bpf: Fix memlock accounting for sock_hash
authorAndrey Ignatov <rdna@fb.com>
Fri, 12 Jun 2020 00:08:57 +0000 (17:08 -0700)
committerGreg Kroah-Hartman <gregkh@linuxfoundation.org>
Wed, 24 Jun 2020 15:49:11 +0000 (17:49 +0200)
commit14b3a7fef8d6b37315de543dc2122939cbd0066f
tree547f568887a82add4525be0efe0ff2f726dbb696
parent032f8eb4b9cec71a6e2ec47a846b293401bbea74
bpf: Fix memlock accounting for sock_hash

[ Upstream commit 60e5ca8a64bad8f3e2e20a1e57846e497361c700 ]

Add missed bpf_map_charge_init() in sock_hash_alloc() and
correspondingly bpf_map_charge_finish() on ENOMEM.

It was found accidentally while working on unrelated selftest that
checks "map->memory.pages > 0" is true for all map types.

Before:
# bpftool m l
...
3692: sockhash  name m_sockhash  flags 0x0
key 4B  value 4B  max_entries 8  memlock 0B

After:
# bpftool m l
...
84: sockmap  name m_sockmap  flags 0x0
key 4B  value 4B  max_entries 8  memlock 4096B

Fixes: 604326b41a6f ("bpf, sockmap: convert to generic sk_msg interface")
Signed-off-by: Andrey Ignatov <rdna@fb.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200612000857.2881453-1-rdna@fb.com
Signed-off-by: Sasha Levin <sashal@kernel.org>
net/core/sock_map.c