diff options
| author | Eric Dumazet <edumazet@google.com> | 2022-06-17 20:47:05 -0700 |
|---|---|---|
| committer | David S. Miller <davem@davemloft.net> | 2022-06-19 10:00:02 +0100 |
| commit | 0daf07e527095e64ee8927ce297ab626643e9f51 (patch) | |
| tree | c74d50c59f37ae3e636d6976734762811e118c27 /include/net | |
| parent | ba44f8182ec299c5d1c8a72fc0fde4ec127b5a6d (diff) | |
raw: convert raw sockets to RCU
Using rwlock in networking code is extremely risky.
writers can starve if enough readers are constantly
grabing the rwlock.
I thought rwlock were at fault and sent this patch:
https://lkml.org/lkml/2022/6/17/272
But Peter and Linus essentially told me rwlock had to be unfair.
We need to get rid of rwlock in networking code.
Without this fix, following script triggers soft lockups:
for i in {1..48}
do
ping -f -n -q 127.0.0.1 &
sleep 0.1
done
Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Diffstat (limited to 'include/net')
| -rw-r--r-- | include/net/raw.h | 11 | ||||
| -rw-r--r-- | include/net/rawv6.h | 1 |
2 files changed, 11 insertions, 1 deletions
diff --git a/include/net/raw.h b/include/net/raw.h index 719d3556fc0a..d81eeeb8f1e6 100644 --- a/include/net/raw.h +++ b/include/net/raw.h @@ -33,9 +33,18 @@ int raw_rcv(struct sock *, struct sk_buff *); struct raw_hashinfo { rwlock_t lock; - struct hlist_head ht[RAW_HTABLE_SIZE]; + struct hlist_nulls_head ht[RAW_HTABLE_SIZE]; }; +static inline void raw_hashinfo_init(struct raw_hashinfo *hashinfo) +{ + int i; + + rwlock_init(&hashinfo->lock); + for (i = 0; i < RAW_HTABLE_SIZE; i++) + INIT_HLIST_NULLS_HEAD(&hashinfo->ht[i], i); +} + #ifdef CONFIG_PROC_FS int raw_proc_init(void); void raw_proc_exit(void); diff --git a/include/net/rawv6.h b/include/net/rawv6.h index c48c1298699a..bc70909625f6 100644 --- a/include/net/rawv6.h +++ b/include/net/rawv6.h @@ -3,6 +3,7 @@ #define _NET_RAWV6_H #include <net/protocol.h> +#include <net/raw.h> extern struct raw_hashinfo raw_v6_hashinfo; bool raw_v6_match(struct net *net, struct sock *sk, unsigned short num, |