[dpdk-dev] bnx2x pmd performance expectations

[dpdk-dev] bnx2x pmd performance expectations

[Alexander Belyakov]

Hi,

just tried to forward a lot of tiny packets with testpmd (dpdk-2.2.0)
running on Broadcom Corporation NetXtreme II BCM57810S 10 Gigabit Ethernet
(rev 10) adapter. I see forwarding performance of 2.6Mpps instead of
expected 14.8Mpps. What should be done to achieve better results?

Thank you,
Alexander Belyakov

Re: [dpdk-dev] bnx2x pmd performance expectations

[Chas Williams]

I wouldn't consider myself an expert on this driver but while looking
at some other things, I have noted that RTE_PMD_BNX2X_TX_MAX_BURST is
defined to be 1.  This bursts single packets to bnx2x_tx_encap() but it
looks like bnx2x_tx_encap() is far more capable than that.

On Tue, 2015-12-22 at 14:52 +0300, Alexander Belyakov wrote:
> Hi,
> 
> just tried to forward a lot of tiny packets with testpmd (dpdk-2.2.0)
> running on Broadcom Corporation NetXtreme II BCM57810S 10 Gigabit
> Ethernet
> (rev 10) adapter. I see forwarding performance of 2.6Mpps instead of
> expected 14.8Mpps. What should be done to achieve better results?
> 
> Thank you,
> Alexander Belyakov

Re: [dpdk-dev] bnx2x pmd performance expectations

[Alexander Belyakov]

Thank you for pointing this out. While it seems to me problem here is RX, I
will also look into TX burst limitations.

-a

On 28 December 2015 at 00:17, Chas Williams <3chas3@gmail.com> wrote:

> I wouldn't consider myself an expert on this driver but while looking
> at some other things, I have noted that RTE_PMD_BNX2X_TX_MAX_BURST is
> defined to be 1.  This bursts single packets to bnx2x_tx_encap() but it
> looks like bnx2x_tx_encap() is far more capable than that.
>
> On Tue, 2015-12-22 at 14:52 +0300, Alexander Belyakov wrote:
> > Hi,
> >
> > just tried to forward a lot of tiny packets with testpmd (dpdk-2.2.0)
> > running on Broadcom Corporation NetXtreme II BCM57810S 10 Gigabit
> > Ethernet
> > (rev 10) adapter. I see forwarding performance of 2.6Mpps instead of
> > expected 14.8Mpps. What should be done to achieve better results?
> >
> > Thank you,
> > Alexander Belyakov
>

Re: [dpdk-dev] bnx2x pmd performance expectations

[Harish Patil]

>
>Thank you for pointing this out. While it seems to me problem here is RX,
>I
>will also look into TX burst limitations.
>
>-a
>
>On 28 December 2015 at 00:17, Chas Williams <3chas3@gmail.com> wrote:
>
>> I wouldn't consider myself an expert on this driver but while looking
>> at some other things, I have noted that RTE_PMD_BNX2X_TX_MAX_BURST is
>> defined to be 1.  This bursts single packets to bnx2x_tx_encap() but it
>> looks like bnx2x_tx_encap() is far more capable than that.
>>
>> On Tue, 2015-12-22 at 14:52 +0300, Alexander Belyakov wrote:
>> > Hi,
>> >
>> > just tried to forward a lot of tiny packets with testpmd (dpdk-2.2.0)
>> > running on Broadcom Corporation NetXtreme II BCM57810S 10 Gigabit
>> > Ethernet
>> > (rev 10) adapter. I see forwarding performance of 2.6Mpps instead of
>> > expected 14.8Mpps. What should be done to achieve better results?
>> >
>> > Thank you,
>> > Alexander Belyakov
>>
>

Alex/Chas,
We have HW/FW limitation for PPS. The adapter can only do ~3 MPPS.
However, yes we can revamp the Tx routines a bit to transmit multiple
packets in a single burst at a time using single doorbelling mechanism.

Thanks
Harish



________________________________

This message and any attached documents contain information from the sending company or its parent company(s), subsidiaries, divisions or branch offices that may be confidential. If you are not the intended recipient, you may not read, copy, distribute, or use this information. If you have received this transmission in error, please notify the sender immediately by reply e-mail and then delete this message.