bug#69249: bug in native-compile?

[Robert Boyer <robertstephenboyer@gmail.com>]

[-- Attachment #1: Type: text/plain, Size: 4325 bytes --]

 If so, we'd be interested to see your code to try and improve our compiler.

I hope this helps.  If not, please let me know.



Robert Boyer <robertstephenboyer@gmail.com>
4:19 PM (1 hour ago)
Reply to all
to Andreas, Eli, 69249, Andrea, Stefan, rms
This message is about how important DECLARE and THE are.  If you do not
handle DECLARE and THE right, people will laugh at native-compile.

Here is an SBCL transcript.  I am running on a $100 Lenovo Chromebook.

I got SBCL with the command sudo apt-get install sbcl.

* (declaim (optimize (safety 0) (speed 3) (debug 0)))
NIL
* (defun foo-with-declare (n)
  (let ((i 0)
        (sum 0)
        (max (expt 10 n)))
    (declare (fixnum i sum max))
    (loop (cond ((<= i max)
                 (setq sum (the fixnum (+ sum i))))
                (t (return sum)))
          (incf i))))
FOO-WITH-DECLARE
* (time (foo-with-declare 8))
Evaluation took:
  0.125 seconds of real time
  0.123203 seconds of total run time (0.123193 user, 0.000010 system)
  98.40% CPU
  135,905,516 processor cycles
  0 bytes consed
5000000050000000
* (defun foo-without-declare (n)
  (let ((i 0)
        (sum 0)
        (max (expt 10 n)))
    (loop (cond ((<= i max)
                 (setq sum (+ sum i)))
                (t (return sum)))
          (incf i))))
WARNING: redefining COMMON-LISP-USER::FOO-WITHOUT-DECLARE in DEFUN
FOO-WITHOUT-DECLARE
* (time (foo-without-declare 8))
Evaluation took:
  1.080 seconds of real time
  1.072932 seconds of total run time (1.072932 user, 0.000000 system)
  99.35% CPU
  1,181,369,303 processor cycles
  0 bytes consed
5000000050000000

People go wow about 8X.  How does native-compile do?

There is a great song by Alabama 'If you gonna play in Texas, you gotta
have a fiddle in the band'. Check it out!

https://www.youtube.com/watch?v=w6r4E514nJg

Bob


Bob


On Sun, Feb 18, 2024 at 3:28 PM Andreas Schwab <schwab@linux-m68k.org>
wrote:

> On Feb 18 2024, Robert Boyer wrote:
>
> > Furthermore, such declare forms are necessary for life as we know it
> > in the free world, so that fixnum arithmetic can be used where
> > appropriate!
>
> If you are programming in Emacs Lisp, you need to follow the rules of
> Emacs Lisp.
>
> --
> Andreas Schwab, schwab@linux-m68k.org
> GPG Key fingerprint = 7578 EB47 D4E5 4D69 2510  2552 DF73 E780 A9DA AEC1
> "And now for something completely different."
>


--
Anything I seem to state should be taken as a question.  I am at least 77
and feeble.
Attachments area
Preview YouTube video If You're Gonna Play In Texas
If You're Gonna Play In Texas
<https://www.youtube.com/watch?v=w6r4E514nJg&authuser=0>
Reply allReplyForward
Add reaction

On Sun, Feb 18, 2024 at 5:32 PM Stefan Monnier <monnier@iro.umontreal.ca>
wrote:

> > ;; However, (native-compile "compile-bug.el") results in the following
> > error message:
> >
> > ;; Compiling file
> > /mnt/chromeos/GoogleDrive/MyDrive/Linux/working/compile-bug.el at Fri Feb
> > 16 08:25:19 2024
> > ;; compile-bug.el:2:45: Warning: Unknown defun property ‘fixnum’ in foo
> > ;; compile-bug.el:2:45: Warning: Unknown defun property ‘vector’ in foo
>
> First thing: these are not error messages!  They're *warnings*!
>
> `batch-byte-compile` also gives those warnings, because ELisp does not
> know those declarations.  The native compiler does support some type
> annotations, but they do not take this shape.
>
> I presume you took those from some other Lisp dialect, but that won't do
> you much good.  Scheme is not Clojure is not Common Lisp is not ELisp ...
>
> BTW, A more recent Emacs would give you an additional warning:
>
>     Warning: file has no ‘lexical-binding’ directive on its first line
>
> So, I see no bug.  Did the resulting native-compile misbehave in
> any way?  I presume you're worried about its performance.
> If so, do you have concrete code where you measured the performance and
> were disappointed and where you have good reasons to believe that type
> annotations would make a difference?  If so, we'd be interested to see
> your code to try and improve our compiler.
>
>
>         Stefan
>
>

-- 
Anything I seem to state should be taken as a question.  I am at least 77
and feeble.

[-- Attachment #2: Type: text/html, Size: 21441 bytes --]