• State
  • Locked Locked
  • Replies 1 reply
  • Subscribers 98 subscribers
  • Views 213 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

NetCP hangs if data rate exceeds 1Gb/s

Sebastian D.
Intellectual 265 points

Hi,

we are working on EVM6670 with CCS 5.2 and pdk 6670 1.1.1.4.  We use the PA LLD to send data from the DSP via Ethernet.

QMSS Packets are pushed to PA queue number 5 to add a checksum and contain a route to the EMAC.

We noticed, that when the data rate exceeds a gigabit the PacketDMA does not return/recycle Packets to the FDQ.

Is this problem resulting from the PA being faster than the EMAC?

Is there a workaround or a solution to this problem?

Thaks,

Sebastian

  • 0
    TI__Intellectual 1585 points

    Hi Sebastion, 

    Okay, so you're sending packets to PA Queue number 5 (I'm assuming you mean queue 644 for Multiroute) such that it will eventually go to the EMAC-> SGMII-> Network, correct? 

    So your thinking is right (as it shows in the user guide on http://www.ti.com/lit/ug/sprugs4a/sprugs4a.pdf at page 21). You send the packet out to the queue, tell it to go to 644, it'll send it to the TX DMA (returning the descriptors to FDQ) then send it to the EMAC/GbE/SGMII as an egress packet should. 

    However, the GbE is called the GbE for a reason, it only supports up to 1 Mega byte on the MAC module (as seen in http://www.ti.com/lit/ug/sprugv9b/sprugv9b.pdf on page 25) for the 6670 device. So, you're right. The MAC is too slow for the faster Packet Accelerator. 

    As far as a workaround, what is your goal exactly? Maybe we might be able to find a quicker solution. 

    Hope this helps! Regards, 

    Kat Kelsch