[SRU][X/B][PATCH v2 0/1] CVE-2024-50006

[Alice C. Munduruca]

v2 -> Fixed the structure of the patchset to match the flat hierarchy standard.

[ Impact ]

A deadlock is possible in ext4 when `ext4_ind_migrate` is called
and races with a write syscall. Calling `ext4_journal_stop` in
the correct order mitigates this.

[ Fix ]

xenial: backported patch from upstream.
bionic: cleanly cherry picked patch from upstream.

[ Tests ]

Compile and boot tested.

[ Where problems could occur ]

Given changes are relevant to ext4, it is possible that this
could cause instabilities in filesystems under certain
circumstances. As this is also the status quo, and changes are
minimal, it still seems like regression potential is small.

Artem Sadovnikov (1):
  ext4: fix i_data_sem unlock order in ext4_ind_migrate()

 fs/ext4/migrate.c | 2 +-
 1 file changed, 1 insertions(+), 1 deletions(-)

-- 
2.48.1