Hi Darrick, On Tue, Jun 09, 2020 at 03:12:39PM -0700, Darrick J. Wong wrote: ... > > Oops, I hit send too quickly; the computation looks ok to me. > > Granted that's probably due to having written another btree geometry > calculation function. :D > > Also, I think init_rmapbt_cursor does a similar trick and therefore > suffers from a similar bug. See the comment "Leave enough slack in the > rmapbt that we can insert..." around line 1412 or so? These comments are fine :) I will try to fix them all in the next version. Thanks, Gao Xiang
