Re: [PATCH] PCI: iproc: move quirks to driver

[Bjorn Helgaas <bjorn.helgaas@gmail.com>]

On Wed, Dec 11, 2019 at 5:40 PM Ray Jui <ray.jui@broadcom.com> wrote:
>
>
>
> On 12/11/19 2:34 PM, Bjorn Helgaas wrote:
> > On Wed, Dec 11, 2019 at 05:45:11PM +0000, Wei Liu wrote:
> >> The quirks were originally enclosed by ifdef. That made the quirks not
> >> to be applied when respective drivers were compiled as modules.
> >>
> >> Move the quirks to driver code to fix the issue.
> >>
> >> Signed-off-by: Wei Liu <wei.liu@kernel.org>
> >
> > This straddles the core and native driver boundary, so I applied it to
> > pci/misc for v5.6.  Thanks, I think this is a great solution!  It's
> > always nice when we can encapsulate device-specific things in a
> > driver.
> >
>
> Opps! I was going to review and comment and you are quick, :)
>
> I was going to say, I think it's better to keep this quirk in
> "pcie-iproc.c" instead of "pcie-iproc-platform.c".
>
> The quirk is specific to certain PCIe devices under iProc (activated
> based on device ID), but should not be tied to a specific bus
> architecture (i.e., platform vs BCMA).

I'm happy to move it; that's no problem.

> >> ---
> >>   drivers/pci/controller/pcie-iproc-platform.c | 24 ++++++++++++++++++
> >>   drivers/pci/quirks.c                         | 26 --------------------
> >>   2 files changed, 24 insertions(+), 26 deletions(-)
> >>
> >> diff --git a/drivers/pci/controller/pcie-iproc-platform.c b/drivers/pci/controller/pcie-iproc-platform.c
> >> index ff0a81a632a1..4e6f7cdd9a25 100644
> >> --- a/drivers/pci/controller/pcie-iproc-platform.c
> >> +++ b/drivers/pci/controller/pcie-iproc-platform.c
> >> @@ -19,6 +19,30 @@
> >>   #include "../pci.h"
> >>   #include "pcie-iproc.h"
> >>
> >> +static void quirk_paxc_bridge(struct pci_dev *pdev)
> >> +{
> >> +    /*
> >> +     * The PCI config space is shared with the PAXC root port and the first
> >> +     * Ethernet device.  So, we need to workaround this by telling the PCI
> >> +     * code that the bridge is not an Ethernet device.
> >> +     */
> >> +    if (pdev->hdr_type == PCI_HEADER_TYPE_BRIDGE)
> >> +            pdev->class = PCI_CLASS_BRIDGE_PCI << 8;
> >> +
> >> +    /*
> >> +     * MPSS is not being set properly (as it is currently 0).  This is
> >> +     * because that area of the PCI config space is hard coded to zero, and
> >> +     * is not modifiable by firmware.  Set this to 2 (e.g., 512 byte MPS)
> >> +     * so that the MPS can be set to the real max value.
> >> +     */
> >> +    pdev->pcie_mpss = 2;
> >> +}
> >> +DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0x16cd, quirk_paxc_bridge);
> >> +DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0x16f0, quirk_paxc_bridge);
> >> +DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0xd750, quirk_paxc_bridge);
> >> +DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0xd802, quirk_paxc_bridge);
> >> +DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0xd804, quirk_paxc_bridge);
> >> +
> >>   static const struct of_device_id iproc_pcie_of_match_table[] = {
> >>      {
> >>              .compatible = "brcm,iproc-pcie",
> >> diff --git a/drivers/pci/quirks.c b/drivers/pci/quirks.c
> >> index 4937a088d7d8..202837ed68c9 100644
> >> --- a/drivers/pci/quirks.c
> >> +++ b/drivers/pci/quirks.c
> >> @@ -2381,32 +2381,6 @@ DECLARE_PCI_FIXUP_ENABLE(PCI_VENDOR_ID_BROADCOM,
> >>                       PCI_DEVICE_ID_TIGON3_5719,
> >>                       quirk_brcm_5719_limit_mrrs);
> >>
> >> -#ifdef CONFIG_PCIE_IPROC_PLATFORM
> >> -static void quirk_paxc_bridge(struct pci_dev *pdev)
> >> -{
> >> -    /*
> >> -     * The PCI config space is shared with the PAXC root port and the first
> >> -     * Ethernet device.  So, we need to workaround this by telling the PCI
> >> -     * code that the bridge is not an Ethernet device.
> >> -     */
> >> -    if (pdev->hdr_type == PCI_HEADER_TYPE_BRIDGE)
> >> -            pdev->class = PCI_CLASS_BRIDGE_PCI << 8;
> >> -
> >> -    /*
> >> -     * MPSS is not being set properly (as it is currently 0).  This is
> >> -     * because that area of the PCI config space is hard coded to zero, and
> >> -     * is not modifiable by firmware.  Set this to 2 (e.g., 512 byte MPS)
> >> -     * so that the MPS can be set to the real max value.
> >> -     */
> >> -    pdev->pcie_mpss = 2;
> >> -}
> >> -DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0x16cd, quirk_paxc_bridge);
> >> -DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0x16f0, quirk_paxc_bridge);
> >> -DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0xd750, quirk_paxc_bridge);
> >> -DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0xd802, quirk_paxc_bridge);
> >> -DECLARE_PCI_FIXUP_EARLY(PCI_VENDOR_ID_BROADCOM, 0xd804, quirk_paxc_bridge);
> >> -#endif
> >> -
> >>   /*
> >>    * Originally in EDAC sources for i82875P: Intel tells BIOS developers to
> >>    * hide device 6 which configures the overflow device access containing the
> >> --
> >> 2.20.1
> >>