netdev - [PATCH net-next 5/5] tls: rx: periodically flush socket backlog

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: <20220705235926.1035407-6-kuba@kernel.org>
Date:   Tue,  5 Jul 2022 16:59:26 -0700
From:   Jakub Kicinski <kuba@...nel.org>
To:     davem@...emloft.net
Cc:     netdev@...r.kernel.org, edumazet@...gle.com, pabeni@...hat.com,
        john.fastabend@...il.com, borisp@...dia.com,
        linux-doc@...r.kernel.org, linux-kselftest@...r.kernel.org,
        maximmi@...dia.com, Jakub Kicinski <kuba@...nel.org>
Subject: [PATCH net-next 5/5] tls: rx: periodically flush socket backlog

We continuously hold the socket lock during large reads and writes.
This may inflate RTT and negatively impact TCP performance.
Flush the backlog periodically. I tried to pick a flush period (128kB)
which gives significant benefit but the max Bps rate is not yet visibly
impacted.

Signed-off-by: Jakub Kicinski <kuba@...nel.org>
---
 net/core/sock.c  |  1 +
 net/tls/tls_sw.c | 23 +++++++++++++++++++++++
 2 files changed, 24 insertions(+)

diff --git a/net/core/sock.c b/net/core/sock.c
index 92a0296ccb18..4cb957d934a2 100644
--- a/net/core/sock.c
+++ b/net/core/sock.c
@@ -2870,6 +2870,7 @@ void __sk_flush_backlog(struct sock *sk)
 	__release_sock(sk);
 	spin_unlock_bh(&sk->sk_lock.slock);
 }
+EXPORT_SYMBOL_GPL(__sk_flush_backlog);
 
 /**
  * sk_wait_data - wait for data to arrive at sk_receive_queue
diff --git a/net/tls/tls_sw.c b/net/tls/tls_sw.c
index 7592b6519953..79043bc3da39 100644
--- a/net/tls/tls_sw.c
+++ b/net/tls/tls_sw.c
@@ -1738,6 +1738,24 @@ static int process_rx_list(struct tls_sw_context_rx *ctx,
 	return copied ? : err;
 }
 
+static void
+tls_read_flush_backlog(struct sock *sk, struct tls_prot_info *prot,
+		       size_t len_left, size_t decrypted, ssize_t done,
+		       size_t *flushed_at)
+{
+	size_t max_rec;
+
+	if (len_left <= decrypted)
+		return;
+
+	max_rec = prot->overhead_size - prot->tail_size + TLS_MAX_PAYLOAD_SIZE;
+	if (done - *flushed_at < SZ_128K && tcp_inq(sk) > max_rec)
+		return;
+
+	*flushed_at = done;
+	sk_flush_backlog(sk);
+}
+
 int tls_sw_recvmsg(struct sock *sk,
 		   struct msghdr *msg,
 		   size_t len,
@@ -1750,6 +1768,7 @@ int tls_sw_recvmsg(struct sock *sk,
 	struct sk_psock *psock;
 	unsigned char control = 0;
 	ssize_t decrypted = 0;
+	size_t flushed_at = 0;
 	struct strp_msg *rxm;
 	struct tls_msg *tlm;
 	struct sk_buff *skb;
@@ -1839,6 +1858,10 @@ int tls_sw_recvmsg(struct sock *sk,
 		if (err <= 0)
 			goto recv_end;
 
+		/* periodically flush backlog, and feed strparser */
+		tls_read_flush_backlog(sk, prot, len, to_decrypt,
+				       decrypted + copied, &flushed_at);
+
 		ctx->recv_pkt = NULL;
 		__strp_unpause(&ctx->strp);
 		__skb_queue_tail(&ctx->rx_list, skb);
-- 
2.36.1