>The way I see it this fix only reduces the probability of a race condition, but
>still leaves room for other threads to face an inconsistent memory. The
>probability of an error depends on the architecture; our experience is that
>you'll face a much greater risk on PowerPC compared to Intel.

Please explain what race condition you see in more detail.  cma_dev_cnt is only
changed under lock, and by the time it is set, all other initialization has
completed.

- Sean

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to [email protected]
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Reply via email to