Fix unchecked return values

Coverity complained about unchecked return values.

Different approaches were used to handle the different cases.

* cmd/zdb/zdb.c: VERIFY0 was used since the existing code had no error
  handling.

* cmd/zed/agents/zfs_retire.c: We silently return since this is a void
  function and other nvlist lookup failures already silently return.

* cmd/zpool_influxdb/zpool_influxdb.c: `(void) fnvlist_lookup_*()` was
  used since there is no error handling for the return from
  `get_vdev_desc()`.

* cmd/ztest.c: VERIFY0 was used since we want failures if something goes
  wrong in ztest.

* module/zfs/dsl_dir.c: We dismiss the return value with `(void)`
  because there is no guarantee that the zap entry will always be there.
  For example, old pools imported readonly would not have it and we do
  not want to fail here because of that.

* module/zfs/zfs_fm.c: `(void) fnvlist_lookup_*()` was used since the
  allocations sleep and thus can never fail.

* module/zfs/zvol.c: We dismiss the return value with `(void)` because
  we do not need it. This matches what is already done in the analogous
  `zfs_replay_write2()`.

Signed-off-by: Richard Yao <richard.yao@alumni.stonybrook.edu>

cmd/zdb/zdb.c

@@ -4418,9 +4418,9 @@ dump_l2arc_log_blocks(int fd, l2arc_dev_hdr_phys_t l2dhdr,
 		default:
 			abd = abd_alloc_for_io(asize, B_TRUE);
 			abd_copy_from_buf_off(abd, &this_lb, 0, asize);
-			zio_decompress_data(L2BLK_GET_COMPRESS(
+			VERIFY0(zio_decompress_data(L2BLK_GET_COMPRESS(
 			    (&lbps[0])->lbp_prop), abd, &this_lb,
-			    asize, sizeof (this_lb), NULL);
+			    asize, sizeof (this_lb), NULL));
 			abd_free(abd);
 			break;
 		}
@@ -7586,7 +7586,7 @@ dump_mos_leaks(spa_t *spa)
 		} else {
 			dmu_object_info_t doi;
 			const char *name;
-			dmu_object_info(mos, object, &doi);
+			VERIFY0(dmu_object_info(mos, object, &doi));
 			if (doi.doi_type & DMU_OT_NEWTYPE) {
 				dmu_object_byteswap_t bswap =
 				    DMU_OT_BYTESWAP(doi.doi_type);

cmd/zed/agents/zfs_retire.c

@@ -326,7 +326,10 @@ zfs_retire_recv(fmd_hdl_t *hdl, fmd_event_t *ep, nvlist_t *nvl,
 
 	fmd_hdl_debug(hdl, "zfs_retire_recv: '%s'", class);
 
-	nvlist_lookup_uint64(nvl, FM_EREPORT_PAYLOAD_ZFS_VDEV_STATE, &state);
+	if (nvlist_lookup_uint64(nvl, FM_EREPORT_PAYLOAD_ZFS_VDEV_STATE,
+	    &state) != 0) {
+		return;
+	}
 
 	/*
 	 * If this is a resource notifying us of device removal then simply

cmd/zpool_influxdb/zpool_influxdb.c

@@ -302,9 +302,9 @@ get_vdev_desc(nvlist_t *nvroot, const char *parent_name)
 	char *vdev_type = (char *)"unknown";
 	uint64_t vdev_id = UINT64_MAX;
 	char *vdev_path = NULL;
-	nvlist_lookup_string(nvroot, ZPOOL_CONFIG_TYPE, &vdev_type);
-	nvlist_lookup_uint64(nvroot, ZPOOL_CONFIG_ID, &vdev_id);
-	nvlist_lookup_string(nvroot, ZPOOL_CONFIG_PATH, &vdev_path);
+	vdev_type = fnvlist_lookup_string(nvroot, ZPOOL_CONFIG_TYPE);
+	vdev_id = fnvlist_lookup_uint64(nvroot, ZPOOL_CONFIG_ID);
+	vdev_path = fnvlist_lookup_string(nvroot, ZPOOL_CONFIG_PATH);
 
 	if (parent_name == NULL) {
 		s = escape_string(vdev_type);

cmd/ztest.c

@@ -2214,7 +2214,7 @@ ztest_replay_write(void *arg1, void *arg2, boolean_t byteswap)
 		dmu_write(os, lr->lr_foid, offset, length, data, tx);
 	} else {
 		memcpy(abuf->b_data, data, length);
-		dmu_assign_arcbuf_by_dbuf(db, offset, abuf, tx);
+		VERIFY0(dmu_assign_arcbuf_by_dbuf(db, offset, abuf, tx));
 	}
 
 	(void) ztest_log_write(zd, tx, lr);

module/zfs/dsl_dir.c

@@ -270,7 +270,7 @@ dsl_dir_hold_obj(dsl_pool_t *dp, uint64_t ddobj,
 
 		if (dsl_dir_is_zapified(dd)) {
 			inode_timespec_t t = {0};
-			zap_lookup(dp->dp_meta_objset, ddobj,
+			(void) zap_lookup(dp->dp_meta_objset, ddobj,
 			    DD_FIELD_SNAPSHOTS_CHANGED,
 			    sizeof (uint64_t),
 			    sizeof (inode_timespec_t) / sizeof (uint64_t),

module/zfs/zfs_fm.c

@@ -1389,17 +1389,17 @@ zfs_post_state_change(spa_t *spa, vdev_t *vd, uint64_t laststate)
 	aux = fm_nvlist_create(NULL);
 	if (vd && aux) {
 		if (vd->vdev_physpath) {
-			(void) nvlist_add_string(aux,
+			(void) fnvlist_add_string(aux,
 			    FM_EREPORT_PAYLOAD_ZFS_VDEV_PHYSPATH,
 			    vd->vdev_physpath);
 		}
 		if (vd->vdev_enc_sysfs_path) {
-			(void) nvlist_add_string(aux,
+			(void) fnvlist_add_string(aux,
 			    FM_EREPORT_PAYLOAD_ZFS_VDEV_ENC_SYSFS_PATH,
 			    vd->vdev_enc_sysfs_path);
 		}
 
-		(void) nvlist_add_uint64(aux,
+		(void) fnvlist_add_uint64(aux,
 		    FM_EREPORT_PAYLOAD_ZFS_VDEV_LASTSTATE, laststate);
 	}
 
@@ -1496,12 +1496,12 @@ zfs_ereport_zvol_post(const char *subclass, const char *name,
 		return;
 
 	aux = fm_nvlist_create(NULL);
-	nvlist_add_string(aux, FM_EREPORT_PAYLOAD_ZFS_DEVICE_NAME, dev_name);
-	nvlist_add_string(aux, FM_EREPORT_PAYLOAD_ZFS_RAW_DEVICE_NAME,
+	(void) fnvlist_add_string(aux, FM_EREPORT_PAYLOAD_ZFS_DEVICE_NAME, dev_name);
+	(void) fnvlist_add_string(aux, FM_EREPORT_PAYLOAD_ZFS_RAW_DEVICE_NAME,
 	    raw_name);
 	r = strchr(name, '/');
 	if (r && r[1])
-		nvlist_add_string(aux, FM_EREPORT_PAYLOAD_ZFS_VOLUME, &r[1]);
+		(void) fnvlist_add_string(aux, FM_EREPORT_PAYLOAD_ZFS_VOLUME, &r[1]);
 
 	zfs_post_common(spa, NULL, FM_RSRC_CLASS, subclass, aux);
 	fm_nvlist_destroy(aux, FM_NVA_FREE);

module/zfs/zvol.c

@@ -429,7 +429,7 @@ zvol_replay_truncate(void *arg1, void *arg2, boolean_t byteswap)
 	if (error != 0) {
 		dmu_tx_abort(tx);
 	} else {
-		zil_replaying(zv->zv_zilog, tx);
+		(void) zil_replaying(zv->zv_zilog, tx);
 		dmu_tx_commit(tx);
 		error = dmu_free_long_range(zv->zv_objset, ZVOL_OBJ, offset,
 		    length);
@@ -475,7 +475,7 @@ zvol_replay_write(void *arg1, void *arg2, boolean_t byteswap)
 		dmu_tx_abort(tx);
 	} else {
 		dmu_write(os, ZVOL_OBJ, offset, length, data, tx);
-		zil_replaying(zv->zv_zilog, tx);
+		(void) zil_replaying(zv->zv_zilog, tx);
 		dmu_tx_commit(tx);
 	}