Re: Avoid treating the stack as an array [Re: "Back & Forth" is back!]

On Mon, 16 Sep 2024 12:47:10 +0000, dxf wrote:

On 16/09/2024 8:13 pm, mhx wrote:

[..]
FORTH> tnb
\ no locals:  5ns/call.
\    locals:  18.2ns/call.
\    globals: 6ns/call.
\ no locals2: 21.9ns/call. ok
>
This appears not to be a good idea.
The root cause is piling up too many
items on the F-stack (exceeding the
hardware FPU stack limits).

>
FVALUEs may be the way to go for hardware stack.
Is this any better?
>
: tri_mf ( f: x a b c -- mv)
  3 fpick ( x) 3 fpick ( x a) f>=
  3 fpick ( x) 2 fpick ( x b) f< and if
    fdrop               \ x a b
    frot 2 fpick  f-    \ a b x-a
    fswap frot f-       \ x-a b-a
    f/ exit
  then
  3 fpick ( x) 2 fpick ( x b) f>=
  3 fpick ( x) 1 fpick ( x c) f< and if
    frot fdrop          \ x b c
    frot fover fswap f- \ b c c-x
    fswap frot f-       \ c-x c-b
    f/ exit
  then
  fdrop fdrop fdrop fdrop  0e
;

Your solution gives the best speed compared to others. With gforth under
wsl, I find 59ns/call
Here is the code:
\ here is your definition
: tri_mf ( f: x a b c -- mv)
  3 fpick ( x) 3 fpick ( x a) f>=
  3 fpick ( x) 2 fpick ( x b) f< and if
    fdrop               \ x a b
    frot 2 fpick  f-    \ a b x-a
    fswap frot f-       \ x-a b-a
    f/ exit
  then
  3 fpick ( x) 2 fpick ( x b) f>=
  3 fpick ( x) 1 fpick ( x c) f< and if
    frot fdrop          \ x b c
    frot fover fswap f- \ b c c-x
    fswap frot f-       \ c-x c-b
    f/ exit
  then
  fdrop fdrop fdrop fdrop  0e
;
\ and then the code
: neg_big -1e308  -1e    0e    tri_mf ;
: zero    -1e      0e    1e    tri_mf ;
: pos_big  0e      1e    1e308 tri_mf ;
: fuzzify ( f: x)
    fdup neg_big cr f.
    fdup zero    cr f.
         pos_big cr f.
;
: go 0 do -0.1e neg_big fdrop loop ;
utime 100000000 go utime d>f d>f f- 1e-8 f* f. 0.05871598  ok
utime 100000000 go utime d>f d>f f- 1e-8 f* f. 0.05926772  ok
utime 100000000 go utime d>f d>f f- 1e-8 f* f. 0.05896149  ok
utime 100000000 go utime d>f d>f f- 1e-8 f* f. 0.05899284  ok
Ahmed