Re: [PATCH] PCI: mediatek: Remove the usage of virt_to_phys

["manivannan.sadhasivam@linaro.org" <manivannan.sadhasivam@linaro.org>]

On Mon, Jan 27, 2025 at 06:41:50PM -0600, Bjorn Helgaas wrote:
> On Thu, Jan 23, 2025 at 08:11:19AM +0000, Jianjun Wang (王建军) wrote:
> > On Wed, 2025-01-15 at 23:01 +0530, Manivannan Sadhasivam wrote:
> > > On Tue, Jan 07, 2025 at 01:20:58PM +0800, Jianjun Wang wrote:
> > > > Remove the usage of virt_to_phys, as it will cause sparse warnings
> > > > when
> > > > building on some platforms.
> > > 
> > > Strange. What are those warnings and platforms?
> > 
> > There are some warning messages when building tests with different
> > configs on different platforms (e.g., allmodconfig.arm,
> > allmodconfig.i386, allmodconfig.mips, etc.):
> > 
> > pcie-mediatek.c:399:40: sparse: warning: incorrect type in argument 1
> > (different address spaces)
> > pcie-mediatek.c:399:40: sparse:    expected void const volatile *x
> > pcie-mediatek.c:399:40: sparse:    got void [noderef] __iomem *
> > pcie-mediatek.c:515:44: sparse: warning: incorrect type in argument 1
> > (different address spaces)
> > pcie-mediatek.c:515:44: sparse:    expected void const volatile *x
> > pcie-mediatek.c:515:44: sparse:    got void [noderef] __iomem *
> >
> > > > Signed-off-by: Jianjun Wang <jianjun.wang@mediatek.com>
> > > > ---
> > > >  drivers/pci/controller/pcie-mediatek.c | 19 ++++++++++++-------
> > > >  1 file changed, 12 insertions(+), 7 deletions(-)
> > > > 
> > > > diff --git a/drivers/pci/controller/pcie-mediatek.c
> > > > b/drivers/pci/controller/pcie-mediatek.c
> > > > index 3bcfc4e58ba2..dc1e5fd6c7aa 100644
> > > > --- a/drivers/pci/controller/pcie-mediatek.c
> > > > +++ b/drivers/pci/controller/pcie-mediatek.c
> > > > @@ -178,6 +178,7 @@ struct mtk_pcie_soc {
> > > >   * @phy: pointer to PHY control block
> > > >   * @slot: port slot
> > > >   * @irq: GIC irq
> > > > + * @msg_addr: MSI message address
> > > >   * @irq_domain: legacy INTx IRQ domain
> > > >   * @inner_domain: inner IRQ domain
> > > >   * @msi_domain: MSI IRQ domain
> > > > @@ -198,6 +199,7 @@ struct mtk_pcie_port {
> > > >       struct phy *phy;
> > > >       u32 slot;
> > > >       int irq;
> > > > +     phys_addr_t msg_addr;
> > > >       struct irq_domain *irq_domain;
> > > >       struct irq_domain *inner_domain;
> > > >       struct irq_domain *msi_domain;
> > > > @@ -393,12 +395,10 @@ static struct pci_ops mtk_pcie_ops_v2 = {
> > > >  static void mtk_compose_msi_msg(struct irq_data *data, struct
> > > > msi_msg *msg)
> > > >  {
> > > >       struct mtk_pcie_port *port =
> > > > irq_data_get_irq_chip_data(data);
> > > > -     phys_addr_t addr;
> > > > 
> > > >       /* MT2712/MT7622 only support 32-bit MSI addresses */
> > > > -     addr = virt_to_phys(port->base + PCIE_MSI_VECTOR);
> 
> Thanks for working on this!  We should definitely try to get rid of
> virt_to_phys().
> 
> > > >       msg->address_hi = 0;
> > > > -     msg->address_lo = lower_32_bits(addr);
> > > > +     msg->address_lo = lower_32_bits(port->msg_addr);
> > > > 
> > > >       msg->data = data->hwirq;
> > > > 
> > > > @@ -510,10 +510,8 @@ static int
> > > > mtk_pcie_allocate_msi_domains(struct mtk_pcie_port *port)
> > > >  static void mtk_pcie_enable_msi(struct mtk_pcie_port *port)
> > > >  {
> > > >       u32 val;
> > > > -     phys_addr_t msg_addr;
> > > > 
> > > > -     msg_addr = virt_to_phys(port->base + PCIE_MSI_VECTOR);
> > > > -     val = lower_32_bits(msg_addr);
> > > > +     val = lower_32_bits(port->msg_addr);
> > > >       writel(val, port->base + PCIE_IMSI_ADDR);
> > > > 
> > > >       val = readl(port->base + PCIE_INT_MASK);
> > > > @@ -913,6 +911,7 @@ static int mtk_pcie_parse_port(struct mtk_pcie
> > > > *pcie,
> > > >       struct mtk_pcie_port *port;
> > > >       struct device *dev = pcie->dev;
> > > >       struct platform_device *pdev = to_platform_device(dev);
> > > > +     struct resource *regs;
> > > >       char name[10];
> > > >       int err;
> > > > 
> > > > @@ -921,12 +920,18 @@ static int mtk_pcie_parse_port(struct
> > > > mtk_pcie *pcie,
> > > >               return -ENOMEM;
> > > > 
> > > >       snprintf(name, sizeof(name), "port%d", slot);
> > > > -     port->base = devm_platform_ioremap_resource_byname(pdev,
> > > > name);
> > > > +     regs = platform_get_resource_byname(pdev, IORESOURCE_MEM,
> > > > name);
> > > > +     if (!regs)
> > > > +             return -EINVAL;
> > > > +
> > > > +     port->base = devm_ioremap_resource(dev, regs);
> > > >       if (IS_ERR(port->base)) {
> > > >               dev_err(dev, "failed to map port%d base\n", slot);
> > > >               return PTR_ERR(port->base);
> > > >       }
> > > > 
> > > > +     port->msg_addr = regs->start + PCIE_MSI_VECTOR;
> 
> I think this still assumes that a CPU physical address (regs->start)
> is the same as the PCI bus address used for MSI, so this doesn't seem
> like the right solution to me.
> 
> Apparently they happen to be the same on this platform because (I
> assume) MSIs actually do work, but it's not a good pattern for drivers
> to copy.  I think what we really need is a dma_addr_t, and I think
> there are one or two PCI controller drivers that do that.
> 

I don't see why we would need 'dma_addr_t' here. The MSI address is a static
physical address on this platform and that is not a DMA address. Other drivers
can *only* copy this pattern if they also have the physical address allocated
for MSI.

- Mani

-- 
மணிவண்ணன் சதாசிவம்