On Wed, Feb 26, 2025 at 12:13:42PM -0800, Sean Christopherson wrote:
> I much prefer my (misguided in the original context[*]) approach of marking the
> call_once() COMPLETED if and only if it succeeds.
I have a new appreciation for this approach given our recent
discoveries. I was mistaken in assuming there were no retryable errors
here.
Thanks for the suggestion, I'll merge your propsal with the kvm side and
give it a test.
> diff --git a/include/linux/call_once.h b/include/linux/call_once.h
> index 6261aa0b3fb0..9d47ed50139b 100644
> --- a/include/linux/call_once.h
> +++ b/include/linux/call_once.h
> @@ -26,20 +26,28 @@ do { \
> __once_init((once), #once, &__key); \
> } while (0)
>
> -static inline void call_once(struct once *once, void (*cb)(struct once *))
> +static inline int call_once(struct once *once, int (*cb)(struct once *))
> {
> + int r;
> +
> /* Pairs with atomic_set_release() below. */
> if (atomic_read_acquire(&once->state) == ONCE_COMPLETED)
> - return;
> + return 0;
>
> guard(mutex)(&once->lock);
> WARN_ON(atomic_read(&once->state) == ONCE_RUNNING);
> if (atomic_read(&once->state) != ONCE_NOT_STARTED)
> - return;
> + return -EINVAL;
>
> atomic_set(&once->state, ONCE_RUNNING);
> - cb(once);
> + r = cb(once);
> + if (r) {
> + atomic_set(&once->state, ONCE_NOT_STARTED);
> + return r;
> + }
> +
> atomic_set_release(&once->state, ONCE_COMPLETED);
> + return 0;
> }
>
> #endif /* _LINUX_CALL_ONCE_H */
