Re: [PATCH v5 5/6] ixgbe: Config VF RSS

[Vlad Zolotarov <vladz-RmZWMc9puTNJc61us3aD9laTQe2KTcn/@public.gmane.org>]

On 01/09/15 08:07, Ouyang, Changchun wrote:
>
>> -----Original Message-----
>> From: Vlad Zolotarov [mailto:vladz-RmZWMc9puTNJc61us3aD9laTQe2KTcn/@public.gmane.org]
>> Sent: Thursday, January 8, 2015 5:43 PM
>> To: Ouyang, Changchun; dev-VfR2kkLFssw@public.gmane.org
>> Subject: Re: [dpdk-dev] [PATCH v5 5/6] ixgbe: Config VF RSS
>>
>>
>> On 01/07/15 08:32, Ouyang Changchun wrote:
>>> It needs config RSS and IXGBE_MRQC and IXGBE_VFPSRTYPE to enable VF
>> RSS.
>>> The psrtype will determine how many queues the received packets will
>>> distribute to, and the value of psrtype should depends on both facet:
>>> max VF rxq number which has been negotiated with PF, and the number of
>> rxq specified in config on guest.
>>> Signed-off-by: Changchun Ouyang <changchun.ouyang-ral2JQCrhuEAvxtiuMwx3w@public.gmane.org>
>>>
>>> Changes in v4:
>>>    - the number of rxq from config should be power of 2 and should not
>> bigger than
>>>       max VF rxq number(negotiated between guest and host).
>>>
>>> ---
>>>    lib/librte_pmd_ixgbe/ixgbe_pf.c   |  15 ++++++
>>>    lib/librte_pmd_ixgbe/ixgbe_rxtx.c | 103
>> +++++++++++++++++++++++++++++++++-----
>>>    2 files changed, 106 insertions(+), 12 deletions(-)
>>>
>>> diff --git a/lib/librte_pmd_ixgbe/ixgbe_pf.c
>>> b/lib/librte_pmd_ixgbe/ixgbe_pf.c index dbda9b5..93f6e43 100644
>>> --- a/lib/librte_pmd_ixgbe/ixgbe_pf.c
>>> +++ b/lib/librte_pmd_ixgbe/ixgbe_pf.c
>>> @@ -187,6 +187,21 @@ int ixgbe_pf_host_configure(struct rte_eth_dev
>> *eth_dev)
>>>    	IXGBE_WRITE_REG(hw, IXGBE_MPSAR_LO(hw-
>>> mac.num_rar_entries), 0);
>>>    	IXGBE_WRITE_REG(hw, IXGBE_MPSAR_HI(hw-
>>> mac.num_rar_entries), 0);
>>>
>>> +	/*
>>> +	 * VF RSS can support at most 4 queues for each VF, even if
>>> +	 * 8 queues are available for each VF, it need refine to 4
>>> +	 * queues here due to this limitation, otherwise no queue
>>> +	 * will receive any packet even RSS is enabled.
>>> +	 */
>>> +	if (eth_dev->data->dev_conf.rxmode.mq_mode ==
>> ETH_MQ_RX_VMDQ_RSS) {
>>> +		if (RTE_ETH_DEV_SRIOV(eth_dev).nb_q_per_pool == 8) {
>>> +			RTE_ETH_DEV_SRIOV(eth_dev).active =
>> ETH_32_POOLS;
>>> +			RTE_ETH_DEV_SRIOV(eth_dev).nb_q_per_pool = 4;
>>> +			RTE_ETH_DEV_SRIOV(eth_dev).def_pool_q_idx =
>>> +				dev_num_vf(eth_dev) * 4;
>>> +		}
>>> +	}
>>> +
>>>    	/* set VMDq map to default PF pool */
>>>    	hw->mac.ops.set_vmdq(hw, 0,
>>> RTE_ETH_DEV_SRIOV(eth_dev).def_vmdq_idx);
>>>
>>> diff --git a/lib/librte_pmd_ixgbe/ixgbe_rxtx.c
>>> b/lib/librte_pmd_ixgbe/ixgbe_rxtx.c
>>> index f69abda..e83a9ab 100644
>>> --- a/lib/librte_pmd_ixgbe/ixgbe_rxtx.c
>>> +++ b/lib/librte_pmd_ixgbe/ixgbe_rxtx.c
>>> @@ -3327,6 +3327,68 @@ ixgbe_alloc_rx_queue_mbufs(struct
>> igb_rx_queue *rxq)
>>>    }
>>>
>>>    static int
>>> +ixgbe_config_vf_rss(struct rte_eth_dev *dev) {
>>> +	struct ixgbe_hw *hw;
>>> +	uint32_t mrqc;
>>> +
>>> +	ixgbe_rss_configure(dev);
>>> +
>>> +	hw = IXGBE_DEV_PRIVATE_TO_HW(dev->data->dev_private);
>>> +
>>> +	/* MRQC: enable VF RSS */
>>> +	mrqc = IXGBE_READ_REG(hw, IXGBE_MRQC);
>>> +	mrqc &= ~IXGBE_MRQC_MRQE_MASK;
>>> +	switch (RTE_ETH_DEV_SRIOV(dev).active) {
>>> +	case ETH_64_POOLS:
>>> +		mrqc |= IXGBE_MRQC_VMDQRSS64EN;
>>> +		break;
>>> +
>>> +	case ETH_32_POOLS:
>>> +	case ETH_16_POOLS:
>> Isn't ETH_16_POOLS mode is invalid for VF RSS? It's what both spec states
>> and what u handle in this patch in ixgbe_pf_host_configure().
>> IMHO it would be better to treat this mode value as an error here since if u
>> get it here it indicates of a SW bug.
> I think we discussed it before already,  return err here will break here in the case of max vf number is less than 16.
> If doing that, This make the library seems can't support vf rss in the case of max vf num less than 16.
> So we obviously don't hope it break here.

I don't remember we were discussing these specific lines. However I do 
remember we talked about the previous section of this patch.
I'm afraid u are missing my point here: ixgbe_pf_host_configure() is 
called before ixgbe_config_vf_rss() in the ixgbe_dev_start() flow. This 
means that
RTE_ETH_DEV_SRIOV(dev).active will already be adjusted by your (!!!) 
code in the ixgbe_pf_host_configure() when u get to 
ixgbe_config_vf_rss() and it should not be equal ETH_16_POOLS unless 
there is a bug in your code.

So, unless I've missed something here, don't u think an assert() would 
be appropriate if RTE_ETH_DEV_SRIOV(dev).active equals ETH_16_POOLS?

thanks,
vlad

>
>