usb: dwc2: Fix call location of dwc2_check_core_endianness
authorBruno Meirelles Herrera <bmh@certi.org.br>
Mon, 27 Aug 2018 21:36:38 +0000 (18:36 -0300)
committerFelipe Balbi <felipe.balbi@linux.intel.com>
Wed, 5 Sep 2018 10:12:31 +0000 (13:12 +0300)
commitd9707490077bee0c7060ef5665a90656e1078b66
tree7be07bf74b9834eb2291d1f52221fdc79f988dda
parentb497fff6f59ec4ab2816439e7ab976a90b7bab5c
usb: dwc2: Fix call location of dwc2_check_core_endianness

Some SoC/IP as STM32F469, the snpsid can only be read after clock is
enabled, otherwise it will read as 0, and the dwc2_check_core_endianness
will assume the core and AHB have opposite endianness, leading to the
following error:

[    1.976339] dwc2 50000000.usb: 50000000.usb supply vusb_d not found, using dummy regulator
[    1.986124] dwc2 50000000.usb: Linked as a consumer to regulator.0
[    1.992711] dwc2 50000000.usb: 50000000.usb supply vusb_a not found, using dummy regulator
[    2.003672] dwc2 50000000.usb: dwc2_core_reset: HANG! AHB Idle timeout GRSTCTL GRSTCTL_AHBIDLE
[    2.015176] dwc2: probe of 50000000.usb failed with error -16

The proposed patch changes the location where dwc2_check_core_endianness
is called, allowing the clock peripheral to be enabled first.

Acked-by: Minas Harutyunyan <hminas@synopsys.com>
Tested-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
Signed-off-by: Bruno Meirelles Herrera <bmh@certi.org.br>
Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com>
drivers/usb/dwc2/platform.c