Re: conservative scanning and infinite streams

[Ole Myren Rohne <ole dot rohne at cern dot ch>]

Jim Blandy writes:

 > Are you *sure* the "garbage" is really garbage?  It looks to me like
 > stream-fold, used by stream-for-each, is written (probably
 > accidentally) so that it holds onto the initial stream until it returns.
 > 
 > Have you tried rewriting your code without using stream-for-each?

No, I am not sure it is really garbage... The simplest case gives
inconsistent results on i586-pc-linux-gnu and sparc-sun-solaris2.6.

Silly test program:

(use-modules (ice-9 streams))
(do ((str (make-stream (lambda (i) 
                          (cons i (+ i 1))) 
                       0) 
          (stream-cdr str))) 
          ((stream-null? str)) 
   (and (= 0 (remainder (stream-car str) 10000)) 
   (gc) 
   (display (gc-stats) (current-error-port)) 
   (newline (current-error-port))))

On i586-pc-linux-gnu, cells-allocated is stuck at 67 (bug?) while
gc-time-taken increase at a constant rate per cycle. On
sparc-sun-solaris2.6, cells-allocated starts around 30000, increases
gradually up to 300000, jumps back to 30000 a few times and eventually
it seems to increase without bound (now at 4500000)

	Regards, Ole