On Sat, Oct 21, 2023 at 09:56:13PM -0400, Bruce Momjian wrote:
> I did some more research. It turns out that the source slot/planSlot is
> populating its pg_attribute information via makeTargetEntry() and it
> has no concept of a storage type.
>
> Digging further, I found that we cannot get rid of the the use of
> att->attstorage != TYPSTORAGE_PLAIN in macros ATT_IS_PACKABLE and
> VARLENA_ATT_IS_PACKABLE macros in src/backend/access/common/heaptuple.c
> because there are internal uses of fill_val() that can't handle packed
> varlena headers.
>
> I ended up with a doc patch that adds a C comment about this odd
> behavior and removes doc text about PLAIN storage not using packed
> headers.
Oops, patch attached.
--
Bruce Momjian <br...@momjian.us> https://momjian.us
EDB https://enterprisedb.com
Only you can decide what is important to you.
diff --git a/doc/src/sgml/storage.sgml b/doc/src/sgml/storage.sgml
index 148fb1b49d..3ea4e5526d 100644
--- a/doc/src/sgml/storage.sgml
+++ b/doc/src/sgml/storage.sgml
@@ -456,9 +456,7 @@ for storing <acronym>TOAST</acronym>-able columns on disk:
<listitem>
<para>
<literal>PLAIN</literal> prevents either compression or
- out-of-line storage; furthermore it disables use of single-byte headers
- for varlena types.
- This is the only possible strategy for
+ out-of-line storage. This is the only possible strategy for
columns of non-<acronym>TOAST</acronym>-able data types.
</para>
</listitem>
diff --git a/src/backend/access/common/heaptuple.c b/src/backend/access/common/heaptuple.c
index d6a4ddfd51..c52d40dce0 100644
--- a/src/backend/access/common/heaptuple.c
+++ b/src/backend/access/common/heaptuple.c
@@ -68,7 +68,16 @@
#include "utils/memutils.h"
-/* Does att's datatype allow packing into the 1-byte-header varlena format? */
+/*
+ * Does att's datatype allow packing into the 1-byte-header varlena format?
+ * While functions that use TupleDescAttr() and assign attstorage =
+ * TYPSTORAGE_PLAIN cannot use packed varlena headers, functions that call
+ * TupleDescInitEntry() use typeForm->typstorage (TYPSTORAGE_EXTENDED) and
+ * can use packed varlena headers, e.g.:
+ * CREATE TABLE test(a VARCHAR(10000) STORAGE PLAIN);
+ * INSERT INTO test VALUES (repeat('A',10));
+ * This can be verified with pageinspect.
+ */
#define ATT_IS_PACKABLE(att) \
((att)->attlen == -1 && (att)->attstorage != TYPSTORAGE_PLAIN)
/* Use this if it's already known varlena */
diff --git a/src/backend/utils/adt/rangetypes.c b/src/backend/utils/adt/rangetypes.c
index d65e5625c7..24bad52923 100644
--- a/src/backend/utils/adt/rangetypes.c
+++ b/src/backend/utils/adt/rangetypes.c
@@ -2608,7 +2608,8 @@ range_contains_elem_internal(TypeCacheEntry *typcache, const RangeType *r, Datum
* values into a range object. They are modeled after heaptuple.c's
* heap_compute_data_size() and heap_fill_tuple(), but we need not handle
* null values here. TYPE_IS_PACKABLE must test the same conditions as
- * heaptuple.c's ATT_IS_PACKABLE macro.
+ * heaptuple.c's ATT_IS_PACKABLE macro. See the comments thare for more
+ * details.
*/
/* Does datatype allow packing into the 1-byte-header varlena format? */
