One of the unit tests (Test multi-threaded debugging) fail almost -25%-50% of the time when using the new PTVSD debugger. What's interesting is the same tests pass 100% of the time when using the old PTVSD debugger.
Here's the python code
print("step1")
import sys
import threading
import time
print("step2")
def bar():
time.sleep(2)
print("print")
def foo(x):
while True:
bar() # add breakpoint here <-
threading.Thread(target=lambda: foo(2), name="foo").start()
foo(1)
In the unit test I add a breakpoint to the line bar() inside the loop.
I can see (after placing print statements) that two threads are running the while loop.
However, the tests fail as the debugger fails to break at the break point.
Issue filed upstream microsoft/ptvsd#282
One of the unit tests (
Test multi-threaded debugging) fail almost -25%-50% of the time when using the new PTVSD debugger. What's interesting is the same tests pass 100% of the time when using the old PTVSD debugger.Here's the python code
In the unit test I add a breakpoint to the line
bar()inside the loop.I can see (after placing print statements) that two threads are running the while loop.
However, the tests fail as the debugger fails to break at the break point.
Issue filed upstream microsoft/ptvsd#282