NVMe: per-cpu io queues

The device's IO queues are associated with CPUs, so we can use a per-cpu
variable to map the a qid to a cpu. This provides a convienient way
to optimally assign queues to multiple cpus when the device supports
fewer queues than the host has cpus. The previous implementation may
have assigned these poorly in these situations. This patch addresses
this by sharing queues among cpus that are "close" together and should
have a lower lock contention penalty.

Signed-off-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Matthew Wilcox <matthew.r.wilcox@intel.com>

include/linux/nvme.h

@@ -74,12 +74,16 @@ enum {
 struct nvme_dev {
 	struct list_head node;
 	struct nvme_queue __rcu **queues;
+	unsigned short __percpu *io_queue;
 	u32 __iomem *dbs;
 	struct pci_dev *pci_dev;
 	struct dma_pool *prp_page_pool;
 	struct dma_pool *prp_small_pool;
 	int instance;
-	int queue_count;
+	unsigned queue_count;
+	unsigned online_queues;
+	unsigned max_qid;
+	int q_depth;
 	u32 db_stride;
 	u32 ctrl_config;
 	struct msix_entry *entry;