| msg56313 - (view) |
Author: Jiri Krivanek (kakacek) |
Date: 2007-10-10 14:10 |
| See please attached ZIP archive which contains the two simple example source codes. If you execute bb.py, let it running for a while and then press Ctrl+C then the Python will terminate safely and you will get the following output on your console: E:\KkWebrman\Instalace\Tmp>bb.py init parsing time... ...done started parsing time... ...done parsing time... ...done finishing done E:\KkWebrman\Instalace\Tmp> However, if you execute aa.py then it will not terminate by Ctrl+C and you will get the following output on your console: E:\KkWebrman\Instalace\Tmp>aa.py init parsing time... started finishing The aa.py simply imports bb. My OS is: MS Windows 2000. My Python version is: Python 2.4.4 (#71, Oct 18 2006, 08:34:43) [MSC v.1310 32 bit (Intel)] on win32 |
|
|
| msg56315 - (view) |
Author: Tim Golden (tim.golden) *  |
Date: 2007-10-10 15:51 |
| Do you realise that the code at the bottom of bb.py is executed when you import it from aa.py? In other words, when you run aa.py, the whole of your significant code is running within an import statement. I don't know if it's the cause of the problem (although I remember past strictures on the messiness of threads and imports) but I doubt if it's a fantastic design choice. |
|
|
| msg56321 - (view) |
Author: Jiri Krivanek (kakacek) |
Date: 2007-10-10 19:32 |
| The reason for it is pragmatic: The whole of my application is published as .pyc files. Except of the only one which must stay .py - that one which has to be executable (+x) on Linux and thus it starts with the line: # !/usr/bin/python and then it should only contain: import and is .pyc. So I did a simple change into my application. I start it via a very brief script whic only contains: # !/usr/bin/python import Wow! It stopped working at all! After 4 hours of isolating of the problem I developped an elementary code which demonstrates my problem (attached at the bug report). I expect that the import statement does not return the control until the code being executed by it finishes, does'n it? |
|
|
| msg56322 - (view) |
Author: Jiri Krivanek (kakacek) |
Date: 2007-10-10 19:42 |
| One more hint: There is the coincidence of three facts: 1. It uses the thread (if I remove the thread then it works fine). 2. It is "double imported" (if I remove the outer import then it works fine). 3. There is the strptime() function being used (if I remove this function then it works fine). |
|
|
| msg56771 - (view) |
Author: Facundo Batista (facundobatista) * |
Date: 2007-10-26 02:38 |
| The deadlock happens because strptime has an import inside it, and recursive imports are not allowed in different threads. As a general rule and good coding style, don't run your code when the module is imported, but put it in a function like "main" in the second file,import it and call it from the first one. This will solve your problem. Note that this happens to you with strptime, but could happen with a lot, *a lot*, of functions that do this internal import of something else. So, you'll never be sure. You can follow the python-dev thread titled "Deadlock by a second import in a thread" for more info. |
|
|
| msg56781 - (view) |
Author: Jiri Krivanek (kakacek) |
Date: 2007-10-26 06:36 |
| In the mena time, by intuition, I have resolved my troube exactly the way you recommend. Thanks to you, currently I also know what is he core of the problem. So the issue can be closed... |
|
|