-
Notifications
You must be signed in to change notification settings - Fork 37
PyLib.stopPython() generates a memory core with java 8 #77
Comments
@pabloa I'm curious if you have tried the same logic in plain Java on your machine using the libraries you built from that snapshot of JPY. Are you able to try a similar JUnit case in plain java? I resolved a similar issue when using Java 8 (see #71) and I'll see if I can reproduce this in Scala on my machine. |
I just tried this on my local machine with the latest build of master and it seems to work. It's a slightly different environment (macOS, Python 2.7.13) but it seems to work at its most basic:
Looking at the source for Python 2.7.6 and 2.7.13 there's very minimal difference between their Can you try a similar Scala repl test like I performed above? Don't add any extra modules or libs to the Python or Java (class)paths and make sure you set |
I also just tested with Scala 2.12.1 and Python 2.7.6 on Ubuntu 14.04.5. My JPY debug details on start:
|
Hi, I have the same error on a Windows 7.1 Java 1.8.25 Python 2.7.13 (Anaconda 4.3.0) 64-bit: the PyLib.stopPython() generates a JVM crash with a EXCEPTION_ACCESS_VIOLATION |
Hello @voutilad, I used scala 2.11.8 |
@voutilad Yes I was using the master branch. I'll give a try tomorrow with a more recent JDK. |
@voutilad I've got the same error using the latest JDK (Java(TM) SE Runtime Environment (8.0_121-b13) (build 1.8.0_121-b13)) |
@Treydone and @pabloa I'm still struggling to reproduce the same error using the code in the master branch. When we fixed some issues with the
Replace You should get 1 test run without errors. If it fails, then either you aren't running master (commit has If this test works, but your code does not, it could be few issues:
If it's the GIL lock issue, it could be related to what you're doing with your Python code. In that case it might be worth taking this into a private discussion for a bit to see if we can isolate what could be the cause. Sorry for the long winded message! I really want to help get to the bottom of this. If your project can be shared publicly, please let me know and I can take a look at how you've implemented JPY. |
It seems, the issue remains merely on Windows. AppVeyor builds still crash with unit-level tests that call |
I executed this test with java 1.8.111, Scala 2.11 and jpy 0.9-SNAPSHOT (SHA1 4a8a43c) and it hangs the JVM when the test tries to stop the python engine.
The output is:
This is the content of hs_err_pid14716.log file:
The text was updated successfully, but these errors were encountered: