1 files changed, 62 insertions, 23 deletions

diff --git a/drivers/scsi/storvsc_drv.c b/drivers/scsi/storvsc_drv.c
index a2ec0bc9e9fa..33a4a4dad324 100644
--- a/drivers/scsi/storvsc_drv.c
+++ b/drivers/scsi/storvsc_drv.c
@@ -395,6 +395,12 @@ MODULE_PARM_DESC(storvsc_ringbuffer_size, "Ring buffer size (bytes)");
 
 module_param(storvsc_vcpus_per_sub_channel, int, S_IRUGO);
 MODULE_PARM_DESC(storvsc_vcpus_per_sub_channel, "Ratio of VCPUs to subchannels");
+
+static int ring_avail_percent_lowater = 10;
+module_param(ring_avail_percent_lowater, int, S_IRUGO);
+MODULE_PARM_DESC(ring_avail_percent_lowater,
+		"Select a channel if available ring size > this in percent");
+
 /*
  * Timeout in seconds for all devices managed by this driver.
  */
@@ -1241,7 +1247,7 @@ static struct vmbus_channel *get_og_chn(struct storvsc_device *stor_device,
 {
 	u16 slot = 0;
 	u16 hash_qnum;
-	struct cpumask alloced_mask;
+	const struct cpumask *node_mask;
 	int num_channels, tgt_cpu;
 
 	if (stor_device->num_sc == 0)
@@ -1257,10 +1263,13 @@ static struct vmbus_channel *get_og_chn(struct storvsc_device *stor_device,
 	 * III. Mapping is persistent.
 	 */
 
-	cpumask_and(&alloced_mask, &stor_device->alloced_cpus,
-		    cpumask_of_node(cpu_to_node(q_num)));
+	node_mask = cpumask_of_node(cpu_to_node(q_num));
 
-	num_channels = cpumask_weight(&alloced_mask);
+	num_channels = 0;
+	for_each_cpu(tgt_cpu, &stor_device->alloced_cpus) {
+		if (cpumask_test_cpu(tgt_cpu, node_mask))
+			num_channels++;
+	}
 	if (num_channels == 0)
 		return stor_device->device->channel;
 
@@ -1268,7 +1277,9 @@ static struct vmbus_channel *get_og_chn(struct storvsc_device *stor_device,
 	while (hash_qnum >= num_channels)
 		hash_qnum -= num_channels;
 
-	for_each_cpu(tgt_cpu, &alloced_mask) {
+	for_each_cpu(tgt_cpu, &stor_device->alloced_cpus) {
+		if (!cpumask_test_cpu(tgt_cpu, node_mask))
+			continue;
 		if (slot == hash_qnum)
 			break;
 		slot++;
@@ -1285,9 +1296,9 @@ static int storvsc_do_io(struct hv_device *device,
 {
 	struct storvsc_device *stor_device;
 	struct vstor_packet *vstor_packet;
-	struct vmbus_channel *outgoing_channel;
+	struct vmbus_channel *outgoing_channel, *channel;
 	int ret = 0;
-	struct cpumask alloced_mask;
+	const struct cpumask *node_mask;
 	int tgt_cpu;
 
 	vstor_packet = &request->vstor_packet;
@@ -1301,22 +1312,52 @@ static int storvsc_do_io(struct hv_device *device,
 	/*
 	 * Select an an appropriate channel to send the request out.
 	 */
-
 	if (stor_device->stor_chns[q_num] != NULL) {
 		outgoing_channel = stor_device->stor_chns[q_num];
-		if (outgoing_channel->target_cpu == smp_processor_id()) {
+		if (outgoing_channel->target_cpu == q_num) {
 			/*
 			 * Ideally, we want to pick a different channel if
 			 * available on the same NUMA node.
 			 */
-			cpumask_and(&alloced_mask, &stor_device->alloced_cpus,
-				    cpumask_of_node(cpu_to_node(q_num)));
-			for_each_cpu_wrap(tgt_cpu, &alloced_mask,
-					outgoing_channel->target_cpu + 1) {
-				if (tgt_cpu != outgoing_channel->target_cpu) {
-					outgoing_channel =
-					stor_device->stor_chns[tgt_cpu];
-					break;
+			node_mask = cpumask_of_node(cpu_to_node(q_num));
+			for_each_cpu_wrap(tgt_cpu,
+				 &stor_device->alloced_cpus, q_num + 1) {
+				if (!cpumask_test_cpu(tgt_cpu, node_mask))
+					continue;
+				if (tgt_cpu == q_num)
+					continue;
+				channel = stor_device->stor_chns[tgt_cpu];
+				if (hv_get_avail_to_write_percent(
+							&channel->outbound)
+						> ring_avail_percent_lowater) {
+					outgoing_channel = channel;
+					goto found_channel;
+				}
+			}
+
+			/*
+			 * All the other channels on the same NUMA node are
+			 * busy. Try to use the channel on the current CPU
+			 */
+			if (hv_get_avail_to_write_percent(
+						&outgoing_channel->outbound)
+					> ring_avail_percent_lowater)
+				goto found_channel;
+
+			/*
+			 * If we reach here, all the channels on the current
+			 * NUMA node are busy. Try to find a channel in
+			 * other NUMA nodes
+			 */
+			for_each_cpu(tgt_cpu, &stor_device->alloced_cpus) {
+				if (cpumask_test_cpu(tgt_cpu, node_mask))
+					continue;
+				channel = stor_device->stor_chns[tgt_cpu];
+				if (hv_get_avail_to_write_percent(
+							&channel->outbound)
+						> ring_avail_percent_lowater) {
+					outgoing_channel = channel;
+					goto found_channel;
 				}
 			}
 		}
@@ -1324,7 +1365,7 @@ static int storvsc_do_io(struct hv_device *device,
 		outgoing_channel = get_og_chn(stor_device, q_num);
 	}
 
-
+found_channel:
 	vstor_packet->flags |= REQUEST_COMPLETION_FLAG;
 
 	vstor_packet->vm_srb.length = (sizeof(struct vmscsi_request) -
@@ -1382,9 +1423,6 @@ static int storvsc_device_alloc(struct scsi_device *sdevice)
 
 static int storvsc_device_configure(struct scsi_device *sdevice)
 {
-
-	blk_queue_bounce_limit(sdevice->request_queue, BLK_BOUNCE_ANY);
-
 	blk_queue_rq_timeout(sdevice->request_queue, (storvsc_timeout * HZ));
 
 	/* Ensure there are no gaps in presented sgls */
@@ -1732,8 +1770,9 @@ static int storvsc_probe(struct hv_device *device,
 				(num_cpus - 1) / storvsc_vcpus_per_sub_channel;
 	}
 
-	scsi_driver.can_queue = (max_outstanding_req_per_channel *
-				 (max_sub_channels + 1));
+	scsi_driver.can_queue = max_outstanding_req_per_channel *
+				(max_sub_channels + 1) *
+				(100 - ring_avail_percent_lowater) / 100;
 
 	host = scsi_host_alloc(&scsi_driver,
 			       sizeof(struct hv_host_device));