BTRFS corrupt leaf / open_ctree failed

[SuperManni72]

Hello,

since I have updated from Ubuntu 22.02.1 (Kernel 5.15, btrfs-progs 5.16.2) to Ubuntu 24.04. Now, with Kernel 6.8 and btrfs-progs 6.6.3-1 (and also with Kernel 6.11), I cannot mount my BTRFS disks any more.

dmesg gives me:

[Sun Feb 23 15:08:11 2025] BTRFS info (device dm-5): first mount of filesystem d560d735-e812-49a7-bcc2-d883cbeae2f4
[Sun Feb 23 15:08:11 2025] BTRFS info (device dm-5): using crc32c (crc32c-intel) checksum algorithm
[Sun Feb 23 15:08:11 2025] BTRFS info (device dm-5): disk space caching is enabled
[Sun Feb 23 15:08:12 2025] BTRFS critical (device dm-5): corrupt leaf: block=20387303407616 slot=33 extent bytenr=9103199440896 len=8238972779538548086 invalid extent data backref objectid value 13163
[Sun Feb 23 15:08:12 2025] BTRFS error (device dm-5): read time tree block corruption detected on logical 20387303407616 mirror 1
[Sun Feb 23 15:08:12 2025] BTRFS critical (device dm-5): corrupt leaf: block=20387303407616 slot=33 extent bytenr=9103199440896 len=8238972779538548086 invalid extent data backref objectid value 13163
[Sun Feb 23 15:08:12 2025] BTRFS error (device dm-5): read time tree block corruption detected on logical 20387303407616 mirror 2
[Sun Feb 23 15:08:12 2025] BTRFS error (device dm-5): failed to read block groups: -5
[Sun Feb 23 15:08:12 2025] BTRFS error (device dm-5): open_ctree failed

The following checks have been done:

1. btrfsck --readonly /dev/mapper/data1_crypt

Opening filesystem to check...
Checking filesystem on /dev/mapper/data1_crypt
UUID: d560d735-e812-49a7-bcc2-d883cbeae2f4
[1/7] checking root items
[2/7] checking extents
[3/7] checking free space cache
[4/7] checking fs roots
[5/7] checking only csums items (without verifying data)
[6/7] checking root refs
[7/7] checking quota groups skipped (not enabled on this FS)
found 5883597525113 bytes used, no error found
total csum bytes: 5729361204
total tree bytes: 15929737216
total fs tree bytes: 9227829248
total extent tree bytes: 635011072
btree space waste bytes: 1907771393
file data blocks allocated: 57498707316736
 referenced 46500400185344

2. btrfs-find-root /dev/mapper/data1_crypt

Superblock thinks the generation is 294543
Superblock thinks the level is 1
Found tree root at 19305371582464 gen 294543 level 1
Well block 19305203253248(gen: 294542 level: 1) seems good, but generation/level doesn't match, want gen: 294543 level: 1
Well block 19305194897408(gen: 294541 level: 1) seems good, but generation/level doesn't match, want gen: 294543 level: 1
Well block 19305183739904(gen: 294540 level: 0) seems good, but generation/level doesn't match, want gen: 294543 level: 1
Well block 19305371500544(gen: 294538 level: 0) seems good, but generation/level doesn't match, want gen: 294543 level: 1
Well block 19305367502848(gen: 294538 level: 0) seems good, but generation/level doesn't match, want gen: 294543 level: 1

3. btrfs rescue super-recover -v /dev/mapper/data1_crypt

All Devices:
        Device: id = 2, name = /dev/mapper/data2_crypt
        Device: id = 1, name = /dev/mapper/data1_crypt
Before Recovering:
        [All good supers]:
                device name = /dev/mapper/data2_crypt
                superblock bytenr = 65536
                device name = /dev/mapper/data2_crypt
                superblock bytenr = 67108864
                device name = /dev/mapper/data2_crypt
                superblock bytenr = 274877906944
                device name = /dev/mapper/data1_crypt
                superblock bytenr = 65536
                device name = /dev/mapper/data1_crypt
                superblock bytenr = 67108864
                device name = /dev/mapper/data1_crypt
                superblock bytenr = 274877906944
        [All bad supers]:

All supers are valid, no need to recover

4. smartctl -a

--> No errors, no problems

5. btrfs scrub is done monthly:

sudo btrfs device stats /dev/mapper/data1_crypt
[/dev/mapper/data1_crypt].write_io_errs    0
[/dev/mapper/data1_crypt].read_io_errs     0
[/dev/mapper/data1_crypt].flush_io_errs    0
[/dev/mapper/data1_crypt].corruption_errs  0
[/dev/mapper/data1_crypt].generation_errs  0

What I have tried until now (with a clean boot from Ubuntu 22.04 Live-ISO):

6. Several mount tries (without sucess on Kernel 6.8/6.11):

mount -o recovery /dev/mapper/data1_crypt /media/ddd

mount -vs -t btrfs  -o ro,recovery,errors=continue /dev/mapper/data1_crypt /media/ddd

mount -t btrfs -o degraded,noatime,nodiratime,clear_cache /dev/mapper/data1-crypt /mnt/Daten

7. btrfs rescue zero-log /dev/mapper/data1-crypt

8. btrfs rescue clear-ino-cache /dev/mapper/data1-crypt

Unable to find block group for 0
ERROR: failed to clear ino cache: No space left on device

9. btrfs rescue clear-space-cache v1 /dev/mapper/data1-crypt

Unable to find block group for 0
ERROR: failed to clear free space cache
extent buffer leak: start 20449393967104 len 16384

10. btrfs inspect-internal tree-stats /dev/mapper/data1-crypt

warning, device 2 is missing
Calculating size of root tree
Total size: 2.28MiB
Inline data: 0.00B
Total seeks: 140
Forward seeks: 70
Backward seeks: 70
Avg seek len: 2.09TiB
Seek histogram
16384 - 185712640: 21 ###
290275328 - 50034622464: 21 ###
61473767424 - 490582196224: 21 ###
490944790528 - 932145905664: 21 ###
992462012416 - 1556186071040: 21 ###
1556985511936 - 10512543825920: 21 ###
10512621191168 - 12069176721408: 8 #
Total clusters: 4
Avg cluster size: 36.00KiB
Min cluster size: 32.00KiB
Max cluster size: 48.00KiB
Total disk spread: 11.00TiB
Total read time: 0 s 29 us
Levels: 2
Total nodes: 146
On level 0: 145
On level 1: 1 (avg fanout 145)
Calculating size of extent tree
Segmentation fault

dmesg
[Tue Mar 4 08:50:46 2025] btrfs[7804]: segfault at 0 ip 0000564b44894e65 sp 00007ffe6b099710 error 4 in btrfs[564b4482b000+71000]
[Tue Mar 4 08:50:46 2025] Code: 8d 74 24 30 48 8d ac 24 98 00 00 00 49 89 c4 4c 89 f7 e8 43 a5 f9 ff 31 c0 b9 30 00 00 00 48 89 ef f3 ab 31 f6 48 8d 7c 24 10 <49> 8b 04 24 48 c7 84 24 e8 00 00 00 ff ff ff ff 0f b6 98 dc 00 00

Now, I do not know what I can do else... and hope, someone can help me.

And nevertheless it's strange to me, that I can mount the disk with Kernel 5.15 and no error will be displayed…

Thank you & Best regards,
Markus