Re: inherit support for foreign tables - Mailing list pgsql-hackers

Hello,

Sorry, this was no relation with this patch.

ForeignNext materializes the slot, which would be any of physical
and virtual tuple, when system column was requested. If it was a
virtual one, file_fdw makes this, heap_form_tuple generates the
tuple as DatumTuple. The result is a jumble of virtual and
physical. But the returning slot has tts_tuple so the caller
interprets it as a physical one.

Anyway the requests for xmin/xmax could not be prevented
beforehand in current framework, I did rewrite the physical tuple
header so as to really be a physical one.

This would be another patch, so I will put this into next CF if
this don't get any immediate objection.


> By the way, I tried xmin and xmax for the file_fdw tables.
> 
> postgres=# select tableoid, xmin, xmax, * from passwd1;
> tableoid | xmin |    xmax    |     uname     | pass |  uid  |  gid  | .. 
>    16396 |  244 | 4294967295 | root          | x    |     0 |     0 | root...
>    16396 |  252 | 4294967295 | bin           | x    |     1 |     1 | bin...
>    16396 |  284 | 4294967295 | daemon        | x    |     2 |     2 | daemon...
> 
> The xmin and xmax apparently doesn't look sane. After some
> investigation, I found that they came from the following place in
> heap_form_tuple(), (call stach is show below)


regards,

-- 
Kyotaro Horiguchi
NTT Open Source Software Center
diff --git a/src/backend/executor/nodeForeignscan.c b/src/backend/executor/nodeForeignscan.c
index 9cc5345..728db14 100644
--- a/src/backend/executor/nodeForeignscan.c
+++ b/src/backend/executor/nodeForeignscan.c
@@ -22,6 +22,8 @@ */#include "postgres.h"
+#include "access/transam.h"
+#include "access/htup_details.h"#include "executor/executor.h"#include "executor/nodeForeignscan.h"#include
"foreign/fdwapi.h"
@@ -58,8 +60,21 @@ ForeignNext(ForeignScanState *node)     */    if (plan->fsSystemCol && !TupIsNull(slot))    {
+        bool        was_virtual_tuple = (slot->tts_tuple == NULL);        HeapTuple    tup =
ExecMaterializeSlot(slot);
+        if (was_virtual_tuple)
+        {
+            /*
+             * ExecMaterializeSlot fills the tuple header as a
+             * DatumTupleFields if the slot was a virtual tuple, but a
+             * physical one is needed here. So rewrite the tuple header as
+             * HeapTupleFirelds.
+             */
+            HeapTupleHeaderSetXmin(tup->t_data, FrozenTransactionId);
+            HeapTupleHeaderSetXmax(tup->t_data, InvalidTransactionId);
+            HeapTupleHeaderSetCmin(tup->t_data, FirstCommandId);
+        }        tup->t_tableOid = RelationGetRelid(node->ss.ss_currentRelation);    }