From: Eric Miao <eric.miao-QSEj5FYQhm4dnm+yROfE0A@public.gmane.org>
To: Ben Dooks <ben-linux-elnMNo+KYs3YtjvyW6yDsg@public.gmane.org>
Cc: linux-i2c-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
Jean Delvare <khali-PUYAD+kWke1g9hUCZPvPmw@public.gmane.org>,
linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
Eric Miao <eric.miao-QSEj5FYQhm4dnm+yROfE0A@public.gmane.org>,
Richard Zhao
<richard.zhao-QSEj5FYQhm4dnm+yROfE0A@public.gmane.org>
Subject: Re: [PATCH] i2c: imx: choose the better clock divider
Date: Wed, 15 Jun 2011 16:34:31 +0800 [thread overview]
Message-ID: <BANLkTi=1hMNNWeq_4bcCr_ynT8osY3efaA@mail.gmail.com> (raw)
In-Reply-To: <1308035825-22410-1-git-send-email-eric.miao-QSEj5FYQhm4dnm+yROfE0A@public.gmane.org>
Sorry forgot to include Ben.
Ben, any ideas?
On Tue, Jun 14, 2011 at 3:17 PM, Eric Miao <eric.miao-QSEj5FYQhm4dnm+yROfE0A@public.gmane.org> wrote:
> The original algorithm doesn't perform very well in some cases, e.g.
>
> When the source clock of the I2C controller is 66MHz, and the
> requested rate is 100KHz, it gives a divider of 768 instead of
> the better 640.
>
> Choose a better clock divider so the final clock rate is closer to
> the requested one by comparing the rate distances calculated by
> two adjacent dividers.
>
> Cc: Richard Zhao <richard.zhao-QSEj5FYQhm4dnm+yROfE0A@public.gmane.org>
> Signed-off-by: Eric Miao <eric.miao-QSEj5FYQhm4dnm+yROfE0A@public.gmane.org>
> ---
> drivers/i2c/busses/i2c-imx.c | 46 ++++++++++++++++++++++++++++-------------
> 1 files changed, 31 insertions(+), 15 deletions(-)
>
> diff --git a/drivers/i2c/busses/i2c-imx.c b/drivers/i2c/busses/i2c-imx.c
> index 4c2a62b..cd640ff 100644
> --- a/drivers/i2c/busses/i2c-imx.c
> +++ b/drivers/i2c/busses/i2c-imx.c
> @@ -112,6 +112,8 @@ static u16 __initdata i2c_clk_div[50][2] = {
> { 3072, 0x1E }, { 3840, 0x1F }
> };
>
> +#define I2C_CLK_DIV_NUM ARRAY_SIZE(i2c_clk_div)
> +
> struct imx_i2c_struct {
> struct i2c_adapter adapter;
> struct resource *res;
> @@ -240,22 +242,37 @@ static void i2c_imx_stop(struct imx_i2c_struct *i2c_imx)
> clk_disable(i2c_imx->clk);
> }
>
> -static void __init i2c_imx_set_clk(struct imx_i2c_struct *i2c_imx,
> - unsigned int rate)
> +/* find the index into i2c_clk_div[] array, compare with the two
> + * dividers found, return the one with smaller error
> + */
> +static int find_div(unsigned long i2c_clk_rate, unsigned long rate)
> {
> - unsigned int i2c_clk_rate;
> - unsigned int div;
> + unsigned long div, delta_l, delta_h;
> int i;
>
> - /* Divider value calculation */
> - i2c_clk_rate = clk_get_rate(i2c_imx->clk);
> div = (i2c_clk_rate + rate - 1) / rate;
> +
> if (div < i2c_clk_div[0][0])
> - i = 0;
> - else if (div > i2c_clk_div[ARRAY_SIZE(i2c_clk_div) - 1][0])
> - i = ARRAY_SIZE(i2c_clk_div) - 1;
> - else
> - for (i = 0; i2c_clk_div[i][0] < div; i++);
> + return 0;
> +
> + if (div > i2c_clk_div[I2C_CLK_DIV_NUM - 1][0])
> + return I2C_CLK_DIV_NUM;
> +
> + for (i = 0; i < I2C_CLK_DIV_NUM; i++)
> + if (i2c_clk_div[i][0] > div)
> + break;
> +
> + delta_h = (i2c_clk_rate / i2c_clk_div[i - 1][0]) - rate;
> + delta_l = rate - (i2c_clk_rate / i2c_clk_div[i][0]);
> +
> + return (delta_l < delta_h) ? i : i - 1;
> +}
> +
> +static void __init i2c_imx_set_clk(struct imx_i2c_struct *i2c_imx,
> + unsigned int rate)
> +{
> + unsigned long i2c_clk_rate = clk_get_rate(i2c_imx->clk);
> + int i = find_div(i2c_clk_rate, rate);
>
> /* Store divider value */
> i2c_imx->ifdr = i2c_clk_div[i][1];
> @@ -271,10 +288,9 @@ static void __init i2c_imx_set_clk(struct imx_i2c_struct *i2c_imx,
>
> /* dev_dbg() can't be used, because adapter is not yet registered */
> #ifdef CONFIG_I2C_DEBUG_BUS
> - printk(KERN_DEBUG "I2C: <%s> I2C_CLK=%d, REQ DIV=%d\n",
> - __func__, i2c_clk_rate, div);
> - printk(KERN_DEBUG "I2C: <%s> IFDR[IC]=0x%x, REAL DIV=%d\n",
> - __func__, i2c_clk_div[i][1], i2c_clk_div[i][0]);
> + pr_debug("<%s> I2C_CLK=%ld, REQ RATE=%d, DIV=%d, IFDR[IC]=0x%x\n",
> + __func__, i2c_clk_rate, rate,
> + i2c_clk_div[i][0], i2c_clk_div[i][1]);
> #endif
> }
>
> --
> 1.7.4.1
>
>
next prev parent reply other threads:[~2011-06-15 8:34 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-06-14 7:17 [PATCH] i2c: imx: choose the better clock divider Eric Miao
[not found] ` <1308035825-22410-1-git-send-email-eric.miao-QSEj5FYQhm4dnm+yROfE0A@public.gmane.org>
2011-06-15 8:34 ` Eric Miao [this message]
2011-08-04 16:51 ` Ben Dooks
[not found] ` <20110804165146.GB19115-SMNkleLxa3Z6Wcw2j4pizdi2O/JbrIOy@public.gmane.org>
2011-08-05 16:22 ` Eric Miao
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='BANLkTi=1hMNNWeq_4bcCr_ynT8osY3efaA@mail.gmail.com' \
--to=eric.miao-qsej5fyqhm4dnm+yrofe0a@public.gmane.org \
--cc=ben-linux-elnMNo+KYs3YtjvyW6yDsg@public.gmane.org \
--cc=khali-PUYAD+kWke1g9hUCZPvPmw@public.gmane.org \
--cc=linux-i2c-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
--cc=linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org \
--cc=richard.zhao-QSEj5FYQhm4dnm+yROfE0A@public.gmane.org \
/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;
as well as URLs for NNTP newsgroup(s).