[PATCH 1/3] mmc: sunxi: Check the value returned by clk_round_rate

[PATCH 1/3] mmc: sunxi: Check the value returned by clk_round_rate

[Jean-Francois Moine]

clk_round_rate() may return an error. Check it.

Signed-off-by: Jean-Francois Moine <moinejf@free.fr>
---
 drivers/mmc/host/sunxi-mmc.c | 12 +++++++++---
 1 file changed, 9 insertions(+), 3 deletions(-)

diff --git a/drivers/mmc/host/sunxi-mmc.c b/drivers/mmc/host/sunxi-mmc.c
index 2ee4c21..ba647b7 100644
--- a/drivers/mmc/host/sunxi-mmc.c
+++ b/drivers/mmc/host/sunxi-mmc.c
@@ -656,7 +656,8 @@ static int sunxi_mmc_oclk_onoff(struct sunxi_mmc_host *host, u32 oclk_en)
 static int sunxi_mmc_clk_set_rate(struct sunxi_mmc_host *host,
 				  struct mmc_ios *ios)
 {
-	u32 rate, oclk_dly, rval, sclk_dly;
+	long rate;
+	u32 oclk_dly, rval, sclk_dly;
 	u32 clock = ios->clock;
 	int ret;
 
@@ -666,13 +667,18 @@ static int sunxi_mmc_clk_set_rate(struct sunxi_mmc_host *host,
 		clock <<= 1;
 
 	rate = clk_round_rate(host->clk_mmc, clock);
-	dev_dbg(mmc_dev(host->mmc), "setting clk to %d, rounded %d\n",
+	if (rate < 0) {
+		dev_err(mmc_dev(host->mmc), "error rounding clk to %d: %ld\n",
+			clock, rate);
+		return rate;
+	}
+	dev_dbg(mmc_dev(host->mmc), "setting clk to %d, rounded %ld\n",
 		clock, rate);
 
 	/* setting clock rate */
 	ret = clk_set_rate(host->clk_mmc, rate);
 	if (ret) {
-		dev_err(mmc_dev(host->mmc), "error setting clk to %d: %d\n",
+		dev_err(mmc_dev(host->mmc), "error setting clk to %ld: %d\n",
 			rate, ret);
 		return ret;
 	}
-- 
2.9.2

[PATCH 2/3] mmc: sunxi: Set the 'New Timing' register for 8 bits DDR transfers

[Jean-Francois Moine]

The 'new timing mode' with 8 bits DDR works correctly when the NewTiming
register is set.

Signed-off-by: Jean-Francois Moine <moinejf@free.fr>
---
Note about the 'new timing mode'.

This patch assumes that, when the new mode is used, the clock driver
sets the mode select in the MMC clock and multiplies the clock rate
by 2:
- MMC side:
  - with a timing 8 bits DDR at 50MHz, the MMC driver calls
    clk_set_rate() with a rate 50*2 = 100MHz,
- clock side:
  - the clock driver sets the hardware MMC clock to 100*2 = 200MHz,
  - setting the 'mode select' of the hardware MMC clock divides the
    rate by 2,
- MMC side:
  - setting the MMC clock divider register to 1 divides the rate by 2.
So, the final rate is 50MHz.

For this to work, the clock driver must know when the 'mode select'
is used.
Instead of using a new special clock function, in my driver, the
'mode select' capable clocks are declared as such with an associated
rate (actually, mmc2 in A83T, and mmc{1,2} in H3).
When this rate is asked for (100MHz), the mode select is set.
This works without change in the MMC driver. If a 100MHz rate would be
used without 'mode select' for such clocks, an other mechanism should
be found (information in the low bits of the rate?).
---
 drivers/mmc/host/sunxi-mmc.c | 22 ++++++++++++++++++----
 1 file changed, 18 insertions(+), 4 deletions(-)

diff --git a/drivers/mmc/host/sunxi-mmc.c b/drivers/mmc/host/sunxi-mmc.c
index ba647b7..7f9c31a 100644
--- a/drivers/mmc/host/sunxi-mmc.c
+++ b/drivers/mmc/host/sunxi-mmc.c
@@ -64,6 +64,7 @@
 #define SDXC_REG_CBCR	(0x48) /* SMC CIU Byte Count Register */
 #define SDXC_REG_BBCR	(0x4C) /* SMC BIU Byte Count Register */
 #define SDXC_REG_DBGC	(0x50) /* SMC Debug Enable Register */
+#define SDXC_REG_NTSR	(0x5c) /* SMC NewTiming Set Register */
 #define SDXC_REG_HWRST	(0x78) /* SMC Card Hardware Reset for Register */
 #define SDXC_REG_DMAC	(0x80) /* SMC IDMAC Control Register */
 #define SDXC_REG_DLBA	(0x84) /* SMC IDMAC Descriptor List Base Addre */
@@ -171,6 +172,9 @@
 #define SDXC_SEND_AUTO_STOPCCSD		BIT(9)
 #define SDXC_CEATA_DEV_IRQ_ENABLE	BIT(10)
 
+/* NewTiming Set Register */
+#define SDXC_NEWMODE_ENABLE		BIT(31)
+
 /* IDMA controller bus mod bit field */
 #define SDXC_IDMAC_SOFT_RESET		BIT(0)
 #define SDXC_IDMAC_FIX_BURST		BIT(1)
@@ -661,10 +665,20 @@ static int sunxi_mmc_clk_set_rate(struct sunxi_mmc_host *host,
 	u32 clock = ios->clock;
 	int ret;
 
-	/* 8 bit DDR requires a higher module clock */
-	if (ios->timing == MMC_TIMING_MMC_DDR52 &&
-	    ios->bus_width == MMC_BUS_WIDTH_8)
-		clock <<= 1;
+	/*
+	 * 8 bit DDR requires a higher module clock
+	 * and the 'new mode'
+	 */
+	if (ios->bus_width == MMC_BUS_WIDTH_8) {	/* eMMC */
+		rval = mmc_readl(host, REG_NTSR);
+		if (ios->timing == MMC_TIMING_MMC_DDR52) {
+			clock <<= 1;
+			rval |= SDXC_NEWMODE_ENABLE;
+		} else {
+			rval &= ~SDXC_NEWMODE_ENABLE;
+		}
+		mmc_writel(host, REG_NTSR, rval);
+	}
 
 	rate = clk_round_rate(host->clk_mmc, clock);
 	if (rate < 0) {
-- 
2.9.2

[PATCH 3/3] mmc: sunxi: Add support to the Allwinner A83T

[Jean-Francois Moine]

The rate of the PLL-PERIPH clock is usually set to 1.2GHz in the A83T.
This patch sets the phase delays of the output and sample clocks
accordingly.

Signed-off-by: Jean-Francois Moine <moinejf@free.fr>
---
Note: The impacted phase delays are only for 50MHz.
The phase delays are not used in 50MHz 8 bits DDR (new timing mode).
---
 drivers/mmc/host/sunxi-mmc.c | 11 +++++++++++
 1 file changed, 11 insertions(+), 0 deletions(-)

diff --git a/drivers/mmc/host/sunxi-mmc.c b/drivers/mmc/host/sunxi-mmc.c
index 7f9c31a..e161a64 100644
--- a/drivers/mmc/host/sunxi-mmc.c
+++ b/drivers/mmc/host/sunxi-mmc.c
@@ -961,6 +961,7 @@ static int sunxi_mmc_card_busy(struct mmc_host *mmc)
 static const struct of_device_id sunxi_mmc_of_match[] = {
 	{ .compatible = "allwinner,sun4i-a10-mmc", },
 	{ .compatible = "allwinner,sun5i-a13-mmc", },
+	{ .compatible = "allwinner,sun8i-a83t-mmc", },
 	{ .compatible = "allwinner,sun9i-a80-mmc", },
 	{ /* sentinel */ }
 };
@@ -986,6 +987,14 @@ static const struct sunxi_mmc_clk_delay sunxi_mmc_clk_delays[] = {
 	[SDXC_CLK_50M_DDR_8BIT]	= { .output =  90, .sample = 180 },
 };
 
+static const struct sunxi_mmc_clk_delay sun8i_a83t_mmc_clk_delays[] = {
+	[SDXC_CLK_400K]		= { .output = 180, .sample = 180 },
+	[SDXC_CLK_25M]		= { .output = 180, .sample =  75 },
+	[SDXC_CLK_50M]		= { .output =  90, .sample = 105 },
+	[SDXC_CLK_50M_DDR]	= { .output =  60, .sample = 120 },
+	[SDXC_CLK_50M_DDR_8BIT]	= { .output =  180, .sample = 180 },
+};
+
 static const struct sunxi_mmc_clk_delay sun9i_mmc_clk_delays[] = {
 	[SDXC_CLK_400K]		= { .output = 180, .sample = 180 },
 	[SDXC_CLK_25M]		= { .output = 180, .sample =  75 },
@@ -1007,6 +1016,8 @@ static int sunxi_mmc_resource_request(struct sunxi_mmc_host *host,
 
 	if (of_device_is_compatible(np, "allwinner,sun9i-a80-mmc"))
 		host->clk_delays = sun9i_mmc_clk_delays;
+	else if (of_device_is_compatible(np, "allwinner,sun8i-a83t-mmc"))
+		host->clk_delays = sun8i_a83t_mmc_clk_delays;
 	else
 		host->clk_delays = sunxi_mmc_clk_delays;
 
-- 
2.9.2

[PATCH 0/3] mmc: sunxi: Changes in the host driver

[Jean-Francois Moine]

While testing the eMMC access in a Banana Pi M3, I found and fixed
some problems:
- the driver was trying to set a strange rate when clk_round_rate()
  was returning an error
- the 'new timing mode' could not work because a register was not set
- with a parent clock at 1.2GHz, the output and sample phases were
  not correct for the A83T

Jean-Francois Moine (3):
  mmc: sunxi: Check the value returned by clk_round_rate
  mmc: sunxi: Set the 'New Timing' register for 8 bits DDR transfers
  mmc: sunxi: Add support to the Allwinner A83T

 drivers/mmc/host/sunxi-mmc.c | 45 +++++++++++++++++++++++++++++++++++---------
 1 file changed, 38 insertions(+), 7 deletions(-)

-- 
2.9.2

[PATCH 1/3] mmc: sunxi: Check the value returned by clk_round_rate

[Maxime Ripard]

On Wed, Jul 20, 2016 at 08:01:47PM +0200, Jean-Francois Moine wrote:
> clk_round_rate() may return an error. Check it.
> 
> Signed-off-by: Jean-Francois Moine <moinejf@free.fr>

Acked-by: Maxime Ripard <maxime.ripard@free-electrons.com>

Thanks!
Maxime

-- 
Maxime Ripard, Free Electrons
Embedded Linux and Kernel engineering
http://free-electrons.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: not available
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20160721/4f538229/attachment.sig>

[PATCH 2/3] mmc: sunxi: Set the 'New Timing' register for 8 bits DDR transfers

[Maxime Ripard]

Hi,

On Wed, Jul 20, 2016 at 08:16:28PM +0200, Jean-Francois Moine wrote:
> The 'new timing mode' with 8 bits DDR works correctly when the NewTiming
> register is set.

What does that mode brings to the table?

> 
> Signed-off-by: Jean-Francois Moine <moinejf@free.fr>
> ---
> Note about the 'new timing mode'.
> 
> This patch assumes that, when the new mode is used, the clock driver
> sets the mode select in the MMC clock and multiplies the clock rate
> by 2:
> - MMC side:
>   - with a timing 8 bits DDR at 50MHz, the MMC driver calls
>     clk_set_rate() with a rate 50*2 = 100MHz,
> - clock side:
>   - the clock driver sets the hardware MMC clock to 100*2 = 200MHz,
>   - setting the 'mode select' of the hardware MMC clock divides the
>     rate by 2,
> - MMC side:
>   - setting the MMC clock divider register to 1 divides the rate by 2.
> So, the final rate is 50MHz.

What happens if you actually want to set it to 100MHz?

Maxime

-- 
Maxime Ripard, Free Electrons
Embedded Linux and Kernel engineering
http://free-electrons.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: not available
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20160721/39cd3e0c/attachment.sig>

[PATCH 3/3] mmc: sunxi: Add support to the Allwinner A83T

[Maxime Ripard]

Maxime

On Wed, Jul 20, 2016 at 08:28:47PM +0200, Jean-Francois Moine wrote:
> The rate of the PLL-PERIPH clock is usually set to 1.2GHz in the A83T.

Uh? The datasheet says to set it to 600MHz.

> This patch sets the phase delays of the output and sample clocks
> accordingly.
> 
> Signed-off-by: Jean-Francois Moine <moinejf@free.fr>
> ---
> Note: The impacted phase delays are only for 50MHz.
> The phase delays are not used in 50MHz 8 bits DDR (new timing mode).

Actually, they seem to be, in the new timing mode register.

Maxime

-- 
Maxime Ripard, Free Electrons
Embedded Linux and Kernel engineering
http://free-electrons.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: not available
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20160721/50da17d5/attachment.sig>

[PATCH 3/3] mmc: sunxi: Add support to the Allwinner A83T

[Jean-Francois Moine]

On Thu, 21 Jul 2016 10:58:38 +0200
Maxime Ripard <maxime.ripard@free-electrons.com> wrote:

> On Wed, Jul 20, 2016 at 08:28:47PM +0200, Jean-Francois Moine wrote:
> > The rate of the PLL-PERIPH clock is usually set to 1.2GHz in the A83T.
> 
> Uh? The datasheet says to set it to 600MHz.

Right. But the driver of the SDK for the Banana Pi M3 (A83T) sets it
to 1.2GHz.

> > This patch sets the phase delays of the output and sample clocks
> > accordingly.
> > 
> > Signed-off-by: Jean-Francois Moine <moinejf@free.fr>
> > ---
> > Note: The impacted phase delays are only for 50MHz.
> > The phase delays are not used in 50MHz 8 bits DDR (new timing mode).
> 
> Actually, they seem to be, in the new timing mode register.

In the SDK driver, nothing is set in the new timing mode register.
Anyway, the eMMC works fine in both the Banana Pis M2+ and M3
with no delay.

-- 
Ken ar c'henta?	|	      ** Breizh ha Linux atav! **
Jef		|		http://moinejf.free.fr/

[PATCH 2/3] mmc: sunxi: Set the 'New Timing' register for 8 bits DDR transfers

[Jean-Francois Moine]

On Thu, 21 Jul 2016 10:56:15 +0200
Maxime Ripard <maxime.ripard@free-electrons.com> wrote:

> On Wed, Jul 20, 2016 at 08:16:28PM +0200, Jean-Francois Moine wrote:
> > The 'new timing mode' with 8 bits DDR works correctly when the NewTiming
> > register is set.
> 
> What does that mode brings to the table?

>From my tests, the eMMC of the Banana Pi M3 (A83T) cannot work when the
new mode is not used.

> > 
> > Signed-off-by: Jean-Francois Moine <moinejf@free.fr>
> > ---
> > Note about the 'new timing mode'.
> > 
> > This patch assumes that, when the new mode is used, the clock driver
> > sets the mode select in the MMC clock and multiplies the clock rate
> > by 2:
> > - MMC side:
> >   - with a timing 8 bits DDR at 50MHz, the MMC driver calls
> >     clk_set_rate() with a rate 50*2 = 100MHz,
> > - clock side:
> >   - the clock driver sets the hardware MMC clock to 100*2 = 200MHz,
> >   - setting the 'mode select' of the hardware MMC clock divides the
> >     rate by 2,
> > - MMC side:
> >   - setting the MMC clock divider register to 1 divides the rate by 2.
> > So, the final rate is 50MHz.
> 
> What happens if you actually want to set it to 100MHz?

There is no SDXC_CLK_100M in the mainline driver, and 100MHz is asked
only for 8 bits DDR at 50MHz.

-- 
Ken ar c'henta?	|	      ** Breizh ha Linux atav! **
Jef		|		http://moinejf.free.fr/

[PATCH 3/3] mmc: sunxi: Add support to the Allwinner A83T

[Jean-Francois Moine]

On Thu, 21 Jul 2016 11:18:51 +0200
Jean-Francois Moine <moinejf@free.fr> wrote:

> > On Wed, Jul 20, 2016 at 08:28:47PM +0200, Jean-Francois Moine wrote:
> > > The rate of the PLL-PERIPH clock is usually set to 1.2GHz in the A83T.
> > 
> > Uh? The datasheet says to set it to 600MHz.
> 
> Right. But the driver of the SDK for the Banana Pi M3 (A83T) sets it
> to 1.2GHz.

I tested again, and, with the pll-periph at 600MHz, both the SDcard and
the eMMC are working fine.
So, you may forget about this patch, sorry for the noise.

-- 
Ken ar c'henta?	|	      ** Breizh ha Linux atav! **
Jef		|		http://moinejf.free.fr/

[PATCH 2/3] mmc: sunxi: Set the 'New Timing' register for 8 bits DDR transfers

[Maxime Ripard]

On Thu, Jul 21, 2016 at 11:26:55AM +0200, Jean-Francois Moine wrote:
> On Thu, 21 Jul 2016 10:56:15 +0200
> Maxime Ripard <maxime.ripard@free-electrons.com> wrote:
> 
> > On Wed, Jul 20, 2016 at 08:16:28PM +0200, Jean-Francois Moine wrote:
> > > The 'new timing mode' with 8 bits DDR works correctly when the NewTiming
> > > register is set.
> > 
> > What does that mode brings to the table?
> 
> From my tests, the eMMC of the Banana Pi M3 (A83T) cannot work when the
> new mode is not used.
> 
> > > 
> > > Signed-off-by: Jean-Francois Moine <moinejf@free.fr>
> > > ---
> > > Note about the 'new timing mode'.
> > > 
> > > This patch assumes that, when the new mode is used, the clock driver
> > > sets the mode select in the MMC clock and multiplies the clock rate
> > > by 2:
> > > - MMC side:
> > >   - with a timing 8 bits DDR at 50MHz, the MMC driver calls
> > >     clk_set_rate() with a rate 50*2 = 100MHz,
> > > - clock side:
> > >   - the clock driver sets the hardware MMC clock to 100*2 = 200MHz,
> > >   - setting the 'mode select' of the hardware MMC clock divides the
> > >     rate by 2,
> > > - MMC side:
> > >   - setting the MMC clock divider register to 1 divides the rate by 2.
> > > So, the final rate is 50MHz.
> > 
> > What happens if you actually want to set it to 100MHz?
> 
> There is no SDXC_CLK_100M in the mainline driver, and 100MHz is asked
> only for 8 bits DDR at 50MHz.

You're missing the point.

clk_set_rate is supposed to apply a rate as close as possible as
requested, there's no reason why you would request a rate twice as
high as need.

You want to switch the clock from one mode to another, fine, create a
new function for that. But don't hack an existing one.

Maxime

-- 
Maxime Ripard, Free Electrons
Embedded Linux and Kernel engineering
http://free-electrons.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: not available
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20160729/9761ce2a/attachment.sig>

[PATCH 2/3] mmc: sunxi: Set the 'New Timing' register for 8 bits DDR transfers

[Maxime Ripard]

On Thu, Jul 21, 2016 at 11:26:55AM +0200, Jean-Francois Moine wrote:
> On Thu, 21 Jul 2016 10:56:15 +0200
> Maxime Ripard <maxime.ripard@free-electrons.com> wrote:
> 
> > On Wed, Jul 20, 2016 at 08:16:28PM +0200, Jean-Francois Moine wrote:
> > > The 'new timing mode' with 8 bits DDR works correctly when the NewTiming
> > > register is set.
> > 
> > What does that mode brings to the table?
> 
> From my tests, the eMMC of the Banana Pi M3 (A83T) cannot work when the
> new mode is not used.

That's odd. The one in the Pine64 seems to work just fine, and yet
there's only the new mode on the A64.

Maxime

-- 
Maxime Ripard, Free Electrons
Embedded Linux and Kernel engineering
http://free-electrons.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: not available
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20160729/e29fe3f0/attachment.sig>

[PATCH 2/3] mmc: sunxi: Set the 'New Timing' register for 8 bits DDR transfers

[Jean-Francois Moine]

On Fri, 29 Jul 2016 21:17:30 +0200
Maxime Ripard <maxime.ripard@free-electrons.com> wrote:

> > > What happens if you actually want to set it to 100MHz?
> > 
> > There is no SDXC_CLK_100M in the mainline driver, and 100MHz is asked
> > only for 8 bits DDR at 50MHz.
> 
> You're missing the point.
> 
> clk_set_rate is supposed to apply a rate as close as possible as
> requested, there's no reason why you would request a rate twice as
> high as need.
> 
> You want to switch the clock from one mode to another, fine, create a
> new function for that. But don't hack an existing one.

I will not change the core clock stuff for this marginal case.
Flagging the clock as 'new mode capable' is enough.
Anyway, setting the 'new mode' bit to the clock divides the rate by
two, so, there is no reason to not use it.

-- 
Ken ar c'henta?	|	      ** Breizh ha Linux atav! **
Jef		|		http://moinejf.free.fr/

[PATCH 2/3] mmc: sunxi: Set the 'New Timing' register for 8 bits DDR transfers

[Jean-Francois Moine]

On Fri, 29 Jul 2016 21:36:34 +0200
Maxime Ripard <maxime.ripard@free-electrons.com> wrote:

> On Thu, Jul 21, 2016 at 11:26:55AM +0200, Jean-Francois Moine wrote:
> > On Thu, 21 Jul 2016 10:56:15 +0200
> > Maxime Ripard <maxime.ripard@free-electrons.com> wrote:
> > 
> > > On Wed, Jul 20, 2016 at 08:16:28PM +0200, Jean-Francois Moine wrote:
> > > > The 'new timing mode' with 8 bits DDR works correctly when the NewTiming
> > > > register is set.
> > > 
> > > What does that mode brings to the table?
> > 
> > From my tests, the eMMC of the Banana Pi M3 (A83T) cannot work when the
> > new mode is not used.
> 
> That's odd. The one in the Pine64 seems to work just fine, and yet
> there's only the new mode on the A64.

I spent 2 weeks on this problem. You may try it by yourself.

-- 
Ken ar c'henta?	|	      ** Breizh ha Linux atav! **
Jef		|		http://moinejf.free.fr/

[linux-sunxi] Re: [PATCH 2/3] mmc: sunxi: Set the 'New Timing' register for 8 bits DDR transfers

[Hans de Goede]

Hi,

On 29-07-16 21:17, Maxime Ripard wrote:
> On Thu, Jul 21, 2016 at 11:26:55AM +0200, Jean-Francois Moine wrote:
>> On Thu, 21 Jul 2016 10:56:15 +0200
>> Maxime Ripard <maxime.ripard@free-electrons.com> wrote:
>>
>>> On Wed, Jul 20, 2016 at 08:16:28PM +0200, Jean-Francois Moine wrote:
>>>> The 'new timing mode' with 8 bits DDR works correctly when the NewTiming
>>>> register is set.
>>>
>>> What does that mode brings to the table?
>>
>> From my tests, the eMMC of the Banana Pi M3 (A83T) cannot work when the
>> new mode is not used.
>>
>>>>
>>>> Signed-off-by: Jean-Francois Moine <moinejf@free.fr>
>>>> ---
>>>> Note about the 'new timing mode'.
>>>>
>>>> This patch assumes that, when the new mode is used, the clock driver
>>>> sets the mode select in the MMC clock and multiplies the clock rate
>>>> by 2:
>>>> - MMC side:
>>>>   - with a timing 8 bits DDR at 50MHz, the MMC driver calls
>>>>     clk_set_rate() with a rate 50*2 = 100MHz,
>>>> - clock side:
>>>>   - the clock driver sets the hardware MMC clock to 100*2 = 200MHz,
>>>>   - setting the 'mode select' of the hardware MMC clock divides the
>>>>     rate by 2,
>>>> - MMC side:
>>>>   - setting the MMC clock divider register to 1 divides the rate by 2.
>>>> So, the final rate is 50MHz.
>>>
>>> What happens if you actually want to set it to 100MHz?
>>
>> There is no SDXC_CLK_100M in the mainline driver, and 100MHz is asked
>> only for 8 bits DDR at 50MHz.
>
> You're missing the point.
>
> clk_set_rate is supposed to apply a rate as close as possible as
> requested, there's no reason why you would request a rate twice as
> high as need.

On all SoCs which support it, requesting DDR-8bit needs a twice as high
clock as the mmc base clock (due to the DDR).

If you look at the current code you will see that what actually happens
is that we pass twice as high clock as the mmc-subsys asks for to
clk_set_rate and set the controllers *internal* divider to 2 (instead of
the usual 1), this is not new and not introduced by Jean-Francois' patch.

All Jean-Francois' patch adds is the setting of the new SDXC_REG_NTSR
reg msb bit when we do this. Note no changes to the clock rate calculations
are made. If this fixes the eMMC on A83T boards this seems like a good
patch to me, but IMHO we should only write the SDXC_REG_NTSR on the A83t
and not everywhere as Jean-Francois' patch does.

Jean-Francois, can you submit a v2 of your patch and make the writing of
SDXC_REG_NTSR depend on a new sun8i-a83t-mmc compatible ?

Also you should probably drop the bits about the clock stuff from the
commit message as that just seems to confuse people.

Regards,

Hans

[linux-sunxi] Re: [PATCH 2/3] mmc: sunxi: Set the 'New Timing' register for 8 bits DDR transfers

[Jean-Francois Moine]

On Sat, 30 Jul 2016 12:19:03 +0200
Hans de Goede <hdegoede@redhat.com> wrote:

> Jean-Francois, can you submit a v2 of your patch and make the writing of
> SDXC_REG_NTSR depend on a new sun8i-a83t-mmc compatible ?
> 
> Also you should probably drop the bits about the clock stuff from the
> commit message as that just seems to confuse people.

Hi Hans,

I submitted a new patch (sorry, I forgot the history), but it asks for
some explanation.

- in the old timings, the phase delays are set in the clock.
  That's why there is a function clk_set_phase() which is called from
  the MMC side.

- in the new timings, the delays are in the MMC register SDXC_REG_NTSR
  only.
  In this case, the function clk_set_phase() is of no use
  (also, by test, it seems that the phase delays set by hardware
   reset do work, so, there is no need to set them), but,

- there is a bit in the clock telling that the new timings are used,
  i.e. that the phase delays of the clock must be ignored.

  Not setting this bit prevents the device to work (at least in the
  A83T, the H3 seems more helpful).

-- 
Ken ar c'henta?	|	      ** Breizh ha Linux atav! **
Jef		|		http://moinejf.free.fr/

[linux-sunxi] Re: [PATCH 2/3] mmc: sunxi: Set the 'New Timing' register for 8 bits DDR transfers

[Hans de Goede]

Hi,

On 01-08-16 15:52, Jean-Francois Moine wrote:
> On Sat, 30 Jul 2016 12:19:03 +0200
> Hans de Goede <hdegoede@redhat.com> wrote:
>
>> Jean-Francois, can you submit a v2 of your patch and make the writing of
>> SDXC_REG_NTSR depend on a new sun8i-a83t-mmc compatible ?
>>
>> Also you should probably drop the bits about the clock stuff from the
>> commit message as that just seems to confuse people.
>
> Hi Hans,
>
> I submitted a new patch (sorry, I forgot the history), but it asks for
> some explanation.
>
> - in the old timings, the phase delays are set in the clock.
>   That's why there is a function clk_set_phase() which is called from
>   the MMC side.
>
> - in the new timings, the delays are in the MMC register SDXC_REG_NTSR
>   only.
>   In this case, the function clk_set_phase() is of no use
>   (also, by test, it seems that the phase delays set by hardware
>    reset do work, so, there is no need to set them), but,
>
> - there is a bit in the clock telling that the new timings are used,
>   i.e. that the phase delays of the clock must be ignored.
>
>   Not setting this bit prevents the device to work (at least in the
>   A83T, the H3 seems more helpful).

Thanks for the explanation, it would be good to put this in the commit
msg of v3 of the patch.

Regards,

Hans