On Fri, 17 Jul 2026 at 11:00, Alexander Lakhin <[email protected]> wrote:
>
> Hello Michael and Matthias,
>
> Could you please look at other similar anomalies I have discovered?:
>
> (Assuming the attached OOM injection patch applied.)
> 1) An issue in generate_partition_qual():
> --- a/src/backend/utils/cache/partcache.c
> +++ b/src/backend/utils/cache/partcache.c
> @@ -419,3 +419,5 @@ generate_partition_qual(Relation rel)
>           oldcxt = MemoryContextSwitchTo(rel->rd_partcheckcxt);
> +oom_prob = 0.1;
>           rel->rd_partcheck = copyObject(result);
> +oom_prob = 0;
>           MemoryContextSwitchTo(oldcxt);
[...]
> triggers
> TRAP: failed Assert("cache->cc_tupdesc != NULL"), File: "catcache.c", Line: 
> 1112, PID: 43932

While the Assert is caused by other effects, it would still leave an
incorrect rd_partcheck around in relcaches, so here's a patch that
fixes that issue with rd_partcheckcxt. It also avoids leaking the
context itself into the CacheMemoryContext when an OOM happens, so
that's a neat addition.


Kind regards,

Matthias van de Meent
Databricks (https://www.databricks.com)

Attachment: v1-0001-Fix-corrupted-caches-when-we-OOM-getting-the-part.patch
Description: Binary data

Reply via email to