Re: [edk2-platforms: PATCH 08/14] Marvell/Armada7k8k: Enable PCIE support

["Ard Biesheuvel" <ard.biesheuvel@linaro.org>]

On Thu, 16 May 2019 at 16:22, Marcin Wojtas <mw@semihalf.com> wrote:
>
> Hi Ard,
>
> czw., 16 maj 2019 o 16:16 Ard Biesheuvel <ard.biesheuvel@linaro.org> napisał(a):
> >
> > On Thu, 9 May 2019 at 11:54, Marcin Wojtas <mw@semihalf.com> wrote:
> > >
> > > Wire up the platform libraries to the generic drivers so that we can use
> > > PCI devices and UEFI, and leave the controller initialized so that the
> > > OS can boot it using a generic driver of its own.
> > >
> > > Contributed-under: TianoCore Contribution Agreement 1.1
> > > Signed-off-by: Marcin Wojtas <mw@semihalf.com>
> > > ---
> > >  Silicon/Marvell/Armada7k8k/Armada7k8k.dsc.inc | 17 +++++++++++++++--
> > >  Silicon/Marvell/Armada7k8k/Armada7k8k.fdf     |  5 +++++
> > >  2 files changed, 20 insertions(+), 2 deletions(-)
> > >
> > > diff --git a/Silicon/Marvell/Armada7k8k/Armada7k8k.dsc.inc b/Silicon/Marvell/Armada7k8k/Armada7k8k.dsc.inc
> > > index 545b369..f78a76b 100644
> > > --- a/Silicon/Marvell/Armada7k8k/Armada7k8k.dsc.inc
> > > +++ b/Silicon/Marvell/Armada7k8k/Armada7k8k.dsc.inc
> > > @@ -70,8 +70,10 @@
> > >    IoLib|MdePkg/Library/BaseIoLibIntrinsic/BaseIoLibIntrinsic.inf
> > >    UefiDecompressLib|MdePkg/Library/BaseUefiDecompressLib/BaseUefiDecompressLib.inf
> > >    CpuLib|MdePkg/Library/BaseCpuLib/BaseCpuLib.inf
> > > -  PciLib|MdePkg/Library/BasePciLibCf8/BasePciLibCf8.inf
> > > -  PciCf8Lib|MdePkg/Library/BasePciCf8Lib/BasePciCf8Lib.inf
> > > +  PciLib|MdePkg/Library/BasePciLibPciExpress/BasePciLibPciExpress.inf
> > > +  PciHostBridgeLib|Silicon/Marvell/Armada7k8k/Library/Armada7k8kPciHostBridgeLib/PciHostBridgeLib.inf
> > > +  PciSegmentLib|MdePkg/Library/BasePciSegmentLibPci/BasePciSegmentLibPci.inf
> > > +  PciExpressLib|Silicon/Marvell/Armada7k8k/Library/Armada7k8kPciExpressLib/PciExpressLib.inf
> > >
> > >  # Basic UEFI services libraries
> > >    UefiLib|MdePkg/Library/UefiLib/UefiLib.inf
> > > @@ -407,6 +409,12 @@
> > >    gEfiMdeModulePkgTokenSpaceGuid.PcdFlashNvStorageFtwWorkingSize|0x00010000
> > >    gEfiMdeModulePkgTokenSpaceGuid.PcdFlashNvStorageFtwSpareSize|0x00010000
> > >
> > > +  # PCIE
> > > +  gEfiMdePkgTokenSpaceGuid.PcdPciExpressBaseAddress|0xE0000000
> > > +
> > > +  # SoC Configuration Space
> > > +  gMarvellTokenSpaceGuid.PcdConfigSpaceBaseAddress|0xE0000000
> > > +
> > >  !if $(CAPSULE_ENABLE)
> > >  [PcdsDynamicExDefault.common.DEFAULT]
> > >    gEfiSignedCapsulePkgTokenSpaceGuid.PcdEdkiiSystemFirmwareImageDescriptor|{0x0}|VOID*|0x100
> > > @@ -520,6 +528,11 @@
> > >    MdeModulePkg/Bus/Pci/SdMmcPciHcDxe/SdMmcPciHcDxe.inf
> > >    Silicon/Marvell/Drivers/SdMmc/XenonDxe/XenonDxe.inf
> > >
> > > +  # PCI
> > > +  ArmPkg/Drivers/ArmPciCpuIo2Dxe/ArmPciCpuIo2Dxe.inf
> >
> > This driver requires gArmTokenSpaceGuid.PcdPciIoTranslation to be set
> > to a sane value. Are you sure this is the case for your platforms?
> >
>
> Do you mean the IO space for the controller? If yes, I'll set the PCD
> to according value I use in board description. I don't have an old
> enough endpoint that requires IO space to test :)
>

Yes, it is basically the MMIO address of the I/O space.

Leif kindly gave me a PCIe serial port controller that I use
especially for testing these things at plugfests.