Skip to contentSkip to conversation

nvme: warn only once for legacy uuid attribute

PATCH Superseded Keith Busch 2 participants 2 messages Jul 12, 2023 NVMe

v1 of 2 Superseded Latest is v2 →

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 nvme: warn only once for legacy uuid attribute

From: Keith Busch <[email protected]>

Report the legacy fallback behavior for uuid attributes just once instead of logging repeated warnings for the same condition every time the attribute is read. The old behavior is too spamy on the kernel logs.

---
 drivers/nvme/host/core.c  | 7 +++++++
 drivers/nvme/host/sysfs.c | 8 +++-----
 2 files changed, 10 insertions(+), 5 deletions(-)
drivers/nvme/host/core.c +7 -0
index 47d7ba2827ff2..6984205b3514f 100644
@@ -3597,6 +3597,13 @@ static void nvme_alloc_ns(struct nvme_ctrl *ctrl, struct nvme_ns_info *info)
 	nvme_mpath_add_disk(ns, info->anagrpid);
 	nvme_fault_inject_init(&ns->fault_inject, ns->disk->disk_name);
 
+
+	if (uuid_is_null(&info->ids.uuid) &&
+	    memchr_inv(info->ids.nguid, 0, sizeof(info->ids.nguid))) {
+		dev_warn(ctrl->device,
+			"No UUID available providing old NGUID\n");
+	}
+
 	return;
 
  out_cleanup_ns_from_list:
drivers/nvme/host/sysfs.c +3 -5
index 45e91811f905c..0c636d873e058 100644
@@ -88,14 +88,12 @@ static ssize_t uuid_show(struct device *dev, struct device_attribute *attr,
 {
 	struct nvme_ns_ids *ids = &dev_to_ns_head(dev)->ids;
 
-	/* For backward compatibility expose the NGUID to userspace if
+	/*
+	 * For backward compatibility expose the NGUID to userspace if
 	 * we have no UUID set
 	 */
-	if (uuid_is_null(&ids->uuid)) {
-		dev_warn_ratelimited(dev,
-			"No UUID available providing old NGUID\n");
+	if (uuid_is_null(&ids->uuid))
 		return sysfs_emit(buf, "%pU\n", ids->nguid);
-	}
 	return sysfs_emit(buf, "%pU\n", &ids->uuid);
 }
 static DEVICE_ATTR_RO(uuid);
Signature · 1 line
2.34.1
#1 lore raw
Quoted 5 lines · Keith Busch

From: Keith Busch <[email protected]>

Report the legacy fallback behavior for uuid attributes just once instead of logging repeated warnings for the same condition every time the attribute is read. The old behavior is too spamy on the kernel logs.

Warning on allocation just seems wrong - no uuid is perfectly fine. I think what we should do is warn once when reading the attribute and then never again.