On 10/1/20 1:14 AM, Laine Stump wrote:
Lack of this one function (which is called for each active tap
device
every time libvirtd is started) is the one thing preventing a
"WITHOUT_LIBNL" build of libvirt from being useful. With this
alternate implementation, guests using standard tap devices will work
properly.
Signed-off-by: Laine Stump <laine(a)redhat.com>
---
src/util/virnetdev.c | 23 ++++++++++++++++++++++-
1 file changed, 22 insertions(+), 1 deletion(-)
diff --git a/src/util/virnetdev.c b/src/util/virnetdev.c
index 5221bada7b..c43823c747 100644
--- a/src/util/virnetdev.c
+++ b/src/util/virnetdev.c
@@ -915,9 +915,30 @@ virNetDevGetMaster(const char *ifname, char **master)
return 0;
}
+#elif defined(__linux__)
-#else
+/* libnl isn't available, so we can't use netlink.
+ * Fall back to using sysfs
+ */
+int
+virNetDevGetMaster(const char *ifname, char **master)
+{
+ g_autofree char *path = NULL;
+ g_autofree char *canonical = NULL;
+
+ if (virNetDevSysfsFile(&path, ifname, "master") < 0)
+ return -1;
+ if (!(canonical = virFileCanonicalizePath(path)))
+ return -1;
+
+ *master = g_path_get_basename(canonical);
+
+ VIR_DEBUG("IFLA_MASTER for %s is %s", ifname, *master ? *master :
"(none)");
+ return 0;
+}
+
+#else
int
virNetDevGetMaster(const char *ifname G_GNUC_UNUSED,
Whoa, building without LIBNL should fail, n'est-ce pas? I'm not saying
your patch is wrong, it's just that I'm surprised we haven't caught this
earlier.
Michal