Xen project Mailing List

##- Please type your reply above this line -##

You are registered as a CC on this support request (809). Reply to this email to add a comment to the request.

Jason V. (SoftIron, Inc.)

Jun 18, 11:26 PDT

This ticket has been closed in accordance with our policy for inactive tickets. Should you require further support, please initiate a new ticket via support@xxxxxxxxxxxx

Julien Grall

July 28, 2017, 02:28 PDT

> ##- Please type your reply above this line -##
>
> Your request (809) has been updated. To add additional comments, reply
> to this email.
>
> Alan Ott
>
> *Alan Ott* (SoftIron, Inc.)
>
> Jul 27, 12:41 PDT

Hi Alan,

> Hi Julien, Sorry this ticket fell through the cracks for so long. I was
> just assigned it today.
>
> My understanding is that there is a bug in the closed-source AMD SCP
> firmware which prevents reliable reboot. I do not believe this to have
> anything to do with EDK2. Feel free to try it out, but the problem as I
> know it is in the AMD SCP firmware. I could always be wrong, but unless
> this was fixed very recently, edk2 still has the issue.
>
> We do not have an updated version of the AMI firmware.
>
> Have you tried the edk2 firmware on that box, and does it work?

At the moment, the two softiron we have are dead (see ticket 837) so we
cannot do testing.

However, is there any plan for Sotfiron to fix officially this bug? This
is a rather annoying one as we have random failure in testing and losing
nearly a day every time...

Cheers,

--
Julien Grall

Alan Ott (SoftIron, Inc.)

July 27, 2017, 12:41 PDT

Hi Julien, Sorry this ticket fell through the cracks for so long. I was just assigned it today.

My understanding is that there is a bug in the closed-source AMD SCP firmware which prevents reliable reboot. I do not believe this to have anything to do with EDK2. Feel free to try it out, but the problem as I know it is in the AMD SCP firmware. I could always be wrong, but unless this was fixed very recently, edk2 still has the issue.

We do not have an updated version of the AMI firmware.

Have you tried the edk2 firmware on that box, and does it work?

Also, feel free to find me on #linaro-enterprise as alan_o.

Alan.

Julien Grall

June 23, 2017, 02:32 PDT

Hello,

In reference to my previous e-mail on the 15th June, I would like to
know if you have any update on the reboot failure?

Regards,

Julien Grall

June 23, 2017, 02:32 PDT

Hello,

In reference to my previous e-mail on the 15th June, I would like to
know if you have any update on the reboot failure?

Regards,

Ian Jackson

June 15, 2017, 04:56 PDT

Julien Grall writes ("Random reboot failure on Softiron overdrive 3000"):
> The Xen Project recently acquired a couple of overdrive 3000 for
> automatic testing. We have a couple of issues with the current
> firmware used (BL3-1: Built : 16:00:55, Apr 8 2016):
>
> 1) Failure to reboot (see the only we have on the serial console below).
> I have been told this is due to PCIe link training failure.
...
> Whilst the latter is not critical, the former prevents us to get reliable
> testing on the Overdrive 3000. I spoke with UEFI developers, they told me the
> two bugs should be fixed in upstream EDK2. Would it be possible that Softiron
> provides an updated firmware?

Thanks, Julien.

FYI, we have budget set aside for purchasing a further pair of
Softiron ARM64 machines, but this is of course dependent on us getting
the existing ones to work reliably.

Thanks,
Ian.

Julien Grall

June 15, 2017, 04:29 PDT

Hello,

The Xen Project recently acquired a couple of overdrive 3000 for
automatic testing. We have a couple of issues with the current
firmware used (BL3-1: Built : 16:00:55, Apr 8 2016):

1) Failure to reboot (see the only we have on the serial console below).
I have been told this is due to PCIe link training failure.

Jun 12 12:47:32.084419 [ 182.968398] reboot: Restarting system
Jun 12 12:47:32.092383 INFO: PSCI Power Domain Map:
Jun 12 12:47:32.092412 INFO: Domain Node : Level 1, parent_node -1, State ON (0x0)
Jun 12 12:47:32.100393 INFO: Domain Node : Level 1, parent_node -1, State ON (0x0)
Jun 12 12:47:32.100427 INFO: Domain Node : Level 1, parent_node -1, State ON (0x0)
Jun 12 12:47:32.108409 INFO: Domain Node : Level 1, parent_node -1, State ON (0x0)
Jun 12 12:47:32.116391 INFO: CPU Node : MPID 0x0, parent_node 0, State ON (0x0)
Jun 12 12:47:32.116424 INFO: CPU Node : MPID 0x1, parent_node 0, State ON (0x0)
Jun 12 12:47:32.124413 INFO: CPU Node : MPID 0x100, parent_node 1, State ON (0x0)
Jun 12 12:47:32.132395 INFO: CPU Node : MPID 0x101, parent_node 1, State ON (0x0)
Jun 12 12:47:32.140389 INFO: CPU Node : MPID 0x200, parent_node 2, State ON (0x0)
Jun 12 12:47:32.140422 INFO: CPU Node : MPID 0x201, parent_node 2, State ON (0x0)
Jun 12 12:47:32.148398 INFO: CPU Node : MPID 0x300, parent_node 3, State ON (0x0)
Jun 12 12:47:32.156385 INFO: CPU Node : MPID 0x301, parent_node 3, State ON (0x0)

2) Linux is complaining about with:
efi: [Firmware Bug]: IRQ flags corrupted (0x00000040=>0x00000000) by EFI get_next_variable

Whilst the latter is not critical, the former prevents us to get reliable
testing on the Overdrive 3000. I spoke with UEFI developers, they told me the
two bugs should be fixed in upstream EDK2. Would it be possible that Softiron
provides an updated firmware?

Many thanks,

--
Julien Grall
IMPORTANT NOTICE: The contents of this email and any attachments are confidential and may also be privileged. If you are not the intended recipient, please notify the sender immediately and do not disclose the contents to any other person, use it for any purpose, or store or copy the information in any medium. Thank you.

This email is a service from SoftIron, Inc.. Delivered by Zendesk

[Wg-test-framework] Random reboot failure on Softiron overdrive 3000