Horton

Re: Buffer locking is special (hints, checksums, AIO writes)

Mikhail Nikalayeu <mihailnikalayeu@gmail.com>

View thread
From: Mihail Nikalayeu <mihailnikalayeu@gmail.com>
To: Andres Freund <andres@anarazel.de>
Cc: pgsql-hackers@postgresql.org, Melanie Plageman <melanieplageman@gmail.com>, Thomas Munro <thomas.munro@gmail.com>, Heikki Linnakangas <hlinnaka@iki.fi>, Noah Misch <noah@leadboat.com>
Date: 2025-08-23T10:31:53Z
Lists: pgsql-hackers

Commits

Same data as JSON: GET /api/v1/messages/:b64id/commits the thread's linked commits as JSON, with link sources. API reference →

  1. Require share-exclusive lock to set hint bits and to flush

  2. lwlock: Remove ForEachLWLockHeldByMe

  3. bufmgr: Implement buffer content locks independently of lwlocks

  4. bufmgr: Change BufferDesc.state to be a 64-bit atomic

  5. heapam: Add batch mode mvcc check and use it in page mode

  6. freespace: Don't modify page without any lock

  7. heapam: Move logic to handle HEAP_MOVED into a helper function

  8. bufmgr: Optimize & harmonize LockBufHdr(), LWLockWaitListLock()

  9. bufmgr: Add one-entry cache for private refcount

  10. bufmgr: Separate keys for private refcount infrastructure

  11. Add pg_atomic_unlocked_write_u64

  12. Rename BUFFERPIN wait event class to BUFFER

  13. bufmgr: Turn BUFFER_LOCK_* into an enum

  14. lwlock: Fix, currently harmless, bug in LWLockWakeup()

  15. bufmgr: Use atomic sub for unpinning buffers

  16. bufmgr: Allow some buffer state modifications while holding header lock

  17. bufmgr: Fix valgrind checking for buffers pinned in StrategyGetBuffer()

  18. bufmgr: Don't lock buffer header in StrategyGetBuffer()

  19. bufmgr: fewer calls to BufferDescriptorGetContentLock

  20. bufmgr: Fix signedness of mask variable in BufferSync()

  21. bufmgr: Introduce FlushUnlockedBuffer

  22. Improve ReadRecentBuffer() scalability

Hello, Andres!

Andres Freund <andres@anarazel.de>:
>
>    As part of that introduce a new lock level in-between share and exclusive
>    locks (provisionally called share-exclusive, but I hate it). The new lock
>    level allows other share lockers, but can only be held by one backend.
>
>    This allows to change the rules so that:
>
>    1) Share lockers are not allowed to modify buffers anymore
>    2) Hint bits need the new lock mode (conditionally upgrading the lock in SetHintBits())
>    3) Write IO needs to the new lock level

IIUC, it may be mapped to existing locking system:

BUFFER_LOCK_SHARE          ---->      AccessShareLock
new lock mode                           ---->     ExclusiveLock
BUFFER_LOCK_EXCLUSIVE   ---->     AccessExclusiveLock

So, it all may be named:

BUFFER_LOCK_ACCESS_SHARE
BUFFER_LOCK_EXCLUSIVE
BUFFER_LOCK_ACCESS_EXCLUSIVE

being more consistent with table-level locking system.

Greetings,
Mikhail.