[PATCH V5] ethdev: fix one address occupies two indexes in MAC addrs

[lihuisong (C)]

在 2023/1/18 16:38, Thomas Monjalon 写道:
> 18/01/2023 09:26, Thomas Monjalon:
>> 20/10/2022 11:31, Huisong Li:
>>> The dev->data->mac_addrs[0] will be changed to a new MAC address when
>>> applications modify the default MAC address by .mac_addr_set(). However,
>>> if the new default one has been added as a non-default MAC address by
>>> .mac_addr_add(), the .mac_addr_set() doesn't remove it from the mac_addrs
>>> list. As a result, one MAC address occupies two indexes in the list. Like:
>>> add(MAC1)
>>> add(MAC2)
>>> add(MAC3)
>>> add(MAC4)
>>> set_default(MAC3)
>>> default=MAC3, filters=MAC1, MAC2, MAC3, MAC4
>> I agree it may be simpler to ensure that the addresses are uniques.
>>
>>> In addition, some PMDs, such as i40e, ice, hns3 and so on, do remove the
>>> old default MAC when set default MAC. If user continues to do
>>> set_default(MAC5), and the mac_addrs list is default=MAC5, filters=(MAC1,
>>> MAC2, MAC3, MAC4). At this moment, user can still see MAC3 from the list,
>>> but packets with MAC3 aren't actually received by the PMD.
>> If MAC3 is not removed with rte_eth_dev_mac_addr_remove() by the app,
>> MAC3 packets should be received.
>> The MAC address should not be removed by the PMD.
> Sorry for the confusion. I mean the opposite.
> With the new behavior, if we want the old default address to be kept
> in the list, we need to add it explicitly with rte_eth_dev_mac_addr_add.
> This is a behavior change and must be highlighted in the release notes.
Hi Thomas,

Sorry for your confusion. Maybe my cmmmit log is misleading.
Please ignore the second paragraph description in my commit.

It has nothing to do with the old defalut address.

The purpose of this patch is to solve the problem that one
address occupies two entries in MAC list.
All network engines have this problem. If the new defalut address
has been added as a non-default MAC address to the MAC list, this
address will occupy two entries in MAC list.

It's still a problem for some network engines, like, hns3 and i40e.
That's what I described in the second paragraph in my commit.

/Huisong
>>> So this patch adds a remove operation in set default MAC API documents
>>> this behavior.
>> Let's be clear here: only the new default address is removed from the rest of the list.
>>
>>> Fixes: 854d8ad4ef68 ("ethdev: add default mac address modifier")
>>> Cc: stable at dpdk.org
>>> --- a/lib/ethdev/ethdev_driver.h
>>> +++ b/lib/ethdev/ethdev_driver.h
>>> @@ -116,7 +116,12 @@ struct rte_eth_dev_data {
>>>   
>>>   	uint64_t rx_mbuf_alloc_failed; /**< Rx ring mbuf allocation failures */
>>>   
>>> -	/** Device Ethernet link address. @see rte_eth_dev_release_port() */
>>> +	/**
>>> +	 * Device Ethernet link address. The index zero of the array is as the
>> It should be "addresses" as there can be multiple.
>>
>> What means "as" above?
>> Can we say the first entry (index zero) is the default address?
>>
>>> +	 * index of the default address, and other indexes can't be the same
>> You can split the sentence in 2 instead of ", and".
>> indexes -> entries
>> can't -> cannot
>>
>>> +	 * as the address corresponding to index 0.
>> simpler: as the default address.
>>
>>> +	 * @see rte_eth_dev_release_port()
>> Why referencing this function here?
>>
>>> +	 */
>>> --- a/lib/ethdev/rte_ethdev.c
>>> +++ b/lib/ethdev/rte_ethdev.c
>>> @@ -4498,7 +4498,10 @@ rte_eth_dev_mac_addr_remove(uint16_t port_id, struct rte_ether_addr *addr)
>>>   int
>>>   rte_eth_dev_default_mac_addr_set(uint16_t port_id, struct rte_ether_addr *addr)
>>>   {
>>> +	uint64_t mac_pool_sel_bk = 0;
>>>   	struct rte_eth_dev *dev;
>>> +	uint32_t pool;
>>> +	int index;
>>>   	int ret;
>>>   
>>>   	RTE_ETH_VALID_PORTID_OR_ERR_RET(port_id, -ENODEV);
>>> @@ -4517,16 +4520,50 @@ rte_eth_dev_default_mac_addr_set(uint16_t port_id, struct rte_ether_addr *addr)
>>>   	if (*dev->dev_ops->mac_addr_set == NULL)
>>>   		return -ENOTSUP;
>>>   
>>> +	/*
>>> +	 * If the address has been added as a non-default MAC address by
>>> +	 * rte_eth_dev_mac_addr_add API, it should be removed from
>>> +	 * dev->data->mac_addrs[].
>>> +	 */
>> Make is simpler:
>> "Keep address unique in dev->data->mac_addrs[]."
>>
>>> +	index = eth_dev_get_mac_addr_index(port_id, addr);
>>> +	if (index > 0) {
>>> +		/* Remove address in dev data structure */
>>> +		mac_pool_sel_bk = dev->data->mac_pool_sel[index];
>>> +		ret = rte_eth_dev_mac_addr_remove(port_id, addr);
>>> +		if (ret < 0) {
>>> +			RTE_ETHDEV_LOG(ERR, "Delete MAC address from the MAC list of ethdev port %u.\n",
>>> +				       port_id);
>> It is not clear with this log that it failed.
>>
>>> +			return ret;
>>> +		}
>>> +		/* Reset pool bitmap */
>>> +		dev->data->mac_pool_sel[index] = 0;
>> mac_pool_sel[index] is already reset in rte_eth_dev_mac_addr_remove().
>>
>>> +	}
>>>   	ret = (*dev->dev_ops->mac_addr_set)(dev, addr);
>>>   	if (ret < 0)
>>> -		return ret;
>>> +		goto out;
>>>   
>>>   	/* Update default address in NIC data structure */
>>>   	rte_ether_addr_copy(addr, &dev->data->mac_addrs[0]);
>>>   
>>>   	return 0;
>>> -}
>>>   
>>> +out:
>>> +	if (index > 0) {
>>> +		pool = 0;
>>> +		do {
>>> +			if (mac_pool_sel_bk & UINT64_C(1)) {
>>> +				if (rte_eth_dev_mac_addr_add(port_id, addr,
>>> +							     pool) != 0)
>>> +					RTE_ETHDEV_LOG(ERR, "failed to restore MAC pool id(%u) in port %u.\n",
>>> +						       pool, port_id);
>>> +			}
>>> +			mac_pool_sel_bk >>= 1;
>>> +			pool++;
>>> +		} while (mac_pool_sel_bk != 0);
>>> +	}
>>> +
>>> +	return ret;
>>> +}
>> Can we avoid this rollback by removing the address after mac_addr_set() succeed?
>
>
> .