[PATCH V2 1/3] clk: mvebu: add armada-370-xp specific clocks
Gregory CLEMENT
gregory.clement at free-electrons.com
Tue Oct 30 22:46:22 EST 2012
On 10/30/2012 12:20 PM, Mike Turquette wrote:
> Quoting Gregory CLEMENT (2012-10-01 14:12:04)
>> Add Armada 370/XP specific clocks: core clocks and CPU clocks.
>>
>> The CPU clocks are only for Armada XP for the SMP mode.
>>
>> The core clocks are clocks which have their rate set during reset. The
>> code was written with the other SoCs of the mvebu family in
>> mind. Adding them should be pretty straight forward. For a new
>> SoC, only 3 binding have to be added:
>> - one to provide the tclk frequency
>> - one to provde the pclk frequency
>> - and one to provide the ratio between the pclk and the children
>> clocks
>>
>> Signed-off-by: Gregory CLEMENT <gregory.clement at free-electrons.com>
>
> Thanks for sending v2 Gregory. Functionally this series looks good but
> unfortunately there are a few style and white space nitpicks below.
Hi Mike,
You don't review the last version: I have sent a third version tow weeks
ago (October th 15th). But all your comment should applied for this last
version.
The changelog was
V2 -> V3:
- Rebased on top of v3.7-rc1
- Fixed a typo in device trees
- Fixed warning from checkpatch
Thanks for your review, I will send a new version in few hours.
>
>> diff --git a/Documentation/devicetree/bindings/clock/mvebu-cpu-clock.txt b/Documentation/devicetree/bindings/clock/mvebu-cpu-clock.txt
>> new file mode 100644
>> index 0000000..c524618
>> --- /dev/null
>> +++ b/Documentation/devicetree/bindings/clock/mvebu-cpu-clock.txt
>> @@ -0,0 +1,21 @@
>> +Device Tree Clock bindings for cpu clock of Marvell EBU platforms
>> +
>> +Required properties:
>> +- compatible : shall be one of the following:
>> + "marvell,armada-xp-cpu-clockctrl" - cpu clocks for Armada XP
>> +- reg : Address and length of the clock complex register set
>> +- #clock-cells : should be set to 1.
>> +- clocks : shall be the input parent clock phandle for the clock.
>> +
>> +cpuclk: clock-complex at d0018700 {
>> + #clock-cells = <1>;
>> + compatible = "marvell,armada-xp-cpu-clockctrl";
>> + reg = <0xd0018700 0xA0>;
>> + clocks = <&coreclk 1>;
>> +}
>> +
>> +cpu at 0 {
>> + compatible = "marvell,sheeva-v7";
>
> Unnecessary white space before the tab in the line above.
>
>> diff --git a/drivers/clk/mvebu/clk-core.c b/drivers/clk/mvebu/clk-core.c
>> new file mode 100644
>> index 0000000..5792cb5
>> --- /dev/null
>> +++ b/drivers/clk/mvebu/clk-core.c
>> @@ -0,0 +1,312 @@
>> +/*
>> + * Marvell EBU clock core handling defined at reset
>> + *
>> + * Copyright (C) 2012 Marvell
>> + *
>> + * Gregory CLEMENT <gregory.clement at free-electrons.com>
>> + *
>> + * This file is licensed under the terms of the GNU General Public
>> + * License version 2. This program is licensed "as is" without any
>> + * warranty of any kind, whether express or implied.
>> + */
>> +#include <linux/kernel.h>
>> +#include <linux/clk.h>
>> +#include <linux/clkdev.h>
>> +#include <linux/clk-provider.h>
>> +#include <linux/of_address.h>
>> +#include <linux/io.h>
>> +#include <linux/of.h>
>> +
>> +/* Sample At Reset is a 64 bit bitfiled split in two register of 32
>> + * bits*/
>
> Can you make this proper comment block style per CodingStyle convention?
>
>> +
>> +#define SARL 0 /* Low part [0:31] */
>> +#define SARL_AXP_PCLK_FREQ_OPT 21
>> +#define SARL_AXP_PCLK_FREQ_OPT_MASK 0x7
>> +#define SARL_A370_PCLK_FREQ_OPT 11
>> +#define SARL_A370_PCLK_FREQ_OPT_MASK 0xF
>> +#define SARL_AXP_FAB_FREQ_OPT 24
>> +#define SARL_AXP_FAB_FREQ_OPT_MASK 0xF
>> +#define SARL_A370_FAB_FREQ_OPT 15
>> +#define SARL_A370_FAB_FREQ_OPT_MASK 0x1F
>> +#define SARL_A370_TCLK_FREQ_OPT 20
>> +#define SARL_A370_TCLK_FREQ_OPT_MASK 0x1
>> +#define SARH 4 /* High part [32:63] */
>> +#define SARH_AXP_PCLK_FREQ_OPT (52-32)
>> +#define SARH_AXP_PCLK_FREQ_OPT_MASK 0x1
>> +#define SARH_AXP_PCLK_FREQ_OPT_SHIFT 3
>> +#define SARH_AXP_FAB_FREQ_OPT (51-32)
>> +#define SARH_AXP_FAB_FREQ_OPT_MASK 0x1
>> +#define SARH_AXP_FAB_FREQ_OPT_SHIFT 4
>> +
>> +u32 *sar_reg;
>> +int sar_reg_size;
>> +
>> +enum core_clk {
>> + tclk, pclk, nbclk, hclk, dramclk, clk_max
>> +};
>> +
>> +struct core_clk_fn {
>> + u32(*get_tclk_freq) (void);
>> + u32(*get_pck_freq) (void);
>> + const int *(*get_fab_freq_opt) (void);
>> +};
>> +
>> +/* Ratio between VCO and each of the member in the following order:
>> + CPU clock, L2 clock, DRAM controler clock, DDR clcok */
>
> ditto
>
>> +static const int reset_core_ratio[32][4] = {
>> + [0x01] = {1, 2, 2, 2},
>> + [0x02] = {2, 2, 6, 3},
>> + [0x03] = {2, 2, 3, 3},
>> + [0x04] = {1, 2, 3, 3},
>> + [0x05] = {1, 2, 4, 2},
>> + [0x06] = {1, 1, 2, 2},
>> + [0x07] = {2, 3, 6, 6},
>> + [0x09] = {1, 2, 6, 3},
>> + [0x0A] = {2, 4, 10, 5},
>> + [0x0C] = {1, 2, 4, 4},
>> + [0x0F] = {2, 2, 5, 5},
>> + [0x13] = {1, 1, 2, 1},
>> + [0x14] = {2, 3, 6, 3},
>> + [0x1B] = {1, 1, 1, 1},
>> +};
>> +
>> +static struct clk *clks[clk_max];
>> +
>> +static struct clk_onecell_data clk_data;
>> +
>> +/* Frequency in MHz*/
>> +static u32 armada_370_pclk[] = { 400, 533, 667, 800, 1000, 1067, 1200 };
>> +
>> +static u32 armada_xp_pclk[] = { 1000, 1066, 1200, 1333, 1500, 1666,
>> + 1800, 2000, 667, 0, 800, 1600
>> +};
>> +
>> +static u32 armada_370_tclk[] = { 166, 200 };
>> +
>> +static const int *__init armada_370_get_fab_freq_opt(void)
>> +{
>> + u8 fab_freq_opt = 0;
>> +
>> + fab_freq_opt = ((sar_reg[0] >> SARL_A370_FAB_FREQ_OPT) &
>> + SARL_A370_FAB_FREQ_OPT_MASK);
>> +
>> + if (reset_core_ratio[fab_freq_opt][0] == 0)
>> + return NULL;
>> + else
>> + return reset_core_ratio[fab_freq_opt];
>> +}
>> +
>> +static u32 __init armada_370_get_pck_freq(void)
>> +{
>> + u32 cpu_freq;
>> + u8 cpu_freq_select = 0;
>> +
>> + cpu_freq_select = ((sar_reg[0] >> SARL_A370_PCLK_FREQ_OPT) &
>> + SARL_A370_PCLK_FREQ_OPT_MASK);
>> + if (cpu_freq_select > ARRAY_SIZE(armada_370_pclk)) {
>> + pr_err("CPU freq select unsuported %d\n", cpu_freq_select);
>> + cpu_freq = 0;
>> + } else
>> + cpu_freq = armada_370_pclk[cpu_freq_select];
>> +
>> + return cpu_freq * 1000 * 1000;
>> +}
>> +
>> +static u32 __init armada_370_get_tclk_freq(void)
>> +{
>> + u32 tclk_freq;
>> + u8 tclk_freq_select = 0;
>> +
>> + tclk_freq_select = ((sar_reg[0] >> SARL_A370_TCLK_FREQ_OPT) &
>> + SARL_A370_TCLK_FREQ_OPT_MASK);
>> + if (tclk_freq_select > ARRAY_SIZE(armada_370_tclk)) {
>> + pr_err("TCLK freq select unsuported %d\n", tclk_freq_select);
>> + tclk_freq = 0;
>> + } else
>> + tclk_freq = armada_370_tclk[tclk_freq_select];
>> +
>> + return tclk_freq * 1000 * 1000;
>> +}
>> +
>> +static const int *__init armada_xp_get_fab_freq_opt(void)
>> +{
>> + u8 fab_freq_opt = 0;
>> +
>> + fab_freq_opt = ((sar_reg[0] >> SARL_AXP_FAB_FREQ_OPT) &
>> + SARL_AXP_FAB_FREQ_OPT_MASK);
>> + /* The upper bit is not contiguous to the other ones
>> + * and located in the high part of the SAR
>> + * registers */
>
> ditto
>
>> + fab_freq_opt |= (((sar_reg[1] >> SARH_AXP_FAB_FREQ_OPT) &
>> + SARH_AXP_FAB_FREQ_OPT_MASK)
>> + << SARH_AXP_FAB_FREQ_OPT_SHIFT);
>> +
>> + if (reset_core_ratio[fab_freq_opt][0] == 0)
>> + return NULL;
>> + else
>> + return reset_core_ratio[fab_freq_opt];
>> +}
>> +
>> +static u32 __init armada_xp_get_pck_freq(void)
>> +{
>> + u32 cpu_freq;
>> + u8 cpu_freq_select = 0;
>> +
>> + cpu_freq_select = ((sar_reg[0] >> SARL_AXP_PCLK_FREQ_OPT) &
>> + SARL_AXP_PCLK_FREQ_OPT_MASK);
>> + /* The upper bit is not contiguous to the other ones
>> + * and located in the high part of the SAR
>> + * registers */
>
> ditto
>
>> + cpu_freq_select |= (((sar_reg[1] >> SARH_AXP_PCLK_FREQ_OPT) &
>> + SARH_AXP_PCLK_FREQ_OPT_MASK)
>> + << SARH_AXP_PCLK_FREQ_OPT_SHIFT);
>> + if (cpu_freq_select > ARRAY_SIZE(armada_xp_pclk)) {
>> + pr_err("CPU freq select unsuported: %d\n", cpu_freq_select);
>> + cpu_freq = 0;
>> + } else
>> + cpu_freq = armada_xp_pclk[cpu_freq_select];
>> +
>> + return cpu_freq * 1000 * 1000;
>> +}
>> +
>> +/* For Armada XP TCLK frequency is fix: 250MHz */
>> +static u32 __init armada_xp_get_tclk_freq(void)
>> +{
>> + return 250 * 1000 * 1000;
>> +}
>> +
>> +void __init of_core_clk_setup(struct device_node *node,
>> + struct core_clk_fn clk_fn)
>> +{
>> + struct clk *clk;
>> + unsigned long rate;
>> + const char *clk_name;
>> + int i;
>> +
>> + /* clock 0 is tclk */
>> + of_property_read_string_index(node, "clock-output-names", tclk,
>> + &clk_name);
>> + rate = clk_fn.get_tclk_freq();
>> + if (rate != 0)
>> + clk = clk_register_fixed_rate(NULL, clk_name, NULL,
>> + CLK_IS_ROOT, rate);
>> + else {
>> + pr_err("Invalid freq for %s\n", clk_name);
>> + return;
>> + }
>> +
>> + if (WARN_ON(IS_ERR(clk)))
>> + return;
>> + clks[tclk] = clk;
>> +
>> + /* clock 1 is pclk */
>> + of_property_read_string_index(node, "clock-output-names", pclk,
>> + &clk_name);
>> + rate = clk_fn.get_pck_freq();
>> + if (rate != 0)
>> + clk = clk_register_fixed_rate(NULL, clk_name, NULL,
>> + CLK_IS_ROOT, rate);
>> + else {
>> + pr_err("Invalid freq for %s\n", clk_name);
>> + return;
>> + }
>> + if (WARN_ON(IS_ERR(clk)))
>> + return;
>> + clks[pclk] = clk;
>> +
>> + /* the clocks 2 to 4 are nbclk, hclk and dramclk and are all
>> + * derivated from the clock 1: pclk */
>
> ditto
>
>> +
>> + for (i = nbclk; i <= dramclk; i++) {
>> + const int *ratio = clk_fn.get_fab_freq_opt();
>> + const char *parent_clk_name;
>> +
>> + of_property_read_string_index(node, "clock-output-names",
>> + i, &clk_name);
>> + of_property_read_string_index(node, "clock-output-names",
>> + pclk, &parent_clk_name);
>> +
>> + if (ratio != NULL)
>> + clk = clk_register_fixed_factor(NULL, clk_name,
>> + parent_clk_name, 0,
>> + ratio[0],
>> + ratio[i - nbclk + 1]);
>> + else {
>> + pr_err("Invalid clk ratio for %s\n", clk_name);
>> + return;
>> + }
>> +
>> + if (WARN_ON(IS_ERR(clk)))
>> + return;
>> + clks[i] = clk;
>> + }
>> + clk_data.clk_num = ARRAY_SIZE(clks);
>> + clk_data.clks = clks;
>> + of_clk_add_provider(node, of_clk_src_onecell_get, &clk_data);
>> +}
>> +
>> +static struct core_clk_fn armada_370_clk_fn = {
>> + .get_tclk_freq = armada_370_get_tclk_freq,
>> + .get_pck_freq = armada_370_get_pck_freq,
>> + .get_fab_freq_opt = armada_370_get_fab_freq_opt,
>> +};
>> +
>> +static struct core_clk_fn armada_xp_clk_fn = {
>> + .get_tclk_freq = armada_xp_get_tclk_freq,
>> + .get_pck_freq = armada_xp_get_pck_freq,
>> + .get_fab_freq_opt = armada_xp_get_fab_freq_opt,
>> +};
>> +
>> +static const __initconst struct of_device_id clk_match[] = {
>> + {
>> + .compatible = "marvell,armada-370-core-clockctrl",
>> + .data = &armada_370_clk_fn,
>> + },
>
> Indention is a bit weird here. The members are just indented with one
> space and there is an extra white space in front of the closing bracket.
> Can you re-align?
>
>> +
>> + {
>> + .compatible = "marvell,armada-xp-core-clockctrl",
>> + .data = &armada_xp_clk_fn,
>> + },
>
> ditto
>
>> + {
>> + /* sentinel */
>> + }
>
> ditto
>
>> +};
>> +
>> +void __init mvebu_core_clocks_init(void)
>> +{
>> + struct device_node *np;
>> + struct resource res;
>> + void __iomem *sar_base;
>> + int i;
>> +
>> + np = of_find_node_by_name(NULL, "mvebu-sar");
>> +
>> + if (!np)
>> + goto err;
>> +
>> + if (of_address_to_resource(np, 0, &res))
>> + goto err;
>> +
>> + sar_reg_size = resource_size(&res);
>> + sar_reg = kmalloc(sar_reg_size, GFP_KERNEL);
>> +
>> + sar_base = ioremap(res.start, sar_reg_size);
>> + if (sar_base == NULL)
>> + goto err;
>> + for (i = 0; i < sar_reg_size; i += sizeof(*sar_reg))
>> + sar_reg[i] = readl(sar_base + i);
>> +
>> + iounmap(sar_base);
>> +
>> + for_each_matching_node(np, clk_match) {
>> + const struct of_device_id *match = of_match_node(clk_match, np);
>> + struct core_clk_fn *clk_fn = match->data;
>> + of_core_clk_setup(np, *clk_fn);
>> + }
>> +
>> + return;
>> +err:
>> + pr_err("%s:SAR base adresse not set in DT\n", __func__);
>> + return;
>> +}
>> diff --git a/drivers/clk/mvebu/clk-core.h b/drivers/clk/mvebu/clk-core.h
>> new file mode 100644
>> index 0000000..a04f80b
>> --- /dev/null
>> +++ b/drivers/clk/mvebu/clk-core.h
>> @@ -0,0 +1,19 @@
>> +/*
>> + * * Marvell EBU clock core handling defined at reset
>> + *
>> + * Copyright (C) 2012 Marvell
>> + *
>> + * Gregory CLEMENT <gregory.clement at free-electrons.com>
>> + *
>> + * This file is licensed under the terms of the GNU General Public
>> + * License version 2. This program is licensed "as is" without any
>> + * warranty of any kind, whether express or implied.
>> + */
>> +
>> +#ifndef __MVEBU_CLK_CORE_H
>> +#define __MVEBU_CLK_CORE_H
>> +
>> +void __init of_core_clk_setup(struct device_node *node);
>> +void __init mvebu_core_clocks_init(void);
>> +
>> +#endif
>> diff --git a/drivers/clk/mvebu/clk-cpu.c b/drivers/clk/mvebu/clk-cpu.c
>> new file mode 100644
>> index 0000000..6b5c841
>> --- /dev/null
>> +++ b/drivers/clk/mvebu/clk-cpu.c
>> @@ -0,0 +1,155 @@
>> +/*
>> + * Marvell MVEBU CPU clock handling.
>> + *
>> + * Copyright (C) 2012 Marvell
>> + *
>> + * Gregory CLEMENT <gregory.clement at free-electrons.com>
>> + *
>> + * This file is licensed under the terms of the GNU General Public
>> + * License version 2. This program is licensed "as is" without any
>> + * warranty of any kind, whether express or implied.
>> + */
>> +#include <linux/kernel.h>
>> +#include <linux/clk.h>
>
> clk-provider.h includes clk.h, so this is redundant.
>
> Regards,
> Mike
>
--
Gregory Clement, Free Electrons
Kernel, drivers, real-time and embedded Linux
development, consulting, training and support.
http://free-electrons.com
More information about the devicetree-discuss
mailing list