Re: [PATCH v9 06/10] usb: dwc3: qcom: Add support to read IRQ's related to multiport

[Krishna Kurapati PSSNV]

On 7/21/2023 2:51 PM, Johan Hovold wrote:
> Again, please remember to trim your replies.
>
> On Fri, Jul 21, 2023 at 01:49:37PM +0530, Krishna Kurapati PSSNV wrote:
> > On 7/21/2023 1:44 PM, Johan Hovold wrote:
> > > On Mon, Jul 03, 2023 at 12:29:41AM +0530, Krishna Kurapati PSSNV wrote:
> > > > On 6/27/2023 8:01 PM, Johan Hovold wrote:
>
> [...]
>
> > > > > So the above should all be merged in either a single helper looking up
> > > > > all the interrupts for a port and resused for the non-MP case.
>
> >    How about the implementation in the attached patches. This way we
> > don't need to know if we are multiport capable or not.
>
> As I wrote above, I think you should instead add a common helper for
> setting up all the interrupts for a port. For example, along the lines
> of:
>
> 	dwc3_setup_port_irq(int index)
> 	{
> 		if (index == 0)
> 			try non-mp name
> 		else
> 			generate mp name
>
> 		lookup and request hs irqs
> 		lookup and request ss irq
> 		lookup and request power irq
> 	}
>
> 	dwc3_setup_irq()
> 	{
> 		determine if MP (num_ports)
>
> 		for each port
> 			dwc3_setup_port_irq(port index)
> 	}
> The port irq helper can either be told using a second argument that this
> is a non-mp controller, or you can first try looking up one of the
> non-mp names.


I think I did something similar. I prepared a helper to request IRQ in 
the patch and the main logic would reside in setup_irq where i would try 
and get IRQ's.

Irrespective of whether we are MP capable or not, how about we read all 
IRQ's like in the patch attached previously. And the implementation 
facilitates addition of ACPI to multiport also if required. I am just 
trying to cover all cases like this by declaring IRQ info in global section.

static int dwc3_qcom_prep_irq(struct dwc3_qcom *qcom, char *irq_name,
                               char *disp_name, int irq)
{
       int ret;

       /* Keep wakeup interrupts disabled until suspend */
       irq_set_status_flags(irq, IRQ_NOAUTOEN);
       ret = devm_request_threaded_irq(/* Give inouts here*/);
       if (ret)
              dev_err(qcom->dev, "%s failed: %d\n", irq_name, ret);

       return ret;
}


static int dwc3_qcom_setup_irq(struct platform_device *pdev)
{
  for (DP_IRQ[ i = 0-3] {
     try getting dp_irq_i using MP_IRQ strings
     if ((ret < 0)  and (i == 0))
	try getting dp_irq_i using NON_MP_IRQ strings

     call prep_irq accordingly.
  }

  //Run same loop for DM and SS
}

The second patch was just enabling IRQ's for all ports to support wakeup.

> My mailer discarded your second patch, but you cannot assume that the
> devices connected to each port are of the same type (e.g. HS or SS)
> based on what is connected to the first port.
Are you referring to enabling IRQ's for different ports before going to 
suspend ? Meaning get the speed of enum on each port and accordingly 
enable IRQ's ?

Regards,
Krishna,