Improve test_os._kill (failing on slow machines)

[briancurtin]

BPO	8405
Nosy	@vstinner, @bitdancer, @briancurtin
Files	fix_test_os.diff: patch against trunk issue8405.diff: trunk

^{Note: these values reflect the state of the issue at the time it was migrated and might not reflect the current state.}

Show more details

GitHub fields:

assignee = 'https://github.com/briancurtin'
closed_at = <Date 2010-05-28.16:09:28.111>
created_at = <Date 2010-04-15.02:42:01.217>
labels = ['type-bug', 'library', 'OS-windows']
title = 'Improve test_os._kill (failing on slow machines)'
updated_at = <Date 2010-05-28.16:30:04.465>
user = 'https://github.com/briancurtin'

bugs.python.org fields:

activity = <Date 2010-05-28.16:30:04.465>
actor = 'vstinner'
assignee = 'brian.curtin'
closed = True
closed_date = <Date 2010-05-28.16:09:28.111>
closer = 'brian.curtin'
components = ['Library (Lib)', 'Windows']
creation = <Date 2010-04-15.02:42:01.217>
creator = 'brian.curtin'
dependencies = []
files = ['16928', '17473']
hgrepos = []
issue_num = 8405
keywords = ['patch']
message_count = 6.0
messages = ['103179', '103245', '103317', '106611', '106676', '106677']
nosy_count = 3.0
nosy_names = ['vstinner', 'r.david.murray', 'brian.curtin']
pr_nums = []
priority = 'normal'
resolution = 'fixed'
stage = 'resolved'
status = 'closed'
superseder = None
type = 'behavior'
url = 'https://bugs.python.org/issue8405'
versions = ['Python 2.7', 'Python 3.2']

[briancurtin]

test_os._kill is used by test_kill_sigterm and test_kill_int and is failing on a slow Windows buildbot due to timing issues between the process starting and the signal being sent. I've checked in a few small time.sleep hacks in the meantime to see if that would help the bot, but I'd like to get a solid fix in there.

The attached patch adds a loop to the test function to see that even if the subprocess doesn't respond right away, it will have a few chances to respond and run the test. If the expected message isn't piped from the subprocess after 5 retries, the test fails.

[bitdancer]

I like this version of the patch better. Is there any reason not to drop the initial sleep(0.5) and loop 10 times instead? Any place we can cut down the sleeps so the test suite runs faster on fast machines is good.

As someone else said, it *ought* to be possible to do this without ctypes. I'll see if I can find some time to play a few subprocess games on my windows KVM instance.

[vstinner]

With the patch, the test have 1.0 second to succeed. I would prefer 5.0 or 10.0 seconds: it should not waste the 10 seconds because it should succeed if a new milliseconds. But if the machine is "slow" (which is the topic of the issue), you can give some more seconds to the test. So max=100, sleep=0.1

[briancurtin]

I just noticed the other day that a buildbot failed because of this issue. Attached is a patch which removes the unconditional 0.5 sleep, and increases the loop to run 100 times. It should cover the worst case of a super slow buildbot, but is still typically complete in tenths of a second in the common case.

[briancurtin]

Committed to trunk in r81584 and py3k in r81585.

[vstinner]

Committed to trunk in r81584 and py3k in r81585

"sparc solaris10 gcc trunk" buildbot slave doesn't compile anymore. I'm not sure that it's related, so I prefer to not reopen the issue :-)

http://www.python.org/dev/buildbot/trunk/builders/sparc%20solaris10%20gcc%20trunk/builds/891

-----------
...
building '_struct' extension
gcc -fPIC -fno-strict-aliasing -g -O2 -g -O0 -Wall -Wstrict-prototypes -I. -IInclude -I./Include -I/usr/local/include -I/home2/buildbot/slave/trunk.loewis-sun/build/Include -I/home2/buildbot/slave/trunk.loewis-sun/build -c /home2/buildbot/slave/trunk.loewis-sun/build/Modules/_struct.c -o build/temp.solaris-2.10-sun4u-2.7-pydebug/home2/buildbot/slave/trunk.loewis-sun/build/Modules/_struct.o
gcc -shared build/temp.solaris-2.10-sun4u-2.7-pydebug/home2/buildbot/slave/trunk.loewis-sun/build/Modules/_struct.o -L/usr/local/lib -o build/lib.solaris-2.10-sun4u-2.7-pydebug/_struct.so
Assertion failed: min >= 0, file Python/getargs.c, line 1826
*** Error code 134
The following command caused the error:
case $MAKEFLAGS in \
*s*) CC='gcc' LDSHARED='gcc -shared' LDFLAGS='' OPT='-g -O0 -Wall -Wstrict-prototypes' ./python -E ./setup.py -q build;; \
*) CC='gcc' LDSHARED='gcc -shared' LDFLAGS='' OPT='-g -O0 -Wall -Wstrict-prototypes' ./python -E ./setup.py build;; \
esac
make: Fatal error: Command failed for target `sharedmods'
program finished with exit code 1
-----------