Le ven. 29 mars 2019 à 12:22, Thierry Reding
<thierry.reding@xxxxxxxxx> a écrit :
>
> On Mon, Jul 30, 2018 at 11:42:13AM +0200, Nicolas Chauvet wrote:
> > This patch disable the use of nv_msi_ht_cap_quirk_leaf quirk on non-x86
> > NVIDIA devices such as Tegra
> >
> > This fixes the following output:
> > "pci 0000:00:01.0: nv_msi_ht_cap_quirk didn't locate host bridge"
> > as experienced on a Trimslice device with pci host bridge enabled
> >
> > Signed-off-by: Nicolas Chauvet <kwizart@xxxxxxxxx>
> > ---
> > drivers/pci/quirks.c | 3 +++
> > 1 file changed, 3 insertions(+)
> >
> > diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c
> > index f439de848658..9a2bba5d84c7 100644
> > --- a/drivers/pci/quirks.c
> > +++ b/drivers/pci/quirks.c
> > @@ -2777,8 +2777,11 @@ static void nv_msi_ht_cap_quirk_leaf(struct pci_dev *dev)
> > {
> > return __nv_msi_ht_cap_quirk(dev, 0);
> > }
> > +/* This quirk is only relevant on x86 NVIDIA pci host bridges */
> > +#if IS_ENABLED(CONFIG_X86)
> > DECLARE_PCI_FIXUP_FINAL(PCI_VENDOR_ID_NVIDIA, PCI_ANY_ID, nv_msi_ht_cap_quirk_leaf);
> > DECLARE_PCI_FIXUP_RESUME_EARLY(PCI_VENDOR_ID_NVIDIA, PCI_ANY_ID, nv_msi_ht_cap_quirk_leaf);
> > +#endif
>
> I think we probably want to turn this around. HyperTransport is not
> specific to x86, so I think we probably want to make this:
>
> #if !IS_ENABLED(CONFIG_ARM) && !IS_ENABLED(CONFIG_ARM64)
>
> to exclude it only on 32-bit and 64-bit ARM builds.
I've assumed there is no NVIDIA (PCI/MSI/HT) related chipset elsewhere than x86
And if there are any in the future (either for ppc64, riscv or else)
that it may not require the quirk...
Is this false assumption ? If the case, I'm sending a v3 with the
requested change.
Thx for the review.
--
-
Nicolas (kwizart)
