Re: WAL Consistency checking for hash indexes

On Mon, Mar 13, 2017 at 10:36 AM, Ashutosh Sharma <ashu(dot)coek88(at)gmail(dot)com> wrote:
> Couple of review comments,,
>
> You may also need to update the documentation as now we are also going
> to support wal consistency check for hash index. The current
> documentation does not include hash index.
>
> + only records originating from those resource managers. Currently,
> + the supported resource managers are <literal>heap</>,
> + <literal>heap2</>, <literal>btree</>, <literal>gin</>,
> + <literal>gist</>, <literal>sequence</>, <literal>spgist</>,
> + <literal>brin</>, and <literal>generic</>. Only

Did that, committed this. Also ran pgindent over hash_mask() and
fixed an instance of dubious capitalization.

> Following comment in hash_mask() may require changes if patch for
> 'Microvacuum support for Hash Index - [1]' gets committed.
>
> + /*
> + * In hash bucket and overflow pages, it is possible to modify the
> + * LP_FLAGS without emitting any WAL record. Hence, mask the line
> + * pointer flags.
> + * See hashgettuple() for details.
> + */

If that's so, then that patch is responsible for updating these
comments (and the code, if necessary).

--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company