classification
Title: test_subprocess is far too slow
Type: Stage:
Components: Library (Lib) Versions: Python 2.4
process
Status: closed Resolution: fixed
Dependencies: Superseder:
Assigned To: astrand Nosy List: anthonybaxter, astrand, georg.brandl, mwh, tim_one (5)
Priority: normal Keywords

Created on 2005-02-17 11:10 by mwh, last changed 2005-03-04 10:17 by mwh.

Messages (10)
msg24321 - (view) Author: Michael Hudson (mwh) Date: 2005-02-17 11:10 
test_subprocess takes multiple minutes.  I'm pretty
sure it's "test_no_leaking".  It should either be sped
up or only tested when some -u argument is passed to
regrtest.
msg24322 - (view) Author: Peter Åstrand (astrand) Date: 2005-02-17 11:50 
Logged In: YES 
user_id=344921

Tell me a bit about your type of OS and hardware. On my
machine (P4 2.66 GHz with Linux), the test takes 28 seconds.
msg24323 - (view) Author: Michael Hudson (mwh) Date: 2005-02-17 12:09 
Logged In: YES 
user_id=6656

Bog standard linux pc -- p3 933, 384 megs of ram.

"$ time ./python ../Lib/test/regrtest.py test_subprocess"
reports 2 minutes 7.  This is a debug build, a release build
might be quicker.  A run of the entire test suite takes a
hair over nine minutes, so 20-odd % of the time seems to be
test_subprocess.

It also takes ages on my old-ish ibook (600 Mhz G3, also 384
megs of ram), but that's at home and I can't time it.
msg24324 - (view) Author: Anthony Baxter (anthonybaxter) Date: 2005-02-17 13:16 
Logged In: YES 
user_id=29957

I'm with mwh. This test should be enabled with an option -
I'd noticed that test_subprocess took a long time, but
hadn't bothered to track it down.

Before releases (and at other times) I always try to make
sure that I run 'make testall', but for day-to-day use, this
is too slow.
msg24325 - (view) Author: Tim Peters (tim_one) Date: 2005-02-17 15:13 
Logged In: YES 
user_id=31435

Peculiar -- test_subprocess takes about 8 seconds on my 
WinXP box release build, about 10 seconds debug build, and 
Windows is notorious for slow process creation.  That's a 
zippy 3.4 GHz P4, but clock rate alone can't account for such 
a large speed difference.
msg24326 - (view) Author: Peter Åstrand (astrand) Date: 2005-02-17 15:17 
Logged In: YES 
user_id=344921

tim_one: test_subprocess only creates 65 processes on
Windows, compared to 1026 for UNIX.
msg24327 - (view) Author: Peter Åstrand (astrand) Date: 2005-03-03 20:53 
Logged In: YES 
user_id=344921

Fixed in HEAD revisions:
regrtest.py: 1.166
1.19

Fixed in 2.4 revisions:
regrtest.py: 1.165.2.1
test_subprocess.py: 1.15.2.2
msg24328 - (view) Author: Georg Brandl (georg.brandl) Date: 2005-03-03 20:54 
Logged In: YES 
user_id=1188172

Shouldn't it be closed then?
msg24329 - (view) Author: Peter Åstrand (astrand) Date: 2005-03-03 20:58 
Logged In: YES 
user_id=344921

Forgot to close.
msg24330 - (view) Author: Michael Hudson (mwh) Date: 2005-03-04 10:17 
Logged In: YES 
user_id=6656

Thank you!
History
Date User Action Args
2005-02-17 11:10:13mwhcreate