RE: [PATCH v2 4/5] clk: socfpga: agilex: add clock driver for the Agilex5

From: Rabara, Niravkumar L
Date: Sun Aug 13 2023 - 08:56:49 EST




> -----Original Message-----
> From: Stephen Boyd <sboyd@xxxxxxxxxx>
> Sent: Thursday, 10 August, 2023 5:27 AM
> To: Rabara, Niravkumar L <niravkumar.l.rabara@xxxxxxxxx>
> Cc: Ng, Adrian Ho Yin <adrian.ho.yin.ng@xxxxxxxxx>; andrew@xxxxxxx;
> conor+dt@xxxxxxxxxx; devicetree@xxxxxxxxxxxxxxx; dinguyen@xxxxxxxxxx;
> krzysztof.kozlowski+dt@xxxxxxxxxx; linux-clk@xxxxxxxxxxxxxxx; linux-
> kernel@xxxxxxxxxxxxxxx; Turquette, Mike <mturquette@xxxxxxxxxxxx>;
> netdev@xxxxxxxxxxxxxxx; p.zabel@xxxxxxxxxxxxxx; richardcochran@xxxxxxxxx;
> robh+dt@xxxxxxxxxx; wen.ping.teh@xxxxxxxxx
> Subject: Re: [PATCH v2 4/5] clk: socfpga: agilex: add clock driver for the Agilex5
>
> Quoting niravkumar.l.rabara@xxxxxxxxx (2023-07-31 18:02:33)
> > diff --git a/drivers/clk/socfpga/clk-agilex.c
> > b/drivers/clk/socfpga/clk-agilex.c
> > index 74d21bd82710..3dcd0f233c17 100644
> > --- a/drivers/clk/socfpga/clk-agilex.c
> > +++ b/drivers/clk/socfpga/clk-agilex.c
> > @@ -1,6 +1,6 @@
> > // SPDX-License-Identifier: GPL-2.0
> > /*
> > - * Copyright (C) 2019, Intel Corporation
> > + * Copyright (C) 2019-2023, Intel Corporation
> > */
> > #include <linux/slab.h>
> > #include <linux/clk-provider.h>
> > @@ -9,6 +9,7 @@
> > #include <linux/platform_device.h>
> >
> > #include <dt-bindings/clock/agilex-clock.h>
> > +#include <dt-bindings/clock/intel,agilex5-clkmgr.h>
> >
> > #include "stratix10-clk.h"
> >
> > @@ -41,6 +42,67 @@ static const struct clk_parent_data mpu_free_mux[] = {
> > .name = "f2s-free-clk", },
> > };
> >
> > +static const struct clk_parent_data core0_free_mux[] = {
> > + { .fw_name = "main_pll_c1",
> > + .name = "main_pll_c1", },
>
> We're adding support for something new? Only set .fw_name in that case, as
> .name will never be used. To do even better, set only .index so that we don't do
> any string comparisons.
>
Yes we are adding Agilex5 SoCFPGA platform specific clocks.
I will remove .name and only keep .fw_name in next version of this patch.

> > + { .fw_name = "peri_pll_c0",
> > + .name = "peri_pll_c0", },
> > + { .fw_name = "osc1",
> > + .name = "osc1", },
> > + { .fw_name = "cb-intosc-hs-div2-clk",
> > + .name = "cb-intosc-hs-div2-clk", },
> > + { .fw_name = "f2s-free-clk",
> > + .name = "f2s-free-clk", },
> > +};
> > +
> [...]
> > +
> > static int n5x_clk_register_c_perip(const struct n5x_perip_c_clock *clks,
> > int nums, struct
> > stratix10_clock_data *data) { @@ -535,6 +917,51 @@ static int
> > n5x_clkmgr_init(struct platform_device *pdev)
> > return 0;
> > }
> >
> > +static int agilex5_clkmgr_init(struct platform_device *pdev) {
> > + struct device_node *np = pdev->dev.of_node;
> > + struct device *dev = &pdev->dev;
> > + struct stratix10_clock_data *clk_data;
>
> Maybe call this stratix_data so that clk_data.clk_data is stratix_data.clk_data.

Will fix this in next version.

>
> > + struct resource *res;
> > + void __iomem *base;
> > + int i, num_clks;
> > +
> > + res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
> > + base = devm_ioremap_resource(dev, res);
>
> This is a single function call, devm_platform_ioremap_resource().i

Noted. Will fix in next version.

>
> > + if (IS_ERR(base))
> > + return PTR_ERR(base);
> > +
> > + num_clks = AGILEX5_NUM_CLKS;
> > +
> > + clk_data = devm_kzalloc(dev, struct_size(clk_data, clk_data.hws,
> > + num_clks), GFP_KERNEL);
> > + if (!clk_data)
> > + return -ENOMEM;
> > +
> > + for (i = 0; i < num_clks; i++)
> > + clk_data->clk_data.hws[i] = ERR_PTR(-ENOENT);
> > +
> > + clk_data->base = base;
> > + clk_data->clk_data.num = num_clks;
> > +
> > + agilex_clk_register_pll(agilex_pll_clks, ARRAY_SIZE(agilex_pll_clks),
> > + clk_data);
> > +
> > + agilex_clk_register_c_perip(agilex5_main_perip_c_clks,
> > + ARRAY_SIZE(agilex5_main_perip_c_clks),
> > + clk_data);
> > +
> > + agilex_clk_register_cnt_perip(agilex5_main_perip_cnt_clks,
> > + ARRAY_SIZE(agilex5_main_perip_cnt_clks),
> > + clk_data);
> > +
> > + agilex_clk_register_gate(agilex5_gate_clks,
> > + ARRAY_SIZE(agilex5_gate_clks),
> > + clk_data);
> > +
> > + of_clk_add_hw_provider(np, of_clk_hw_onecell_get,
> > + &clk_data->clk_data);
>
> devm? Or when is this provider removed?

Will fix in next version.

>
> > + return 0;
> > +}
> > +
> > static int agilex_clkmgr_probe(struct platform_device *pdev) {
> > int (*probe_func)(struct platform_device *init_func);