Re: [PATCH v11 06/13] usb: dwc3: core: Refactor PHY logic to support Multiport Controller

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 





On 9/1/2023 6:43 AM, Wesley Cheng wrote:
Hi Krishna,

          if (dwc->usb2_phy)
              otg_set_vbus(dwc->usb2_phy->otg, false);
-        phy_set_mode(dwc->usb2_generic_phy, PHY_MODE_USB_DEVICE);
-        phy_set_mode(dwc->usb3_generic_phy, PHY_MODE_USB_DEVICE);
+        phy_set_mode(dwc->usb2_generic_phy[0], PHY_MODE_USB_DEVICE);
+        phy_set_mode(dwc->usb3_generic_phy[0], PHY_MODE_USB_DEVICE);

Throughout this patch, you are looping across all PHYs irrespective of if we are in device mode or not.  This is the only exception where you are setting only PHY index 0 (for both SS and HS PHYs).  Do you think we should also only modify PHY index#0 for other PHY related sequences?

Hi Wesley,

Multiport controllers are host only capable currently. So if the GHWPARAMS indicate we are DRD/peripheral capable, we set num_usb2/3_ports to "1" unconditionally. So there would not be any looping necessary here.

          if (ret)
@@ -1804,9 +1887,12 @@ static int dwc3_read_port_info(struct dwc3 *dwc)
      dev_dbg(dwc->dev, "hs-ports: %u ss-ports: %u\n",
              dwc->num_usb2_ports, dwc->num_usb3_ports);
-
      iounmap(base);
+    if ((dwc->num_usb2_ports > DWC3_MAX_PORTS) ||
+        (dwc->num_usb3_ports > DWC3_MAX_PORTS))
+        return -ENOMEM;
+

Shouldn't this be more applicable to be included in patch#4 in this series?

The read_port_info function was only initially intended to read only port count and later the macro was added. So the check was put here.

Regards,
Krishna,



[Index of Archives]     [Device Tree Compilter]     [Device Tree Spec]     [Linux Driver Backports]     [Video for Linux]     [Linux USB Devel]     [Linux PCI Devel]     [Linux Audio Users]     [Linux Kernel]     [Linux SCSI]     [XFree86]     [Yosemite Backpacking]


  Powered by Linux