Sujet : Re: 80286 protected mode
De : mitchalsup (at) *nospam* aol.com (MitchAlsup1)
Groupes : comp.archDate : 10. Oct 2024, 22:30:38
Autres entêtes
Organisation : Rocksolid Light
Message-ID : <35cb536e6310a38f0269788881cffdaf@www.novabbs.org>
References : 1 2 3 4 5 6 7 8 9 10 11 12
User-Agent : Rocksolid Light
On Thu, 10 Oct 2024 19:21:20 +0000, David Brown wrote:
On 10/10/2024 20:38, MitchAlsup1 wrote:
This is more a symptom of bad ISA design/evolution than of libc
writers needing superpowers.
>
No, it is not. It has absolutely /nothing/ to do with the ISA.
>
For example, if ISA contains an MM instruction which is the
embodiment of memmove() then absolutely no heroics are needed
of desired in the libc call.
>
>
The existence of a dedicated assembly instruction does not let you write
an efficient memmove() in standard C.
{
memmove( p, q, size );
}
Where the compiler produces the MM instruction itself. Looks damn
close to standard C to me !!
OR
for( int i = 0, i < size; i++ )
p[i] = q[i];
Which gets compiled to memcpy()--also looks to be standard C.
OR
p_struct = q_struct;
gets compiled to::
memmove( &p_struct, &q_struct, sizeof( q_struct ) );
also looks to be std C.
The thing is you are no longer writing memmove(), you are simply
teaching the compiler to recognizes its _use_ cases directly. In
addition, these will always be within spitting distance of as fast
as one can perform those activities.
That's why I said there was no
connection between the two concepts.
>
For some targets, it can be helpful to write memmove() in assembly or
using inline assembly, rather than in non-portable C (which is the
common case).
>
Thus, it IS a symptom of ISA evolution that one has to rewrite
memmove() every time wider SIMD registers are available.
>
It is not that simple.
>
There can often be trade-offs between the speed of memmove() and
memcpy() on large transfers, and the overhead in setting things up that
is proportionally more costly for small transfers.
Given that we are talking about GBOoO machines here, the several
AGEN units[1,2,3] have plenty of calculation BW to determine order
without wasting cycles getting started.
But given LBIO machine, the ability to process memory to memory moves
at cache port width is always an advantage except for cases needing
only 1 read or 1 write--if you build the HW with these in mind.
Often that can be
eliminated when the compiler optimises the functions inline - when the
compiler knows the size of the move/copy, it can optimise directly.
In HW they should always be optimized.
Date | Sujet | # | | Auteur |
16 Apr 24 | Re: Whether something is RISC or not (Re: PDP-8 theology, not Concertina II Progress) | 237 | | Lawrence D'Oliveiro |
16 Apr 24 | Re: Whether something is RISC or not (Re: PDP-8 theology, not Concertina II Progress) | 236 | | David Brown |
16 Apr 24 | Re: Whether something is RISC or not (Re: PDP-8 theology, not Concertina II Progress) | 1 | | MitchAlsup1 |
26 May 24 | Re: Whether something is RISC or not (Re: PDP-8 theology, not Concertina II Progress) | 1 | | MitchAlsup1 |
1 Oct 24 | Re: Whether something is RISC or not (Re: PDP-8 theology, not Concertina II Progress) | 233 | | MitchAlsup1 |
1 Oct 24 | Re: Whether something is RISC or not (Re: PDP-8 theology, not Concertina II Progress) | 232 | | Thomas Koenig |
1 Oct 24 | Re: Whether something is RISC or not (Re: PDP-8 theology, not Concertina II Progress) | 225 | | MitchAlsup1 |
2 Oct 24 | Re: Whether something is RISC or not (Re: PDP-8 theology, not Concertina II Progress) | 223 | | Brett |
3 Oct 24 | Re: Whether something is RISC or not (Re: PDP-8 theology, not Concertina II Progress) | 222 | | Lawrence D'Oliveiro |
3 Oct 24 | Re: Whether something is RISC or not (Re: PDP-8 theology, not Concertina II Progress) | 1 | | Brett |
3 Oct 24 | Re: Whether something is RISC or not (Re: PDP-8 theology, not Concertina II Progress) | 1 | | Anton Ertl |
3 Oct 24 | Re: Whether something is RISC or not (Re: PDP-8 theology, not Concertina II Progress) | 219 | | David Brown |
3 Oct 24 | Byte ordering (was: Whether something is RISC or not) | 218 | | Anton Ertl |
3 Oct 24 | Re: Byte ordering (was: Whether something is RISC or not) | 1 | | David Brown |
4 Oct 24 | Re: Byte ordering (was: Whether something is RISC or not) | 215 | | Lawrence D'Oliveiro |
4 Oct 24 | Re: Byte ordering | 1 | | Lynn Wheeler |
4 Oct 24 | Re: Byte ordering (was: Whether something is RISC or not) | 211 | | David Brown |
4 Oct 24 | Re: Byte ordering (was: Whether something is RISC or not) | 210 | | Anton Ertl |
4 Oct 24 | Re: Byte ordering | 5 | | BGB |
5 Oct 24 | Re: Byte ordering | 4 | | MitchAlsup1 |
5 Oct 24 | Re: Byte ordering | 2 | | BGB |
5 Oct 24 | Re: Byte ordering | 1 | | Lawrence D'Oliveiro |
5 Oct 24 | Re: Byte ordering | 1 | | Lawrence D'Oliveiro |
5 Oct 24 | Re: Byte ordering (was: Whether something is RISC or not) | 13 | | Lawrence D'Oliveiro |
5 Oct 24 | Re: Byte ordering (was: Whether something is RISC or not) | 12 | | Brett |
5 Oct 24 | Re: Byte ordering (was: Whether something is RISC or not) | 11 | | Anton Ertl |
5 Oct 24 | Re: Byte ordering (was: Whether something is RISC or not) | 10 | | Michael S |
6 Oct 24 | Re: Byte ordering | 1 | | Terje Mathisen |
6 Oct 24 | Re: Byte ordering (was: Whether something is RISC or not) | 8 | | Brett |
7 Oct 24 | Re: Byte ordering (was: Whether something is RISC or not) | 7 | | Lawrence D'Oliveiro |
7 Oct 24 | Re: Byte ordering (was: Whether something is RISC or not) | 6 | | Brett |
7 Oct 24 | Re: Byte ordering (was: Whether something is RISC or not) | 5 | | Michael S |
7 Oct 24 | Re: Byte ordering | 2 | | Stefan Monnier |
7 Oct 24 | Re: Byte ordering | 1 | | Michael S |
7 Oct 24 | Re: Byte ordering (was: Whether something is RISC or not) | 2 | | Lawrence D'Oliveiro |
8 Oct 24 | Re: Byte ordering | 1 | | Terje Mathisen |
6 Oct 24 | Re: Byte ordering | 191 | | David Brown |
6 Oct 24 | Re: Byte ordering | 190 | | Anton Ertl |
6 Oct 24 | Re: Byte ordering | 189 | | John Dallman |
7 Oct 24 | Re: Byte ordering | 20 | | Lawrence D'Oliveiro |
8 Oct 24 | Re: Byte ordering | 19 | | John Dallman |
9 Oct 24 | VMS/NT memory management (was: Byte ordering) | 1 | | Stefan Monnier |
15 Oct 24 | Re: Byte ordering | 2 | | Lawrence D'Oliveiro |
15 Oct 24 | Re: Byte ordering | 1 | | MitchAlsup1 |
15 Oct 24 | Re: Byte ordering | 15 | | Lawrence D'Oliveiro |
15 Oct 24 | Re: Byte ordering | 3 | | Michael S |
15 Oct 24 | Re: Byte ordering | 1 | | John Dallman |
18 Oct 24 | Re: Byte ordering | 1 | | Lawrence D'Oliveiro |
15 Oct 24 | Re: Byte ordering | 9 | | John Dallman |
16 Oct 24 | Re: Byte ordering | 7 | | George Neuner |
16 Oct 24 | Re: Byte ordering | 6 | | Terje Mathisen |
16 Oct 24 | Re: Byte ordering | 5 | | David Brown |
17 Oct 24 | Re: Byte ordering | 2 | | George Neuner |
17 Oct 24 | Re: Byte ordering | 1 | | David Brown |
17 Oct 24 | Re: clouds, not Byte ordering | 2 | | John Levine |
17 Oct 24 | Re: clouds, not Byte ordering | 1 | | David Brown |
18 Oct 24 | Re: Byte ordering | 1 | | Lawrence D'Oliveiro |
16 Oct 24 | Re: Byte ordering | 2 | | Paul A. Clayton |
18 Oct 24 | Re: Microkernels & Capabilities (was Re: Byte ordering) | 1 | | Lawrence D'Oliveiro |
7 Oct 24 | 80286 protected mode | 168 | | Anton Ertl |
7 Oct 24 | Re: 80286 protected mode | 5 | | Lars Poulsen |
7 Oct 24 | Re: 80286 protected mode | 4 | | Terje Mathisen |
7 Oct 24 | Re: 80286 protected mode | 1 | | Michael S |
7 Oct 24 | Re: 80286 protected mode | 2 | | Lawrence D'Oliveiro |
8 Oct 24 | Re: 80286 protected mode | 1 | | Terje Mathisen |
7 Oct 24 | Re: 80286 protected mode | 3 | | Brett |
7 Oct 24 | Re: 80286 protected mode | 2 | | Michael S |
7 Oct 24 | Re: 80286 protected mode | 1 | | Brett |
7 Oct 24 | Re: 80286 protected mode | 1 | | Lawrence D'Oliveiro |
8 Oct 24 | Re: 80286 protected mode | 152 | | MitchAlsup1 |
8 Oct 24 | Re: 80286 protected mode | 4 | | Lawrence D'Oliveiro |
8 Oct 24 | Re: 80286 protected mode | 3 | | MitchAlsup1 |
9 Oct 24 | Re: 80286 protected mode | 1 | | David Brown |
15 Oct 24 | Re: 80286 protected mode | 1 | | Lawrence D'Oliveiro |
8 Oct 24 | Re: 80286 protected mode | 147 | | Anton Ertl |
8 Oct 24 | Re: 80286 protected mode | 1 | | Robert Finch |
9 Oct 24 | Re: 80286 protected mode | 145 | | David Brown |
9 Oct 24 | Re: 80286 protected mode | 79 | | MitchAlsup1 |
9 Oct 24 | Re: 80286 protected mode | 78 | | David Brown |
9 Oct 24 | Re: 80286 protected mode | 77 | | Stephen Fuld |
10 Oct 24 | Re: 80286 protected mode | 2 | | MitchAlsup1 |
10 Oct 24 | Re: 80286 protected mode | 1 | | David Brown |
10 Oct 24 | Re: 80286 protected mode | 1 | | David Brown |
11 Oct 24 | Re: 80286 protected mode | 73 | | Tim Rentsch |
15 Oct 24 | Re: 80286 protected mode | 72 | | Stefan Monnier |
15 Oct 24 | Re: 80286 protected mode | 30 | | MitchAlsup1 |
16 Oct 24 | Re: 80286 protected mode | 25 | | MitchAlsup1 |
16 Oct 24 | Re: C and turtles, 80286 protected mode | 13 | | John Levine |
16 Oct 24 | Re: C and turtles, 80286 protected mode | 7 | | MitchAlsup1 |
16 Oct 24 | Re: C and turtles, 80286 protected mode | 6 | | John Levine |
17 Oct 24 | Re: C and turtles, 80286 protected mode | 5 | | Thomas Koenig |
20 Oct 24 | Re: C and turtles, 80286 protected mode | 4 | | Lawrence D'Oliveiro |
20 Oct 24 | Re: C and turtles, 80286 protected mode | 3 | | George Neuner |
22 Oct 24 | Re: C and turtles, 80286 protected mode | 2 | | Tim Rentsch |
22 Oct 24 | Re: C and turtles, 80286 protected mode | 1 | | George Neuner |
16 Oct 24 | Re: C and turtles, 80286 protected mode | 1 | | David Brown |
16 Oct 24 | Re: C and turtles, 80286 protected mode | 4 | | Paul A. Clayton |
17 Oct 24 | Re: C and turtles, 80286 protected mode | 1 | | David Brown |
20 Oct 24 | Re: C and turtles, 80286 protected mode | 2 | | Lawrence D'Oliveiro |
20 Oct 24 | Re: C and turtles, 80286 protected mode | 1 | | Paul A. Clayton |
16 Oct 24 | Re: 80286 protected mode | 7 | | Thomas Koenig |
17 Oct 24 | Re: 80286 protected mode | 3 | | George Neuner |
17 Oct 24 | Re: 80286 protected mode | 1 | | Tim Rentsch |
16 Oct 24 | Re: 80286 protected mode | 3 | | David Brown |
17 Oct 24 | Re: 80286 protected mode | 1 | | Tim Rentsch |
16 Oct 24 | Re: 80286 protected mode | 41 | | David Brown |
9 Oct 24 | Re: 80286 protected mode | 51 | | Thomas Koenig |
13 Oct 24 | Re: 80286 protected mode | 14 | | Anton Ertl |
8 Oct 24 | Re: 80286 protected mode | 6 | | John Levine |
6 Oct 24 | Re: Byte ordering (was: Whether something is RISC or not) | 2 | | Michael S |
4 Oct 24 | Re: Byte ordering (was: Whether something is RISC or not) | 1 | | John Dallman |
2 Oct 24 | Re: Whether something is RISC or not (Re: PDP-8 theology, not Concertina II Progress) | 1 | | Thomas Koenig |
2 Oct 24 | Re: Whether something is RISC or not (Re: PDP-8 theology, not Concertina II Progress) | 5 | | David Schultz |
3 Oct 24 | Re: Whether something is RISC or not (Re: PDP-8 theology, not Concertina II Progress) | 1 | | Lawrence D'Oliveiro |