exist and show up from time to time, even on the simplest of drives.

Signature · 2 lines
Damien Le Moal
Western Digital Research
#48 lore raw

Re: [PATCH v10 05/41] btrfs: check and enable ZONED mode

Quoted 9 lines · Damien Le Moal

Well, not really. HA drives, per specifications, are backward compatible. If they are partitioned, the block layer will force a regular drive mode use, hiding their zoned interface, which is completely optional to use in the first place.

If by "untested territory" you mean the possibility of hitting drive FW bugs coming from the added complexity of internal GC, then I would argue that this is a common territory for any FS on any drive, especially SSDs: device FW bugs do exist and show up from time to time, even on the simplest of drives.

Also note that most cheaper hard drives are using SMR internally, you'll run into the same GC "issues" as with a HA drive that is used for random writes.

#49 lore raw

Re: [PATCH v10 05/41] btrfs: check and enable ZONED mode

Quoted 6 lines · On 2020/12/01 20:11, [email protected] wrote:
Quoted 5 lines · Damien Le Moal

Well, not really. HA drives, per specifications, are backward compatible. If they are partitioned, the block layer will force a regular drive mode use, hiding their zoned interface, which is completely optional to use in the first place.

Quoted 8 lines
Quoted 4 lines

If by "untested territory" you mean the possibility of hitting drive FW bugs coming from the added complexity of internal GC, then I would argue that this is a common territory for any FS on any drive, especially SSDs: device FW bugs do exist and show up from time to time, even on the simplest of drives.

Also note that most cheaper hard drives are using SMR internally, you'll run into the same GC "issues" as with a HA drive that is used for random writes.

Indeed. Good point !

Signature · 2 lines
Damien Le Moal
Western Digital Research
#50 lore raw

PATCH v10 6/41 btrfs: introduce max_zone_append_size

In reply to Naohiro Aota

The zone append write command has a maximum IO size restriction it accepts. This is because a zone append write command cannot be split, as we ask the device to place the data into a specific target zone and the device responds with the actual written location of the data.

Introduce max_zone_append_size to zone_info and fs_info to track the value, so we can limit all I/O to a zoned block device that we want to write using the zone append command to the device's limits.

---
 fs/btrfs/ctree.h |  3 +++
 fs/btrfs/zoned.c | 17 +++++++++++++++--
 fs/btrfs/zoned.h |  1 +
 3 files changed, 19 insertions(+), 2 deletions(-)
fs/btrfs/ctree.h +3 -0
index 453f41ca024e..c70d3fcc62c2 100644
@@ -954,6 +954,9 @@ struct btrfs_fs_info {
 		u64 zoned;
 	};
 
+	/* Max size to emit ZONE_APPEND write command */
+	u64 max_zone_append_size;
+
 #ifdef CONFIG_BTRFS_FS_REF_VERIFY
 	spinlock_t ref_verify_lock;
 	struct rb_root block_tree;
fs/btrfs/zoned.c +15 -2
index 1223d5b0e411..2897432eb43c 100644
@@ -48,6 +48,7 @@ int btrfs_get_dev_zone_info(struct btrfs_device *device)
 {
 	struct btrfs_zoned_device_info *zone_info = NULL;
 	struct block_device *bdev = device->bdev;
+	struct request_queue *queue = bdev_get_queue(bdev);
 	sector_t nr_sectors = bdev->bd_part->nr_sects;
 	sector_t sector = 0;
 	struct blk_zone *zones = NULL;
@@ -69,6 +70,8 @@ int btrfs_get_dev_zone_info(struct btrfs_device *device)
 	ASSERT(is_power_of_2(zone_sectors));
 	zone_info->zone_size = (u64)zone_sectors << SECTOR_SHIFT;
 	zone_info->zone_size_shift = ilog2(zone_info->zone_size);
+	zone_info->max_zone_append_size =
+		(u64)queue_max_zone_append_sectors(queue) << SECTOR_SHIFT;
 	zone_info->nr_zones = nr_sectors >> ilog2(bdev_zone_sectors(bdev));
 	if (!IS_ALIGNED(nr_sectors, zone_sectors))
 		zone_info->nr_zones++;
@@ -188,6 +191,7 @@ int btrfs_check_zoned_mode(struct btrfs_fs_info *fs_info)
 	u64 zoned_devices = 0;
 	u64 nr_devices = 0;
 	u64 zone_size = 0;
+	u64 max_zone_append_size = 0;
 	const bool incompat_zoned = btrfs_is_zoned(fs_info);
 	int ret = 0;
 
@@ -201,10 +205,13 @@ int btrfs_check_zoned_mode(struct btrfs_fs_info *fs_info)
 		model = bdev_zoned_model(device->bdev);
 		if (model == BLK_ZONED_HM ||
 		    (model == BLK_ZONED_HA && incompat_zoned)) {
+			struct btrfs_zoned_device_info *zone_info =
+				device->zone_info;
+
 			zoned_devices++;
 			if (!zone_size) {
-				zone_size = device->zone_info->zone_size;
-			} else if (device->zone_info->zone_size != zone_size) {
+				zone_size = zone_info->zone_size;
+			} else if (zone_info->zone_size != zone_size) {
 				btrfs_err(fs_info,
 					  "zoned: unequal block device zone sizes: have %llu found %llu",
 					  device->zone_info->zone_size,
@@ -212,6 +219,11 @@ int btrfs_check_zoned_mode(struct btrfs_fs_info *fs_info)
 				ret = -EINVAL;
 				goto out;
 			}
+			if (!max_zone_append_size ||
+			    (zone_info->max_zone_append_size &&
+			     zone_info->max_zone_append_size < max_zone_append_size))
+				max_zone_append_size =
+					zone_info->max_zone_append_size;
 		}
 		nr_devices++;
 	}
@@ -255,6 +267,7 @@ int btrfs_check_zoned_mode(struct btrfs_fs_info *fs_info)
 	}
 
 	fs_info->zone_size = zone_size;
+	fs_info->max_zone_append_size = max_zone_append_size;
 
 	btrfs_info(fs_info, "zoned mode enabled with zone size %llu",
 		   fs_info->zone_size);
fs/btrfs/zoned.h +1 -0
index bcb1cb99a4f3..52aa6af5d8dc 100644
@@ -13,6 +13,7 @@ struct btrfs_zoned_device_info {
 	 */
 	u64 zone_size;
 	u8  zone_size_shift;
+	u64 max_zone_append_size;
 	u32 nr_zones;
 	unsigned long *seq_zones;
 	unsigned long *empty_zones;
Signature · 1 line
2.27.0
#51 lore raw

Re: [PATCH v10 06/41] btrfs: introduce max_zone_append_size

Quoted 5 lines · Naohiro Aota

The zone append write command has a maximum IO size restriction it accepts. This is because a zone append write command cannot be split, as we ask the device to place the data into a specific target zone and the device responds with the actual written location of the data.

Quoted 7 lines

Introduce max_zone_append_size to zone_info and fs_info to track the value, so we can limit all I/O to a zoned block device that we want to write using the zone append command to the device's limits.

---

Looks good except for - what happens when we replace or add a new zone device with a different queue_max_zone_append_sectors(queue) value. ?

Nit: IMHO some parts of patch-4, 5 and 6 could have been in one patch. Now it's fine as they are already at v10 and have rb.

Thanks, Anand

Quoted 97 lines
  fs/btrfs/ctree.h |  3 +++
  fs/btrfs/zoned.c | 17 +++++++++++++++--
  fs/btrfs/zoned.h |  1 +
  3 files changed, 19 insertions(+), 2 deletions(-)
fs/btrfs/ctree.h +3 -0
index 453f41ca024e..c70d3fcc62c2 100644
@@ -954,6 +954,9 @@ struct btrfs_fs_info {
  		u64 zoned;
  	};
 
+	/* Max size to emit ZONE_APPEND write command */
+	u64 max_zone_append_size;
+
  #ifdef CONFIG_BTRFS_FS_REF_VERIFY
  	spinlock_t ref_verify_lock;
  	struct rb_root block_tree;
fs/btrfs/zoned.c +15 -2
index 1223d5b0e411..2897432eb43c 100644
@@ -48,6 +48,7 @@ int btrfs_get_dev_zone_info(struct btrfs_device *device)
  {
  	struct btrfs_zoned_device_info *zone_info = NULL;
  	struct block_device *bdev = device->bdev;
+	struct request_queue *queue = bdev_get_queue(bdev);
  	sector_t nr_sectors = bdev->bd_part->nr_sects;
  	sector_t sector = 0;
  	struct blk_zone *zones = NULL;
@@ -69,6 +70,8 @@ int btrfs_get_dev_zone_info(struct btrfs_device *device)
  	ASSERT(is_power_of_2(zone_sectors));
  	zone_info->zone_size = (u64)zone_sectors << SECTOR_SHIFT;
  	zone_info->zone_size_shift = ilog2(zone_info->zone_size);
+	zone_info->max_zone_append_size =
+		(u64)queue_max_zone_append_sectors(queue) << SECTOR_SHIFT;
  	zone_info->nr_zones = nr_sectors >> ilog2(bdev_zone_sectors(bdev));
  	if (!IS_ALIGNED(nr_sectors, zone_sectors))
  		zone_info->nr_zones++;
@@ -188,6 +191,7 @@ int btrfs_check_zoned_mode(struct btrfs_fs_info *fs_info)
  	u64 zoned_devices = 0;
  	u64 nr_devices = 0;
  	u64 zone_size = 0;
+	u64 max_zone_append_size = 0;
  	const bool incompat_zoned = btrfs_is_zoned(fs_info);
  	int ret = 0;
 
@@ -201,10 +205,13 @@ int btrfs_check_zoned_mode(struct btrfs_fs_info *fs_info)
  		model = bdev_zoned_model(device->bdev);
  		if (model == BLK_ZONED_HM ||
  		    (model == BLK_ZONED_HA && incompat_zoned)) {
+			struct btrfs_zoned_device_info *zone_info =
+				device->zone_info;
+
  			zoned_devices++;
  			if (!zone_size) {
-				zone_size = device->zone_info->zone_size;
-			} else if (device->zone_info->zone_size != zone_size) {
+				zone_size = zone_info->zone_size;
+			} else if (zone_info->zone_size != zone_size) {
  				btrfs_err(fs_info,
  					  "zoned: unequal block device zone sizes: have %llu found %llu",
  					  device->zone_info->zone_size,
@@ -212,6 +219,11 @@ int btrfs_check_zoned_mode(struct btrfs_fs_info *fs_info)
  				ret = -EINVAL;
  				goto out;
  			}
+			if (!max_zone_append_size ||
+			    (zone_info->max_zone_append_size &&
+			     zone_info->max_zone_append_size < max_zone_append_size))
+				max_zone_append_size =
+					zone_info->max_zone_append_size;
  		}
  		nr_devices++;
  	}
@@ -255,6 +267,7 @@ int btrfs_check_zoned_mode(struct btrfs_fs_info *fs_info)
  	}
 
  	fs_info->zone_size = zone_size;
+	fs_info->max_zone_append_size = max_zone_append_size;
 
  	btrfs_info(fs_info, "zoned mode enabled with zone size %llu",
  		   fs_info->zone_size);
fs/btrfs/zoned.h +1 -0
index bcb1cb99a4f3..52aa6af5d8dc 100644
@@ -13,6 +13,7 @@ struct btrfs_zoned_device_info {
  	 */
  	u64 zone_size;
  	u8  zone_size_shift;
+	u64 max_zone_append_size;
  	u32 nr_zones;
  	unsigned long *seq_zones;
  	unsigned long *empty_zones;
#52 lore raw

Re: [PATCH v10 06/41] btrfs: introduce max_zone_append_size

Quoted 10 lines · Anand Jain
Quoted 9 lines · Naohiro Aota

The zone append write command has a maximum IO size restriction it accepts. This is because a zone append write command cannot be split, as we ask the device to place the data into a specific target zone and the device responds with the actual written location of the data.

Introduce max_zone_append_size to zone_info and fs_info to track the value, so we can limit all I/O to a zoned block device that we want to write using the zone append command to the device's limits.

Quoted 6 lines
Quoted 3 lines

---

Looks good except for - what happens when we replace or add a new zone device with a different queue_max_zone_append_sectors(queue) value. ?

The max zone seems to be a constraint for all devices so yeah it should be recalculated.

#53 lore raw

PATCH v10 7/41 btrfs: disallow space_cache in ZONED mode

In reply to Naohiro Aota

As updates to the space cache v1 are in-place, the space cache cannot be located over sequential zones and there is no guarantees that the device will have enough conventional zones to store this cache. Resolve this problem by disabling completely the space cache v1. This does not introduces any problems with sequential block groups: all the free space is located after the allocation pointer and no free space before the pointer. There is no need to have such cache.

Note: we can technically use free-space-tree (space cache v2) on ZONED mode. But, since ZONED mode now always allocate extents in a block group sequentially regardless of underlying device zone type, it's no use to enable and maintain the tree.

For the same reason, NODATACOW is also disabled.

In summary, ZONED will disable:

| Disabled features | Reason                                              |
|-------------------+-----------------------------------------------------|
| RAID/Dup          | Cannot handle two zone append writes to different   |
|                   | zones                                               |
|-------------------+-----------------------------------------------------|
| space_cache (v1)  | In-place updating                                   |
| NODATACOW         | In-place updating                                   |
|-------------------+-----------------------------------------------------|
| fallocate         | Reserved extent will be a write hole                |
|-------------------+-----------------------------------------------------|
| MIXED_BG          | Allocated metadata region will be write holes for   |
|                   | data writes                                         |
---
 fs/btrfs/super.c | 13 +++++++++++--
 fs/btrfs/zoned.c | 18 ++++++++++++++++++
 fs/btrfs/zoned.h |  6 ++++++
 3 files changed, 35 insertions(+), 2 deletions(-)
fs/btrfs/super.c +11 -2
index 3312fe08168f..1adbbeebc649 100644
@@ -525,8 +525,15 @@ int btrfs_parse_options(struct btrfs_fs_info *info, char *options,
 	cache_gen = btrfs_super_cache_generation(info->super_copy);
 	if (btrfs_fs_compat_ro(info, FREE_SPACE_TREE))
 		btrfs_set_opt(info->mount_opt, FREE_SPACE_TREE);
-	else if (cache_gen)
-		btrfs_set_opt(info->mount_opt, SPACE_CACHE);
+	else if (cache_gen) {
+		if (btrfs_is_zoned(info)) {
+			btrfs_info(info,
+			"zoned: clearing existing space cache");
+			btrfs_set_super_cache_generation(info->super_copy, 0);
+		} else {
+			btrfs_set_opt(info->mount_opt, SPACE_CACHE);
+		}
+	}
 
 	/*
 	 * Even the options are empty, we still need to do extra check
@@ -985,6 +992,8 @@ int btrfs_parse_options(struct btrfs_fs_info *info, char *options,
 		ret = -EINVAL;
 
 	}
+	if (!ret)
+		ret = btrfs_check_mountopts_zoned(info);
 	if (!ret && btrfs_test_opt(info, SPACE_CACHE))
 		btrfs_info(info, "disk space caching is enabled");
 	if (!ret && btrfs_test_opt(info, FREE_SPACE_TREE))
fs/btrfs/zoned.c +18 -0
index 2897432eb43c..d6b8165e2c91 100644
@@ -274,3 +274,21 @@ int btrfs_check_zoned_mode(struct btrfs_fs_info *fs_info)
 out:
 	return ret;
 }
+
+int btrfs_check_mountopts_zoned(struct btrfs_fs_info *info)
+{
+	if (!btrfs_is_zoned(info))
+		return 0;
+
+	/*
+	 * Space cache writing is not COWed. Disable that to avoid write
+	 * errors in sequential zones.
+	 */
+	if (btrfs_test_opt(info, SPACE_CACHE)) {
+		btrfs_err(info,
+			  "zoned: space cache v1 is not supported");
+		return -EINVAL;
+	}
+
+	return 0;
+}
fs/btrfs/zoned.h +6 -0
index 52aa6af5d8dc..81c00a3ed202 100644
@@ -25,6 +25,7 @@ int btrfs_get_dev_zone(struct btrfs_device *device, u64 pos,
 int btrfs_get_dev_zone_info(struct btrfs_device *device);
 void btrfs_destroy_dev_zone_info(struct btrfs_device *device);
 int btrfs_check_zoned_mode(struct btrfs_fs_info *fs_info);
+int btrfs_check_mountopts_zoned(struct btrfs_fs_info *info);
 #else /* CONFIG_BLK_DEV_ZONED */
 static inline int btrfs_get_dev_zone(struct btrfs_device *device, u64 pos,
 				     struct blk_zone *zone)
@@ -48,6 +49,11 @@ static inline int btrfs_check_zoned_mode(struct btrfs_fs_info *fs_info)
 	return -EOPNOTSUPP;
 }
 
+static inline int btrfs_check_mountopts_zoned(struct btrfs_fs_info *info)
+{
+	return 0;
+}
+
 #endif
 
 static inline bool btrfs_dev_is_sequential(struct btrfs_device *device, u64 pos)
Signature · 1 line
2.27.0
#54 lore raw

Re: [PATCH v10 07/41] btrfs: disallow space_cache in ZONED mode

Quoted 52 lines
fs/btrfs/super.c +2 -0
@@ -985,6 +992,8 @@ int btrfs_parse_options(struct btrfs_fs_info *info, char *options,
  		ret = -EINVAL;
 
  	}
+	if (!ret)
+		ret = btrfs_check_mountopts_zoned(info);
  	if (!ret && btrfs_test_opt(info, SPACE_CACHE))
  		btrfs_info(info, "disk space caching is enabled");
  	if (!ret && btrfs_test_opt(info, FREE_SPACE_TREE))
fs/btrfs/zoned.c +18 -0
index 2897432eb43c..d6b8165e2c91 100644
@@ -274,3 +274,21 @@ int btrfs_check_zoned_mode(struct btrfs_fs_info *fs_info)
  out:
  	return ret;
  }
+
+int btrfs_check_mountopts_zoned(struct btrfs_fs_info *info)
+{
+	if (!btrfs_is_zoned(info))
+		return 0;
+
+	/*
+	 * Space cache writing is not COWed. Disable that to avoid write
+	 * errors in sequential zones.
+	 */
+	if (btrfs_test_opt(info, SPACE_CACHE)) {
+		btrfs_err(info,
+			  "zoned: space cache v1 is not supported");
+		return -EINVAL;
+	}
+
+	return 0;
+}
fs/btrfs/zoned.h +1 -0
index 52aa6af5d8dc..81c00a3ed202 100644
@@ -25,6 +25,7 @@ int btrfs_get_dev_zone(struct btrfs_device *device, u64 pos,
  int btrfs_get_dev_zone_info(struct btrfs_device *device);
  void btrfs_destroy_dev_zone_info(struct btrfs_device *device);
  int btrfs_check_zoned_mode(struct btrfs_fs_info *fs_info);
+int btrfs_check_mountopts_zoned(struct btrfs_fs_info *info);
  #else /* CONFIG_BLK_DEV_ZONED */
  static inline int btrfs_get_dev_zone(struct btrfs_device *device, u64 pos,
  				     struct blk_zone *zone)
@@ -48,6 +49,11 @@ static inline int btrfs_check_zoned_mode(struct btrfs_fs_info *fs_info)
  	return -EOPNOTSUPP;

fs/btrfs/zoned.h static inline int btrfs_check_zoned_mode(struct btrfs_fs_info *fs_info) · line 59

Quoted 10 lines
fs/btrfs/zoned.h +5 -0
  }
 
+static inline int btrfs_check_mountopts_zoned(struct btrfs_fs_info *info)
+{
+	return 0;
+}
+
  #endif
 
  static inline bool btrfs_dev_is_sequential(struct btrfs_device *device, u64 pos)

The whole of the above code can be replaced by..

-------------------
@@ -810,8 +810,15 @@ int btrfs_parse_options(struct btrfs_fs_info *info, char *options,

                        break;
                case Opt_space_cache:
                case Opt_space_cache_version:
                        if (token == Opt_space_cache ||
                            strcmp(args[0].from, "v1") == 0) {
+                               if (btrfs_is_zoned(info)) {
+                                       btrfs_err(info,
+                                       "zoned: space cache v1 is not supported");
+                                       ret = -EINVAL;
+                                       goto out;
+                               }
                                btrfs_clear_opt(info->mount_opt,
                                                FREE_SPACE_TREE);
                                btrfs_set_and_info(info, SPACE_CACHE,

-------------------

Thanks.

#55 lore raw

Re: [PATCH v10 07/41] btrfs: disallow space_cache in ZONED mode

Quoted 80 lines · Anand Jain
Quoted 54 lines
fs/btrfs/super.c +2 -0
@@ -985,6 +992,8 @@ int btrfs_parse_options(struct btrfs_fs_info 
 *info, char *options,
           ret = -EINVAL;
       }
+    if (!ret)
+        ret = btrfs_check_mountopts_zoned(info);
       if (!ret && btrfs_test_opt(info, SPACE_CACHE))
           btrfs_info(info, "disk space caching is enabled");
       if (!ret && btrfs_test_opt(info, FREE_SPACE_TREE))
fs/btrfs/zoned.c +0 -0
index 2897432eb43c..d6b8165e2c91 100644
@@ -274,3 +274,21 @@ int btrfs_check_zoned_mode(struct btrfs_fs_info 
 *fs_info)

  out:
      return ret;
  }

fs/btrfs/zoned.c +18 -0
+
+int btrfs_check_mountopts_zoned(struct btrfs_fs_info *info)
+{
+    if (!btrfs_is_zoned(info))
+        return 0;
+
+    /*
+     * Space cache writing is not COWed. Disable that to avoid write
+     * errors in sequential zones.
+     */
+    if (btrfs_test_opt(info, SPACE_CACHE)) {
+        btrfs_err(info,
+              "zoned: space cache v1 is not supported");
+        return -EINVAL;
+    }
+
+    return 0;
+}
fs/btrfs/zoned.h +1 -0
index 52aa6af5d8dc..81c00a3ed202 100644
@@ -25,6 +25,7 @@ int btrfs_get_dev_zone(struct btrfs_device *device, 
 u64 pos,
   int btrfs_get_dev_zone_info(struct btrfs_device *device);
   void btrfs_destroy_dev_zone_info(struct btrfs_device *device);
   int btrfs_check_zoned_mode(struct btrfs_fs_info *fs_info);
+int btrfs_check_mountopts_zoned(struct btrfs_fs_info *info);
   #else /* CONFIG_BLK_DEV_ZONED */
   static inline int btrfs_get_dev_zone(struct btrfs_device *device, u64 pos,

                       struct blk_zone *zone)

fs/btrfs/zoned.h +0 -0
@@ -48,6 +49,11 @@ static inline int btrfs_check_zoned_mode(struct 
 btrfs_fs_info *fs_info)

fs/btrfs/zoned.h static inline int btrfs_check_zoned_mode(struct · line 59

Quoted 10 lines
fs/btrfs/zoned.h +5 -0
       return -EOPNOTSUPP;
   }
+static inline int btrfs_check_mountopts_zoned(struct btrfs_fs_info 
 *info)
+{
+    return 0;
+}
+
   #endif
   static inline bool btrfs_dev_is_sequential(struct btrfs_device *device, u64 pos)

The whole of the above code can be replaced by..

-------------------

(quoted diff) +1 -0
@@ -810,8 +810,15 @@ int btrfs_parse_options(struct btrfs_fs_info *info, 
 char *options,
                         break;
                 case Opt_space_cache:
                 case Opt_space_cache_version:
                         if (token == Opt_space_cache ||
                             strcmp(args[0].from, "v1") == 0) {
+                               if (btrfs_is_zoned(info)) {

Review int btrfs_parse_options(struct btrfs_fs_info *info,

Quoted 10 lines
(quoted diff) +5 -1
+                                       btrfs_err(info,
+                                       "zoned: space cache v1 is not 
 supported");
+                                       ret = -EINVAL;
+                                       goto out;
+                               }
                                 btrfs_clear_opt(info->mount_opt,
                                                 FREE_SPACE_TREE);
                                 btrfs_set_and_info(info, SPACE_CACHE,
-------------------

Later patches add more flag-checks to btrfs_check_mountopts_zoned(), I have no preference for either this way or adding those checks individually in btrfs_parse_options(). So you may ignore this.

Thanks.

Quoted 1 line

Thanks.

#56 lore raw

PATCH v10 8/41 btrfs: disallow NODATACOW in ZONED mode

In reply to Naohiro Aota

NODATACOW implies overwriting the file data on a device, which is impossible in sequential required zones. Disable NODATACOW globally with mount option and per-file NODATACOW attribute by masking FS_NOCOW_FL.

---
 fs/btrfs/ioctl.c | 13 +++++++++++++
 fs/btrfs/zoned.c |  5 +++++
 2 files changed, 18 insertions(+)
fs/btrfs/ioctl.c +13 -0
index ab408a23ba32..d13b522e7bb2 100644
@@ -193,6 +193,15 @@ static int check_fsflags(unsigned int old_flags, unsigned int flags)
 	return 0;
 }
 
+static int check_fsflags_compatible(struct btrfs_fs_info *fs_info,
+				    unsigned int flags)
+{
+	if (btrfs_is_zoned(fs_info) && (flags & FS_NOCOW_FL))
+		return -EPERM;
+
+	return 0;
+}
+
 static int btrfs_ioctl_setflags(struct file *file, void __user *arg)
 {
 	struct inode *inode = file_inode(file);
@@ -230,6 +239,10 @@ static int btrfs_ioctl_setflags(struct file *file, void __user *arg)
 	if (ret)
 		goto out_unlock;
 
+	ret = check_fsflags_compatible(fs_info, fsflags);
+	if (ret)
+		goto out_unlock;
+
 	binode_flags = binode->flags;
 	if (fsflags & FS_SYNC_FL)
 		binode_flags |= BTRFS_INODE_SYNC;
fs/btrfs/zoned.c +5 -0
index d6b8165e2c91..bd153932606e 100644
@@ -290,5 +290,10 @@ int btrfs_check_mountopts_zoned(struct btrfs_fs_info *info)
 		return -EINVAL;
 	}
 
+	if (btrfs_test_opt(info, NODATACOW)) {
+		btrfs_err(info, "zoned: NODATACOW not supported");
+		return -EINVAL;
+	}
+
 	return 0;
 }
Signature · 1 line
2.27.0
Reviewed-by on v10 8/41
#57 lore raw

Re: [PATCH v10 08/41] btrfs: disallow NODATACOW in ZONED mode

Quoted 7 lines · Naohiro Aota

NODATACOW implies overwriting the file data on a device, which is impossible in sequential required zones. Disable NODATACOW globally with mount option and per-file NODATACOW attribute by masking FS_NOCOW_FL.

Looks good.

A nit below.

Quoted 10 lines
---
  fs/btrfs/ioctl.c | 13 +++++++++++++
  fs/btrfs/zoned.c |  5 +++++
  2 files changed, 18 insertions(+)
fs/btrfs/ioctl.c +0 -0
index ab408a23ba32..d13b522e7bb2 100644

fs/btrfs/ioctl.c static int check_fsflags(unsigned int old_flags, unsigned int flags) · line 202

Quoted 10 lines
fs/btrfs/ioctl.c +5 -0
@@ -193,6 +193,15 @@ static int check_fsflags(unsigned int old_flags, unsigned int flags)
  	return 0;
  }
 
+static int check_fsflags_compatible(struct btrfs_fs_info *fs_info,
+				    unsigned int flags)
+{
+	if (btrfs_is_zoned(fs_info) && (flags & FS_NOCOW_FL))
 
 
+		return -EPERM;

nit:
Should it be -EINVAL instead? I am not sure. May be David can fix
while integrating.

Thanks.

Quoted 34 lines
fs/btrfs/ioctl.c +4 -0
+
+	return 0;
+}
+
  static int btrfs_ioctl_setflags(struct file *file, void __user *arg)
  {
  	struct inode *inode = file_inode(file);
fs/btrfs/ioctl.c +4 -0
@@ -230,6 +239,10 @@ static int btrfs_ioctl_setflags(struct file *file, void __user *arg)
  	if (ret)
  		goto out_unlock;
 
+	ret = check_fsflags_compatible(fs_info, fsflags);
+	if (ret)
+		goto out_unlock;
+
  	binode_flags = binode->flags;
  	if (fsflags & FS_SYNC_FL)
  		binode_flags |= BTRFS_INODE_SYNC;
fs/btrfs/zoned.c +5 -0
index d6b8165e2c91..bd153932606e 100644
@@ -290,5 +290,10 @@ int btrfs_check_mountopts_zoned(struct btrfs_fs_info *info)
  		return -EINVAL;
  	}
 
+	if (btrfs_test_opt(info, NODATACOW)) {
+		btrfs_err(info, "zoned: NODATACOW not supported");
+		return -EINVAL;
+	}
+
  	return 0;
  }
#58 lore raw

Re: [PATCH v10 08/41] btrfs: disallow NODATACOW in ZONED mode

Quoted 11 lines · Anand Jain

fs/btrfs/ioctl.c static int check_fsflags(unsigned int old_flags, unsigned int flags) · line 200

Quoted 6 lines · Naohiro Aota
fs/btrfs/ioctl.c +4 -0
+				    unsigned int flags)
+{
+	if (btrfs_is_zoned(fs_info) && (flags & FS_NOCOW_FL))
 
 
+		return -EPERM;

nit:
Should it be -EINVAL instead? I am not sure. May be David can fix
while integrating.

IIRC we've discussed that in some previous iteration. EPERM should be interpreted as that it's not permitted right now, but otherwise it is a valid operation/flag. The constraint is the zoned device.

As an example: deleting default subvolume is not permitted (EPERM), but otherwise subvolume deletion is a valid operation.

So, EINVAL is for invalid combination of parameters or a request for something that does not make sense at all.

#59 lore raw

Re: [PATCH v10 08/41] btrfs: disallow NODATACOW in ZONED mode

Quoted 17 lines · David Sterba
Quoted 11 lines · Anand Jain

fs/btrfs/ioctl.c static int check_fsflags(unsigned int old_flags, unsigned int flags) · line 200

Quoted 6 lines · Naohiro Aota
fs/btrfs/ioctl.c +4 -0
+				    unsigned int flags)
+{
+	if (btrfs_is_zoned(fs_info) && (flags & FS_NOCOW_FL))
 
 
+		return -EPERM;

nit:
Should it be -EINVAL instead? I am not sure. May be David can fix
while integrating.

IIRC we've discussed that in some previous iteration. EPERM should be interpreted as that it's not permitted right now, but otherwise it is a valid operation/flag. The constraint is the zoned device.

Quoted 6 lines

As an example: deleting default subvolume is not permitted (EPERM), but otherwise subvolume deletion is a valid operation.

So, EINVAL is for invalid combination of parameters or a request for something that does not make sense at all.

Ok. Thanks.

#60 lore raw

PATCH v10 9/41 btrfs: disable fallocate in ZONED mode

In reply to Naohiro Aota

fallocate() is implemented by reserving actual extent instead of reservations. This can result in exposing the sequential write constraint of host-managed zoned block devices to the application, which would break the POSIX semantic for the fallocated file. To avoid this, report fallocate() as not supported when in ZONED mode for now.

In the future, we may be able to implement "in-memory" fallocate() in ZONED mode by utilizing space_info->bytes_may_use or so.

---
 fs/btrfs/file.c | 4 ++++
 1 file changed, 4 insertions(+)
fs/btrfs/file.c +4 -0
index 0ff659455b1e..68938a43081e 100644
@@ -3341,6 +3341,10 @@ static long btrfs_fallocate(struct file *file, int mode,
 	alloc_end = round_up(offset + len, blocksize);
 	cur_offset = alloc_start;
 
+	/* Do not allow fallocate in ZONED mode */
+	if (btrfs_is_zoned(btrfs_sb(inode->i_sb)))
+		return -EOPNOTSUPP;
+
 	/* Make sure we aren't being give some crap mode */
 	if (mode & ~(FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE |
 		     FALLOC_FL_ZERO_RANGE))
Signature · 1 line
2.27.0
Reviewed-by on v10 9/41
#61 lore raw

Re: [PATCH v10 09/41] btrfs: disable fallocate in ZONED mode

Quoted 6 lines · Naohiro Aota

fallocate() is implemented by reserving actual extent instead of reservations. This can result in exposing the sequential write constraint of host-managed zoned block devices to the application, which would break the POSIX semantic for the fallocated file. To avoid this, report fallocate() as not supported when in ZONED mode for now.

Quoted 6 lines

In the future, we may be able to implement "in-memory" fallocate() in ZONED mode by utilizing space_info->bytes_may_use or so.

Looks good.

Thanks.

#62 lore raw

PATCH v10 10/41 btrfs: disallow mixed-bg in ZONED mode

In reply to Naohiro Aota

Placing both data and metadata in a block group is impossible in ZONED mode. For data, we can allocate a space for it and write it immediately after the allocation. For metadata, however, we cannot do so, because the logical addresses are recorded in other metadata buffers to build up the trees. As a result, a data buffer can be placed after a metadata buffer, which is not written yet. Writing out the data buffer will break the sequential write rule.

This commit check and disallow MIXED_BG with ZONED mode.

---
 fs/btrfs/zoned.c | 7 +++++++
 1 file changed, 7 insertions(+)
fs/btrfs/zoned.c +7 -0
index bd153932606e..f87d35cb9235 100644
@@ -266,6 +266,13 @@ int btrfs_check_zoned_mode(struct btrfs_fs_info *fs_info)
 		goto out;
 	}
 
+	if (btrfs_fs_incompat(fs_info, MIXED_GROUPS)) {
+		btrfs_err(fs_info,
+			  "zoned: mixed block groups not supported");
+		ret = -EINVAL;
+		goto out;
+	}
+
 	fs_info->zone_size = zone_size;
 	fs_info->max_zone_append_size = max_zone_append_size;
 
Signature · 1 line
2.27.0
Reviewed-by on v10 10/41
#63 lore raw

Re: [PATCH v10 10/41] btrfs: disallow mixed-bg in ZONED mode

Quoted 8 lines · Naohiro Aota

Placing both data and metadata in a block group is impossible in ZONED mode. For data, we can allocate a space for it and write it immediately after the allocation. For metadata, however, we cannot do so, because the logical addresses are recorded in other metadata buffers to build up the trees. As a result, a data buffer can be placed after a metadata buffer, which is not written yet. Writing out the data buffer will break the sequential write rule.

Quoted 5 lines

This commit check and disallow MIXED_BG with ZONED mode.

---

Thanks.

#64 lore raw

PATCH v10 11/41 btrfs: implement log-structured superblock for ZONED mode

In reply to Naohiro Aota

Superblock (and its copies) is the only data structure in btrfs which has a fixed location on a device. Since we cannot overwrite in a sequential write required zone, we cannot place superblock in the zone. One easy solution is limiting superblock and copies to be placed only in conventional zones. However, this method has two downsides: one is reduced number of superblock copies. The location of the second copy of superblock is 256GB, which is in a sequential write required zone on typical devices in the market today. So, the number of superblock and copies is limited to be two. Second downside is that we cannot support devices which have no conventional zones at all.

To solve these two problems, we employ superblock log writing. It uses two zones as a circular buffer to write updated superblocks. Once the first zone is filled up, start writing into the second buffer. Then, when the both zones are filled up and before start writing to the first zone again, it reset the first zone.

We can determine the position of the latest superblock by reading write pointer information from a device. One corner case is when the both zones are full. For this situation, we read out the last superblock of each zone, and compare them to determine which zone is older.

The following zones are reserved as the circular buffer on ZONED btrfs.

- The primary superblock: zones 0 and 1
- The first copy: zones 16 and 17
- The second copy: zones 1024 or zone at 256GB which is minimum, and next to it

If these reserved zones are conventional, superblock is written fixed at the start of the zone without logging.

---
 fs/btrfs/block-group.c |   9 ++
 fs/btrfs/disk-io.c     |  41 ++++-
 fs/btrfs/scrub.c       |   3 +
 fs/btrfs/volumes.c     |  21 ++-
 fs/btrfs/zoned.c       | 329 +++++++++++++++++++++++++++++++++++++++++
 fs/btrfs/zoned.h       |  44 ++++++
 6 files changed, 435 insertions(+), 12 deletions(-)
fs/btrfs/block-group.c +9 -0
index c0f1d6818df7..6b4831824f51 100644
@@ -1723,6 +1723,7 @@ int btrfs_rmap_block(struct btrfs_fs_info *fs_info, u64 chunk_start,
 static int exclude_super_stripes(struct btrfs_block_group *cache)
 {
 	struct btrfs_fs_info *fs_info = cache->fs_info;
+	const bool zoned = btrfs_is_zoned(fs_info);
 	u64 bytenr;
 	u64 *logical;
 	int stripe_len;
@@ -1744,6 +1745,14 @@ static int exclude_super_stripes(struct btrfs_block_group *cache)
 		if (ret)
 			return ret;
 
+		/* Shouldn't have super stripes in sequential zones */
+		if (zoned && nr) {
+			btrfs_err(fs_info,
+				  "zoned: block group %llu must not contain super block",
+				  cache->start);
+			return -EUCLEAN;
+		}
+
 		while (nr--) {
 			u64 len = min_t(u64, stripe_len,
 				cache->start + cache->length - logical[nr]);
fs/btrfs/disk-io.c +34 -7

89 lines not on this page · raw message

fs/btrfs/scrub.c +3 -0
index cf63f1e27a27..aa1b36cf5c88 100644
@@ -20,6 +20,7 @@
 #include "rcu-string.h"
 #include "raid56.h"
 #include "block-group.h"
+#include "zoned.h"
 
 /*
  * This is only the first step towards a full-features scrub. It reads all
@@ -3704,6 +3705,8 @@ static noinline_for_stack int scrub_supers(struct scrub_ctx *sctx,
 		if (bytenr + BTRFS_SUPER_INFO_SIZE >
 		    scrub_dev->commit_total_bytes)
 			break;
+		if (!btrfs_check_super_location(scrub_dev, bytenr))
+			continue;
 
 		ret = scrub_pages(sctx, bytenr, BTRFS_SUPER_INFO_SIZE, bytenr,
 				  scrub_dev, BTRFS_EXTENT_FLAG_SUPER, gen, i,
fs/btrfs/volumes.c +16 -5
index 10827892c086..db884b96a5ea 100644
@@ -1282,7 +1282,8 @@ void btrfs_release_disk_super(struct btrfs_super_block *super)
 }
 
 static struct btrfs_super_block *btrfs_read_disk_super(struct block_device *bdev,
-						       u64 bytenr)
+						       u64 bytenr,
+						       u64 bytenr_orig)
 {
 	struct btrfs_super_block *disk_super;
 	struct page *page;
@@ -1313,7 +1314,7 @@ static struct btrfs_super_block *btrfs_read_disk_super(struct block_device *bdev
 	/* align our pointer to the offset of the super block */
 	disk_super = p + offset_in_page(bytenr);
 
-	if (btrfs_super_bytenr(disk_super) != bytenr ||
+	if (btrfs_super_bytenr(disk_super) != bytenr_orig ||
 	    btrfs_super_magic(disk_super) != BTRFS_MAGIC) {
 		btrfs_release_disk_super(p);
 		return ERR_PTR(-EINVAL);
@@ -1348,7 +1349,8 @@ struct btrfs_device *btrfs_scan_one_device(const char *path, fmode_t flags,
 	bool new_device_added = false;
 	struct btrfs_device *device = NULL;
 	struct block_device *bdev;
-	u64 bytenr;
+	u64 bytenr, bytenr_orig;
+	int ret;
 
 	lockdep_assert_held(&uuid_mutex);
 
@@ -1358,14 +1360,18 @@ struct btrfs_device *btrfs_scan_one_device(const char *path, fmode_t flags,
 	 * So, we need to add a special mount option to scan for
 	 * later supers, using BTRFS_SUPER_MIRROR_MAX instead
 	 */
-	bytenr = btrfs_sb_offset(0);
 	flags |= FMODE_EXCL;
 
 	bdev = blkdev_get_by_path(path, flags, holder);
 	if (IS_ERR(bdev))
 		return ERR_CAST(bdev);
 
-	disk_super = btrfs_read_disk_super(bdev, bytenr);
+	bytenr_orig = btrfs_sb_offset(0);
+	ret = btrfs_sb_log_location_bdev(bdev, 0, READ, &bytenr);
+	if (ret)
+		return ERR_PTR(ret);
+
+	disk_super = btrfs_read_disk_super(bdev, bytenr, bytenr_orig);
 	if (IS_ERR(disk_super)) {
 		device = ERR_CAST(disk_super);
 		goto error_bdev_put;
@@ -2029,6 +2035,11 @@ void btrfs_scratch_superblocks(struct btrfs_fs_info *fs_info,
 		if (IS_ERR(disk_super))
 			continue;
 
+		if (bdev_is_zoned(bdev)) {
+			btrfs_reset_sb_log_zones(bdev, copy_num);
+			continue;
+		}
+
 		memset(&disk_super->magic, 0, sizeof(disk_super->magic));
 
 		page = virt_to_page(disk_super);
fs/btrfs/zoned.c +329 -0

350 lines not on this page · raw message

fs/btrfs/zoned.h +44 -0

72 lines not on this page · raw message

Signature · 1 line
2.27.0
#65 lore raw

Re: [PATCH v10 11/41] btrfs: implement log-structured superblock for ZONED mode

I love your patch! Perhaps something to improve:

Hi Naohiro,

I love your patch! Perhaps something to improve:

[auto build test WARNING on xfs-linux/for-next]
[also build test WARNING on v5.10-rc3]
[cannot apply to kdave/for-next block/for-next next-20201110] [If your patch is applied to the wrong git tree, kindly drop us a note. And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]

url:    https://github.com/0day-ci/linux/commits/Naohiro-Aota/btrfs-zoned-block-device-support/20201110-193227
base:   https://git.kernel.org/pub/scm/fs/xfs/xfs-linux.git for-next
config: arc-allyesconfig (attached as .config)
compiler: arceb-elf-gcc (GCC) 9.3.0

reproduce (this is a W=1 build):
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross chmod +x ~/bin/make.cross
# https://github.com/0day-ci/linux/commit/9c0a9c59b898ad314f5610c22069f439465b2fec
git remote add linux-review https://github.com/0day-ci/linux
git fetch --no-tags linux-review Naohiro-Aota/btrfs-zoned-block-device-support/20201110-193227 git checkout 9c0a9c59b898ad314f5610c22069f439465b2fec
# save the attached .config to linux build tree
COMPILER_INSTALL_PATH=$HOME/0day COMPILER=gcc-9.3.0 make.cross ARCH=arc

If you fix the issue, kindly add following tag as appropriate

All warnings (new ones prefixed by >>):

fs/btrfs/zoned.c: In function 'btrfs_sb_log_location_bdev':

>> fs/btrfs/zoned.c:511:6: warning: variable 'zone_size' set but not used [-Wunused-but-set-variable]
     511 |  u64 zone_size;
         |      ^~~~~~~~~

vim +/zone_size +511 fs/btrfs/zoned.c

503
504 int btrfs_sb_log_location_bdev(struct block_device *bdev, int mirror, int rw,
505 u64 *bytenr_ret)
506 {
507 struct blk_zone zones[BTRFS_NR_SB_LOG_ZONES];
508 unsigned int zone_sectors;
509 u32 sb_zone;
510 int ret;

 > 511		u64 zone_size;

512 u8 zone_sectors_shift;
513 sector_t nr_sectors = bdev->bd_part->nr_sects;
514 u32 nr_zones;
515
516 if (!bdev_is_zoned(bdev)) {
517 *bytenr_ret = btrfs_sb_offset(mirror);
518 return 0;
519 }
520
521 ASSERT(rw == READ || rw == WRITE);
522
523 zone_sectors = bdev_zone_sectors(bdev);
524 if (!is_power_of_2(zone_sectors))
525 return -EINVAL;
526 zone_size = zone_sectors << SECTOR_SHIFT;
527 zone_sectors_shift = ilog2(zone_sectors);
528 nr_zones = nr_sectors >> zone_sectors_shift;
529
530 sb_zone = sb_zone_number(zone_sectors_shift + SECTOR_SHIFT, mirror);
531 if (sb_zone + 1 >= nr_zones)
532 return -ENOENT;
533
534 ret = blkdev_report_zones(bdev, sb_zone << zone_sectors_shift,
535 BTRFS_NR_SB_LOG_ZONES, copy_zone_info_cb,
536 zones);
537 if (ret < 0)
538 return ret;
539 if (ret != BTRFS_NR_SB_LOG_ZONES)
540 return -EIO;
541
542 return sb_log_location(bdev, zones, rw, bytenr_ret);
543 }
544

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/[email protected]

#66 lore raw

Re: [PATCH v10 11/41] btrfs: implement log-structured superblock for ZONED mode

In reply to Naohiro Aota
I love your patch! Perhaps something to improve:

Hi Naohiro,

I love your patch! Perhaps something to improve:

[auto build test WARNING on xfs-linux/for-next]
[also build test WARNING on v5.10-rc3]
[cannot apply to kdave/for-next block/for-next next-20201110] [If your patch is applied to the wrong git tree, kindly drop us a note. And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]

url:    https://github.com/0day-ci/linux/commits/Naohiro-Aota/btrfs-zoned-block-device-support/20201110-193227
base:   https://git.kernel.org/pub/scm/fs/xfs/xfs-linux.git for-next
config: x86_64-randconfig-s022-20201110 (attached as .config)
compiler: gcc-9 (Debian 9.3.0-15) 9.3.0
reproduce:
        # apt-get install sparse
        # sparse version: v0.6.3-76-gf680124b-dirty
        # https://github.com/0day-ci/linux/commit/9c0a9c59b898ad314f5610c22069f439465b2fec
        git remote add linux-review https://github.com/0day-ci/linux
        git fetch --no-tags linux-review Naohiro-Aota/btrfs-zoned-block-device-support/20201110-193227
        git checkout 9c0a9c59b898ad314f5610c22069f439465b2fec
        # save the attached .config to linux build tree
        make W=1 C=1 CF='-fdiagnostic-prefix -D__CHECK_ENDIAN__' ARCH=x86_64 

If you fix the issue, kindly add following tag as appropriate

"sparse warnings: (new ones prefixed by >>)"

>> fs/btrfs/zoned.c:85:29: sparse: sparse: restricted __le64 degrades to integer

fs/btrfs/zoned.c:85:52: sparse: sparse: restricted __le64 degrades to integer

vim +85 fs/btrfs/zoned.c

    25	
    26	static int sb_write_pointer(struct block_device *bdev, struct blk_zone *zones,
    27				    u64 *wp_ret)
    28	{
    29		bool empty[BTRFS_NR_SB_LOG_ZONES];
    30		bool full[BTRFS_NR_SB_LOG_ZONES];
    31		sector_t sector;
    32	
    33		ASSERT(zones[0].type != BLK_ZONE_TYPE_CONVENTIONAL &&
    34		       zones[1].type != BLK_ZONE_TYPE_CONVENTIONAL);
    35	
    36		empty[0] = (zones[0].cond == BLK_ZONE_COND_EMPTY);
    37		empty[1] = (zones[1].cond == BLK_ZONE_COND_EMPTY);
    38		full[0] = (zones[0].cond == BLK_ZONE_COND_FULL);
    39		full[1] = (zones[1].cond == BLK_ZONE_COND_FULL);
    40	
    41		/*
    42		 * Possible state of log buffer zones
    43		 *
    44		 *   E I F
    45		 * E * x 0
    46		 * I 0 x 0
    47		 * F 1 1 C
    48		 *
    49		 * Row: zones[0]
    50		 * Col: zones[1]
    51		 * State:
    52		 *   E: Empty, I: In-Use, F: Full
    53		 * Log position:
    54		 *   *: Special case, no superblock is written
    55		 *   0: Use write pointer of zones[0]
    56		 *   1: Use write pointer of zones[1]
    57		 *   C: Compare SBs from zones[0] and zones[1], use the newer one
    58		 *   x: Invalid state
    59		 */
    60	
    61		if (empty[0] && empty[1]) {
    62			/* Special case to distinguish no superblock to read */
    63			*wp_ret = zones[0].start << SECTOR_SHIFT;
    64			return -ENOENT;
    65		} else if (full[0] && full[1]) {
    66			/* Compare two super blocks */
    67			struct address_space *mapping = bdev->bd_inode->i_mapping;
    68			struct page *page[BTRFS_NR_SB_LOG_ZONES];
    69			struct btrfs_super_block *super[BTRFS_NR_SB_LOG_ZONES];
    70			int i;
    71	
    72			for (i = 0; i < BTRFS_NR_SB_LOG_ZONES; i++) {
    73				u64 bytenr = ((zones[i].start + zones[i].len) << SECTOR_SHIFT) -
    74					BTRFS_SUPER_INFO_SIZE;
    75	
    76				page[i] = read_cache_page_gfp(mapping, bytenr >> PAGE_SHIFT, GFP_NOFS);
    77				if (IS_ERR(page[i])) {
    78					if (i == 1)
    79						btrfs_release_disk_super(super[0]);
    80					return PTR_ERR(page[i]);
    81				}
    82				super[i] = page_address(page[i]);
    83			}
    84	
  > 85			if (super[0]->generation > super[1]->generation)
    86				sector = zones[1].start;
    87			else
    88				sector = zones[0].start;
    89	
    90			for (i = 0; i < BTRFS_NR_SB_LOG_ZONES; i++)
    91				btrfs_release_disk_super(super[i]);
    92		} else if (!full[0] && (empty[1] || full[1])) {
    93			sector = zones[0].wp;
    94		} else if (full[0]) {
    95			sector = zones[1].wp;
    96		} else {
    97			return -EUCLEAN;
    98		}
    99		*wp_ret = sector << SECTOR_SHIFT;

100 return 0;
101 }
102

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/[email protected]

#67 lore raw

Re: [PATCH v10 11/41] btrfs: implement log-structured superblock for ZONED mode

In reply to Naohiro Aota
Quoted 22 lines · Naohiro Aota

Superblock (and its copies) is the only data structure in btrfs which has a fixed location on a device. Since we cannot overwrite in a sequential write required zone, we cannot place superblock in the zone. One easy solution is limiting superblock and copies to be placed only in conventional zones. However, this method has two downsides: one is reduced number of superblock copies. The location of the second copy of superblock is 256GB, which is in a sequential write required zone on typical devices in the market today. So, the number of superblock and copies is limited to be two. Second downside is that we cannot support devices which have no conventional zones at all.

To solve these two problems, we employ superblock log writing. It uses two zones as a circular buffer to write updated superblocks. Once the first zone is filled up, start writing into the second buffer. Then, when the both zones are filled up and before start writing to the first zone again, it reset the first zone.

We can determine the position of the latest superblock by reading write pointer information from a device. One corner case is when the both zones are full. For this situation, we read out the last superblock of each zone, and compare them to determine which zone is older.

Quoted 6 lines

The following zones are reserved as the circular buffer on ZONED btrfs.

- The primary superblock: zones 0 and 1
- The first copy: zones 16 and 17
- The second copy: zones 1024 or zone at 256GB which is minimum, and next to it

I was thinking about that, again. We need a specification. The above is too vague.

- supported zone sizes
eg. if device has 256M, how does it work? I think we can support zones from some range (256M-1G), where filling the zone will start filing the other zone, leaving the remaining space empty if needed, effectively reserving the logical range [0..2G] for superblock

- related to the above, is it necessary to fill the whole zone? if both zones are filled, assuming 1G zone size, do we really expect the user to wait until 2G of data are read?
with average reading speed 150MB/s, reading 2G will take about 13 seconds, just to find the latest copy of the superblock(!)

- what are exact offsets of the superblocks
primary (64K), ie. not from the beginning
as partitioning is not supported, nor bootloaders, we don't need to worry about overwriting them

- what is an application supposed to do when there's a garbage after a sequence of valid superblocks (all zeros can be considered a valid termination block)

The idea is to provide enough information for a 3rd party tool to read the superblock (blkid, progs) and decouple the format from current hardware capabilities. If the zones are going to be large in the future we might consider allowing further flexibility, or fix the current zone maximum to 1G and in the future add a separate incompat bit that would extend the maximum to say 10G.

#68 lore raw

Re: [PATCH v10 11/41] btrfs: implement log-structured superblock for ZONED mode

Quoted 54 lines · David Sterba
Quoted 22 lines · Naohiro Aota

Superblock (and its copies) is the only data structure in btrfs which has a fixed location on a device. Since we cannot overwrite in a sequential write required zone, we cannot place superblock in the zone. One easy solution is limiting superblock and copies to be placed only in conventional zones. However, this method has two downsides: one is reduced number of superblock copies. The location of the second copy of superblock is 256GB, which is in a sequential write required zone on typical devices in the market today. So, the number of superblock and copies is limited to be two. Second downside is that we cannot support devices which have no conventional zones at all.

To solve these two problems, we employ superblock log writing. It uses two zones as a circular buffer to write updated superblocks. Once the first zone is filled up, start writing into the second buffer. Then, when the both zones are filled up and before start writing to the first zone again, it reset the first zone.

We can determine the position of the latest superblock by reading write pointer information from a device. One corner case is when the both zones are full. For this situation, we read out the last superblock of each zone, and compare them to determine which zone is older.

Quoted 6 lines

The following zones are reserved as the circular buffer on ZONED btrfs.

- The primary superblock: zones 0 and 1
- The first copy: zones 16 and 17
- The second copy: zones 1024 or zone at 256GB which is minimum, and next to it

I was thinking about that, again. We need a specification. The above is too vague.

- supported zone sizes
eg. if device has 256M, how does it work? I think we can support zones from some range (256M-1G), where filling the zone will start filing the other zone, leaving the remaining space empty if needed, effectively reserving the logical range [0..2G] for superblock

- related to the above, is it necessary to fill the whole zone? if both zones are filled, assuming 1G zone size, do we really expect the user to wait until 2G of data are read?
with average reading speed 150MB/s, reading 2G will take about 13 seconds, just to find the latest copy of the superblock(!)

- what are exact offsets of the superblocks
primary (64K), ie. not from the beginning
as partitioning is not supported, nor bootloaders, we don't need to worry about overwriting them

- what is an application supposed to do when there's a garbage after a sequence of valid superblocks (all zeros can be considered a valid termination block)

Quoted 7 lines

The idea is to provide enough information for a 3rd party tool to read the superblock (blkid, progs) and decouple the format from current hardware capabilities. If the zones are going to be large in the future we might consider allowing further flexibility, or fix the current zone maximum to 1G and in the future add a separate incompat bit that would extend the maximum to say 10G.

We don't need to do that. All we need to do for finding the valid superblock is a report zones call, get the write pointer and then read from write-pointer - sizeof(struct brtfs_super_block). There is no need for scanning a whole zone. The last thing that was written will be right before the write pointer.

#69 lore raw

Re: [PATCH v10 11/41] btrfs: implement log-structured superblock for ZONED mode

In reply to Naohiro Aota
Quoted 24 lines · Naohiro Aota

Superblock (and its copies) is the only data structure in btrfs which has a fixed location on a device. Since we cannot overwrite in a sequential write required zone, we cannot place superblock in the zone. One easy solution is limiting superblock and copies to be placed only in conventional zones. However, this method has two downsides: one is reduced number of superblock copies. The location of the second copy of superblock is 256GB, which is in a sequential write required zone on typical devices in the market today. So, the number of superblock and copies is limited to be two. Second downside is that we cannot support devices which have no conventional zones at all.

To solve these two problems, we employ superblock log writing. It uses two zones as a circular buffer to write updated superblocks. Once the first zone is filled up, start writing into the second buffer. Then, when the both zones are filled up and before start writing to the first zone again, it reset the first zone.

We can determine the position of the latest superblock by reading write pointer information from a device. One corner case is when the both zones are full. For this situation, we read out the last superblock of each zone, and compare them to determine which zone is older.

Quoted 6 lines

The following zones are reserved as the circular buffer on ZONED btrfs.

- The primary superblock: zones 0 and 1
- The first copy: zones 16 and 17
- The second copy: zones 1024 or zone at 256GB which is minimum, and next to it

Superblock log approach needs a non-deterministic and inconsistent number of blocks to be read to find copy #0. And, to use 4K bytes we are reserving a lot more space. But I don't know any better way. I am just checking with you...

At the time of mkfs, is it possible to format the block device to add conventional zones as needed to support our sb LBAs?
OR
For superblock zones why not reset the write pointer before the transaction commit?

Thanks.

Quoted 6 lines

If these reserved zones are conventional, superblock is written fixed at the start of the zone without logging.

#70 lore raw

Re: [PATCH v10 11/41] btrfs: implement log-structured superblock for ZONED mode

Quoted 32 lines · Anand Jain
Quoted 24 lines · Naohiro Aota

Superblock (and its copies) is the only data structure in btrfs which has a fixed location on a device. Since we cannot overwrite in a sequential write required zone, we cannot place superblock in the zone. One easy solution is limiting superblock and copies to be placed only in conventional zones. However, this method has two downsides: one is reduced number of superblock copies. The location of the second copy of superblock is 256GB, which is in a sequential write required zone on typical devices in the market today. So, the number of superblock and copies is limited to be two.  Second downside is that we cannot support devices which have no conventional zones at all.

To solve these two problems, we employ superblock log writing. It uses two zones as a circular buffer to write updated superblocks. Once the first zone is filled up, start writing into the second buffer. Then, when the both zones are filled up and before start writing to the first zone again, it reset the first zone.

We can determine the position of the latest superblock by reading write pointer information from a device. One corner case is when the both zones are full. For this situation, we read out the last superblock of each zone, and compare them to determine which zone is older.

Quoted 6 lines

The following zones are reserved as the circular buffer on ZONED btrfs.

- The primary superblock: zones 0 and 1
- The first copy: zones 16 and 17
- The second copy: zones 1024 or zone at 256GB which is minimum, and next to it

Quoted 7 lines

Superblock log approach needs a non-deterministic and inconsistent number of blocks to be read to find copy #0. And, to use 4K bytes we are reserving a lot more space. But I don't know any better way. I am just checking with you...

At the time of mkfs, is it possible to format the block device to add conventional zones as needed to support our sb LBAs?

No. The number of conventional zones (if any) are a drive characteristic and one cannot assume that the number can be modified.

Quoted 4 lines

 OR
For superblock zones why not reset the write pointer before the transaction commit?

A write pointer reset is equivalent to clearing the contents of the zone, so we would lose the previous information there.

HTH.

Cheers,

Hannes

Signature · 4 lines
Dr. Hannes Reinecke                Kernel Storage Architect
[email protected]                              +49 911 74053 688
SUSE Software Solutions GmbH, Maxfeldstr. 5, 90409 Nürnberg
HRB 36809 (AG Nürnberg), Geschäftsführer: Felix Imendörffer
#71 lore raw

PATCH v10 12/41 btrfs: implement zoned chunk allocator

In reply to Naohiro Aota

This commit implements a zoned chunk/dev_extent allocator. The zoned allocator aligns the device extents to zone boundaries, so that a zone reset affects only the device extent and does not change the state of blocks in the neighbor device extents.

Also, it checks that a region allocation is not overlapping any of the super block zones, and ensures the region is empty.

---
 fs/btrfs/volumes.c | 136 ++++++++++++++++++++++++++++++++++++++++++
 fs/btrfs/volumes.h |   1 +
 fs/btrfs/zoned.c   | 144 +++++++++++++++++++++++++++++++++++++++++++++
 fs/btrfs/zoned.h   |  34 +++++++++++
 4 files changed, 315 insertions(+)
fs/btrfs/volumes.c +136 -0
index db884b96a5ea..7831cf6c6da4 100644
@@ -1416,6 +1416,21 @@ static bool contains_pending_extent(struct btrfs_device *device, u64 *start,
 	return false;
 }
 
+static inline u64 dev_extent_search_start_zoned(struct btrfs_device *device,
+						u64 start)
+{
+	u64 tmp;
+
+	if (device->zone_info->zone_size > SZ_1M)
+		tmp = device->zone_info->zone_size;
+	else
+		tmp = SZ_1M;
+	if (start < tmp)
+		start = tmp;
+
+	return btrfs_align_offset_to_zone(device, start);
+}
+
 static u64 dev_extent_search_start(struct btrfs_device *device, u64 start)
 {
 	switch (device->fs_devices->chunk_alloc_policy) {
@@ -1426,11 +1441,57 @@ static u64 dev_extent_search_start(struct btrfs_device *device, u64 start)
 		 * make sure to start at an offset of at least 1MB.
 		 */
 		return max_t(u64, start, SZ_1M);
+	case BTRFS_CHUNK_ALLOC_ZONED:
+		return dev_extent_search_start_zoned(device, start);
 	default:
 		BUG();
 	}
 }
 
+static bool dev_extent_hole_check_zoned(struct btrfs_device *device,
+					u64 *hole_start, u64 *hole_size,
+					u64 num_bytes)
+{
+	u64 zone_size = device->zone_info->zone_size;
+	u64 pos;
+	int ret;
+	int changed = 0;
+
+	ASSERT(IS_ALIGNED(*hole_start, zone_size));
+
+	while (*hole_size > 0) {
+		pos = btrfs_find_allocatable_zones(device, *hole_start,
+						   *hole_start + *hole_size,
+						   num_bytes);
+		if (pos != *hole_start) {
+			*hole_size = *hole_start + *hole_size - pos;
+			*hole_start = pos;
+			changed = 1;
+			if (*hole_size < num_bytes)
+				break;
+		}
+
+		ret = btrfs_ensure_empty_zones(device, pos, num_bytes);
+
+		/* Range is ensured to be empty */
+		if (!ret)
+			return changed;
+
+		/* Given hole range was invalid (outside of device) */
+		if (ret == -ERANGE) {
+			*hole_start += *hole_size;
+			*hole_size = 0;
+			return 1;
+		}
+
+		*hole_start += zone_size;
+		*hole_size -= zone_size;
+		changed = 1;
+	}
+
+	return changed;
+}
+
 /**
  * dev_extent_hole_check - check if specified hole is suitable for allocation
  * @device:	the device which we have the hole
@@ -1463,6 +1524,10 @@ static bool dev_extent_hole_check(struct btrfs_device *device, u64 *hole_start,
 	case BTRFS_CHUNK_ALLOC_REGULAR:
 		/* No extra check */
 		break;
+	case BTRFS_CHUNK_ALLOC_ZONED:
+		changed |= dev_extent_hole_check_zoned(device, hole_start,
+						       hole_size, num_bytes);
+		break;
 	default:
 		BUG();
 	}
@@ -1517,6 +1582,9 @@ static int find_free_dev_extent_start(struct btrfs_device *device,
 
 	search_start = dev_extent_search_start(device, search_start);
 
+	WARN_ON(device->zone_info &&
+		!IS_ALIGNED(num_bytes, device->zone_info->zone_size));
+
 	path = btrfs_alloc_path();
 	if (!path)
 		return -ENOMEM;
@@ -4907,6 +4975,37 @@ static void init_alloc_chunk_ctl_policy_regular(
 	ctl->dev_extent_min = BTRFS_STRIPE_LEN * ctl->dev_stripes;
 }
 
+static void init_alloc_chunk_ctl_policy_zoned(
+				      struct btrfs_fs_devices *fs_devices,
+				      struct alloc_chunk_ctl *ctl)
+{
+	u64 zone_size = fs_devices->fs_info->zone_size;
+	u64 limit;
+	int min_num_stripes = ctl->devs_min * ctl->dev_stripes;
+	int min_data_stripes = (min_num_stripes - ctl->nparity) / ctl->ncopies;
+	u64 min_chunk_size = min_data_stripes * zone_size;
+	u64 type = ctl->type;
+
+	ctl->max_stripe_size = zone_size;
+	if (type & BTRFS_BLOCK_GROUP_DATA) {
+		ctl->max_chunk_size = round_down(BTRFS_MAX_DATA_CHUNK_SIZE,
+						 zone_size);
+	} else if (type & BTRFS_BLOCK_GROUP_METADATA) {
+		ctl->max_chunk_size = ctl->max_stripe_size;
+	} else if (type & BTRFS_BLOCK_GROUP_SYSTEM) {
+		ctl->max_chunk_size = 2 * ctl->max_stripe_size;
+		ctl->devs_max = min_t(int, ctl->devs_max,
+				      BTRFS_MAX_DEVS_SYS_CHUNK);
+	}
+
+	/* We don't want a chunk larger than 10% of writable space */
+	limit = max(round_down(div_factor(fs_devices->total_rw_bytes, 1),
+			       zone_size),
+		    min_chunk_size);
+	ctl->max_chunk_size = min(limit, ctl->max_chunk_size);
+	ctl->dev_extent_min = zone_size * ctl->dev_stripes;
+}
+
 static void init_alloc_chunk_ctl(struct btrfs_fs_devices *fs_devices,
 				 struct alloc_chunk_ctl *ctl)
 {
@@ -4927,6 +5026,9 @@ static void init_alloc_chunk_ctl(struct btrfs_fs_devices *fs_devices,
 	case BTRFS_CHUNK_ALLOC_REGULAR:
 		init_alloc_chunk_ctl_policy_regular(fs_devices, ctl);
 		break;
+	case BTRFS_CHUNK_ALLOC_ZONED:
+		init_alloc_chunk_ctl_policy_zoned(fs_devices, ctl);
+		break;
 	default:
 		BUG();
 	}
@@ -5053,6 +5155,38 @@ static int decide_stripe_size_regular(struct alloc_chunk_ctl *ctl,
 	return 0;
 }
 
+static int decide_stripe_size_zoned(struct alloc_chunk_ctl *ctl,
+				    struct btrfs_device_info *devices_info)
+{
+	u64 zone_size = devices_info[0].dev->zone_info->zone_size;
+	/* Number of stripes that count for block group size */
+	int data_stripes;
+
+	/*
+	 * It should hold because:
+	 *    dev_extent_min == dev_extent_want == zone_size * dev_stripes
+	 */
+	ASSERT(devices_info[ctl->ndevs - 1].max_avail == ctl->dev_extent_min);
+
+	ctl->stripe_size = zone_size;
+	ctl->num_stripes = ctl->ndevs * ctl->dev_stripes;
+	data_stripes = (ctl->num_stripes - ctl->nparity) / ctl->ncopies;
+
+	/* stripe_size is fixed in ZONED. Reduce ndevs instead. */
+	if (ctl->stripe_size * data_stripes > ctl->max_chunk_size) {
+		ctl->ndevs = div_u64(div_u64(ctl->max_chunk_size * ctl->ncopies,
+					     ctl->stripe_size) + ctl->nparity,
+				     ctl->dev_stripes);
+		ctl->num_stripes = ctl->ndevs * ctl->dev_stripes;
+		data_stripes = (ctl->num_stripes - ctl->nparity) / ctl->ncopies;
+		ASSERT(ctl->stripe_size * data_stripes <= ctl->max_chunk_size);
+	}
+
+	ctl->chunk_size = ctl->stripe_size * data_stripes;
+
+	return 0;
+}
+
 static int decide_stripe_size(struct btrfs_fs_devices *fs_devices,
 			      struct alloc_chunk_ctl *ctl,
 			      struct btrfs_device_info *devices_info)
@@ -5080,6 +5214,8 @@ static int decide_stripe_size(struct btrfs_fs_devices *fs_devices,
 	switch (fs_devices->chunk_alloc_policy) {
 	case BTRFS_CHUNK_ALLOC_REGULAR:
 		return decide_stripe_size_regular(ctl, devices_info);
+	case BTRFS_CHUNK_ALLOC_ZONED:
+		return decide_stripe_size_zoned(ctl, devices_info);
 	default:
 		BUG();
 	}
fs/btrfs/volumes.h +1 -0
index 9c07b97a2260..0249aca668fb 100644
@@ -213,6 +213,7 @@ BTRFS_DEVICE_GETSET_FUNCS(bytes_used);
 
 enum btrfs_chunk_allocation_policy {
 	BTRFS_CHUNK_ALLOC_REGULAR,
+	BTRFS_CHUNK_ALLOC_ZONED,
 };
 
 struct btrfs_fs_devices {
fs/btrfs/zoned.c +144 -0

164 lines not on this page · raw message

fs/btrfs/zoned.h +34 -0
index de9d7dd8c351..ec2391c52d8b 100644
@@ -35,6 +35,11 @@ int btrfs_sb_log_location(struct btrfs_device *device, int mirror, int rw,
 			  u64 *bytenr_ret);
 void btrfs_advance_sb_log(struct btrfs_device *device, int mirror);
 int btrfs_reset_sb_log_zones(struct block_device *bdev, int mirror);
+u64 btrfs_find_allocatable_zones(struct btrfs_device *device, u64 hole_start,
+				 u64 hole_end, u64 num_bytes);
+int btrfs_reset_device_zone(struct btrfs_device *device, u64 physical,
+			    u64 length, u64 *bytes);
+int btrfs_ensure_empty_zones(struct btrfs_device *device, u64 start, u64 size);
 #else /* CONFIG_BLK_DEV_ZONED */
 static inline int btrfs_get_dev_zone(struct btrfs_device *device, u64 pos,
 				     struct blk_zone *zone)
@@ -87,6 +92,26 @@ static inline int btrfs_reset_sb_log_zones(struct block_device *bdev,
 	return 0;
 }
 
+static inline u64 btrfs_find_allocatable_zones(struct btrfs_device *device,
+					       u64 hole_start, u64 hole_end,
+					       u64 num_bytes)
+{
+	return hole_start;
+}
+
+static inline int btrfs_reset_device_zone(struct btrfs_device *device,
+					  u64 physical, u64 length, u64 *bytes)
+{
+	*bytes = 0;
+	return 0;
+}
+
+static inline int btrfs_ensure_empty_zones(struct btrfs_device *device,
+					   u64 start, u64 size)
+{
+	return 0;
+}
+
 #endif
 
 static inline bool btrfs_dev_is_sequential(struct btrfs_device *device, u64 pos)
@@ -165,4 +190,13 @@ static inline bool btrfs_check_super_location(struct btrfs_device *device,
 	       !btrfs_dev_is_sequential(device, pos);
 }
 
+static inline u64 btrfs_align_offset_to_zone(struct btrfs_device *device,
+					     u64 pos)
+{
+	if (!device->zone_info)
+		return pos;
+
+	return ALIGN(pos, device->zone_info->zone_size);
+}
+
 #endif
Signature · 1 line
2.27.0
#72 lore raw

Re: [PATCH v10 12/41] btrfs: implement zoned chunk allocator

Quoted 9 lines · Naohiro Aota

This commit implements a zoned chunk/dev_extent allocator. The zoned allocator aligns the device extents to zone boundaries, so that a zone reset affects only the device extent and does not change the state of blocks in the neighbor device extents.

Also, it checks that a region allocation is not overlapping any of the super block zones, and ensures the region is empty.

Looks good.

Chunks and stripes are aligned to the zone_size. I guess zone_size won't change after the block device has been formatted with it? For testing, what if the device image is dumped onto another zoned device with a different zone_size?

A small nit is below.

Quoted 17 lines
fs/btrfs/volumes.c +17 -0
+static void init_alloc_chunk_ctl_policy_zoned(
+				      struct btrfs_fs_devices *fs_devices,
+				      struct alloc_chunk_ctl *ctl)
+{
+	u64 zone_size = fs_devices->fs_info->zone_size;
+	u64 limit;
+	int min_num_stripes = ctl->devs_min * ctl->dev_stripes;
+	int min_data_stripes = (min_num_stripes - ctl->nparity) / ctl->ncopies;
+	u64 min_chunk_size = min_data_stripes * zone_size;
+	u64 type = ctl->type;
+
+	ctl->max_stripe_size = zone_size;
+	if (type & BTRFS_BLOCK_GROUP_DATA) {
+		ctl->max_chunk_size = round_down(BTRFS_MAX_DATA_CHUNK_SIZE,
+						 zone_size);
+	} else if (type & BTRFS_BLOCK_GROUP_METADATA) {
+		ctl->max_chunk_size = ctl->max_stripe_size;

fs/btrfs/volumes.c static void init_alloc_chunk_ctl_policy_regular( · line 5002

Quoted 10 lines
fs/btrfs/volumes.c +8 -0
+	} else if (type & BTRFS_BLOCK_GROUP_SYSTEM) {
+		ctl->max_chunk_size = 2 * ctl->max_stripe_size;
+		ctl->devs_max = min_t(int, ctl->devs_max,
+				      BTRFS_MAX_DEVS_SYS_CHUNK);
+	}
+
 
 
+	/* We don't want a chunk larger than 10% of writable space */
+	limit = max(round_down(div_factor(fs_devices->total_rw_bytes, 1),

What's the purpose of dev_factor here?

Thanks.

#73 lore raw

Re: [PATCH v10 12/41] btrfs: implement zoned chunk allocator

Quoted 9 lines · Anand Jain
Quoted 8 lines · Naohiro Aota

This commit implements a zoned chunk/dev_extent allocator. The zoned allocator aligns the device extents to zone boundaries, so that a zone reset affects only the device extent and does not change the state of blocks in the neighbor device extents.

Also, it checks that a region allocation is not overlapping any of the super block zones, and ensures the region is empty.

Quoted 8 lines
Quoted 1 line

Looks good.

Chunks and stripes are aligned to the zone_size. I guess zone_size won't change after the block device has been formatted with it? For testing, what if the device image is dumped onto another zoned device with a different zone_size?

Zone size is a drive characteristic, so it never change on the same device.

Dump/restore on another device with a different zone_size should be banned, because we cannot ensure device extents are aligned to zone boundaries.

Quoted 28 lines

A small nit is below.

Quoted 23 lines
fs/btrfs/volumes.c +23 -0
+static void init_alloc_chunk_ctl_policy_zoned(
+				      struct btrfs_fs_devices *fs_devices,
+				      struct alloc_chunk_ctl *ctl)
+{
+	u64 zone_size = fs_devices->fs_info->zone_size;
+	u64 limit;
+	int min_num_stripes = ctl->devs_min * ctl->dev_stripes;
+	int min_data_stripes = (min_num_stripes - ctl->nparity) / ctl->ncopies;
+	u64 min_chunk_size = min_data_stripes * zone_size;
+	u64 type = ctl->type;
+
+	ctl->max_stripe_size = zone_size;
+	if (type & BTRFS_BLOCK_GROUP_DATA) {
+		ctl->max_chunk_size = round_down(BTRFS_MAX_DATA_CHUNK_SIZE,
+						 zone_size);
+	} else if (type & BTRFS_BLOCK_GROUP_METADATA) {
+		ctl->max_chunk_size = ctl->max_stripe_size;
+	} else if (type & BTRFS_BLOCK_GROUP_SYSTEM) {
+		ctl->max_chunk_size = 2 * ctl->max_stripe_size;
+		ctl->devs_max = min_t(int, ctl->devs_max,
+				      BTRFS_MAX_DEVS_SYS_CHUNK);
+	}
+
Quoted 4 lines

fs/btrfs/volumes.c static void init_alloc_chunk_ctl_policy_regular( · line 5002

Quoted 2 lines
fs/btrfs/volumes.c +2 -0
+	/* We don't want a chunk larger than 10% of writable space */
+	limit = max(round_down(div_factor(fs_devices->total_rw_bytes, 1),

What's the purpose of dev_factor here?

This one follows the same limitation as in regular allocator (init_alloc_chunk_ctl_policy_regular).

Quoted 3 lines

Thanks.

#74 lore raw

Re: [PATCH v10 12/41] btrfs: implement zoned chunk allocator

Quoted 19 lines · Naohiro Aota
Quoted 9 lines · Anand Jain
Quoted 8 lines · Naohiro Aota

This commit implements a zoned chunk/dev_extent allocator. The zoned allocator aligns the device extents to zone boundaries, so that a zone reset affects only the device extent and does not change the state of blocks in the neighbor device extents.

Also, it checks that a region allocation is not overlapping any of the super block zones, and ensures the region is empty.

Quoted 8 lines
Quoted 1 line

Looks good.

Chunks and stripes are aligned to the zone_size. I guess zone_size won't change after the block device has been formatted with it? For testing, what if the device image is dumped onto another zoned device with a different zone_size?

Quoted 4 lines

Zone size is a drive characteristic, so it never change on the same device.

Dump/restore on another device with a different zone_size should be banned, because we cannot ensure device extents are aligned to zone boundaries.

Fair enough. Do we have any checks to fail such mount? Sorry if I have missed it somewhere in the patch? Thanks.

#75 lore raw

Re: [PATCH v10 12/41] btrfs: implement zoned chunk allocator

Quoted 22 lines · Anand Jain
Quoted 19 lines · Naohiro Aota
Quoted 17 lines · Anand Jain
Quoted 9 lines · Naohiro Aota

This commit implements a zoned chunk/dev_extent allocator. The zoned allocator aligns the device extents to zone boundaries, so that a zone reset affects only the device extent and does not change the state of blocks in the neighbor device extents.

Also, it checks that a region allocation is not overlapping any of the super block zones, and ensures the region is empty.

Looks good.

Chunks and stripes are aligned to the zone_size. I guess zone_size won't change after the block device has been formatted with it? For testing, what if the device image is dumped onto another zoned device with a different zone_size?

Quoted 7 lines
Quoted 4 lines

Zone size is a drive characteristic, so it never change on the same device.

Dump/restore on another device with a different zone_size should be banned, because we cannot ensure device extents are aligned to zone boundaries.

Fair enough. Do we have any checks to fail such mount? Sorry if I have missed it somewhere in the patch? Thanks.

We have a check in verify_one_dev_extent() to confirm that a device extent's position and size are aligned to zone size (patch 13).

#76 lore raw

Re: [PATCH v10 12/41] btrfs: implement zoned chunk allocator

In reply to Naohiro Aota
Quoted 19 lines · Naohiro Aota
Quoted 9 lines · Anand Jain
Quoted 8 lines · Naohiro Aota

This commit implements a zoned chunk/dev_extent allocator. The zoned allocator aligns the device extents to zone boundaries, so that a zone reset affects only the device extent and does not change the state of blocks in the neighbor device extents.

Also, it checks that a region allocation is not overlapping any of the super block zones, and ensures the region is empty.

Quoted 8 lines
Quoted 1 line

Looks good.

Chunks and stripes are aligned to the zone_size. I guess zone_size won't change after the block device has been formatted with it? For testing, what if the device image is dumped onto another zoned device with a different zone_size?

Quoted 4 lines

Zone size is a drive characteristic, so it never change on the same device.

Dump/restore on another device with a different zone_size should be banned, because we cannot ensure device extents are aligned to zone boundaries.

Does this mean 'btrfs replace' is banned as well? Or is it allowed to a similar-enough device? What about 'add' followed by 'remove'?

#77 lore raw

Re: [PATCH v10 12/41] btrfs: implement zoned chunk allocator

Quoted 20 lines · Graham Cobb
Quoted 19 lines · Naohiro Aota
Quoted 17 lines · Anand Jain
Quoted 9 lines · Naohiro Aota

This commit implements a zoned chunk/dev_extent allocator. The zoned allocator aligns the device extents to zone boundaries, so that a zone reset affects only the device extent and does not change the state of blocks in the neighbor device extents.

Also, it checks that a region allocation is not overlapping any of the super block zones, and ensures the region is empty.

Looks good.

Chunks and stripes are aligned to the zone_size. I guess zone_size won't change after the block device has been formatted with it? For testing, what if the device image is dumped onto another zoned device with a different zone_size?

Quoted 7 lines
Quoted 4 lines

Zone size is a drive characteristic, so it never change on the same device.

Dump/restore on another device with a different zone_size should be banned, because we cannot ensure device extents are aligned to zone boundaries.

Does this mean 'btrfs replace' is banned as well? Or is it allowed to a similar-enough device? What about 'add' followed by 'remove'?

Replacing is allowed if the zone size is the same. Adding a disk is the same. This restriction is checked in btrfs_init_new_device() (patch 5).

#78 lore raw

Re: [PATCH v10 12/41] btrfs: implement zoned chunk allocator

In reply to Naohiro Aota
Quoted 46 lines
fs/btrfs/volumes.c +17 -0
index db884b96a5ea..7831cf6c6da4 100644
@@ -1416,6 +1416,21 @@ static bool contains_pending_extent(struct btrfs_device *device, u64 *start,
  	return false;
  }
 
+static inline u64 dev_extent_search_start_zoned(struct btrfs_device *device,
+						u64 start)
+{
+	u64 tmp;
+
+	if (device->zone_info->zone_size > SZ_1M)
+		tmp = device->zone_info->zone_size;
+	else
+		tmp = SZ_1M;
+	if (start < tmp)
+		start = tmp;
+
+	return btrfs_align_offset_to_zone(device, start);
+}
+
  static u64 dev_extent_search_start(struct btrfs_device *device, u64 start)
  {
  	switch (device->fs_devices->chunk_alloc_policy) {
@@ -1426,11 +1441,57 @@ static u64 dev_extent_search_start(struct btrfs_device *device, u64 start)
  		 * make sure to start at an offset of at least 1MB.
  		 */
  		return max_t(u64, start, SZ_1M);
+	case BTRFS_CHUNK_ALLOC_ZONED:
+		return dev_extent_search_start_zoned(device, start);
  	default:
  		BUG();
  	}
  }
 
 
@@ -165,4 +190,13 @@ static inline bool btrfs_check_super_location(struct btrfs_device *device,
  	       !btrfs_dev_is_sequential(device, pos);
  }
 

fs/btrfs/volumes.c static inline bool btrfs_check_super_location(struct btrfs_device *device, · line 202

Quoted 10 lines
fs/btrfs/volumes.c +9 -0
+static inline u64 btrfs_align_offset_to_zone(struct btrfs_device *device,
+					     u64 pos)
+{
+	if (!device->zone_info)
+		return pos;
+
+	return ALIGN(pos, device->zone_info->zone_size);
+}
+
  #endif

Small functions (such as above) can be opened coded to make the reviewing easier. The btrfs_align_offset_to_zone() and dev_extent_search_start_zoned() can be open coded and merged into the parent function dev_extent_search_start() as below...

dev_extent_search_start()
::

	case BTRFS_CHUNK_ALLOC_ZONED:
		start = max_t(u64, start,
			      max_t(u64, device->zone_info->zone_size, SZ_1M));

         return ALIGN(start, device->zone_info->zone_size);

As of now we don't allow mix of zoned with regular device in a btrfs (those are verified during mount and device add/replace). So we don't have to check for the same again in btrfs_align_offset_to_zone().

Thanks.

#79 lore raw

PATCH v10 13/41 btrfs: verify device extent is aligned to zone

In reply to Naohiro Aota

Add a check in verify_one_dev_extent() to check if a device extent on a zoned block device is aligned to the respective zone boundary.

---
 fs/btrfs/volumes.c | 14 ++++++++++++++
 1 file changed, 14 insertions(+)
fs/btrfs/volumes.c +14 -0
index 7831cf6c6da4..c0e27c1e2559 100644
@@ -7783,6 +7783,20 @@ static int verify_one_dev_extent(struct btrfs_fs_info *fs_info,
 		ret = -EUCLEAN;
 		goto out;
 	}
+
+	if (dev->zone_info) {
+		u64 zone_size = dev->zone_info->zone_size;
+
+		if (!IS_ALIGNED(physical_offset, zone_size) ||
+		    !IS_ALIGNED(physical_len, zone_size)) {
+			btrfs_err(fs_info,
+"zoned: dev extent devid %llu physical offset %llu len %llu is not aligned to device zone",
+				  devid, physical_offset, physical_len);
+			ret = -EUCLEAN;
+			goto out;
+		}
+	}
+
 out:
 	free_extent_map(em);
 	return ret;
Signature · 1 line
2.27.0
Reviewed-by on v10 13/41
#80 lore raw

Re: [PATCH v10 13/41] btrfs: verify device extent is aligned to zone

Quoted 25 lines · Naohiro Aota

Add a check in verify_one_dev_extent() to check if a device extent on a zoned block device is aligned to the respective zone boundary.

---
  fs/btrfs/volumes.c | 14 ++++++++++++++
  1 file changed, 14 insertions(+)
fs/btrfs/volumes.c +7 -0
index 7831cf6c6da4..c0e27c1e2559 100644
@@ -7783,6 +7783,20 @@ static int verify_one_dev_extent(struct btrfs_fs_info *fs_info,
  		ret = -EUCLEAN;
  		goto out;
  	}
+
+	if (dev->zone_info) {
+		u64 zone_size = dev->zone_info->zone_size;
+
+		if (!IS_ALIGNED(physical_offset, zone_size) ||
+		    !IS_ALIGNED(physical_len, zone_size)) {
+			btrfs_err(fs_info,

fs/btrfs/volumes.c static int verify_one_dev_extent(struct btrfs_fs_info *fs_info, · line 7802

Quoted 10 lines
fs/btrfs/volumes.c +7 -0
+"zoned: dev extent devid %llu physical offset %llu len %llu is not aligned to device zone",
+				  devid, physical_offset, physical_len);
+			ret = -EUCLEAN;
+			goto out;
+		}
+	}
+
  out:
  	free_extent_map(em);
  	return ret;

Looks good.

#81 lore raw

PATCH v10 14/41 btrfs: load zone's alloction offset

In reply to Naohiro Aota

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() 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.

This commit supports the case all the zones in a block group are sequential. The next patch will handle the case having a conventional zone.

---
 fs/btrfs/block-group.c |  15 ++++
 fs/btrfs/block-group.h |   6 ++
 fs/btrfs/zoned.c       | 154 +++++++++++++++++++++++++++++++++++++++++
 fs/btrfs/zoned.h       |   7 ++
 4 files changed, 182 insertions(+)
fs/btrfs/block-group.c +15 -0
index 6b4831824f51..ffc64dfbe09e 100644
@@ -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, "zoned: 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 */
fs/btrfs/block-group.h +6 -0
index adfd7583a17b..14e3043c9ce7 100644
@@ -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)
fs/btrfs/zoned.c +154 -0

171 lines not on this page · raw message

fs/btrfs/zoned.h +7 -0
index ec2391c52d8b..e3338a2f1be9 100644
@@ -40,6 +40,7 @@ u64 btrfs_find_allocatable_zones(struct btrfs_device *device, u64 hole_start,
 int btrfs_reset_device_zone(struct btrfs_device *device, u64 physical,
 			    u64 length, u64 *bytes);
 int btrfs_ensure_empty_zones(struct btrfs_device *device, u64 start, u64 size);
+int btrfs_load_block_group_zone_info(struct btrfs_block_group *cache);
 #else /* CONFIG_BLK_DEV_ZONED */
 static inline int btrfs_get_dev_zone(struct btrfs_device *device, u64 pos,
 				     struct blk_zone *zone)
@@ -112,6 +113,12 @@ static inline int btrfs_ensure_empty_zones(struct btrfs_device *device,
 	return 0;
 }
 
+static inline int btrfs_load_block_group_zone_info(
+	struct btrfs_block_group *cache)
+{
+	return 0;
+}
+
 #endif
 
 static inline bool btrfs_dev_is_sequential(struct btrfs_device *device, u64 pos)
Signature · 1 line
2.27.0
Reviewed-by on v10 14/41
#82 lore raw

Re: [PATCH v10 14/41] btrfs: load zone's alloction offset

Quoted 280 lines · Naohiro Aota

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() 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.

This commit supports the case all the zones in a block group are sequential. The next patch will handle the case having a conventional zone.

---
  fs/btrfs/block-group.c |  15 ++++
  fs/btrfs/block-group.h |   6 ++
  fs/btrfs/zoned.c       | 154 +++++++++++++++++++++++++++++++++++++++++
  fs/btrfs/zoned.h       |   7 ++
  4 files changed, 182 insertions(+)
fs/btrfs/block-group.c +15 -0
index 6b4831824f51..ffc64dfbe09e 100644
@@ -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, "zoned: 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 */
fs/btrfs/block-group.h +6 -0
index adfd7583a17b..14e3043c9ce7 100644
@@ -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)
fs/btrfs/zoned.c +154 -0
index ed5de1c138d7..69d3412c4fef 100644
@@ -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)
 
  /* Number of superblock log zones */
  #define BTRFS_NR_SB_LOG_ZONES 2
@@ -777,3 +783,151 @@ 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, "zoned: block group %llu len %llu unaligned to zone size %llu",
+			  logical, length, fs_info->zone_size);
+		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, "zoned: offline/readonly zone %llu on device %s (devid %llu)",
+				  physical >> device->zone_info->zone_size_shift,
+				  rcu_str_deref(device->name), device->devid);
+			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 do not have a write pointer, we
+		 * cannot determine alloc_offset from the pointer
+		 */
+		ret = -EINVAL;
+		goto out;
+	}
+
+	switch (map->type & BTRFS_BLOCK_GROUP_PROFILE_MASK) {
+	case 0: /* single */
+		cache->alloc_offset = alloc_offsets[0];
+		break;
+	case BTRFS_BLOCK_GROUP_DUP:
+	case BTRFS_BLOCK_GROUP_RAID1:
+	case BTRFS_BLOCK_GROUP_RAID0:
+	case BTRFS_BLOCK_GROUP_RAID10:
+	case BTRFS_BLOCK_GROUP_RAID5:
+	case BTRFS_BLOCK_GROUP_RAID6:
+		/* non-SINGLE profiles are not supported yet */
+	default:
+		btrfs_err(fs_info, "zoned: profile %s not supported",
+			  btrfs_bg_type_to_raid_name(map->type));
+		ret = -EINVAL;
+		goto out;
+	}
+
+out:
+	kfree(alloc_offsets);
+	free_extent_map(em);
+
+	return ret;
+}
fs/btrfs/zoned.h +1 -0
index ec2391c52d8b..e3338a2f1be9 100644
@@ -40,6 +40,7 @@ u64 btrfs_find_allocatable_zones(struct btrfs_device *device, u64 hole_start,
  int btrfs_reset_device_zone(struct btrfs_device *device, u64 physical,
  			    u64 length, u64 *bytes);
  int btrfs_ensure_empty_zones(struct btrfs_device *device, u64 start, u64 size);
+int btrfs_load_block_group_zone_info(struct btrfs_block_group *cache);
  #else /* CONFIG_BLK_DEV_ZONED */
  static inline int btrfs_get_dev_zone(struct btrfs_device *device, u64 pos,
  				     struct blk_zone *zone)
@@ -112,6 +113,12 @@ static inline int btrfs_ensure_empty_zones(struct btrfs_device *device,
  	return 0;
  }

fs/btrfs/zoned.h static inline int btrfs_ensure_empty_zones(struct btrfs_device *device, · line 124

Quoted 10 lines
fs/btrfs/zoned.h +6 -0
 
+static inline int btrfs_load_block_group_zone_info(
+	struct btrfs_block_group *cache)
+{
+	return 0;
+}
+
  #endif
 
  static inline bool btrfs_dev_is_sequential(struct btrfs_device *device, u64 pos)

looks good.

#83 lore raw

PATCH v10 15/41 btrfs: emulate write pointer for conventional zones

In reply to Naohiro Aota

Conventional zones do not have a write pointer, so we cannot use it to determine the allocation offset if a block group contains a conventional zone.

But instead, we can consider the end of the last allocated extent in the block group as an allocation offset.

---
 fs/btrfs/zoned.c | 80 ++++++++++++++++++++++++++++++++++++++++++++----
 1 file changed, 74 insertions(+), 6 deletions(-)
fs/btrfs/zoned.c +74 -6
index 69d3412c4fef..9bf40300e428 100644
@@ -784,6 +784,61 @@ int btrfs_ensure_empty_zones(struct btrfs_device *device, u64 start, u64 size)
 	return 0;
 }
 
+static int emulate_write_pointer(struct btrfs_block_group *cache,
+				 u64 *offset_ret)
+{
+	struct btrfs_fs_info *fs_info = cache->fs_info;
+	struct btrfs_root *root = fs_info->extent_root;
+	struct btrfs_path *path;
+	struct btrfs_key key;
+	struct btrfs_key found_key;
+	int ret;
+	u64 length;
+
+	path = btrfs_alloc_path();
+	if (!path)
+		return -ENOMEM;
+
+	key.objectid = cache->start + cache->length;
+	key.type = 0;
+	key.offset = 0;
+
+	ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
+	/* We should not find the exact match */
+	if (ret <= 0) {
+		ret = -EUCLEAN;
+		goto out;
+	}
+
+	ret = btrfs_previous_extent_item(root, path, cache->start);
+	if (ret) {
+		if (ret == 1) {
+			ret = 0;
+			*offset_ret = 0;
+		}
+		goto out;
+	}
+
+	btrfs_item_key_to_cpu(path->nodes[0], &found_key, path->slots[0]);
+
+	if (found_key.type == BTRFS_EXTENT_ITEM_KEY)
+		length = found_key.offset;
+	else
+		length = fs_info->nodesize;
+
+	if (!(found_key.objectid >= cache->start &&
+	       found_key.objectid + length <= cache->start + cache->length)) {
+		ret = -EUCLEAN;
+		goto out;
+	}
+	*offset_ret = found_key.objectid + length - cache->start;
+	ret = 0;
+
+out:
+	btrfs_free_path(path);
+	return ret;
+}
+
 int btrfs_load_block_group_zone_info(struct btrfs_block_group *cache)
 {
 	struct btrfs_fs_info *fs_info = cache->fs_info;
@@ -798,6 +853,7 @@ int btrfs_load_block_group_zone_info(struct btrfs_block_group *cache)
 	int i;
 	unsigned int nofs_flag;
 	u64 *alloc_offsets = NULL;
+	u64 emulated_offset = 0;
 	u32 num_sequential = 0, num_conventional = 0;
 
 	if (!btrfs_is_zoned(fs_info))
@@ -899,12 +955,16 @@ int btrfs_load_block_group_zone_info(struct btrfs_block_group *cache)
 	}
 
 	if (num_conventional > 0) {
-		/*
-		 * Since conventional zones do not have a write pointer, we
-		 * cannot determine alloc_offset from the pointer
-		 */
-		ret = -EINVAL;
-		goto out;
+		ret = emulate_write_pointer(cache, &emulated_offset);
+		if (ret || map->num_stripes == num_conventional) {
+			if (!ret)
+				cache->alloc_offset = emulated_offset;
+			else
+				btrfs_err(fs_info,
+			"zoned: failed to emulate write pointer of bg %llu",
+					  cache->start);
+			goto out;
+		}
 	}
 
 	switch (map->type & BTRFS_BLOCK_GROUP_PROFILE_MASK) {
@@ -926,6 +986,14 @@ int btrfs_load_block_group_zone_info(struct btrfs_block_group *cache)
 	}
 
 out:
+	/* An extent is allocated after the write pointer */
+	if (num_conventional && emulated_offset > cache->alloc_offset) {
+		btrfs_err(fs_info,
+			  "zoned: got wrong write pointer in BG %llu: %llu > %llu",
+			  logical, emulated_offset, cache->alloc_offset);
+		ret = -EIO;
+	}
+
 	kfree(alloc_offsets);
 	free_extent_map(em);
 
Signature · 1 line
2.27.0
#84 lore raw

PATCH v10 16/41 btrfs: track unusable bytes for zones

In reply to Naohiro Aota

In zoned btrfs a region that was once written then freed is not usable until we reset the underlying zone. So we need to distinguish such unusable space from usable free space.

Therefore we need to introduce the "zone_unusable" field to the block group structure, and "bytes_zone_unusable" to the space_info structure to track the unusable space.

Pinned bytes are always reclaimed to the unusable space. But, when an allocated region is returned before using e.g., the block group becomes read-only between allocation time and reservation time, we can safely return the region to the block group. For the situation, this commit introduces "btrfs_add_free_space_unused". This behaves the same as btrfs_add_free_space() on regular btrfs. On zoned btrfs, it rewinds the allocation offset.

---
 fs/btrfs/block-group.c      | 21 ++++++++++-----
 fs/btrfs/block-group.h      |  1 +
 fs/btrfs/extent-tree.c      | 10 ++++++-
 fs/btrfs/free-space-cache.c | 52 +++++++++++++++++++++++++++++++++++++
 fs/btrfs/free-space-cache.h |  2 ++
 fs/btrfs/space-info.c       | 13 ++++++----
 fs/btrfs/space-info.h       |  4 ++-
 fs/btrfs/sysfs.c            |  2 ++
 fs/btrfs/zoned.c            | 24 +++++++++++++++++
 fs/btrfs/zoned.h            |  3 +++
 10 files changed, 119 insertions(+), 13 deletions(-)
fs/btrfs/block-group.c +15 -6
index ffc64dfbe09e..723b7c183cd9 100644
@@ -1080,12 +1080,17 @@ int btrfs_remove_block_group(struct btrfs_trans_handle *trans,
 		WARN_ON(block_group->space_info->total_bytes
 			< block_group->length);
 		WARN_ON(block_group->space_info->bytes_readonly
-			< block_group->length);
+			< block_group->length - block_group->zone_unusable);
+		WARN_ON(block_group->space_info->bytes_zone_unusable
+			< block_group->zone_unusable);
 		WARN_ON(block_group->space_info->disk_total
 			< block_group->length * factor);
 	}
 	block_group->space_info->total_bytes -= block_group->length;
-	block_group->space_info->bytes_readonly -= block_group->length;
+	block_group->space_info->bytes_readonly -=
+		(block_group->length - block_group->zone_unusable);
+	block_group->space_info->bytes_zone_unusable -=
+		block_group->zone_unusable;
 	block_group->space_info->disk_total -= block_group->length * factor;
 
 	spin_unlock(&block_group->space_info->lock);
@@ -1229,7 +1234,7 @@ static int inc_block_group_ro(struct btrfs_block_group *cache, int force)
 	}
 
 	num_bytes = cache->length - cache->reserved - cache->pinned -
-		    cache->bytes_super - cache->used;
+		    cache->bytes_super - cache->zone_unusable - cache->used;
 
 	/*
 	 * Data never overcommits, even in mixed mode, so do just the straight
@@ -1973,6 +1978,8 @@ static int read_one_block_group(struct btrfs_fs_info *info,
 		btrfs_free_excluded_extents(cache);
 	}
 
+	btrfs_calc_zone_unusable(cache);
+
 	ret = btrfs_add_block_group_cache(info, cache);
 	if (ret) {
 		btrfs_remove_free_space_cache(cache);
@@ -1980,7 +1987,8 @@ static int read_one_block_group(struct btrfs_fs_info *info,
 	}
 	trace_btrfs_add_block_group(info, cache, 0);
 	btrfs_update_space_info(info, cache->flags, cache->length,
-				cache->used, cache->bytes_super, &space_info);
+				cache->used, cache->bytes_super,
+				cache->zone_unusable, &space_info);
 
 	cache->space_info = space_info;
 
@@ -2217,7 +2225,7 @@ int btrfs_make_block_group(struct btrfs_trans_handle *trans, u64 bytes_used,
 	 */
 	trace_btrfs_add_block_group(fs_info, cache, 1);
 	btrfs_update_space_info(fs_info, cache->flags, size, bytes_used,
-				cache->bytes_super, &cache->space_info);
+				cache->bytes_super, 0, &cache->space_info);
 	btrfs_update_global_block_rsv(fs_info);
 
 	link_block_group(cache);
@@ -2325,7 +2333,8 @@ void btrfs_dec_block_group_ro(struct btrfs_block_group *cache)
 	spin_lock(&cache->lock);
 	if (!--cache->ro) {
 		num_bytes = cache->length - cache->reserved -
-			    cache->pinned - cache->bytes_super - cache->used;
+			    cache->pinned - cache->bytes_super -
+			    cache->zone_unusable - cache->used;
 		sinfo->bytes_readonly -= num_bytes;
 		list_del_init(&cache->ro_list);
 	}
fs/btrfs/block-group.h +1 -0
index 14e3043c9ce7..5be47f4bfea7 100644
@@ -189,6 +189,7 @@ struct btrfs_block_group {
 	 * allocation. This is used only with ZONED mode enabled.
 	 */
 	u64 alloc_offset;
+	u64 zone_unusable;
 };
 
 static inline u64 btrfs_block_group_end(struct btrfs_block_group *block_group)
fs/btrfs/extent-tree.c +9 -1
index 3b21fee13e77..09439782b9a8 100644
@@ -34,6 +34,7 @@
 #include "block-group.h"
 #include "discard.h"
 #include "rcu-string.h"
+#include "zoned.h"
 
 #undef SCRAMBLE_DELAYED_REFS
 
@@ -2765,6 +2766,9 @@ fetch_cluster_info(struct btrfs_fs_info *fs_info,
 {
 	struct btrfs_free_cluster *ret = NULL;
 
+	if (btrfs_is_zoned(fs_info))
+		return NULL;
+
 	*empty_cluster = 0;
 	if (btrfs_mixed_space_info(space_info))
 		return ret;
@@ -2846,7 +2850,11 @@ static int unpin_extent_range(struct btrfs_fs_info *fs_info,
 		space_info->max_extent_size = 0;
 		percpu_counter_add_batch(&space_info->total_bytes_pinned,
 			    -len, BTRFS_TOTAL_BYTES_PINNED_BATCH);
-		if (cache->ro) {
+		if (btrfs_is_zoned(fs_info)) {
+			/* Need reset before reusing in a zoned block group */
+			space_info->bytes_zone_unusable += len;
+			readonly = true;
+		} else if (cache->ro) {
 			space_info->bytes_readonly += len;
 			readonly = true;
 		}
fs/btrfs/free-space-cache.c +52 -0

87 lines not on this page · raw message

fs/btrfs/free-space-cache.h +2 -0
index e3d5e0ad8f8e..469382529f7e 100644
@@ -116,6 +116,8 @@ int __btrfs_add_free_space(struct btrfs_fs_info *fs_info,
 			   enum btrfs_trim_state trim_state);
 int btrfs_add_free_space(struct btrfs_block_group *block_group,
 			 u64 bytenr, u64 size);
+int btrfs_add_free_space_unused(struct btrfs_block_group *block_group,
+				u64 bytenr, u64 size);
 int btrfs_add_free_space_async_trimmed(struct btrfs_block_group *block_group,
 				       u64 bytenr, u64 size);
 int btrfs_remove_free_space(struct btrfs_block_group *block_group,
fs/btrfs/space-info.c +8 -5
index 64099565ab8f..bbbf3c1412a4 100644
@@ -163,6 +163,7 @@ u64 __pure btrfs_space_info_used(struct btrfs_space_info *s_info,
 	ASSERT(s_info);
 	return s_info->bytes_used + s_info->bytes_reserved +
 		s_info->bytes_pinned + s_info->bytes_readonly +
+		s_info->bytes_zone_unusable +
 		(may_use_included ? s_info->bytes_may_use : 0);
 }
 
@@ -257,7 +258,7 @@ int btrfs_init_space_info(struct btrfs_fs_info *fs_info)
 
 void btrfs_update_space_info(struct btrfs_fs_info *info, u64 flags,
 			     u64 total_bytes, u64 bytes_used,
-			     u64 bytes_readonly,
+			     u64 bytes_readonly, u64 bytes_zone_unusable,
 			     struct btrfs_space_info **space_info)
 {
 	struct btrfs_space_info *found;
@@ -273,6 +274,7 @@ void btrfs_update_space_info(struct btrfs_fs_info *info, u64 flags,
 	found->bytes_used += bytes_used;
 	found->disk_used += bytes_used * factor;
 	found->bytes_readonly += bytes_readonly;
+	found->bytes_zone_unusable += bytes_zone_unusable;
 	if (total_bytes > 0)
 		found->full = 0;
 	btrfs_try_granting_tickets(info, found);
@@ -422,10 +424,10 @@ static void __btrfs_dump_space_info(struct btrfs_fs_info *fs_info,
 		   info->total_bytes - btrfs_space_info_used(info, true),
 		   info->full ? "" : "not ");
 	btrfs_info(fs_info,
-		"space_info total=%llu, used=%llu, pinned=%llu, reserved=%llu, may_use=%llu, readonly=%llu",
+		"space_info total=%llu, used=%llu, pinned=%llu, reserved=%llu, may_use=%llu, readonly=%llu zone_unusable=%llu",
 		info->total_bytes, info->bytes_used, info->bytes_pinned,
 		info->bytes_reserved, info->bytes_may_use,
-		info->bytes_readonly);
+		info->bytes_readonly, info->bytes_zone_unusable);
 
 	DUMP_BLOCK_RSV(fs_info, global_block_rsv);
 	DUMP_BLOCK_RSV(fs_info, trans_block_rsv);
@@ -454,9 +456,10 @@ void btrfs_dump_space_info(struct btrfs_fs_info *fs_info,
 	list_for_each_entry(cache, &info->block_groups[index], list) {
 		spin_lock(&cache->lock);
 		btrfs_info(fs_info,
-			"block group %llu has %llu bytes, %llu used %llu pinned %llu reserved %s",
+			"block group %llu has %llu bytes, %llu used %llu pinned %llu reserved %llu zone_unusable %s",
 			cache->start, cache->length, cache->used, cache->pinned,
-			cache->reserved, cache->ro ? "[readonly]" : "");
+			cache->reserved, cache->zone_unusable,
+			cache->ro ? "[readonly]" : "");
 		spin_unlock(&cache->lock);
 		btrfs_dump_free_space(cache, bytes);
 	}
fs/btrfs/space-info.h +3 -1
index 5646393b928c..ee003ffba956 100644
@@ -17,6 +17,8 @@ struct btrfs_space_info {
 	u64 bytes_may_use;	/* number of bytes that may be used for
 				   delalloc/allocations */
 	u64 bytes_readonly;	/* total bytes that are read only */
+	u64 bytes_zone_unusable;	/* total bytes that are unusable until
+					   resetting the device zone */
 
 	u64 max_extent_size;	/* This will hold the maximum extent size of
 				   the space info if we had an ENOSPC in the
@@ -119,7 +121,7 @@ DECLARE_SPACE_INFO_UPDATE(bytes_pinned, "pinned");
 int btrfs_init_space_info(struct btrfs_fs_info *fs_info);
 void btrfs_update_space_info(struct btrfs_fs_info *info, u64 flags,
 			     u64 total_bytes, u64 bytes_used,
-			     u64 bytes_readonly,
+			     u64 bytes_readonly, u64 bytes_zone_unusable,
 			     struct btrfs_space_info **space_info);
 struct btrfs_space_info *btrfs_find_space_info(struct btrfs_fs_info *info,
 					       u64 flags);
fs/btrfs/sysfs.c +2 -0
index 828006020bbd..ea679803da9b 100644
@@ -635,6 +635,7 @@ SPACE_INFO_ATTR(bytes_pinned);
 SPACE_INFO_ATTR(bytes_reserved);
 SPACE_INFO_ATTR(bytes_may_use);
 SPACE_INFO_ATTR(bytes_readonly);
+SPACE_INFO_ATTR(bytes_zone_unusable);
 SPACE_INFO_ATTR(disk_used);
 SPACE_INFO_ATTR(disk_total);
 BTRFS_ATTR(space_info, total_bytes_pinned,
@@ -648,6 +649,7 @@ static struct attribute *space_info_attrs[] = {
 	BTRFS_ATTR_PTR(space_info, bytes_reserved),
 	BTRFS_ATTR_PTR(space_info, bytes_may_use),
 	BTRFS_ATTR_PTR(space_info, bytes_readonly),
+	BTRFS_ATTR_PTR(space_info, bytes_zone_unusable),
 	BTRFS_ATTR_PTR(space_info, disk_used),
 	BTRFS_ATTR_PTR(space_info, disk_total),
 	BTRFS_ATTR_PTR(space_info, total_bytes_pinned),
fs/btrfs/zoned.c +24 -0
index 9bf40300e428..5ee26b9fe5b1 100644
@@ -999,3 +999,27 @@ int btrfs_load_block_group_zone_info(struct btrfs_block_group *cache)
 
 	return ret;
 }
+
+void btrfs_calc_zone_unusable(struct btrfs_block_group *cache)
+{
+	u64 unusable, free;
+
+	if (!btrfs_is_zoned(cache->fs_info))
+		return;
+
+	WARN_ON(cache->bytes_super != 0);
+	unusable = cache->alloc_offset - cache->used;
+	free = cache->length - cache->alloc_offset;
+
+	/* We only need ->free_space in ALLOC_SEQ BGs */
+	cache->last_byte_to_unpin = (u64)-1;
+	cache->cached = BTRFS_CACHE_FINISHED;
+	cache->free_space_ctl->free_space = free;
+	cache->zone_unusable = unusable;
+
+	/*
+	 * Should not have any excluded extents. Just
+	 * in case, though.
+	 */
+	btrfs_free_excluded_extents(cache);
+}
fs/btrfs/zoned.h +3 -0
index e3338a2f1be9..c86cde1978cd 100644
@@ -41,6 +41,7 @@ int btrfs_reset_device_zone(struct btrfs_device *device, u64 physical,
 			    u64 length, u64 *bytes);
 int btrfs_ensure_empty_zones(struct btrfs_device *device, u64 start, u64 size);
 int btrfs_load_block_group_zone_info(struct btrfs_block_group *cache);
+void btrfs_calc_zone_unusable(struct btrfs_block_group *cache);
 #else /* CONFIG_BLK_DEV_ZONED */
 static inline int btrfs_get_dev_zone(struct btrfs_device *device, u64 pos,
 				     struct blk_zone *zone)
@@ -119,6 +120,8 @@ static inline int btrfs_load_block_group_zone_info(
 	return 0;
 }
 
+static inline void btrfs_calc_zone_unusable(struct btrfs_block_group *cache) { }
+
 #endif
 
 static inline bool btrfs_dev_is_sequential(struct btrfs_device *device, u64 pos)
Signature · 1 line
2.27.0
#85 lore raw

PATCH v10 17/41 btrfs: do sequential extent allocation in ZONED mode

In reply to Naohiro Aota

This commit implements a sequential extent allocator for the ZONED mode. This allocator just needs to check if there is enough space in the block group. Therefor the allocator never manages bitmaps or clusters. Also add ASSERTs to the corresponding functions.

Actually, with zone append writing, it is unnecessary to track the allocation offset. It only needs to check space availability. But, by tracking the offset and returning the offset as an allocated region, we can skip modification of ordered extents and checksum information when there is no IO reordering.

---
 fs/btrfs/block-group.c      |  4 ++
 fs/btrfs/extent-tree.c      | 85 ++++++++++++++++++++++++++++++++++---
 fs/btrfs/free-space-cache.c |  6 +++
 3 files changed, 89 insertions(+), 6 deletions(-)
fs/btrfs/block-group.c +4 -0
index 723b7c183cd9..232885261c37 100644
@@ -683,6 +683,10 @@ int btrfs_cache_block_group(struct btrfs_block_group *cache, int load_cache_only
 	struct btrfs_caching_control *caching_ctl;
 	int ret = 0;
 
+	/* Allocator for ZONED btrfs does not use the cache at all */
+	if (btrfs_is_zoned(fs_info))
+		return 0;
+
 	caching_ctl = kzalloc(sizeof(*caching_ctl), GFP_NOFS);
 	if (!caching_ctl)
 		return -ENOMEM;
fs/btrfs/extent-tree.c +79 -6

147 lines not on this page · raw message

fs/btrfs/free-space-cache.c +6 -0
index f6434794cb0b..2161d0ad5cf0 100644
@@ -2903,6 +2903,8 @@ u64 btrfs_find_space_for_alloc(struct btrfs_block_group *block_group,
 	u64 align_gap_len = 0;
 	enum btrfs_trim_state align_gap_trim_state = BTRFS_TRIM_STATE_UNTRIMMED;
 
+	ASSERT(!btrfs_is_zoned(block_group->fs_info));
+
 	spin_lock(&ctl->tree_lock);
 	entry = find_free_space(ctl, &offset, &bytes_search,
 				block_group->full_stripe_len, max_extent_size);
@@ -3034,6 +3036,8 @@ u64 btrfs_alloc_from_cluster(struct btrfs_block_group *block_group,
 	struct rb_node *node;
 	u64 ret = 0;
 
+	ASSERT(!btrfs_is_zoned(block_group->fs_info));
+
 	spin_lock(&cluster->lock);
 	if (bytes > cluster->max_size)
 		goto out;
@@ -3810,6 +3814,8 @@ int btrfs_trim_block_group(struct btrfs_block_group *block_group,
 	int ret;
 	u64 rem = 0;
 
+	ASSERT(!btrfs_is_zoned(block_group->fs_info));
+
 	*trimmed = 0;
 
 	spin_lock(&block_group->lock);
Signature · 1 line
2.27.0
#86 lore raw

PATCH v10 18/41 btrfs: reset zones of unused block groups

In reply to Naohiro Aota

For an ZONED volume, a block group maps to a zone of the device. For deleted unused block groups, the zone of the block group can be reset to rewind the zone write pointer at the start of the zone.

---
 fs/btrfs/block-group.c |  8 ++++++--
 fs/btrfs/extent-tree.c | 17 ++++++++++++-----
 fs/btrfs/zoned.h       | 16 ++++++++++++++++
 3 files changed, 34 insertions(+), 7 deletions(-)
fs/btrfs/block-group.c +6 -2
index 232885261c37..31511e59ca74 100644
@@ -1470,8 +1470,12 @@ void btrfs_delete_unused_bgs(struct btrfs_fs_info *fs_info)
 		if (!async_trim_enabled && btrfs_test_opt(fs_info, DISCARD_ASYNC))
 			goto flip_async;
 
-		/* DISCARD can flip during remount */
-		trimming = btrfs_test_opt(fs_info, DISCARD_SYNC);
+		/*
+		 * DISCARD can flip during remount. In ZONED mode, we need
+		 * to reset sequential required zones.
+		 */
+		trimming = btrfs_test_opt(fs_info, DISCARD_SYNC) ||
+				btrfs_is_zoned(fs_info);
 
 		/* Implicit trim during transaction commit. */
 		if (trimming)
fs/btrfs/extent-tree.c +12 -5
index ab0ce3ba2b89..11e6483372c3 100644
@@ -1331,6 +1331,9 @@ int btrfs_discard_extent(struct btrfs_fs_info *fs_info, u64 bytenr,
 
 		stripe = bbio->stripes;
 		for (i = 0; i < bbio->num_stripes; i++, stripe++) {
+			struct btrfs_device *dev = stripe->dev;
+			u64 physical = stripe->physical;
+			u64 length = stripe->length;
 			u64 bytes;
 			struct request_queue *req_q;
 
@@ -1338,14 +1341,18 @@ int btrfs_discard_extent(struct btrfs_fs_info *fs_info, u64 bytenr,
 				ASSERT(btrfs_test_opt(fs_info, DEGRADED));
 				continue;
 			}
+
 			req_q = bdev_get_queue(stripe->dev->bdev);
-			if (!blk_queue_discard(req_q))
+			/* Zone reset in ZONED mode */
+			if (btrfs_can_zone_reset(dev, physical, length))
+				ret = btrfs_reset_device_zone(dev, physical,
+							      length, &bytes);
+			else if (blk_queue_discard(req_q))
+				ret = btrfs_issue_discard(dev->bdev, physical,
+							  length, &bytes);
+			else
 				continue;
 
-			ret = btrfs_issue_discard(stripe->dev->bdev,
-						  stripe->physical,
-						  stripe->length,
-						  &bytes);
 			if (!ret) {
 				discarded_bytes += bytes;
 			} else if (ret != -EOPNOTSUPP) {
fs/btrfs/zoned.h +16 -0
index c86cde1978cd..6a07af0c7f6d 100644
@@ -209,4 +209,20 @@ static inline u64 btrfs_align_offset_to_zone(struct btrfs_device *device,
 	return ALIGN(pos, device->zone_info->zone_size);
 }
 
+static inline bool btrfs_can_zone_reset(struct btrfs_device *device,
+					u64 physical, u64 length)
+{
+	u64 zone_size;
+
+	if (!btrfs_dev_is_sequential(device, physical))
+		return false;
+
+	zone_size = device->zone_info->zone_size;
+	if (!IS_ALIGNED(physical, zone_size) ||
+	    !IS_ALIGNED(length, zone_size))
+		return false;
+
+	return true;
+}
+
 #endif
Signature · 1 line
2.27.0
#87 lore raw

PATCH v10 19/41 btrfs: redirty released extent buffers in ZONED mode

In reply to Naohiro Aota

Tree manipulating operations like merging nodes often release once-allocated tree nodes. Btrfs cleans such nodes so that pages in the node are not uselessly written out. On ZONED volumes, however, such optimization blocks the following IOs as the cancellation of the write out of the freed blocks breaks the sequential write sequence expected by the device.

This patch introduces a list of clean and unwritten extent buffers that have been released in a transaction. Btrfs redirty the buffer so that btree_write_cache_pages() can send proper bios to the devices.

Besides it clears the entire content of the extent buffer not to confuse raw block scanners e.g. btrfsck. By clearing the content, csum_dirty_buffer() complains about bytenr mismatch, so avoid the checking and checksum using newly introduced buffer flag EXTENT_BUFFER_NO_CHECK.

---
 fs/btrfs/disk-io.c     |  8 ++++++++
 fs/btrfs/extent-tree.c | 12 +++++++++++-
 fs/btrfs/extent_io.c   |  4 ++++
 fs/btrfs/extent_io.h   |  2 ++
 fs/btrfs/transaction.c | 10 ++++++++++
 fs/btrfs/transaction.h |  3 +++
 fs/btrfs/tree-log.c    |  6 ++++++
 fs/btrfs/zoned.c       | 37 +++++++++++++++++++++++++++++++++++++
 fs/btrfs/zoned.h       |  7 +++++++
 9 files changed, 88 insertions(+), 1 deletion(-)
fs/btrfs/disk-io.c +8 -0
index 509085a368bb..c0180fbd5c78 100644
@@ -462,6 +462,12 @@ static int csum_dirty_buffer(struct btrfs_fs_info *fs_info, struct page *page)
 		return 0;
 
 	found_start = btrfs_header_bytenr(eb);
+
+	if (test_bit(EXTENT_BUFFER_NO_CHECK, &eb->bflags)) {
+		WARN_ON(found_start != 0);
+		return 0;
+	}
+
 	/*
 	 * Please do not consolidate these warnings into a single if.
 	 * It is useful to know what went wrong.
@@ -4616,6 +4622,8 @@ void btrfs_cleanup_one_transaction(struct btrfs_transaction *cur_trans,
 				     EXTENT_DIRTY);
 	btrfs_destroy_pinned_extent(fs_info, &cur_trans->pinned_extents);
 
+	btrfs_free_redirty_list(cur_trans);
+
 	cur_trans->state =TRANS_STATE_COMPLETED;
 	wake_up(&cur_trans->commit_wait);
 }
fs/btrfs/extent-tree.c +11 -1
index 11e6483372c3..99640dacf8e6 100644
@@ -3422,8 +3422,10 @@ void btrfs_free_tree_block(struct btrfs_trans_handle *trans,
 
 		if (root->root_key.objectid != BTRFS_TREE_LOG_OBJECTID) {
 			ret = check_ref_cleanup(trans, buf->start);
-			if (!ret)
+			if (!ret) {
+				btrfs_redirty_list_add(trans->transaction, buf);
 				goto out;
+			}
 		}
 
 		pin = 0;
@@ -3435,6 +3437,13 @@ void btrfs_free_tree_block(struct btrfs_trans_handle *trans,
 			goto out;
 		}
 
+		if (btrfs_is_zoned(fs_info)) {
+			btrfs_redirty_list_add(trans->transaction, buf);
+			pin_down_extent(trans, cache, buf->start, buf->len, 1);
+			btrfs_put_block_group(cache);
+			goto out;
+		}
+
 		WARN_ON(test_bit(EXTENT_BUFFER_DIRTY, &buf->bflags));
 
 		btrfs_add_free_space(cache, buf->start, buf->len);
@@ -4771,6 +4780,7 @@ btrfs_init_new_buffer(struct btrfs_trans_handle *trans, struct btrfs_root *root,
 	__btrfs_tree_lock(buf, nest);
 	btrfs_clean_tree_block(buf);
 	clear_bit(EXTENT_BUFFER_STALE, &buf->bflags);
+	clear_bit(EXTENT_BUFFER_NO_CHECK, &buf->bflags);
 
 	btrfs_set_lock_blocking_write(buf);
 	set_extent_buffer_uptodate(buf);
fs/btrfs/extent_io.c +4 -0
index 60f5f68d892d..e91c504fe973 100644
@@ -24,6 +24,7 @@
 #include "rcu-string.h"
 #include "backref.h"
 #include "disk-io.h"
+#include "zoned.h"
 
 static struct kmem_cache *extent_state_cache;
 static struct kmem_cache *extent_buffer_cache;
@@ -4959,6 +4960,7 @@ __alloc_extent_buffer(struct btrfs_fs_info *fs_info, u64 start,
 
 	btrfs_leak_debug_add(&fs_info->eb_leak_lock, &eb->leak_list,
 			     &fs_info->allocated_ebs);
+	INIT_LIST_HEAD(&eb->release_list);
 
 	spin_lock_init(&eb->refs_lock);
 	atomic_set(&eb->refs, 1);
@@ -5744,6 +5746,8 @@ void write_extent_buffer(const struct extent_buffer *eb, const void *srcv,
 	char *src = (char *)srcv;
 	unsigned long i = start >> PAGE_SHIFT;
 
+	WARN_ON(test_bit(EXTENT_BUFFER_NO_CHECK, &eb->bflags));
+
 	if (check_eb_range(eb, start, len))
 		return;
 
fs/btrfs/extent_io.h +2 -0
index f39d02e7f7ef..5f2ccfd0205e 100644
@@ -30,6 +30,7 @@ enum {
 	EXTENT_BUFFER_IN_TREE,
 	/* write IO error */
 	EXTENT_BUFFER_WRITE_ERR,
+	EXTENT_BUFFER_NO_CHECK,
 };
 
 /* these are flags for __process_pages_contig */
@@ -107,6 +108,7 @@ struct extent_buffer {
 	 */
 	wait_queue_head_t read_lock_wq;
 	struct page *pages[INLINE_EXTENT_BUFFER_PAGES];
+	struct list_head release_list;
 #ifdef CONFIG_BTRFS_DEBUG
 	int spinning_writers;
 	atomic_t spinning_readers;
fs/btrfs/transaction.c +10 -0
index 52ada47aff50..a8561536cd0d 100644
@@ -22,6 +22,7 @@
 #include "qgroup.h"
 #include "block-group.h"
 #include "space-info.h"
+#include "zoned.h"
 
 #define BTRFS_ROOT_TRANS_TAG 0
 
@@ -336,6 +337,8 @@ static noinline int join_transaction(struct btrfs_fs_info *fs_info,
 	spin_lock_init(&cur_trans->dirty_bgs_lock);
 	INIT_LIST_HEAD(&cur_trans->deleted_bgs);
 	spin_lock_init(&cur_trans->dropped_roots_lock);
+	INIT_LIST_HEAD(&cur_trans->releasing_ebs);
+	spin_lock_init(&cur_trans->releasing_ebs_lock);
 	list_add_tail(&cur_trans->list, &fs_info->trans_list);
 	extent_io_tree_init(fs_info, &cur_trans->dirty_pages,
 			IO_TREE_TRANS_DIRTY_PAGES, fs_info->btree_inode);
@@ -2345,6 +2348,13 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans)
 		goto scrub_continue;
 	}
 
+	/*
+	 * At this point, we should have written all the tree blocks
+	 * allocated in this transaction. So it's now safe to free the
+	 * redirtyied extent buffers.
+	 */
+	btrfs_free_redirty_list(cur_trans);
+
 	ret = write_all_supers(fs_info, 0);
 	/*
 	 * the super is written, we can safely allow the tree-loggers
fs/btrfs/transaction.h +3 -0
index 858d9153a1cd..380e0aaa15b3 100644
@@ -92,6 +92,9 @@ struct btrfs_transaction {
 	 */
 	atomic_t pending_ordered;
 	wait_queue_head_t pending_wait;
+
+	spinlock_t releasing_ebs_lock;
+	struct list_head releasing_ebs;
 };
 
 #define __TRANS_FREEZABLE	(1U << 0)
fs/btrfs/tree-log.c +6 -0
index 56cbc1706b6f..5f585cf57383 100644
@@ -20,6 +20,7 @@
 #include "inode-map.h"
 #include "block-group.h"
 #include "space-info.h"
+#include "zoned.h"
 
 /* magic values for the inode_only field in btrfs_log_inode:
  *
@@ -2742,6 +2743,8 @@ static noinline int walk_down_log_tree(struct btrfs_trans_handle *trans,
 						free_extent_buffer(next);
 						return ret;
 					}
+					btrfs_redirty_list_add(
+						trans->transaction, next);
 				} else {
 					if (test_and_clear_bit(EXTENT_BUFFER_DIRTY, &next->bflags))
 						clear_extent_buffer_dirty(next);
@@ -3277,6 +3280,9 @@ static void free_log_tree(struct btrfs_trans_handle *trans,
 	clear_extent_bits(&log->dirty_log_pages, 0, (u64)-1,
 			  EXTENT_DIRTY | EXTENT_NEW | EXTENT_NEED_WAIT);
 	extent_io_tree_release(&log->log_csum_range);
+
+	if (trans && log->node)
+		btrfs_redirty_list_add(trans->transaction, log->node);
 	btrfs_put_root(log);
 }
 
fs/btrfs/zoned.c +37 -0
index 5ee26b9fe5b1..b56bfeaf8744 100644
@@ -10,6 +10,7 @@
 #include "rcu-string.h"
 #include "disk-io.h"
 #include "block-group.h"
+#include "transaction.h"
 
 /* Maximum number of zones to report per blkdev_report_zones() call */
 #define BTRFS_REPORT_NR_ZONES   4096
@@ -1023,3 +1024,39 @@ void btrfs_calc_zone_unusable(struct btrfs_block_group *cache)
 	 */
 	btrfs_free_excluded_extents(cache);
 }
+
+void btrfs_redirty_list_add(struct btrfs_transaction *trans,
+			    struct extent_buffer *eb)
+{
+	struct btrfs_fs_info *fs_info = eb->fs_info;
+
+	if (!btrfs_is_zoned(fs_info) ||
+	    btrfs_header_flag(eb, BTRFS_HEADER_FLAG_WRITTEN) ||
+	    !list_empty(&eb->release_list))
+		return;
+
+	set_extent_buffer_dirty(eb);
+	set_extent_bits_nowait(&trans->dirty_pages, eb->start,
+			       eb->start + eb->len - 1, EXTENT_DIRTY);
+	memzero_extent_buffer(eb, 0, eb->len);
+	set_bit(EXTENT_BUFFER_NO_CHECK, &eb->bflags);
+
+	spin_lock(&trans->releasing_ebs_lock);
+	list_add_tail(&eb->release_list, &trans->releasing_ebs);
+	spin_unlock(&trans->releasing_ebs_lock);
+	atomic_inc(&eb->refs);
+}
+
+void btrfs_free_redirty_list(struct btrfs_transaction *trans)
+{
+	spin_lock(&trans->releasing_ebs_lock);
+	while (!list_empty(&trans->releasing_ebs)) {
+		struct extent_buffer *eb;
+
+		eb = list_first_entry(&trans->releasing_ebs,
+				      struct extent_buffer, release_list);
+		list_del_init(&eb->release_list);
+		free_extent_buffer(eb);
+	}
+	spin_unlock(&trans->releasing_ebs_lock);
+}
fs/btrfs/zoned.h +7 -0
index 6a07af0c7f6d..a7de80c313be 100644
@@ -42,6 +42,9 @@ int btrfs_reset_device_zone(struct btrfs_device *device, u64 physical,
 int btrfs_ensure_empty_zones(struct btrfs_device *device, u64 start, u64 size);
 int btrfs_load_block_group_zone_info(struct btrfs_block_group *cache);
 void btrfs_calc_zone_unusable(struct btrfs_block_group *cache);
+void btrfs_redirty_list_add(struct btrfs_transaction *trans,
+			    struct extent_buffer *eb);
+void btrfs_free_redirty_list(struct btrfs_transaction *trans);
 #else /* CONFIG_BLK_DEV_ZONED */
 static inline int btrfs_get_dev_zone(struct btrfs_device *device, u64 pos,
 				     struct blk_zone *zone)
@@ -122,6 +125,10 @@ static inline int btrfs_load_block_group_zone_info(
 
 static inline void btrfs_calc_zone_unusable(struct btrfs_block_group *cache) { }
 
+static inline void btrfs_redirty_list_add(struct btrfs_transaction *trans,
+					  struct extent_buffer *eb) { }
+static inline void btrfs_free_redirty_list(struct btrfs_transaction *trans) { }
+
 #endif
 
 static inline bool btrfs_dev_is_sequential(struct btrfs_device *device, u64 pos)
Signature · 1 line
2.27.0
#88 lore raw

PATCH v10 20/41 btrfs: extract page adding function

In reply to Naohiro Aota

This commit extract page adding to bio part from submit_extent_page(). The page is added only when bio_flags are the same, contiguous and the added page fits in the same stripe as pages in the bio.

Condition checkings are reordered to allow early return to avoid possibly heavy btrfs_bio_fits_in_stripe() calling.

---
 fs/btrfs/extent_io.c | 56 ++++++++++++++++++++++++++++++++------------
 1 file changed, 41 insertions(+), 15 deletions(-)
fs/btrfs/extent_io.c +41 -15
index e91c504fe973..868ae0874a34 100644
@@ -3012,6 +3012,44 @@ struct bio *btrfs_bio_clone_partial(struct bio *orig, int offset, int size)
 	return bio;
 }
 
+/**
+ * btrfs_bio_add_page	-	attempt to add a page to bio
+ * @bio:	destination bio
+ * @page:	page to add to the bio
+ * @logical:	offset of the new bio or to check whether we are adding
+ *              a contiguous page to the previous one
+ * @pg_offset:	starting offset in the page
+ * @size:	portion of page that we want to write
+ * @prev_bio_flags:  flags of previous bio to see if we can merge the current one
+ * @bio_flags:	flags of the current bio to see if we can merge them
+ *
+ * Attempt to add a page to bio considering stripe alignment etc. Return
+ * true if successfully page added. Otherwise, return false.
+ */
+static bool btrfs_bio_add_page(struct bio *bio, struct page *page, u64 logical,
+			       unsigned int size, unsigned int pg_offset,
+			       unsigned long prev_bio_flags,
+			       unsigned long bio_flags)
+{
+	sector_t sector = logical >> SECTOR_SHIFT;
+	bool contig;
+
+	if (prev_bio_flags != bio_flags)
+		return false;
+
+	if (prev_bio_flags & EXTENT_BIO_COMPRESSED)
+		contig = bio->bi_iter.bi_sector == sector;
+	else
+		contig = bio_end_sector(bio) == sector;
+	if (!contig)
+		return false;
+
+	if (btrfs_bio_fits_in_stripe(page, size, bio, bio_flags))
+		return false;
+
+	return bio_add_page(bio, page, size, pg_offset) == size;
+}
+
 /*
  * @opf:	bio REQ_OP_* and REQ_* flags as one value
  * @wbc:	optional writeback control for io accounting
@@ -3040,27 +3078,15 @@ static int submit_extent_page(unsigned int opf,
 	int ret = 0;
 	struct bio *bio;
 	size_t page_size = min_t(size_t, size, PAGE_SIZE);
-	sector_t sector = offset >> 9;
 	struct extent_io_tree *tree = &BTRFS_I(page->mapping->host)->io_tree;
 
 	ASSERT(bio_ret);
 
 	if (*bio_ret) {
-		bool contig;
-		bool can_merge = true;
-
 		bio = *bio_ret;
-		if (prev_bio_flags & EXTENT_BIO_COMPRESSED)
-			contig = bio->bi_iter.bi_sector == sector;
-		else
-			contig = bio_end_sector(bio) == sector;
-
-		if (btrfs_bio_fits_in_stripe(page, page_size, bio, bio_flags))
-			can_merge = false;
-
-		if (prev_bio_flags != bio_flags || !contig || !can_merge ||
-		    force_bio_submit ||
-		    bio_add_page(bio, page, page_size, pg_offset) < page_size) {
+		if (force_bio_submit ||
+		    !btrfs_bio_add_page(bio, page, offset, page_size, pg_offset,
+					prev_bio_flags, bio_flags)) {
 			ret = submit_one_bio(bio, mirror_num, prev_bio_flags);
 			if (ret < 0) {
 				*bio_ret = NULL;
Signature · 1 line
2.27.0
#89 lore raw

PATCH v10 21/41 btrfs: use bio_add_zone_append_page for zoned btrfs

In reply to Naohiro Aota

Zoned device has its own hardware restrictions e.g. max_zone_append_size when using REQ_OP_ZONE_APPEND. To follow the restrictions, use bio_add_zone_append_page() instead of bio_add_page(). We need target device to use bio_add_zone_append_page(), so this commit reads the chunk information to memoize the target device to btrfs_io_bio(bio)->device.

Currently, zoned btrfs only supports SINGLE profile. In the feature, btrfs_io_bio can hold extent_map and check the restrictions for all the devices the bio will be mapped.

---
 fs/btrfs/extent_io.c | 30 +++++++++++++++++++++++++++---
 1 file changed, 27 insertions(+), 3 deletions(-)
fs/btrfs/extent_io.c +27 -3
index 868ae0874a34..b9b366f4d942 100644
@@ -3033,6 +3033,7 @@ static bool btrfs_bio_add_page(struct bio *bio, struct page *page, u64 logical,
 {
 	sector_t sector = logical >> SECTOR_SHIFT;
 	bool contig;
+	int ret;
 
 	if (prev_bio_flags != bio_flags)
 		return false;
@@ -3047,7 +3048,12 @@ static bool btrfs_bio_add_page(struct bio *bio, struct page *page, u64 logical,
 	if (btrfs_bio_fits_in_stripe(page, size, bio, bio_flags))
 		return false;
 
-	return bio_add_page(bio, page, size, pg_offset) == size;
+	if (bio_op(bio) == REQ_OP_ZONE_APPEND)
+		ret = bio_add_zone_append_page(bio, page, size, pg_offset);
+	else
+		ret = bio_add_page(bio, page, size, pg_offset);
+
+	return ret == size;
 }
 
 /*
@@ -3078,7 +3084,9 @@ static int submit_extent_page(unsigned int opf,
 	int ret = 0;
 	struct bio *bio;
 	size_t page_size = min_t(size_t, size, PAGE_SIZE);
-	struct extent_io_tree *tree = &BTRFS_I(page->mapping->host)->io_tree;
+	struct btrfs_inode *inode = BTRFS_I(page->mapping->host);
+	struct extent_io_tree *tree = &inode->io_tree;
+	struct btrfs_fs_info *fs_info = inode->root->fs_info;
 
 	ASSERT(bio_ret);
 
@@ -3109,11 +3117,27 @@ static int submit_extent_page(unsigned int opf,
 	if (wbc) {
 		struct block_device *bdev;
 
-		bdev = BTRFS_I(page->mapping->host)->root->fs_info->fs_devices->latest_bdev;
+		bdev = fs_info->fs_devices->latest_bdev;
 		bio_set_dev(bio, bdev);
 		wbc_init_bio(wbc, bio);
 		wbc_account_cgroup_owner(wbc, page, page_size);
 	}
+	if (btrfs_is_zoned(fs_info) &&
+	    bio_op(bio) == REQ_OP_ZONE_APPEND) {
+		struct extent_map *em;
+		struct map_lookup *map;
+
+		em = btrfs_get_chunk_map(fs_info, offset, page_size);
+		if (IS_ERR(em))
+			return PTR_ERR(em);
+
+		map = em->map_lookup;
+		/* We only support SINGLE profile for now */
+		ASSERT(map->num_stripes == 1);
+		btrfs_io_bio(bio)->device = map->stripes[0].dev;
+
+		free_extent_map(em);
+	}
 
 	*bio_ret = bio;
 
Signature · 1 line
2.27.0
#90 lore raw

PATCH v10 22/41 btrfs: handle REQ_OP_ZONE_APPEND as writing

In reply to Naohiro Aota

ZONED btrfs uses REQ_OP_ZONE_APPEND bios for writing to actual devices. Let btrfs_end_bio() and btrfs_op be aware of it.

---
 fs/btrfs/disk-io.c |  4 ++--
 fs/btrfs/inode.c   | 10 +++++-----
 fs/btrfs/volumes.c |  8 ++++----
 fs/btrfs/volumes.h |  1 +
 4 files changed, 12 insertions(+), 11 deletions(-)
fs/btrfs/disk-io.c +2 -2
index c0180fbd5c78..8acf1ed75889 100644
@@ -652,7 +652,7 @@ static void end_workqueue_bio(struct bio *bio)
 	fs_info = end_io_wq->info;
 	end_io_wq->status = bio->bi_status;
 
-	if (bio_op(bio) == REQ_OP_WRITE) {
+	if (btrfs_op(bio) == BTRFS_MAP_WRITE) {
 		if (end_io_wq->metadata == BTRFS_WQ_ENDIO_METADATA)
 			wq = fs_info->endio_meta_write_workers;
 		else if (end_io_wq->metadata == BTRFS_WQ_ENDIO_FREE_SPACE)
@@ -827,7 +827,7 @@ blk_status_t btrfs_submit_metadata_bio(struct inode *inode, struct bio *bio,
 	int async = check_async_write(fs_info, BTRFS_I(inode));
 	blk_status_t ret;
 
-	if (bio_op(bio) != REQ_OP_WRITE) {
+	if (btrfs_op(bio) != BTRFS_MAP_WRITE) {
 		/*
 		 * called for a read, do the setup so that checksum validation
 		 * can happen in the async kernel threads
fs/btrfs/inode.c +5 -5
index 936c3137c646..591ca539e444 100644
@@ -2192,7 +2192,7 @@ blk_status_t btrfs_submit_data_bio(struct inode *inode, struct bio *bio,
 	if (btrfs_is_free_space_inode(BTRFS_I(inode)))
 		metadata = BTRFS_WQ_ENDIO_FREE_SPACE;
 
-	if (bio_op(bio) != REQ_OP_WRITE) {
+	if (btrfs_op(bio) != BTRFS_MAP_WRITE) {
 		ret = btrfs_bio_wq_end_io(fs_info, bio, metadata);
 		if (ret)
 			goto out;
@@ -7526,7 +7526,7 @@ static void btrfs_dio_private_put(struct btrfs_dio_private *dip)
 	if (!refcount_dec_and_test(&dip->refs))
 		return;
 
-	if (bio_op(dip->dio_bio) == REQ_OP_WRITE) {
+	if (btrfs_op(dip->dio_bio) == BTRFS_MAP_WRITE) {
 		__endio_write_update_ordered(BTRFS_I(dip->inode),
 					     dip->logical_offset,
 					     dip->bytes,
@@ -7695,7 +7695,7 @@ static inline blk_status_t btrfs_submit_dio_bio(struct bio *bio,
 {
 	struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb);
 	struct btrfs_dio_private *dip = bio->bi_private;
-	bool write = bio_op(bio) == REQ_OP_WRITE;
+	bool write = btrfs_op(bio) == BTRFS_MAP_WRITE;
 	blk_status_t ret;
 
 	/* Check btrfs_submit_bio_hook() for rules about async submit. */
@@ -7746,7 +7746,7 @@ static struct btrfs_dio_private *btrfs_create_dio_private(struct bio *dio_bio,
 							  struct inode *inode,
 							  loff_t file_offset)
 {
-	const bool write = (bio_op(dio_bio) == REQ_OP_WRITE);
+	const bool write = (btrfs_op(dio_bio) == BTRFS_MAP_WRITE);
 	const bool csum = !(BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM);
 	size_t dip_size;
 	struct btrfs_dio_private *dip;
@@ -7777,7 +7777,7 @@ static struct btrfs_dio_private *btrfs_create_dio_private(struct bio *dio_bio,
 static blk_qc_t btrfs_submit_direct(struct inode *inode, struct iomap *iomap,
 		struct bio *dio_bio, loff_t file_offset)
 {
-	const bool write = (bio_op(dio_bio) == REQ_OP_WRITE);
+	const bool write = (btrfs_op(dio_bio) == BTRFS_MAP_WRITE);
 	const bool csum = !(BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM);
 	struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb);
 	const bool raid56 = (btrfs_data_alloc_profile(fs_info) &
fs/btrfs/volumes.c +4 -4
index c0e27c1e2559..683b3ed06226 100644
@@ -6451,7 +6451,7 @@ static void btrfs_end_bio(struct bio *bio)
 			struct btrfs_device *dev = btrfs_io_bio(bio)->device;
 
 			ASSERT(dev->bdev);
-			if (bio_op(bio) == REQ_OP_WRITE)
+			if (btrfs_op(bio) == BTRFS_MAP_WRITE)
 				btrfs_dev_stat_inc_and_print(dev,
 						BTRFS_DEV_STAT_WRITE_ERRS);
 			else if (!(bio->bi_opf & REQ_RAHEAD))
@@ -6564,10 +6564,10 @@ blk_status_t btrfs_map_bio(struct btrfs_fs_info *fs_info, struct bio *bio,
 	atomic_set(&bbio->stripes_pending, bbio->num_stripes);
 
 	if ((bbio->map_type & BTRFS_BLOCK_GROUP_RAID56_MASK) &&
-	    ((bio_op(bio) == REQ_OP_WRITE) || (mirror_num > 1))) {
+	    ((btrfs_op(bio) == BTRFS_MAP_WRITE) || (mirror_num > 1))) {
 		/* In this case, map_length has been set to the length of
 		   a single stripe; not the whole write */
-		if (bio_op(bio) == REQ_OP_WRITE) {
+		if (btrfs_op(bio) == BTRFS_MAP_WRITE) {
 			ret = raid56_parity_write(fs_info, bio, bbio,
 						  map_length);
 		} else {
@@ -6590,7 +6590,7 @@ blk_status_t btrfs_map_bio(struct btrfs_fs_info *fs_info, struct bio *bio,
 		dev = bbio->stripes[dev_nr].dev;
 		if (!dev || !dev->bdev || test_bit(BTRFS_DEV_STATE_MISSING,
 						   &dev->dev_state) ||
-		    (bio_op(first_bio) == REQ_OP_WRITE &&
+		    (btrfs_op(first_bio) == BTRFS_MAP_WRITE &&
 		    !test_bit(BTRFS_DEV_STATE_WRITEABLE, &dev->dev_state))) {
 			bbio_error(bbio, first_bio, logical);
 			continue;
fs/btrfs/volumes.h +1 -0
index 0249aca668fb..cff1f7689eac 100644
@@ -410,6 +410,7 @@ static inline enum btrfs_map_op btrfs_op(struct bio *bio)
 	case REQ_OP_DISCARD:
 		return BTRFS_MAP_DISCARD;
 	case REQ_OP_WRITE:
+	case REQ_OP_ZONE_APPEND:
 		return BTRFS_MAP_WRITE;
 	default:
 		WARN_ON_ONCE(1);
Signature · 1 line
2.27.0
#91 lore raw

PATCH v10 23/41 btrfs: split ordered extent when bio is sent

In reply to Naohiro Aota

For a zone append write, the device decides the location the data is written to. Therefore we cannot ensure that two bios are written consecutively on the device. In order to ensure that a ordered extent maps to a contiguous region on disk, we need to maintain a "one bio == one ordered extent" rule.

This commit implements the splitting of an ordered extent and extent map on bio submission to adhere to the rule.

---
 fs/btrfs/inode.c        | 89 +++++++++++++++++++++++++++++++++++++++++
 fs/btrfs/ordered-data.c | 76 +++++++++++++++++++++++++++++++++++
 fs/btrfs/ordered-data.h |  2 +
 3 files changed, 167 insertions(+)
fs/btrfs/inode.c +89 -0
index 591ca539e444..df85d8dea37c 100644
@@ -2158,6 +2158,86 @@ static blk_status_t btrfs_submit_bio_start(void *private_data, struct bio *bio,
 	return btrfs_csum_one_bio(BTRFS_I(inode), bio, 0, 0);
 }
 
+int extract_ordered_extent(struct inode *inode, struct bio *bio,
+			   loff_t file_offset)
+{
+	struct btrfs_ordered_extent *ordered;
+	struct extent_map *em = NULL, *em_new = NULL;
+	struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree;
+	u64 start = (u64)bio->bi_iter.bi_sector << SECTOR_SHIFT;
+	u64 len = bio->bi_iter.bi_size;
+	u64 end = start + len;
+	u64 ordered_end;
+	u64 pre, post;
+	int ret = 0;
+
+	ordered = btrfs_lookup_ordered_extent(BTRFS_I(inode), file_offset);
+	if (WARN_ON_ONCE(!ordered))
+		return -EIO;
+
+	/* No need to split */
+	if (ordered->disk_num_bytes == len)
+		goto out;
+
+	/* We cannot split once end_bio'd ordered extent */
+	if (WARN_ON_ONCE(ordered->bytes_left != ordered->disk_num_bytes)) {
+		ret = -EINVAL;
+		goto out;
+	}
+
+	/* We cannot split a compressed ordered extent */
+	if (WARN_ON_ONCE(ordered->disk_num_bytes != ordered->num_bytes)) {
+		ret = -EINVAL;
+		goto out;
+	}
+
+	/* We cannot split a waited ordered extent */
+	if (WARN_ON_ONCE(wq_has_sleeper(&ordered->wait))) {
+		ret = -EINVAL;
+		goto out;
+	}
+
+	ordered_end = ordered->disk_bytenr + ordered->disk_num_bytes;
+	/* bio must be in one ordered extent */
+	if (WARN_ON_ONCE(start < ordered->disk_bytenr || end > ordered_end)) {
+		ret = -EINVAL;
+		goto out;
+	}
+
+	/* Checksum list should be empty */
+	if (WARN_ON_ONCE(!list_empty(&ordered->list))) {
+		ret = -EINVAL;
+		goto out;
+	}
+
+	pre = start - ordered->disk_bytenr;
+	post = ordered_end - end;
+
+	btrfs_split_ordered_extent(ordered, pre, post);
+
+	read_lock(&em_tree->lock);
+	em = lookup_extent_mapping(em_tree, ordered->file_offset, len);
+	if (!em) {
+		read_unlock(&em_tree->lock);
+		ret = -EIO;
+		goto out;
+	}
+	read_unlock(&em_tree->lock);
+
+	ASSERT(!test_bit(EXTENT_FLAG_COMPRESSED, &em->flags));
+	em_new = create_io_em(BTRFS_I(inode), em->start + pre, len,
+			      em->start + pre, em->block_start + pre, len,
+			      len, len, BTRFS_COMPRESS_NONE,
+			      BTRFS_ORDERED_REGULAR);
+	free_extent_map(em_new);
+
+out:
+	free_extent_map(em);
+	btrfs_put_ordered_extent(ordered);
+
+	return ret;
+}
+
 /*
  * extent_io.c submission hook. This does the right thing for csum calculation
  * on write, or reading the csums from the tree before a read.
@@ -2192,6 +2272,15 @@ blk_status_t btrfs_submit_data_bio(struct inode *inode, struct bio *bio,
 	if (btrfs_is_free_space_inode(BTRFS_I(inode)))
 		metadata = BTRFS_WQ_ENDIO_FREE_SPACE;
 
+	if (bio_op(bio) == REQ_OP_ZONE_APPEND) {
+		struct page *page = bio_first_bvec_all(bio)->bv_page;
+		loff_t file_offset = page_offset(page);
+
+		ret = extract_ordered_extent(inode, bio, file_offset);
+		if (ret)
+			goto out;
+	}
+
 	if (btrfs_op(bio) != BTRFS_MAP_WRITE) {
 		ret = btrfs_bio_wq_end_io(fs_info, bio, metadata);
 		if (ret)
fs/btrfs/ordered-data.c +76 -0

82 lines not on this page · raw message

fs/btrfs/ordered-data.h +2 -0
index c3a2325e64a4..e346b03bd66a 100644
@@ -193,6 +193,8 @@ void btrfs_wait_ordered_roots(struct btrfs_fs_info *fs_info, u64 nr,
 void btrfs_lock_and_flush_ordered_range(struct btrfs_inode *inode, u64 start,
 					u64 end,
 					struct extent_state **cached_state);
+void btrfs_split_ordered_extent(struct btrfs_ordered_extent *ordered, u64 pre,
+				u64 post);
 int __init ordered_data_init(void);
 void __cold ordered_data_exit(void);
 
Signature · 1 line
2.27.0
#92 lore raw

Re: [PATCH v10 23/41] btrfs: split ordered extent when bio is sent

I love your patch! Perhaps something to improve:

Hi Naohiro,

I love your patch! Perhaps something to improve:

[auto build test WARNING on xfs-linux/for-next]
[also build test WARNING on v5.10-rc3]
[cannot apply to kdave/for-next block/for-next next-20201110] [If your patch is applied to the wrong git tree, kindly drop us a note. And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]

url:    https://github.com/0day-ci/linux/commits/Naohiro-Aota/btrfs-zoned-block-device-support/20201110-193227
base:   https://git.kernel.org/pub/scm/fs/xfs/xfs-linux.git for-next
config: arc-allyesconfig (attached as .config)
compiler: arceb-elf-gcc (GCC) 9.3.0

reproduce (this is a W=1 build):
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross chmod +x ~/bin/make.cross
# https://github.com/0day-ci/linux/commit/c2b1e52b104fa60d0c731cc5016be18e98ec71d2
git remote add linux-review https://github.com/0day-ci/linux
git fetch --no-tags linux-review Naohiro-Aota/btrfs-zoned-block-device-support/20201110-193227 git checkout c2b1e52b104fa60d0c731cc5016be18e98ec71d2
# save the attached .config to linux build tree
COMPILER_INSTALL_PATH=$HOME/0day COMPILER=gcc-9.3.0 make.cross ARCH=arc

If you fix the issue, kindly add following tag as appropriate

All warnings (new ones prefixed by >>):

>> fs/btrfs/inode.c:2161:5: warning: no previous prototype for 'extract_ordered_extent' [-Wmissing-prototypes]

2161 | int extract_ordered_extent(struct inode *inode, struct bio *bio,
| ^~~~~~~~~~~~~~~~~~~~~~

vim +/extract_ordered_extent +2161 fs/btrfs/inode.c

2160

> 2161	int extract_ordered_extent(struct inode *inode, struct bio *bio,
  2162				   loff_t file_offset)
  2163	{
  2164		struct btrfs_ordered_extent *ordered;
  2165		struct extent_map *em = NULL, *em_new = NULL;
  2166		struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree;
  2167		u64 start = (u64)bio->bi_iter.bi_sector << SECTOR_SHIFT;
  2168		u64 len = bio->bi_iter.bi_size;
  2169		u64 end = start + len;
  2170		u64 ordered_end;
  2171		u64 pre, post;
  2172		int ret = 0;
  2173	
  2174		ordered = btrfs_lookup_ordered_extent(BTRFS_I(inode), file_offset);
  2175		if (WARN_ON_ONCE(!ordered))
  2176			return -EIO;
  2177	
  2178		/* No need to split */
  2179		if (ordered->disk_num_bytes == len)
  2180			goto out;
  2181	
  2182		/* We cannot split once end_bio'd ordered extent */
  2183		if (WARN_ON_ONCE(ordered->bytes_left != ordered->disk_num_bytes)) {
  2184			ret = -EINVAL;
  2185			goto out;
  2186		}
  2187	
  2188		/* We cannot split a compressed ordered extent */
  2189		if (WARN_ON_ONCE(ordered->disk_num_bytes != ordered->num_bytes)) {
  2190			ret = -EINVAL;
  2191			goto out;
  2192		}
  2193	
  2194		/* We cannot split a waited ordered extent */
  2195		if (WARN_ON_ONCE(wq_has_sleeper(&ordered->wait))) {
  2196			ret = -EINVAL;
  2197			goto out;
  2198		}
  2199	
  2200		ordered_end = ordered->disk_bytenr + ordered->disk_num_bytes;
  2201		/* bio must be in one ordered extent */
  2202		if (WARN_ON_ONCE(start < ordered->disk_bytenr || end > ordered_end)) {
  2203			ret = -EINVAL;
  2204			goto out;
  2205		}
  2206	
  2207		/* Checksum list should be empty */
  2208		if (WARN_ON_ONCE(!list_empty(&ordered->list))) {
  2209			ret = -EINVAL;
  2210			goto out;
  2211		}
  2212	
  2213		pre = start - ordered->disk_bytenr;
  2214		post = ordered_end - end;
  2215	
  2216		btrfs_split_ordered_extent(ordered, pre, post);
  2217	
  2218		read_lock(&em_tree->lock);
  2219		em = lookup_extent_mapping(em_tree, ordered->file_offset, len);
  2220		if (!em) {
  2221			read_unlock(&em_tree->lock);
  2222			ret = -EIO;
  2223			goto out;
  2224		}
  2225		read_unlock(&em_tree->lock);
  2226	
  2227		ASSERT(!test_bit(EXTENT_FLAG_COMPRESSED, &em->flags));
  2228		em_new = create_io_em(BTRFS_I(inode), em->start + pre, len,
  2229				      em->start + pre, em->block_start + pre, len,
  2230				      len, len, BTRFS_COMPRESS_NONE,
  2231				      BTRFS_ORDERED_REGULAR);

2232 free_extent_map(em_new);
2233
2234 out:
2235 free_extent_map(em);
2236 btrfs_put_ordered_extent(ordered);
2237
2238 return ret;
2239 }
2240

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all(a)lists.01.org

#93 lore raw

Re: [PATCH v10 23/41] btrfs: split ordered extent when bio is sent

In reply to Naohiro Aota
I love your patch! Perhaps something to improve:

Hi Naohiro,

I love your patch! Perhaps something to improve:

[auto build test WARNING on xfs-linux/for-next]
[also build test WARNING on v5.10-rc3]
[cannot apply to kdave/for-next block/for-next next-20201110] [If your patch is applied to the wrong git tree, kindly drop us a note. And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]

url:    https://github.com/0day-ci/linux/commits/Naohiro-Aota/btrfs-zoned-block-device-support/20201110-193227
base:   https://git.kernel.org/pub/scm/fs/xfs/xfs-linux.git for-next
config: powerpc-randconfig-r022-20201110 (attached as .config)
compiler: clang version 12.0.0 (https://github.com/llvm/llvm-project 4d81c8adb6ed9840257f6cb6b93f60856d422a15)

reproduce (this is a W=1 build):
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross chmod +x ~/bin/make.cross
# install powerpc cross compiling tool for clang build
# apt-get install binutils-powerpc-linux-gnu
# https://github.com/0day-ci/linux/commit/c2b1e52b104fa60d0c731cc5016be18e98ec71d2
git remote add linux-review https://github.com/0day-ci/linux
git fetch --no-tags linux-review Naohiro-Aota/btrfs-zoned-block-device-support/20201110-193227 git checkout c2b1e52b104fa60d0c731cc5016be18e98ec71d2
# save the attached .config to linux build tree
COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross ARCH=powerpc

If you fix the issue, kindly add following tag as appropriate

All warnings (new ones prefixed by >>):

>> fs/btrfs/inode.c:2161:5: warning: no previous prototype for function 'extract_ordered_extent' [-Wmissing-prototypes]
   int extract_ordered_extent(struct inode *inode, struct bio *bio,
       ^

fs/btrfs/inode.c:2161:1: note: declare 'static' if the function is not intended to be used outside of this translation unit int extract_ordered_extent(struct inode *inode, struct bio *bio,
^
static
1 warning generated.

vim +/extract_ordered_extent +2161 fs/btrfs/inode.c

2160

> 2161	int extract_ordered_extent(struct inode *inode, struct bio *bio,
  2162				   loff_t file_offset)
  2163	{
  2164		struct btrfs_ordered_extent *ordered;
  2165		struct extent_map *em = NULL, *em_new = NULL;
  2166		struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree;
  2167		u64 start = (u64)bio->bi_iter.bi_sector << SECTOR_SHIFT;
  2168		u64 len = bio->bi_iter.bi_size;
  2169		u64 end = start + len;
  2170		u64 ordered_end;
  2171		u64 pre, post;
  2172		int ret = 0;
  2173	
  2174		ordered = btrfs_lookup_ordered_extent(BTRFS_I(inode), file_offset);
  2175		if (WARN_ON_ONCE(!ordered))
  2176			return -EIO;
  2177	
  2178		/* No need to split */
  2179		if (ordered->disk_num_bytes == len)
  2180			goto out;
  2181	
  2182		/* We cannot split once end_bio'd ordered extent */
  2183		if (WARN_ON_ONCE(ordered->bytes_left != ordered->disk_num_bytes)) {
  2184			ret = -EINVAL;
  2185			goto out;
  2186		}
  2187	
  2188		/* We cannot split a compressed ordered extent */
  2189		if (WARN_ON_ONCE(ordered->disk_num_bytes != ordered->num_bytes)) {
  2190			ret = -EINVAL;
  2191			goto out;
  2192		}
  2193	
  2194		/* We cannot split a waited ordered extent */
  2195		if (WARN_ON_ONCE(wq_has_sleeper(&ordered->wait))) {
  2196			ret = -EINVAL;
  2197			goto out;
  2198		}
  2199	
  2200		ordered_end = ordered->disk_bytenr + ordered->disk_num_bytes;
  2201		/* bio must be in one ordered extent */
  2202		if (WARN_ON_ONCE(start < ordered->disk_bytenr || end > ordered_end)) {
  2203			ret = -EINVAL;
  2204			goto out;
  2205		}
  2206	
  2207		/* Checksum list should be empty */
  2208		if (WARN_ON_ONCE(!list_empty(&ordered->list))) {
  2209			ret = -EINVAL;
  2210			goto out;
  2211		}
  2212	
  2213		pre = start - ordered->disk_bytenr;
  2214		post = ordered_end - end;
  2215	
  2216		btrfs_split_ordered_extent(ordered, pre, post);
  2217	
  2218		read_lock(&em_tree->lock);
  2219		em = lookup_extent_mapping(em_tree, ordered->file_offset, len);
  2220		if (!em) {
  2221			read_unlock(&em_tree->lock);
  2222			ret = -EIO;
  2223			goto out;
  2224		}
  2225		read_unlock(&em_tree->lock);
  2226	
  2227		ASSERT(!test_bit(EXTENT_FLAG_COMPRESSED, &em->flags));
  2228		em_new = create_io_em(BTRFS_I(inode), em->start + pre, len,
  2229				      em->start + pre, em->block_start + pre, len,
  2230				      len, len, BTRFS_COMPRESS_NONE,
  2231				      BTRFS_ORDERED_REGULAR);

2232 free_extent_map(em_new);
2233
2234 out:
2235 free_extent_map(em);
2236 btrfs_put_ordered_extent(ordered);
2237
2238 return ret;
2239 }
2240

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all(a)lists.01.org

#94 lore raw

Re: [PATCH v10 23/41] btrfs: split ordered extent when bio is sent

In reply to Naohiro Aota
I love your patch! Perhaps something to improve:

Hi Naohiro,

I love your patch! Perhaps something to improve:

[auto build test WARNING on xfs-linux/for-next]
[also build test WARNING on v5.10-rc3]
[cannot apply to kdave/for-next block/for-next next-20201110] [If your patch is applied to the wrong git tree, kindly drop us a note. And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]

url:    https://github.com/0day-ci/linux/commits/Naohiro-Aota/btrfs-zoned-block-device-support/20201110-193227
base:   https://git.kernel.org/pub/scm/fs/xfs/xfs-linux.git for-next
config: x86_64-randconfig-s022-20201110 (attached as .config)
compiler: gcc-9 (Debian 9.3.0-15) 9.3.0
reproduce:
        # apt-get install sparse
        # sparse version: v0.6.3-76-gf680124b-dirty
        # https://github.com/0day-ci/linux/commit/c2b1e52b104fa60d0c731cc5016be18e98ec71d2
        git remote add linux-review https://github.com/0day-ci/linux
        git fetch --no-tags linux-review Naohiro-Aota/btrfs-zoned-block-device-support/20201110-193227
        git checkout c2b1e52b104fa60d0c731cc5016be18e98ec71d2
        # save the attached .config to linux build tree
        make W=1 C=1 CF='-fdiagnostic-prefix -D__CHECK_ENDIAN__' ARCH=x86_64 

If you fix the issue, kindly add following tag as appropriate

"sparse warnings: (new ones prefixed by >>)"

>> fs/btrfs/inode.c:2161:5: sparse: sparse: symbol 'extract_ordered_extent' was not declared. Should it be static?
>> fs/btrfs/inode.c:2279:21: sparse: sparse: incorrect type in assignment (different base types) @@     expected restricted blk_status_t [usertype] ret @@     got int @@
>> fs/btrfs/inode.c:2279:21: sparse:     expected restricted blk_status_t [usertype] ret
>> fs/btrfs/inode.c:2279:21: sparse:     got int

Please review and possibly fold the followup patch.

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all(a)lists.01.org

#95 lore raw

[RFC PATCH] btrfs: extract_ordered_extent() can be static

In reply to Naohiro Aota
Reported-by: kernel test robot <[email protected]>
---
 inode.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
fs/btrfs/inode.c +2 -2
index df85d8dea37c9e..ee3f3ab1b964b6 100644
@@ -2158,8 +2158,8 @@ static blk_status_t btrfs_submit_bio_start(void *private_data, struct bio *bio,
 	return btrfs_csum_one_bio(BTRFS_I(inode), bio, 0, 0);
 }
 
-int extract_ordered_extent(struct inode *inode, struct bio *bio,
-			   loff_t file_offset)
+static int extract_ordered_extent(struct inode *inode, struct bio *bio,
+				  loff_t file_offset)
 {
 	struct btrfs_ordered_extent *ordered;
 	struct extent_map *em = NULL, *em_new = NULL;
#96 lore raw

PATCH v10 23/41 btrfs: split ordered extent when bio is sent

In reply to Naohiro Aota

For a zone append write, the device decides the location the data is written to. Therefore we cannot ensure that two bios are written consecutively on the device. In order to ensure that a ordered extent maps to a contiguous region on disk, we need to maintain a "one bio == one ordered extent" rule.

This commit implements the splitting of an ordered extent and extent map on bio submission to adhere to the rule.

[testbot] made extract_ordered_extent static

---
 fs/btrfs/inode.c        | 89 +++++++++++++++++++++++++++++++++++++++++
 fs/btrfs/ordered-data.c | 76 +++++++++++++++++++++++++++++++++++
 fs/btrfs/ordered-data.h |  2 +
 3 files changed, 167 insertions(+)
fs/btrfs/inode.c +89 -0
index 591ca539e444..ee3f3ab1b964 100644
@@ -2158,6 +2158,86 @@ static blk_status_t btrfs_submit_bio_start(void *private_data, struct bio *bio,
 	return btrfs_csum_one_bio(BTRFS_I(inode), bio, 0, 0);
 }
 
+static int extract_ordered_extent(struct inode *inode, struct bio *bio,
+				  loff_t file_offset)
+{
+	struct btrfs_ordered_extent *ordered;
+	struct extent_map *em = NULL, *em_new = NULL;
+	struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree;
+	u64 start = (u64)bio->bi_iter.bi_sector << SECTOR_SHIFT;
+	u64 len = bio->bi_iter.bi_size;
+	u64 end = start + len;
+	u64 ordered_end;
+	u64 pre, post;
+	int ret = 0;
+
+	ordered = btrfs_lookup_ordered_extent(BTRFS_I(inode), file_offset);
+	if (WARN_ON_ONCE(!ordered))
+		return -EIO;
+
+	/* No need to split */
+	if (ordered->disk_num_bytes == len)
+		goto out;
+
+	/* We cannot split once end_bio'd ordered extent */
+	if (WARN_ON_ONCE(ordered->bytes_left != ordered->disk_num_bytes)) {
+		ret = -EINVAL;
+		goto out;
+	}
+
+	/* We cannot split a compressed ordered extent */
+	if (WARN_ON_ONCE(ordered->disk_num_bytes != ordered->num_bytes)) {
+		ret = -EINVAL;
+		goto out;
+	}
+
+	/* We cannot split a waited ordered extent */
+	if (WARN_ON_ONCE(wq_has_sleeper(&ordered->wait))) {
+		ret = -EINVAL;
+		goto out;
+	}
+
+	ordered_end = ordered->disk_bytenr + ordered->disk_num_bytes;
+	/* bio must be in one ordered extent */
+	if (WARN_ON_ONCE(start < ordered->disk_bytenr || end > ordered_end)) {
+		ret = -EINVAL;
+		goto out;
+	}
+
+	/* Checksum list should be empty */
+	if (WARN_ON_ONCE(!list_empty(&ordered->list))) {
+		ret = -EINVAL;
+		goto out;
+	}
+
+	pre = start - ordered->disk_bytenr;
+	post = ordered_end - end;
+
+	btrfs_split_ordered_extent(ordered, pre, post);
+
+	read_lock(&em_tree->lock);
+	em = lookup_extent_mapping(em_tree, ordered->file_offset, len);
+	if (!em) {
+		read_unlock(&em_tree->lock);
+		ret = -EIO;
+		goto out;
+	}
+	read_unlock(&em_tree->lock);
+
+	ASSERT(!test_bit(EXTENT_FLAG_COMPRESSED, &em->flags));
+	em_new = create_io_em(BTRFS_I(inode), em->start + pre, len,
+			      em->start + pre, em->block_start + pre, len,
+			      len, len, BTRFS_COMPRESS_NONE,
+			      BTRFS_ORDERED_REGULAR);
+	free_extent_map(em_new);
+
+out:
+	free_extent_map(em);
+	btrfs_put_ordered_extent(ordered);
+
+	return ret;
+}
+
 /*
  * extent_io.c submission hook. This does the right thing for csum calculation
  * on write, or reading the csums from the tree before a read.
@@ -2192,6 +2272,15 @@ blk_status_t btrfs_submit_data_bio(struct inode *inode, struct bio *bio,
 	if (btrfs_is_free_space_inode(BTRFS_I(inode)))
 		metadata = BTRFS_WQ_ENDIO_FREE_SPACE;
 
+	if (bio_op(bio) == REQ_OP_ZONE_APPEND) {
+		struct page *page = bio_first_bvec_all(bio)->bv_page;
+		loff_t file_offset = page_offset(page);
+
+		ret = extract_ordered_extent(inode, bio, file_offset);
+		if (ret)
+			goto out;
+	}
+
 	if (btrfs_op(bio) != BTRFS_MAP_WRITE) {
 		ret = btrfs_bio_wq_end_io(fs_info, bio, metadata);
 		if (ret)
fs/btrfs/ordered-data.c +76 -0

82 lines not on this page · raw message

fs/btrfs/ordered-data.h +2 -0
index c3a2325e64a4..e346b03bd66a 100644
@@ -193,6 +193,8 @@ void btrfs_wait_ordered_roots(struct btrfs_fs_info *fs_info, u64 nr,
 void btrfs_lock_and_flush_ordered_range(struct btrfs_inode *inode, u64 start,
 					u64 end,
 					struct extent_state **cached_state);
+void btrfs_split_ordered_extent(struct btrfs_ordered_extent *ordered, u64 pre,
+				u64 post);
 int __init ordered_data_init(void);
 void __cold ordered_data_exit(void);
 
Signature · 1 line
2.27.0
#97 lore raw

PATCH v10 24/41 btrfs: extend btrfs_rmap_block for specifying a device

In reply to Naohiro Aota

btrfs_rmap_block currently reverse-maps the physical addresses on all devices to the corresponding logical addresses.

This commit extends the function to match to a specified device. The old functionality of querying all devices is left intact by specifying NULL as target device.

We pass block_device instead of btrfs_device to __btrfs_rmap_block. This function is intended to reverse-map the result of bio, which only have block_device.

This commit also exports the function for later use.

---
 fs/btrfs/block-group.c            | 20 ++++++++++++++------
 fs/btrfs/block-group.h            |  8 +++-----
 fs/btrfs/tests/extent-map-tests.c |  2 +-
 3 files changed, 18 insertions(+), 12 deletions(-)
fs/btrfs/block-group.c +14 -6
index 31511e59ca74..04bb0602f1cc 100644
@@ -1646,8 +1646,11 @@ static void set_avail_alloc_bits(struct btrfs_fs_info *fs_info, u64 flags)
 }
 
 /**
- * btrfs_rmap_block - Map a physical disk address to a list of logical addresses
+ * btrfs_rmap_block - Map a physical disk address to a list of logical
+ *                    addresses
  * @chunk_start:   logical address of block group
+ * @bdev:	   physical device to resolve. Can be NULL to indicate any
+ *                 device.
  * @physical:	   physical address to map to logical addresses
  * @logical:	   return array of logical addresses which map to @physical
  * @naddrs:	   length of @logical
@@ -1657,9 +1660,9 @@ static void set_avail_alloc_bits(struct btrfs_fs_info *fs_info, u64 flags)
  * Used primarily to exclude those portions of a block group that contain super
  * block copies.
  */
-EXPORT_FOR_TESTS
 int btrfs_rmap_block(struct btrfs_fs_info *fs_info, u64 chunk_start,
-		     u64 physical, u64 **logical, int *naddrs, int *stripe_len)
+		     struct block_device *bdev, u64 physical, u64 **logical,
+		     int *naddrs, int *stripe_len)
 {
 	struct extent_map *em;
 	struct map_lookup *map;
@@ -1677,6 +1680,7 @@ int btrfs_rmap_block(struct btrfs_fs_info *fs_info, u64 chunk_start,
 	map = em->map_lookup;
 	data_stripe_length = em->orig_block_len;
 	io_stripe_size = map->stripe_len;
+	chunk_start = em->start;
 
 	/* For RAID5/6 adjust to a full IO stripe length */
 	if (map->type & BTRFS_BLOCK_GROUP_RAID56_MASK)
@@ -1691,14 +1695,18 @@ int btrfs_rmap_block(struct btrfs_fs_info *fs_info, u64 chunk_start,
 	for (i = 0; i < map->num_stripes; i++) {
 		bool already_inserted = false;
 		u64 stripe_nr;
+		u64 offset;
 		int j;
 
 		if (!in_range(physical, map->stripes[i].physical,
 			      data_stripe_length))
 			continue;
 
+		if (bdev && map->stripes[i].dev->bdev != bdev)
+			continue;
+
 		stripe_nr = physical - map->stripes[i].physical;
-		stripe_nr = div64_u64(stripe_nr, map->stripe_len);
+		stripe_nr = div64_u64_rem(stripe_nr, map->stripe_len, &offset);
 
 		if (map->type & BTRFS_BLOCK_GROUP_RAID10) {
 			stripe_nr = stripe_nr * map->num_stripes + i;
@@ -1712,7 +1720,7 @@ int btrfs_rmap_block(struct btrfs_fs_info *fs_info, u64 chunk_start,
 		 * instead of map->stripe_len
 		 */
 
-		bytenr = chunk_start + stripe_nr * io_stripe_size;
+		bytenr = chunk_start + stripe_nr * io_stripe_size + offset;
 
 		/* Ensure we don't add duplicate addresses */
 		for (j = 0; j < nr; j++) {
@@ -1754,7 +1762,7 @@ static int exclude_super_stripes(struct btrfs_block_group *cache)
 
 	for (i = 0; i < BTRFS_SUPER_MIRROR_MAX; i++) {
 		bytenr = btrfs_sb_offset(i);
-		ret = btrfs_rmap_block(fs_info, cache->start,
+		ret = btrfs_rmap_block(fs_info, cache->start, NULL,
 				       bytenr, &logical, &nr, &stripe_len);
 		if (ret)
 			return ret;
fs/btrfs/block-group.h +3 -5
index 5be47f4bfea7..9a4009eaaecb 100644
@@ -275,6 +275,9 @@ void check_system_chunk(struct btrfs_trans_handle *trans, const u64 type);
 u64 btrfs_get_alloc_profile(struct btrfs_fs_info *fs_info, u64 orig_flags);
 void btrfs_put_block_group_cache(struct btrfs_fs_info *info);
 int btrfs_free_block_groups(struct btrfs_fs_info *info);
+int btrfs_rmap_block(struct btrfs_fs_info *fs_info, u64 chunk_start,
+		       struct block_device *bdev, u64 physical, u64 **logical,
+		       int *naddrs, int *stripe_len);
 
 static inline u64 btrfs_data_alloc_profile(struct btrfs_fs_info *fs_info)
 {
@@ -301,9 +304,4 @@ static inline int btrfs_block_group_done(struct btrfs_block_group *cache)
 void btrfs_freeze_block_group(struct btrfs_block_group *cache);
 void btrfs_unfreeze_block_group(struct btrfs_block_group *cache);
 
-#ifdef CONFIG_BTRFS_FS_RUN_SANITY_TESTS
-int btrfs_rmap_block(struct btrfs_fs_info *fs_info, u64 chunk_start,
-		     u64 physical, u64 **logical, int *naddrs, int *stripe_len);
-#endif
-
 #endif /* BTRFS_BLOCK_GROUP_H */
fs/btrfs/tests/extent-map-tests.c +1 -1
index 57379e96ccc9..c0aefe6dee0b 100644
@@ -507,7 +507,7 @@ static int test_rmap_block(struct btrfs_fs_info *fs_info,
 		goto out_free;
 	}
 
-	ret = btrfs_rmap_block(fs_info, em->start, btrfs_sb_offset(1),
+	ret = btrfs_rmap_block(fs_info, em->start, NULL, btrfs_sb_offset(1),
 			       &logical, &out_ndaddrs, &out_stripe_len);
 	if (ret || (out_ndaddrs == 0 && test->expected_mapped_addr)) {
 		test_err("didn't rmap anything but expected %d",
Signature · 1 line
2.27.0
#98 lore raw

PATCH v10 25/41 btrfs: use ZONE_APPEND write for ZONED btrfs

In reply to Naohiro Aota

This commit enables zone append writing for zoned btrfs. When using zone append, a bio is issued to the start of a target zone and the device decides to place it inside the zone. Upon completion the device reports the actual written position back to the host.

Three parts are necessary to enable zone append in btrfs. First, modify the bio to use REQ_OP_ZONE_APPEND in btrfs_submit_bio_hook() and adjust the bi_sector to point the beginning of the zone.

Secondly, records the returned physical address (and disk/partno) to the ordered extent in end_bio_extent_writepage() after the bio has been completed. We cannot resolve the physical address to the logical address because we can neither take locks nor allocate a buffer in this end_bio context. So, we need to record the physical address to resolve it later in btrfs_finish_ordered_io().

And finally, rewrites the logical addresses of the extent mapping and checksum data according to the physical address (using __btrfs_rmap_block). If the returned address matches the originally allocated address, we can skip this rewriting process.

---
 fs/btrfs/extent_io.c    | 12 +++++++-
 fs/btrfs/file.c         |  2 +-
 fs/btrfs/inode.c        |  4 +++
 fs/btrfs/ordered-data.c |  3 ++
 fs/btrfs/ordered-data.h |  8 +++++
 fs/btrfs/volumes.c      | 15 +++++++++
 fs/btrfs/zoned.c        | 68 +++++++++++++++++++++++++++++++++++++++++
 fs/btrfs/zoned.h        | 11 +++++++
 8 files changed, 121 insertions(+), 2 deletions(-)
fs/btrfs/extent_io.c +11 -1
index b9b366f4d942..7f94fef3647b 100644
@@ -2743,6 +2743,7 @@ static void end_bio_extent_writepage(struct bio *bio)
 	u64 start;
 	u64 end;
 	struct bvec_iter_all iter_all;
+	bool first_bvec = true;
 
 	ASSERT(!bio_flagged(bio, BIO_CLONED));
 	bio_for_each_segment_all(bvec, bio, iter_all) {
@@ -2769,6 +2770,11 @@ static void end_bio_extent_writepage(struct bio *bio)
 		start = page_offset(page);
 		end = start + bvec->bv_offset + bvec->bv_len - 1;
 
+		if (first_bvec) {
+			btrfs_record_physical_zoned(inode, start, bio);
+			first_bvec = false;
+		}
+
 		end_extent_writepage(page, error, start, end);
 		end_page_writeback(page);
 	}
@@ -3525,6 +3531,7 @@ static noinline_for_stack int __extent_writepage_io(struct btrfs_inode *inode,
 	size_t blocksize;
 	int ret = 0;
 	int nr = 0;
+	int opf = REQ_OP_WRITE;
 	const unsigned int write_flags = wbc_to_write_flags(wbc);
 	bool compressed;
 
@@ -3537,6 +3544,9 @@ static noinline_for_stack int __extent_writepage_io(struct btrfs_inode *inode,
 		return 1;
 	}
 
+	if (btrfs_is_zoned(inode->root->fs_info))
+		opf = REQ_OP_ZONE_APPEND;
+
 	/*
 	 * we don't want to touch the inode after unlocking the page,
 	 * so we update the mapping writeback index now
@@ -3597,7 +3607,7 @@ static noinline_for_stack int __extent_writepage_io(struct btrfs_inode *inode,
 			       page->index, cur, end);
 		}
 
-		ret = submit_extent_page(REQ_OP_WRITE | write_flags, wbc,
+		ret = submit_extent_page(opf | write_flags, wbc,
 					 page, offset, iosize, pg_offset,
 					 &epd->bio,
 					 end_bio_extent_writepage,
fs/btrfs/file.c +1 -1
index 68938a43081e..bdc268c91334 100644
@@ -2226,7 +2226,7 @@ int btrfs_sync_file(struct file *file, loff_t start, loff_t end, int datasync)
 	 * the current transaction commits before the ordered extents complete
 	 * and a power failure happens right after that.
 	 */
-	if (full_sync) {
+	if (full_sync || btrfs_is_zoned(fs_info)) {
 		ret = btrfs_wait_ordered_range(inode, start, len);
 	} else {
 		/*
fs/btrfs/inode.c +4 -0
index df85d8dea37c..fe15441278de 100644
@@ -51,6 +51,7 @@
 #include "delalloc-space.h"
 #include "block-group.h"
 #include "space-info.h"
+#include "zoned.h"
 
 struct btrfs_iget_args {
 	u64 ino;
@@ -2676,6 +2677,9 @@ static int btrfs_finish_ordered_io(struct btrfs_ordered_extent *ordered_extent)
 	bool clear_reserved_extent = true;
 	unsigned int clear_bits;
 
+	if (ordered_extent->disk)
+		btrfs_rewrite_logical_zoned(ordered_extent);
+
 	start = ordered_extent->file_offset;
 	end = start + ordered_extent->num_bytes - 1;
 
fs/btrfs/ordered-data.c +3 -0
index 35ef25e39561..1a3b06713d0f 100644
@@ -199,6 +199,9 @@ static int __btrfs_add_ordered_extent(struct btrfs_inode *inode, u64 file_offset
 	entry->compress_type = compress_type;
 	entry->truncated_len = (u64)-1;
 	entry->qgroup_rsv = ret;
+	entry->physical = (u64)-1;
+	entry->disk = NULL;
+	entry->partno = (u8)-1;
 	if (type != BTRFS_ORDERED_IO_DONE && type != BTRFS_ORDERED_COMPLETE)
 		set_bit(type, &entry->flags);
 
fs/btrfs/ordered-data.h +8 -0
index e346b03bd66a..084c609afd83 100644
@@ -127,6 +127,14 @@ struct btrfs_ordered_extent {
 	struct completion completion;
 	struct btrfs_work flush_work;
 	struct list_head work_list;
+
+	/*
+	 * used to reverse-map physical address returned from ZONE_APPEND
+	 * write command in a workqueue context.
+	 */
+	u64 physical;
+	struct gendisk *disk;
+	u8 partno;
 };
 
 /*
fs/btrfs/volumes.c +15 -0
index 683b3ed06226..c8187d704c89 100644
@@ -6503,6 +6503,21 @@ static void submit_stripe_bio(struct btrfs_bio *bbio, struct bio *bio,
 	btrfs_io_bio(bio)->device = dev;
 	bio->bi_end_io = btrfs_end_bio;
 	bio->bi_iter.bi_sector = physical >> 9;
+	/*
+	 * For zone append writing, bi_sector must point the beginning of the
+	 * zone
+	 */
+	if (bio_op(bio) == REQ_OP_ZONE_APPEND) {
+		if (btrfs_dev_is_sequential(dev, physical)) {
+			u64 zone_start = round_down(physical,
+						    fs_info->zone_size);
+
+			bio->bi_iter.bi_sector = zone_start >> SECTOR_SHIFT;
+		} else {
+			bio->bi_opf &= ~REQ_OP_ZONE_APPEND;
+			bio->bi_opf |= REQ_OP_WRITE;
+		}
+	}
 	btrfs_debug_in_rcu(fs_info,
 	"btrfs_map_bio: rw %d 0x%x, sector=%llu, dev=%lu (%s id %llu), size=%u",
 		bio_op(bio), bio->bi_opf, (u64)bio->bi_iter.bi_sector,
fs/btrfs/zoned.c +68 -0

71 lines not on this page · raw message

fs/btrfs/zoned.h +11 -0
index a7de80c313be..2872a0cbc847 100644
@@ -45,6 +45,9 @@ void btrfs_calc_zone_unusable(struct btrfs_block_group *cache);
 void btrfs_redirty_list_add(struct btrfs_transaction *trans,
 			    struct extent_buffer *eb);
 void btrfs_free_redirty_list(struct btrfs_transaction *trans);
+void btrfs_record_physical_zoned(struct inode *inode, u64 file_offset,
+				 struct bio *bio);
+void btrfs_rewrite_logical_zoned(struct btrfs_ordered_extent *ordered);
 #else /* CONFIG_BLK_DEV_ZONED */
 static inline int btrfs_get_dev_zone(struct btrfs_device *device, u64 pos,
 				     struct blk_zone *zone)
@@ -129,6 +132,14 @@ static inline void btrfs_redirty_list_add(struct btrfs_transaction *trans,
 					  struct extent_buffer *eb) { }
 static inline void btrfs_free_redirty_list(struct btrfs_transaction *trans) { }
 
+static inline void btrfs_record_physical_zoned(struct inode *inode,
+					       u64 file_offset, struct bio *bio)
+{
+}
+
+static inline void btrfs_rewrite_logical_zoned(
+				struct btrfs_ordered_extent *ordered) { }
+
 #endif
 
 static inline bool btrfs_dev_is_sequential(struct btrfs_device *device, u64 pos)
Signature · 1 line
2.27.0
#99 lore raw

PATCH v10 26/41 btrfs: enable zone append writing for direct IO

In reply to Naohiro Aota

Likewise to buffered IO, enable zone append writing for direct IO when its used on a zoned block device.

---
 fs/btrfs/inode.c | 17 +++++++++++++++++
 1 file changed, 17 insertions(+)
fs/btrfs/inode.c +17 -0
index fe15441278de..445cb6ba4a59 100644
@@ -7542,6 +7542,9 @@ static int btrfs_dio_iomap_begin(struct inode *inode, loff_t start,
 	iomap->bdev = fs_info->fs_devices->latest_bdev;
 	iomap->length = len;
 
+	if (write && btrfs_is_zoned(fs_info) && fs_info->max_zone_append_size)
+		iomap->flags |= IOMAP_F_ZONE_APPEND;
+
 	free_extent_map(em);
 
 	return 0;
@@ -7779,6 +7782,8 @@ static void btrfs_end_dio_bio(struct bio *bio)
 	if (err)
 		dip->dio_bio->bi_status = err;
 
+	btrfs_record_physical_zoned(dip->inode, dip->logical_offset, bio);
+
 	bio_put(bio);
 	btrfs_dio_private_put(dip);
 }
@@ -7933,6 +7938,18 @@ static blk_qc_t btrfs_submit_direct(struct inode *inode, struct iomap *iomap,
 		bio->bi_end_io = btrfs_end_dio_bio;
 		btrfs_io_bio(bio)->logical = file_offset;
 
+		WARN_ON_ONCE(write && btrfs_is_zoned(fs_info) &&
+			     fs_info->max_zone_append_size &&
+			     bio_op(bio) != REQ_OP_ZONE_APPEND);
+
+		if (bio_op(bio) == REQ_OP_ZONE_APPEND) {
+			ret = extract_ordered_extent(inode, bio, file_offset);
+			if (ret) {
+				bio_put(bio);
+				goto out_err;
+			}
+		}
+
 		ASSERT(submit_len >= clone_len);
 		submit_len -= clone_len;
 
Signature · 1 line
2.27.0
#100 lore raw

PATCH v10 27/41 btrfs: introduce dedicated data write path for ZONED mode

In reply to Naohiro Aota

If more than one IO is issued for one file extent, these IO can be written to separate regions on a device. Since we cannot map one file extent to such a separate area, we need to follow the "one IO == one ordered extent" rule.

The Normal buffered, uncompressed, not pre-allocated write path (used by cow_file_range()) sometimes does not follow this rule. It can write a part of an ordered extent when specified a region to write e.g., when its called from fdatasync().

Introduces a dedicated (uncompressed buffered) data write path for ZONED mode. This write path will CoW the region and write it at once.

---
 fs/btrfs/inode.c | 34 ++++++++++++++++++++++++++++++++--
 1 file changed, 32 insertions(+), 2 deletions(-)
fs/btrfs/inode.c +32 -2
index 445cb6ba4a59..991ef2bf018f 100644
@@ -1350,6 +1350,29 @@ static int cow_file_range_async(struct btrfs_inode *inode,
 	return 0;
 }
 
+static noinline int run_delalloc_zoned(struct btrfs_inode *inode,
+				       struct page *locked_page, u64 start,
+				       u64 end, int *page_started,
+				       unsigned long *nr_written)
+{
+	int ret;
+
+	ret = cow_file_range(inode, locked_page, start, end,
+			     page_started, nr_written, 0);
+	if (ret)
+		return ret;
+
+	if (*page_started)
+		return 0;
+
+	__set_page_dirty_nobuffers(locked_page);
+	account_page_redirty(locked_page);
+	extent_write_locked_range(&inode->vfs_inode, start, end, WB_SYNC_ALL);
+	*page_started = 1;
+
+	return 0;
+}
+
 static noinline int csum_exist_in_range(struct btrfs_fs_info *fs_info,
 					u64 bytenr, u64 num_bytes)
 {
@@ -1820,17 +1843,24 @@ int btrfs_run_delalloc_range(struct btrfs_inode *inode, struct page *locked_page
 {
 	int ret;
 	int force_cow = need_force_cow(inode, start, end);
+	const bool do_compress = inode_can_compress(inode) &&
+		inode_need_compress(inode, start, end);
+	const bool zoned = btrfs_is_zoned(inode->root->fs_info);
 
 	if (inode->flags & BTRFS_INODE_NODATACOW && !force_cow) {
+		ASSERT(!zoned);
 		ret = run_delalloc_nocow(inode, locked_page, start, end,
 					 page_started, 1, nr_written);
 	} else if (inode->flags & BTRFS_INODE_PREALLOC && !force_cow) {
+		ASSERT(!zoned);
 		ret = run_delalloc_nocow(inode, locked_page, start, end,
 					 page_started, 0, nr_written);
-	} else if (!inode_can_compress(inode) ||
-		   !inode_need_compress(inode, start, end)) {
+	} else if (!do_compress && !zoned) {
 		ret = cow_file_range(inode, locked_page, start, end,
 				     page_started, nr_written, 1);
+	} else if (!do_compress && zoned) {
+		ret = run_delalloc_zoned(inode, locked_page, start, end,
+					 page_started, nr_written);
 	} else {
 		set_bit(BTRFS_INODE_HAS_ASYNC_EXTENT, &inode->runtime_flags);
 		ret = cow_file_range_async(inode, wbc, locked_page, start, end,
Signature · 1 line
2.27.0
#101 lore raw

PATCH v10 28/41 btrfs: serialize meta IOs on ZONED mode

In reply to Naohiro Aota

We cannot use zone append for writing metadata, because the B-tree nodes have references to each other using the logical address. Without knowing the address in advance, we cannot construct the tree in the first place. So we need to serialize write IOs for metadata.

We cannot add a mutex around allocation and submission because metadata blocks are allocated in an earlier stage to build up B-trees.

Add a zoned_meta_io_lock and hold it during metadata IO submission in btree_write_cache_pages() to serialize IOs. Furthermore, this add a per-block group metadata IO submission pointer "meta_write_pointer" to ensure sequential writing, which can be caused when writing back blocks in an unfinished transaction.

---
 fs/btrfs/block-group.h |  1 +
 fs/btrfs/ctree.h       |  1 +
 fs/btrfs/disk-io.c     |  1 +
 fs/btrfs/extent_io.c   | 27 ++++++++++++++++++++++-
 fs/btrfs/zoned.c       | 50 ++++++++++++++++++++++++++++++++++++++++++
 fs/btrfs/zoned.h       | 32 +++++++++++++++++++++++++++
 6 files changed, 111 insertions(+), 1 deletion(-)
fs/btrfs/block-group.h +1 -0
index 9a4009eaaecb..44f68e12f863 100644
@@ -190,6 +190,7 @@ struct btrfs_block_group {
 	 */
 	u64 alloc_offset;
 	u64 zone_unusable;
+	u64 meta_write_pointer;
 };
 
 static inline u64 btrfs_block_group_end(struct btrfs_block_group *block_group)
fs/btrfs/ctree.h +1 -0
index c70d3fcc62c2..8138e932b7cc 100644
@@ -956,6 +956,7 @@ struct btrfs_fs_info {
 
 	/* Max size to emit ZONE_APPEND write command */
 	u64 max_zone_append_size;
+	struct mutex zoned_meta_io_lock;
 
 #ifdef CONFIG_BTRFS_FS_REF_VERIFY
 	spinlock_t ref_verify_lock;
fs/btrfs/disk-io.c +1 -0
index 8acf1ed75889..66f90ebfc01f 100644
@@ -2652,6 +2652,7 @@ void btrfs_init_fs_info(struct btrfs_fs_info *fs_info)
 	mutex_init(&fs_info->delete_unused_bgs_mutex);
 	mutex_init(&fs_info->reloc_mutex);
 	mutex_init(&fs_info->delalloc_root_mutex);
+	mutex_init(&fs_info->zoned_meta_io_lock);
 	seqlock_init(&fs_info->profiles_lock);
 
 	INIT_LIST_HEAD(&fs_info->dirty_cowonly_roots);
fs/btrfs/extent_io.c +26 -1
index 7f94fef3647b..d26c827f39c6 100644
@@ -25,6 +25,7 @@
 #include "backref.h"
 #include "disk-io.h"
 #include "zoned.h"
+#include "block-group.h"
 
 static struct kmem_cache *extent_state_cache;
 static struct kmem_cache *extent_buffer_cache;
@@ -3995,6 +3996,7 @@ int btree_write_cache_pages(struct address_space *mapping,
 				   struct writeback_control *wbc)
 {
 	struct extent_buffer *eb, *prev_eb = NULL;
+	struct btrfs_block_group *cache = NULL;
 	struct extent_page_data epd = {
 		.bio = NULL,
 		.extent_locked = 0,
@@ -4029,6 +4031,7 @@ int btree_write_cache_pages(struct address_space *mapping,
 		tag = PAGECACHE_TAG_TOWRITE;
 	else
 		tag = PAGECACHE_TAG_DIRTY;
+	btrfs_zoned_meta_io_lock(fs_info);
 retry:
 	if (wbc->sync_mode == WB_SYNC_ALL)
 		tag_pages_for_writeback(mapping, index, end);
@@ -4071,12 +4074,30 @@ int btree_write_cache_pages(struct address_space *mapping,
 			if (!ret)
 				continue;
 
+			if (!btrfs_check_meta_write_pointer(fs_info, eb,
+							    &cache)) {
+				/*
+				 * If for_sync, this hole will be filled with
+				 * trasnsaction commit.
+				 */
+				if (wbc->sync_mode == WB_SYNC_ALL &&
+				    !wbc->for_sync)
+					ret = -EAGAIN;
+				else
+					ret = 0;
+				done = 1;
+				free_extent_buffer(eb);
+				break;
+			}
+
 			prev_eb = eb;
 			ret = lock_extent_buffer_for_io(eb, &epd);
 			if (!ret) {
+				btrfs_revert_meta_write_pointer(cache, eb);
 				free_extent_buffer(eb);
 				continue;
 			} else if (ret < 0) {
+				btrfs_revert_meta_write_pointer(cache, eb);
 				done = 1;
 				free_extent_buffer(eb);
 				break;
@@ -4109,10 +4130,12 @@ int btree_write_cache_pages(struct address_space *mapping,
 		index = 0;
 		goto retry;
 	}
+	if (cache)
+		btrfs_put_block_group(cache);
 	ASSERT(ret <= 0);
 	if (ret < 0) {
 		end_write_bio(&epd, ret);
-		return ret;
+		goto out;
 	}
 	/*
 	 * If something went wrong, don't allow any metadata write bio to be
@@ -4147,6 +4170,8 @@ int btree_write_cache_pages(struct address_space *mapping,
 		ret = -EROFS;
 		end_write_bio(&epd, ret);
 	}
+out:
+	btrfs_zoned_meta_io_unlock(fs_info);
 	return ret;
 }
 
fs/btrfs/zoned.c +50 -0

59 lines not on this page · raw message

fs/btrfs/zoned.h +32 -0
index 2872a0cbc847..41d786a97e40 100644
@@ -48,6 +48,11 @@ void btrfs_free_redirty_list(struct btrfs_transaction *trans);
 void btrfs_record_physical_zoned(struct inode *inode, u64 file_offset,
 				 struct bio *bio);
 void btrfs_rewrite_logical_zoned(struct btrfs_ordered_extent *ordered);
+bool btrfs_check_meta_write_pointer(struct btrfs_fs_info *fs_info,
+				    struct extent_buffer *eb,
+				    struct btrfs_block_group **cache_ret);
+void btrfs_revert_meta_write_pointer(struct btrfs_block_group *cache,
+				     struct extent_buffer *eb);
 #else /* CONFIG_BLK_DEV_ZONED */
 static inline int btrfs_get_dev_zone(struct btrfs_device *device, u64 pos,
 				     struct blk_zone *zone)
@@ -140,6 +145,19 @@ static inline void btrfs_record_physical_zoned(struct inode *inode,
 static inline void btrfs_rewrite_logical_zoned(
 				struct btrfs_ordered_extent *ordered) { }
 
+static inline bool btrfs_check_meta_write_pointer(struct btrfs_fs_info *fs_info,
+			       struct extent_buffer *eb,
+			       struct btrfs_block_group **cache_ret)
+{
+	return true;
+}
+
+static inline void btrfs_revert_meta_write_pointer(
+						struct btrfs_block_group *cache,
+						struct extent_buffer *eb)
+{
+}
+
 #endif
 
 static inline bool btrfs_dev_is_sequential(struct btrfs_device *device, u64 pos)
@@ -243,4 +261,18 @@ static inline bool btrfs_can_zone_reset(struct btrfs_device *device,
 	return true;
 }
 
+static inline void btrfs_zoned_meta_io_lock(struct btrfs_fs_info *fs_info)
+{
+	if (!btrfs_is_zoned(fs_info))
+		return;
+	mutex_lock(&fs_info->zoned_meta_io_lock);
+}
+
+static inline void btrfs_zoned_meta_io_unlock(struct btrfs_fs_info *fs_info)
+{
+	if (!btrfs_is_zoned(fs_info))
+		return;
+	mutex_unlock(&fs_info->zoned_meta_io_lock);
+}
+
 #endif
Signature · 1 line
2.27.0
#102 lore raw

PATCH v10 29/41 btrfs: wait existing extents before truncating

In reply to Naohiro Aota

When truncating a file, file buffers which have already been allocated but not yet written may be truncated. Truncating these buffers could cause breakage of a sequential write pattern in a block group if the truncated blocks are for example followed by blocks allocated to another file. To avoid this problem, always wait for write out of all unwritten buffers before proceeding with the truncate execution.

---
 fs/btrfs/inode.c | 10 ++++++++++
 1 file changed, 10 insertions(+)
fs/btrfs/inode.c +10 -0
index 991ef2bf018f..992aa963592d 100644
@@ -4955,6 +4955,16 @@ static int btrfs_setsize(struct inode *inode, struct iattr *attr)
 		btrfs_drew_write_unlock(&root->snapshot_lock);
 		btrfs_end_transaction(trans);
 	} else {
+		struct btrfs_fs_info *fs_info = btrfs_sb(inode->i_sb);
+
+		if (btrfs_is_zoned(fs_info)) {
+			ret = btrfs_wait_ordered_range(
+				inode,
+				ALIGN(newsize, fs_info->sectorsize),
+				(u64)-1);
+			if (ret)
+				return ret;
+		}
 
 		/*
 		 * We're truncating a file that used to have good data down to
Signature · 1 line
2.27.0
#103 lore raw

PATCH v10 30/41 btrfs: avoid async metadata checksum on ZONED mode

In reply to Naohiro Aota

In ZONED, btrfs uses per-FS zoned_meta_io_lock to serialize the metadata write IOs.

Even with these serialization, write bios sent from btree_write_cache_pages can be reordered by async checksum workers as these workers are per CPU and not per zone.

To preserve write BIO ordering, we can disable async metadata checksum on ZONED. This does not result in lower performance with HDDs as a single CPU core is fast enough to do checksum for a single zone write stream with the maximum possible bandwidth of the device. If multiple zones are being written simultaneously, HDD seek overhead lowers the achievable maximum bandwidth, resulting again in a per zone checksum serialization not affecting performance.

---
 fs/btrfs/disk-io.c | 2 ++
 1 file changed, 2 insertions(+)
fs/btrfs/disk-io.c +2 -0
index 66f90ebfc01f..9490dbbbdb2a 100644
@@ -813,6 +813,8 @@ static blk_status_t btree_submit_bio_start(void *private_data, struct bio *bio,
 static int check_async_write(struct btrfs_fs_info *fs_info,
 			     struct btrfs_inode *bi)
 {
+	if (btrfs_is_zoned(fs_info))
+		return 0;
 	if (atomic_read(&bi->sync_writers))
 		return 0;
 	if (test_bit(BTRFS_FS_CSUM_IMPL_FAST, &fs_info->flags))
Signature · 1 line
2.27.0
#104 lore raw

PATCH v10 31/41 btrfs: mark block groups to copy for device-replace

In reply to Naohiro Aota

This is the 1/4 patch to support device-replace in ZONED mode.

We have two types of I/Os during the device-replace process. One is an I/O to "copy" (by the scrub functions) all the device extents on the source device to the destination device. The other one is an I/O to "clone" (by handle_ops_on_dev_replace()) new incoming write I/Os from users to the source device into the target device.

Cloning incoming I/Os can break the sequential write rule in the target device. When writing is mapped in the middle of a block group, the I/O is directed in the middle of a target device zone, which breaks the sequential write rule.

However, the cloning function cannot be merely disabled since incoming I/Os targeting already copied device extents must be cloned so that the I/O is executed on the target device.

We cannot use dev_replace->cursor_{left,right} to determine whether bio is going to not yet copied region. Since we have a time gap between finishing btrfs_scrub_dev() and rewriting the mapping tree in btrfs_dev_replace_finishing(), we can have a newly allocated device extent which is never cloned nor copied.

So the point is to copy only already existing device extents. This patch introduces mark_block_group_to_copy() to mark existing block groups as a target of copying. Then, handle_ops_on_dev_replace() and dev-replace can check the flag to do their job.

---
 fs/btrfs/block-group.h |   1 +
 fs/btrfs/dev-replace.c | 183 +++++++++++++++++++++++++++++++++++++++++
 fs/btrfs/dev-replace.h |   3 +
 fs/btrfs/scrub.c       |  17 ++++
 4 files changed, 204 insertions(+)
fs/btrfs/block-group.h +1 -0
index 44f68e12f863..ccbcf37eae9c 100644
@@ -95,6 +95,7 @@ struct btrfs_block_group {
 	unsigned int iref:1;
 	unsigned int has_caching_ctl:1;
 	unsigned int removed:1;
+	unsigned int to_copy:1;
 
 	int disk_cache_state;
 
fs/btrfs/dev-replace.c +183 -0

201 lines not on this page · raw message

fs/btrfs/dev-replace.h +3 -0
index 60b70dacc299..3911049a5f23 100644
@@ -18,5 +18,8 @@ int btrfs_dev_replace_cancel(struct btrfs_fs_info *fs_info);
 void btrfs_dev_replace_suspend_for_unmount(struct btrfs_fs_info *fs_info);
 int btrfs_resume_dev_replace_async(struct btrfs_fs_info *fs_info);
 int __pure btrfs_dev_replace_is_ongoing(struct btrfs_dev_replace *dev_replace);
+bool btrfs_finish_block_group_to_copy(struct btrfs_device *srcdev,
+				      struct btrfs_block_group *cache,
+				      u64 physical);
 
 #endif
fs/btrfs/scrub.c +17 -0
index aa1b36cf5c88..d0d7db3c8b0b 100644
@@ -3500,6 +3500,17 @@ int scrub_enumerate_chunks(struct scrub_ctx *sctx,
 		if (!cache)
 			goto skip;
 
+
+		if (sctx->is_dev_replace && btrfs_fs_incompat(fs_info, ZONED)) {
+			spin_lock(&cache->lock);
+			if (!cache->to_copy) {
+				spin_unlock(&cache->lock);
+				ro_set = 0;
+				goto done;
+			}
+			spin_unlock(&cache->lock);
+		}
+
 		/*
 		 * Make sure that while we are scrubbing the corresponding block
 		 * group doesn't get its logical address and its device extents
@@ -3631,6 +3642,12 @@ int scrub_enumerate_chunks(struct scrub_ctx *sctx,
 
 		scrub_pause_off(fs_info);
 
+		if (sctx->is_dev_replace &&
+		    !btrfs_finish_block_group_to_copy(dev_replace->srcdev,
+						      cache, found_key.offset))
+			ro_set = 0;
+
+done:
 		down_write(&dev_replace->rwsem);
 		dev_replace->cursor_left = dev_replace->cursor_right;
 		dev_replace->item_needs_writeback = 1;
Signature · 1 line
2.27.0
#105 lore raw

Re: [PATCH v10 31/41] btrfs: mark block groups to copy for device-replace

I love your patch! Perhaps something to improve:

Hi Naohiro,

I love your patch! Perhaps something to improve:

[auto build test WARNING on xfs-linux/for-next]
[also build test WARNING on v5.10-rc3]
[cannot apply to kdave/for-next block/for-next next-20201110] [If your patch is applied to the wrong git tree, kindly drop us a note. And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]

url:    https://github.com/0day-ci/linux/commits/Naohiro-Aota/btrfs-zoned-block-device-support/20201110-193227
base:   https://git.kernel.org/pub/scm/fs/xfs/xfs-linux.git for-next
config: arc-allyesconfig (attached as .config)
compiler: arceb-elf-gcc (GCC) 9.3.0

reproduce (this is a W=1 build):
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross chmod +x ~/bin/make.cross
# https://github.com/0day-ci/linux/commit/355d6db6997f390f363f08fa9bfbf8b38eb891a2
git remote add linux-review https://github.com/0day-ci/linux
git fetch --no-tags linux-review Naohiro-Aota/btrfs-zoned-block-device-support/20201110-193227 git checkout 355d6db6997f390f363f08fa9bfbf8b38eb891a2
# save the attached .config to linux build tree
COMPILER_INSTALL_PATH=$HOME/0day COMPILER=gcc-9.3.0 make.cross ARCH=arc

If you fix the issue, kindly add following tag as appropriate

All warnings (new ones prefixed by >>):

fs/btrfs/dev-replace.c: In function 'mark_block_group_to_copy':

>> fs/btrfs/dev-replace.c:452:20: warning: variable 'length' set but not used [-Wunused-but-set-variable]
     452 |  u64 chunk_offset, length;
         |                    ^~~~~~

vim +/length +452 fs/btrfs/dev-replace.c

440
441 static int mark_block_group_to_copy(struct btrfs_fs_info *fs_info,
442 struct btrfs_device *src_dev)
443 {
444 struct btrfs_path *path;
445 struct btrfs_key key;
446 struct btrfs_key found_key;
447 struct btrfs_root *root = fs_info->dev_root;
448 struct btrfs_dev_extent *dev_extent = NULL;
449 struct btrfs_block_group *cache;
450 struct btrfs_trans_handle *trans;
451 int ret = 0;

 > 452		u64 chunk_offset, length;

453
454 /* Do not use "to_copy" on non-ZONED for now */
455 if (!btrfs_is_zoned(fs_info))
456 return 0;
457
458 mutex_lock(&fs_info->chunk_mutex);
459
460 /* Ensure we don't have pending new block group */
461 spin_lock(&fs_info->trans_lock);
462 while (fs_info->running_transaction &&
463 !list_empty(&fs_info->running_transaction->dev_update_list)) {
464 spin_unlock(&fs_info->trans_lock);
465 mutex_unlock(&fs_info->chunk_mutex);
466 trans = btrfs_attach_transaction(root);
467 if (IS_ERR(trans)) {
468 ret = PTR_ERR(trans);
469 mutex_lock(&fs_info->chunk_mutex);
470 if (ret == -ENOENT)
471 continue;
472 else
473 goto unlock;
474 }
475
476 ret = btrfs_commit_transaction(trans);
477 mutex_lock(&fs_info->chunk_mutex);
478 if (ret)
479 goto unlock;
480
481 spin_lock(&fs_info->trans_lock);
482 }
483 spin_unlock(&fs_info->trans_lock);
484
485 path = btrfs_alloc_path();
486 if (!path) {
487 ret = -ENOMEM;
488 goto unlock;
489 }
490
491 path->reada = READA_FORWARD;
492 path->search_commit_root = 1;
493 path->skip_locking = 1;
494
495 key.objectid = src_dev->devid;
496 key.offset = 0;
497 key.type = BTRFS_DEV_EXTENT_KEY;
498
499 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
500 if (ret < 0)
501 goto free_path;
502 if (ret > 0) {
503 if (path->slots[0] >=
504 btrfs_header_nritems(path->nodes[0])) {
505 ret = btrfs_next_leaf(root, path);
506 if (ret < 0)
507 goto free_path;
508 if (ret > 0) {
509 ret = 0;
510 goto free_path;
511 }
512 } else {
513 ret = 0;
514 }
515 }
516
517 while (1) {
518 struct extent_buffer *l = path->nodes[0];
519 int slot = path->slots[0];
520
521 btrfs_item_key_to_cpu(l, &found_key, slot);
522
523 if (found_key.objectid != src_dev->devid)
524 break;
525
526 if (found_key.type != BTRFS_DEV_EXTENT_KEY)
527 break;
528
529 if (found_key.offset < key.offset)
530 break;
531
532 dev_extent = btrfs_item_ptr(l, slot, struct btrfs_dev_extent);
533 length = btrfs_dev_extent_length(l, dev_extent);
534
535 chunk_offset = btrfs_dev_extent_chunk_offset(l, dev_extent);
536
537 cache = btrfs_lookup_block_group(fs_info, chunk_offset);
538 if (!cache)
539 goto skip;
540
541 spin_lock(&cache->lock);
542 cache->to_copy = 1;
543 spin_unlock(&cache->lock);
544
545 btrfs_put_block_group(cache);
546
547 skip:
548 ret = btrfs_next_item(root, path);
549 if (ret != 0) {
550 if (ret > 0)
551 ret = 0;
552 break;
553 }
554 }
555
556 free_path:
557 btrfs_free_path(path);
558 unlock:
559 mutex_unlock(&fs_info->chunk_mutex);
560
561 return ret;
562 }
563

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all(a)lists.01.org

#106 lore raw

Re: [PATCH v10 31/41] btrfs: mark block groups to copy for device-replace

In reply to Naohiro Aota
I love your patch! Perhaps something to improve:

Hi Naohiro,

I love your patch! Perhaps something to improve:

[auto build test WARNING on xfs-linux/for-next]
[also build test WARNING on v5.10-rc3]
[cannot apply to kdave/for-next block/for-next next-20201110] [If your patch is applied to the wrong git tree, kindly drop us a note. And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]

url:    https://github.com/0day-ci/linux/commits/Naohiro-Aota/btrfs-zoned-block-device-support/20201110-193227
base:   https://git.kernel.org/pub/scm/fs/xfs/xfs-linux.git for-next
config: arc-allyesconfig (attached as .config)
compiler: arceb-elf-gcc (GCC) 9.3.0

reproduce (this is a W=1 build):
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross chmod +x ~/bin/make.cross
# https://github.com/0day-ci/linux/commit/355d6db6997f390f363f08fa9bfbf8b38eb891a2
git remote add linux-review https://github.com/0day-ci/linux
git fetch --no-tags linux-review Naohiro-Aota/btrfs-zoned-block-device-support/20201110-193227 git checkout 355d6db6997f390f363f08fa9bfbf8b38eb891a2
# save the attached .config to linux build tree
COMPILER_INSTALL_PATH=$HOME/0day COMPILER=gcc-9.3.0 make.cross ARCH=arc

If you fix the issue, kindly add following tag as appropriate

All warnings (new ones prefixed by >>):

fs/btrfs/dev-replace.c: In function 'mark_block_group_to_copy':

>> fs/btrfs/dev-replace.c:452:20: warning: variable 'length' set but not used [-Wunused-but-set-variable]
     452 |  u64 chunk_offset, length;
         |                    ^~~~~~

vim +/length +452 fs/btrfs/dev-replace.c

440
441 static int mark_block_group_to_copy(struct btrfs_fs_info *fs_info,
442 struct btrfs_device *src_dev)
443 {
444 struct btrfs_path *path;
445 struct btrfs_key key;
446 struct btrfs_key found_key;
447 struct btrfs_root *root = fs_info->dev_root;
448 struct btrfs_dev_extent *dev_extent = NULL;
449 struct btrfs_block_group *cache;
450 struct btrfs_trans_handle *trans;
451 int ret = 0;

 > 452		u64 chunk_offset, length;

453
454 /* Do not use "to_copy" on non-ZONED for now */
455 if (!btrfs_is_zoned(fs_info))
456 return 0;
457
458 mutex_lock(&fs_info->chunk_mutex);
459
460 /* Ensure we don't have pending new block group */
461 spin_lock(&fs_info->trans_lock);
462 while (fs_info->running_transaction &&
463 !list_empty(&fs_info->running_transaction->dev_update_list)) {
464 spin_unlock(&fs_info->trans_lock);
465 mutex_unlock(&fs_info->chunk_mutex);
466 trans = btrfs_attach_transaction(root);
467 if (IS_ERR(trans)) {
468 ret = PTR_ERR(trans);
469 mutex_lock(&fs_info->chunk_mutex);
470 if (ret == -ENOENT)
471 continue;
472 else
473 goto unlock;
474 }
475
476 ret = btrfs_commit_transaction(trans);
477 mutex_lock(&fs_info->chunk_mutex);
478 if (ret)
479 goto unlock;
480
481 spin_lock(&fs_info->trans_lock);
482 }
483 spin_unlock(&fs_info->trans_lock);
484
485 path = btrfs_alloc_path();
486 if (!path) {
487 ret = -ENOMEM;
488 goto unlock;
489 }
490
491 path->reada = READA_FORWARD;
492 path->search_commit_root = 1;
493 path->skip_locking = 1;
494
495 key.objectid = src_dev->devid;
496 key.offset = 0;
497 key.type = BTRFS_DEV_EXTENT_KEY;
498
499 ret = btrfs_search_slot(NULL, root, &key, path, 0, 0);
500 if (ret < 0)
501 goto free_path;
502 if (ret > 0) {
503 if (path->slots[0] >=
504 btrfs_header_nritems(path->nodes[0])) {
505 ret = btrfs_next_leaf(root, path);
506 if (ret < 0)
507 goto free_path;
508 if (ret > 0) {
509 ret = 0;
510 goto free_path;
511 }
512 } else {
513 ret = 0;
514 }
515 }
516
517 while (1) {
518 struct extent_buffer *l = path->nodes[0];
519 int slot = path->slots[0];
520
521 btrfs_item_key_to_cpu(l, &found_key, slot);
522
523 if (found_key.objectid != src_dev->devid)
524 break;
525
526 if (found_key.type != BTRFS_DEV_EXTENT_KEY)
527 break;
528
529 if (found_key.offset < key.offset)
530 break;
531
532 dev_extent = btrfs_item_ptr(l, slot, struct btrfs_dev_extent);
533 length = btrfs_dev_extent_length(l, dev_extent);
534
535 chunk_offset = btrfs_dev_extent_chunk_offset(l, dev_extent);
536
537 cache = btrfs_lookup_block_group(fs_info, chunk_offset);
538 if (!cache)
539 goto skip;
540
541 spin_lock(&cache->lock);
542 cache->to_copy = 1;
543 spin_unlock(&cache->lock);
544
545 btrfs_put_block_group(cache);
546
547 skip:
548 ret = btrfs_next_item(root, path);
549 if (ret != 0) {
550 if (ret > 0)
551 ret = 0;
552 break;
553 }
554 }
555
556 free_path:
557 btrfs_free_path(path);
558 unlock:
559 mutex_unlock(&fs_info->chunk_mutex);
560
561 return ret;
562 }
563

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all(a)lists.01.org

#107 lore raw

PATCH v10 32/41 btrfs: implement cloning for ZONED device-replace

In reply to Naohiro Aota

This is 2/4 patch to implement device-replace for ZONED mode.

On zoned mode, a block group must be either copied (from the source device to the destination device) or cloned (to the both device).

This commit implements the cloning part. If a block group targeted by an IO is marked to copy, we should not clone the IO to the destination device, because the block group is eventually copied by the replace process.

This commit also handles cloning of device reset.

---
 fs/btrfs/extent-tree.c | 57 +++++++++++++++++++++++++++++++-----------
 fs/btrfs/scrub.c       |  2 +-
 fs/btrfs/volumes.c     | 33 ++++++++++++++++++++++--
 fs/btrfs/zoned.c       | 11 ++++++++
 4 files changed, 85 insertions(+), 18 deletions(-)
fs/btrfs/extent-tree.c +42 -15
index 99640dacf8e6..2ee21076b641 100644
@@ -35,6 +35,7 @@
 #include "discard.h"
 #include "rcu-string.h"
 #include "zoned.h"
+#include "dev-replace.h"
 
 #undef SCRAMBLE_DELAYED_REFS
 
@@ -1298,6 +1299,46 @@ static int btrfs_issue_discard(struct block_device *bdev, u64 start, u64 len,
 	return ret;
 }
 
+static int do_discard_extent(struct btrfs_bio_stripe *stripe, u64 *bytes)
+{
+	struct btrfs_device *dev = stripe->dev;
+	struct btrfs_fs_info *fs_info = dev->fs_info;
+	struct btrfs_dev_replace *dev_replace = &fs_info->dev_replace;
+	u64 phys = stripe->physical;
+	u64 len = stripe->length;
+	u64 discarded = 0;
+	int ret = 0;
+
+	/* Zone reset in ZONED mode */
+	if (btrfs_can_zone_reset(dev, phys, len)) {
+		u64 src_disc;
+
+		ret = btrfs_reset_device_zone(dev, phys, len, &discarded);
+		if (ret)
+			goto out;
+
+		if (!btrfs_dev_replace_is_ongoing(dev_replace) ||
+		    dev != dev_replace->srcdev)
+			goto out;
+
+		src_disc = discarded;
+
+		/* send to replace target as well */
+		ret = btrfs_reset_device_zone(dev_replace->tgtdev, phys, len,
+					      &discarded);
+		discarded += src_disc;
+	} else if (blk_queue_discard(bdev_get_queue(stripe->dev->bdev))) {
+		ret = btrfs_issue_discard(dev->bdev, phys, len, &discarded);
+	} else {
+		ret = 0;
+		*bytes = 0;
+	}
+
+out:
+	*bytes = discarded;
+	return ret;
+}
+
 int btrfs_discard_extent(struct btrfs_fs_info *fs_info, u64 bytenr,
 			 u64 num_bytes, u64 *actual_bytes)
 {
@@ -1331,28 +1372,14 @@ int btrfs_discard_extent(struct btrfs_fs_info *fs_info, u64 bytenr,
 
 		stripe = bbio->stripes;
 		for (i = 0; i < bbio->num_stripes; i++, stripe++) {
-			struct btrfs_device *dev = stripe->dev;
-			u64 physical = stripe->physical;
-			u64 length = stripe->length;
 			u64 bytes;
-			struct request_queue *req_q;
 
 			if (!stripe->dev->bdev) {
 				ASSERT(btrfs_test_opt(fs_info, DEGRADED));
 				continue;
 			}
 
-			req_q = bdev_get_queue(stripe->dev->bdev);
-			/* Zone reset in ZONED mode */
-			if (btrfs_can_zone_reset(dev, physical, length))
-				ret = btrfs_reset_device_zone(dev, physical,
-							      length, &bytes);
-			else if (blk_queue_discard(req_q))
-				ret = btrfs_issue_discard(dev->bdev, physical,
-							  length, &bytes);
-			else
-				continue;
-
+			ret = do_discard_extent(stripe, &bytes);
 			if (!ret) {
 				discarded_bytes += bytes;
 			} else if (ret != -EOPNOTSUPP) {
fs/btrfs/scrub.c +1 -1
index d0d7db3c8b0b..371bb6437cab 100644
@@ -3501,7 +3501,7 @@ int scrub_enumerate_chunks(struct scrub_ctx *sctx,
 			goto skip;
 
 
-		if (sctx->is_dev_replace && btrfs_fs_incompat(fs_info, ZONED)) {
+		if (sctx->is_dev_replace && btrfs_is_zoned(fs_info)) {
 			spin_lock(&cache->lock);
 			if (!cache->to_copy) {
 				spin_unlock(&cache->lock);
fs/btrfs/volumes.c +31 -2
index c8187d704c89..434fc6f758cc 100644
@@ -5969,9 +5969,29 @@ static int get_extra_mirror_from_replace(struct btrfs_fs_info *fs_info,
 	return ret;
 }
 
+static bool is_block_group_to_copy(struct btrfs_fs_info *fs_info, u64 logical)
+{
+	struct btrfs_block_group *cache;
+	bool ret;
+
+	/* non-ZONED mode does not use "to_copy" flag */
+	if (!btrfs_is_zoned(fs_info))
+		return false;
+
+	cache = btrfs_lookup_block_group(fs_info, logical);
+
+	spin_lock(&cache->lock);
+	ret = cache->to_copy;
+	spin_unlock(&cache->lock);
+
+	btrfs_put_block_group(cache);
+	return ret;
+}
+
 static void handle_ops_on_dev_replace(enum btrfs_map_op op,
 				      struct btrfs_bio **bbio_ret,
 				      struct btrfs_dev_replace *dev_replace,
+				      u64 logical,
 				      int *num_stripes_ret, int *max_errors_ret)
 {
 	struct btrfs_bio *bbio = *bbio_ret;
@@ -5984,6 +6004,15 @@ static void handle_ops_on_dev_replace(enum btrfs_map_op op,
 	if (op == BTRFS_MAP_WRITE) {
 		int index_where_to_add;
 
+		/*
+		 * a block group which have "to_copy" set will
+		 * eventually copied by dev-replace process. We can
+		 * avoid cloning IO here.
+		 */
+		if (is_block_group_to_copy(dev_replace->srcdev->fs_info,
+					   logical))
+			return;
+
 		/*
 		 * duplicate the write operations while the dev replace
 		 * procedure is running. Since the copying of the old disk to
@@ -6379,8 +6408,8 @@ static int __btrfs_map_block(struct btrfs_fs_info *fs_info,
 
 	if (dev_replace_is_ongoing && dev_replace->tgtdev != NULL &&
 	    need_full_stripe(op)) {
-		handle_ops_on_dev_replace(op, &bbio, dev_replace, &num_stripes,
-					  &max_errors);
+		handle_ops_on_dev_replace(op, &bbio, dev_replace, logical,
+					  &num_stripes, &max_errors);
 	}
 
 	*bbio_ret = bbio;
fs/btrfs/zoned.c +11 -0
index d345c07f5fdf..8bf5df03ceb8 100644
@@ -11,6 +11,7 @@
 #include "disk-io.h"
 #include "block-group.h"
 #include "transaction.h"
+#include "dev-replace.h"
 
 /* Maximum number of zones to report per blkdev_report_zones() call */
 #define BTRFS_REPORT_NR_ZONES   4096
@@ -891,6 +892,8 @@ int btrfs_load_block_group_zone_info(struct btrfs_block_group *cache)
 	for (i = 0; i < map->num_stripes; i++) {
 		bool is_sequential;
 		struct blk_zone zone;
+		struct btrfs_dev_replace *dev_replace = &fs_info->dev_replace;
+		int dev_replace_is_ongoing = 0;
 
 		device = map->stripes[i].dev;
 		physical = map->stripes[i].physical;
@@ -917,6 +920,14 @@ int btrfs_load_block_group_zone_info(struct btrfs_block_group *cache)
 		 */
 		btrfs_dev_clear_zone_empty(device, physical);
 
+		down_read(&dev_replace->rwsem);
+		dev_replace_is_ongoing =
+			btrfs_dev_replace_is_ongoing(dev_replace);
+		if (dev_replace_is_ongoing && dev_replace->tgtdev != NULL)
+			btrfs_dev_clear_zone_empty(dev_replace->tgtdev,
+						   physical);
+		up_read(&dev_replace->rwsem);
+
 		/*
 		 * The group is mapped to a sequential zone. Get the zone write
 		 * pointer to determine the allocation offset within the zone.
Signature · 1 line
2.27.0
#108 lore raw

PATCH v10 33/41 btrfs: implement copying for ZONED device-replace

In reply to Naohiro Aota

This is 3/4 patch to implement device-replace on ZONED mode.

This commit implement copying. So, it track the write pointer during device replace process. Device-replace's copying is smart to copy only used extents on source device, we have to fill the gap to honor the sequential write rule in the target device.

Device-replace process in ZONED mode must copy or clone all the extents in the source device exactly once. So, we need to use to ensure allocations started just before the dev-replace process to have their corresponding extent information in the B-trees. finish_extent_writes_for_zoned() implements that functionality, which basically is the removed code in the commit 042528f8d840 ("Btrfs: fix block group remaining RO forever after error during device replace").

---
 fs/btrfs/scrub.c | 86 ++++++++++++++++++++++++++++++++++++++++++++++++
 fs/btrfs/zoned.c | 12 +++++++
 fs/btrfs/zoned.h |  8 +++++
 3 files changed, 106 insertions(+)
fs/btrfs/scrub.c +86 -0
index 371bb6437cab..aaf7882dee06 100644
@@ -169,6 +169,7 @@ struct scrub_ctx {
 	int			pages_per_rd_bio;
 
 	int			is_dev_replace;
+	u64			write_pointer;
 
 	struct scrub_bio        *wr_curr_bio;
 	struct mutex            wr_lock;
@@ -1623,6 +1624,25 @@ static int scrub_write_page_to_dev_replace(struct scrub_block *sblock,
 	return scrub_add_page_to_wr_bio(sblock->sctx, spage);
 }
 
+static int fill_writer_pointer_gap(struct scrub_ctx *sctx, u64 physical)
+{
+	int ret = 0;
+	u64 length;
+
+	if (!btrfs_is_zoned(sctx->fs_info))
+		return 0;
+
+	if (sctx->write_pointer < physical) {
+		length = physical - sctx->write_pointer;
+
+		ret = btrfs_zoned_issue_zeroout(sctx->wr_tgtdev,
+						sctx->write_pointer, length);
+		if (!ret)
+			sctx->write_pointer = physical;
+	}
+	return ret;
+}
+
 static int scrub_add_page_to_wr_bio(struct scrub_ctx *sctx,
 				    struct scrub_page *spage)
 {
@@ -1645,6 +1665,13 @@ static int scrub_add_page_to_wr_bio(struct scrub_ctx *sctx,
 	if (sbio->page_count == 0) {
 		struct bio *bio;
 
+		ret = fill_writer_pointer_gap(sctx,
+					      spage->physical_for_dev_replace);
+		if (ret) {
+			mutex_unlock(&sctx->wr_lock);
+			return ret;
+		}
+
 		sbio->physical = spage->physical_for_dev_replace;
 		sbio->logical = spage->logical;
 		sbio->dev = sctx->wr_tgtdev;
@@ -1706,6 +1733,10 @@ static void scrub_wr_submit(struct scrub_ctx *sctx)
 	 * doubled the write performance on spinning disks when measured
 	 * with Linux 3.5 */
 	btrfsic_submit_bio(sbio->bio);
+
+	if (btrfs_is_zoned(sctx->fs_info))
+		sctx->write_pointer = sbio->physical +
+			sbio->page_count * PAGE_SIZE;
 }
 
 static void scrub_wr_bio_end_io(struct bio *bio)
@@ -2973,6 +3004,21 @@ static noinline_for_stack int scrub_raid56_parity(struct scrub_ctx *sctx,
 	return ret < 0 ? ret : 0;
 }
 
+static void sync_replace_for_zoned(struct scrub_ctx *sctx)
+{
+	if (!btrfs_is_zoned(sctx->fs_info))
+		return;
+
+	sctx->flush_all_writes = true;
+	scrub_submit(sctx);
+	mutex_lock(&sctx->wr_lock);
+	scrub_wr_submit(sctx);
+	mutex_unlock(&sctx->wr_lock);
+
+	wait_event(sctx->list_wait,
+		   atomic_read(&sctx->bios_in_flight) == 0);
+}
+
 static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx,
 					   struct map_lookup *map,
 					   struct btrfs_device *scrub_dev,
@@ -3105,6 +3151,14 @@ static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx,
 	 */
 	blk_start_plug(&plug);
 
+	if (sctx->is_dev_replace &&
+	    btrfs_dev_is_sequential(sctx->wr_tgtdev, physical)) {
+		mutex_lock(&sctx->wr_lock);
+		sctx->write_pointer = physical;
+		mutex_unlock(&sctx->wr_lock);
+		sctx->flush_all_writes = true;
+	}
+
 	/*
 	 * now find all extents for each stripe and scrub them
 	 */
@@ -3292,6 +3346,9 @@ static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx,
 			if (ret)
 				goto out;
 
+			if (sctx->is_dev_replace)
+				sync_replace_for_zoned(sctx);
+
 			if (extent_logical + extent_len <
 			    key.objectid + bytes) {
 				if (map->type & BTRFS_BLOCK_GROUP_RAID56_MASK) {
@@ -3414,6 +3471,25 @@ static noinline_for_stack int scrub_chunk(struct scrub_ctx *sctx,
 	return ret;
 }
 
+static int finish_extent_writes_for_zoned(struct btrfs_root *root,
+					  struct btrfs_block_group *cache)
+{
+	struct btrfs_fs_info *fs_info = cache->fs_info;
+	struct btrfs_trans_handle *trans;
+
+	if (!btrfs_is_zoned(fs_info))
+		return 0;
+
+	btrfs_wait_block_group_reservations(cache);
+	btrfs_wait_nocow_writers(cache);
+	btrfs_wait_ordered_roots(fs_info, U64_MAX, cache->start, cache->length);
+
+	trans = btrfs_join_transaction(root);
+	if (IS_ERR(trans))
+		return PTR_ERR(trans);
+	return btrfs_commit_transaction(trans);
+}
+
 static noinline_for_stack
 int scrub_enumerate_chunks(struct scrub_ctx *sctx,
 			   struct btrfs_device *scrub_dev, u64 start, u64 end)
@@ -3569,6 +3645,16 @@ int scrub_enumerate_chunks(struct scrub_ctx *sctx,
 		 * group is not RO.
 		 */
 		ret = btrfs_inc_block_group_ro(cache, sctx->is_dev_replace);
+		if (!ret && sctx->is_dev_replace) {
+			ret = finish_extent_writes_for_zoned(root, cache);
+			if (ret) {
+				btrfs_dec_block_group_ro(cache);
+				scrub_pause_off(fs_info);
+				btrfs_put_block_group(cache);
+				break;
+			}
+		}
+
 		if (ret == 0) {
 			ro_set = 1;
 		} else if (ret == -ENOSPC && !sctx->is_dev_replace) {
fs/btrfs/zoned.c +12 -0
index 8bf5df03ceb8..cce4ddfff5d2 100644
@@ -1189,3 +1189,15 @@ void btrfs_revert_meta_write_pointer(struct btrfs_block_group *cache,
 	ASSERT(cache->meta_write_pointer == eb->start + eb->len);
 	cache->meta_write_pointer = eb->start;
 }
+
+int btrfs_zoned_issue_zeroout(struct btrfs_device *device, u64 physical,
+			      u64 length)
+{
+	if (!btrfs_dev_is_sequential(device, physical))
+		return -EOPNOTSUPP;
+
+	return blkdev_issue_zeroout(device->bdev,
+				    physical >> SECTOR_SHIFT,
+				    length >> SECTOR_SHIFT,
+				    GFP_NOFS, 0);
+}
fs/btrfs/zoned.h +8 -0
index 41d786a97e40..40204f8310ca 100644
@@ -53,6 +53,8 @@ bool btrfs_check_meta_write_pointer(struct btrfs_fs_info *fs_info,
 				    struct btrfs_block_group **cache_ret);
 void btrfs_revert_meta_write_pointer(struct btrfs_block_group *cache,
 				     struct extent_buffer *eb);
+int btrfs_zoned_issue_zeroout(struct btrfs_device *device, u64 physical,
+			      u64 length);
 #else /* CONFIG_BLK_DEV_ZONED */
 static inline int btrfs_get_dev_zone(struct btrfs_device *device, u64 pos,
 				     struct blk_zone *zone)
@@ -158,6 +160,12 @@ static inline void btrfs_revert_meta_write_pointer(
 {
 }
 
+static inline int btrfs_zoned_issue_zeroout(struct btrfs_device *device,
+					    u64 physical, u64 length)
+{
+	return -EOPNOTSUPP;
+}
+
 #endif
 
 static inline bool btrfs_dev_is_sequential(struct btrfs_device *device, u64 pos)
Signature · 1 line
2.27.0
#109 lore raw

PATCH v10 34/41 btrfs: support dev-replace in ZONED mode

In reply to Naohiro Aota

This is 4/4 patch to implement device-replace on ZONED mode.

Even after the copying is done, the write pointers of the source device and the destination device may not be synchronized. For example, when the last allocated extent is freed before device-replace process, the extent is not copied, leaving a hole there.

This patch synchronize the write pointers by writing zeros to the destination device.

---
 fs/btrfs/scrub.c | 36 +++++++++++++++++++++++++
 fs/btrfs/zoned.c | 69 ++++++++++++++++++++++++++++++++++++++++++++++++
 fs/btrfs/zoned.h |  9 +++++++
 3 files changed, 114 insertions(+)
fs/btrfs/scrub.c +36 -0
index aaf7882dee06..0e2211b9c810 100644
@@ -3019,6 +3019,31 @@ static void sync_replace_for_zoned(struct scrub_ctx *sctx)
 		   atomic_read(&sctx->bios_in_flight) == 0);
 }
 
+static int sync_write_pointer_for_zoned(struct scrub_ctx *sctx, u64 logical,
+					u64 physical, u64 physical_end)
+{
+	struct btrfs_fs_info *fs_info = sctx->fs_info;
+	int ret = 0;
+
+	if (!btrfs_is_zoned(fs_info))
+		return 0;
+
+	wait_event(sctx->list_wait, atomic_read(&sctx->bios_in_flight) == 0);
+
+	mutex_lock(&sctx->wr_lock);
+	if (sctx->write_pointer < physical_end) {
+		ret = btrfs_sync_zone_write_pointer(sctx->wr_tgtdev, logical,
+						    physical,
+						    sctx->write_pointer);
+		if (ret)
+			btrfs_err(fs_info, "failed to recover write pointer");
+	}
+	mutex_unlock(&sctx->wr_lock);
+	btrfs_dev_clear_zone_empty(sctx->wr_tgtdev, physical);
+
+	return ret;
+}
+
 static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx,
 					   struct map_lookup *map,
 					   struct btrfs_device *scrub_dev,
@@ -3416,6 +3441,17 @@ static noinline_for_stack int scrub_stripe(struct scrub_ctx *sctx,
 	blk_finish_plug(&plug);
 	btrfs_free_path(path);
 	btrfs_free_path(ppath);
+
+	if (sctx->is_dev_replace && ret >= 0) {
+		int ret2;
+
+		ret2 = sync_write_pointer_for_zoned(sctx, base + offset,
+						    map->stripes[num].physical,
+						    physical_end);
+		if (ret2)
+			ret = ret2;
+	}
+
 	return ret < 0 ? ret : 0;
 }
 
fs/btrfs/zoned.c +69 -0

78 lines not on this page · raw message

fs/btrfs/zoned.h +9 -0
index 40204f8310ca..5b61500a0aa9 100644
@@ -55,6 +55,8 @@ void btrfs_revert_meta_write_pointer(struct btrfs_block_group *cache,
 				     struct extent_buffer *eb);
 int btrfs_zoned_issue_zeroout(struct btrfs_device *device, u64 physical,
 			      u64 length);
+int btrfs_sync_zone_write_pointer(struct btrfs_device *tgt_dev, u64 logical,
+				  u64 physical_start, u64 physical_pos);
 #else /* CONFIG_BLK_DEV_ZONED */
 static inline int btrfs_get_dev_zone(struct btrfs_device *device, u64 pos,
 				     struct blk_zone *zone)
@@ -166,6 +168,13 @@ static inline int btrfs_zoned_issue_zeroout(struct btrfs_device *device,
 	return -EOPNOTSUPP;
 }
 
+static inline int btrfs_sync_zone_write_pointer(struct btrfs_device *tgt_dev,
+						u64 logical, u64 physical_start,
+						u64 physical_pos)
+{
+	return -EOPNOTSUPP;
+}
+
 #endif
 
 static inline bool btrfs_dev_is_sequential(struct btrfs_device *device, u64 pos)
Signature · 1 line
2.27.0
#110 lore raw

PATCH v10 35/41 btrfs: enable relocation in ZONED mode

In reply to Naohiro Aota

To serialize allocation and submit_bio, we introduced mutex around them. As a result, preallocation must be completely disabled to avoid a deadlock.

Since current relocation process relies on preallocation to move file data extents, it must be handled in another way. In ZONED mode, we just truncate the inode to the size that we wanted to pre-allocate. Then, we flush dirty pages on the file before finishing relocation process. run_delalloc_zoned() will handle all the allocation and submit IOs to the underlying layers.

---
 fs/btrfs/relocation.c | 35 +++++++++++++++++++++++++++++++++--
 1 file changed, 33 insertions(+), 2 deletions(-)
fs/btrfs/relocation.c +33 -2
index 3602806d71bd..44b697b881b6 100644
@@ -2603,6 +2603,32 @@ static noinline_for_stack int prealloc_file_extent_cluster(
 	if (ret)
 		return ret;
 
+	/*
+	 * In ZONED mode, we cannot preallocate the file region. Instead, we
+	 * dirty and fiemap_write the region.
+	 */
+
+	if (btrfs_is_zoned(inode->root->fs_info)) {
+		struct btrfs_root *root = inode->root;
+		struct btrfs_trans_handle *trans;
+
+		end = cluster->end - offset + 1;
+		trans = btrfs_start_transaction(root, 1);
+		if (IS_ERR(trans))
+			return PTR_ERR(trans);
+
+		inode->vfs_inode.i_ctime = current_time(&inode->vfs_inode);
+		i_size_write(&inode->vfs_inode, end);
+		ret = btrfs_update_inode(trans, root, &inode->vfs_inode);
+		if (ret) {
+			btrfs_abort_transaction(trans, ret);
+			btrfs_end_transaction(trans);
+			return ret;
+		}
+
+		return btrfs_end_transaction(trans);
+	}
+
 	inode_lock(&inode->vfs_inode);
 	for (nr = 0; nr < cluster->nr; nr++) {
 		start = cluster->boundary[nr] - offset;
@@ -2799,6 +2825,8 @@ static int relocate_file_extent_cluster(struct inode *inode,
 		}
 	}
 	WARN_ON(nr != cluster->nr);
+	if (btrfs_is_zoned(fs_info) && !ret)
+		ret = btrfs_wait_ordered_range(inode, 0, (u64)-1);
 out:
 	kfree(ra);
 	return ret;
@@ -3434,8 +3462,12 @@ static int __insert_orphan_inode(struct btrfs_trans_handle *trans,
 	struct btrfs_path *path;
 	struct btrfs_inode_item *item;
 	struct extent_buffer *leaf;
+	u64 flags = BTRFS_INODE_NOCOMPRESS | BTRFS_INODE_PREALLOC;
 	int ret;
 
+	if (btrfs_is_zoned(trans->fs_info))
+		flags &= ~BTRFS_INODE_PREALLOC;
+
 	path = btrfs_alloc_path();
 	if (!path)
 		return -ENOMEM;
@@ -3450,8 +3482,7 @@ static int __insert_orphan_inode(struct btrfs_trans_handle *trans,
 	btrfs_set_inode_generation(leaf, item, 1);
 	btrfs_set_inode_size(leaf, item, 0);
 	btrfs_set_inode_mode(leaf, item, S_IFREG | 0600);
-	btrfs_set_inode_flags(leaf, item, BTRFS_INODE_NOCOMPRESS |
-					  BTRFS_INODE_PREALLOC);
+	btrfs_set_inode_flags(leaf, item, flags);
 	btrfs_mark_buffer_dirty(leaf);
 out:
 	btrfs_free_path(path);
Signature · 1 line
2.27.0
#111 lore raw

PATCH v10 36/41 btrfs: relocate block group to repair IO failure in ZONED

In reply to Naohiro Aota

When btrfs find a checksum error and if the file system has a mirror of the damaged data, btrfs read the correct data from the mirror and write the data to damaged blocks. This repairing, however, is against the sequential write required rule.

We can consider three methods to repair an IO failure in ZONED mode:
(1) Reset and rewrite the damaged zone
(2) Allocate new device extent and replace the damaged device extent to the new extent
(3) Relocate the corresponding block group

Method (1) is most similar to a behavior done with regular devices. However, it also wipes non-damaged data in the same device extent, and so it unnecessary degrades non-damaged data.

Method (2) is much like device replacing but done in the same device. It is safe because it keeps the device extent until the replacing finish. However, extending device replacing is non-trivial. It assumes "src_dev->physical == dst_dev->physical". Also, the extent mapping replacing function should be extended to support replacing device extent position in one device.

Method (3) invokes relocation of the damaged block group, so it is straightforward to implement. It relocates all the mirrored device extents, so it is, potentially, a more costly operation than method (1) or (2). But it relocates only using extents which reduce the total IO size.

Let's apply method (3) for now. In the future, we can extend device-replace and apply method (2).

For protecting a block group gets relocated multiple time with multiple IO errors, this commit introduces "relocating_repair" bit to show it's now relocating to repair IO failures. Also it uses a new kthread "btrfs-relocating-repair", not to block IO path with relocating process.

This commit also supports repairing in the scrub process.

---
 fs/btrfs/block-group.h |  1 +
 fs/btrfs/extent_io.c   |  3 ++
 fs/btrfs/scrub.c       |  3 ++
 fs/btrfs/volumes.c     | 71 ++++++++++++++++++++++++++++++++++++++++++
 fs/btrfs/volumes.h     |  1 +
 5 files changed, 79 insertions(+)
fs/btrfs/block-group.h +1 -0
index ccbcf37eae9c..25f67fe24746 100644
@@ -96,6 +96,7 @@ struct btrfs_block_group {
 	unsigned int has_caching_ctl:1;
 	unsigned int removed:1;
 	unsigned int to_copy:1;
+	unsigned int relocating_repair:1;
 
 	int disk_cache_state;
 
fs/btrfs/extent_io.c +3 -0
index d26c827f39c6..c11cf531ba86 100644
@@ -2268,6 +2268,9 @@ int repair_io_failure(struct btrfs_fs_info *fs_info, u64 ino, u64 start,
 	ASSERT(!(fs_info->sb->s_flags & SB_RDONLY));
 	BUG_ON(!mirror_num);
 
+	if (btrfs_is_zoned(fs_info))
+		return btrfs_repair_one_zone(fs_info, logical);
+
 	bio = btrfs_io_bio_alloc(1);
 	bio->bi_iter.bi_size = 0;
 	map_length = length;
fs/btrfs/scrub.c +3 -0
index 0e2211b9c810..e6a8df8a8f4f 100644
@@ -861,6 +861,9 @@ static int scrub_handle_errored_block(struct scrub_block *sblock_to_check)
 	have_csum = sblock_to_check->pagev[0]->have_csum;
 	dev = sblock_to_check->pagev[0]->dev;
 
+	if (btrfs_is_zoned(fs_info) && !sctx->is_dev_replace)
+		return btrfs_repair_one_zone(fs_info, logical);
+
 	/*
 	 * We must use GFP_NOFS because the scrub task might be waiting for a
 	 * worker task executing this function and in turn a transaction commit
fs/btrfs/volumes.c +71 -0

74 lines not on this page · raw message

fs/btrfs/volumes.h +1 -0
index cff1f7689eac..7c1ad6901791 100644
@@ -584,5 +584,6 @@ void btrfs_scratch_superblocks(struct btrfs_fs_info *fs_info,
 int btrfs_bg_type_to_factor(u64 flags);
 const char *btrfs_bg_type_to_raid_name(u64 flags);
 int btrfs_verify_dev_extents(struct btrfs_fs_info *fs_info);
+int btrfs_repair_one_zone(struct btrfs_fs_info *fs_info, u64 logical);
 
 #endif
Signature · 1 line
2.27.0
#112 lore raw

PATCH v10 37/41 btrfs: split alloc_log_tree()

In reply to Naohiro Aota

This is a preparation for the next patch. This commit split alloc_log_tree() to allocating tree structure part (remains in alloc_log_tree()) and allocating tree node part (moved in btrfs_alloc_log_tree_node()). The latter part is also exported to be used in the next patch.

---
 fs/btrfs/disk-io.c | 33 +++++++++++++++++++++++++++------
 fs/btrfs/disk-io.h |  2 ++
 2 files changed, 29 insertions(+), 6 deletions(-)
fs/btrfs/disk-io.c +27 -6
index 9490dbbbdb2a..97e3deb46cf1 100644
@@ -1211,7 +1211,6 @@ static struct btrfs_root *alloc_log_tree(struct btrfs_trans_handle *trans,
 					 struct btrfs_fs_info *fs_info)
 {
 	struct btrfs_root *root;
-	struct extent_buffer *leaf;
 
 	root = btrfs_alloc_root(fs_info, BTRFS_TREE_LOG_OBJECTID, GFP_NOFS);
 	if (!root)
@@ -1221,6 +1220,14 @@ static struct btrfs_root *alloc_log_tree(struct btrfs_trans_handle *trans,
 	root->root_key.type = BTRFS_ROOT_ITEM_KEY;
 	root->root_key.offset = BTRFS_TREE_LOG_OBJECTID;
 
+	return root;
+}
+
+int btrfs_alloc_log_tree_node(struct btrfs_trans_handle *trans,
+			      struct btrfs_root *root)
+{
+	struct extent_buffer *leaf;
+
 	/*
 	 * DON'T set SHAREABLE bit for log trees.
 	 *
@@ -1233,26 +1240,33 @@ static struct btrfs_root *alloc_log_tree(struct btrfs_trans_handle *trans,
 
 	leaf = btrfs_alloc_tree_block(trans, root, 0, BTRFS_TREE_LOG_OBJECTID,
 			NULL, 0, 0, 0, BTRFS_NESTING_NORMAL);
-	if (IS_ERR(leaf)) {
-		btrfs_put_root(root);
-		return ERR_CAST(leaf);
-	}
+	if (IS_ERR(leaf))
+		return PTR_ERR(leaf);
 
 	root->node = leaf;
 
 	btrfs_mark_buffer_dirty(root->node);
 	btrfs_tree_unlock(root->node);
-	return root;
+
+	return 0;
 }
 
 int btrfs_init_log_root_tree(struct btrfs_trans_handle *trans,
 			     struct btrfs_fs_info *fs_info)
 {
 	struct btrfs_root *log_root;
+	int ret;
 
 	log_root = alloc_log_tree(trans, fs_info);
 	if (IS_ERR(log_root))
 		return PTR_ERR(log_root);
+
+	ret = btrfs_alloc_log_tree_node(trans, log_root);
+	if (ret) {
+		btrfs_put_root(log_root);
+		return ret;
+	}
+
 	WARN_ON(fs_info->log_root_tree);
 	fs_info->log_root_tree = log_root;
 	return 0;
@@ -1264,11 +1278,18 @@ int btrfs_add_log_tree(struct btrfs_trans_handle *trans,
 	struct btrfs_fs_info *fs_info = root->fs_info;
 	struct btrfs_root *log_root;
 	struct btrfs_inode_item *inode_item;
+	int ret;
 
 	log_root = alloc_log_tree(trans, fs_info);
 	if (IS_ERR(log_root))
 		return PTR_ERR(log_root);
 
+	ret = btrfs_alloc_log_tree_node(trans, log_root);
+	if (ret) {
+		btrfs_put_root(log_root);
+		return ret;
+	}
+
 	log_root->last_trans = trans->transid;
 	log_root->root_key.offset = root->root_key.objectid;
 
fs/btrfs/disk-io.h +2 -0
index fee69ced58b4..b82ae3711c42 100644
@@ -115,6 +115,8 @@ blk_status_t btrfs_wq_submit_bio(struct btrfs_fs_info *fs_info, struct bio *bio,
 			extent_submit_bio_start_t *submit_bio_start);
 blk_status_t btrfs_submit_bio_done(void *private_data, struct bio *bio,
 			  int mirror_num);
+int btrfs_alloc_log_tree_node(struct btrfs_trans_handle *trans,
+			      struct btrfs_root *root);
 int btrfs_init_log_root_tree(struct btrfs_trans_handle *trans,
 			     struct btrfs_fs_info *fs_info);
 int btrfs_add_log_tree(struct btrfs_trans_handle *trans,
Signature · 1 line
2.27.0
#113 lore raw

PATCH v10 38/41 btrfs: extend zoned allocator to use dedicated tree-log block group

In reply to Naohiro Aota

This is the 1/3 patch to enable tree log on ZONED mode.

The tree-log feature does not work on ZONED mode as is. Blocks for a tree-log tree are allocated mixed with other metadata blocks, and btrfs writes and syncs the tree-log blocks to devices at the time of fsync(), which is different timing from a global transaction commit. As a result, both writing tree-log blocks and writing other metadata blocks become non-sequential writes that ZONED mode must avoid.

We can introduce a dedicated block group for tree-log blocks so that tree-log blocks and other metadata blocks can be separated write streams. As a result, each write stream can now be written to devices separately. "fs_info->treelog_bg" tracks the dedicated block group and btrfs assign "treelog_bg" on-demand on tree-log block allocation time.

This commit extends the zoned block allocator to use the block group.

---
 fs/btrfs/block-group.c |  7 +++++
 fs/btrfs/ctree.h       |  2 ++
 fs/btrfs/extent-tree.c | 63 +++++++++++++++++++++++++++++++++++++++---
 3 files changed, 68 insertions(+), 4 deletions(-)
fs/btrfs/block-group.c +7 -0
index 04bb0602f1cc..d222f54eb0c1 100644
@@ -939,6 +939,13 @@ int btrfs_remove_block_group(struct btrfs_trans_handle *trans,
 	btrfs_return_cluster_to_free_space(block_group, cluster);
 	spin_unlock(&cluster->refill_lock);
 
+	if (btrfs_is_zoned(fs_info)) {
+		spin_lock(&fs_info->treelog_bg_lock);
+		if (fs_info->treelog_bg == block_group->start)
+			fs_info->treelog_bg = 0;
+		spin_unlock(&fs_info->treelog_bg_lock);
+	}
+
 	path = btrfs_alloc_path();
 	if (!path) {
 		ret = -ENOMEM;
fs/btrfs/ctree.h +2 -0
index 8138e932b7cc..2fd7e58343ce 100644
@@ -957,6 +957,8 @@ struct btrfs_fs_info {
 	/* Max size to emit ZONE_APPEND write command */
 	u64 max_zone_append_size;
 	struct mutex zoned_meta_io_lock;
+	spinlock_t treelog_bg_lock;
+	u64 treelog_bg;
 
 #ifdef CONFIG_BTRFS_FS_REF_VERIFY
 	spinlock_t ref_verify_lock;
fs/btrfs/extent-tree.c +59 -4

141 lines not on this page · raw message

Signature · 1 line
2.27.0
#114 lore raw

PATCH v10 38/41 btrfs: extend zoned allocator to use dedicated tree-log block group

This is the 1/3 patch to enable tree log on ZONED mode.

The tree-log feature does not work on ZONED mode as is. Blocks for a tree-log tree are allocated mixed with other metadata blocks, and btrfs writes and syncs the tree-log blocks to devices at the time of fsync(), which is different timing from a global transaction commit. As a result, both writing tree-log blocks and writing other metadata blocks become non-sequential writes that ZONED mode must avoid.

We can introduce a dedicated block group for tree-log blocks so that tree-log blocks and other metadata blocks can be separated write streams. As a result, each write stream can now be written to devices separately. "fs_info->treelog_bg" tracks the dedicated block group and btrfs assign "treelog_bg" on-demand on tree-log block allocation time.

This commit extends the zoned block allocator to use the block group.

---
 fs/btrfs/block-group.c |  7 ++++
 fs/btrfs/ctree.h       |  2 ++
 fs/btrfs/extent-tree.c | 79 +++++++++++++++++++++++++++++++++++++++---
 3 files changed, 84 insertions(+), 4 deletions(-)

I forgot to merge a patch to add the lock description...

fs/btrfs/block-group.c +7 -0
index 04bb0602f1cc..d222f54eb0c1 100644
@@ -939,6 +939,13 @@ int btrfs_remove_block_group(struct btrfs_trans_handle *trans,
 	btrfs_return_cluster_to_free_space(block_group, cluster);
 	spin_unlock(&cluster->refill_lock);
 
+	if (btrfs_is_zoned(fs_info)) {
+		spin_lock(&fs_info->treelog_bg_lock);
+		if (fs_info->treelog_bg == block_group->start)
+			fs_info->treelog_bg = 0;
+		spin_unlock(&fs_info->treelog_bg_lock);
+	}
+
 	path = btrfs_alloc_path();
 	if (!path) {
 		ret = -ENOMEM;
fs/btrfs/ctree.h +2 -0
index 8138e932b7cc..2fd7e58343ce 100644
@@ -957,6 +957,8 @@ struct btrfs_fs_info {
 	/* Max size to emit ZONE_APPEND write command */
 	u64 max_zone_append_size;
 	struct mutex zoned_meta_io_lock;
+	spinlock_t treelog_bg_lock;
+	u64 treelog_bg;
 
 #ifdef CONFIG_BTRFS_FS_REF_VERIFY
 	spinlock_t ref_verify_lock;
fs/btrfs/extent-tree.c +75 -4

163 lines not on this page · raw message

Signature · 1 line
2.27.0
#115 lore raw

PATCH v10 39/41 btrfs: serialize log transaction on ZONED mode

In reply to Naohiro Aota

This is the 2/3 patch to enable tree-log on ZONED mode.

Since we can start more than one log transactions per subvolume simultaneously, nodes from multiple transactions can be allocated interleaved. Such mixed allocation results in non-sequential writes at the time of log transaction commit. The nodes of the global log root tree (fs_info->log_root_tree), also have the same mixed allocation problem.

This patch serializes log transactions by waiting for a committing transaction when someone tries to start a new transaction, to avoid the mixed allocation problem. We must also wait for running log transactions from another subvolume, but there is no easy way to detect which subvolume root is running a log transaction. So, this patch forbids starting a new log transaction when other subvolumes already allocated the global log root tree.

---
 fs/btrfs/tree-log.c | 22 +++++++++++++++++++++-
 1 file changed, 21 insertions(+), 1 deletion(-)
fs/btrfs/tree-log.c +21 -1
index 5f585cf57383..505de1cc1394 100644
@@ -106,6 +106,7 @@ static noinline int replay_dir_deletes(struct btrfs_trans_handle *trans,
 				       struct btrfs_root *log,
 				       struct btrfs_path *path,
 				       u64 dirid, int del_all);
+static void wait_log_commit(struct btrfs_root *root, int transid);
 
 /*
  * tree logging is a special write ahead log used to make sure that
@@ -140,16 +141,25 @@ static int start_log_trans(struct btrfs_trans_handle *trans,
 			   struct btrfs_log_ctx *ctx)
 {
 	struct btrfs_fs_info *fs_info = root->fs_info;
+	const bool zoned = btrfs_is_zoned(fs_info);
 	int ret = 0;
 
 	mutex_lock(&root->log_mutex);
 
+again:
 	if (root->log_root) {
+		int index = (root->log_transid + 1) % 2;
+
 		if (btrfs_need_log_full_commit(trans)) {
 			ret = -EAGAIN;
 			goto out;
 		}
 
+		if (zoned && atomic_read(&root->log_commit[index])) {
+			wait_log_commit(root, root->log_transid - 1);
+			goto again;
+		}
+
 		if (!root->log_start_pid) {
 			clear_bit(BTRFS_ROOT_MULTI_LOG_TASKS, &root->state);
 			root->log_start_pid = current->pid;
@@ -158,7 +168,9 @@ static int start_log_trans(struct btrfs_trans_handle *trans,
 		}
 	} else {
 		mutex_lock(&fs_info->tree_log_mutex);
-		if (!fs_info->log_root_tree)
+		if (zoned && fs_info->log_root_tree)
+			ret = -EAGAIN;
+		else if (!fs_info->log_root_tree)
 			ret = btrfs_init_log_root_tree(trans, fs_info);
 		mutex_unlock(&fs_info->tree_log_mutex);
 		if (ret)
@@ -193,14 +205,22 @@ static int start_log_trans(struct btrfs_trans_handle *trans,
  */
 static int join_running_log_trans(struct btrfs_root *root)
 {
+	const bool zoned = btrfs_is_zoned(root->fs_info);
 	int ret = -ENOENT;
 
 	if (!test_bit(BTRFS_ROOT_HAS_LOG_TREE, &root->state))
 		return ret;
 
 	mutex_lock(&root->log_mutex);
+again:
 	if (root->log_root) {
+		int index = (root->log_transid + 1) % 2;
+
 		ret = 0;
+		if (zoned && atomic_read(&root->log_commit[index])) {
+			wait_log_commit(root, root->log_transid - 1);
+			goto again;
+		}
 		atomic_inc(&root->log_writers);
 	}
 	mutex_unlock(&root->log_mutex);
Signature · 1 line
2.27.0
#116 lore raw

PATCH v10 40/41 btrfs: reorder log node allocation

In reply to Naohiro Aota

This is the 3/3 patch to enable tree-log on ZONED mode.

The allocation order of nodes of "fs_info->log_root_tree" and nodes of "root->log_root" is not the same as the writing order of them. So, the writing causes unaligned write errors.

This patch reorders the allocation of them by delaying allocation of the root node of "fs_info->log_root_tree," so that the node buffers can go out sequentially to devices.

---
 fs/btrfs/disk-io.c  |  7 -------
 fs/btrfs/tree-log.c | 24 ++++++++++++++++++------
 2 files changed, 18 insertions(+), 13 deletions(-)
fs/btrfs/disk-io.c +0 -7
index 97e3deb46cf1..e896dd564434 100644
@@ -1255,18 +1255,11 @@ int btrfs_init_log_root_tree(struct btrfs_trans_handle *trans,
 			     struct btrfs_fs_info *fs_info)
 {
 	struct btrfs_root *log_root;
-	int ret;
 
 	log_root = alloc_log_tree(trans, fs_info);
 	if (IS_ERR(log_root))
 		return PTR_ERR(log_root);
 
-	ret = btrfs_alloc_log_tree_node(trans, log_root);
-	if (ret) {
-		btrfs_put_root(log_root);
-		return ret;
-	}
-
 	WARN_ON(fs_info->log_root_tree);
 	fs_info->log_root_tree = log_root;
 	return 0;
fs/btrfs/tree-log.c +18 -6
index 505de1cc1394..15f9e8a461ee 100644
@@ -3140,6 +3140,16 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans,
 	list_add_tail(&root_log_ctx.list, &log_root_tree->log_ctxs[index2]);
 	root_log_ctx.log_transid = log_root_tree->log_transid;
 
+	mutex_lock(&fs_info->tree_log_mutex);
+	if (!log_root_tree->node) {
+		ret = btrfs_alloc_log_tree_node(trans, log_root_tree);
+		if (ret) {
+			mutex_unlock(&fs_info->tree_log_mutex);
+			goto out;
+		}
+	}
+	mutex_unlock(&fs_info->tree_log_mutex);
+
 	/*
 	 * Now we are safe to update the log_root_tree because we're under the
 	 * log_mutex, and we're a current writer so we're holding the commit
@@ -3289,12 +3299,14 @@ static void free_log_tree(struct btrfs_trans_handle *trans,
 		.process_func = process_one_buffer
 	};
 
-	ret = walk_log_tree(trans, log, &wc);
-	if (ret) {
-		if (trans)
-			btrfs_abort_transaction(trans, ret);
-		else
-			btrfs_handle_fs_error(log->fs_info, ret, NULL);
+	if (log->node) {
+		ret = walk_log_tree(trans, log, &wc);
+		if (ret) {
+			if (trans)
+				btrfs_abort_transaction(trans, ret);
+			else
+				btrfs_handle_fs_error(log->fs_info, ret, NULL);
+		}
 	}
 
 	clear_extent_bits(&log->dirty_log_pages, 0, (u64)-1,
Signature · 1 line
2.27.0
#117 lore raw

PATCH v10 41/41 btrfs: enable to mount ZONED incompat flag

In reply to Naohiro Aota

This final patch adds the ZONED incompat flag to BTRFS_FEATURE_INCOMPAT_SUPP and enables btrfs to mount ZONED flagged file system.

---
 fs/btrfs/ctree.h | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
fs/btrfs/ctree.h +2 -1
index 2fd7e58343ce..935b3470a069 100644
@@ -302,7 +302,8 @@ struct btrfs_super_block {
 	 BTRFS_FEATURE_INCOMPAT_SKINNY_METADATA |	\
 	 BTRFS_FEATURE_INCOMPAT_NO_HOLES	|	\
 	 BTRFS_FEATURE_INCOMPAT_METADATA_UUID	|	\
-	 BTRFS_FEATURE_INCOMPAT_RAID1C34)
+	 BTRFS_FEATURE_INCOMPAT_RAID1C34	|	\
+	 BTRFS_FEATURE_INCOMPAT_ZONED)
 
 #define BTRFS_FEATURE_INCOMPAT_SAFE_SET			\
 	(BTRFS_FEATURE_INCOMPAT_EXTENDED_IREF)
Signature · 1 line
2.27.0
#118 lore raw
In reply to Naohiro Aota
Quoted 4 lines · Naohiro Aota

This series adds zoned block device support to btrfs.

This series is also available on github.
Kernel https://github.com/naota/linux/tree/btrfs-zoned-v10

This branch is not reachable. Should it be

https://github.com/naota/linux/tree/btrfs-zoned-for-v10 ?

But the commits in this branch are at a pre-fixups stage.

Thanks, Anand

Quoted 434 lines

Userland https://github.com/naota/btrfs-progs/tree/btrfs-zoned
xfstests https://github.com/naota/fstests/tree/btrfs-zoned

Userland tool depends on patched util-linux (libblkid and wipefs) to handle log-structured superblock. To ease the testing, pre-compiled static linked userland tools are available here:
https://wdc.app.box.com/s/fnhqsb3otrvgkstq66o6bvdw6tk525kp

This v10 still leaves the following issues left for later fix. But, the first part of the series should be good shape to be merged.
- Bio submission path & splitting an ordered extent
- Redirtying freed tree blocks
- Switch to keeping it dirty
- Not working correctly for now
- Dedicated tree-log block group
- We need tree-log for zoned device
- Dbench (32 clients) is 85% slower with "-o notreelog"
- Need to separate tree-log block group from other metadata space_info
- Relocation
- Use normal write command for relocation
- Relocated device extents must be reset
- It should be discarded on regular btrfs too though

Changes from v9:
- Extract iomap_dio_bio_opflags() to set the proper bi_opf flag
- write pointer emulation
- Rewrite using btrfs_previous_extent_item()
- Convert ASSERT() to runtime check
- Exclude regular superblock positions
- Fix an error on writing to conventional zones
- Take the transaction lock in mark_block_group_to_copy()
- Rename 'hmzoned_devices' to 'zoned_devices' in btrfs_check_zoned_mode()
- Add do_discard_extent() helper
- Move zoned check into fetch_cluster_info()
- Drop setting bdev to bio in btrfs_bio_add_page() (will fix later once we support multiple devices)
- Subtract bytes_zone_unusable properly when removing a block group
- Add "struct block_device *bdev" directly to btrfs_rmap_block()
- Rename btrfs_zone_align to btrfs_align_offset_to_zone
- Add comment to use pr_info in place of btrfs_info
- Add comment for superblock log zones
- Fix coding style
- Fix typos

btrfs-progs and xfstests series will follow.

This version of ZONED btrfs switched from normal write command to zone append write command. You do not need to specify LBA (at the write pointer) to write for zone append write command. Instead, you only select a zone to write with its start LBA. Then the device (NVMe ZNS), or the emulation of zone append command in the sd driver in the case of SAS or SATA HDDs, automatically writes the data at the write pointer position and return the written LBA as a command reply.

The benefit of using the zone append write command is that write command issuing order does not matter. So, we can eliminate block group lock and utilize asynchronous checksum, which can reorder the IOs.

Eliminating the lock improves performance. In particular, on a workload with massive competing to the same zone [1], we observed 36% performance improvement compared to normal write.

[1] Fio running 16 jobs with 4KB random writes for 5 minutes

However, there are some limitations. We cannot use the non-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.

For the same reason, we cannot issue multiple IOs for one ordered extent. Thus, the size of an ordered extent is limited under max_zone_append_size. This limitation will cause fragmentation and increased usage of metadata. In the future, we can add optimization to merge ordered extents after end_bio.

* Patch series description

A zoned block device consists of a number of zones. Zones are either conventional and accepting random writes or sequential and requiring that writes be issued in LBA order from each zone write pointer position. This patch series ensures that the sequential write constraint of sequential zones is respected while fundamentally not changing BtrFS block and I/O management for block stored in conventional zones.

To achieve this, the default chunk size of btrfs is changed on zoned block devices so that chunks are always aligned to a zone. Allocation of blocks within a chunk is changed so that the allocation is always sequential from the beginning of the chunks. To do so, an allocation pointer is added to block groups and used as the allocation hint. The allocation changes also ensure that blocks freed below the allocation pointer are ignored, resulting in sequential block allocation regardless of the chunk usage.

The zone of a chunk is reset to allow reuse of the zone only when the block group is being freed, that is, when all the chunks of the block group are unused.

For btrfs volumes composed of multiple zoned disks, a restriction is added to ensure that all disks have the same zone size. This restriction matches the existing constraint that all chunks in a block group must have the same size.

* Enabling tree-log

The tree-log feature does not work on ZONED mode as is. Blocks for a tree-log tree are allocated mixed with other metadata blocks, and btrfs writes and syncs the tree-log blocks to devices at the time of fsync(), which is different timing than a global transaction commit. As a result, both writing tree-log blocks and writing other metadata blocks become non-sequential writes which ZONED mode must avoid.

This series introduces a dedicated block group for tree-log blocks to create two metadata writing streams, one for tree-log blocks and the other for metadata blocks. As a result, each write stream can now be written to devices separately and sequentially.

* Log-structured superblock

Superblock (and its copies) is the only data structure in btrfs which has a fixed location on a device. Since we cannot overwrite in a sequential write required zone, we cannot place superblock in the zone.

This series implements superblock log writing. It uses two zones as a circular buffer to write updated superblocks. Once the first zone is filled up, start writing into the second zone. The first zone will be reset once both zones are filled. We can determine the postion of the latest superblock by reading the write pointer information from a device.

* Patch series organization

Patches 1 and 2 are preparing patches for block and iomap layer.

Patch 3 introduces the ZONED incompatible feature flag to indicate that the btrfs volume was formatted for use on zoned block devices.

Patches 4 to 6 implement functions to gather information on the zones of the device (zones type, write pointer position, and max_zone_append_size).

Patches 7 to 10 disable features which are not compatible with the sequential write constraints of zoned block devices. These includes space_cache, NODATACOW, fallocate, and MIXED_BG.

Patch 11 implements the log-structured superblock writing.

Patches 12 and 13 tweak the device extent allocation for ZONED mode and add verification to check if a device extent is properly aligned to zones.

Patches 14 to 17 implements sequential block allocator for ZONED mode.

Patch 18 implement a zone reset for unused block groups.

Patches 19 to 30 implement the writing path for several types of IO (non-compressed data, direct IO, and metadata). These include re-dirtying once-freed metadata blocks to prevent write holes.

Patches 31 to 40 tweak some btrfs features work with ZONED mode. These include device-replace, relocation, repairing IO error, and tree-log.

Finally, patch 41 adds the ZONED feature to the list of supported features.

* Patch testing note

** Zone-aware util-linux

Since the log-structured superblock feature changed the location of superblock magic, the current util-linux (libblkid) cannot detect ZONED btrfs anymore. You need to apply a to-be posted patch to util-linux to make it "zone aware".

** Testing device

You need devices with zone append writing command support to run ZONED btrfs.

Other than real devices, null_blk supports zone append write command. You can use memory backed null_blk to run the test on it. Following script creates 12800 MB /dev/nullb0.

     sysfs=/sys/kernel/config/nullb/nullb0
     size=12800 # MB
     
     # drop nullb0
     if [[ -d $sysfs ]]; then
             echo 0 > "${sysfs}"/power
             rmdir $sysfs
     fi
     lsmod | grep -q null_blk && rmmod null_blk
     modprobe null_blk nr_devices=0
     
     mkdir "${sysfs}"
     
     echo "${size}" > "${sysfs}"/size
     echo 1 > "${sysfs}"/zoned
     echo 0 > "${sysfs}"/zone_nr_conv
     echo 1 > "${sysfs}"/memory_backed
     
     echo 1 > "${sysfs}"/power
     udevadm settle

Zoned SCSI devices such as SMR HDDs or scsi_debug also support the zone append command as an emulated command within the SCSI sd driver. This emulation is completely transparent to the user and provides the same semantic as a NVMe ZNS native drive support.

Also, there is a qemu patch available to enable NVMe ZNS device.

** xfstests

We ran xfstests on ZONED btrfs, and, if we omit some cases that are known to fail currently, all test cases pass.

Cases that can be ignored:
1) failing also with the regular btrfs on regular devices,
2) trying to test fallocate feature without testing with

    "_require_xfs_io_command "falloc"",

3) trying to test incompatible features for ZONED btrfs (e.g. RAID5/6)
4) trying to use incompatible setup for ZONED btrfs (e.g. dm-linear not aligned to zone boundary, swap)
5) trying to create a file system with too small size, (we require at least 9 zones to initiate a ZONED btrfs)
6) dropping original MKFS_OPTIONS ("-O zoned"), so it cannot create ZONED

    btrfs (btrfs/003)

7) having ENOSPC which incurred by larger metadata block group size

I will send a patch series for xfstests to handle these cases (2-6) properly.

Patched xfstests is available here:

https://github.com/naota/fstests/tree/btrfs-zoned

Also, you need to apply the following patch if you run xfstests with tcmu devices. xfstests btrfs/003 failed to "_devmgt_add" after "_devmgt_remove" without this patch.

https://marc.info/?l=linux-scsi&m=156498625421698&w=2

v9 https://lore.kernel.org/linux-btrfs/[email protected]/
v8 https://lore.kernel.org/linux-btrfs/[email protected]/
v7 https://lore.kernel.org/linux-btrfs/[email protected]/
v6 https://lore.kernel.org/linux-btrfs/[email protected]/
v5 https://lore.kernel.org/linux-btrfs/[email protected]/
v4 https://lwn.net/Articles/797061/
v3 https://lore.kernel.org/linux-btrfs/[email protected]/
v2 https://lore.kernel.org/linux-btrfs/[email protected]/
v1 https://lore.kernel.org/linux-btrfs/[email protected]/

Changelog
v9
- Direct-IO path now follow several hardware restrictions (other than max_zone_append_size) by using ZONE_APPEND support of iomap
- introduces union of fs_info->zone_size and fs_info->zoned [Johannes]
- and use btrfs_is_zoned(fs_info) in place of btrfs_fs_incompat(fs_info, ZONED)
- print if zoned is enabled or not when printing module info [Johannes]
- drop patch of disabling inode_cache on ZONED
- moved for_teelog flag to a proper location [Johannes]
- Code style fixes [Johannes]
- Add comment about adding physical layer things to ordered extent structure
- Pass file_offset explicitly to extract_ordered_extent() instead of determining it from bio
- Bug fixes
- write out fsync region so that the logical address of ordered extents and checksums are properly finalized
- free zone_info at umount time
- fix superblock log handling when entering zones[1] in the first time
- fixes double free of log-tree roots [Johannes]
- Drop erroneous ASSERT in do_allocation_zoned()
v8
- Use bio_add_hw_page() to build up bio to honor hardware restrictions
- add bio_add_zone_append_page() as a wrapper of the function
- Split file extent on submitting bio
- If bio_add_zone_append_page() fails, split the file extent and send out bio
- so, we can ensure one bio == one file extent
- Fix build bot issues
- Rebased on misc-next
v7:
- Use zone append write command instead of normal write command
- Bio issuing order does not matter
- No need to use lock anymore
- Can use asynchronous checksum
- Removed RAID support for now
- Rename HMZONED to ZONED
- Split some patches
- Rebased on kdave/for-5.9-rc3 + iomap direct IO
v6:
- Use bitmap helpers (Johannes)
- Code cleanup (Johannes)
- Rebased on kdave/for-5.5
- Enable the tree-log feature.
- Treat conventional zones as sequential zones, so we can now allow mixed allocation of conventional zone and sequential write required zone to construct a block group.
- Implement log-structured superblock
- No need for one conventional zone at the beginning of a device.
- Fix deadlock of direct IO writing
- Fix building with !CONFIG_BLK_DEV_ZONED (Johannes)
- Fix leak of zone_info (Johannes)
v5:
- Rebased on kdave/for-5.5
- Enable the tree-log feature.
- Treat conventional zones as sequential zones, so we can now allow mixed allocation of conventional zone and sequential write required zone to construct a block group.
- Implement log-structured superblock
- No need for one conventional zone at the beginning of a device.
- Fix deadlock of direct IO writing
- Fix building with !CONFIG_BLK_DEV_ZONED (Johannes)
- Fix leak of zone_info (Johannes)
v4:
- Move memory allcation of zone informattion out of btrfs_get_dev_zones() (Anand)
- Add disabled features table in commit log (Anand)
- Ensure "max_chunk_size >= devs_min * data_stripes * zone_size"
v3:
- Serialize allocation and submit_bio instead of bio buffering in

    btrfs_map_bio().

-- Disable async checksum/submit in HMZONED mode
- Introduce helper functions and hmzoned.c/h (Josef, David)
- Add support for repairing IO failure
- Add support for NOCOW direct IO write (Josef)
- Disable preallocation entirely
-- Disable INODE_MAP_CACHE
-- relocation is reworked not to rely on preallocation in HMZONED mode
- Disable NODATACOW
-Disable MIXED_BG
- Device extent that cover super block position is banned (David)
v2:
- Add support for dev-replace
-- To support dev-replace, moved submit_buffer one layer up. It now
handles bio instead of btrfs_bio.
-- Mark unmirrored Block Group readonly only when there are writable
mirrored BGs. Necessary to handle degraded RAID.
- Expire worker use vanilla delayed_work instead of btrfs's async-thread
- Device extent allocator now ensure that region is on the same zone type.
- Add delayed allocation shrinking.
- Rename btrfs_drop_dev_zonetypes() to btrfs_destroy_dev_zonetypes
- Fix
-- Use SECTOR_SHIFT (Nikolay)
-- Use btrfs_err (Nikolay)

Johannes Thumshirn (1):
block: add bio_add_zone_append_page

Naohiro Aota (40):
iomap: support REQ_OP_ZONE_APPEND
btrfs: introduce ZONED feature flag
btrfs: get zone information of zoned block devices
btrfs: check and enable ZONED mode
btrfs: introduce max_zone_append_size
btrfs: disallow space_cache in ZONED mode
btrfs: disallow NODATACOW in ZONED mode
btrfs: disable fallocate in ZONED mode
btrfs: disallow mixed-bg in ZONED mode
btrfs: implement log-structured superblock for ZONED mode
btrfs: implement zoned chunk allocator
btrfs: verify device extent is aligned to zone
btrfs: load zone's alloction offset
btrfs: emulate write pointer for conventional zones
btrfs: track unusable bytes for zones
btrfs: do sequential extent allocation in ZONED mode
btrfs: reset zones of unused block groups
btrfs: redirty released extent buffers in ZONED mode
btrfs: extract page adding function
btrfs: use bio_add_zone_append_page for zoned btrfs
btrfs: handle REQ_OP_ZONE_APPEND as writing
btrfs: split ordered extent when bio is sent
btrfs: extend btrfs_rmap_block for specifying a device
btrfs: use ZONE_APPEND write for ZONED btrfs
btrfs: enable zone append writing for direct IO
btrfs: introduce dedicated data write path for ZONED mode
btrfs: serialize meta IOs on ZONED mode
btrfs: wait existing extents before truncating
btrfs: avoid async metadata checksum on ZONED mode
btrfs: mark block groups to copy for device-replace
btrfs: implement cloning for ZONED device-replace
btrfs: implement copying for ZONED device-replace
btrfs: support dev-replace in ZONED mode
btrfs: enable relocation in ZONED mode
btrfs: relocate block group to repair IO failure in ZONED
btrfs: split alloc_log_tree()
btrfs: extend zoned allocator to use dedicated tree-log block group
btrfs: serialize log transaction on ZONED mode
btrfs: reorder log node allocation
btrfs: enable to mount ZONED incompat flag

  block/bio.c                       |   38 +
  fs/btrfs/Makefile                 |    1 +
  fs/btrfs/block-group.c            |   84 +-
  fs/btrfs/block-group.h            |   18 +-
  fs/btrfs/ctree.h                  |   20 +-
  fs/btrfs/dev-replace.c            |  195 +++++
  fs/btrfs/dev-replace.h            |    3 +
  fs/btrfs/disk-io.c                |   93 ++-
  fs/btrfs/disk-io.h                |    2 +
  fs/btrfs/extent-tree.c            |  218 ++++-
  fs/btrfs/extent_io.c              |  130 ++-
  fs/btrfs/extent_io.h              |    2 +
  fs/btrfs/file.c                   |    6 +-
  fs/btrfs/free-space-cache.c       |   58 ++
  fs/btrfs/free-space-cache.h       |    2 +
  fs/btrfs/inode.c                  |  164 +++-
  fs/btrfs/ioctl.c                  |   13 +
  fs/btrfs/ordered-data.c           |   79 ++
  fs/btrfs/ordered-data.h           |   10 +
  fs/btrfs/relocation.c             |   35 +-
  fs/btrfs/scrub.c                  |  145 ++++
  fs/btrfs/space-info.c             |   13 +-
  fs/btrfs/space-info.h             |    4 +-
  fs/btrfs/super.c                  |   19 +-
  fs/btrfs/sysfs.c                  |    4 +
  fs/btrfs/tests/extent-map-tests.c |    2 +-
  fs/btrfs/transaction.c            |   10 +
  fs/btrfs/transaction.h            |    3 +
  fs/btrfs/tree-log.c               |   52 +-
  fs/btrfs/volumes.c                |  322 +++++++-
  fs/btrfs/volumes.h                |    7 +
  fs/btrfs/zoned.c                  | 1272 +++++++++++++++++++++++++++++
  fs/btrfs/zoned.h                  |  295 +++++++
  fs/iomap/direct-io.c              |   41 +-
  include/linux/bio.h               |    2 +
  include/linux/iomap.h             |    1 +
  include/uapi/linux/btrfs.h        |    1 +
  37 files changed, 3246 insertions(+), 118 deletions(-)

create mode 100644 fs/btrfs/zoned.c
create mode 100644 fs/btrfs/zoned.h

#119 lore raw
Quoted 11 lines · Anand Jain
Quoted 4 lines · Naohiro Aota

This series adds zoned block device support to btrfs.

This series is also available on github.
Kernel https://github.com/naota/linux/tree/btrfs-zoned-v10

This branch is not reachable. Should it be

https://github.com/naota/linux/tree/btrfs-zoned-for-v10 ?

But the commits in this branch are at a pre-fixups stage.

Sorry, I forgot to push the one. I have pushed the fixuped version to "btrfs-zoned-v10" branch. Thanks

Quoted 435 lines

Thanks, Anand

Quoted 430 lines

Userland https://github.com/naota/btrfs-progs/tree/btrfs-zoned
xfstests https://github.com/naota/fstests/tree/btrfs-zoned

Userland tool depends on patched util-linux (libblkid and wipefs) to handle log-structured superblock. To ease the testing, pre-compiled static linked userland tools are available here:
https://wdc.app.box.com/s/fnhqsb3otrvgkstq66o6bvdw6tk525kp

This v10 still leaves the following issues left for later fix. But, the first part of the series should be good shape to be merged.
- Bio submission path & splitting an ordered extent
- Redirtying freed tree blocks
- Switch to keeping it dirty
- Not working correctly for now
- Dedicated tree-log block group
- We need tree-log for zoned device
- Dbench (32 clients) is 85% slower with "-o notreelog"
- Need to separate tree-log block group from other metadata space_info
- Relocation
- Use normal write command for relocation
- Relocated device extents must be reset
- It should be discarded on regular btrfs too though

Changes from v9:
- Extract iomap_dio_bio_opflags() to set the proper bi_opf flag
- write pointer emulation
- Rewrite using btrfs_previous_extent_item()
- Convert ASSERT() to runtime check
- Exclude regular superblock positions
- Fix an error on writing to conventional zones
- Take the transaction lock in mark_block_group_to_copy()
- Rename 'hmzoned_devices' to 'zoned_devices' in btrfs_check_zoned_mode()
- Add do_discard_extent() helper
- Move zoned check into fetch_cluster_info()
- Drop setting bdev to bio in btrfs_bio_add_page() (will fix later once we support multiple devices)
- Subtract bytes_zone_unusable properly when removing a block group
- Add "struct block_device *bdev" directly to btrfs_rmap_block()
- Rename btrfs_zone_align to btrfs_align_offset_to_zone
- Add comment to use pr_info in place of btrfs_info
- Add comment for superblock log zones
- Fix coding style
- Fix typos

btrfs-progs and xfstests series will follow.

This version of ZONED btrfs switched from normal write command to zone append write command. You do not need to specify LBA (at the write pointer) to write for zone append write command. Instead, you only select a zone to write with its start LBA. Then the device (NVMe ZNS), or the emulation of zone append command in the sd driver in the case of SAS or SATA HDDs, automatically writes the data at the write pointer position and return the written LBA as a command reply.

The benefit of using the zone append write command is that write command issuing order does not matter. So, we can eliminate block group lock and utilize asynchronous checksum, which can reorder the IOs.

Eliminating the lock improves performance. In particular, on a workload with massive competing to the same zone [1], we observed 36% performance improvement compared to normal write.

[1] Fio running 16 jobs with 4KB random writes for 5 minutes

However, there are some limitations. We cannot use the non-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.

For the same reason, we cannot issue multiple IOs for one ordered extent. Thus, the size of an ordered extent is limited under max_zone_append_size. This limitation will cause fragmentation and increased usage of metadata. In the future, we can add optimization to merge ordered extents after end_bio.

* Patch series description

A zoned block device consists of a number of zones. Zones are either conventional and accepting random writes or sequential and requiring that writes be issued in LBA order from each zone write pointer position. This patch series ensures that the sequential write constraint of sequential zones is respected while fundamentally not changing BtrFS block and I/O management for block stored in conventional zones.

To achieve this, the default chunk size of btrfs is changed on zoned block devices so that chunks are always aligned to a zone. Allocation of blocks within a chunk is changed so that the allocation is always sequential from the beginning of the chunks. To do so, an allocation pointer is added to block groups and used as the allocation hint. The allocation changes also ensure that blocks freed below the allocation pointer are ignored, resulting in sequential block allocation regardless of the chunk usage.

The zone of a chunk is reset to allow reuse of the zone only when the block group is being freed, that is, when all the chunks of the block group are unused.

For btrfs volumes composed of multiple zoned disks, a restriction is added to ensure that all disks have the same zone size. This restriction matches the existing constraint that all chunks in a block group must have the same size.

* Enabling tree-log

The tree-log feature does not work on ZONED mode as is. Blocks for a tree-log tree are allocated mixed with other metadata blocks, and btrfs writes and syncs the tree-log blocks to devices at the time of fsync(), which is different timing than a global transaction commit. As a result, both writing tree-log blocks and writing other metadata blocks become non-sequential writes which ZONED mode must avoid.

This series introduces a dedicated block group for tree-log blocks to create two metadata writing streams, one for tree-log blocks and the other for metadata blocks. As a result, each write stream can now be written to devices separately and sequentially.

* Log-structured superblock

Superblock (and its copies) is the only data structure in btrfs which has a fixed location on a device. Since we cannot overwrite in a sequential write required zone, we cannot place superblock in the zone.

This series implements superblock log writing. It uses two zones as a circular buffer to write updated superblocks. Once the first zone is filled up, start writing into the second zone. The first zone will be reset once both zones are filled. We can determine the postion of the latest superblock by reading the write pointer information from a device.

* Patch series organization

Patches 1 and 2 are preparing patches for block and iomap layer.

Patch 3 introduces the ZONED incompatible feature flag to indicate that the btrfs volume was formatted for use on zoned block devices.

Patches 4 to 6 implement functions to gather information on the zones of the device (zones type, write pointer position, and max_zone_append_size).

Patches 7 to 10 disable features which are not compatible with the sequential write constraints of zoned block devices. These includes space_cache, NODATACOW, fallocate, and MIXED_BG.

Patch 11 implements the log-structured superblock writing.

Patches 12 and 13 tweak the device extent allocation for ZONED mode and add verification to check if a device extent is properly aligned to zones.

Patches 14 to 17 implements sequential block allocator for ZONED mode.

Patch 18 implement a zone reset for unused block groups.

Patches 19 to 30 implement the writing path for several types of IO (non-compressed data, direct IO, and metadata). These include re-dirtying once-freed metadata blocks to prevent write holes.

Patches 31 to 40 tweak some btrfs features work with ZONED mode. These include device-replace, relocation, repairing IO error, and tree-log.

Finally, patch 41 adds the ZONED feature to the list of supported features.

* Patch testing note

** Zone-aware util-linux

Since the log-structured superblock feature changed the location of superblock magic, the current util-linux (libblkid) cannot detect ZONED btrfs anymore. You need to apply a to-be posted patch to util-linux to make it "zone aware".

** Testing device

You need devices with zone append writing command support to run ZONED btrfs.

Other than real devices, null_blk supports zone append write command. You can use memory backed null_blk to run the test on it. Following script creates 12800 MB /dev/nullb0.

    sysfs=/sys/kernel/config/nullb/nullb0
    size=12800 # MB
    # drop nullb0
    if [[ -d $sysfs ]]; then
            echo 0 > "${sysfs}"/power
            rmdir $sysfs
    fi
    lsmod | grep -q null_blk && rmmod null_blk
    modprobe null_blk nr_devices=0
    mkdir "${sysfs}"
    echo "${size}" > "${sysfs}"/size
    echo 1 > "${sysfs}"/zoned
    echo 0 > "${sysfs}"/zone_nr_conv
    echo 1 > "${sysfs}"/memory_backed
    echo 1 > "${sysfs}"/power
    udevadm settle

Zoned SCSI devices such as SMR HDDs or scsi_debug also support the zone append command as an emulated command within the SCSI sd driver. This emulation is completely transparent to the user and provides the same semantic as a NVMe ZNS native drive support.

Also, there is a qemu patch available to enable NVMe ZNS device.

** xfstests

We ran xfstests on ZONED btrfs, and, if we omit some cases that are known to fail currently, all test cases pass.

Cases that can be ignored:
1) failing also with the regular btrfs on regular devices,
2) trying to test fallocate feature without testing with "_require_xfs_io_command "falloc"",
3) trying to test incompatible features for ZONED btrfs (e.g. RAID5/6)
4) trying to use incompatible setup for ZONED btrfs (e.g. dm-linear not aligned to zone boundary, swap)
5) trying to create a file system with too small size, (we require at least 9 zones to initiate a ZONED btrfs)
6) dropping original MKFS_OPTIONS ("-O zoned"), so it cannot create ZONED btrfs (btrfs/003)
7) having ENOSPC which incurred by larger metadata block group size

I will send a patch series for xfstests to handle these cases (2-6) properly.

Patched xfstests is available here:

https://github.com/naota/fstests/tree/btrfs-zoned

Also, you need to apply the following patch if you run xfstests with tcmu devices. xfstests btrfs/003 failed to "_devmgt_add" after "_devmgt_remove" without this patch.

https://marc.info/?l=linux-scsi&m=156498625421698&w=2

v9 https://lore.kernel.org/linux-btrfs/[email protected]/
v8 https://lore.kernel.org/linux-btrfs/[email protected]/
v7 https://lore.kernel.org/linux-btrfs/[email protected]/
v6 https://lore.kernel.org/linux-btrfs/[email protected]/
v5 https://lore.kernel.org/linux-btrfs/[email protected]/
v4 https://lwn.net/Articles/797061/
v3 https://lore.kernel.org/linux-btrfs/[email protected]/
v2 https://lore.kernel.org/linux-btrfs/[email protected]/
v1 https://lore.kernel.org/linux-btrfs/[email protected]/

Changelog
v9
- Direct-IO path now follow several hardware restrictions (other than max_zone_append_size) by using ZONE_APPEND support of iomap
- introduces union of fs_info->zone_size and fs_info->zoned [Johannes]
- and use btrfs_is_zoned(fs_info) in place of btrfs_fs_incompat(fs_info, ZONED)
- print if zoned is enabled or not when printing module info [Johannes]
- drop patch of disabling inode_cache on ZONED
- moved for_teelog flag to a proper location [Johannes]
- Code style fixes [Johannes]
- Add comment about adding physical layer things to ordered extent structure
- Pass file_offset explicitly to extract_ordered_extent() instead of determining it from bio
- Bug fixes
- write out fsync region so that the logical address of ordered extents and checksums are properly finalized
- free zone_info at umount time
- fix superblock log handling when entering zones[1] in the first time
- fixes double free of log-tree roots [Johannes]
- Drop erroneous ASSERT in do_allocation_zoned()
v8
- Use bio_add_hw_page() to build up bio to honor hardware restrictions
- add bio_add_zone_append_page() as a wrapper of the function
- Split file extent on submitting bio
- If bio_add_zone_append_page() fails, split the file extent and send out bio
- so, we can ensure one bio == one file extent
- Fix build bot issues
- Rebased on misc-next
v7:
- Use zone append write command instead of normal write command
- Bio issuing order does not matter
- No need to use lock anymore
- Can use asynchronous checksum
- Removed RAID support for now
- Rename HMZONED to ZONED
- Split some patches
- Rebased on kdave/for-5.9-rc3 + iomap direct IO
v6:
- Use bitmap helpers (Johannes)
- Code cleanup (Johannes)
- Rebased on kdave/for-5.5
- Enable the tree-log feature.
- Treat conventional zones as sequential zones, so we can now allow mixed allocation of conventional zone and sequential write required zone to construct a block group.
- Implement log-structured superblock
- No need for one conventional zone at the beginning of a device.
- Fix deadlock of direct IO writing
- Fix building with !CONFIG_BLK_DEV_ZONED (Johannes)
- Fix leak of zone_info (Johannes)
v5:
- Rebased on kdave/for-5.5
- Enable the tree-log feature.
- Treat conventional zones as sequential zones, so we can now allow mixed allocation of conventional zone and sequential write required zone to construct a block group.
- Implement log-structured superblock
- No need for one conventional zone at the beginning of a device.
- Fix deadlock of direct IO writing
- Fix building with !CONFIG_BLK_DEV_ZONED (Johannes)
- Fix leak of zone_info (Johannes)
v4:
- Move memory allcation of zone informattion out of btrfs_get_dev_zones() (Anand)
- Add disabled features table in commit log (Anand)
- Ensure "max_chunk_size >= devs_min * data_stripes * zone_size"
v3:
- Serialize allocation and submit_bio instead of bio buffering in btrfs_map_bio().
-- Disable async checksum/submit in HMZONED mode
- Introduce helper functions and hmzoned.c/h (Josef, David)
- Add support for repairing IO failure
- Add support for NOCOW direct IO write (Josef)
- Disable preallocation entirely
-- Disable INODE_MAP_CACHE
-- relocation is reworked not to rely on preallocation in HMZONED mode
- Disable NODATACOW
-Disable MIXED_BG
- Device extent that cover super block position is banned (David)
v2:
- Add support for dev-replace
-- To support dev-replace, moved submit_buffer one layer up. It now
handles bio instead of btrfs_bio.
-- Mark unmirrored Block Group readonly only when there are writable
mirrored BGs. Necessary to handle degraded RAID.
- Expire worker use vanilla delayed_work instead of btrfs's async-thread
- Device extent allocator now ensure that region is on the same zone type.
- Add delayed allocation shrinking.
- Rename btrfs_drop_dev_zonetypes() to btrfs_destroy_dev_zonetypes
- Fix
-- Use SECTOR_SHIFT (Nikolay)
-- Use btrfs_err (Nikolay)

Johannes Thumshirn (1):
block: add bio_add_zone_append_page

Naohiro Aota (40):
iomap: support REQ_OP_ZONE_APPEND
btrfs: introduce ZONED feature flag
btrfs: get zone information of zoned block devices
btrfs: check and enable ZONED mode
btrfs: introduce max_zone_append_size
btrfs: disallow space_cache in ZONED mode
btrfs: disallow NODATACOW in ZONED mode
btrfs: disable fallocate in ZONED mode
btrfs: disallow mixed-bg in ZONED mode
btrfs: implement log-structured superblock for ZONED mode
btrfs: implement zoned chunk allocator
btrfs: verify device extent is aligned to zone
btrfs: load zone's alloction offset
btrfs: emulate write pointer for conventional zones
btrfs: track unusable bytes for zones
btrfs: do sequential extent allocation in ZONED mode
btrfs: reset zones of unused block groups
btrfs: redirty released extent buffers in ZONED mode
btrfs: extract page adding function
btrfs: use bio_add_zone_append_page for zoned btrfs
btrfs: handle REQ_OP_ZONE_APPEND as writing
btrfs: split ordered extent when bio is sent
btrfs: extend btrfs_rmap_block for specifying a device
btrfs: use ZONE_APPEND write for ZONED btrfs
btrfs: enable zone append writing for direct IO
btrfs: introduce dedicated data write path for ZONED mode
btrfs: serialize meta IOs on ZONED mode
btrfs: wait existing extents before truncating
btrfs: avoid async metadata checksum on ZONED mode
btrfs: mark block groups to copy for device-replace
btrfs: implement cloning for ZONED device-replace
btrfs: implement copying for ZONED device-replace
btrfs: support dev-replace in ZONED mode
btrfs: enable relocation in ZONED mode
btrfs: relocate block group to repair IO failure in ZONED
btrfs: split alloc_log_tree()
btrfs: extend zoned allocator to use dedicated tree-log block group
btrfs: serialize log transaction on ZONED mode
btrfs: reorder log node allocation
btrfs: enable to mount ZONED incompat flag

 block/bio.c                       |   38 +
 fs/btrfs/Makefile                 |    1 +
 fs/btrfs/block-group.c            |   84 +-
 fs/btrfs/block-group.h            |   18 +-
 fs/btrfs/ctree.h                  |   20 +-
 fs/btrfs/dev-replace.c            |  195 +++++
 fs/btrfs/dev-replace.h            |    3 +
 fs/btrfs/disk-io.c                |   93 ++-
 fs/btrfs/disk-io.h                |    2 +
 fs/btrfs/extent-tree.c            |  218 ++++-
 fs/btrfs/extent_io.c              |  130 ++-
 fs/btrfs/extent_io.h              |    2 +
 fs/btrfs/file.c                   |    6 +-
 fs/btrfs/free-space-cache.c       |   58 ++
 fs/btrfs/free-space-cache.h       |    2 +
 fs/btrfs/inode.c                  |  164 +++-
 fs/btrfs/ioctl.c                  |   13 +
 fs/btrfs/ordered-data.c           |   79 ++
 fs/btrfs/ordered-data.h           |   10 +
 fs/btrfs/relocation.c             |   35 +-
 fs/btrfs/scrub.c                  |  145 ++++
 fs/btrfs/space-info.c             |   13 +-
 fs/btrfs/space-info.h             |    4 +-
 fs/btrfs/super.c                  |   19 +-
 fs/btrfs/sysfs.c                  |    4 +
 fs/btrfs/tests/extent-map-tests.c |    2 +-
 fs/btrfs/transaction.c            |   10 +
 fs/btrfs/transaction.h            |    3 +
 fs/btrfs/tree-log.c               |   52 +-
 fs/btrfs/volumes.c                |  322 +++++++-
 fs/btrfs/volumes.h                |    7 +
 fs/btrfs/zoned.c                  | 1272 +++++++++++++++++++++++++++++
 fs/btrfs/zoned.h                  |  295 +++++++
 fs/iomap/direct-io.c              |   41 +-
 include/linux/bio.h               |    2 +
 include/linux/iomap.h             |    1 +
 include/uapi/linux/btrfs.h        |    1 +
 37 files changed, 3246 insertions(+), 118 deletions(-)
 create mode 100644 fs/btrfs/zoned.c
 create mode 100644 fs/btrfs/zoned.h
#120 lore raw
In reply to Naohiro Aota
Quoted 5 lines · Naohiro Aota

Johannes Thumshirn (1):
block: add bio_add_zone_append_page

Naohiro Aota (40):
iomap: support REQ_OP_ZONE_APPEND

From that one

Quoted 9 lines

btrfs: introduce ZONED feature flag
btrfs: get zone information of zoned block devices
btrfs: check and enable ZONED mode
btrfs: introduce max_zone_append_size
btrfs: disallow space_cache in ZONED mode
btrfs: disallow NODATACOW in ZONED mode
btrfs: disable fallocate in ZONED mode
btrfs: disallow mixed-bg in ZONED mode
btrfs: implement log-structured superblock for ZONED mode

up to this patch added to misc-next. There's still open question regarding the superblock copies, we had a chat about that with Johanness so he'll tell you more.