Rocksolid Light - comp.os.vms - Re: SOBGEQ vs SOBGTR

SOBGEQ vs SOBGTR

<urbmvf$108iv$1@dont-email.me>

https://news.novabbs.org/computers/article-flat.php?id=33559&group=comp.os.vms#33559

Path: i2pn2.org!i2pn.org!eternal-september.org!feeder3.eternal-september.org!news.eternal-september.org!.POSTED!not-for-mail
From: ldo@nz.invalid (Lawrence D'Oliveiro)
Newsgroups: comp.os.vms
Subject: SOBGEQ vs SOBGTR
Date: Sat, 24 Feb 2024 03:14:55 -0000 (UTC)
Organization: A noiseless patient Spider
Lines: 9
Message-ID: <urbmvf$108iv$1@dont-email.me>
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Injection-Date: Sat, 24 Feb 2024 03:14:55 -0000 (UTC)
Injection-Info: dont-email.me; posting-host="90cd705a0b104282cd54d6df08a2cdc8";
logging-data="1057375"; mail-complaints-to="abuse@eternal-september.org"; posting-account="U2FsdGVkX19l74g7MkxR1F0qegl5navU"
User-Agent: Pan/0.155 (Kherson; fc5a80b8)
Cancel-Lock: sha1:/lVjFCZoA1Y8CyZdO/y+ioexlI0=

by: Lawrence D'Oliv - Sat, 24 Feb 2024 03:14 UTC

... set up loop count in Rn ...
BR 9000$
1000$:
... body of loop ...
9000$: SOBGEQ Rn, 1000$

Why branch and use SOBGEQ instead of SOBGTR? So that, if the loop
count is initially zero, the branch falls right through without
executing the body of the loop.

Re: SOBGEQ vs SOBGTR

<urbnq5$109l4$1@dont-email.me>

copy mid

https://news.novabbs.org/computers/article-flat.php?id=33561&group=comp.os.vms#33561

copy link Newsgroups: comp.os.vms

Path: i2pn2.org!i2pn.org!eternal-september.org!feeder3.eternal-september.org!news.eternal-september.org!.POSTED!not-for-mail
From: arne@vajhoej.dk (Arne Vajhøj)
Newsgroups: comp.os.vms
Subject: Re: SOBGEQ vs SOBGTR
Date: Fri, 23 Feb 2024 22:29:08 -0500
Organization: A noiseless patient Spider
Lines: 45
Message-ID: <urbnq5$109l4$1@dont-email.me>
References: <urbmvf$108iv$1@dont-email.me>
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
Injection-Date: Sat, 24 Feb 2024 03:29:09 -0000 (UTC)
Injection-Info: dont-email.me; posting-host="927881bcdcbd985d655aa867c10f2f7d";
logging-data="1058468"; mail-complaints-to="abuse@eternal-september.org"; posting-account="U2FsdGVkX18aInQybtDkIJ6HAPPJB1NwsFBFJCzCayY="
User-Agent: Mozilla Thunderbird
Cancel-Lock: sha1:Idpu0ucF+EIqgEhSLI4QnAsufbc=
Content-Language: en-US
In-Reply-To: <urbmvf$108iv$1@dont-email.me>

by: Arne Vajhøj - Sat, 24 Feb 2024 03:29 UTC

On 2/23/2024 10:14 PM, Lawrence D'Oliveiro wrote:
> ... set up loop count in Rn ...
> BR 9000$
> 1000$:
> ... body of loop ...
> 9000$: SOBGEQ Rn, 1000$
>
> Why branch and use SOBGEQ instead of SOBGTR? So that, if the loop
> count is initially zero, the branch falls right through without
> executing the body of the loop.

With the details provided then it will be pure
speculation.

n = whatever
loop:
if n <= 0 goto endloop
...
n = n - 1
goto loop
endloop:

has more instructions than:

n = whatever
loop:
...
n = n - 1
if n > 0 goto loop

but the second has the problem of not handling the n = 0
case properly.

n = whatever
goto checkfirst
loop:
...
n = n - 1
checkfirst:
if n > 0 goto loop

looks like a hack to fix that problem.

Arne

Re: SOBGEQ vs SOBGTR

<urcp60$16pfi$2@dont-email.me>

copy mid

https://news.novabbs.org/computers/article-flat.php?id=33564&group=comp.os.vms#33564

copy link Newsgroups: comp.os.vms

Path: i2pn2.org!i2pn.org!eternal-september.org!feeder3.eternal-september.org!news.eternal-september.org!.POSTED!not-for-mail
From: arne@vajhoej.dk (Arne Vajhøj)
Newsgroups: comp.os.vms
Subject: Re: SOBGEQ vs SOBGTR
Date: Sat, 24 Feb 2024 07:58:41 -0500
Organization: A noiseless patient Spider
Lines: 50
Message-ID: <urcp60$16pfi$2@dont-email.me>
References: <urbmvf$108iv$1@dont-email.me> <urbnq5$109l4$1@dont-email.me>
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 8bit
Injection-Date: Sat, 24 Feb 2024 12:58:40 -0000 (UTC)
Injection-Info: dont-email.me; posting-host="927881bcdcbd985d655aa867c10f2f7d";
logging-data="1271282"; mail-complaints-to="abuse@eternal-september.org"; posting-account="U2FsdGVkX182NZGlGhwrFXvvRLZhwEj0UbY1jaTYiBQ="
User-Agent: Mozilla Thunderbird
Cancel-Lock: sha1:SmozntBrcmZHlGMzBzUvxvgvlpU=
Content-Language: en-US
In-Reply-To: <urbnq5$109l4$1@dont-email.me>

by: Arne Vajhøj - Sat, 24 Feb 2024 12:58 UTC

On 2/23/2024 10:29 PM, Arne Vajhøj wrote:
> On 2/23/2024 10:14 PM, Lawrence D'Oliveiro wrote:
>>          ... set up loop count in Rn ...
>>          BR      9000$
>> 1000$:
>>          ... body of loop ...
>> 9000$: SOBGEQ Rn, 1000$
>>
>> Why branch and use SOBGEQ instead of SOBGTR? So that, if the loop
>> count is initially zero, the branch falls right through without
>> executing the body of the loop.
>
> With the details provided then it will be pure
> speculation.
>
>     n = whatever
> loop:
>     if n <= 0 goto endloop
>     ...
>     n = n - 1
>     goto loop
> endloop:
>
> has more instructions than:
>
>     n = whatever
> loop:
>     ...
>     n = n - 1
>     if n > 0 goto loop
>
> but the second has the problem of not handling the n = 0
> case properly.
>
>     n = whatever
>     goto checkfirst
> loop:
>     ...
>     n = n - 1
> checkfirst:
>     if n > 0 goto loop
>
> looks like a hack to fix that problem.

The entire thing reminds me of Fortran 66 vs 77
on DO loops.

Arne

Torque is cheap.

computers / comp.os.vms / Re: SOBGEQ vs SOBGTR

Subject	Author
SOBGEQ vs SOBGTR	Lawrence D'Oliveiro
Re: SOBGEQ vs SOBGTR	Arne Vajhøj
Re: SOBGEQ vs SOBGTR	Arne Vajhøj