[PATCH v2 2/7] nvme-multipath: numa support for marginal paths

[Jesse Taube]

FPIN LI (link integrity) messages are received when the attached
fabric detects hardware errors. In response to these messages I/O
should be directed away from the affected ports, and only used
if no other non-marginal paths are available.
To handle this a new controller flag 'NVME_CTRL_MARGINAL' is added
which will cause the multipath scheduler to skip these paths when
checking for 'optimized' paths.

Signed-off-by: Jesse Taube <jtaubepe@xxxxxxxxxx>
---
This is a distinct change from the previous commit
which treated marginal paths as non-optimized but still usable.
This changes the priority of marginal paths to be lower than
non-optimized paths.
V10 -> V1:
 - New commit
V1 -> V2:
 - Rewrite
---
 drivers/nvme/host/multipath.c | 33 ++++++++++++++++++++++++++++++++-
 1 file changed, 32 insertions(+), 1 deletion(-)

diff --git a/drivers/nvme/host/multipath.c b/drivers/nvme/host/multipath.c
index 9b9a657fa330..dbf09cdda815 100644
--- a/drivers/nvme/host/multipath.c
+++ b/drivers/nvme/host/multipath.c
@@ -305,16 +305,45 @@ static bool nvme_path_is_disabled(struct nvme_ns *ns)
 	return false;
 }
 
+static bool nvme_path_is_usable(struct nvme_ns *ns)
+{
+	/* Only NVME_ANA_OPTIMIZED and NVME_ANA_NONOPTIMIZED are usable */
+	return !nvme_path_is_disabled(ns) &&
+		(ns->ana_state == NVME_ANA_OPTIMIZED ||
+		 ns->ana_state == NVME_ANA_NONOPTIMIZED);
+}
+
+static bool nvme_all_paths_marginal(struct nvme_ns_head *head)
+{
+	struct nvme_ns *ns;
+
+	list_for_each_entry_srcu(ns, &head->list, siblings,
+				 srcu_read_lock_held(&head->srcu)) {
+		/* skip paths which can not be used */
+		if (!nvme_path_is_usable(ns))
+			continue;
+		if (!nvme_ctrl_is_marginal(ns->ctrl))
+			return false;
+	}
+
+	return true;
+}
+
 static struct nvme_ns *__nvme_find_path(struct nvme_ns_head *head, int node)
 {
 	int found_distance = INT_MAX, fallback_distance = INT_MAX, distance;
 	struct nvme_ns *found = NULL, *fallback = NULL, *ns;
+	bool need_marginal = nvme_all_paths_marginal(head);
 
 	list_for_each_entry_srcu(ns, &head->list, siblings,
 				 srcu_read_lock_held(&head->srcu)) {
 		if (nvme_path_is_disabled(ns))
 			continue;
 
+		/* Skip marginal paths unless we need to use them */
+		if (!need_marginal && nvme_ctrl_is_marginal(ns->ctrl))
+			continue;
+
 		if (ns->ctrl->numa_node != NUMA_NO_NODE &&
 		    READ_ONCE(head->subsys->iopolicy) == NVME_IOPOLICY_NUMA)
 			distance = node_distance(node, ns->ctrl->numa_node);
@@ -339,6 +368,7 @@ static struct nvme_ns *__nvme_find_path(struct nvme_ns_head *head, int node)
 		}
 	}
 
+	/* No optimized path found, use the fallback */
 	if (!found)
 		found = fallback;
 	if (found)
@@ -444,7 +474,8 @@ static struct nvme_ns *nvme_queue_depth_path(struct nvme_ns_head *head)
 static inline bool nvme_path_is_optimized(struct nvme_ns *ns)
 {
 	return nvme_ctrl_state(ns->ctrl) == NVME_CTRL_LIVE &&
-		ns->ana_state == NVME_ANA_OPTIMIZED;
+		ns->ana_state == NVME_ANA_OPTIMIZED &&
+		!nvme_ctrl_is_marginal(ns->ctrl);
 }
 
 static struct nvme_ns *nvme_numa_path(struct nvme_ns_head *head)
-- 
2.55.0