Hernan Ponce de Leon
Hernan Ponce de Leon
We do have full support to this now, right? If so, this can be closed
Closing resource issues. They will remain available even if closed.
Does this one make sense still?
We have problems (i.e., exception) to handle the function pointer, but at least we do not return a wrong result anymore. Closing.
This issue is not reproducible anymore.
@xeren why do you say we should treat it as a spinloop? Which problem did you encounter? I actually run this test with herd7 and it reports that the program...
Yes, this is the way to deal with such situations in litmus test (at least to be consistent with herd). The question is, is there an "issue" to be solved?...
Is the following pattern the only deadlocking scenario? ``` rcu_read_lock(); ... synchronize_rcu(); ... rcu_read_unlock(); ``` If so, I think it is fair to assume kernel developer will not write such...
I don't think this issues brings much value. If we ever revise the idea of properly supporting CFGs, we can still access the resource via "closed issues".
I take as an outcome that the inclusion holds (and thus `sloc & sref` in the definition of `mutordatom` is redundant; there are quite a few redundancies in the formal...