josephoswaldgg@hotmail.com <josephoswald@gmail.com> wrote:
+---------------
| > When running under cmucl, I get the following messages (repeatedly
| > with the memory size constantly increasing).
| > What am I doing wrong here? ...
| 
| This is incredibly ugly and un-Lispy code. I'll try to point out a few
| major issues.  [...[trimmed]...
+---------------

One additional issue is that the CMUCL interpreter tends to generate
quite a lot of garbage. You'll do a *lot* better if you compile your
code before you try to worry about the garbage-generation rate, e.g.:

    cmu> (defun delay (n) (dotimes (i n)))

    DELAY
    cmu> (time (delay 1000000))
    ; Compiling LAMBDA NIL: 
    ; Compiling Top-Level Form: 
    ; [GC threshold exceeded with 12,013,880 bytes in use.  Commencing GC.]
    ; [GC completed with 1,079,768 bytes retained and 10,934,112 bytes freed.]
    ; [GC will next occur when at least 13,079,768 bytes are in use.]
    ; [GC threshold exceeded with 13,090,712 bytes in use.  Commencing GC.]
    ; [GC completed with 1,079,776 bytes retained and 12,010,936 bytes freed.]
    ; [GC will next occur when at least 13,079,776 bytes are in use.]
    ; [GC threshold exceeded with 13,089,240 bytes in use.  Commencing GC.]
    ; [GC completed with 1,089,832 bytes retained and 11,999,408 bytes freed.]
    ; [GC will next occur when at least 13,089,832 bytes are in use.]
    ; [GC threshold exceeded with 13,099,296 bytes in use.  Commencing GC.]
    ; [GC completed with 1,089,832 bytes retained and 12,009,464 bytes freed.]
    ; [GC will next occur when at least 13,089,832 bytes are in use.]

    ; Evaluation took:
    ;   4.03f0 seconds of real time
    ;   4.0f0 seconds of user run time
    ;   0.03f0 seconds of system run time
    ;   7,266,606,333 CPU cycles
    ;   [Run times include 0.1f0 seconds GC run time]
    ;   133 page faults and
    ;   48,010,536 bytes consed.
    ; 
    NIL
    cmu> (compile 'delay)
    ; Compiling LAMBDA (N): 
    ; Compiling Top-Level Form: 

    DELAY
    NIL
    NIL
    cmu> (time (delay 1000000))
    ; Compiling LAMBDA NIL: 
    ; Compiling Top-Level Form: 

    ; Evaluation took:
    ;   0.01f0 seconds of real time
    ;   0.01f0 seconds of user run time
    ;   0.0f0 seconds of system run time
    ;   26,082,042 CPU cycles
    ;   0 page faults and
    ;   0 bytes consed.
    ; 
    NIL
    cmu> 

Compare the number of "bytes consed" (total size of all objects
allocated by the GC during the execution of "time") in the two
of these: ~48 MB versus *zero*!

Also consider the runtime: ~4 seconds vs. 0.01.


-Rob

-----
Rob Warnock			<rpw3@rpw3.org>
627 26th Avenue			<URL:http://rpw3.org/>
San Mateo, CA 94403		(650)572-2607