Hello Jan and the ext4 team,

This patch series follows up on the previous discussion regarding
converting hard J_ASSERT panics into graceful journal aborts.

In v1, we addressed a specific panic on unlock. Per Jan's suggestion,
I have audited fs/jbd2/transaction.c for other low-hanging fruit
where state machine invariants are enforced by J_ASSERT inside
functions that natively support error returns.

Changes in v2:

    Patch 1: The original fix, unmodified. Collected the Reviewed-by
    tags from the v1 thread.

    Patch 2: New patch resulting from the broader audit. It systematically
    replaces J_ASSERTs with WARN_ON_ONCE and graceful -EINVAL returns
    across 6 core transaction lifecycle functions.

For Patch 2, careful attention was paid to ensuring spinlocks are safely
dropped before triggering jbd2_journal_abort(), reference counts
remain balanced, and no memory is leaked on the error paths.

Call-chain tracing confirms that upstream VFS callers (including ext4,
ocfs2, and others) already cleanly intercept these error codes to
abort the filesystem handle rather than crashing the server.

Milos Nikic (2):
  jbd2: gracefully abort instead of panicking on unlocked buffer
  jbd2: gracefully abort on transaction state corruptions

 fs/jbd2/transaction.c | 102 +++++++++++++++++++++++++++++++++---------
 1 file changed, 80 insertions(+), 22 deletions(-)

-- 
2.53.0