On Thu, 14 Mar 2019 18:16:32 +0000
Peter Robinson <pbrobinson(a)gmail.com> wrote:
> Just tested with Kernel 5.0.1-300.fc29.armv7hl, same problem,
> of 3 reboots one did not manage to initialise network.
So one of the patches I referenced above, "usb: dwc3: exynos: Fix
error handling of clk_prepare_enable" is in 5.0 and just landed in
the other problem is basically summarised in the following
But should be mitigatable with "rd.driver.pre=xhci-plat-hcd "
Sadly not, I boot with
(because I want the 5000M mode).
One thing to note, since my initial testing I've switched from the HK
power bricks to HK's ODROID power cable wired to a single 12V/20A bench
While I'm confident in the power setup up to the HK power cable itself
and have measured voltage under load up to the power distributors, I
have not measured yet on the PCB, my current voltmeter leads are a tad
too. The HK power cable seems a wee bit too much on the thin
side for my taste.
If you want I can re-test 5.0.1-300.fc29.armv7hl with one of HK's
In case you want it, I've dumped the console log at
once as generated by conserver(8), and once after running through
strings(1) as I found the control codes annoying when trying to view
in a browser ;-)
Until I lost network connectivity, I triggered the reboots by ssh-ing
lines 0020 to 0220 is the shutdown of the first systemctl reboot via ssh
lines 0221 to 1348 is the system coming up fine and going down after
I triggered again a clean reboot
lines 1349 to 2462 same again
At line 2464 the boot where I had no network starts.
At line 3174 of that boot I get the first
Timeout while waiting for setup device command
At line 3459 I trigger a clean reboot via serial
lines 3487 to 4446 the system boots but again the USB network
fails to init.
lines 4464 to end is a boot triggerd via watchdog, not helping either.
As before removing power fixed the issue.
Anything else I can test for you?
Patrick C. F. Ernzer | e-mail: pcfe(a)pcfe.net
Abstinence makes the church grow fondlers