Re: Setting pd_lower in GIN metapage

On Mon, Sep 25, 2017 at 10:13 AM, Michael Paquier
<michael(dot)paquier(at)gmail(dot)com> wrote:
> On Sun, Sep 24, 2017 at 2:25 PM, Amit Kapila <amit(dot)kapila16(at)gmail(dot)com> wrote:
>> Added and updated the comments for both btree and hash index patches.
>
> I don't have real complaints about this patch, this looks fine to me.
>
> + * Currently, the advantage of setting pd_lower is in limited cases like
> + * during wal_consistency_checking or while logging for unlogged relation
> + * as for all other purposes, we initialize the metapage. Note, it also
> + * helps in page masking by allowing to mask unused space.
> I would have reworked this comment a bit, say like that:
> Setting pd_lower is useful for two cases which make use of WAL
> compressibility even if the meta page is initialized at replay:
> - Logging of init forks for unlogged relations.
> - wal_consistency_checking logs extra full-page writes, and this
> allows masking of the unused space of the page.
>
> Now I often get complains that I suck at this exercise ;)
>

I understand that and I think there are always multiple ways to write
same information. It might be better to pass this patch series to
committer if you don't see any mistake because he might anyway change
some comments before committing.

--
With Regards,
Amit Kapila.
EnterpriseDB: http://www.enterprisedb.com