RE: [Patch] Optimize dropping of relation buffers using dlist - Mailing list pgsql-hackers

From: Jamison, Kirk/ジャミソン カーク <k.jamison@fujitsu.com>
> > (6)
> > +                    bufHdr->tag.blockNum >=
> > firstDelBlock[j])
> > +                    InvalidateBuffer(bufHdr);    /*
> > releases spinlock */
> >
> > The right side of >= should be cur_block.
> 
> Fixed.

>= should be =, shouldn't it?

Please measure and let us see just the recovery performance again because the critical part of the patch was modified.
Ifthe performance is good as the previous one, and there's no review interaction with others in progress, I'll mark the
patchas ready for committer in a few days.
 


 Regards
Takayuki Tsunakawa