[U-Boot] [PATCH v4 25/28] arm: socfpga: arria10: Added drivers for Arria10 clock manager
Marek Vasut
marex at denx.de
Fri Feb 17 21:10:59 UTC 2017
On 02/17/2017 09:56 AM, Ley Foon Tan wrote:
> On Mon, Jan 23, 2017 at 12:18 PM, Marek Vasut <marex at denx.de> wrote:
>> On 01/10/2017 06:20 AM, Chee Tien Fong wrote:
>>> From: Tien Fong Chee <tien.fong.chee at intel.com>
>>>
>>> The drivers is restructured such common functions, gen5 functions, and
>>> arria10 functions are moved to clock_manager.c, clock_manager_gen5 and
>>> clock_manager_arria10 respectively.
>>>
>>> Signed-off-by: Tien Fong Chee <tien.fong.chee at intel.com>
>>> Cc: Marek Vasut <marex at denx.de>
>>> Cc: Dinh Nguyen <dingnuyen at kernel.org>
>>> Cc: Ching Liang See <chin.liang.see at intel.com>
>>> Cc: Tien Fong <skywindctf at gmail.com>
>>> ---
>>> arch/arm/mach-socfpga/clock_manager.c | 752 +++++++---------
>>> arch/arm/mach-socfpga/clock_manager_arria10.c | 954 +++++++++++++++++++++
>>> .../{clock_manager.c => clock_manager_gen5.c} | 240 +-----
>>> arch/arm/mach-socfpga/include/mach/clock_manager.h | 356 ++++++--
>>> 4 files changed, 1573 insertions(+), 729 deletions(-)
>>> create mode 100644 arch/arm/mach-socfpga/clock_manager_arria10.c
>>> copy arch/arm/mach-socfpga/{clock_manager.c => clock_manager_gen5.c} (62%)
>>>
>>> diff --git a/arch/arm/mach-socfpga/clock_manager.c b/arch/arm/mach-socfpga/clock_manager.c
>>> index aa71636..d209f7d 100644
>>> --- a/arch/arm/mach-socfpga/clock_manager.c
>>> +++ b/arch/arm/mach-socfpga/clock_manager.c
>>> @@ -1,5 +1,5 @@
>>> /*
>>> - * Copyright (C) 2013 Altera Corporation <www.altera.com>
>>> + * Copyright (C) 2013-2016 Altera Corporation <www.altera.com>
>>> *
>>> * SPDX-License-Identifier: GPL-2.0+
>>> */
>>> @@ -7,416 +7,287 @@
>>> #include <common.h>
>>> #include <asm/io.h>
>>> #include <asm/arch/clock_manager.h>
>>> +#include <fdtdec.h>
>>>
>>> DECLARE_GLOBAL_DATA_PTR;
>>>
>>> +/* Function prototypes */
>>> +/* Common prototypes */
>>> +unsigned int cm_get_l4_sp_clk_hz(void);
>>> +unsigned int cm_get_qspi_controller_clk_hz(void);
>>> +unsigned int cm_get_mmc_controller_clk_hz(void);
>>> +unsigned int cm_get_spi_controller_clk_hz(void);
>>> +static void cm_print_clock_quick_summary(void);
>>> +int do_showclocks(cmd_tbl_t *cmdtp, int flag, int argc, char * const argv[]);
>>> +void cm_wait_for_lock(uint32_t mask);
>>> +void cm_wait_for_fsm(void);
>>> +unsigned int cm_get_main_vco_clk_hz(void);
>>> +unsigned int cm_get_per_vco_clk_hz(void);
>>> +unsigned long cm_get_mpu_clk_hz(void);
>>> +
>>> static const struct socfpga_clock_manager *clock_manager_base =
>>> (struct socfpga_clock_manager *)SOCFPGA_CLKMGR_ADDRESS;
>>>
>>> -static void cm_wait_for_lock(uint32_t mask)
>>> +/* Common functions */
>>> +int set_cpu_clk_info(void)
>>> {
>>> - register uint32_t inter_val;
>>> - uint32_t retry = 0;
>>> - do {
>>> - inter_val = readl(&clock_manager_base->inter) & mask;
>>> - if (inter_val == mask)
>>> - retry++;
>>> - else
>>> - retry = 0;
>>> - if (retry >= 10)
>>> - break;
>>> - } while (1);
>>> -}
>>> + /* Calculate the clock frequencies required for drivers */
>>> + cm_get_l4_sp_clk_hz();
>>> + cm_get_mmc_controller_clk_hz();
>>>
>>> -/* function to poll in the fsm busy bit */
>>> -static void cm_wait_for_fsm(void)
>>> -{
>>> - while (readl(&clock_manager_base->stat) & CLKMGR_STAT_BUSY)
>>> - ;
>>> -}
>>> + gd->bd->bi_arm_freq = cm_get_mpu_clk_hz() / 1000000;
>>> + gd->bd->bi_dsp_freq = 0;
>>>
>>> -/*
>>> - * function to write the bypass register which requires a poll of the
>>> - * busy bit
>>> - */
>>> -static void cm_write_bypass(uint32_t val)
>>> -{
>>> - writel(val, &clock_manager_base->bypass);
>>> - cm_wait_for_fsm();
>>> -}
>>> +#if defined(CONFIG_TARGET_SOCFPGA_GEN5)
>>> + gd->bd->bi_ddr_freq = cm_get_sdram_clk_hz() / 1000000;
>>> +#elif defined(CONFIG_TARGET_SOCFPGA_ARRIA10)
>>> + gd->bd->bi_ddr_freq = 0;
>>
>> What ? This cannot work ...
> For A10, ddr controller is on fpga side and we don't need the ddr freq
> for ddr setup.
And if the user wants to know ? Any chance we can pass this info to the
user ? I presume this is still a hard-IP controller, not some soft core
thing ...
[...]
--
Best regards,
Marek Vasut
More information about the U-Boot
mailing list