• State TI Thinks Resolved
  • Locked Locked
  • Replies 9 replies
  • Answers 1 answer
  • Subscribers 33 subscribers
  • Views 1266 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.

Original question:

Linux/DP83TC811S-Q1: DP83TC811S-Q1 driver porting bring up issue

Linux/DP83TC811S-Q1: No SGMII signal output after installed dp83tc811.ko & qcom_emac.ko

user5947325
Prodigy 130 points
Part Number: DP83TC811S-Q1
Other Parts Discussed in Thread: DP83TC811, DP83TC811SEVM, DP83TC811EVM

Tool/software: Linux

Hi,

After installed driver modules, kernel tried but can't connect external phy dp83tc811s. Detailed as follows:

---- 1 ----

root@mdm9607:/usr/lib/modules# insmod dp83tc811.ko

root@mdm9607:/usr/lib/modules# insmod qcom_emac.ko

Here we got log from kernel:

[ 20.463954] [emac_probe] Start probe, dma_set_mask_and_coherent.adpt = 0x0
[ 20.503511] [emac_probe] Start probe, alloc_etherdev. adpt = 0x0
[ 20.524359] [emac_probe] Middile point1, adpt = 0xca47b500
[ 20.528811] [emac_probe] Start probe, set IRQ. adpt = 0xca47b500
[ 20.567384] [emac_probe] Start probe, emac_get_resources. adpt = 0xca47b500
[ 20.580071] [msm_emac_pinctrl_init] After get pin resource, output reset for 20ms.
[ 20.649560] [emac_probe] Start probe, msm_emac_ldo_init.adpt = 0xca47b500
[ 20.686913] [emac_probe] Start probe, emac_clks_phase1_init.adpt = 0xca47b500
[ 20.693288] [emac_probe] Start probe, emac_set_ethtool_ops.adpt = 0xca47b500
[ 20.713167] [emac_probe] Start probe, emac_phy_config_internal.adpt = 0xca47b500
[ 20.737083] [emac_phy_config_internal] Phy mode = 3.
[ 20.754374] [emac_probe] Start probe, emac_clks_phase2_init.adpt = 0xca47b500
[ 20.773631] [emac_probe] Start probe, emac_reg_r32.adpt = 0xca47b500
[ 20.782877] [emac_probe] Start probe, emac_reg_r32.hw_ver = 0x10030000
[ 20.810093] [emac_probe] Start probe, emac_init_adapter.adpt = 0xca47b500
[ 20.837950] [emac_probe] Start probe, adpt->gpio_on.adpt = 0xca47b500
[ 20.868009] [emac_probe] Start probe, emac_phy_config_external.adpt = 0xca47b500
[ 20.958950] libphy: emac-mdio: probed
[ 20.961597] [get_phy_device] mii_bus = 0xca4b6800, addr(the No.x phy) = 0, is_c45 = 0, MII_PHYSID1 = 0x2, MII_PHYSID2 = 0x3
[ 21.025501] [get_phy_device] phy_id = 0x200196ff
[ 21.047171] phydrv->phy_id = 0x2000a251, phydev->phy_id = 0x200196ff, phydrv->phy_id_mask = 0xfffffff0
[ 21.109920] phydrv->phy_id = 0xffffffff, phydev->phy_id = 0x200196ff, phydrv->phy_id_mask = 0xffffffff
[ 21.158556] phydrv->phy_id = 0x200196ff, phydev->phy_id = 0x200196ff, phydrv->phy_id_mask = 0xfffffff0
[ 21.187210] [emac_phy_config_external] Phydev = 0xca4b6c00.
[ 21.191756] [emac_phy_config_external] Phy id = 0x200196ff.
[ 21.217417] [emac_probe] Start probe, emac_hw_reset_mac.adpt = 0xca47b500
[ 21.234241] [emac_probe] Start probe, emac_work_thread.

---- 2 ----

Now we can find eth0 in system, and can operate it for some case.

root@mdm9607:~# ifconfig eth0
eth0 Link encap:Ethernet HWaddr 00:00:00:00:00:00
BROADCAST MULTICAST MTU:1500 Metric:1
RX packets:0 errors:0 dropped:0 overruns:0 frame:0
TX packets:0 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:0 (0.0 B) TX bytes:0 (0.0 B)
Interrupt:47

---- 3 ----

For other more cases we can't operate eth0 successfully.

root@mdm9607:~# ethtool eth0

Settings for eth0:
Cannot get device settings: Invalid argument --- CANT GET INFORMATION HERE.
Supports Wake-on: pg
Wake-on: p
Current message level: 0x00007fff (32767)
drv probe link timer ifdown ifup rx_err tx_err tx_queued intr tx_done rx_status pktdata hw wol
Link detected: no
root@mdm9607:~# ethtool -i eth0
driver: eth0 --- HERE DRIVER IS NOT DP83TC811.
version: Revision: 1.1.0.0
firmware-version:
bus-info: eth0
supports-statistics: yes
supports-test: no
supports-eeprom-access: no
supports-register-dump: yes
supports-priv-flags: no

root@mdm9607:~# ifconfig eth0 hw ether aa:11:22:33:44:55

root@mdm9607:~# [ 1891.737532] qcom-emac 7c40000.qcom,emac eth0: TX queues 1, TX descriptors 512
[ 1891.743637] qcom-emac 7c40000.qcom,emac eth0: RX queues 1, Rx descriptors 256
[ 1891.765380] [emac_mac_up] Calling phy_connect_direct.
[ 1891.769701] qcom-emac 7c40000.qcom,emac eth0: could not connect phy --- HERE WE CANT CONNECT PHY.
[ 1891.800917] QTI:Netlink Query to Kernel Success
[ 1892.016804] qcom-emac 7c40000.qcom,emac eth0: TX queues 1, TX descriptors 512
[ 1892.022916] qcom-emac 7c40000.qcom,emac eth0: RX queues 1, Rx descriptors 256
[ 1892.057049] [emac_mac_up] Calling phy_connect_direct.
[ 1892.061370] qcom-emac 7c40000.qcom,emac eth0: could not connect phy --- HERE WE CANT CONNECT PHY.
[ 1892.096782] QTI:Netlink Query to Kernel Success

---- 4 ----

I tried to check port configuration and signals for SGMII.

root@mdm9607:~# cat /sys/bus/mdio_bus/devices/7c40000.qcom,emac\:00/phy_id
0x200196ff --- THIS ID IS NOT MATCH SPECIFICATION PDF.
root@mdm9607:~# cat /sys/bus/mdio_bus/devices/7c40000.qcom,emac\:00/phy_interface
sgmii --- THIS PORT CONFIGURED AS SGMII, IT'S OK.

The signal from TX_M / TX_P and RX_M / RX_P we grasped by an osilloscope, were ANALOG signals and there is NO DIFFERENTAL.

Now we don't have more ideas and project is delaying :.<

Please help us~~~

  • 0
    TI__Expert 5670 points
    Hi user5947325 ,

    It looks like MDC/MDIO communication to the device is not working correctly.

    Can you share the schematic?

    Regards,

    Hung Nguyen
  • 0 in reply to Hung Nguyen92
    Prodigy 130 points

    Dear Nguyen,

      Thank you for your fast reply.

      We are using DP83TC811SEVM board, and Qualcomm Mdm9607 Internal Emac to contorl.

      We also have another DP83TC811EVM board (NO S for chip name), and from 2 board we can read phy chip ID through MDC / MDIO.

      For DP83TC811SEVM board, we got ID is 0x200196ff, so I still wonder it not like described in datasheet. This is suspected we also confused. If we have some user application to test MDC / MDIO?

      We use DP83TC811SEVM  board so if you need mdm9607 schematic?

  • 0 in reply to user5947325
    Prodigy 130 points

    Dear Nguyen,

     I steped into kernel and tried to get how phy chip get up, by printk during mac up. Log as below:

    [   23.084743] [emac_mac_up] Calling phy_connect_direct. adpt->phy.is_ext_phy_connect = 0

    [   23.091629] [phy_attach_direct] Calling driver->probe. d->driver = 0xbf013740

    [   23.156078] [phy_attach_direct] Get if attached. phydev->attached_dev = 0

    [   23.161837] [phy_attach_direct] Try to get bus module, bus_module = 0

    [   23.206421] [phy_attach_direct] Calling phy_init_hw.

    [   23.210444] [phy_init_hw] Calling phydev->drv->config_init.

    [   23.234476] [PR] Reg = 0x1, value = 0xc7

       --- Here we read PHY chip register, used phy_read in linux kernel.

       --- Phy register ID is 1, BMSR Register 0x0001 – Basic Mode Status Register.

    [   23.237513] [phy_connect_direct] Called phy_attach_direct. rc = -95

    [   23.243627] qcom-emac 7c40000.qcom,emac eth0: could not connect phy

    [   23.334244] [emac_mac_up] Error req irq exit! Cant connect phy, called adpt->phy.ops.down

        --- [PR] : This line we add code in kernel/include/linux/phy.h, function phy_read.

    This is information update, for analysis reference.

    Regards,

  • 0 in reply to user5947325
    Prodigy 130 points
    Dear Nguyen,
    I did further code explore and find that the chip driver has not hooked in kernel, instead of general phy driver did the job.
    The code I added, and which has not printed yet, even after insmod ko file.
    static int dp83811_config_intr(struct phy_device *phydev)
    {
    int misr_status, err;

    printk(KERN_ALERT "[dp83811_config_intr] Calling func.\n"); // Add this line to indicate if dp83tc811 driver effected.
    ...
    Please give me some guidance :..<
    Regards,
  • 0 in reply to user5947325
    TI__Expert 5670 points
    Hi,

    I have limited expertise in Linux driver. I would suggest asking for support from Linux experts in Linux forum. For example: lore.kernel.org/.../

    As for MDC/MDIO tool, please follow this link for both software and required hardware: www.ti.com/.../USB-2-MDIO

    Regards,

    Hung Nguyen
  • 0 in reply to user5947325
    Prodigy 130 points
    This is not a proper init function I put code, init function should be dp83811_config_init (intr is interrupt :..< )
  • 0 in reply to Hung Nguyen92
    Prodigy 130 points
    Thanks dear Nguyen, I will go to linux forum for linux mechanism. Still if I have chip related issues I'd like to get your help. Is that OK?
  • 0 in reply to user5947325
    TI__Expert 5670 points
    Hi John,

    Yes, we can definitely help to resolve chip related issues.

    I will go ahead and close this thread. Kindly open a new thread when you encounter chip related issue.

    Regards,

    Hung Nguyen
  • 0 in reply to Hung Nguyen92
    Prodigy 130 points

    Hi, Dear Nguyen,

      Thank u anyway, I got some support from kernel side and still not working.

      I'll collect information about this application and feedback by another thread.

    Regards,