Atom E3815 processor - reporting incorrect PCI device id

[Mike Hibbett]

Hi all,

I have two B3 stepping E3815 processors, one is reporting a PCI device id of 0x0f00 ( as expected ) but another is reporting 0x0000. I can work round that in the BAYTRAIL_FSP CoreBoot build but cannot install Linux, and I'm guessing that having a wonky device id may be involved. Has anyone else seen this and know the work around? I can't see anything about it on the Intel IBL.

Cheers,

Mike.

[Sean McNeil]

Hi Mike,

This is due to a bad or missing IFD. You need to have the correct 
information in the bootStrap settings.

Cheers,
Sean

On 06/17/2014 10:27 PM, Mike Hibbett wrote:
> Hi all,
>
> I have two B3 stepping E3815 processors, one is reporting a PCI device id of 0x0f00 ( as expected ) but another is reporting 0x0000. I can work round that in the BAYTRAIL_FSP CoreBoot build but cannot install Linux, and I'm guessing that having a wonky device id may be involved. Has anyone else seen this and know the work around? I can't see anything about it on the Intel IBL.
>
> Cheers,
>
> Mike.

[Mike Hibbett]

Hi Sean,

I would have thought so too, but the devices are using the same IFD.

Do you know what the ifd parameter is that affects this?

Cheers,

Mike

Sent with AquaMail for Android
http://www.aqua-mail.com

On 18 June 2014 01:08:25 Sean McNeil  wrote:
> Hi Mike,
>
> This is due to a bad or missing IFD. You need to have the correct information in the bootStrap settings.
>
> Cheers,
> Sean
>
> On 06/17/2014 10:27 PM, Mike Hibbett wrote:
> > Hi all,
> >
> > I have two B3 stepping E3815 processors, one is reporting a PCI device id of 0x0f00 ( as expected ) but another is reporting 0x0000. I can work round that in the BAYTRAIL_FSP CoreBoot build but cannot install Linux, and I'm guessing that having a wonky device id may be involved. Has anyone else seen this and know the work around? I can't see anything about it on the Intel IBL.
> >
> > Cheers,
> >
> > Mike.

[Sean McNeil]

Hi Mike,

I believe it is one of the Reserved bits at 85:84 or 82, but not 
certain. You can check ILB 514482.

Are the boards identical?
Are the sizes of flash identical?
Any hardware straps connected?

or are you replacing the CPU in the same board? If unique boards, I 
would dump back the flash and verify.

Sean

On 06/18/2014 12:22 PM, Mike Hibbett wrote:
> Hi Sean,
>
> I would have thought so too, but the devices are using the same IFD.
>
> Do you know what the ifd parameter is that affects this?
>
> Cheers,
>
> Mike
>
> Sent with AquaMail for Android
> http://www.aqua-mail.com
>
> On 18 June 2014 01:08:25 Sean McNeil  wrote:
> > Hi Mike,
> >
> > This is due to a bad or missing IFD. You need to have the correct information in the bootStrap settings.
> >
> > Cheers,
> > Sean
> >
> > On 06/17/2014 10:27 PM, Mike Hibbett wrote:
> > > Hi all,
> > >
> > > I have two B3 stepping E3815 processors, one is reporting a PCI device id of 0x0f00 ( as expected ) but another is reporting 0x0000. I can work round that in the BAYTRAIL_FSP CoreBoot build but cannot install Linux, and I'm guessing that having a wonky device id may be involved. Has anyone else seen this and know the work around? I can't see anything about it on the Intel IBL.
> > >
> > > Cheers,
> > >
> > > Mike.

[Mike Hibbett]

Thanks Sean. I was looking at that document last night, didn't see anything obvious, I suspect it's a hard strap related issue. I'll double check.

Thanks,

Mike

-----Original Message-----
From: Sean McNeil [mailto:[e-mail]] 
Sent: 18 June 2014 06:58
To: Mike Hibbett; [e-mail]
Subject: Re: [coreboot] Atom E3815 processor - reporting incorrect PCI device id

Hi Mike,

I believe it is one of the Reserved bits at 85:84 or 82, but not certain. You can check ILB 514482.

Are the boards identical?
Are the sizes of flash identical?
Any hardware straps connected?

or are you replacing the CPU in the same board? If unique boards, I would dump back the flash and verify.

Sean

On 06/18/2014 12:22 PM, Mike Hibbett wrote:
> Hi Sean,
>
> I would have thought so too, but the devices are using the same IFD.
>
> Do you know what the ifd parameter is that affects this?
>
> Cheers,
>
> Mike
>
> Sent with AquaMail for Android
> http://www.aqua-mail.com
>
> On 18 June 2014 01:08:25 Sean McNeil  wrote:
> > Hi Mike,
> >
> > This is due to a bad or missing IFD. You need to have the correct information in the bootStrap settings.
> >
> > Cheers,
> > Sean
> >
> > On 06/17/2014 10:27 PM, Mike Hibbett wrote:
> > > Hi all,
> > >
> > > I have two B3 stepping E3815 processors, one is reporting a PCI device id of 0x0f00 ( as expected ) but another is reporting 0x0000. I can work round that in the BAYTRAIL_FSP CoreBoot build but cannot install Linux, and I'm guessing that having a wonky device id may be involved. Has anyone else seen this and know the work around? I can't see anything about it on the Intel IBL.
> > >
> > > Cheers,
> > >
> > > Mike.