Fix assign_record_type_typmod().
authorJeff Davis <jdavis@postgresql.org>
Fri, 9 Jul 2021 21:15:48 +0000 (14:15 -0700)
committerJeff Davis <jdavis@postgresql.org>
Sat, 10 Jul 2021 17:26:38 +0000 (10:26 -0700)
commitdd0e37cc1598050ec38fa289908487d4f5c96dca
treefdcc3bc228a3c24d70d8fe6cf4969023a8e2007b
parent9f6be2e79f88c65cef7cfeb0580e8b2ba74974f6
Fix assign_record_type_typmod().

If an error occurred in the wrong place, it was possible to leave an
unintialized entry in the hash table, leading to a crash. Fixed.

Also, be more careful about the order of operations so that an
allocation error doesn't leak memory in CacheMemoryContext or
unnecessarily advance NextRecordTypmod.

Backpatch through version 11. Earlier versions (prior to 35ea75632a5)
do not exhibit the problem, because an uninitialized hash entry
contains a valid empty list.

Author: Sait Talha Nisanci <Sait.Nisanci@microsoft.com>
Reviewed-by: Andres Freund
Discussion: https://postgr.es/m/HE1PR8303MB009069D476225B9A9E194B8891779@HE1PR8303MB0090.EURPRD83.prod.outlook.com
Backpatch-through: 11
src/backend/utils/cache/typcache.c