Bruce Momjian <bruce@momjian.us> writes:
> Tom Lane wrote:
>> I've applied this but I'm now having some second thoughts about it,
>> because I'm seeing an actual *decrease* in pgbench numbers from the
>> immediately prior CVS HEAD code.
> The attached patch requires the new row to fit, and 10% to be free on
> the page. Would someone test that?
At the moment, I cannot replicate any consistent difference between
CVS head with the patch, without the patch, with the patch plus your
BLCKSZ/10 limit addition, or with a variant BLCKSZ/32 limit addition.
That's whether I use HEAD's broken version of pgbench or one from late
July. So I'm feeling a tad frustrated ... but I have no evidence in
favor of changing what is in CVS, and accordingly recommend that we
leave well enough alone for 8.2.
regards, tom lane