* Tom Lane (tgl@sss.pgh.pa.us) wrote:
> [ trimming -novice from the cc list ]
>
> Jeff Janes <jeff.janes@gmail.com> writes:
> > On Tue, Mar 1, 2016 at 7:40 AM, Tom Lane <tgl@sss.pgh.pa.us> wrote:
> >> (Problem is reproducible in 9.5 and HEAD, but not 9.4.)
>
> > Bisects down to:
>
> > 606c0123d627b37d5ac3f7c2c97cd715dde7842f is the first bad commit
> > commit 606c0123d627b37d5ac3f7c2c97cd715dde7842f
> > Author: Simon Riggs <simon@2ndQuadrant.com>
> > Date: Tue Nov 18 10:24:55 2014 +0000
>
> > Reduce btree scan overhead for < and > strategies
>
> Hmm ... just from the commit message, this seems much more likely to be
> the cause than does the buffer locking patch Stephen fingered. Stephen,
> how'd you identify 2ed5b87f as being the problem?
Badly. :)
I didn't expect it to be something that far back and was just going
backwards through reverting/testing patches that looked like
candidates, but forgot to create the index when I tested the revert of
that, which made it look like reverting it 'fixed' it.
Apologies for the noise.
Thanks!
Stephen