You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Not sure if the current test is runnable but with BUF_LEN > ELEMENTS the ring buffer will never loop and thus the writer will never write the same cell twice. This means you don't test the contention resolution algorithm (i.e. readers never spin).
Furthermore, I'm pretty sure loom will just report the data race and then panic. So I'm not entirely sure seqlocks can be proven sound with loom alone.
Not sure if the current test is runnable but with
BUF_LEN > ELEMENTS
the ring buffer will never loop and thus the writer will never write the same cell twice. This means you don't test the contention resolution algorithm (i.e. readers never spin).Furthermore, I'm pretty sure loom will just report the data race and then panic. So I'm not entirely sure seqlocks can be proven sound with loom alone.
Referenced code:
The text was updated successfully, but these errors were encountered: