Endless (logging of) transid failure in btrfsck after chunk-recover with btrfs-progs 3.18 on Linux 3.17.7

Endless (logging of) transid failure in btrfsck after chunk-recover with btrfs-progs 3.18 on Linux 3.17.7

[Karl-Philipp Richter]

[-- Attachment #1: Type: text/plain, Size: 986 bytes --]

Hi,
After running `sudo btrfs rescue chunk-recover -y -v /dev/loop0` with 4
unrecoverable chunks (of 619) I'm having trouble with `btrfsck` which logs

    parent transid verify failed on 4210688 wanted 61241 found 7983179
    Ignoring transid failure

apparently for ever (let it run for 1 hour which produced the entry >
3.5M times, then killed it). There must be something wrong because the
program needs to succeed or fail, right? After aborting `btrfsck`
`btrfs-debug-tree` and `btrfs-zero-log` both fail with

    Check tree block failed, want=16384, have=0
    read block failed check_tree_block
    Couldn't read chunk root
    unable to open /dev/loop0

Please find attached the output of `rescue chunk-recover` and `btrfsck`.
The filesystem is on a `dd` image, mounted via a loop device with
`losetup`. This is the only version of the filesystem I have. I'm
running Ubuntu 14.10 amd64. I can't test 3.18.1 and 3.19rc2.

Any help or hints are appreciated.

Best regards,
Kalle

[-- Attachment #2: chuck-recover_btrfsck_failure.log.tar.xz --]
[-- Type: application/x-xz, Size: 14956 bytes --]