Re: bcachefs mount issue

Viacheslav Dubeyko <slava@xxxxxxxxxxx> · Tue, 6 Aug 2024 11:37:41 +0300

On 8/6/24 03:50, Kent Overstreet wrote:
On Mon, Aug 05, 2024 at 09:35:09PM GMT, Viacheslav Dubeyko wrote:
Hi Kent,

As far as I  can see, I have found a mount issue. I believe that it’s a mkfs tool issue.

ENVIRONMENT:
Linux ssdfs-test-0070 6.10.0 #15 SMP PREEMPT_DYNAMIC Mon Aug  5 19:00:55 MSK 2024 x86_64 x86_64 x86_64 GNU/Linux

I am not sure how to share the mkfs.bcachefs tool version because this tool doesn’t show the version.

You're using an ancient version - I presume from Debian?

Debian hasn't been getting tools updates, you can't get anything modern
because of, I believe, a libsodium transition (?), and modern 1.9.x
versions aren't getting pushed out either.

I'll have to refer you to them - Jonathan, what's going on?

I am using Ubuntu:

lsb_release -a
No LSB modules are available.
Distributor ID:	Ubuntu
Description:	Ubuntu 22.04.3 LTS
Release:	22.04
Codename:	jammy

Thanks,
Slava.

REPRODUCTION PATH:

(1) Format partition by mkfs tool of any file system (for example, NILFS2)
(2) Mount the prepared volume
(3) Execute any file system operations on the volume
(4) Unmount the volume
(5) Format partition by mkfs.bcachefs tool
(6) Try to mount the prepared  bcachefs volume
(7) The bcachefs logic fails too mount the formatted volume

sudo mkfs.nilfs2 -f -b 4096 /dev/sda1
mkfs.nilfs2 (nilfs-utils 2.2.8)
Start writing file system initial data to the device
        Blocksize:4096  Device:/dev/sda1  Device Size:999292928
File system initialization succeeded !!

sudo mount /dev/sda1 /mnt/test/

mount
<skipped>
/dev/sda1 on /mnt/test type nilfs2 (rw,relatime)

Aug  5 19:14:40 ssdfs-test-0070 kernel: [  520.066975] NILFS (sda1): segctord starting. Construction interval = 5 seconds, CP frequency < 30 seconds
Aug  5 19:14:40 ssdfs-test-0070 nilfs_cleanerd[3854]: start
Aug  5 19:14:40 ssdfs-test-0070 nilfs_cleanerd[3854]: pause (clean check)

sudo umount /mnt/test

Aug  5 19:15:18 ssdfs-test-0070 nilfs_cleanerd[3854]: shutdown

sudo mkfs.bcachefs -f --block_size=4096 /dev/sda1
External UUID: 483fb669-63aa-4f41-b0ba-61eb2446c2fe
Internal UUID: 43527890-f6c8-43f1-bdd9-1c4936f71a8d
Device index: 0
Label:
Version: 14
Oldest version on disk: 14
Created: Mon Aug  5 19:20:32 2024
Squence number: 0
Block_size: 4.0K
Btree node size: 128.0K
Error action: ro
Clean: 0
Features: new_siphash,new_extent_overwrite,btree_ptr_v2,extents_above_btree_updates,btree_updates_journalled,new_varint,journal_no_flush,alloc_v2,extents_across_btree_nodes
Compat features:
Metadata replicas: 1
Data replicas: 1
Metadata checksum type: crc32c (1)
Data checksum type: crc32c (1)
Compression type: none (0)
Foreground write target: none
Background write target: none
Promote target: none
Metadata target:                none
String hash type: siphash (2)
32 bit inodes: 1
GC reserve percentage: 8%
Root reserve percentage: 0%
Devices: 1 live, 1 total
Sections: members
Superblock size: 816

Members (size 64):
   Device 0:
     UUID: 2c54ddfc-f50c-4d15-aa80-7d23474de3e6
     Size: 953.0M
     Bucket size: 128.0K
     First bucket: 0
     Buckets: 7624
     Last mount: (never)
     State: rw
     Group: (none)
     Data allowed: journal,btree,user
     Has data: (none)
     Replacement policy: lru
     Discard: 0
initializing new filesystem
going read-write
mounted with opts: (null)

sudo mount /dev/sda1 /mnt/test/

mount
<skipped>
/dev/sda1 on /mnt/test type nilfs2 (rw,relatime) <— completely unexpected

Aug  5 19:21:13 ssdfs-test-0070 kernel: [  912.678991] NILFS (sda1): broken superblock, retrying with spare superblock (blocksize = 1024)
Aug  5 19:21:13 ssdfs-test-0070 kernel: [  912.679835] NILFS (sda1): broken superblock, retrying with spare superblock (blocksize = 4096)
Aug  5 19:21:13 ssdfs-test-0070 kernel: [  912.706795] NILFS (sda1): segctord starting. Construction interval = 5 seconds, CP frequency < 30 seconds
Aug  5 19:21:13 ssdfs-test-0070 nilfs_cleanerd[4751]: start
Aug  5 19:21:13 ssdfs-test-0070 nilfs_cleanerd[4751]: pause (clean check)

sudo umount /mnt/test

sudo mount -t bcachefs /dev/sda1 /mnt/test/
mount: /mnt/test: wrong fs type, bad option, bad superblock on /dev/sda1, missing codepage or helper program, or other error.

Aug  5 19:23:01 ssdfs-test-0070 kernel: [ 1020.311715] bcachefs (sda1): mounting version 0.14: btree_ptr_sectors_written opts=noshard_inode_numbers,journal_reclaim_delay=1000,nojournal_transaction_names
Aug  5 19:23:01 ssdfs-test-0070 kernel: [ 1020.311753] bcachefs (sda1): recovering from clean shutdown, journal seq 4
Aug  5 19:23:01 ssdfs-test-0070 kernel: [ 1020.311782] bcachefs (sda1): Version upgrade required:
Aug  5 19:23:01 ssdfs-test-0070 kernel: [ 1020.311782] Doing incompatible version upgrade from 0.14: btree_ptr_sectors_written to 1.7: mi_btree_bitmap
Aug  5 19:23:01 ssdfs-test-0070 kernel: [ 1020.311782]   running recovery passes: check_allocations,check_alloc_info,check_lrus,check_btree_backpointers,check_backpointers_to_extents,check_extents_to_backpointers,check_alloc_to_lru_refs,bucket_gens_init,check_snapshot_trees,check_snapshots,check_subvols,check_subvol_children,delete_dead_snapshots,check_inodes,check_extents,check_indirect_extents,check_dirents,check_xattrs,check_root,check_subvolume_structure,check_directory_structure,check_nlinks,delete_dead_inodes,set_fs_needs_rebalance
Aug  5 19:23:01 ssdfs-test-0070 kernel: [ 1020.431462] bcachefs (sda1): alloc_read... done
Aug  5 19:23:01 ssdfs-test-0070 kernel: [ 1020.431675] bcachefs (sda1): stripes_read... done
Aug  5 19:23:01 ssdfs-test-0070 kernel: [ 1020.431688] bcachefs (sda1): snapshots_read... done
Aug  5 19:23:01 ssdfs-test-0070 kernel: [ 1020.431702] bcachefs (sda1): check_allocations...
Aug  5 19:23:01 ssdfs-test-0070 kernel: [ 1020.469609] dev 0 has wrong free buckets: got 0, should be 7537, fixing
Aug  5 19:23:01 ssdfs-test-0070 kernel: [ 1020.469683]  done
Aug  5 19:23:01 ssdfs-test-0070 kernel: [ 1020.471013] bcachefs (sda1): going read-write
Aug  5 19:23:01 ssdfs-test-0070 kernel: [ 1020.471766] bcachefs (sda1): journal_replay... done
Aug  5 19:23:01 ssdfs-test-0070 kernel: [ 1020.471794] bcachefs (sda1): check_alloc_info... done
Aug  5 19:23:01 ssdfs-test-0070 kernel: [ 1020.472921] bcachefs (sda1): check_lrus... done
Aug  5 19:23:01 ssdfs-test-0070 kernel: [ 1020.473309] bcachefs (sda1): check_btree_backpointers... done
Aug  5 19:23:01 ssdfs-test-0070 kernel: [ 1020.473960] bcachefs (sda1): check_backpointers_to_extents... done
Aug  5 19:23:01 ssdfs-test-0070 kernel: [ 1020.474483] bcachefs (sda1): check_extents_to_backpointers...
Aug  5 19:23:01 ssdfs-test-0070 kernel: [ 1020.474582] missing backpointer for btree=inodes l=1 u64s 11 type btree_ptr_v2 SPOS_MAX len 0 ver 0: seq d7dbe59ccc0e54fa written 24 min_key POS_MIN durability: 1 ptr: 0:78:0 gen 1
Aug  5 19:23:01 ssdfs-test-0070 kernel: [ 1020.474590]   got:   u64s 5 type deleted 0:20447232:0 len 0 ver 0
Aug  5 19:23:01 ssdfs-test-0070 kernel: [ 1020.474595]   want:  u64s 9 type backpointer 0:20447232:0 len 0 ver 0: bucket=0:78:0 btree=inodes l=1 offset=0:0 len=256 pos=SPOS_MAX, shutting down
Aug  5 19:23:01 ssdfs-test-0070 kernel: [ 1020.474645] bcachefs (sda1): inconsistency detected - emergency read only at journal seq 4
Aug  5 19:23:01 ssdfs-test-0070 kernel: [ 1020.474664] bcachefs (sda1): bch2_check_extents_to_backpointers(): error fsck_errors_not_fixed
Aug  5 19:23:01 ssdfs-test-0070 kernel: [ 1020.474682] bcachefs (sda1): bch2_fs_recovery(): error fsck_errors_not_fixed
Aug  5 19:23:01 ssdfs-test-0070 kernel: [ 1020.474692] bcachefs (sda1): bch2_fs_start(): error starting filesystem fsck_errors_not_fixed
Aug  5 19:23:01 ssdfs-test-0070 kernel: [ 1020.474842] bcachefs (sda1): unshutdown complete, journal seq 4
Aug  5 19:23:01 ssdfs-test-0070 kernel: [ 1020.595522] bcachefs: bch2_mount() error: fsck_errors_not_fixed

Thanks,
Slava.