On 4/28/21 3:51 AM, Tim Harvey wrote:
On Tue, Apr 27, 2021 at 10:45 AM Marek Vasut <ma...@denx.de> wrote:
On 4/27/21 7:08 PM, Tim Harvey wrote:
There is no need to set and/or detect mode in of_to_plat and
accessing phy registers at that point before device power domain and
clock are enabled will cause hangs on platforms such as IMX8M Mini.
Move the mode set/detect from of_to_plat into the probe and remove
the unnecessary of_to_plat.
Signed-off-by: Tim Harvey <thar...@gateworks.com>
---
drivers/usb/host/ehci-mx6.c | 42 ++++++++++++++-----------------------
1 file changed, 16 insertions(+), 26 deletions(-)
diff --git a/drivers/usb/host/ehci-mx6.c b/drivers/usb/host/ehci-mx6.c
index 06be9deaaa..c2dfe49012 100644
--- a/drivers/usb/host/ehci-mx6.c
+++ b/drivers/usb/host/ehci-mx6.c
@@ -539,28 +539,6 @@ static int ehci_usb_phy_mode(struct udevice *dev)
return 0;
}
-static int ehci_usb_of_to_plat(struct udevice *dev)
-{
- struct usb_plat *plat = dev_get_plat(dev);
- enum usb_dr_mode dr_mode;
-
- dr_mode = usb_get_dr_mode(dev_ofnode(dev));
-
- switch (dr_mode) {
- case USB_DR_MODE_HOST:
- plat->init_type = USB_INIT_HOST;
- break;
- case USB_DR_MODE_PERIPHERAL:
- plat->init_type = USB_INIT_DEVICE;
- break;
- case USB_DR_MODE_OTG:
- case USB_DR_MODE_UNKNOWN:
- return ehci_usb_phy_mode(dev);
- };
-
- return 0;
-}
-
[...]
@@ -764,7 +755,6 @@ U_BOOT_DRIVER(usb_mx6) = {
.name = "ehci_mx6",
.id = UCLASS_USB,
.of_match = mx6_usb_ids,
- .of_to_plat = ehci_usb_of_to_plat,
I wonder why it was implemented in of_to_plat originally , maybe there
is some reason for that ?
Marek,
Looking back the commit that added the ehci_usb_ofdata_to_platdata was:
cccbddc38c43 ("usb: ehci-mx6: implement ofdata_to_platdata")
Before that there was a board-specific function that would set the
usb_plat->init_type.
The only reason to set usb_plat->init_type in of_to_plat would be so
that drivers/usb/host/usb-uclass.c would have knowledge of it but I
only see that it is set there in usb_setup_ehci_gadget.
So interaction with Gadget, that's what I was afraid of.
I added Peng, Stefano, and Simon to the thread to see if they see an
issue with doing away with of_to_plat setting the usb_plat->init_type
prior to probe.
I added Lukasz too.
