Re: What integer C type to use

Liste des GroupesRevenir à c arch 
Sujet : Re: What integer C type to use
De : already5chosen (at) *nospam* yahoo.com (Michael S)
Groupes : comp.arch
Date : 15. Mar 2024, 13:01:09
Autres entêtes
Organisation : A noiseless patient Spider
Message-ID : <20240315140109.00006006@yahoo.com>
References : 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
User-Agent : Claws Mail 4.1.1 (GTK 3.24.34; x86_64-w64-mingw32)
On Thu, 14 Mar 2024 23:56:04 -0400
"Paul A. Clayton" <paaronclayton@gmail.com> wrote:

On 3/13/24 3:24 PM, MitchAlsup1 wrote:
Stefan Monnier wrote: 
[snip]
So, short vectors have a fairly free hand at shuffling data across
their vector (e.g. bitmatrix transpose), and they can be
implemented/scheduled/dispatched just like any other instruction,
but the vector length tends to be severely limited and exposed all
over the place. 
 
Consuming OpCode space like nobody's business. 
 
Is that necessarily the case? Excluding the shuffle operations, I
think only loads and stores would need to have length specifiers.
Shuffle operations become much more expensive with larger
'vectors', so providing the same granularity of shuffle for larger
vectors seems questionable. (With scatter/gather, permute/shuffle
may be less useful anyway.)
 
<snip>
 
Am I missing something when I assume that lane-based (SIMD)
operations do not need size information in the instruction? The
extra metadata is not free (perhaps especially as that controls
execution at least for efficiency), but if opcode explosion is so
undesirable using metadata might be preferred.

I would guess that Mitch operates under assumption that we are still
talking about ISA that is very similar to CRAY Y-MP.
I.e. non-load-store, but more like AVX512, where one of the source
operands could be in memory.
That is, I had never seen CRAY Y-MP ISA docs, but would imagine that it
was like that.


Date Sujet#  Auteur
11 Mar 24 * Re: What integer C type to use17MitchAlsup1
12 Mar 24 `* Re: What integer C type to use16David Brown
12 Mar 24  `* Re: What integer C type to use15Michael S
12 Mar 24   +* Re: What integer C type to use13MitchAlsup1
12 Mar 24   i`* Re: What integer C type to use12Michael S
12 Mar 24   i `* Re: What integer C type to use11MitchAlsup1
13 Mar 24   i  +* Re: What integer C type to use9Michael S
13 Mar 24   i  i+- Re: What integer C type to use1MitchAlsup1
13 Mar 24   i  i`* Re: What integer C type to use7Stefan Monnier
13 Mar 24   i  i +* Re: What integer C type to use5MitchAlsup1
15 Mar 24   i  i i`* Re: What integer C type to use4Paul A. Clayton
15 Mar 24   i  i i +- Re: What integer C type to use1Michael S
15 Mar 24   i  i i `* Re: What integer C type to use2MitchAlsup1
15 Mar 24   i  i i  `- Re: What integer C type to use1MitchAlsup1
14 Mar 24   i  i `- Re: What integer C type to use1Michael S
15 Mar 24   i  `- Re: What integer C type to use1Terje Mathisen
13 Mar 24   `- Re: What integer C type to use1Thomas Koenig

Haut de la page

Les messages affichés proviennent d'usenet.

NewsPortal