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.

Linux/BEAGLEBOARD-X15: Wake on LAN

Part Number: BEAGLEBOARD-X15
Other Parts Discussed in Thread: AM5728

Tool/software: Linux

Hi,

I'm planning with power management system but when I use command 

$ echo mem > /sys/power/state or systemctl suspend it will be back when I press start button

1. Is it possible Wake on LAN ?
In schematic Ethernet chip it support Wake On LAN




GPIO7_27, Does it affect the WOL?

I using am57xx-evm-linux-rt-05.00.00.15
  • Our Power Management documentation for Linux on AM57xx (AKA DRA7) does not list wake on LAN as an option. In particular, look at section "Idle Power Management". I will double check.

    Regards,

  • Hello Thepnimit,

    It looks like section "Idle Power Management" needs some updates. Linux Processor SDK 5.0 and later look like they support IO daisy chain wake events.

    Wake on LAN is not enabled by TI software. However, D26 is CTRL_CORE_PAD_UART2_TXD in the AM5728 technical reference manual. It has register fields WAKEUPEVENT and WAKEUPENABLE, so the hardware is capable of supporting IO daisy chain wake event on GPIO7_27.

    I would not be able to support you throughout the process if you wanted to add Wake on LAN support to the software yourself. However, I could try to find some pointers for you if you decided to add Wake on LAN support.

    Regards,
    Nick
  • I found example edit source code chip KSZ8091 for send Magic package Wake on LAN in GITHUB. When I see datasheet it look like KSZ9031RNX just only change register address.

    I haven't tried build new source code edit in /drive/net/phy/micrel.c (may be test next week)

    But if I assume it's worked and KSZ9031RNX send something signal to trigger system through KSZ9031RNX::PME_N2 -> AM572x::GPIO7_27.

    Do you have advise about

    1. How to enable GPIO7_27 in empty mode and receive mode for waiting trigger something signal?

    2. How to system know GPIO7_27 is triggered when it's in suspend mode?

    3. If cannot use GPIO7_27 which GPIO can replace. (In TI Idle Power Management document talked about GPIO0)

    Thanks,

    Thepnimit

  • Hello Thepnimit,

    I am sorry for the delayed response. I just got back from vacation, and will post a longer reply soon.

    Regards,

    Nick

  • Ok, it looks like this should work, but so far I am only able to find documentation on how to do it on AM335x and AM437x. So you will need to do some experimentation.

    The post GPIO wake up from standby has a good template for using gpio-key driver with the pinmuxing to set up a GPIO as a wake source. It looks like GPIO1 is the only GPIO bank in the wakeup domain, so if you wanted to use GPIO7 I suspect you would need to make sure that it never enters idle. (see the post).

    Another example we could look at is an AM437x touch screen being used as a wakeup source. I have not looked too deeply into what device tree / potentially driver changes would be needed to follow that template.

    Regards,

    Nick

  • Hello Thepnimit,

    I am going to mark this thread resolved for now. Feel free to continue the conversation later if needed.

    Regards,

    Nick

  • Hi,

    I think it's impossible now, When system in suspend mode RJ45 is cut off  and chip Ethernet maybe have not power

    This magic package from ping or wol cannot detect

    Thank for help

  • Hello Thepnimit,

    On your schematic, are you saying that VDD_3v3 and ETH0_3V3 are going from 3.3V to 0V? Or is something else going on in suspend mode?

    Regards,

    Nick

  • Hi ,

    I don't sure it going 0V because I don't have tools for detect (it's so small).

    But when I follow picture

    I think it has a source from the same place.

    When system in suspend mode LED on RJ45 off.

    Dose it mean VDD_3v3 all point is cut off?

    Or I think wrong

  • I am sorry about that, I did not mean to click "TI Thinks Resolved".

    I am curious about what is causing the Ethernet to stop working - e.g.,
    * is it because power to the PHY is turned off? (use a multimeter and check if VDD_3V3 and ETH0_3V3 are 3.3V or 0V)
    * is it because or something else? e.g., the reset signal getting toggled, or the clocks turning off? (use an oscilloscope or digital logic analyzer)
    * something else?

    Depending on what causes the PHY to stop responding, we might be able to get it working.

    Regards,

    Nick