Path: csiph.com!fu-berlin.de!uni-berlin.de!individual.net!not-for-mail
From: minforth <minforth@gmx.net>
Newsgroups: comp.lang.forth
Subject: Re: 3dup again
Date: Fri, 3 Oct 2025 11:09:07 +0200
Lines: 24
Message-ID: <mk9i5hFld65U1@mid.individual.net>
References: <2025Sep30.183350@mips.complang.tuwien.ac.at> <mk5b0uFubm8U1@mid.individual.net> <68de4aaa@news.ausics.net> <2025Oct2.224440@mips.complang.tuwien.ac.at> <nnd$36c0a5f5$2a9d9179@1179128ec025831e>
Mime-Version: 1.0
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
X-Trace: individual.net 7WMSfiQMoGMDAiPDrYv1egOeD9MCs8e5VjxPc14kgu8q1lKi10
Cancel-Lock: sha1:Ry8KsVXSdDOq3UTXGMsopkxglMk= sha256:udbVNz8pLZBeZqFmvIAJR6wbG+D0qOhvTx46YPMzpSY=
User-Agent: Mozilla Thunderbird
In-Reply-To: <nnd$36c0a5f5$2a9d9179@1179128ec025831e>
Xref: csiph.com comp.lang.forth:134239

Am 03.10.2025 um 11:02 schrieb albert@spenarnc.xs4all.nl:
> In article <2025Oct2.224440@mips.complang.tuwien.ac.at>,
> Anton Ertl <anton@mips.complang.tuwien.ac.at> wrote:
>> dxf <dxforth@gmail.com> writes:
>>> For 3DUP I believe this is the one to beat:
>>>
>>> : 3DUP ( a b c -- a b c a b c )  dup 2over rot ;
>>>
>>> With NTF/LFX the locals version will break even.
>>
>> As we already discussed in the thread including
>> <2021Sep11.083507@mips.complang.tuwien.ac.at>, NTF/LXF produces the
>> same (optimal for the calling convention used by NTF/LXF) code for
>> 3DUP versions using the data stack, return stack, or locals.  That's
>> because the actual data flow is always the same, and NTF/LXF can see
>> this data flow in all three cases.
> 
> The problem with 3DUP is that it is actually used in context.
> What is the data that is going to 3DUP ped? In my view this
> amounts to double use of data that is in registers (32 in the riscv)
> anyway, after an optimiser does his thing.
> 

Code inlining will mend it.