[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Gcl-devel] Re: load

From: Camm Maguire
Subject: [Gcl-devel] Re: load
Date: 22 Nov 2006 09:50:55 -0500
User-agent: Gnus/5.09 (Gnus v5.9.0) Emacs/21.2


Robert Boyer <address@hidden> writes:

> > My informal yardstick is that I'd like to keep run-gbc <= child run
> > time.
> I don't have any intelligent insight on that ratio.  Sometimes I think
> of GCC as 'just' an assembler for GCL.  Sometimes it think of it as the
> miracle that deals with the complicated x86 instruction set.

Actually, gcc is cpp + gcc + gas (+ possibly ld), so maybe it really
should take longer than gcl.

> > Note this is still larger than the 2.6.7 loads.  Why? -- the source of
> > each function is installed
> If the whole source file were read in as a mere string and individual
> functions were simply remembered by beginning and ending points in that
> string, and not created as consed up objects until first needed, might
> the reading possibly be faster?  Something like the following seems
> so fast:
> (defun get-file-as-string (file)
>   (with-open-file (stream file)
>     (let* ((len (file-length stream))
>            (string (make-string len)))
>       (si::fread string 0 len stream)
>       string)))

Interesting.  The read now processes input text files where the source
element is of the form


055     4!G6479086 7.2
0886   7-1     0¬ªÌÏÁÄ­ÖÅÒÂÏÓŪ,FORMAT,T
ompiled in GCL ~a.~a.~a] -./"

It maybe faster to put the length in first, and then fread the rest.
I think the real load though is on doing all those searches for
signature conflicts for every loaded function.  Might be a good idea
to profile this at some point.

Take care,

> Bob

Camm Maguire                                            address@hidden
"The earth is but one country, and mankind its citizens."  --  Baha'u'llah

reply via email to

[Prev in Thread] Current Thread [Next in Thread]