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.

  • Answer Suggested

Linux/DP83TC811SEVM: DP83TC811SEVM's communication issue

Prodigy 410 points

Replies: 11

Views: 631

Part Number: DP83TC811SEVM

Tool/software: Linux

Hi team,

I am the FAE in Guangzhou, now the customer Gosuncn is evaluating the DP83TC811S using DP83TC811SEVM.

The customer’s situation is shown below:

1. MAC interface SGMII is used in the T-Box project. And the platform is Qualcomm’s 9x07.

2. The Linux driver on the TI.com has been used and the processor can communicate with DP83TC811.

https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/drivers/net/phy/dp83tc811.c

3. The system is established using two DP83TC811SEVMs and two MDM9607 platforms.


5. The Ping test can pass as shown below, but the customer feedback that the communication between the two platforms is not ok.



My question is shown below:

1. Does the pass of Ping test show that the register and phy driver is ok?

2. Could you please deduce the reason why the communication is not ok?

3. Is there any training material about the software? Anyone in our company can provide this kind of help? Many thanks.

  • For the item 5, there is something need to correct, the Ping test between two platforms can't pass. The passed Ping test is for loopback ping test. Thanks.

  • In reply to Chen Jiao:

    Hi chen,

    Software support is usually handled by the company distributing Linux. That appears to be Qualcomm in this case.

    If MDIO works, and the PHY reports link is up, the HW is OK.

    Ping test is not a good test to use because it will not work if Linux is not configured properly.

    If using SGMII, also look at sgmii auto-negotiation complete bit in the registers. If this bit is not 1, then SGMII is not resolving properly.

    Best Regards,

    Rob Rodrigues

    CTS (Clock & Timing Solutions) Systems Engineer

  • In reply to Rob Rodrigues:

    Hi, I am a worker from Gaoxinxing company, in this case the MDIO works indeed, but the PHY reports NOLINK when I connect the two devices using two wires.

    What should we do now? Pls give us some advice. 

    Thanks.

  • In reply to Rob Rodrigues:

    Hi Rob,
    Thanks for the reply.
    We have checked the sgmii auto-negotiation complete bit and its value is 0, the SGMII is not resolving properly.

    As the customer uses the SGMII Mac connection and the DP83TC811SEVM's default state meets this requirement, the register is remain default. I have checked the hardware link and it is ok. The MDIO also works. But the PING test can't pass.

    Could you please tell me if any of the register should be configured additionally in this condition? or if the issue comes from the Linux driver? Many thanks.

    Best regards
    Chen Jiao
  • In reply to Chen Jiao:

    Hi Chen,

    How are you checking the link? Are you using MDIO register 0x1 bit[2] status?

    Best Regards,

    Rob Rodrigues

    CTS (Clock & Timing Solutions) Systems Engineer

  • In reply to Rob Rodrigues:

    Here are two demo devices:
    first device:
    / # ifconfig eth0
    eth0 Link encap:Ethernet HWaddr 00:80:48:BA:D1:30
    inet addr:169.254.4.1 Bcast:169.254.4.255 Mask:255.255.255.0
    inet6 addr: fe80::280:48ff:feba:d130/64 Scope:Link
    UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
    RX packets:0 errors:0 dropped:0 overruns:0 frame:0
    TX packets:10 errors:0 dropped:0 overruns:0 carrier:0
    collisions:0 txqueuelen:1000
    RX bytes:0 (0.0 B) TX bytes:794 (794.0 B)
    Interrupt:108

    second device:
    / # ifconfig eth0
    eth0 Link encap:Ethernet HWaddr 00:80:48:BA:D2:31
    inet addr:169.254.4.1 Bcast:169.254.4.255 Mask:255.255.255.0
    inet6 addr: fe80::280:48ff:feba:d231/64 Scope:Link
    UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
    RX packets:0 errors:0 dropped:0 overruns:0 frame:0
    TX packets:12 errors:0 dropped:0 overruns:0 carrier:0
    collisions:0 txqueuelen:1000
    RX bytes:0 (0.0 B) TX bytes:940 (940.0 B)
    Interrupt:108

    the first device's bridge0 ip addr is 192.168.225.11,second device's ip addr is 192.168.225.22, but we can't ping through each other, is there any method to
    find where the problem is?

    I am confused...............
  • In reply to Rob Rodrigues:

    Hi Rob,
    The PHY reports link is up and the IP address can be got automatically, so I think the HW is OK.
    But now the Ping test can't pass. I have checked the register 0x1 and its value is 0x61.
    0x1[2]=0 represents that link not established, could you please tell me the ‘link’ means software link or HW link?
    And 0x1[3]=0 represents that the device is not able to perform auto-negotiation, but the LED1 jumper on DP83TC811SEVM is set to AUTO mode indeed, could you please tell me the reason of this phenomenon?
    Thanks a lot.
    Best regards
    Chen Jiao
  • In reply to Rob Rodrigues:

    Hi Rob,
    The PHY reports link is up and the IP address can be got automatically, so I think the HW is OK.
    But now the Ping test can't pass. I have checked the register 0x1 and its value is 0x61.
    0x1[2]=0 represents that link not established, could you please tell me the ‘link’ means software link or HW link?
    And 0x1[3]=0 represents that the device is not able to perform auto-negotiation, but the LED1 jumper on DP83TC811SEVM is set to AUTO mode indeed, could you please tell me the reason of this phenomenon?
    Thanks a lot.
    Best regards
    Chen Jiao
  • In reply to Chen Jiao:

    Hi Chen,

    Auto-negotiation is not supported on 100base-t1 devices. By IEEE standard, auto-negotiation does not exist.

    if 0x1[2] = 0, then a HW link is not established between master and slave PHY.

    1 of your EVMs must be set to slave, 1 EVM must be set to master. If both are in AUTO mode, then the link will come up when the master device is powered on and connected to the slave.

    Best Regards,

    Rob Rodrigues

    CTS (Clock & Timing Solutions) Systems Engineer

  • In reply to Rob Rodrigues:

    Hi Rob,
    Thanks for the reply.
    Both EVMs are in AUTO mode, and one is in Slave and the other is in Master mode.
    As the 0x1[2]=0, I want to know that if the 'HW link' means the link of TRD_P and TRD_N between two phys? Does it include the link between Phy and MAC?
    Thanks a lot.

    Best regards
    Chen Jiao

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.