mirror of https://github.com/citusdata/citus.git
* Use CalculateUniformHashRangeIndex in HashPartitionId
INT32_MIN definition can change among different platforms hence it is
possible to get overflow, we would see crashes because of this in debian
distros. We have already solved a similar problem with introducing
CalculateUniformHashRangeIndex method, hence to solve it we can use the
same method, this also removes some duplication and has a single place
to decide that.
* Use PG_INT32_XX instead of INT32_XX to be safer
(cherry picked from commit
|
||
---|---|---|
.. | ||
task_tracker.c | ||
task_tracker_protocol.c | ||
worker_create_or_replace.c | ||
worker_data_fetch_protocol.c | ||
worker_drop_protocol.c | ||
worker_file_access_protocol.c | ||
worker_merge_protocol.c | ||
worker_partition_protocol.c | ||
worker_shard_visibility.c | ||
worker_sql_task_protocol.c | ||
worker_truncate_trigger_protocol.c |