On Thursday 23 April 2020 13:44:13 Bjorn Helgaas wrote: > [+cc Rob] > > On Tue, Apr 21, 2020 at 01:16:57PM +0200, Marek Behún wrote: > > From: Pali Rohár <pali@xxxxxxxxxx> > > > > Register PCIE_CORE_CMD_STATUS_REG is applicable only when the controller > > is configured for Endpoint mode, which is not the case for the current > > version of this driver. > > > > Add a FIXME comment, since this needs to be explained, removed or fixed. > > If it's not applicable, why not just remove it? Problem is hat if I remove this part of code, ath10k cards stops working. So for some unknown reasons setting this "not applicable" register is required. I have in-progress V3 version of this patch series with updated comment and commit message of this patch (include this observation). > > Signed-off-by: Pali Rohár <pali@xxxxxxxxxx> > > --- > > drivers/pci/controller/pci-aardvark.c | 6 ++++++ > > 1 file changed, 6 insertions(+) > > > > diff --git a/drivers/pci/controller/pci-aardvark.c b/drivers/pci/controller/pci-aardvark.c > > index e2d18094d8ca..e893d7d8859f 100644 > > --- a/drivers/pci/controller/pci-aardvark.c > > +++ b/drivers/pci/controller/pci-aardvark.c > > @@ -429,6 +429,12 @@ static void advk_pcie_setup_hw(struct advk_pcie *pcie) > > > > advk_pcie_train_link(pcie); > > > > + /* > > + * FIXME: Following code which access PCIE_CORE_CMD_STATUS_REG register > > + * is suspicious. This register is applicable only when the PCI > > + * controller is configured for Endpoint mode. And not when it > > + * is configured for Root Complex. > > + */ > > reg = advk_readl(pcie, PCIE_CORE_CMD_STATUS_REG); > > reg |= PCIE_CORE_CMD_MEM_ACCESS_EN | > > PCIE_CORE_CMD_IO_ACCESS_EN | > > -- > > 2.24.1 > >
