This series replaces the current read-write lock implementation with the queue read-write locks from Linux. These are fair; under contention both readers and writers will be queued and obtain the lock in FIFO order (due to the fairness of the internal ticket lock).
The implementation is all in C and thus architecture independent. Compared to the Linux implementation some of the memory barrier primitives were changed. The ARM maintainers may want to give this area a careful review. Significant changes in v2: - atomic_cmpxchg() is now arch-dependent as arm/arm64 already provided their own. - Moved rwlocks into their own file. - Removed the special casing of in_irq() in the slow read path. David _______________________________________________ Xen-devel mailing list Xen-devel@lists.xen.org http://lists.xen.org/xen-devel
