bpf: fix struct htab_elem layout

when htab_elem is removed from the bucket list the htab_elem.hash_node.next
field should not be overridden too early otherwise we have a tiny race window
between lookup and delete.
The bug was discovered by manual code analysis and reproducible
only with explicit udelay() in lookup_elem_raw().

Fixes: 6c9059817432 ("bpf: pre-allocate hash map elements")
Reported-by: Jonathan Perry <jonperry@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

diff --git a/kernel/bpf/hashtab.c b/kernel/bpf/hashtab.c
index 3ea87fb..63c86a7 100644 (file)
--- a/kernel/bpf/hashtab.c
+++ b/kernel/bpf/hashtab.c
@@ -45,8 +45,13 @@ enum extra_elem_state {
 struct htab_elem {
        union {
                struct hlist_node hash_node;
-               struct bpf_htab *htab;
-               struct pcpu_freelist_node fnode;
+               struct {
+                       void *padding;
+                       union {
+                               struct bpf_htab *htab;
+                               struct pcpu_freelist_node fnode;
+                       };
+               };
        };
        union {
                struct rcu_head rcu;
@@ -162,7 +167,8 @@ skip_percpu_elems:
                                 offsetof(struct htab_elem, lru_node),
                                 htab->elem_size, htab->map.max_entries);
        else
-               pcpu_freelist_populate(&htab->freelist, htab->elems,
+               pcpu_freelist_populate(&htab->freelist,
+                                      htab->elems + offsetof(struct htab_elem, fnode),
                                       htab->elem_size, htab->map.max_entries);
 
        return 0;
@@ -217,6 +223,11 @@ static struct bpf_map *htab_map_alloc(union bpf_attr *attr)
        int err, i;
        u64 cost;
 
+       BUILD_BUG_ON(offsetof(struct htab_elem, htab) !=
+                    offsetof(struct htab_elem, hash_node.pprev));
+       BUILD_BUG_ON(offsetof(struct htab_elem, fnode.next) !=
+                    offsetof(struct htab_elem, hash_node.pprev));
+
        if (lru && !capable(CAP_SYS_ADMIN))
                /* LRU implementation is much complicated than other
                 * maps.  Hence, limit to CAP_SYS_ADMIN for now.
@@ -582,9 +593,13 @@ static struct htab_elem *alloc_htab_elem(struct bpf_htab *htab, void *key,
        int err = 0;
 
        if (prealloc) {
-               l_new = (struct htab_elem *)pcpu_freelist_pop(&htab->freelist);
-               if (!l_new)
+               struct pcpu_freelist_node *l;
+
+               l = pcpu_freelist_pop(&htab->freelist);
+               if (!l)
                        err = -E2BIG;
+               else
+                       l_new = container_of(l, struct htab_elem, fnode);
        } else {
                if (atomic_inc_return(&htab->count) > htab->map.max_entries) {
                        atomic_dec(&htab->count);