netdev-afxdp: NUMA-aware memory allocation for XSK related memory.

Currently, the AF_XDP socket (XSK) related memory are allocated by main
thread in the main thread's NUMA domain.  With the patch that detects
netdev-linux's NUMA node id, the PMD thread of AF_XDP port will be run on
the AF_XDP netdev's NUMA domain.  If the net device's NUMA domain
is different from the main thread's NUMA domain, we will have two
cross-NUMA memory accesses (netdev <-> memory, memory <-> CPU).

This patch addresses the aforementioned issue by allocating
the memory in the net device's NUMA domain.

Signed-off-by: Yi-Hung Wei <yihung.wei@gmail.com>
Acked-by: William Tu <u9012063@gmail.com>
Signed-off-by: Ilya Maximets <i.maximets@ovn.org>

1 files changed, 32 insertions, 0 deletions

diff --git a/lib/netdev-afxdp.c b/lib/netdev-afxdp.c
index e96328a12..482400d8d 100644
--- a/lib/netdev-afxdp.c
+++ b/lib/netdev-afxdp.c
@@ -26,6 +26,8 @@
 #include <linux/rtnetlink.h>
 #include <linux/if_xdp.h>
 #include <net/if.h>
+#include <numa.h>
+#include <numaif.h>
 #include <poll.h>
 #include <stdlib.h>
 #include <sys/resource.h>
@@ -42,6 +44,7 @@
 #include "openvswitch/list.h"
 #include "openvswitch/thread.h"
 #include "openvswitch/vlog.h"
+#include "ovs-numa.h"
 #include "packets.h"
 #include "socket-util.h"
 #include "util.h"
@@ -667,8 +670,27 @@ netdev_afxdp_reconfigure(struct netdev *netdev)
 {
     struct netdev_linux *dev = netdev_linux_cast(netdev);
     struct rlimit r = {RLIM_INFINITY, RLIM_INFINITY};
+    struct bitmask *old_bm = NULL;
+    int old_policy, numa_id;
     int err = 0;
 
+    /* Allocate all the xsk related memory in the netdev's NUMA domain. */
+    if (numa_available() != -1 && ovs_numa_get_n_numas() > 1) {
+        numa_id = netdev_get_numa_id(netdev);
+        if (numa_id != NETDEV_NUMA_UNSPEC) {
+            old_bm = numa_allocate_nodemask();
+            if (get_mempolicy(&old_policy, old_bm->maskp, old_bm->size + 1,
+                              NULL, 0)) {
+                VLOG_INFO("Failed to get NUMA memory policy: %s.",
+                          ovs_strerror(errno));
+                numa_bitmask_free(old_bm);
+                old_bm = NULL;
+            } else {
+                numa_set_preferred(numa_id);
+            }
+        }
+    }
+
     ovs_mutex_lock(&dev->mutex);
 
     if (netdev->n_rxq == dev->requested_n_rxq
@@ -700,6 +722,16 @@ netdev_afxdp_reconfigure(struct netdev *netdev)
     netdev_change_seq_changed(netdev);
 out:
     ovs_mutex_unlock(&dev->mutex);
+    if (old_bm) {
+        if (set_mempolicy(old_policy, old_bm->maskp, old_bm->size + 1)) {
+            VLOG_WARN("Failed to restore NUMA memory policy: %s.",
+                      ovs_strerror(errno));
+            /* Can't restore correctly.  Try to use localalloc as the most
+             * likely default memory policy. */
+            numa_set_localalloc();
+        }
+        numa_bitmask_free(old_bm);
+    }
     return err;
 }