Re: Remove page-read callback from XLogReaderState.

At Thu, 8 Apr 2021 23:51:34 +1200, Thomas Munro <thomas(dot)munro(at)gmail(dot)com> wrote in
> On Thu, Apr 8, 2021 at 9:46 PM Thomas Munro <thomas(dot)munro(at)gmail(dot)com> wrote:
> > I squashed the patch set into one because half of them were fixups,
> > and the two main patches were really parts of the same change and
> > should go in together.
> >
> > I fixed a few compiler warnings (GCC 10.2 reported several
> > uninitialised variables, comparisons that are always true, etc) and
> > some comments. You can see these in the fixup patch.
>
> Pushed. Luckily there are plenty more improvements possible for
> XLogReader/XLogDecoder in the next cycle.

I'm surprised to see this pushed this soon. Thanks for pushing this!

And thanks for fixing the remaining mistakes including some stupid
ones..

At Thu, 8 Apr 2021 10:04:26 +1200, Thomas Munro <thomas(dot)munro(at)gmail(dot)com> wrote in
> There is a stray elog(HOGE) :-)

Ugggghhhh! This looks like getting slipped-in while investigating
another issue.. Thanks for preventing the repository from being
contaminated by such a thing..

At Thu, 8 Apr 2021 21:46:06 +1200, Thomas Munro <thomas(dot)munro(at)gmail(dot)com> wrote in
> I think maybe it it should really be XLogReaderSetInputData(state,
> tli, data, size) in a later release. In the meantime, I changed it to
> XLogReaderSetInputData(state, size), hope that name make sense...

Sounds better. I didn't like that page-readers are allowed to touch
XLogReaderStats.seg directly. Anyway it would be a small change.

regards.

--
Kyotaro Horiguchi
NTT Open Source Software Center