[+cc Stanimir, Niklas, Jesper]
On Mon, Aug 14, 2017 at 04:38:24PM -0500, Bjorn Helgaas wrote:
> On Thu, Aug 03, 2017 at 04:23:38PM +0800, Zhiqiang Hou wrote:
> > From: Hou Zhiqiang <Zhiqiang.Hou@xxxxxxx>
> >
> > Make the ls1021a's host_init reuse layerscape platform's common
> > host_init function.
> >
> > Signed-off-by: Hou Zhiqiang <Zhiqiang.Hou@xxxxxxx>
> > ---
> > V2:
> > - Removed the disable outbound windows code and the remove duplicate class code
> > fixup code from this patch.
> >
> > drivers/pci/dwc/pci-layerscape.c | 54 ++++++++++++++++++++--------------------
> > 1 file changed, 27 insertions(+), 27 deletions(-)
> >
> > diff --git a/drivers/pci/dwc/pci-layerscape.c b/drivers/pci/dwc/pci-layerscape.c
> > index 09056a6..3533a8c 100644
> > --- a/drivers/pci/dwc/pci-layerscape.c
> > +++ b/drivers/pci/dwc/pci-layerscape.c
> > @@ -107,33 +107,6 @@ static int ls1021_pcie_link_up(struct dw_pcie *pci)
> > return 1;
> > }
> >
> > -static void ls1021_pcie_host_init(struct pcie_port *pp)
> > -{
> > - struct dw_pcie *pci = to_dw_pcie_from_pp(pp);
> > - struct ls_pcie *pcie = to_ls_pcie(pci);
> > - struct device *dev = pci->dev;
> > - u32 index[2];
> > -
> > - pcie->scfg = syscon_regmap_lookup_by_phandle(dev->of_node,
> > - "fsl,pcie-scfg");
> > - if (IS_ERR(pcie->scfg)) {
> > - dev_err(dev, "No syscfg phandle specified\n");
> > - pcie->scfg = NULL;
> > - return;
> > - }
> > -
> > - if (of_property_read_u32_array(dev->of_node,
> > - "fsl,pcie-scfg", index, 2)) {
> > - pcie->scfg = NULL;
> > - return;
> > - }
> > - pcie->index = index[1];
> > -
> > - dw_pcie_setup_rc(pp);
> > -
> > - ls_pcie_drop_msg_tlp(pcie);
> > -}
> > -
> > static int ls_pcie_link_up(struct dw_pcie *pci)
> > {
> > struct ls_pcie *pcie = to_ls_pcie(pci);
> > @@ -160,6 +133,33 @@ static void ls_pcie_host_init(struct pcie_port *pp)
> > dw_pcie_dbi_ro_wr_dis(pci);
> >
> > ls_pcie_drop_msg_tlp(pcie);
> > +
> > + dw_pcie_setup_rc(pp);
> > +}
> > +
> > +static void ls1021_pcie_host_init(struct pcie_port *pp)
> > +{
> > + struct dw_pcie *pci = to_dw_pcie_from_pp(pp);
> > + struct ls_pcie *pcie = to_ls_pcie(pci);
> > + struct device *dev = pci->dev;
> > + u32 index[2];
> > +
> > + pcie->scfg = syscon_regmap_lookup_by_phandle(dev->of_node,
> > + "fsl,pcie-scfg");
> > + if (IS_ERR(pcie->scfg)) {
> > + dev_err(dev, "No syscfg phandle specified\n");
> > + pcie->scfg = NULL;
> > + return;
> > + }
> > +
> > + if (of_property_read_u32_array(dev->of_node,
> > + "fsl,pcie-scfg", index, 2)) {
> > + pcie->scfg = NULL;
> > + return;
> > + }
> > + pcie->index = index[1];
> > +
> > + ls_pcie_host_init(pp);
>
> The changelog suggests that this is a simple refactoring that doesn't
> fix anything.
>
> But because ls1021_pcie_host_init() now calls ls_pcie_host_init(), it
> now calls:
>
> dw_pcie_dbi_ro_wr_en(pci);
> ls_pcie_fix_class(pcie);
> ls_pcie_clear_multifunction(pcie);
> dw_pcie_dbi_ro_wr_dis(pci);
>
> when it did not call them before.
>
> It's fine with me if you want to do that, but it should be mentioned
> in the changelog.
The order of this series is screwed up, which makes it hard to follow.
5192ec7b24dd ("PCI: layerscape: Add support for LS1043a and LS2080a")
added ls_pcie_host_init(), which didn't call dw_pcie_setup_rc(). I
think that was a bug, and you should fix that first.
Then:
- Rebase on pci/host-designware because it conflicts with patches
I've already applied
(https://git.kernel.org/pub/scm/linux/kernel/git/helgaas/pci.git/log/?h=pci/host-designware)
- Make ls_pcie_host_init() call dw_pcie_setup_rc() (fixes
5192ec7b24dd)
- Refactor ls1021_pcie_host_init() to make it call
ls_pcie_host_init() (this will now be no functional change)
- Disable outbound ATUs in ls_pcie_host_init() (bugfix for all
layerscape devices)
- Add DBI write permission accessors and use them in layerscape.
This will make it obvious that PCIE_MISC_CONTROL_1_OFF is moving
from layerscape to designware and that we're actually using these
new interfaces (this will be no functional change).
- Enable write permission in dw_pcie_setup_rc() and remove class
code update from ls_pcie_host_init() This will make it obvious
that fixing dw_pcie_setup_rc() makes ls_pcie_fix_class() obsolete.
Does this dw_pcie_setup_rc() fix mean we can also get rid of the
device class check in qcom_pcie_rd_own_conf()?
I suspect it also means we can drop the write enable in
artpec6_pcie_establish_link().
> > }
> >
> > static int ls_pcie_msi_host_init(struct pcie_port *pp,
> > --
> > 2.1.0.27.g96db324
> >
