[Pythonmac-SIG] subprocess .wait() interrupted system call?

Jim Tittsler jwt at onjapan.net
Sun Sep 3 16:57:29 CEST 2006


I'm having a problem with subprocess .wait() on my G4 Mac running the  
universal Python 2.4.3.  If I do a .wait() call immediately after  
creating the process object (on a very heavily loaded machine), an  
"interrupted system call" exception is raised pointing at the .wait()  
line.  I discovered that if I insert a sleep(1) between the Popen and  
wait, my command runs as expected.  (The command is a rather large  
mimetex[1] executable that I compiled as a universal binary.)

If I just run the Python program fragment, it runs fine without the  
sleep... and I've yet to find a minimal case that triggers the  
behavior I'm seeing.

Does anyone happen to have a clue what my larger program might be  
doing that seems to trigger this race-like behavior of Popen and  
wait?  Or am I misusing Popen() and .wait()?


from subprocess import Popen, PIPE
     cmd = '/usr/local/bin/mimetex'
     process = Popen([cmd, '-d', latex, '-s', str(int(fontsize)-1)],  
bufsize=8092, stdout=PIPE, stderr=PIPE)
     #time.sleep(1)
     returnCode = process.wait()


[1] http://www.forkosh.com/mimetex.html



-- 
Jim Tittsler     http://www.OnJapan.net/      GPG: 0x01159DB6
Python Starship  http://Starship.Python.net/crew/jwt/
Mailman IRC      irc://irc.freenode.net/#mailman





More information about the Pythonmac-SIG mailing list