[PATCH v4 3/3] phy: nuvoton: phy-ma35d1-usb2: extend to dual-port with OTG support

[sashiko-bot at kernel.org]

Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
- [High] The driver fails to clear the MISCFCR0_UHOVRCURH hardware bit when the `nuvoton,oc-active-high` device tree property is absent, inheriting potentially incorrect bootloader state.
- [Medium] The driver ignores the return value of regmap_read(), potentially evaluating uninitialized stack memory if the read fails.
--

--- Patch [3]: [PATCH v4 3/3] phy: nuvoton: phy-ma35d1-usb2: extend to dual-port with OTG support ---
Note: The format of this review is altered due to strict recitation restrictions. Instead of inline patch quoting, a free-form summary of the findings is provided below.

Commit: 678e65678c8fb5e2363eb1d51e7a9a2159ca8dd9
Author: Joey Lu <a0987203069 at gmail.com>
Subject: phy: nuvoton: phy-ma35d1-usb2: extend to dual-port with OTG support

This commit extends the driver to manage both USB ports and updates the init
polling logic to handle host-mode and device-mode clock stability. It also
registers a read-only USB role switch and adds support for the rcalcode and
oc-active-high device tree properties.

[Severity: Medium]
In ma35_usb_phy_init() and ma35_role_sw_get(), the return value of
regmap_read() is not checked before evaluating the 'val' stack variable. If
the register read fails, might the driver evaluate uninitialized memory,
causing it to return an unpredictable USB role or falsely bypass the PHY
initialization?

[Severity: High]
In ma35_usb_phy_probe(), the driver sets the MISCFCR0_UHOVRCURH bit if the
"nuvoton,oc-active-high" property is present, but lacks an else path to clear
it if the property is absent. Does this allow the hardware to incorrectly
inherit an inverted over-current polarity from the bootloader or a prior
kernel instance?

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260720102222.251179-1-a0987203069@gmail.com?part=3