RE: [PATCH v8 2/2] PCI: xilinx-cpm: Add Versal CPM Root Port driver

[Bharat Kumar Gogada]

> Subject: Re: [PATCH v8 2/2] PCI: xilinx-cpm: Add Versal CPM Root Port driver
> 
> On Mon, Jun 08, 2020 at 06:48:58PM +0530, Bharat Kumar Gogada wrote:
> > - Add support for Versal CPM as Root Port.
> > - The Versal ACAP devices include CCIX-PCIe Module (CPM). The integrated
> >   block for CPM along with the integrated bridge can function
> >   as PCIe Root Port.
> > - Bridge error and legacy interrupts in Versal CPM are handled using
> >   Versal CPM specific interrupt line.
> 
> > +static irqreturn_t xilinx_cpm_pcie_intr_handler(int irq, void
> > +*dev_id) {
> > +	struct xilinx_cpm_pcie_port *port = dev_id;
> > +	struct device *dev = port->dev;
> > +	struct irq_data *d;
> > +
> > +	d = irq_domain_get_irq_data(port->cpm_domain, irq);
> > +
> > +	switch (d->hwirq) {
> > +	case XILINX_CPM_PCIE_INTR_CORRECTABLE:
> > +	case XILINX_CPM_PCIE_INTR_NONFATAL:
> > +	case XILINX_CPM_PCIE_INTR_FATAL:
> > +		cpm_pcie_clear_err_interrupts(port);
> > +		fallthrough;
> > +
> > +	default:
> > +		if (intr_cause[d->hwirq].str)
> > +			dev_warn(dev, "%s\n", intr_cause[d->hwirq].str);
> > +		else
> > +			dev_warn(dev, "Unknown interrupt\n");
> 
> Maybe include d->hwirq in the "Unknown interrupt" message?
Hi Bjorn,
Yes, we can add.
> 
> I assume if we take the default case, there's no need to clear the interrupt?
 It's being handled in xilinx_cpm_pcie_event_flow.
> 
> > +	}
> > +
> > +	return IRQ_HANDLED;
> > +}
> 
> > +static int xilinx_cpm_setup_irq(struct xilinx_cpm_pcie_port *port) {
> > +	struct device *dev = port->dev;
> > +	struct platform_device *pdev = to_platform_device(dev);
> > +	int i, irq;
> > +
> > +	port->irq = platform_get_irq(pdev, 0);
> > +	if (port->irq < 0) {
> > +		dev_err(dev, "Unable to find misc IRQ line\n");
> 
> platform_get_irq() already prints an error message; you probably don't need
> another.
> 
> > +		return port->irq;
> > +	}
> > +
> > +	for (i = 0; i < ARRAY_SIZE(intr_cause); i++) {
> > +		int err;
> > +
> > +		if (!intr_cause[i].str)
> > +			continue;
> > +
> > +		irq = irq_create_mapping(port->cpm_domain, i);
> > +		if (WARN_ON(irq <= 0))
> 
> I'm not a huge fan of WARN_ON() inside "if" statements, but ... OK.
> 
> Why do these all need to be WARN_ON() instead of dev_warn()?
Yes, it can be replaced with dev_warn().
Will fix these.
Regards,
Bharat
> 
> > +			return -ENXIO;
> > +
> > +		err = devm_request_irq(dev, irq,
> xilinx_cpm_pcie_intr_handler,
> > +				       0, intr_cause[i].sym, port);
> > +		if (WARN_ON(err))
> > +			return err;
> > +	}
> > +
> > +	port->intx_irq = irq_create_mapping(port->cpm_domain,
> > +					    XILINX_CPM_PCIE_INTR_INTX);
> > +	if (WARN_ON(port->intx_irq <= 0))
> > +		return -ENXIO;
> > +
> > +	/* Plug the INTx chained handler */
> > +	irq_set_chained_handler_and_data(port->intx_irq,
> > +					 xilinx_cpm_pcie_intx_flow, port);