From: Jesper Nilsson <jesper.nilsson@axis.com>
To: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: Camel Guo <camel.guo@axis.com>,
Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>,
<linux-samsung-soc@vger.kernel.org>,
<linux-kernel@vger.kernel.org>, <kernel@axis.com>,
<linux-i2c@vger.kernel.org>,
Alim Akhtar <alim.akhtar@samsung.com>,
<linux-arm-kernel@lists.infradead.org>
Subject: Re: [PATCH] i2c: exynos5: Calculate t_scl_l, t_scl_h according to i2c spec
Date: Mon, 23 Oct 2023 16:06:27 +0200 [thread overview]
Message-ID: <20231023140627.GE11306@axis.com> (raw)
In-Reply-To: <55812f75-3053-4970-bbbe-8ca563f33721@samsung.com>
Thanks Marek for testing. We have been running with this patch for more
than a year now.
The patch has been orphaned for 6 months as Camel no longer works here,
but since I was involved in the creation of it I can resend it.
/Jesper
On Mon, Oct 23, 2023 at 12:34:19PM +0200, Marek Szyprowski wrote:
> On 12.09.2022 10:59, Camel Guo wrote:
> > Previously the duty cycle was divided equally into h_scl_l, t_scl_h.
> > This makes the low period of the SCL clock in Fast Mode is only 1.25us
> > which is way lower than the minimal value (1.3) specified in i2c
> > specification. In order to make sure t_scl_l, t_scl_h always fullfill
> > i2c specification, this commit calculates t_scl_l using this formula:
> >
> > t_scl_l = clk_cycle *
> > ((t_low_min + (scl_clock - t_low_min - t_high_min) / 2) / scl_clock)
> >
> > where:
> > t_low_min is the minimal value of low period of the SCL clock in us;
> > t_high_min is the minimal value of high period of the SCL clock in us;
> > scl_clock is converted from SCL clock frequency into us.
> >
> > Signed-off-by: Camel Guo <camel.guo@axis.com>
>
>
> This sounds reasonable and works fine on all Exynos-based test board I
> have. Feel free to add:
>
> Tested-by: Marek Szyprowski <m.szyprowski@samsung.com>
>
> Reviewed-by: Marek Szyprowski <m.szyprowski@samsung.com>
>
>
> > ---
> > drivers/i2c/busses/i2c-exynos5.c | 34 +++++++++++++++++++++++++++++---
> > 1 file changed, 31 insertions(+), 3 deletions(-)
> >
> > diff --git a/drivers/i2c/busses/i2c-exynos5.c b/drivers/i2c/busses/i2c-exynos5.c
> > index 4a6260d04db2..72bc8adea8d5 100644
> > --- a/drivers/i2c/busses/i2c-exynos5.c
> > +++ b/drivers/i2c/busses/i2c-exynos5.c
> > @@ -267,7 +267,7 @@ static void exynos5_i2c_clr_pend_irq(struct exynos5_i2c *i2c)
> > * exynos5_i2c_set_timing: updates the registers with appropriate
> > * timing values calculated
> > *
> > - * Timing values for operation are calculated against either 100kHz
> > + * Timing values for operation are calculated against 100kHz, 400kHz
> > * or 1MHz controller operating frequency.
> > *
> > * Returns 0 on success, -EINVAL if the cycle length cannot
> > @@ -330,6 +330,23 @@ static int exynos5_i2c_set_timing(struct exynos5_i2c *i2c, bool hs_timings)
> > *
> > * Constraints: 4 <= temp, 0 <= CLK_DIV < 256, 2 <= clk_cycle <= 510
> > *
> > + * To split SCL clock into low, high periods appropriately, one
> > + * proportion factor for each I2C mode is used, which is calculated
> > + * using this formula.
> > + * ```
> > + * ((t_low_min + (scl_clock - t_low_min - t_high_min) / 2) / scl_clock)
> > + * ```
> > + * where:
> > + * t_low_min is the minimal value of low period of the SCL clock in us;
> > + * t_high_min is the minimal value of high period of the SCL clock in us;
> > + * scl_clock is converted from SCL clock frequency into us.
> > + *
> > + * Below are the proportion factors for these I2C modes:
> > + * t_low_min, t_high_min, scl_clock, proportion
> > + * Standard Mode: 4.7us, 4.0us, 10us, 0.535
> > + * Fast Mode: 1.3us, 0.6us, 2.5us, 0.64
> > + * Fast-Plus Mode: 0.5us, 0.26us, 1us, 0.62
> > + *
> > */
> > t_ftl_cycle = (readl(i2c->regs + HSI2C_CONF) >> 16) & 0x7;
> > temp = clkin / op_clk - 8 - t_ftl_cycle;
> > @@ -343,8 +360,19 @@ static int exynos5_i2c_set_timing(struct exynos5_i2c *i2c, bool hs_timings)
> > return -EINVAL;
> > }
> >
> > - t_scl_l = clk_cycle / 2;
> > - t_scl_h = clk_cycle / 2;
> > + /*
> > + * Scale clk_cycle to get t_scl_l using the proption factors for individual I2C modes.
> > + */
> > + if (op_clk <= I2C_MAX_STANDARD_MODE_FREQ)
> > + t_scl_l = clk_cycle * 535 / 1000;
> > + else if (op_clk <= I2C_MAX_FAST_MODE_FREQ)
> > + t_scl_l = clk_cycle * 64 / 100;
> > + else
> > + t_scl_l = clk_cycle * 62 / 100;
> > +
> > + if (t_scl_l > 0xFF)
> > + t_scl_l = 0xFF;
> > + t_scl_h = clk_cycle - t_scl_l;
> > t_start_su = t_scl_l;
> > t_start_hd = t_scl_l;
> > t_stop_su = t_scl_l;
> >
> > base-commit: ce888220d5c7a805e0e155302a318d5d23e62950
>
> Best regards
> --
> Marek Szyprowski, PhD
> Samsung R&D Institute Poland
/^JN - Jesper Nilsson
--
Jesper Nilsson -- jesper.nilsson@axis.com
WARNING: multiple messages have this Message-ID (diff)
From: Jesper Nilsson <jesper.nilsson@axis.com>
To: Marek Szyprowski <m.szyprowski@samsung.com>
Cc: Camel Guo <camel.guo@axis.com>,
Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>,
<linux-samsung-soc@vger.kernel.org>,
<linux-kernel@vger.kernel.org>, <kernel@axis.com>,
<linux-i2c@vger.kernel.org>,
Alim Akhtar <alim.akhtar@samsung.com>,
<linux-arm-kernel@lists.infradead.org>
Subject: Re: [PATCH] i2c: exynos5: Calculate t_scl_l, t_scl_h according to i2c spec
Date: Mon, 23 Oct 2023 16:06:27 +0200 [thread overview]
Message-ID: <20231023140627.GE11306@axis.com> (raw)
In-Reply-To: <55812f75-3053-4970-bbbe-8ca563f33721@samsung.com>
Thanks Marek for testing. We have been running with this patch for more
than a year now.
The patch has been orphaned for 6 months as Camel no longer works here,
but since I was involved in the creation of it I can resend it.
/Jesper
On Mon, Oct 23, 2023 at 12:34:19PM +0200, Marek Szyprowski wrote:
> On 12.09.2022 10:59, Camel Guo wrote:
> > Previously the duty cycle was divided equally into h_scl_l, t_scl_h.
> > This makes the low period of the SCL clock in Fast Mode is only 1.25us
> > which is way lower than the minimal value (1.3) specified in i2c
> > specification. In order to make sure t_scl_l, t_scl_h always fullfill
> > i2c specification, this commit calculates t_scl_l using this formula:
> >
> > t_scl_l = clk_cycle *
> > ((t_low_min + (scl_clock - t_low_min - t_high_min) / 2) / scl_clock)
> >
> > where:
> > t_low_min is the minimal value of low period of the SCL clock in us;
> > t_high_min is the minimal value of high period of the SCL clock in us;
> > scl_clock is converted from SCL clock frequency into us.
> >
> > Signed-off-by: Camel Guo <camel.guo@axis.com>
>
>
> This sounds reasonable and works fine on all Exynos-based test board I
> have. Feel free to add:
>
> Tested-by: Marek Szyprowski <m.szyprowski@samsung.com>
>
> Reviewed-by: Marek Szyprowski <m.szyprowski@samsung.com>
>
>
> > ---
> > drivers/i2c/busses/i2c-exynos5.c | 34 +++++++++++++++++++++++++++++---
> > 1 file changed, 31 insertions(+), 3 deletions(-)
> >
> > diff --git a/drivers/i2c/busses/i2c-exynos5.c b/drivers/i2c/busses/i2c-exynos5.c
> > index 4a6260d04db2..72bc8adea8d5 100644
> > --- a/drivers/i2c/busses/i2c-exynos5.c
> > +++ b/drivers/i2c/busses/i2c-exynos5.c
> > @@ -267,7 +267,7 @@ static void exynos5_i2c_clr_pend_irq(struct exynos5_i2c *i2c)
> > * exynos5_i2c_set_timing: updates the registers with appropriate
> > * timing values calculated
> > *
> > - * Timing values for operation are calculated against either 100kHz
> > + * Timing values for operation are calculated against 100kHz, 400kHz
> > * or 1MHz controller operating frequency.
> > *
> > * Returns 0 on success, -EINVAL if the cycle length cannot
> > @@ -330,6 +330,23 @@ static int exynos5_i2c_set_timing(struct exynos5_i2c *i2c, bool hs_timings)
> > *
> > * Constraints: 4 <= temp, 0 <= CLK_DIV < 256, 2 <= clk_cycle <= 510
> > *
> > + * To split SCL clock into low, high periods appropriately, one
> > + * proportion factor for each I2C mode is used, which is calculated
> > + * using this formula.
> > + * ```
> > + * ((t_low_min + (scl_clock - t_low_min - t_high_min) / 2) / scl_clock)
> > + * ```
> > + * where:
> > + * t_low_min is the minimal value of low period of the SCL clock in us;
> > + * t_high_min is the minimal value of high period of the SCL clock in us;
> > + * scl_clock is converted from SCL clock frequency into us.
> > + *
> > + * Below are the proportion factors for these I2C modes:
> > + * t_low_min, t_high_min, scl_clock, proportion
> > + * Standard Mode: 4.7us, 4.0us, 10us, 0.535
> > + * Fast Mode: 1.3us, 0.6us, 2.5us, 0.64
> > + * Fast-Plus Mode: 0.5us, 0.26us, 1us, 0.62
> > + *
> > */
> > t_ftl_cycle = (readl(i2c->regs + HSI2C_CONF) >> 16) & 0x7;
> > temp = clkin / op_clk - 8 - t_ftl_cycle;
> > @@ -343,8 +360,19 @@ static int exynos5_i2c_set_timing(struct exynos5_i2c *i2c, bool hs_timings)
> > return -EINVAL;
> > }
> >
> > - t_scl_l = clk_cycle / 2;
> > - t_scl_h = clk_cycle / 2;
> > + /*
> > + * Scale clk_cycle to get t_scl_l using the proption factors for individual I2C modes.
> > + */
> > + if (op_clk <= I2C_MAX_STANDARD_MODE_FREQ)
> > + t_scl_l = clk_cycle * 535 / 1000;
> > + else if (op_clk <= I2C_MAX_FAST_MODE_FREQ)
> > + t_scl_l = clk_cycle * 64 / 100;
> > + else
> > + t_scl_l = clk_cycle * 62 / 100;
> > +
> > + if (t_scl_l > 0xFF)
> > + t_scl_l = 0xFF;
> > + t_scl_h = clk_cycle - t_scl_l;
> > t_start_su = t_scl_l;
> > t_start_hd = t_scl_l;
> > t_stop_su = t_scl_l;
> >
> > base-commit: ce888220d5c7a805e0e155302a318d5d23e62950
>
> Best regards
> --
> Marek Szyprowski, PhD
> Samsung R&D Institute Poland
/^JN - Jesper Nilsson
--
Jesper Nilsson -- jesper.nilsson@axis.com
_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
next prev parent reply other threads:[~2023-10-23 14:06 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <CGME20231023103421eucas1p2f1415d69dd979036ef8bcb1a2dc88978@eucas1p2.samsung.com>
2022-09-12 8:59 ` [PATCH] i2c: exynos5: Calculate t_scl_l, t_scl_h according to i2c spec Camel Guo
2022-09-12 8:59 ` Camel Guo
2023-10-23 10:34 ` Marek Szyprowski
2023-10-23 10:34 ` Marek Szyprowski
2023-10-23 14:06 ` Jesper Nilsson [this message]
2023-10-23 14:06 ` Jesper Nilsson
2023-10-23 15:11 ` Wolfram Sang
2023-10-23 15:11 ` Wolfram Sang
2023-10-23 15:12 ` Wolfram Sang
2023-10-23 15:12 ` Wolfram Sang
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=20231023140627.GE11306@axis.com \
--to=jesper.nilsson@axis.com \
--cc=alim.akhtar@samsung.com \
--cc=camel.guo@axis.com \
--cc=kernel@axis.com \
--cc=krzysztof.kozlowski@linaro.org \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-i2c@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-samsung-soc@vger.kernel.org \
--cc=m.szyprowski@samsung.com \
/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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.