Bug 1466044
| Summary: | Crash in keybinder (or its python bindings) | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| Product: | [Fedora] Fedora | Reporter: | Christian Stadelmann <fedora> | ||||||
| Component: | keybinder3 | Assignee: | Patrick Griffis <tingping> | ||||||
| Status: | CLOSED ERRATA | QA Contact: | Fedora Extras Quality Assurance <extras-qa> | ||||||
| Severity: | high | Docs Contact: | |||||||
| Priority: | unspecified | ||||||||
| Version: | 26 | CC: | tingping | ||||||
| Target Milestone: | --- | ||||||||
| Target Release: | --- | ||||||||
| Hardware: | Unspecified | ||||||||
| OS: | Unspecified | ||||||||
| Whiteboard: | |||||||||
| Fixed In Version: | keybinder3-0.3.2-2.fc26 | Doc Type: | If docs needed, set a value | ||||||
| Doc Text: | Story Points: | --- | |||||||
| Clone Of: | Environment: | ||||||||
| Last Closed: | 2017-07-07 23:05:33 UTC | Type: | Bug | ||||||
| Regression: | --- | Mount Type: | --- | ||||||
| Documentation: | --- | CRM: | |||||||
| Verified Versions: | Category: | --- | |||||||
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |||||||
| Cloudforms Team: | --- | Target Upstream Version: | |||||||
| Embargoed: | |||||||||
| Attachments: |
|
||||||||
Created attachment 1292700 [details]
a simple reproducer
Python backtrace:
Traceback (most recent call first):
File "./keybinder_crash/__main__.py", line 38, in do_activate
Keybinder.bind('XF86AudioPlay', foo)
File "./keybinder_crash/__main__.py", line 49, in main
ret = application.run(sys.argv)
File "./keybinder_crash/__main__.py", line 56, in <module>
main()
Is this on Wayland? If so 0.3.2 will be the fix (it simply returns and does nothing on Wayland) (In reply to Patrick Griffis from comment #2) > Is this on Wayland? Yes, this is wayland. > If so 0.3.2 will be the fix (it simply returns and does nothing on Wayland) Hm, thats not nice. Any chance to get this into the documentation? And maybe a warning? (In reply to Christian Stadelmann from comment #3) > (In reply to Patrick Griffis from comment #2) > > If so 0.3.2 will be the fix (it simply returns and does nothing on Wayland) > > Hm, thats not nice. Any chance to get this into the documentation? And maybe > a warning? What documentation? If your code handles bind() failing you can print your own warnings that it failed. Global bindings are a concept that don't really exist in Wayland. A workaround would be for example configuring bindings in Gnome-Shell to run a custom command. (In reply to Patrick Griffis from comment #4) > (In reply to Christian Stadelmann from comment #3) > > (In reply to Patrick Griffis from comment #2) > > > If so 0.3.2 will be the fix (it simply returns and does nothing on Wayland) > > > > Hm, thats not nice. Any chance to get this into the documentation? And maybe > > a warning? > > What documentation? If your code handles bind() failing you can print your > own warnings that it failed. The one shipped by keybinder3-doc on Keybinder.bind(). Having it simply return and doing nothing is bad. It should print a warning to console so anyone having a look at "why did nothing happen?" after a Keybinder.bind() call has a clue on what really happens. > Global bindings are a concept that don't really exist in Wayland. A > workaround would be for example configuring bindings in Gnome-Shell to run a > custom command. is there an API for that? (In reply to Christian Stadelmann from comment #5) > The one shipped by keybinder3-doc on Keybinder.bind(). Ah, looking upstream they didn't document it in the next version either, but that is a discussion to have with them. https://github.com/kupferlauncher/keybinder > Having it simply > return and doing nothing is bad. It should print a warning to console so > anyone having a look at "why did nothing happen?" after a Keybinder.bind() > call has a clue on what really happens. Well like I said, your application *should* be handling the return value anyway. It not working isn't a surprise your code should handle it. > > Global bindings are a concept that don't really exist in Wayland. A > > workaround would be for example configuring bindings in Gnome-Shell to run a > > custom command. > > is there an API for that? No. keybinder3-0.3.2-2.fc26 has been submitted as an update to Fedora 26. https://bodhi.fedoraproject.org/updates/FEDORA-2017-737d3789d3 (In reply to Patrick Griffis from comment #6) > (In reply to Christian Stadelmann from comment #5) > > The one shipped by keybinder3-doc on Keybinder.bind(). > > Ah, looking upstream they didn't document it in the next version either, but > that is a discussion to have with them. > > https://github.com/kupferlauncher/keybinder Ok, will try to contribute docs or add an issue there. > > Having it simply > > return and doing nothing is bad. It should print a warning to console so > > anyone having a look at "why did nothing happen?" after a Keybinder.bind() > > call has a clue on what really happens. > > Well like I said, your application *should* be handling the return value > anyway. It not working isn't a surprise your code should handle it. You're right. I'll fix that upstream at https://github.com/exaile/exaile/pull/402/commits/e13fe377645361eb4ae2c93dfe2c4e0174e7c7a3. Thanks for getting keybinder updated so fast! (In reply to Christian Stadelmann from comment #8) > You're right. I'll fix that upstream at > https://github.com/exaile/exaile/pull/402/commits/ > e13fe377645361eb4ae2c93dfe2c4e0174e7c7a3. Looks good. keybinder3-0.3.2-2.fc26 has been pushed to the Fedora 26 testing repository. If problems still persist, please make note of it in this bug report. See https://fedoraproject.org/wiki/QA:Updates_Testing for instructions on how to install test updates. You can provide feedback for this update here: https://bodhi.fedoraproject.org/updates/FEDORA-2017-737d3789d3 keybinder3-0.3.2-2.fc26 has been pushed to the Fedora 26 stable repository. If problems still persist, please make note of it in this bug report. |
Created attachment 1292699 [details] A full backtrace from gdb attached to python Description of problem: Every time I call Keybinder.bind(), it crashes with SIGSEGV. Version-Release number of selected component (if applicable): keybinder3-0.3.0-7.fc26.x86_64 How reproducible: always with the python script I'll attach Steps to Reproduce: 1. start python script, which calls Keybinder.bind() Actual results: Crash. Full backtrace attached. Truncated backtrace: Thread 1 "python2" received signal SIGSEGV, Segmentation fault. XkbUseExtension (dpy=0x5555559a38b0, major_rtrn=major_rtrn@entry=0x0, minor_rtrn=minor_rtrn@entry=0x0) at XKBUse.c:652 652 dpy->free_funcs->xkb = _XkbFreeInfo; #0 0x00007fffe8dbd154 in XkbUseExtension (dpy=0x5555559a38b0, major_rtrn=major_rtrn@entry=0x0, minor_rtrn=minor_rtrn@entry=0x0) at XKBUse.c:652 #1 0x00007fffe8db7a79 in XkbGetUpdatedMap (dpy=0x5555559a38b0, which=7, xkb=0x555555b95d20) at XKBGetMap.c:574 #2 0x00007fffe8db7afd in XkbGetMap (dpy=0x5555559a38b0, which=which@entry=7, deviceSpec=deviceSpec@entry=256) at XKBGetMap.c:603 #3 0x00007fffd5cb468f in grab_ungrab (rootwin=rootwin@entry=0x5555559a6000 [GdkWaylandWindow], keyval=269025044, modifiers=0, grab=grab@entry=1) at bind.c:203 #4 0x00007fffd5cb484f in do_grab_key (binding=binding@entry=0x555555b64530) at bind.c:326 #5 0x00007fffd5cb4a20 in keybinder_bind_full (keystring=0x555555b8fc00 "XF86AudioPlay", handler=0x7ffff7ee6050, user_data=0x555555d0c2d0, notify=0x7fffefcbc660 <_pygi_invoke_closure_free>) at bind.c:586 #6 0x00007fffef0f5bde in ffi_call_unix64 () at ../src/x86/unix64.S:76 #7 0x00007fffef0f554f in ffi_call (cif=cif@entry=0x555555d6c9a8, fn=<optimized out>, rvalue=<optimized out>, rvalue@entry=0x7fffffffc778, avalue=<optimized out>) at ../src/x86/ffi64.c:525 #8 0x00007fffefcbece4 in pygi_invoke_c_callable (function_cache=0x555555d6c900, state=<optimized out>, py_args=<optimized out>, py_kwargs=<optimized out>) at pygi-invoke.c:679 #9 0x00007fffefcc0838 in pygi_function_cache_invoke (function_cache=<optimized out>, py_args=<optimized out>, py_kwargs=<optimized out>) at pygi-cache.c:861 #10 0x00007ffff7a2bea3 in PyObject_Call () at /lib64/libpython2.7.so.1.0 Expected results: No crash Additional info: Upstream has released 0.3.2 on February 28. You might want to update to 0.3.2 which might fix this bug.