Fix segfault on m1 #828
Fix segfault on m1 #828
Conversation
| Some(c) => c, | ||
| None => return, | ||
| }; | ||
| CFRetain(client.inner() as _); |
There was a problem hiding this comment.
I think the original issue was because of self.client.replace(client);. Did you test without this call to CFRetain? If not, can you quickly check it please?
There was a problem hiding this comment.
yes, at first I think it was because of self.client.replace(client) too, and I did the test, it's not.
and from this crash report
it seems it is the dispatch_worker trying to release the client
There was a problem hiding this comment.
How weird... Oh well, that answers my question at least. Can you add a comment above the call to CFRetain saying something like this:
// Without this call,
clientis freed during the execution of the program. It must be kept!
|
Thanks for this! Just one thing to try out on your side and then I think we're all good. :) |
Mark3K
force-pushed
the
fix-segfault-on-m1
branch
from
195ce09 to
dbd169f
Compare
|
Perfect, thanks! Once CI pass, I'll merge and make a new release. |
Fix #816
I don't know why the client from this API(
IOHIDEventSystemClientCreate) is auto released(that's why I create the client every time refresh the components)so I manually retain the client, it seems fix the segmentation fault