* correct way for handling PCIe in endpoint mode
@ 2009-02-10 12:39 Scott Coulter
2009-02-10 16:22 ` Ira Snyder
0 siblings, 1 reply; 2+ messages in thread
From: Scott Coulter @ 2009-02-10 12:39 UTC (permalink / raw)
To: linuxppc-dev
Hi all,
Maybe a silly questions, but what is the suggested way for handling a
processor which has a PCIe interface, but the interface is not
configured to be the root complex (endpoint only)?
Should the PCIe interface appear in the dts? If so, with what
parameters?
Should the kernel config contain CONFIG_PCI, etc. ?
Thanks,
Scott
___________________________________________________________________
Scott N. Coulter
Senior Software Engineer
=20
Cyclone Microsystems =20
370 James Street Phone: 203.786.5536 ext. 118
New Haven, CT 06513-3051 Email: scott.coulter@cyclone.com
U.S.A. Web: http://www.cyclone.com
___________________________________________________________________
^ permalink raw reply [flat|nested] 2+ messages in thread
* Re: correct way for handling PCIe in endpoint mode
2009-02-10 12:39 correct way for handling PCIe in endpoint mode Scott Coulter
@ 2009-02-10 16:22 ` Ira Snyder
0 siblings, 0 replies; 2+ messages in thread
From: Ira Snyder @ 2009-02-10 16:22 UTC (permalink / raw)
To: Scott Coulter; +Cc: linuxppc-dev
On Tue, Feb 10, 2009 at 07:39:47AM -0500, Scott Coulter wrote:
>
> Hi all,
>
> Maybe a silly questions, but what is the suggested way for handling a
> processor which has a PCIe interface, but the interface is not
> configured to be the root complex (endpoint only)?
>
> Should the PCIe interface appear in the dts? If so, with what
> parameters?
>
> Should the kernel config contain CONFIG_PCI, etc. ?
>
I'm in a similar situation, except I'm using PCI instead of PCIe. I've
tried both removing the PCI interface from the dts, and disabling
CONFIG_PCI. Both work fine.
You probably don't want to leave the PCI interface in the dts and also
leave CONFIG_PCI enabled. I ran into some really strange issues this
way... (Linux ran through the quirks and tried to re-initialize a
network card, even though the driver was disabled, etc.)
I think the suggested way is to remove the PCI interface from the dts,
that way you can potentially share a kernel between similar boards with
and without PCI root interfaces, just by changing the dts.
Ira
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2009-02-10 16:22 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2009-02-10 12:39 correct way for handling PCIe in endpoint mode Scott Coulter
2009-02-10 16:22 ` Ira Snyder
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox