Re: [PATCH 08/12] net: stmmac: Add glue layer for StarFive JH7100 SoC
From: Andrew Lunn <andrew@lunn.ch>
Date: 2023-02-11 16:11:57
Also in:
linux-devicetree, linux-riscv, lkml, netdev
From: Andrew Lunn <andrew@lunn.ch>
Date: 2023-02-11 16:11:57
Also in:
linux-devicetree, linux-riscv, lkml, netdev
+ +#define JH7100_SYSMAIN_REGISTER28 0x70 +/* The value below is not a typo, just really bad naming by StarFive ¯\_(ツ)_/¯ */ +#define JH7100_SYSMAIN_REGISTER49 0xc8
Seems like the comment should be one line earlier? There is value in basing the names on the datasheet, but you could append something meaningful on the end: #define JH7100_SYSMAIN_REGISTER49_DLYCHAIN 0xc8 ???
+ if (!of_property_read_u32(np, "starfive,gtxclk-dlychain", >xclk_dlychain)) {
+ ret = regmap_write(sysmain, JH7100_SYSMAIN_REGISTER49, gtxclk_dlychain);
+ if (ret)
+ return dev_err_probe(dev, ret, "error selecting gtxclk delay chain\n");
+ }You should probably document that if starfive,gtxclk-dlychain is not found in the DT blob, the value for the delay chain is undefined. It would actually be better to define it, set it to 0 for example. That way, you know you don't have any dependency on the bootloader for example. Andrew _______________________________________________ linux-arm-kernel mailing list linux-arm-kernel@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-arm-kernel