Re: [PATCHv5 for-rc1 v5 4/8] RDMA/rxe: Implement dellink in rxe

[Bob Pearson <rpearsonhpe@gmail.com>]

On 6/21/23 01:17, Zhu Yanjun wrote:
> 
> 在 2023/6/21 11:23, Bob Pearson 写道:
>> On 6/20/23 21:13, Zhu Yanjun wrote:
>>> 在 2023/6/21 4:21, Bob Pearson 写道:
>>>> On 4/28/23 04:39, Zhu Yanjun wrote:
>>>>> From: Zhu Yanjun <yanjun.zhu@linux.dev>
>>>>>
>>>>> When running "rdma link del" command, dellink function will be called.
>>>>> If the sock refcnt is greater than the refcnt needed for udp tunnel,
>>>>> the sock refcnt will be decreased by 1.
>>>>>
>>>>> If equal, the last rdma link is removed. The udp tunnel will be
>>>>> destroyed.
>>>>>
>>>>> Tested-by: Rain River <rain.1986.08.12@gmail.com>
>>>>> Signed-off-by: Zhu Yanjun <yanjun.zhu@linux.dev>
>>>>> ---
>>>>>    drivers/infiniband/sw/rxe/rxe.c     | 12 +++++++++++-
>>>>>    drivers/infiniband/sw/rxe/rxe_net.c | 17 +++++++++++++++--
>>>>>    drivers/infiniband/sw/rxe/rxe_net.h |  1 +
>>>>>    3 files changed, 27 insertions(+), 3 deletions(-)
>>>>>
>>>>> diff --git a/drivers/infiniband/sw/rxe/rxe.c b/drivers/infiniband/sw/rxe/rxe.c
>>>>> index 0ce6adb43cfc..ebfabc6d6b76 100644
>>>>> --- a/drivers/infiniband/sw/rxe/rxe.c
>>>>> +++ b/drivers/infiniband/sw/rxe/rxe.c
>>>>> @@ -166,10 +166,12 @@ void rxe_set_mtu(struct rxe_dev *rxe, unsigned int ndev_mtu)
>>>>>    /* called by ifc layer to create new rxe device.
>>>>>     * The caller should allocate memory for rxe by calling ib_alloc_device.
>>>>>     */
>>>>> +static struct rdma_link_ops rxe_link_ops;
>>>>>    int rxe_add(struct rxe_dev *rxe, unsigned int mtu, const char *ibdev_name)
>>>>>    {
>>>>>        rxe_init(rxe);
>>>>>        rxe_set_mtu(rxe, mtu);
>>>>> +    rxe->ib_dev.link_ops = &rxe_link_ops;
>>>>>          return rxe_register_device(rxe, ibdev_name);
>>>>>    }
>>>>> @@ -206,9 +208,17 @@ static int rxe_newlink(const char *ibdev_name, struct net_device *ndev)
>>>>>        return err;
>>>>>    }
>>>>>    -struct rdma_link_ops rxe_link_ops = {
>>>>> +static int rxe_dellink(struct ib_device *dev)
>>>>> +{
>>>>> +    rxe_net_del(dev);
>>>>> +
>>>>> +    return 0;
>>>>> +}
>>>>> +
>>>>> +static struct rdma_link_ops rxe_link_ops = {
>>>>>        .type = "rxe",
>>>>>        .newlink = rxe_newlink,
>>>>> +    .dellink = rxe_dellink,
>>>>>    };
>>>>>      static int __init rxe_module_init(void)
>>>>> diff --git a/drivers/infiniband/sw/rxe/rxe_net.c b/drivers/infiniband/sw/rxe/rxe_net.c
>>>>> index 3ca92e062800..4cc7de7b115b 100644
>>>>> --- a/drivers/infiniband/sw/rxe/rxe_net.c
>>>>> +++ b/drivers/infiniband/sw/rxe/rxe_net.c
>>>>> @@ -530,6 +530,21 @@ int rxe_net_add(const char *ibdev_name, struct net_device *ndev)
>>>>>        return 0;
>>>>>    }
>>>>>    +#define SK_REF_FOR_TUNNEL    2
>>>>> +void rxe_net_del(struct ib_device *dev)
>>>>> +{
>>>>> +    if (refcount_read(&recv_sockets.sk6->sk->sk_refcnt) > SK_REF_FOR_TUNNEL)
>>>>> +        __sock_put(recv_sockets.sk6->sk);
>>>>> +    else
>>>>> +        rxe_release_udp_tunnel(recv_sockets.sk6);
>>>>> +
>>>>> +    if (refcount_read(&recv_sockets.sk4->sk->sk_refcnt) > SK_REF_FOR_TUNNEL)
>>>>> +        __sock_put(recv_sockets.sk4->sk);
>>>>> +    else
>>>>> +        rxe_release_udp_tunnel(recv_sockets.sk4);
>>>>> +}
>>>>> +#undef SK_REF_FOR_TUNNEL
>>>>> +
>>>>>    static void rxe_port_event(struct rxe_dev *rxe,
>>>>>                   enum ib_event_type event)
>>>>>    {
>>>>> @@ -689,8 +704,6 @@ int rxe_register_notifier(void)
>>>>>      void rxe_net_exit(void)
>>>>>    {
>>>>> -    rxe_release_udp_tunnel(recv_sockets.sk6);
>>>>> -    rxe_release_udp_tunnel(recv_sockets.sk4);
>>>>>        unregister_netdevice_notifier(&rxe_net_notifier);
>>>>>    }
>>>> These calls are moved to rxe_net_del which is called by an explicit unlink command.
>>>> But if rxe_net_init fails and returns an error code this will never happen.
>>>> This will result in leaking resources.
>>> Thanks a lot. Bob.
>>>
>>> Sure, if ipv6 tunnel fails to be created, the resource related with ipv4 should be released.
>>>
>>> I will fix it in the latest version.
>>>
>>> Zhu Yanjun
>> I haven't had a chance to test netns yet. I am sure it works but I will test it.
> Yes. Please. It is an interesting feature.
>> The only other thing I noticed are some stylistic differences with the rest of
>> the rxe driver. You use
>>
>> struct rxe_dev *rdev;
>>
>> elsewhere it is
>>
>> struct rxe_dev *rxe;
>>
>> Yours is more like the mlx drivers where they use dev for ib_device and mdev for mlx_device.
>> rxe tries to use ibdev ibqp, ibmr, etc for the ib objects and no prefix for the driver
>> specific ones. It's less typing that way.
> 
> 
> Got you. I think we should use rxe instead of rdev. I will fix it in the latest commits.
> 
> 
>>
>> With a couple of exceptions all the printk's are now in the form
>>
>> rxe_[type]_[obj](obj, "message", ...) or rxe_[type] if there isn't an obj to refer to.
>>
>> where type = info, err, warn, or dbg and obj = rxe, ah, pd, qp, cq, etc. These are basically
>> adapted from the siw driver.
> 
> If I can get you correctly, you mean that we should use rxe_dbg_qp, .... to replace pr_err ....
> 
> I have questions:
> 
> 1). What benefit will this bring?
> 
> 2). If the log is in module __init or  module __exit functions, we should use pr_xxx? Because obj does not exist in these __init and __exit functions.
I think that is the way the driver is now. Go ahead.
> 
> Best Regards,
> 
> Zhu Yanjun
> 
>>
>> Regards,
>>
>> Bob
>>
>>>> Bob
>>>>>    diff --git a/drivers/infiniband/sw/rxe/rxe_net.h b/drivers/infiniband/sw/rxe/rxe_net.h
>>>>> index a222c3eeae12..f48f22f3353b 100644
>>>>> --- a/drivers/infiniband/sw/rxe/rxe_net.h
>>>>> +++ b/drivers/infiniband/sw/rxe/rxe_net.h
>>>>> @@ -17,6 +17,7 @@ struct rxe_recv_sockets {
>>>>>    };
>>>>>      int rxe_net_add(const char *ibdev_name, struct net_device *ndev);
>>>>> +void rxe_net_del(struct ib_device *dev);
>>>>>      int rxe_register_notifier(void);
>>>>>    int rxe_net_init(void);