[dpdk-dev] [PATCH] ethdev: forbid MTU set before device configure

[Andrew Rybchenko]

On 11/1/21 12:39 PM, Jiang, YuX wrote:
>> -----Original Message-----
>> From: Yigit, Ferruh <ferruh.yigit at intel.com>
>> Sent: Friday, October 29, 2021 5:22 PM
>> To: Andrew Rybchenko <andrew.rybchenko at oktetlabs.ru>; Jiang, YuX
>> <yux.jiang at intel.com>; Thomas Monjalon <thomas at monjalon.net>; Ajit
>> Khaparde <ajit.khaparde at broadcom.com>; Huisong Li
>> <lihuisong at huawei.com>; Ananyev, Konstantin
>> <konstantin.ananyev at intel.com>; Xu, Rosen <rosen.xu at intel.com>
>> Cc: dev at dpdk.org; Ivan Ilchenko <ivan.ilchenko at oktetlabs.ru>
>> Subject: Re: [dpdk-dev] [PATCH] ethdev: forbid MTU set before device
>> configure
>>
>> On 10/29/2021 7:19 AM, Andrew Rybchenko wrote:
>>> On 10/29/21 8:58 AM, Jiang, YuX wrote:
>>>>> -----Original Message-----
>>>>> From: dev <dev-bounces at dpdk.org> On Behalf Of Andrew Rybchenko
>>>>> Sent: Friday, October 22, 2021 6:18 PM
>>>>> To: Thomas Monjalon <thomas at monjalon.net>; Yigit, Ferruh
>>>>> <ferruh.yigit at intel.com>; Ajit Khaparde
>>>>> <ajit.khaparde at broadcom.com>; Huisong Li <lihuisong at huawei.com>;
>>>>> Ananyev, Konstantin <konstantin.ananyev at intel.com>; Xu, Rosen
>>>>> <rosen.xu at intel.com>
>>>>> Cc: dev at dpdk.org; Ivan Ilchenko <ivan.ilchenko at oktetlabs.ru>
>>>>> Subject: [dpdk-dev] [PATCH] ethdev: forbid MTU set before device
>>>>> configure
>>>>>
>>>>> From: Ivan Ilchenko <ivan.ilchenko at oktetlabs.ru>
>>>>>
>>>>> rte_eth_dev_configure() always sets MTU to either
>>>>> dev_conf.rxmode.mtu or RTE_ETHER_MTU if application doesn't provide
>> the value.
>>>>> So, there is no point to allow rte_eth_dev_set_mtu() before since
>>>>> set value will be overwritten on configure anyway.
>>>>>
>>>>> Fixes: 1bb4a528c41f ("ethdev: fix max Rx packet length")
>>>>>
>>>>> Signed-off-by: Ivan Ilchenko <ivan.ilchenko at oktetlabs.ru>
>>>>> Signed-off-by: Andrew Rybchenko <andrew.rybchenko at oktetlabs.ru>
>>>>> ---
>>>>>    lib/ethdev/rte_ethdev.c | 7 +++++++
>>>>>    1 file changed, 7 insertions(+)
>>>>>
>>>>> diff --git a/lib/ethdev/rte_ethdev.c b/lib/ethdev/rte_ethdev.c index
>>>>> 4ea5a657e0..0d7dd68dc1 100644
>>>>> --- a/lib/ethdev/rte_ethdev.c
>>>>> +++ b/lib/ethdev/rte_ethdev.c
>>>>> @@ -3728,6 +3728,13 @@ rte_eth_dev_set_mtu(uint16_t port_id,
>>>>> uint16_t
>>>>> mtu)
>>>>>                return ret;
>>>>>        }
>>>>>
>>>>> +    if (dev->data->dev_configured == 0) {
>>>>> +        RTE_ETHDEV_LOG(ERR,
>>>>> +            "Port %u must be configured before MTU set\n",
>>>>> +            port_id);
>>>>> +        return -EINVAL;
>>>>> +    }
>>>>> +
>>>>>        ret = (*dev->dev_ops->mtu_set)(dev, mtu);
>>>>>        if (ret == 0)
>>>>>            dev->data->mtu = mtu;
>>>>> --
>>>>> 2.30.2
>>>> Hi Ivan,
>>>>
>>>> We meet an issue based on this patch when test dpdk21.11-rc1.
>>>> BugID: https://bugs.dpdk.org/show_bug.cgi?id=864, could you pls have a
>> look?
>>>>
>>>
>>> Hi, I think the problem is fixed by [1].
>>>
>>> [1]
>>> https://patches.dpdk.org/project/dpdk/patch/20211024164237.2116166-1-
>> a
>>> ndrew.rybchenko at oktetlabs.ru/
>>
>> That patch is merged to next-net, Yu can you please verify the issue with
>> latest next-net?
> 
> Hi Ferruh,
> We verify patch https://patches.dpdk.org/project/dpdk/patch/20211024164237.2116166-1-andrew.rybchenko@oktetlabs.ru/ on baseline dpdk main branch 6cc51b1293ceac4a77d4bf7ac91a8bbd59e1f78c
> Test failed on nic XXV710 for 25GbE SFP28 158b/os Red Hat Enterprise Linux 8.4.
> Need we verify it on next-net?
> 

What exactly goes wrong in the test if the patch is applied?

Andrew.