RE: [PATCH net-next 3/3] net: phy: realtek: add LED support for RTL8261C/D

[Javen <javen_xu@realsil.com.cn>]

>The RTL8261C/D use the same LED register block as the RTL8221B: one control
>register per LED at VEND2 0xd032 + 2 * index selecting the link speeds that light
>it, activity blink bits in LCR6 and per-LED polarity and enable bits in LCR7.
>Compared to the RTL8221B there are four LEDs instead of three, and two more
>speed bits for 5G and 10G.
>
>Reuse the existing RTL822X_VND2_* definitions and add the two speed bits
>and the enable bit. The callbacks follow rtl822xb_led_*() closely, with two
>additions:
>
> - led_polarity_set() honours active-low LEDs. The polarity bit in LCR7
>   doubles as the manual on/off level in led_brightness_set(), so the
>   configured polarity is cached in the private data and restored by
>   led_hw_control_set().
>
> - led_hw_control_set() and led_brightness_set() set the output enable
>   bit, which firmware normally leaves set already.
>
>The callbacks duplicate most of rtl822xb_led_*(), differing only in the LED
>count, the speed mask and the LCR7 handling. Making the RTL8221B code
>table-driven and sharing it would be the right end state, but that touches a
>working driver and wants testing on an RTL8221B, so it is left for a follow-up
>series.
>
>The callbacks are wired to the RTL8261C_CG and RTL8261CE_CG entries, which
>between them cover the parts this was tested on. RTL8261D_VM is left out for
>lack of hardware.
>
>Tested on a Gemtek W1700K (Airoha AN7581) with an RTL8261CE: index 2
>drives a yellow and index 3 a green LED on each 10G port. Under hardware
>control the link speed bits, the activity blink and the combined
>TRIGGER_NETDEV_LINK all take effect and offloading is confirmed through the
>netdev trigger. The polarity bit was verified to be active-high when clear on that
>board. Also tested on AN7581 and MT7988A boards with an RTL8261C and
>RTL8261D.
>
>Signed-off-by: Chad Monroe <chad@monroe.io>
>Co-developed-by: Simon Polack <spolack+git@mailbox.org>
>Signed-off-by: Simon Polack <spolack+git@mailbox.org>
>Assisted-by: LLM
>---
> drivers/net/phy/realtek/realtek_main.c | 228
>+++++++++++++++++++++++++
> 1 file changed, 228 insertions(+)
>
>diff --git a/drivers/net/phy/realtek/realtek_main.c
>b/drivers/net/phy/realtek/realtek_main.c
>index 04d397f0a2d8..b1f020aa7760 100644
>--- a/drivers/net/phy/realtek/realtek_main.c
>+++ b/drivers/net/phy/realtek/realtek_main.c
>@@ -183,12 +183,15 @@
> #define RTL822X_VND2_LCR_LINK_100              BIT(1)
> #define RTL822X_VND2_LCR_LINK_1000             BIT(2)
> #define RTL822X_VND2_LCR_LINK_2500             BIT(5)
>+#define RTL822X_VND2_LCR_LINK_5000             BIT(6)
>+#define RTL822X_VND2_LCR_LINK_10000            BIT(4)
>
> #define RTL822X_VND2_LCR6                      0xd040
> #define RTL822X_VND2_LED_ACT(x)                        BIT(x)
>
> #define RTL822X_VND2_LCR7                      0xd044
> #define RTL822X_VND2_LED_POLAR(x)              BIT(x)
>+#define RTL822X_VND2_LED_EN(x)                 BIT((x) + 4)
>
> #define RTL8224_MII_RTCT                       0x11
> #define RTL8224_MII_RTCT_ENABLE                        BIT(0)
>@@ -305,6 +308,7 @@
> #define FW_SUB_MAGIC_8261D             0x32363144
> #define RTL8261X_POLL_TIMEOUT_MS       100
> #define RTL8261X_MAX_MMD_DEV           31
>+#define RTL8261X_LED_COUNT             4

Hi,

We kindly suggest dropping this patch for the same reason as in Patch 2/3.

LED config is related to HW config and will be configured via firmware.

BRs,
Javen