fabrics: fix mem leak at nvmf_discovery_config_file()

Valgrind revealed a mem leak caused due to not invoking
parser_cleanup for the corresponding parser_init at
nvmf_discovery_config_file():

==16682== 256 bytes in 1 blocks are definitely lost in loss record 1 of 2
==16682==    at 0x4848C31: calloc (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==16682==    by 0x40881A: cb_parser_init (fabrics.c:254)
==16682==    by 0x487B8A2: nvmf_discovery_config_file (fabrics.c:2554)
==16682==    by 0x4099F6: fabrics_discovery (fabrics.c:574)
==16682==    by 0x446358: handle_plugin (plugin.c:190)
==16682==    by 0x407760: main (nvme.c:11029)
==16682==
==16682== 472 bytes in 1 blocks are still reachable in loss record 2 of 2
==16682==    at 0x4841984: malloc (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==16682==    by 0x494C7CA: __fopen_internal (in /lib64/libc.so.6)
==16682==    by 0x408802: cb_parser_init (fabrics.c:248)
==16682==    by 0x487B8A2: nvmf_discovery_config_file (fabrics.c:2554)
==16682==    by 0x4099F6: fabrics_discovery (fabrics.c:574)
==16682==    by 0x446358: handle_plugin (plugin.c:190)
==16682==    by 0x407760: main (nvme.c:11029)

Fix the same.

Signed-off-by: Martin George <[email protected]>

Author: martin-gpy

libnvme/src/nvme/fabrics.c

@@ -2588,6 +2588,8 @@ int nvmf_discovery_config_file(struct nvme_global_ctx *ctx,
 		nvme_free_ctrl(c);
 	} while (!err);
 
+	fctx->parser_cleanup(fctx, fctx->user_data);
+
 	if (err != -EOF)
 		return err;