[tahoe-lafs-trac-stream] [tahoe-lafs] #1274: eliminate pywin32 dependency

[tahoe-lafs]

#1274: eliminate pywin32 dependency
------------------------------+---------------------------------------------
     Reporter:  davidsarah    |       Owner:  warner                                       
         Type:  defect        |      Status:  new                                          
     Priority:  major         |   Milestone:  1.8.2                                        
    Component:  code-storage  |     Version:  1.8.0                                        
   Resolution:                |    Keywords:  pywin32 windows win64 docs-needed news-needed
Launchpad Bug:                |  
------------------------------+---------------------------------------------

Comment (by davidsarah):

 Replying to [comment:20 warner]:
 > Replying to [comment:14 davidsarah]:
 > >
 > > * change {{{test_runner.py}}} and {{{iputil.py}}} to use the
 > >   {{{subprocess}}} module.
 >
 > The usual problem with using {{{subprocess}}} is {{{SIGCHLD}}}:
 Twisted's
 > reactor and {{{subprocess.py}}} fight over who gets to register a
 handler.
 > The reactor isn't strictly running during a {{{trial}}} unit test
 context. It
 > sort of is, but {{{reactor.run()}}} isn't called, and I've had problems
 using
 > {{{reactor.spawnProcess}}} or {{{utils.getProcessOutputAndValue}}} from
 tests
 > (the "potential zombie child" warning), which I've addressed by
 explicitly
 > invoking the twisted call that installs its SIGCHLD handler.
 >
 > But it almost certainly is running when {{{iputil.py}}} gets used. So if
 > {{{subprocess.py}}} takes over the SIGCHLD handler, we need to look
 carefully
 > and make sure that Twisted gets it back again afterwards. I don't know
 if we
 > use {{{getProcessOutputAndValue}}} anywhere outside of {{{iputil.py}}},
 so a
 > problem there may not bite us until we add a call like that later (like
 maybe
 > a call to {{{du -s}}} to measure storage usage), and I'd like to avoid
 that
 > uncertainty.

 With this patch, we no longer use anything that calls
 {{{reactor.spawnProcess}}} (either in test or non-test code). Does that
 resolve the problem?

 {{{iputil.py}}} is the only place where we previously used
 {{{getProcessOutput}}} and now uses {{{subprocess.Popen}}}. Note that it
 is done in a {{{deferToThread}}} (I'm not sure whether that helps, since
 the SIGCHLD handler is process-global).

-- 
Ticket URL: <http://tahoe-lafs.org/trac/tahoe-lafs/ticket/1274#comment:21>
tahoe-lafs <http://tahoe-lafs.org>
secure decentralized storage