From: 吉藤英明
Date: Sat, 7 Oct 2017 08:57:49 +0900
> Hi,
>
> 2017-10-07 4:06 GMT+09:00 Wei Wang :
>> diff --git a/net/ipv6/route.c b/net/ipv6/route.c
>> index 941c062389d2..aeb349aea429 100644
>> --- a/net/ipv6/route.c
>> +++ b/net/ipv6/route.c
> :
>> @@ -1625,12 +1643,17 @@ struct rt6_info *ip6_po
On Fri, Oct 6, 2017 at 4:57 PM, 吉藤英明 wrote:
> Hi,
>
> 2017-10-07 4:06 GMT+09:00 Wei Wang :
>> diff --git a/net/ipv6/route.c b/net/ipv6/route.c
>> index 941c062389d2..aeb349aea429 100644
>> --- a/net/ipv6/route.c
>> +++ b/net/ipv6/route.c
> :
>> @@ -1625,12 +1643,17 @@ struct rt6_info *ip6_pol_rout
Hi,
2017-10-07 4:06 GMT+09:00 Wei Wang :
> diff --git a/net/ipv6/route.c b/net/ipv6/route.c
> index 941c062389d2..aeb349aea429 100644
> --- a/net/ipv6/route.c
> +++ b/net/ipv6/route.c
:
> @@ -1625,12 +1643,17 @@ struct rt6_info *ip6_pol_route(struct net *net,
> struct fib6_table *table,
>
From: Wei Wang
With rwlock, it is safe to call dst_hold() in the read thread because
read thread is guaranteed to be separated from write thread.
However, after we replace rwlock with rcu, it is no longer safe to use
dst_hold(). A dst might already have been deleted but is waiting for the
rcu gra