Blob Blame History Raw
From a60e14ce4c1aeaa98458b6421c670ca58a4c099a Mon Sep 17 00:00:00 2001
From: Jakub Sitnicki <jakub@cloudflare.com>
Date: Thu, 6 Feb 2020 12:16:51 +0100
Subject: [PATCH 9/9] bpf, sockhash: Synchronize_rcu before free'ing map
Git-commit: 0b2dc83906cf1e694e48003eae5df8fa63f76fd9
Patch-mainline: v5.6-rc1
References: git-fixes

We need to have a synchronize_rcu before free'ing the sockhash because any
outstanding psock references will have a pointer to the map and when they
use it, this could trigger a use after free.

This is a sister fix for sockhash, following commit 2bb90e5cc90e ("bpf:
sockmap, synchronize_rcu before free'ing map") which addressed sockmap,
which comes from a manual audit.

Fixes: 604326b41a6fb ("bpf, sockmap: convert to generic sk_msg interface")
Signed-off-by: Jakub Sitnicki <jakub@cloudflare.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: John Fastabend <john.fastabend@gmail.com>
Link: https://lore.kernel.org/bpf/20200206111652.694507-3-jakub@cloudflare.com
Acked-by: Denis Kirjanov <denis.kirjanov@suse.com>
---
 net/core/sock_map.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/net/core/sock_map.c b/net/core/sock_map.c
index 992504478e68..085cef5857bb 100644
--- a/net/core/sock_map.c
+++ b/net/core/sock_map.c
@@ -250,6 +250,7 @@ static void sock_map_free(struct bpf_map *map)
 	}
 	raw_spin_unlock_bh(&stab->lock);
 
+	/* wait for psock readers accessing its map link */
 	synchronize_rcu();
 
 	bpf_map_area_free(stab->sks);
@@ -877,6 +878,9 @@ static void sock_hash_free(struct bpf_map *map)
 		raw_spin_unlock_bh(&bucket->lock);
 	}
 
+	/* wait for psock readers accessing its map link */
+	synchronize_rcu();
+
 	bpf_map_area_free(htab->buckets);
 	kfree(htab);
 }
-- 
2.16.4