Suspend / Resume Issue

    Dears,

    I'm using LE8.0.2 on Braswel E8000( Bios freshly resetted & Only one USB Hub is connected for the first test

    after deep test, I found one issue about suspend

    When in a shell, I type systemctl suspend. My system entered in suspend mode and immediately after it wakes up

    A workaround is to disable XHC in /proc/acpi/wakeup or to have no usb plugged.

    My concern is I cannot use this workaround because I need usb to wake up my system

    After investigation, I installed ubuntu LTS 16.02 on the same system ( and again with a Bios freshly reseted ).

    What a surprise to see that suspend works perfectly !

    Could you have some guidance to fix / found the correct workaround ?

    Thanks in advance

    JP

    And in case of Ubuntu what happened with USB devices? Could you wake up from USB or it was disabled too?

    I have similar system withe same problem - USB Flirc receiver can't wake system.From what I understand the problem lies in BIOS/EFI -it looks like the ACPI wakeup tables are wrong. At /proc/acpi/wakeup there is XHC1 enabled only for S4 wakeup and not S3.

    There is no workaround I know of.

    Hi,

    The purpose of my test was to determine if it's linked to BIOS or SW.

    Again the test case is simple (I'm not targetted at this stage a usb wakeup , it's just simple test step/step)

    XHC1 is enabled in /proc/acpi/wakeup

    1/ Bios freshly reset -> LE freshly installed -> only USB key connected -> systemctl suspend -> (Immediately wakeup )

    2/ Bios freshly reset -> Ubuntu freshly installed -> only USB key connected -> systemctl suspend -> system suspended !

    regarding this test, I 'm not confident that the root cause is Bios

    Honestly I'm pure HW specialist with some SW basics.

    If you provide some guidande I can investigate by myself .but alone I cannot :(

    I'm 200% convinced we can found the rootcause toghether

    thanks in advance

    JP

    I wrote what is root cause in my case :)

    And your test was little too simple with Ubuntu. Did you check if XHC1 is enabled or disabled? Also appropriate PCI bus power where USB is connected. Also is there any script in Ubuntu which is run on suspend/resume to disable USB. I tested Ubuntu too and resume from USB didn't work. Which means it was disabled somehow. And because of that it is useless to me.

    And even disabling USB in my case suspend/resume is hit and miss - it doesn't work 100%. Don't know why.

    Btw: my system is SolidPC from Solidrun if that matters. And I talk about this issue with one of their specialist (who told ACPI wakeup tables are wrong - probably). They didn't investigate issue I assume because didn't get any more feedback.

    You are right !

    I was focused only on suspend , not including the wakeup

    It doesn't work for me too in ubuntu

    I'll try something alternative, I hope it works... we stay in touch

    regards

    JP

    Hi ,

    I succeed to solve the suspend issue

    The solution arose from the spurious wakeup after the suspend

    After deep night check, the issue is only linked to HW issue on the carrier board

    Please have a look on solidPC Carrier Schematic

    on page 2 top right the power supply of USB (V5S_USB) is provided by current limiter ( NCP380) and by V5S

    Which means 5V suspended !

    this is extremely logical, when LE is entering in suspend , the internal PSU switch off this line and then no power are applied on USB devices.

    the spurious wake up is just generated by the delay of NCP380.

    On my own carrier , I change from V5S by V5A ( which is always applied ) and the suspend / resume by USB is working like a charm

    Regards

    JP

    Nice find. Maybe I could change that on my board too?

    But if I understand correctly then Flirc IR receiver can't resume system because it is powered off?

    If you Flir is USB device, no chance ! it 's not powered

    If you are using Solid run board, unfortunately, there is no permanent 5V on the board

    (Hopefully on my carrier , Ive put a secondary 5V power supply for internal use)

    I'll report this HW bug to SolidRun if I found the relevant person to advice

    REgards

    JP

    So nothing you found is useful for me ;)

    I already pass this info on IRC to a guy I talk about this issue. Waiting for a comment.

    I just noticed we talk on solidrun forum too :P

    I don't think I can :(

    Did you use rainconf tool to set the required bit?

    And you should mention that you are using your own carrier because few post above you wrote resume with USB device doesn't work on standard carrier.

    And btw: Why did you made your own carrier? What is special on it?

    Seems I did make resume by cec here: Suspend Issue - Page 2 - SolidRun Community

    That's the point !

    I strictly followed the procedure and just after "rainconf -C 0005 /dev/ttyACM0"

    my RC wasn't reponding anymore to any key (and I'm testing on SR carrier & my own carrier)

    Are you sure 0005 is the accurate parameter ?

    Why my own carrier? because I need a solution for with minimum connectivity (1 eth, 1hdmi, 1 PS, 2USB in a box of 60x70x30mm )

    Regards

    JP

    As you can see it was working as described back in a days. I will try to reproduce again but as you already figure out bios settings matter. And I don't remember how I have it set. So it will be again try & errors. Not to mention I need to switch hdmi cables again (my AVR doesn't pass CEC). So it is a messy job. Would need some n*m hdmi matrix switch :P

    I flashed BIOS with latest one. Also flashed CEC firmware. CEC works.

    But suspend doesn't work correctly. Maybe it has something to do with BIOS settings. Who knows but I'm tired of this mouse catching game when there are only few people involved.

    Maybe you should open support ticket and fix this issues with solidrun engineers.