netdev - Re: [PATCH net v2 1/3] net: ipv6: add fib6_nh_release

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 for Android: free password hash cracker in your pocket

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

Message-ID: <b75dd859-266e-16d7-f37a-1c349fccacd4@nvidia.com>
Date:   Sun, 28 Nov 2021 22:14:12 +0200
From:   Nikolay Aleksandrov <nikolay@...dia.com>
To:     David Ahern <dsahern@...il.com>,
        Nikolay Aleksandrov <razor@...ckwall.org>,
        netdev@...r.kernel.org
Cc:     idosch@...sch.org, davem@...emloft.net, kuba@...nel.org
Subject: Re: [PATCH net v2 1/3] net: ipv6: add fib6_nh_release_dsts stub

On 28/11/2021 21:21, David Ahern wrote:
> On 11/22/21 8:15 AM, Nikolay Aleksandrov wrote:
>> diff --git a/net/ipv6/route.c b/net/ipv6/route.c
>> index 3ae25b8ffbd6..42d60c76d30a 100644
>> --- a/net/ipv6/route.c
>> +++ b/net/ipv6/route.c
>> @@ -3680,6 +3680,25 @@ void fib6_nh_release(struct fib6_nh *fib6_nh)
>>  	fib_nh_common_release(&fib6_nh->nh_common);
>>  }
>>  
>> +void fib6_nh_release_dsts(struct fib6_nh *fib6_nh)
>> +{
>> +	int cpu;
>> +
>> +	if (!fib6_nh->rt6i_pcpu)
>> +		return;
>> +
>> +	for_each_possible_cpu(cpu) {
>> +		struct rt6_info *pcpu_rt, **ppcpu_rt;
>> +
>> +		ppcpu_rt = per_cpu_ptr(fib6_nh->rt6i_pcpu, cpu);
>> +		pcpu_rt = xchg(ppcpu_rt, NULL);
>> +		if (pcpu_rt) {
>> +			dst_dev_put(&pcpu_rt->dst);
>> +			dst_release(&pcpu_rt->dst);
>> +		}
>> +	}
>> +}
>> +
> 
> this duplicates fib6_nh_release. Can you send a follow on to have it use
> this new function?
> 

It duplicates a part of it but in a safe way because the fib6_nh could still be visible,
while fib6_nh_release does it in a way that assumes it's not. I could re-use
this helper in fib6_nh_release though, since it doesn't matter how the entries are
freed there. I'm guessing that is what you meant?
I'll take care of that and of the few possible optimizations for nexthop in net-next.

Thanks,
 Nik