Bring over illumos ZFS FMA logic -- phase 1

This first phase brings over the ZFS SLM module, zfs_mod.c, to handle
auto operations in response to disk events. Disk event monitoring is
provided from libudev and generates the expected payload schema for
zfs_mod. This work leverages the recently added devid and phys_path
strings in the vdev label.

Signed-off-by: Brian Behlendorf <[email protected]>
Signed-off-by: Don Brady <[email protected]>
Signed-off-by: Tony Hutter <[email protected]>
Closes #4673

2 files changed, 50 insertions, 19 deletions

diff --git a/module/zfs/vdev.c b/module/zfs/vdev.c
index 75f6e5ce1..dcf56d8df 100644
--- a/module/zfs/vdev.c
+++ b/module/zfs/vdev.c
@@ -3373,19 +3373,6 @@ vdev_set_state(vdev_t *vd, boolean_t isopen, vdev_state_t state, vdev_aux_t aux)
 	    vd->vdev_ops->vdev_op_leaf)
 		vd->vdev_ops->vdev_op_close(vd);
 
-	/*
-	 * If we have brought this vdev back into service, we need
-	 * to notify fmd so that it can gracefully repair any outstanding
-	 * cases due to a missing device.  We do this in all cases, even those
-	 * that probably don't correlate to a repaired fault.  This is sure to
-	 * catch all cases, and we let the zfs-retire agent sort it out.  If
-	 * this is a transient state it's OK, as the retire agent will
-	 * double-check the state of the vdev before repairing it.
-	 */
-	if (state == VDEV_STATE_HEALTHY && vd->vdev_ops->vdev_op_leaf &&
-	    vd->vdev_prevstate != state)
-		zfs_post_state_change(spa, vd);
-
 	if (vd->vdev_removed &&
 	    state == VDEV_STATE_CANT_OPEN &&
 	    (aux == VDEV_AUX_OPEN_FAILED || vd->vdev_checkremove)) {
@@ -3466,6 +3453,20 @@ vdev_set_state(vdev_t *vd, boolean_t isopen, vdev_state_t state, vdev_aux_t aux)
 		vd->vdev_removed = B_FALSE;
 	}
 
+	/*
+	 * Notify ZED of any significant state-change on a leaf vdev.
+	 *
+	 * We ignore transitions from a closed state to healthy unless
+	 * the parent was degraded.
+	 */
+	if (vd->vdev_ops->vdev_op_leaf &&
+	    ((save_state > VDEV_STATE_CLOSED) ||
+	    (vd->vdev_state < VDEV_STATE_HEALTHY) ||
+	    (vd->vdev_parent != NULL &&
+	    vd->vdev_parent->vdev_prevstate == VDEV_STATE_DEGRADED))) {
+		zfs_post_state_change(spa, vd, save_state);
+	}
+
 	if (!isopen && vd->vdev_parent)
 		vdev_propagate_state(vd->vdev_parent);
 }
diff --git a/module/zfs/zfs_fm.c b/module/zfs/zfs_fm.c
index 0871fa95b..0d508c0b8 100644
--- a/module/zfs/zfs_fm.c
+++ b/module/zfs/zfs_fm.c
@@ -848,7 +848,8 @@ zfs_ereport_post_checksum(spa_t *spa, vdev_t *vd,
 }
 
 static void
-zfs_post_common(spa_t *spa, vdev_t *vd, const char *type, const char *name)
+zfs_post_common(spa_t *spa, vdev_t *vd, const char *type, const char *name,
+    nvlist_t *aux)
 {
 #ifdef _KERNEL
 	nvlist_t *resource;
@@ -883,6 +884,13 @@ zfs_post_common(spa_t *spa, vdev_t *vd, const char *type, const char *name)
 		if (vd->vdev_fru != NULL)
 			VERIFY0(nvlist_add_string(resource,
 			    FM_EREPORT_PAYLOAD_ZFS_VDEV_FRU, vd->vdev_fru));
+		/* also copy any optional payload data */
+		if (aux) {
+			nvpair_t *elem = NULL;
+
+			while ((elem = nvlist_next_nvpair(aux, elem)) != NULL)
+				(void) nvlist_add_nvpair(resource, elem);
+		}
 	}
 
 	zfs_zevent_post(resource, NULL, zfs_zevent_post_cb);
@@ -898,7 +906,7 @@ zfs_post_common(spa_t *spa, vdev_t *vd, const char *type, const char *name)
 void
 zfs_post_remove(spa_t *spa, vdev_t *vd)
 {
-	zfs_post_common(spa, vd, FM_RSRC_CLASS, FM_RESOURCE_REMOVED);
+	zfs_post_common(spa, vd, FM_RSRC_CLASS, FM_RESOURCE_REMOVED, NULL);
 }
 
 /*
@@ -909,7 +917,7 @@ zfs_post_remove(spa_t *spa, vdev_t *vd)
 void
 zfs_post_autoreplace(spa_t *spa, vdev_t *vd)
 {
-	zfs_post_common(spa, vd, FM_RSRC_CLASS, FM_RESOURCE_AUTOREPLACE);
+	zfs_post_common(spa, vd, FM_RSRC_CLASS, FM_RESOURCE_AUTOREPLACE, NULL);
 }
 
 /*
@@ -919,9 +927,31 @@ zfs_post_autoreplace(spa_t *spa, vdev_t *vd)
  * open because the device was not found (fault.fs.zfs.device).
  */
 void
-zfs_post_state_change(spa_t *spa, vdev_t *vd)
+zfs_post_state_change(spa_t *spa, vdev_t *vd, uint64_t laststate)
 {
-	zfs_post_common(spa, vd, FM_RSRC_CLASS, FM_RESOURCE_STATECHANGE);
+#ifdef _KERNEL
+	nvlist_t *aux;
+
+	/*
+	 * Add optional supplemental keys to payload
+	 */
+	aux = fm_nvlist_create(NULL);
+	if (vd && aux) {
+		if (vd->vdev_physpath) {
+			(void) nvlist_add_string(aux,
+			    FM_EREPORT_PAYLOAD_ZFS_VDEV_PHYSPATH,
+			    vd->vdev_physpath);
+		}
+		(void) nvlist_add_uint64(aux,
+		    FM_EREPORT_PAYLOAD_ZFS_VDEV_LASTSTATE, laststate);
+	}
+
+	zfs_post_common(spa, vd, FM_RSRC_CLASS, FM_RESOURCE_STATECHANGE,
+	    aux);
+
+	if (aux)
+		fm_nvlist_destroy(aux, FM_NVA_FREE);
+#endif
 }
 
 /*
@@ -933,7 +963,7 @@ zfs_post_state_change(spa_t *spa, vdev_t *vd)
 void
 zfs_post_sysevent(spa_t *spa, vdev_t *vd, const char *name)
 {
-	zfs_post_common(spa, vd, FM_SYSEVENT_CLASS, name);
+	zfs_post_common(spa, vd, FM_SYSEVENT_CLASS, name, NULL);
 }
 
 #if defined(_KERNEL) && defined(HAVE_SPL)