[Zope-dev] Re: OSX: zopectl test results in "Interrupted system call"

Tres Seaver tseaver at palladion.com
Sun Aug 21 15:49:13 EDT 2005


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Stefan H. Holek wrote:
> Since r37701 "zopectl test" forks off a child process [1]. On OSX I  now
> reliably get a traceback tacked onto the end of every test report:
> 
> Traceback (most recent call last):
>   File "/Zope8/lib/python/Zope2/Startup/zopectl.py", line 312, in ?
>     main()
>   File "/Zope8/lib/python/Zope2/Startup/zopectl.py", line 281, in main
>     c.onecmd(" ".join(options.args))
>   File "/usr/local/python2.3/lib/python2.3/cmd.py", line 210, in onecmd
>     return func(arg)
>   File "/Zope8/lib/python/Zope2/Startup/zopectl.py", line 266, in  do_test
>     os.waitpid(pid, 0)
> OSError: [Errno 4] Interrupted system call
> 
> Tested on two OSX machines. Appears to work fine on Linux. Anybody  else
> seeing this? Ideas?
> 
> Thanks,
> Stefan
> 
> 
> [1] <http://svn.zope.org/Zope/branches/Zope-2_8-branch/lib/python/
> Zope2/Startup/zopectl.py?rev=37701&r1=30475&r2=37701>

Just for clarity:  spawing a child process is the only way to keep the
testrunner from killing the zopectl shell.  I have no idea why OSX would
cause EINTR in that case, but I would rather not revert here (I run
tests *all day long* from within the zopectl shell, interleaving with
'fg', 'start', 'stop', etc.)

If anyone has a fix for OSX which doesn't remove the child process, I'm
all for that (but can't do the work myself).


Tres.
- --
===================================================================
Tres Seaver          +1 202-558-7113          tseaver at palladion.com
Palladion Software   "Excellence by Design"    http://palladion.com
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.5 (GNU/Linux)
Comment: Using GnuPG with Thunderbird - http://enigmail.mozdev.org

iD8DBQFDCNq4+gerLs4ltQ4RAj+2AKC8d56tq+xDnhcafxJoZ8MPeIdmsACg2UZt
TIWxF4C9IVgQXTV9q7D3NSM=
=4u12
-----END PGP SIGNATURE-----



More information about the Zope-Dev mailing list