lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CACPK8Xct9rWEu40XHWjo_uY9P7dJ=U5LBM+O3KjmjNL+9yD5QQ@mail.gmail.com>
Date:   Fri, 8 Oct 2021 04:06:19 +0000
From:   Joel Stanley <joel@....id.au>
To:     Ryan Chen <ryan_chen@...eedtech.com>
Cc:     Michael Turquette <mturquette@...libre.com>,
        Stephen Boyd <sboyd@...nel.org>,
        Andrew Jeffery <andrew@...id.au>, linux-clk@...r.kernel.org,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: Re: [PATCHv3] clk:aspeed:Fix AST2600 hpll calculate formula

On Sat, 25 Sept 2021 at 02:24, Ryan Chen <ryan_chen@...eedtech.com> wrote:
>

A few notes on process:

> v2 -> v3: change else than if to directly else if
> v1 -> v2: add Fixes commit hash

As this is normally information for reviewers to know what you've
changed since the last version, we normally put this below the --- in
the patch, which means it is not included in the commit message.

Also we put a space between the PATCH and v3 in the subject. If you
use the tools, it will generate this for you:

git format-patch -v3 -1 --to=...

>
> AST2600 HPLL calculate formula [SCU200]
> HPLL Numerator(M): have fixed value depend on SCU strap.
> M = SCU500[10] ? 0x5F : SCU500[8] ? 0xBF : SCU200[12:0]

I recommend adding to the commit message the text from my first review:

>From the datasheet:

CPU frequency selection
000 1.2GHz
001 1.6GHz
010 1.2GHz
011 1.6GHz
100 800MHz
101 800MHz
110 800MHz
111 800MHz

So when the system is running at 800MHz or 1.6GHz, the value for the
numerator (m) in SCU204 is incorrect, and must be overridden.

>
> if SCU500[10] = 1, M=0x5F.
> else if SCU500[10]=0 & SCU500[8]=1, M=0xBF.
> others (SCU510[10]=0 and SCU510[8]=0)
> depend on SCU200[12:0] (default 0x8F) register setting.
>
> HPLL Denumerator (N) =  SCU200[18:13] (default 0x2)
> HPLL Divider (P)         =      SCU200[22:19] (default 0x0)
>
> Fixes: d3d04f6c330a ("clk: Add support for AST2600 SoC")
> Signed-off-by: Ryan Chen <ryan_chen@...eedtech.com>
> ---
>  drivers/clk/clk-ast2600.c | 28 +++++++++++++++++++++++++++-
>  1 file changed, 27 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/clk/clk-ast2600.c b/drivers/clk/clk-ast2600.c
> index 085d0a18b2b6..d30188355aaf 100644
> --- a/drivers/clk/clk-ast2600.c
> +++ b/drivers/clk/clk-ast2600.c
> @@ -169,6 +169,32 @@ static const struct clk_div_table ast2600_div_table[] = {
>  };
>
>  /* For hpll/dpll/epll/mpll */
> +static struct clk_hw *ast2600_calc_hpll(const char *name, u32 val)
> +{
> +       unsigned int mult, div;
> +       u32 hwstrap = readl(scu_g6_base + ASPEED_G6_STRAP1);
> +
> +       if (val & BIT(24)) {
> +               /* Pass through mode */
> +               mult = div = 1;
> +       } else {
> +               /* F = 25Mhz * [(M + 2) / (n + 1)] / (p + 1) */
> +               u32 m = val  & 0x1fff;
> +               u32 n = (val >> 13) & 0x3f;
> +               u32 p = (val >> 19) & 0xf;
> +

Add a comment:

/* If the CPU is running at 800Mhz. */

> +               if (hwstrap & BIT(10))
> +                       m = 0x5F;

/* If the CPU is running at 1600Mhz. */

> +               else if (hwstrap & BIT(8))
> +                       m = 0xBF;


Or you could copy what I suggested in the first patch, and write it
like this, which I think is clear:

ff (hwstrap & BIT(10)) {
    /* CPU running at 800MHz */
   m = 95;
} else if (hwstrap & BIT(10)) {
    /* CPU running at 1.6GHz */
  m  = 191;
} else {
   /* CPU running at 1.2Ghz */
  m = val  & 0x1fff;
}

Please send a v4 with some more comments so we don't need to go back
to the datasheet each time we read this code. Feel free to cut/paste
my suggestion and use it as-is.

Cheers,

Joel

> +
> +               mult = (m + 1) / (n + 1);
> +               div = (p + 1);
> +       }
> +       return clk_hw_register_fixed_factor(NULL, name, "clkin", 0,
> +                       mult, div);
> +};
> +
>  static struct clk_hw *ast2600_calc_pll(const char *name, u32 val)
>  {
>         unsigned int mult, div;
> @@ -716,7 +742,7 @@ static void __init aspeed_g6_cc(struct regmap *map)
>          * and we assume that it is enabled
>          */
>         regmap_read(map, ASPEED_HPLL_PARAM, &val);
> -       aspeed_g6_clk_data->hws[ASPEED_CLK_HPLL] = ast2600_calc_pll("hpll", val);
> +       aspeed_g6_clk_data->hws[ASPEED_CLK_HPLL] = ast2600_calc_hpll("hpll", val);
>
>         regmap_read(map, ASPEED_MPLL_PARAM, &val);
>         aspeed_g6_clk_data->hws[ASPEED_CLK_MPLL] = ast2600_calc_pll("mpll", val);
> --
> 2.17.1
>

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ