Re: [Sbcl-devel] [Sbcl-commits] master: Simplify %SET-FILL-POINTER transform

[Stas B. <sta...@gm...>]

Also, always transforming makes it quite larger, 89 bytes vs 30 for a full call.
And there is really no way to restrict that, due to the known declaration.
And the error changes from
#(A B C) is not an array with a fill pointer.

to

The value
  #(A B C)
is not of type
  (AND VECTOR (NOT SIMPLE-ARRAY))


On Mon, Jul 17, 2017 at 3:19 PM, Douglas Katzman via Sbcl-commits
<sbc...@li...> wrote:
> The branch "master" has been updated in SBCL:
>        via  5a31f05ed1ab75c3bc8ca9c0559ec76d409982c4 (commit)
>       from  4fc6d2eb98d551727c7a361fde6e0a5ecd08584f (commit)
>
> - Log -----------------------------------------------------------------
> commit 5a31f05ed1ab75c3bc8ca9c0559ec76d409982c4
> Author: Douglas Katzman <do...@go...>
> Date:   Mon Jul 17 08:19:14 2017 -0400
>
>     Simplify %SET-FILL-POINTER transform
> ---
>  src/compiler/array-tran.lisp | 8 ++------
>  1 file changed, 2 insertions(+), 6 deletions(-)
>
> diff --git a/src/compiler/array-tran.lisp b/src/compiler/array-tran.lisp
> index 7a9185e..4052216 100644
> --- a/src/compiler/array-tran.lisp
> +++ b/src/compiler/array-tran.lisp
> @@ -1296,12 +1296,8 @@
>                "The array type is ambiguous; must call ~
>                 ARRAY-HAS-FILL-POINTER-P at runtime.")))))))
>
> -(deftransform %set-fill-pointer ((vector fill-pointer) (vector (eql 0)))
> -  (or (let ((type (lvar-type vector)))
> -        (when (and (= (length (array-type-dimensions-or-give-up type)) 1)
> -                   (eq (conservative-array-type-complexp type) t))
> -          `(setf (%array-fill-pointer vector) 0)))
> -      (give-up-ir1-transform)))
> +(deftransform %set-fill-pointer ((vector fill-pointer) (complex-vector (eql 0)))
> +  `(setf (%array-fill-pointer vector) 0))
>
>  (deftransform check-bound ((array dimension index) * * :node node)
>    ;; This is simply to avoid multiple evaluation of INDEX by the
>
> -----------------------------------------------------------------------
>
>
> hooks/post-receive
> --
> SBCL
>
> ------------------------------------------------------------------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> _______________________________________________
> Sbcl-commits mailing list
> Sbc...@li...
> https://lists.sourceforge.net/lists/listinfo/sbcl-commits



-- 
With best regards, Stas.