From 6a2659e2e940b405895c4e19a683aa7fa846a785 Mon Sep 17 00:00:00 2001
From: Willy Tarreau <w@1wt.eu>
Date: Mon, 2 May 2022 10:46:14 +0200
Subject: [PATCH] tcp: drop the hash_32() part from the index calculation

commit e8161345ddbb66e449abde10d2fdce93f867eba9 upstream.

In commit 190cc82489f4 ("tcp: change source port randomizarion at
connect() time"), the table_perturb[] array was introduced and an
index was taken from the port_offset via hash_32(). But it turns
out that hash_32() performs a multiplication while the input here
comes from the output of SipHash in secure_seq, that is well
distributed enough to avoid the need for yet another hash.

Suggested-by: Amit Klein <aksecurity@gmail.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Willy Tarreau <w@1wt.eu>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
---
 net/ipv4/inet_hashtables.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/net/ipv4/inet_hashtables.c b/net/ipv4/inet_hashtables.c
index 25766b9bdeff2..19369fc9bcda8 100644
--- a/net/ipv4/inet_hashtables.c
+++ b/net/ipv4/inet_hashtables.c
@@ -640,7 +640,7 @@ int __inet_hash_connect(struct inet_timewait_death_row *death_row,
 
 	net_get_random_once(table_perturb,
 			    INET_TABLE_PERTURB_SIZE * sizeof(*table_perturb));
-	index = hash_32(port_offset, INET_TABLE_PERTURB_SHIFT);
+	index = port_offset & (INET_TABLE_PERTURB_SIZE - 1);
 
 	offset = READ_ONCE(table_perturb[index]) + (port_offset >> 32);
 	offset %= remaining;