Re: amcheck support for BRIN indexes

Arseniy Mukhin <arseniy.mukhin.dev@gmail.com>

View thread

From: Arseniy Mukhin <arseniy.mukhin.dev@gmail.com>

To: Tomas Vondra <tomas@vondra.me>

Cc: Álvaro Herrera <alvherre@kurilemu.de>, PostgreSQL Hackers <pgsql-hackers@lists.postgresql.org>

Date: 2025-07-22T15:43:20Z

Lists: pgsql-hackers

Attachments

v8-0001-brin-refactoring.patch (text/x-patch) patch v8-0001
v8-0004-amcheck-brin_index_check-heap-all-indexed.patch (text/x-patch) patch v8-0004
v8-0002-amcheck-brin_index_check-index-structure-check.patch (text/x-patch) patch v8-0002
v8-0003-amcheck-common_verify-register-snapshot-with-indc.patch (text/x-patch) patch v8-0003

Hi,

While reviewing gist amcheck patch [1] I realized that brin amcheck
also must check if current snapshot is OK with index indcheckxmin (as
btree, gist do it). Currently this check is contained in btree amcheck
code, but other AMs need it for heapallindexed as well, so I moved it
from btree to verify_common (0003 patch).

Also I returned a consistentFn approach in heapallindexed as it seems
more preferable. But it's not a big deal to return to the within_range
approach if needed.


[1] https://www.postgresql.org/message-id/flat/41F2A10C-4577-413B-9140-BE81CCE04A60%40yandex-team.ru#dc22ff33596f63f554cc551958131cde


Best regards,
Arseniy Mukhin