Path: ...!eternal-september.org!feeder3.eternal-september.org!news.eternal-september.org!eternal-september.org!.POSTED!not-for-mail
From: Terje Mathisen <terje.mathisen@tmsw.no>
Newsgroups: comp.arch
Subject: Re: Cost of handling misaligned access
Date: Tue, 18 Feb 2025 22:09:54 +0100
Organization: A noiseless patient Spider
Lines: 37
Message-ID: <vp2sv2$1skve$1@dont-email.me>
References: <5lNnP.1313925$2xE6.991023@fx18.iad> <vnosj6$t5o0$1@dont-email.me>
 <2025Feb3.075550@mips.complang.tuwien.ac.at> <volg1m$31ca1$1@dont-email.me>
 <voobnc$3l2dl$1@dont-email.me>
 <0fc4cc997441e25330ff5c8735247b54@www.novabbs.org>
 <vp0m3f$1cth6$1@dont-email.me>
 <74142fbdc017bc560d75541f3f3c5118@www.novabbs.org>
 <20250218150739.0000192a@yahoo.com>
 <0357b097bbbf6b87de9bc91dd16757e3@www.novabbs.org>
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: quoted-printable
Injection-Date: Tue, 18 Feb 2025 22:09:55 +0100 (CET)
Injection-Info: dont-email.me; posting-host="8e33d5afbcf681e220a218745a8a45b1";
	logging-data="1987566"; mail-complaints-to="abuse@eternal-september.org";	posting-account="U2FsdGVkX1/f5mRYUDwflfOvd9QdSq/2Q4D0HI2vNqBT9CIKfdTYrA=="
User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:128.0) Gecko/20100101
 Firefox/128.0 SeaMonkey/2.53.20
Cancel-Lock: sha1:TAXdUA3PfXI9IoPzwIHUCYfqWZs=
In-Reply-To: <0357b097bbbf6b87de9bc91dd16757e3@www.novabbs.org>
Bytes: 2669

MitchAlsup1 wrote:
> On Tue, 18 Feb 2025 13:07:39 +0000, Michael S wrote:
>=20
>> On Tue, 18 Feb 2025 02:55:33 +0000
>> mitchalsup@aol.com (MitchAlsup1) wrote:
>>>
>>> It takes Round Nearest Odd to perform Kahan-Babashuka Summation.
>>>
>>
>> Are you aware of any widespread hardware that supplies Round to Neares=
t
>> with tie broken to Odd? Or of any widespread language that can request=

>> such rounding mode?
>=20
> No, No
>=20
>> Until both, implementing RNO on niche HW looks to me as wastage of bot=
h
>> HW resources and of space in your datasheet.
>=20
> They way I implement it, it is only an additional 10=C2=B1 gates.

With discrete logic, it should be identical to RNE, except for flipping=20
the ulp bit when deciding upon the rounding direction, right?

With a full 4-bit lookup table you need a few more gates, but that is=20
still the obvious way to implement rounding in SW. (It is only ceil()=20
and floor() that requires the sign bit as input, the remaining rounding=20
modes can make do with ulp+guard+sticky.

Terje

--=20
- <Terje.Mathisen at tmsw.no>
"almost all programming can be viewed as an exercise in caching"