Re: Add 64-bit XIDs into PostgreSQL 15

Evgeny <evorop@gmail.com>

From: Evgeny <evorop@gmail.com>

To: pgsql-hackers@lists.postgresql.org

Date: 2024-12-08T17:17:17Z

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 →

Add SLRU tests for 64-bit page case
- a60b8a58f435 17.0 landed
Make use FullTransactionId in 2PC filenames
- 5a1dfde8334b 17.0 landed
Use larger segment file names for pg_notify
- 2cdf131c46e6 17.0 landed
Index SLRUs by 64-bit integers rather than by 32-bit integers
- 4ed8f0913bfd 17.0 landed

Attachments

v56-0001-Use-64-bit-format-output-for-multixact-offsets.patch (text/x-patch) patch v56-0001
v56-0002-Use-64-bit-multixact-offsets.patch (text/x-patch) patch v56-0002
v56-0003-Make-pg_upgrade-convert-multixact-offsets.patch (text/x-patch) patch v56-0003
v56-0004-Get-rid-of-MultiXactMemberFreezeThreshold-call.patch (text/x-patch) patch v56-0004
v56-0005-Support-64-bit-integer-GUCs.patch (text/x-patch) patch v56-0005
v56-0006-Use-64-bit-format-to-output-XIDs.patch (text/x-patch) patch v56-0006
v56-0007-Use-64-bit-FullTransactionId-instead-of-Epoch-xi.patch (text/x-patch) patch v56-0007
v56-0008-Use-64-bit-SLRU-pages-in-callers.patch (text/x-patch) patch v56-0008
v56-0009-Use-64-bit-XIDs.patch (text/x-patch) patch v56-0009
v56-0010-Add-initdb-option-to-initialize-cluster-with-non.patch (text/x-patch) patch v56-0010
v56-0011-README.XID64.patch (text/x-patch) patch v56-0011
v56-0012-Fixed-the-stale-xid64-problem-at-the-heapam_tupl.patch (text/x-patch) patch v56-0012

**

*Hello! *

*

The next edition of the xid64 patch has been implemented.

In the presented  version:

1) gathered patches MXIDOFFSETv9 [1], GUC64v2 [2], xid64 (v52-v55) (the 
latest versions from the relevant threads have been used);

2) the resulted patches have been rebased onto the current master 
(3f9b9621766);

3) implemented the HeapTupleCopyRawXidsFromPage. This function avoids 
substitution of xids and returns their real values even in a case of a 
frozen tuple;

4) fixed issues revealed by tests;

5) in addition, fixed the "stale xid64" problem at the 
heapam_tuple_satisfies_snapshot function. This issue mostly appeared in 
practice as the "missing chunk" error during operations on TOASTs of the 
pg_statistic table.

*

**

*Looking forward to cooperation and review,*

*

Evgeny Voropaev,

TantorLabs, LLC.

<evorop@gmail.com>

<evgeny.voropaev@tantorlabs.ru>

*

**

*[1] MXIDOFFSETv9 - 
https://www.postgresql.org/message-id/CACG%3Dezajc_Pcqmy6fcq-N8%2BLzCRMzOzJzez2_BgHEu-6RVJtKQ%40mail.gmail.com*

**

**

*[2] GUC64v2 - 
https://www.postgresql.org/message-id/4c0c1924-a68c-4b3f-bf81-5af13b1a9087%40gmail.com*

**

**

**

**