Re: [PATCH 2/3] mailbox: qcom-cpucp-mbox: Add support for SC7280 CPUCP mailbox controller

[Bjorn Andersson]

On Thu, Oct 17, 2024 at 05:22:36PM +0530, Shivnandan Kumar wrote:
> On 10/6/2024 8:03 AM, Bjorn Andersson wrote:
> > On Tue, Sep 24, 2024 at 10:39:40AM GMT, Shivnandan Kumar wrote:
[..]
> > >   static int qcom_cpucp_mbox_probe(struct platform_device *pdev)
> > >   {
> > > +	const struct qcom_cpucp_mbox_desc *desc;
> > >   	struct device *dev = &pdev->dev;
> > >   	struct qcom_cpucp_mbox *cpucp;
> > >   	struct mbox_controller *mbox;
> > > +	struct resource *res;
> > >   	int irq, ret;
> > > 
> > > +	desc = device_get_match_data(&pdev->dev);
> > > +	if (!desc)
> > > +		return -EINVAL;
> > > +
> > >   	cpucp = devm_kzalloc(dev, sizeof(*cpucp), GFP_KERNEL);
> > >   	if (!cpucp)
> > >   		return -ENOMEM;
> > > 
> > > -	cpucp->rx_base = devm_of_iomap(dev, dev->of_node, 0, NULL);
> > > -	if (IS_ERR(cpucp->rx_base))
> > > -		return PTR_ERR(cpucp->rx_base);
> > > +	cpucp->desc = desc;
> > > +
> > > +	if (desc->v2_mbox) {
> > > +		cpucp->rx_base = devm_of_iomap(dev, dev->of_node, 0, NULL);
> > > +		if (IS_ERR(cpucp->rx_base))
> > > +			return PTR_ERR(cpucp->rx_base);
> > > +	/* Legacy mailbox quirks due to shared region with EPSS register space */
> > 
> > Why can't we have the same code in both cases?
> > 
> 
> 
> RX address space share region with EPSS. Due to which devm_of_iomap returns
> -EBUSY.
> 

I assumed that was the case, and that explains why the legacy system
needs a different code path.

But, couldn't you use the same for the v2 solution, so we avoid having
two different code paths?

Regards,
Bjorn