Horton

Re: index prefetching

Peter Geoghegan <pg@bowt.ie>

View thread View on postgr.es
From: Peter Geoghegan <pg@bowt.ie>
To: Tomas Vondra <tomas@vondra.me>
Cc: Andres Freund <andres@anarazel.de>, Alexandre Felipe <o.alexandre.felipe@gmail.com>, Thomas Munro <thomas.munro@gmail.com>, Nazir Bilal Yavuz <byavuz81@gmail.com>, Robert Haas <robertmhaas@gmail.com>, Melanie Plageman <melanieplageman@gmail.com>, PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>, Georgios <gkokolatos@protonmail.com>, Konstantin Knizhnik <knizhnik@garret.ru>, Dilip Kumar <dilipbalaut@gmail.com>
Date: 2026-03-12T17:27:00Z
Lists: pgsql-hackers 
On Thu, Mar 12, 2026 at 10:43 AM Tomas Vondra <tomas@vondra.me> wrote:
> I took a quick look at 0001 and 0002 patches, and I think it's mostly
> ready to go.

Pushed both patches just now. Thanks for the review!

> I only have a couple very minor comments:
>
> 1) v12-0001-Use-simple-hash-for-PrivateRefCount.patch

> - the "common/hashfn.h" include seems in the wrong place

Fixed in the pushed patch.

> 2) v12-0002-Don-t-allocate-_bt_search-stack.patch
>
> - seems like a reasonable optimization for most callers
> - _bt_search could set "stack_in = new_stack;" in the first if block, I
> think it's nicer

Agreed. Done that way in the commit that I pushed.

> - My only concern is that _bt_search is part of the public API (at least
> technically, as it's in nbtree.h), and we're removing _bt_freestack()
> from the API so the callers won't have a convenient way to free it.
> Maybe that's not an issue in practice, because doing a local version of
> freestack is easy.

I don't think that it's an issue. _bt_search is very much an
nbtree-internal function (amcheck notwithstanding), and as you point
out it's easy enough for an extension to create its own version of
_bt_freestack in the unlikely event that it needs to.

--
Peter Geoghegan

Commits

  1. read stream: Split decision about look ahead for AIO and combining

  2. read_stream: Only increase read-ahead distance when waiting for IO

  3. aio: io_uring: Trigger async processing for large IOs

  4. heapam: Keep buffer pins across index scan resets.

  5. heapam: Track heap block in IndexFetchHeapData.

  6. Move heapam_handler.c index scan code to new file.

  7. Rename heapam_index_fetch_tuple argument for clarity.

  8. Optimize fast-path FK checks with batched index probes

  9. read_stream: Prevent distance from decaying too quickly

  10. read_stream: Issue IO synchronously while in fast path

  11. bufmgr: Return whether WaitReadBuffers() needed to wait

  12. aio: io_uring: Allow IO methods to check if IO completed in the background

  13. bufmgr: Make UnlockReleaseBuffer() more efficient

  14. Add fake LSN support to hash index AM.

  15. Make IndexScanInstrumentation a pointer in executor scan nodes.

  16. Use fake LSNs to improve nbtree dropPin behavior.

  17. Move fake LSN infrastructure out of GiST.

  18. Use simplehash for backend-private buffer pin refcounts.

  19. nbtree: Avoid allocating _bt_search stack.

  20. bufmgr: Fix use of wrong variable in GetPrivateRefCountEntrySlow()

  21. Conditional locking in pgaio_worker_submit_internal

  22. Reduce ExecSeqScan* code size using pg_assume()

  23. Fix rare bug in read_stream.c's split IO handling.

  24. Remove HeapBitmapScan's skip_fetch optimization

  25. Optimize nbtree backwards scans.

  26. Fix multiranges to behave more like dependent types.

  27. Add EXPLAIN (MEMORY) to report planner memory consumption

  28. Optimize nbtree backward scan boundary cases.

  29. Increment xactCompletionCount during subtransaction abort.

  30. Add nbtree Valgrind buffer lock checks.

  31. Add nbtree high key "continuescan" optimization.

  32. Reduce pinning and buffer content locking for btree scans.

  33. Teach btree to handle ScalarArrayOpExpr quals natively.