Re: [PATCH v3 4/5] CI: Add an sjg-lab job for the Intel Edison

On Mon, Jul 27, 2026 at 02:53:27PM -0600, Simon Glass wrote:
> On Mon, 20 Jul 2026 at 14:03, Andy Shevchenko
> <[email protected]> wrote:
> > On Mon, Jul 20, 2026 at 10:14:01AM -0600, Simon Glass wrote:
> > > On Sun, 5 Jul 2026 at 04:42, Simon Glass <[email protected]> wrote:

...

> > > > +# Intel Edison (Merrifield/Tangier); U-Boot sent over DFU into eMMC u-boot0
> > > > +edison:
> > > > +  variables:
> > > > +    ROLE: edison
> > > > +  <<: *sjg_lab_dfn
> > >
> > > Unfortunately somehow the board seems to have died - does not appear
> > > on USB anymore...perhaps it doesn't like all the cycling in the lab?
> >
> > Do you have external power supply connected?
> > It might be this:
> > https://edison-fw.github.io/edison-wiki/edison-fixup
> 
> Thanks for the pointer...I do see the board power up though ("Ready to
> receive application") so that should mean that power is OK? I don't
> have external power, just relying on USB power (which I can turn on
> and off for this port).
> 
> I basically see this at present:
> 
> $ ub-int edison
> Building U-Boot in current dir for edison
> Bootstrapping U-Boot from dir /tmp/b/edison
> USB resource not available; retrying reset with longer delays
> Recovering Edison via the xfstk BootROM downloader
> 
> XFSTK Downloader Solo 0.0.0
> Copyright (c) 2015 Intel Corporation
> Build date and time: May  2 2020 10:22:27
> 
> Intel SoC Device Detection Failed: Attempt #0
> Intel SoC Device Detection Failed: Attempt #1
> Intel SoC Device Detection Failed: Attempt #2
> Intel SoC Device Detection Failed: Attempt #3
> Intel SoC Device Detection Failed: Attempt #4
> Intel SoC Device Detection Failed: Attempt #5
> Intel SoC Device Detection Failed: Attempt #6
> Intel SoC Device Detection Failed: Attempt #7
> Intel SoC Device Detection Failed: Attempt #8
> Intel SoC Device Detection Failed: Attempt #9

Hmm... Have you checked with `lsusb`? Is it there with the proper IDs?

> I suppose something could have happened to the relays or cables...

Often we saw shitty USB cables. Other than that I have no ideas, but
in case you will attend Kernel Recipes in Paris I can bring a new one
to you. I'm not sure about LPC, though.

Ah, one thing is the eMMC that theoretically can go away. But I never heard
of a such in conjunction with Intel Edison.

-- 
With Best Regards,
Andy Shevchenko