Spyder: Spyder 4.1 quits when closing browser tab from vpython
Description
What steps will reproduce the problem?
When using the python code below, the display is correct, but closing the browser tabe with the vpython display will cause Spyder to crash without error message. This only occurs for Windows 10, not macOS. Only Firefox has been tested as a browser. Versions used: python 3.7.7, Spyder 4.1.2, vpython 7.6.1. This bug does NOT occur with Spyder 4.0.1!
from vpython import canvas, color, sphere, vector
def provoke():
scene=canvas(title="view",width=800,height=800)
scene.background=color.white
c = vector(0.8, 0.0, 0.6)
p = vector(0,0,0)
s = sphere(pos=p, radius=8.0, color=c)
provoke()
Versions
- Spyder version: 4.0.1
- Python version: 3.7.5
- Qt version: 5.9.6
- PyQt5 version: 5.9.2
- Operating System: Windows 10
Dependencies
atomicwrites >=1.2.0 : 1.3.0 (OK)
chardet >=2.0.0 : 3.0.4 (OK)
cloudpickle >=0.5.0 : 1.3.0 (OK)
diff_match_patch >=20181111 : 20181111 (OK)
intervaltree : None (OK)
IPython >=4.0 : 7.13.0 (OK)
jedi =0.14.1 : 0.14.1 (OK)
nbconvert >=4.0 : 5.6.1 (OK)
numpydoc >=0.6.0 : 0.9.2 (OK)
pexpect >=4.4.0 : 4.8.0 (OK)
pickleshare >=0.4 : 0.7.5 (OK)
psutil >=0.3 : 5.7.0 (OK)
pygments >=2.0 : 2.6.1 (OK)
pylint >=0.25 : 2.4.4 (OK)
pyls >=0.31.2;<0.32.0 : 0.31.9 (OK)
zmq >=17 : 18.1.1 (OK)
qdarkstyle >=2.7 : 2.8 (OK)
qtawesome >=0.5.7 : 0.7.0 (OK)
qtconsole >=4.6.0 : 4.7.2 (OK)
qtpy >=1.5.0 : 1.9.0 (OK)
rtree >=0.8.3 : 0.9.3 (OK)
sphinx >=0.6.6 : 2.4.4 (OK)
spyder_kernels >=1.8.1;<2.0.0: 1.8.1 (OK)
watchdog : None (OK)
cython >=0.21 : None (NOK)
matplotlib >=2.0.0 : None (NOK)
numpy >=1.7 : 1.18.1 (OK)
pandas >=0.13.1 : None (NOK)
scipy >=0.17.0 : 1.4.1 (OK)
sympy >=0.7.3 : None (NOK)
RussAtkinson
All 16 comments
@dalthviz, please take a look at this one.
ccordoba12
on 13 Apr 2020
@ccordoba12 @RussAtkinson after testing this I can confirm Spyder closes if you close the tab. This is the traceback when running with --debug-info verbose:
Handling signal: 0
Traceback (most recent call last):
File "C:\Users\Daniel\Documents\Spyder\spyder\spyder\widgets\dock.py", line 35, in eventFilter
def eventFilter(self, obj, event):
KeyboardInterrupt
dalthviz
on 15 Apr 2020
Pinging @mwcraig about this one. Any pointers on how we could avoid the crash?
ccordoba12
on 15 Apr 2020
After more testing seems like vpython raises a KeyboardInterrupt when closing the tab (you can see the traceback running the code as a script using the Execute in an external system terminal with the Interact with the Python console after execution option checked):
>>>
Traceback (most recent call last):
File "<stdin>", line 1, in <module>
File "C:\Users\Daniel\Anaconda3\envs\develop\lib\site-packages\vpython\no_notebook.py", line 27, in signal_handler
stop_server()
File "C:\Users\Daniel\Anaconda3\envs\develop\lib\site-packages\vpython\no_notebook.py", line 327, in stop_server
raise KeyboardInterrupt
KeyboardInterrupt
@dalthviz, what happens if we catch the KeyboardInterrupt in dock.py?
ccordoba12
on 15 Apr 2020
I changed the eventFilter code to this:
def eventFilter(self, obj, event):
"""Filter mouse press events.
Events that are captured and not propagated return True. Events that
are not captured and are propagated return False.
"""
try:
event_type = event.type()
if event_type == QEvent.MouseButtonPress:
self.tab_pressed(event)
return False
return False
except Exception:
return True
But Spyder keeps closing. Should the handle be done somewhere else?
dalthviz
on 15 Apr 2020
I really don't know. Let's wait for @mwcraig input before proceeding further.
ccordoba12
on 15 Apr 2020
Ok :+1:
Also, as @RussAtkinson explains, this doesn't happen with Spyder 4.0.1. Maybe the handling of the KeyboardInterrupt error changed between spyder-kernels 1.8.1 and 1.9.0?
dalthviz
on 15 Apr 2020
Ok, that means this is a problem with our kernel. Therefore, I think you need to check the PRs that were merged between 1.8.1 and 1.9.0 and try to detect what code in them could have introduced this bug (I'm not really sure which one could be the culprit, sorry).
ccordoba12
on 15 Apr 2020
Also, for some reason this happens with Windows 10 but not macOS. That
might help focus the search on OS dependent differences.
On Wed, Apr 15, 2020 at 8:36 AM Carlos Cordoba notifications@github.com
wrote:
Ok, that means this is a problem with our kernel. Therefore, I think you
need to check the PRs that were merged between 1.8.1 and 1.9.0 and try to
detect what code in them could have introduced this bug (I'm not really
sure which one could be the culprit, sorry).—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
https://github.com/spyder-ide/spyder/issues/12300#issuecomment-614208539,
or unsubscribe
https://github.com/notifications/unsubscribe-auth/AJPQDYWK55YQMTE3XDRVEKLRMX5EVANCNFSM4MGWZRRA
.
RussAtkinson
on 15 Apr 2020
Notes:
- Seems like after the close of the browser tab in Spyder 4.0.1 the kernel is restarted. However running the script again causes Spyder to crash unless you use a dedicated console:
2020-04-16 13:16:49,254 [INFO] [traitlets] -> KernelRestarter: restarting kernel (1/5), keep random ports
2020-04-16 13:16:49,254 [WARNING] [traitlets] -> kernel restarted
2020-04-16 13:16:49,256 [DEBUG] [traitlets] -> Starting kernel: ['C:\\Users\\Daniel\\Anaconda3\\envs\\spyder4.0.1\\pythonw.exe', '-m', 'spyder_kernels.console', '-f', 'C:\\Users\\Daniel\\AppData\\Roaming\\jupyter\\runtime\\kernel-bd2ae532a18c.json']
2020-04-16 13:16:49,259 [DEBUG] [traitlets] -> Connecting to: tcp://127.0.0.1:60641
2020-04-16 13:16:52,256 [DEBUG] [traitlets] -> KernelRestarter: restart apparently succeeded
- In Spyder 4.1.0 and Spyder 4.1.1 the kernel fails to be restarted but Spyder doesn't close (although the Console where the script was run ends up stuck either using the current console or a dedicated one to run the script):
2020-04-16 13:10:06,020 [INFO] [traitlets] -> KernelRestarter: restarting kernel (1/5), keep random ports
2020-04-16 13:10:06,020 [WARNING] [traitlets] -> kernel restarted
Traceback (most recent call last):
File "C:\Users\Daniel\Anaconda3\envs\spyder4.1.1\lib\site-packages\psutil\_pswindows.py", line 679, in wrapper
return fun(self, *args, **kwargs)
File "C:\Users\Daniel\Anaconda3\envs\spyder4.1.1\lib\site-packages\psutil\_pswindows.py", line 933, in create_time
user, system, created = cext.proc_times(self.pid)
ProcessLookupError: [Errno 3] No such process (originated from GetExitCodeProcess != STILL_ACTIVE)
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "C:\Users\Daniel\Anaconda3\envs\spyder4.1.1\lib\site-packages\psutil\__init__.py", line 373, in _init
self.create_time()
File "C:\Users\Daniel\Anaconda3\envs\spyder4.1.1\lib\site-packages\psutil\__init__.py", line 723, in create_time
self._create_time = self._proc.create_time()
File "C:\Users\Daniel\Anaconda3\envs\spyder4.1.1\lib\site-packages\psutil\_pswindows.py", line 681, in wrapper
raise convert_oserror(err, pid=self.pid, name=self._name)
psutil.NoSuchProcess: psutil.NoSuchProcess process no longer exists (pid=1536)
During handling of the above exception, another exception occurred:
Traceback (most recent call last):
File "C:\Users\Daniel\Anaconda3\envs\spyder4.1.1\lib\site-packages\qtconsole\manager.py", line 27, in poll
super(QtKernelRestarter, self).poll()
File "C:\Users\Daniel\Anaconda3\envs\spyder4.1.1\lib\site-packages\jupyter_client\restarter.py", line 113, in poll
self.kernel_manager.restart_kernel(now=True, newports=newports)
File "C:\Users\Daniel\Anaconda3\envs\spyder4.1.1\lib\site-packages\jupyter_client\manager.py", line 411, in restart_kernel
self.shutdown_kernel(now=now, restart=True)
File "C:\Users\Daniel\Anaconda3\envs\spyder4.1.1\lib\site-packages\jupyter_client\manager.py", line 371, in shutdown_kernel
self._kill_kernel()
File "C:\Users\Daniel\Anaconda3\envs\spyder4.1.1\lib\site-packages\spyder\plugins\ipythonconsole\utils\manager.py", line 78, in _kill_kernel
self.kill_proc_tree(self.kernel.pid)
File "C:\Users\Daniel\Anaconda3\envs\spyder4.1.1\lib\site-packages\spyder\plugins\ipythonconsole\utils\manager.py", line 44, in kill_proc_tree
parent = psutil.Process(pid)
File "C:\Users\Daniel\Anaconda3\envs\spyder4.1.1\lib\site-packages\psutil\__init__.py", line 346, in __init__
self._init(pid)
File "C:\Users\Daniel\Anaconda3\envs\spyder4.1.1\lib\site-packages\psutil\__init__.py", line 386, in _init
raise NoSuchProcess(pid, None, msg)
psutil.NoSuchProcess: psutil.NoSuchProcess no process found with pid 1536
- In Spyder 4.1.2 Spyder closes either using the current console or a dedicated one to run the script:
Handling signal: 0
Traceback (most recent call last):
File "C:\Users\Daniel\Anaconda3\envs\develop\lib\site-packages\spyder\widgets\dock.py", line 35, in eventFilter
def eventFilter(self, obj, event):
KeyboardInterrupt
After a git bisect 1335db55fe70249529e54f72ff768a99743a2d87 show as the first bad commit (taking as bad if the console gets stuck, Spyder doesn't start, or Spyder closes with the close of the browser tab after running the vpython script with a dedicated console). I think it is related with the addition of conda envs activation. Maybe this issue is being caused by the created class SpyderKernelManager that overrides QtKernelManager (?)
dalthviz
on 16 Apr 2020
I think it is related with the addition of conda envs activation
I think it is too. But I don't understand why the kernel is not restarted correctly in this case.
@goanpeca?
ccordoba12
on 16 Apr 2020
No idea. I cannot look a this right now. Maybe next week.
goanpeca
on 16 Apr 2020
Ok, please do.
ccordoba12
on 17 Apr 2020
This will be fixed in the next VPython release (don't know when that's going to be released though).
ccordoba12
on 13 Jun 2020
Related issues
hedeqing
·
3Comments
yogu220172
·
3Comments
impact27
·
3Comments
hgijeon
·
3Comments
gabrielclow
·
3Comments