From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <e@yhbt.net>
X-Spam-Checker-Version: SpamAssassin 3.4.2 (2018-09-13) on dcvr.yhbt.net
X-Spam-Level: 
X-Spam-Status: No, score=-4.0 required=3.0 tests=ALL_TRUSTED,BAYES_00
	shortcircuit=no autolearn=ham autolearn_force=no version=3.4.2
Received: from localhost (dcvr.yhbt.net [127.0.0.1])
	by dcvr.yhbt.net (Postfix) with ESMTP id EE1261F5AE;
	Mon, 20 Jul 2020 20:27:30 +0000 (UTC)
Date: Mon, 20 Jul 2020 20:27:30 +0000
From: Eric Wong <e@yhbt.net>
To: Jean Boussier <jean.boussier@shopify.com>
Cc: unicorn-public@yhbt.net
Subject: Re: [PATCH] Add early hints support
Message-ID: <20200720202730.GA16172@dcvr>
References: <20200716105037.GA26605@dcvr>
 <242F0859-0F83-4F14-A0FF-5BE392BB01E6@shopify.com>
 <20200716121643.GA26942@dcvr>
 <C0BBAE49-63DA-4629-996B-29B3EB21B90E@shopify.com>
 <20200717011955.GA23522@dcvr>
 <624157B8-884C-42EB-8515-AFC6807D6F8D@shopify.com>
 <20200720100958.GA7043@dcvr>
 <F908D215-8646-450C-BA7F-9704F1E9D738@shopify.com>
 <20200720105538.GA13340@dcvr>
 <4274A592-FC06-4886-8013-429ABF689695@shopify.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
In-Reply-To: <4274A592-FC06-4886-8013-429ABF689695@shopify.com>
List-Id: <unicorn-public.yhbt.net>

Jean Boussier <jean.boussier@shopify.com> wrote:
> > upcase seems VERY compelling in a micro benchmark since it can
> > go straight into opt_case_dispatch.  But I worry the extra
> > garbage might have a different effect in a real app, especially
> > with more headers.
> 
> It is indeed debatable. The thing is that on match, that Regexp
> will generate a MatchData and a garbage string. So I'd think
> the extra GC pressure would be totally negligible.

Ah good point.  I take it Ruby doesn't lazy-allocate or have
optimizations to quickly recycle unused MatchData?  (Maybe it
could, if somebody who hacks on the VM is reading this).

> But I agree it's hard to tell without a larger more representative benchmark.

Yes, I would accept patches to speed these up with comprehensive
benchmark :)

> > Maybe:
> > 
> > 	 (val || ''.freeze).include?("\n".freeze)
> > 
> > Can work for those buggy apps, though...
> 
> Yes. or:
> 
>    vall && val.include?...

Yes.