From 0c928f7a37bd2834659972df30aaeebbc4bc70c9 Mon Sep 17 00:00:00 2001
From: Ameer Hamza <ahamza@ixsystems.com>
Date: Fri, 25 Jul 2025 03:47:46 +0500
Subject: [PATCH] ZED: Fix device type detection and pool iteration logic

During hotplug REMOVED events, devid matching fails for partition-based
spares because devid information is not stored in pool config for
partitioned devices. However, when devid is populated by the hotplug
event, the original code skipped the search logic entirely, skipping
vdev_guid matching and resulting in wrong device type detection that
caused spares to be incorrectly identified as l2arc devices.
Additionally, fix zfs_agent_iter_pool() to use the return value from
zfs_agent_iter_vdev() instead of relying on search parameters, which
was previously ignored. Also add pool_guid optimization to enable
targeted pool searching when pool_guid is available.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Alexander Motin <mav@FreeBSD.org>
Signed-off-by: Ameer Hamza <ahamza@ixsystems.com>
Closes #17545
---
 cmd/zed/agents/zfs_agents.c | 67 ++++++++++++++++++++-----------------
 1 file changed, 36 insertions(+), 31 deletions(-)

diff --git a/cmd/zed/agents/zfs_agents.c b/cmd/zed/agents/zfs_agents.c
index 8718dbde0..c0590edc7 100644
--- a/cmd/zed/agents/zfs_agents.c
+++ b/cmd/zed/agents/zfs_agents.c
@@ -134,11 +134,13 @@ zfs_agent_iter_vdev(zpool_handle_t *zhp, nvlist_t *nvl, void *arg)
 	 * of blkid cache and L2ARC VDEV does not contain pool guid in its
 	 * blkid, so this is a special case for L2ARC VDEV.
 	 */
-	else if (gsp->gs_vdev_guid != 0 && gsp->gs_devid == NULL &&
+	else if (gsp->gs_vdev_guid != 0 &&
 	    nvlist_lookup_uint64(nvl, ZPOOL_CONFIG_GUID, &vdev_guid) == 0 &&
 	    gsp->gs_vdev_guid == vdev_guid) {
-		(void) nvlist_lookup_string(nvl, ZPOOL_CONFIG_DEVID,
-		    &gsp->gs_devid);
+		if (gsp->gs_devid == NULL) {
+			(void) nvlist_lookup_string(nvl, ZPOOL_CONFIG_DEVID,
+			    &gsp->gs_devid);
+		}
 		(void) nvlist_lookup_uint64(nvl, ZPOOL_CONFIG_EXPANSION_TIME,
 		    &gsp->gs_vdev_expandtime);
 		return (B_TRUE);
@@ -156,22 +158,28 @@ zfs_agent_iter_pool(zpool_handle_t *zhp, void *arg)
 	/*
 	 * For each vdev in this pool, look for a match by devid
 	 */
-	if ((config = zpool_get_config(zhp, NULL)) != NULL) {
-		if (nvlist_lookup_nvlist(config, ZPOOL_CONFIG_VDEV_TREE,
-		    &nvl) == 0) {
-			(void) zfs_agent_iter_vdev(zhp, nvl, gsp);
-		}
-	}
-	/*
-	 * if a match was found then grab the pool guid
-	 */
-	if (gsp->gs_vdev_guid && gsp->gs_devid) {
-		(void) nvlist_lookup_uint64(config, ZPOOL_CONFIG_POOL_GUID,
-		    &gsp->gs_pool_guid);
-	}
+	boolean_t found = B_FALSE;
+	uint64_t pool_guid;
 
+	/* Get pool configuration and extract pool GUID */
+	if ((config = zpool_get_config(zhp, NULL)) == NULL ||
+	    nvlist_lookup_uint64(config, ZPOOL_CONFIG_POOL_GUID,
+	    &pool_guid) != 0)
+		goto out;
+
+	/* Skip this pool if we're looking for a specific pool */
+	if (gsp->gs_pool_guid != 0 && pool_guid != gsp->gs_pool_guid)
+		goto out;
+
+	if (nvlist_lookup_nvlist(config, ZPOOL_CONFIG_VDEV_TREE, &nvl) == 0)
+		found = zfs_agent_iter_vdev(zhp, nvl, gsp);
+
+	if (found && gsp->gs_pool_guid == 0)
+		gsp->gs_pool_guid = pool_guid;
+
+out:
 	zpool_close(zhp);
-	return (gsp->gs_devid != NULL && gsp->gs_vdev_guid != 0);
+	return (found);
 }
 
 void
@@ -233,20 +241,17 @@ zfs_agent_post_event(const char *class, const char *subclass, nvlist_t *nvl)
 		 * For multipath, spare and l2arc devices ZFS_EV_VDEV_GUID or
 		 * ZFS_EV_POOL_GUID may be missing so find them.
 		 */
-		if (devid == NULL || pool_guid == 0 || vdev_guid == 0) {
-			if (devid == NULL)
-				search.gs_vdev_guid = vdev_guid;
-			else
-				search.gs_devid = devid;
-			zpool_iter(g_zfs_hdl, zfs_agent_iter_pool, &search);
-			if (devid == NULL)
-				devid = search.gs_devid;
-			if (pool_guid == 0)
-				pool_guid = search.gs_pool_guid;
-			if (vdev_guid == 0)
-				vdev_guid = search.gs_vdev_guid;
-			devtype = search.gs_vdev_type;
-		}
+		search.gs_devid = devid;
+		search.gs_vdev_guid = vdev_guid;
+		search.gs_pool_guid = pool_guid;
+		zpool_iter(g_zfs_hdl, zfs_agent_iter_pool, &search);
+		if (devid == NULL)
+			devid = search.gs_devid;
+		if (pool_guid == 0)
+			pool_guid = search.gs_pool_guid;
+		if (vdev_guid == 0)
+			vdev_guid = search.gs_vdev_guid;
+		devtype = search.gs_vdev_type;
 
 		/*
 		 * We want to avoid reporting "remove" events coming from