netdev - [PATCH v2 bpf-next 1/5] af_unix: Refactor unix_next

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20220113002849.4384-2-kuniyu@amazon.co.jp>
Date:   Thu, 13 Jan 2022 09:28:45 +0900
From:   Kuniyuki Iwashima <kuniyu@...zon.co.jp>
To:     "David S. Miller" <davem@...emloft.net>,
        Jakub Kicinski <kuba@...nel.org>,
        Alexei Starovoitov <ast@...nel.org>,
        Daniel Borkmann <daniel@...earbox.net>,
        Andrii Nakryiko <andrii@...nel.org>
CC:     Martin KaFai Lau <kafai@...com>,
        Benjamin Herrenschmidt <benh@...zon.com>,
        Kuniyuki Iwashima <kuniyu@...zon.co.jp>,
        Kuniyuki Iwashima <kuni1840@...il.com>, <bpf@...r.kernel.org>,
        <netdev@...r.kernel.org>
Subject: [PATCH v2 bpf-next 1/5] af_unix: Refactor unix_next_socket().

Currently, unix_next_socket() is overloaded depending on the 2nd argument.
If it is NULL, unix_next_socket() returns the first socket in the hash.  If
not NULL, it returns the next socket in the same hash list or the first
socket in the next non-empty hash list.

This patch refactors unix_next_socket() into two functions unix_get_first()
and unix_get_next().  unix_get_first() newly acquires a lock and returns
the first socket in the list.  unix_get_next() returns the next socket in a
list or releases a lock and falls back to unix_get_first().

In the following patch, bpf iter holds entire sockets in a list and always
releases the lock before .show().  It always calls unix_get_first() to
acquire a lock in each iteration.  So, this patch makes the change easier
to follow.

Signed-off-by: Kuniyuki Iwashima <kuniyu@...zon.co.jp>
---
 net/unix/af_unix.c | 51 +++++++++++++++++++++++++++-------------------
 1 file changed, 30 insertions(+), 21 deletions(-)

diff --git a/net/unix/af_unix.c b/net/unix/af_unix.c
index c19569819866..e1c4082accdb 100644
--- a/net/unix/af_unix.c
+++ b/net/unix/af_unix.c
@@ -3240,49 +3240,58 @@ static struct sock *unix_from_bucket(struct seq_file *seq, loff_t *pos)
 	return sk;
 }
 
-static struct sock *unix_next_socket(struct seq_file *seq,
-				     struct sock *sk,
-				     loff_t *pos)
+static struct sock *unix_get_first(struct seq_file *seq, loff_t *pos)
 {
 	unsigned long bucket = get_bucket(*pos);
+	struct sock *sk;
 
-	while (sk > (struct sock *)SEQ_START_TOKEN) {
-		sk = sk_next(sk);
-		if (!sk)
-			goto next_bucket;
-		if (sock_net(sk) == seq_file_net(seq))
-			return sk;
-	}
-
-	do {
+	while (bucket < ARRAY_SIZE(unix_socket_table)) {
 		spin_lock(&unix_table_locks[bucket]);
+
 		sk = unix_from_bucket(seq, pos);
 		if (sk)
 			return sk;
 
-next_bucket:
-		spin_unlock(&unix_table_locks[bucket++]);
-		*pos = set_bucket_offset(bucket, 1);
-	} while (bucket < ARRAY_SIZE(unix_socket_table));
+		spin_unlock(&unix_table_locks[bucket]);
+
+		*pos = set_bucket_offset(++bucket, 1);
+	}
 
 	return NULL;
 }
 
+static struct sock *unix_get_next(struct seq_file *seq, struct sock *sk,
+				  loff_t *pos)
+{
+	unsigned long bucket = get_bucket(*pos);
+
+	for (sk = sk_next(sk); sk; sk = sk_next(sk))
+		if (sock_net(sk) == seq_file_net(seq))
+			return sk;
+
+	spin_unlock(&unix_table_locks[bucket]);
+
+	*pos = set_bucket_offset(++bucket, 1);
+
+	return unix_get_first(seq, pos);
+}
+
 static void *unix_seq_start(struct seq_file *seq, loff_t *pos)
 {
 	if (!*pos)
 		return SEQ_START_TOKEN;
 
-	if (get_bucket(*pos) >= ARRAY_SIZE(unix_socket_table))
-		return NULL;
-
-	return unix_next_socket(seq, NULL, pos);
+	return unix_get_first(seq, pos);
 }
 
 static void *unix_seq_next(struct seq_file *seq, void *v, loff_t *pos)
 {
 	++*pos;
-	return unix_next_socket(seq, v, pos);
+
+	if (v == SEQ_START_TOKEN)
+		return unix_get_first(seq, pos);
+
+	return unix_get_next(seq, v, pos);
 }
 
 static void unix_seq_stop(struct seq_file *seq, void *v)
-- 
2.30.2