2016-04-19 08:17:23

by Peng Fan

Subject: [PATCH 1/3] arm: dts: imx6qdl: add clocks property for ocotp node

Add clocks property for ocotp node.

Signed-off-by: Peng Fan <[email protected]>
Cc: Shawn Guo <[email protected]>
Cc: Sascha Hauer <[email protected]>
Cc: Rob Herring <[email protected]>
---
arch/arm/boot/dts/imx6qdl.dtsi | 1 +
1 file changed, 1 insertion(+)

diff --git a/arch/arm/boot/dts/imx6qdl.dtsi b/arch/arm/boot/dts/imx6qdl.dtsi
index b42822a..6b2ef6c 100644
--- a/arch/arm/boot/dts/imx6qdl.dtsi
+++ b/arch/arm/boot/dts/imx6qdl.dtsi
@@ -1100,6 +1100,7 @@
ocotp: ocotp@021bc000 {
compatible = "fsl,imx6q-ocotp", "syscon";
reg = <0x021bc000 0x4000>;
+ clocks = <&clks IMX6QDL_CLK_IIM>;
};

tzasc@021d0000 { /* TZASC1 */
--
1.8.4.5

2016-04-19 08:17:33

by Peng Fan

Subject: [PATCH 3/3] nvmem: imx-ocotp: handling clock

Before access ocotp nvmem area, the clock should be enabled.
Or, `hexdump nvmem` will hang the system. So, use such flow:
"
1. clock_enable_prepare
2. read nvmem ocotp area
3. clock_disable_unprepare
"

Signed-off-by: Peng Fan <[email protected]>
Cc: Srinivas Kandagatla <[email protected]>
Cc: Maxime Ripard <[email protected]>
Cc: Shawn Guo <[email protected]>
---
drivers/nvmem/imx-ocotp.c | 10 ++++++++++
1 file changed, 10 insertions(+)

diff --git a/drivers/nvmem/imx-ocotp.c b/drivers/nvmem/imx-ocotp.c
index d7796eb..55095c0 100644
--- a/drivers/nvmem/imx-ocotp.c
+++ b/drivers/nvmem/imx-ocotp.c
@@ -15,6 +15,7 @@
* http://www.gnu.org/copyleft/gpl.html
*/

+#include <linux/clk.h>
#include <linux/device.h>
#include <linux/io.h>
#include <linux/module.h>
@@ -27,6 +28,7 @@

struct ocotp_priv {
struct device *dev;
+ struct clk *clk;
void __iomem *base;
unsigned int nregs;
};
@@ -46,11 +48,15 @@ static int imx_ocotp_read(void *context, const void *reg, size_t reg_size,
if (count > (priv->nregs - index))
count = priv->nregs - index;

+ clk_prepare_enable(priv->clk);
+
for (i = index; i < (index + count); i++) {
*(u32 *)val = readl(priv->base + 0x400 + i * 0x10);
val += 4;
}

+ clk_disable_unprepare(priv->clk);
+
return 0;
}

@@ -112,6 +118,10 @@ static int imx_ocotp_probe(struct platform_device *pdev)
if (IS_ERR(priv->base))
return PTR_ERR(priv->base);

+ priv->clk = devm_clk_get(&pdev->dev, NULL);
+ if (IS_ERR(priv->clk))
+ return PTR_ERR(priv->clk);
+
of_id = of_match_device(imx_ocotp_dt_ids, dev);
priv->nregs = (unsigned int)of_id->data;
imx_ocotp_regmap_config.max_register = 4 * priv->nregs - 4;
--
1.8.4.5

2016-04-19 08:17:27

by Peng Fan

Subject: [PATCH 2/3] arm: dts: imx6sl: add clocks property for ocotp node

Add clocks property for ocotp node.

Signed-off-by: Peng Fan <[email protected]>
Cc: Shawn Guo <[email protected]>
Cc: Sascha Hauer <[email protected]>
Cc: Rob Herring <[email protected]>
---
arch/arm/boot/dts/imx6sl.dtsi | 1 +
1 file changed, 1 insertion(+)

diff --git a/arch/arm/boot/dts/imx6sl.dtsi b/arch/arm/boot/dts/imx6sl.dtsi
index d12b250..b37da94 100644
--- a/arch/arm/boot/dts/imx6sl.dtsi
+++ b/arch/arm/boot/dts/imx6sl.dtsi
@@ -853,6 +853,7 @@
ocotp: ocotp@021bc000 {
compatible = "fsl,imx6sl-ocotp", "syscon";
reg = <0x021bc000 0x4000>;
+ clocks = <&clks IMX6SL_CLK_OCOTP>;
};

audmux: audmux@021d8000 {
--
1.8.4.5

2016-04-19 10:42:19

by Fabio Estevam

Subject: Re: [PATCH 3/3] nvmem: imx-ocotp: handling clock

Hi Peng,

On Tue, Apr 19, 2016 at 5:33 AM, Peng Fan <[email protected]> wrote:

> @@ -46,11 +48,15 @@ static int imx_ocotp_read(void *context, const void *reg, size_t reg_size,
> if (count > (priv->nregs - index))
> count = priv->nregs - index;
>
> + clk_prepare_enable(priv->clk);

clk_prepare_enable() may fail, so you should better check its return
value and propagate it in the case of error.

2016-04-19 13:23:13

by Peng Fan

Subject: Re: [PATCH 3/3] nvmem: imx-ocotp: handling clock

Hi Fabio,
On Tue, Apr 19, 2016 at 07:42:17AM -0300, Fabio Estevam wrote:
>Hi Peng,
>
>On Tue, Apr 19, 2016 at 5:33 AM, Peng Fan <[email protected]> wrote:
>
>> @@ -46,11 +48,15 @@ static int imx_ocotp_read(void *context, const void *reg, size_t reg_size,
>> if (count > (priv->nregs - index))
>> count = priv->nregs - index;
>>
>> + clk_prepare_enable(priv->clk);
>
>clk_prepare_enable() may fail, so you should better check its return
>value and propagate it in the case of error.

Thanks for correcting me. Will fix it in V2.

Thanks,
Peng

--