[qt/qtqa/gerrit]: Summary of bulk changes made

Git repository change summary for qt/qtqa/gerrit
Pushed by mirror-service into branch 'upstream/master'.
Changed from e6955a51409dc489da7552975f7a565b7df53133 to e70cefe8a283c5bbf8b2d55f2788bab73bbcfbef
Acknowledgement was received that this change introduces only existing code that has been pushed to another public open source repository.

This change contains the following new commits:

Git commit eb4a7eb6a3db126f9f1c1cb7b4dfa3dc7aa33a7b by Adithya Chakilam on 30/07/2026 at 06:00..
index: recover from crashes between NoteDb write and index write

Gerrit's BatchUpdate writes changes to NoteDb (git) and then to the
search index as two separate, non-atomic steps. If the process is
killed between those two steps, the change is durably committed to git
but the index is stale, with no automatic recovery. Introduce a WAL
protocol to close this gap.

Release-Notes: Recovery of missed change index updates after a crash/user interrupt can now be enabled via `index.staleChangeRecovery`
Change-Id: Id1248cc9ef4d8a81cf10650c8bd1e7d563bd6a8f
https://invent.kde.org/qt/qtqa/gerrit/-/commit/eb4a7eb6a3db126f9f1c1cb7b4dfa3dc7aa33a7b

Git commit ace60d47d05689fb9074b7fb5fbbc0b8960cf1da by Gerrit Code Review (on behalf of Adithya C) on 07/08/2026 at 15:08..
Merge "index: recover from crashes between NoteDb write and index write"
https://invent.kde.org/qt/qtqa/gerrit/-/commit/ace60d47d05689fb9074b7fb5fbbc0b8960cf1da

Git commit e70cefe8a283c5bbf8b2d55f2788bab73bbcfbef by Gerrit Code Review (on behalf of Kaushik Lingarkar) on 07/08/2026 at 16:13..
Update git submodules

* Update plugins/replication from branch 'master'
  to f0b8c225e060c49072c5a000d5462cb46f5c5b53
  - Distributor: Skip re-fire of already pending tasks
    
    On every cycle the distributor streams all tasks under waiting/ and
    calls fireFromStorage() for each one, which allocates a fresh
    ReplicationState and consolidates it into the pending PushOne for that
    URI. PushOne deduplicates refs, since refBatchesToPush is a Set, but
    its stateMap is a ListMultimap that appends unconditionally. A task
    that stays in waiting/ across cycles therefore grows the pending
    task's notification list by one entry per ref per cycle while adding
    no work at all.
    
    The distributor already snapshots the ref updates carried by pending
    push tasks, in order to prune queue entries whose stored counterpart
    is gone. Each streamed update is removed from that snapshot, leaving
    only the prunable ones behind by the end. Reuse that removal as the
    test, i.e a non-null result means a pending task on this node is already
    holding the update, so re-firing it would append a redundant state. Skip
    it, and let only updates with no pending task reach fireFromStorage().
    Tasks written by another primary are absent from the snapshot and
    continue to be fired, so cluster distribution is unaffected.
    
    This change removes the repeat re-fires when Prune.TRUE rather than
    every possible source of stateMap growth. The startup replay passes
    Prune.FALSE and is deliberately left uncovered. It walks each waiting/
    file exactly once, so no task is fired twice into the same pending push
    and the growth is not reachable there today.
    
    Change-Id: Ib9e77f9cf99cab3d5a6f110f0578ae02eafe9198
    
  - Distributor: Add tests for firing of tasks from storage
    
    Add test coverage to ensure the distributor fires an update which no
    pending task on this node is holding, and the startup replay fires
    unconditionally, without looking at the pending push tasks.
    
    Change-Id: If7fb97cbe9e64dc21dbfb5593541ed2367dad1f5
https://invent.kde.org/qt/qtqa/gerrit/-/commit/e70cefe8a283c5bbf8b2d55f2788bab73bbcfbef