[PATCH 1/2] clk: sunxi-ng: Provide a default reset hook

The reset hook was left implemented. Provide a dumb implementation so that
client drivers can depend on it.

Signed-off-by: Maxime Ripard <redacted>
---
 drivers/clk/sunxi-ng/ccu_reset.c | 12 ++++++++++++
 1 file changed, 12 insertions(+)

diff --git a/drivers/clk/sunxi-ng/ccu_reset.c b/drivers/clk/sunxi-ng/ccu_reset.c
index 6c31d48783a7..1dc4e98ea802 100644
--- a/drivers/clk/sunxi-ng/ccu_reset.c
+++ b/drivers/clk/sunxi-ng/ccu_reset.c

@@ -8,6 +8,7 @@
  * the License, or (at your option) any later version.
  */
 
+#include <linux/delay.h>
 #include <linux/io.h>
 #include <linux/reset-controller.h>
 

@@ -49,7 +50,18 @@ static int ccu_reset_deassert(struct reset_controller_dev *rcdev,
 	return 0;
 }
 
+static int ccu_reset_reset(struct reset_controller_dev *rcdev,
+			   unsigned long id)
+{
+	ccu_reset_assert(rcdev, id);
+	udelay(10);
+	ccu_reset_deassert(rcdev, id);
+
+	return 0;
+}
+
 const struct reset_control_ops ccu_reset_ops = {
 	.assert		= ccu_reset_assert,
 	.deassert	= ccu_reset_deassert,
+	.reset		= ccu_reset_reset,
 };

-- 
2.13.5

We might be into some troubles if the bootloader misconfigured the MMC
controller.

We currently only de-assert the reset line at probe time, which means that
if the device was already out of reset, we're going to keep whatever state
was set already.

Switch to a reset instead of the deassert to have a device in a pristine
state when we start operating.

Signed-off-by: Maxime Ripard <redacted>
---
 drivers/mmc/host/sunxi-mmc.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/mmc/host/sunxi-mmc.c b/drivers/mmc/host/sunxi-mmc.c
index da5f46a14497..53c970fe0873 100644
--- a/drivers/mmc/host/sunxi-mmc.c
+++ b/drivers/mmc/host/sunxi-mmc.c

@@ -1242,7 +1242,7 @@ static int sunxi_mmc_resource_request(struct sunxi_mmc_host *host,
 	}
 
 	if (!IS_ERR(host->reset)) {
-		ret = reset_control_deassert(host->reset);
+		ret = reset_control_reset(host->reset);
 		if (ret) {
 			dev_err(&pdev->dev, "reset err %d\n", ret);
 			goto error_disable_clk_sample;

-- 
2.13.5

On 25 August 2017 at 11:49, Maxime Ripard
[off-list ref] wrote:

We might be into some troubles if the bootloader misconfigured the MMC
controller.

We currently only de-assert the reset line at probe time, which means that
if the device was already out of reset, we're going to keep whatever state
was set already.

Switch to a reset instead of the deassert to have a device in a pristine
state when we start operating.

Signed-off-by: Maxime Ripard <redacted>

Thanks, applied for next!

Kind regards
Uffe

在 2017-08-25 17:49，Maxime Ripard 写道：

The reset hook was left implemented. Provide a dumb implementation so 
that
client drivers can depend on it.

I think some clients should also convert to use this.

For example:

- The TCON driver in sun4i-drm. (It now uses reset_control_status, but
   this is also not implemented in CCU.)
- The EPHY in dwmac-sun8i driver. (It uses explicit assert then 
deassert)

P.S. maybe the status hook should also be implemented?

On Sun, Aug 27, 2017 at 9:01 PM,  [off-list ref] wrote:

在 2017-08-25 17:49，Maxime Ripard 写道：

quoted

The reset hook was left implemented. Provide a dumb implementation so that
client drivers can depend on it.

I think some clients should also convert to use this.

For example:

- The TCON driver in sun4i-drm. (It now uses reset_control_status, but
  this is also not implemented in CCU.)

I have patches for both sides of this.

- The EPHY in dwmac-sun8i driver. (It uses explicit assert then deassert)

P.S. maybe the status hook should also be implemented?

Yes.

ChenYu