From: Marek Vasut <marex@denx.de>
To: u-boot@lists.denx.de
Subject: [U-Boot] [PATCH V2 8/9] mips: ath79: Add AR934x support
Date: Sun, 22 May 2016 04:14:58 +0200 [thread overview]
Message-ID: <57411622.7070100@denx.de> (raw)
In-Reply-To: <BLU436-SMTP18358AF49DD6C2F53636ABFF4D0@phx.gbl>
On 05/22/2016 03:44 AM, Wills Wang wrote:
>
>
> On 05/22/2016 12:49 AM, Marek Vasut wrote:
>> On 05/21/2016 06:22 PM, Wills Wang wrote:
>>>
>>> On 05/07/2016 02:10 AM, Marek Vasut wrote:
>>> [...]
>>>> +
>>>> +static void ar934x_srif_pll_cfg(void __iomem *pll_reg_base, const u32
>>>> srif_val)
>>>> +{
>>>> + u32 reg;
>>>> + do {
>>>> + writel(0x10810f00, pll_reg_base + 0x4);
>>>> + writel(srif_val, pll_reg_base + 0x0);
>>>> + writel(0xd0810f00, pll_reg_base + 0x4);
>>>> + writel(0x03000000, pll_reg_base + 0x8);
>>>> + writel(0xd0800f00, pll_reg_base + 0x4);
>>>> +
>>>> + clrbits_be32(pll_reg_base + 0x8, BIT(30));
>>>> + udelay(5);
>>>> + setbits_be32(pll_reg_base + 0x8, BIT(30));
>>>> + udelay(5);
>>>> +
>>>> + wait_for_bit("clk", pll_reg_base + 0xc, BIT(3), 1, 10, 0);
>>>> +
>>>> + clrbits_be32(pll_reg_base + 0x8, BIT(30));
>>>> + udelay(5);
>>>> +
>>>> + /* Check if CPU SRIF PLL locked. */
>>>> + reg = readl(pll_reg_base + 0x8);
>>>> + reg = (reg & 0x7ffff8) >> 3;
>>>> + } while (reg >= 0x40000);
>>>> +}
>>>> +
>>>> +void ar934x_pll_init(const u16 cpu_mhz, const u16 ddr_mhz, const u16
>>>> ahb_mhz)
>>>> +{
>>> Is it possible to addthe uniform entry for platform PLL initialization,
>>> such as "pll_init"?
>> Yes, this should happen at some point.
>>
>>> [...]
>>> +
>>> +static void ar934x_update_clock(void)
>>> +{
>>> + void __iomem *regs;
>>> + u32 ctrl, cpu, cpupll, ddr, ddrpll;
>>> + u32 cpudiv, ddrdiv, busdiv;
>>> + u32 cpuclk, ddrclk, busclk;
>>> +
>>> + regs = map_physmem(AR71XX_PLL_BASE, AR71XX_PLL_SIZE,
>>> + MAP_NOCACHE);
>>> +
>>> + cpu = readl(regs + AR934X_PLL_CPU_CONFIG_REG);
>>> + ddr = readl(regs + AR934X_PLL_DDR_CONFIG_REG);
>>> + ctrl = readl(regs + AR934X_PLL_CPU_DDR_CLK_CTRL_REG);
>>> +
>>> + cpupll = ar934x_cpupll_to_hz(cpu);
>>> + ddrpll = ar934x_ddrpll_to_hz(ddr);
>>> +
>>> + if (ctrl & AR934X_PLL_CLK_CTRL_CPU_PLL_BYPASS)
>>> + cpuclk = ar934x_get_xtal();
>>> + else if (ctrl & AR934X_PLL_CLK_CTRL_CPUCLK_FROM_CPUPLL)
>>> + cpuclk = cpupll;
>>> + else
>>> + cpuclk = ddrpll;
>>> +
>>> + if (ctrl & AR934X_PLL_CLK_CTRL_DDR_PLL_BYPASS)
>>> + ddrclk = ar934x_get_xtal();
>>> + else if (ctrl & AR934X_PLL_CLK_CTRL_DDRCLK_FROM_DDRPLL)
>>> + ddrclk = ddrpll;
>>> + else
>>> + ddrclk = cpupll;
>>> +
>>> + if (ctrl & AR934X_PLL_CLK_CTRL_AHB_PLL_BYPASS)
>>> + busclk = ar934x_get_xtal();
>>> + else if (ctrl & AR934X_PLL_CLK_CTRL_AHBCLK_FROM_DDRPLL)
>>> + busclk = ddrpll;
>>> + else
>>> + busclk = cpupll;
>>> +
>>> + cpudiv = (ctrl >> AR934X_PLL_CLK_CTRL_CPU_POST_DIV_SHIFT) &
>>> + AR934X_PLL_CLK_CTRL_CPU_POST_DIV_MASK;
>>> + ddrdiv = (ctrl >> AR934X_PLL_CLK_CTRL_DDR_POST_DIV_SHIFT) &
>>> + AR934X_PLL_CLK_CTRL_DDR_POST_DIV_MASK;
>>> + busdiv = (ctrl >> AR934X_PLL_CLK_CTRL_AHB_POST_DIV_SHIFT) &
>>> + AR934X_PLL_CLK_CTRL_AHB_POST_DIV_MASK;
>>> +
>>> + gd->cpu_clk = cpuclk / (cpudiv + 1);
>>> + gd->mem_clk = ddrclk / (ddrdiv + 1);
>>> + gd->bus_clk = busclk / (busdiv + 1);
>>> +}
>>> +
>>> +ulong get_bus_freq(ulong dummy)
>>> +{
>>> + ar934x_update_clock();
>>>
>>> Here i think clock need not be update on each call.
>>>> + return gd->bus_clk;
>>>> +}
>>>> +
>>>> +ulong get_ddr_freq(ulong dummy)
>>>> +{
>>>> + ar934x_update_clock();
>>> Same as above.
>>>> + return gd->mem_clk;
>>>> +}
>>>> +
>>>> +int do_ar934x_showclk(cmd_tbl_t *cmdtp, int flag, int argc, char *
>>>> const argv[])
>>>> +{
>>>> + ar934x_update_clock();
>>> Same as above.
>>>
>>> [...]
>>> +
>>> +static const struct ar934x_mem_config ar934x_mem_config[] = {
>>> + [AR934X_SDRAM] = { 0x7fbe8cd0, 0x959f66a8, 0x33, 0, 0x1f1f },
>>> + [AR934X_DDR1] = { 0x7fd48cd0, 0x99d0e6a8, 0x33, 0, 0x14 },
>>> + [AR934X_DDR2] = { 0xc7d48cd0, 0x9dd0e6a8, 0x33, 0, 0x10012 },
>>> +};
>>> +
>>> +void ar934x_ddr_init(const u16 cpu_mhz, const u16 ddr_mhz, const u16
>>> ahb_mhz)
>>> +{
>>>
>>> Is it possible to use the uniform entry "ddr_init" for DDR
>>> initialization?
>> No, it doesn't allow the board to specify the per-board parameters.
>>
> In ar934x_ddr_init, these parameters was never used.
Yet. These params will be needed when the PMU LDO tuning is added, so
keep those.
>>>> + void __iomem *ddr_regs;
>>>> + const struct ar934x_mem_config *memcfg;
>>>> + int memtype;
>>>> + u32 reg, cycle, ctl;
>>>> +
>>>> + ddr_regs = map_physmem(AR71XX_DDR_CTRL_BASE, AR71XX_DDR_CTRL_SIZE,
>>>> + MAP_NOCACHE);
>>>> +
>>> [...]
>>>> diff --git a/arch/mips/mach-ath79/include/mach/ath79.h
>>>> b/arch/mips/mach-ath79/include/mach/ath79.h
>>>> index 2c6c118..17af082 100644
>>>> --- a/arch/mips/mach-ath79/include/mach/ath79.h
>>>> +++ b/arch/mips/mach-ath79/include/mach/ath79.h
>>>> @@ -143,4 +143,7 @@ static inline int soc_is_qca956x(void)
>>>> int ath79_eth_reset(void);
>>>> int ath79_usb_reset(void);
>>>> +void ar934x_pll_init(const u16 cpu_mhz, const u16 ddr_mhz, const
>>>> u16 ahb_mhz);
>>> I think if we can add a common header for a consistent interface for
>>> platform clock initialization,
>>> such as "clk.h".
>>>> +void ar934x_ddr_init(const u16 cpu_mhz, const u16 ddr_mhz, const u16
>>>> ahb_mhz);
>>>> +
>>> I think it should be moved into mach/ddr.h.
>> Sure, further patches are welcome.
>>
>>>> #endif /* __ASM_MACH_ATH79_H */
>>
>
--
Best regards,
Marek Vasut
next prev parent reply other threads:[~2016-05-22 2:14 UTC|newest]
Thread overview: 19+ messages / expand[flat|nested] mbox.gz Atom feed top
2016-05-06 18:10 [U-Boot] [PATCH 1/9] mips: Add MIPS 74Kc tune Marek Vasut
2016-05-06 18:10 ` [U-Boot] [PATCH 2/9] mips: ath79: Fix ar71xx_regs.h indent Marek Vasut
2016-05-06 18:10 ` [U-Boot] [PATCH 3/9] mips: ath79: Fix compiler warning on const assignment Marek Vasut
2016-05-06 18:10 ` [U-Boot] [PATCH 4/9] mips: ath79: dts: Add generic-ehci node Marek Vasut
2016-05-06 18:10 ` [U-Boot] [PATCH V2 5/9] mips: ath79: Add support for ungating USB on ar933x and ar934x Marek Vasut
2016-05-06 18:10 ` [U-Boot] [PATCH V2 6/9] mips: ath79: dts: Add ethernet MAC nodes for ar933x Marek Vasut
2016-05-06 18:10 ` [U-Boot] [PATCH V2 7/9] mips: ath79: Add support for ungating ethernet on ar933x and ar934x Marek Vasut
2016-05-06 18:10 ` [U-Boot] [PATCH V2 8/9] mips: ath79: Add AR934x support Marek Vasut
2016-05-21 16:22 ` Wills Wang
2016-05-21 16:49 ` Marek Vasut
2016-05-22 1:44 ` Wills Wang
2016-05-22 2:14 ` Marek Vasut [this message]
2016-05-06 18:10 ` [U-Boot] [PATCH 9/9] mips: ath79: Add support for TPLink WDR4300 Marek Vasut
2016-05-21 16:29 ` Wills Wang
2016-05-21 16:47 ` Marek Vasut
2016-05-22 1:49 ` Wills Wang
2016-05-22 2:02 ` Marek Vasut
2016-05-08 11:28 ` [U-Boot] [PATCH 1/9] mips: Add MIPS 74Kc tune Daniel Schwierzeck
2016-05-08 11:58 ` Marek Vasut
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=57411622.7070100@denx.de \
--to=marex@denx.de \
--cc=u-boot@lists.denx.de \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox