nbft: fix tcp/dhcp address fallback retry

libnvme introduced a specific error code (NVME_CONNECT_ADDRNOTAVAIL) when the address is not available. Previously, the generic error code (NVME_CONNENCT_WRITE) was return in this case. Signed-off-by: Stuart Hayes <[email protected]> [dwagner: added commit message] Signed-off-by: Daniel Wagner <[email protected]>
linux-nvme · Jan 22, 2024 · 64de246 · 64de246
1 parent b5a741b
commit 64de246
Showing 1 changed file with 1 addition and 1 deletion.
diff --git a/nbft.c b/nbft.c
@@ -164,7 +164,7 @@ int discover_from_nbft(nvme_root_t r, char *hostnqn_arg, char *hostid_arg,
 				 * obtains a different local IP address than the
 				 * firmware had. Retry without host_traddr.
 				 */
-				if (ret == -1 && errno == ENVME_CONNECT_WRITE &&
+				if (ret == -1 && errno == ENVME_CONNECT_ADDRNOTAVAIL &&
 				    !strcmp((*ss)->transport, "tcp") &&
 				    strlen(hfi->tcp_info.dhcp_server_ipaddr) > 0) {
 					nvme_free_ctrl(c);