* Linux PowerPC, PPC4xx
@ 2019-10-16 13:09 Carlo Pisani
0 siblings, 0 replies; 3+ messages in thread
From: Carlo Pisani @ 2019-10-16 13:09 UTC (permalink / raw)
To: linuxppc-dev
hi
I am a student, I represent a group of friends running a couple of
opensource projects(1), and we are lost with this(2) problem.
I wrote here(2) a couple of years ago, we are still working with kernel
4.11.0 and there is broken support for initializing the PCI.
The PCI initialization of the PPC405GP seems wrong and every
kernel >= 2.6.26 is not able to correctly address the PDC20265
an interesting note is:
kernel 2.6.26 can be compiled with arch=ppc and arch=powerpc
when compiled with arch=ppc the promise PDC20265 chip is correctly
managed; while when compiled with arch=powerpc the PDC20265 is not
correctly managed
any idea? advice? help? suggestion?
a good place to discuss it?
thanks
Carlo
(1) http://www.downthebunker.com/reloaded/space/viewforum.php?f=37
(2) https://bugzilla.kernel.org/show_bug.cgi?id=195933
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: Linux PowerPC, PPC4xx
2019-10-16 13:19 Carlo Pisani
@ 2019-10-17 2:11 ` Benjamin Herrenschmidt
0 siblings, 0 replies; 3+ messages in thread
From: Benjamin Herrenschmidt @ 2019-10-17 2:11 UTC (permalink / raw)
To: Carlo Pisani, paulus, mpe, linuxppc-dev, alistair, mporter, sfr,
michael, david, agust, scottwood, patchwork.ozlabs
On Wed, 2019-10-16 at 15:19 +0200, Carlo Pisani wrote:
> hi
> I am a student, I represent a group of friends running a couple of
> opensource projects(1), and we are lost with this(2) problem.
>
> I wrote here(2) a couple of years ago, we are still working with
> kernel
> 4.11.0 and there is broken support for initializing the PCI.
>
> The PCI initialization of the PPC405GP seems wrong and every
> kernel >= 2.6.26 is not able to correctly address the PDC20265
>
> an interesting note is:
> kernel 2.6.26 can be compiled with arch=ppc and arch=powerpc
>
> when compiled with arch=ppc the promise PDC20265 chip is correctly
> managed; while when compiled with arch=powerpc the PDC20265 is not
> correctly managed
>
> any idea? advice? help? suggestion?
> a good place to discuss it?
On powermac we have a quirk to force these controllers into native
mode. You can try that.
Look for pmac_pci_fixup_pciata(). You could copy that to
powerpc/pci_32.c along with
DECLARE_PCI_FIXUP_EARLY(PCI_ANY_ID, PCI_ANY_ID, pmac_pci_fixup_pciata);
(maybe rename pmac to ppc) and remove the line that checks for
machine_is(powermac)
Cheers,
Ben.
^ permalink raw reply [flat|nested] 3+ messages in thread
* Linux PowerPC, PPC4xx
@ 2019-10-16 13:19 Carlo Pisani
2019-10-17 2:11 ` Benjamin Herrenschmidt
0 siblings, 1 reply; 3+ messages in thread
From: Carlo Pisani @ 2019-10-16 13:19 UTC (permalink / raw)
To: benh, paulus, mpe, linuxppc-dev, alistair, mporter, sfr, michael,
david, agust, scottwood, patchwork.ozlabs
hi
I am a student, I represent a group of friends running a couple of
opensource projects(1), and we are lost with this(2) problem.
I wrote here(2) a couple of years ago, we are still working with kernel
4.11.0 and there is broken support for initializing the PCI.
The PCI initialization of the PPC405GP seems wrong and every
kernel >= 2.6.26 is not able to correctly address the PDC20265
an interesting note is:
kernel 2.6.26 can be compiled with arch=ppc and arch=powerpc
when compiled with arch=ppc the promise PDC20265 chip is correctly
managed; while when compiled with arch=powerpc the PDC20265 is not
correctly managed
any idea? advice? help? suggestion?
a good place to discuss it?
thanks
Carlo
(1) http://www.downthebunker.com/reloaded/space/viewforum.php?f=37
(2) https://bugzilla.kernel.org/show_bug.cgi?id=195933
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2019-10-17 2:13 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2019-10-16 13:09 Linux PowerPC, PPC4xx Carlo Pisani
2019-10-16 13:19 Carlo Pisani
2019-10-17 2:11 ` Benjamin Herrenschmidt
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).