On Wed, Aug 4, 2021 at 3:05 PM Thierry Delisle <tdelisle@xxxxxxxxxxxx> wrote: > > I have attached an atomic stack implementation I wrote. I believe it would > be applicable here. It is very similar except the kernel side no longer > needs a retry loop, the looping is moved to the user-space after the pop. > Using it instead of the code you have in enqueue_idle_worker means the > timeout is no longer needed. Thanks, Thierry! Your implementation seems reasonable - I'll test it and will use it in the future versions of the patchset if everything is OK. > > > - ``uint64_t idle_server_tid_ptr``: points to a pointer variable in the > > userspace that points to an idle server, i.e. a server in IDLE > state waiting > > in sys_umcg_wait(); read-only; workers must have this field set; > not used > > in servers. > > > > When a worker's blocking operation in the kernel completes, the kernel > > changes the worker's state from ``BLOCKED`` to ``IDLE``, adds the > worker > > to the list of idle workers, and checks whether > > ``*idle_server_tid_ptr`` is not zero. If not, the kernel tries to > cmpxchg() > > it with zero; if cmpxchg() succeeds, the kernel will then wake the > server. > > See `State transitions`_ below for more details. > > In this case, I believe cmpxchg is not necessary and xchg suffices. Right! I'll need to roll another atomic_xchg helper for this and the atomic stack above...
