On further reflection, we'd better do the same in int.c.
We previously heard of the same problem in int24div(), so there's not a
good reason to suppose the problem is confined to cases involving int8.
Branch
------
REL8_2_STABLE
Details
-------
http://git.postgresql.org/pg/commitdiff/1b5a7779cbe37a750f799b6314ceb43b8a0e5aba
Modified Files
--------------
src/backend/utils/adt/int.c | 38 ++++++++++++++++++++++++++++++++++++++
1 files changed, 38 insertions(+), 0 deletions(-)