Re: Binary search in fmgr_isbuiltin() is a bottleneck.

Andres Freund <andres(at)anarazel(dot)de> writes:
> On 2017-09-27 11:50:56 -0400, Tom Lane wrote:
>> Robert Haas <robertmhaas(at)gmail(dot)com> writes:
>>> I suppose an even better approach would be to build a perfect hash
>>> table at compile time so that nothing needs to be built at run-time at
>>> all, but I'm not sure it's worth the trouble.

>> Yeah, I was wondering about that too. It would likely mean adding a
>> compile time dependency on gperf or similar tool, but we could take
>> our standard approach of shipping the output in tarballs, so that only
>> developers would really need to install that tool.

> I'd been wondering about that too, but I'm not sure I buy that it's
> worth the effort. The only real argument I see is that there's probably
> multiple cases where it'd be potentially beneficial, not just here.

The other question that ought to be answered is whether a gperf hash
table would be faster. In principle it could be because of
guaranteed-no-collisions, but I have no experience with how fast the
constructed hash functions might be compared to our regular one.

To me, the real takeaway from this thread is that fmgr_isbuiltin()
needs optimization at all; I'd have guessed it didn't matter. But
now that we know that it does, it's worth looking hard at what we
can squeeze out of it.

regards, tom lane