Re: [HACKERS] UPDATE of partition key
Robert Haas <robertmhaas@gmail.com>
On Fri, Jan 12, 2018 at 5:12 AM, Amit Khandekar <amitdkhan.pg@gmail.com> wrote:
> The reason why I am having map_required field inside a structure along
> with the map, as against a separate array, is so that we can do the
> on-demand allocation for both per-leaf array and per-subplan array.
Putting the map_required field inside the structure with the map makes
it completely silly to do the 0/1/2 thing, because the whole structure
is going to be on the same cache line anyway. It won't save anything
to access the flag instead of a pointer in the same struct. Also,
the uint8 will be followed by 7 bytes of padding, because the pointer
that follows will need to begin on an 8-byte boundary (at least, on
64-bit machines), so this will use more memory.
What I suggest is:
#define MT_CONVERSION_REQUIRED_UNKNOWN 0
#define MT_CONVERSION_REQUIRED_YES 1
#define MT_CONVERSION_REQUIRED_NO 2
In ModifyTableState:
uint8 *mt_per_leaf_tupconv_required;
TupleConversionMap **mt_per_leaf_tupconv_maps;
In PartitionTupleRouting:
int *subplan_partition_offsets;
When you initialize the ModifyTableState, do this:
mtstate->mt_per_leaf_tupconv_required = palloc0(sizeof(uint8) *
numResultRelInfos);
mtstate->mt_per_leaf_tupconv_maps = palloc0(sizeof(TupleConversionMap
*) * numResultRelInfos);
When somebody needs a map, then
(1) if they need it by subplan index, first use
subplan_partition_offsets to convert it to a per-leaf index
(2) then write a function that takes the per-leaf index and does this:
switch (mtstate->mt_per_leaf_tupconv_required[leaf_part_index])
{
case MT_CONVERSION_REQUIRED_UNKNOWN:
map = convert_tuples_by_name(...);
if (map == NULL)
mtstate->mt_per_leaf_tupconv_required[leaf_part_index] =
MT_CONVERSION_REQUIRED_NO;
else
{
mtstate->mt_per_leaf_tupconv_required[leaf_part_index] =
MT_CONVERSION_REQUIRED_YES;
mtstate->mt_per_leaf_tupconv_maps[leaf_part_index] = map;
}
return map;
case MT_CONVERSION_REQUIRED_YES:
return mtstate->mt_per_leaf_tupconv_maps[leaf_part_index];
case MT_CONVERSION_REQUIRED_NO:
return NULL;
}
--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company
Commits
-
Avoid referencing off the end of subplan_partition_offsets.
- 945f71db8452 11.0 landed
-
Allow UPDATE to move rows between partitions.
- 2f178441044b 11.0 landed
-
Remove useless lookup of root partitioned rel in ExecInitModifyTable().
- dca48d145e0e 11.0 cited
-
Factor error generation out of ExecPartitionCheck.
- 19c47e7c8202 11.0 landed
-
Minor preparatory refactoring for UPDATE row movement.
- ef6087ee5fa8 11.0 landed
-
Simplify and encapsulate tuple routing support code.
- cc6337d2fed5 11.0 landed
-
Avoid coercing a whole-row variable that is already coerced.
- 1c497fa72df7 11.0 landed
-
Use ResultRelInfo ** rather than ResultRelInfo * for tuple routing.
- 60f7c0abef03 11.0 landed
-
Make RelationGetPartitionDispatchInfo expand depth-first.
- 77b6b5e9ceca 11.0 cited
-
Expand partitioned tables in PartDesc order.
- 30833ba154e0 11.0 cited
-
Use a real RT index when setting up partition tuple routing.
- f81a91db4d1c 10.0 cited
-
Fix transition tables for partition/inheritance.
- 501ed02cf6f4 10.0 cited
-
Fix confusion about number of subplans in partitioned INSERT setup.
- 78a030a44196 10.0 cited
-
Prevent BEFORE triggers from violating partitioning constraints.
- 15ce775faa42 10.0 cited
-
Fire per-statement triggers on partitioned tables.
- e180c8aa8caf 10.0 cited
-
Fix reporting of violations in ExecConstraints, again.
- c0a8ae7be392 10.0 cited
-
Don't scan partitioned tables.
- d3cc37f1d801 10.0 cited
-
Allow FDWs to push down quals without breaking EvalPlanQual rechecks.
- 5fc4c26db512 9.6.0 cited