Gabriel Rossetti wrote:
Terry Reedy wrote:
Gabriel Rossetti wrote:
I ran these tests on linux 2.6 (ubuntu 8.04) using python 2.5.2.
Have you tried the much newer 2.6? 2.5.3 will be out soon with some
bug fixes.
Thanks for the reply Terry, I just tried the pyserial example with
python 2.6 and it still has the same problem, here's the output (I
just copy/pasted the example in a running python2.6 interactive
interpreter) :
test1 at 09:19am :
$ ps waux | grep python2.6
1000 6730 0.2 0.4 6176 4120 pts/12 S+ 09:17 0:00
python2.6
if I try the SIGUSR1 method :
$ kill -SIGUSR1 6730
####################################################################
# DEBUG: The object count is : 12536
####################################################################
test2 at 09:25am :
$ ps waux | grep python2.6
1000 6730 0.0 0.9 12360 10168 pts/12 S+ 09:17 0:00
python2.6
and the SIGUSR1 method gives me :
####################################################################
# DEBUG: The object count is : 25089
####################################################################
I also tried freeing the received string (del t) explicitly but the
results are unchanged.
Thank you,
Gabriel
Apparently the memory goes up when I use the SIGUSR1 code, If I don't
call it, it seams to work fine, my original memory leak must come from
somewhere else then..... sorry for the spam.
Gabriel
--
http://mail.python.org/mailman/listinfo/python-list
