On Thu, 2019-12-12 at 00:44 +0800, Philipp Zabel wrote:

Hi,

On Thu, 2019-12-12 at 00:22 +0800, Jiaxin Yu wrote:

quoted

From: "yong.liang" <redacted>

Add reset controller API in watchdog driver.
Besides watchdog, MTK toprgu module alsa provide sub-system (eg, audio,
camera, codec and connectivity) software reset functionality.

Do any of the listed sub-systems use the reset_control_reset()
functionality? Is there no delay requirement between assert and
deassert? Otherwise it would be safer not to implement the .reset()
operation at all.

quoted

Audio use this reset API to reset audio, and it works well.

    It is no need to add delay between assert and deassert.

quoted

Signed-off-by: yong.liang <redacted>
---
 drivers/watchdog/Kconfig   |   1 +
 drivers/watchdog/mtk_wdt.c | 109 ++++++++++++++++++++++++++++++++++++-
 2 files changed, 109 insertions(+), 1 deletion(-)

diff --git a/drivers/watchdog/Kconfig b/drivers/watchdog/Kconfig
index 2e07caab9db2..629249fe5305 100644
--- a/drivers/watchdog/Kconfig
+++ b/drivers/watchdog/Kconfig

@@ -717,6 +717,7 @@ config MEDIATEK_WATCHDOG
 	tristate "Mediatek SoCs watchdog support"
 	depends on ARCH_MEDIATEK || COMPILE_TEST
 	select WATCHDOG_CORE
+	select RESET_CONTROLLER
 	help
 	  Say Y here to include support for the watchdog timer
 	  in Mediatek SoCs.

diff --git a/drivers/watchdog/mtk_wdt.c b/drivers/watchdog/mtk_wdt.c
index 9c3d0033260d..667380031dfd 100644
--- a/drivers/watchdog/mtk_wdt.c
+++ b/drivers/watchdog/mtk_wdt.c

@@ -9,6 +9,9 @@
  * Based on sunxi_wdt.c
  */
 
+#include <dt-bindings/reset-controller/mt2712-resets.h>
+#include <dt-bindings/reset-controller/mt8183-resets.h>
+#include <linux/delay.h>
 #include <linux/err.h>
 #include <linux/init.h>
 #include <linux/io.h>

@@ -16,10 +19,12 @@
 #include <linux/module.h>
 #include <linux/moduleparam.h>
 #include <linux/of.h>
+#include <linux/of_device.h>
 #include <linux/platform_device.h>
+#include <linux/reset-controller.h>
+#include <linux/slab.h>

What is this required for?

  It is no use. I will remove it.

quoted

 #include <linux/types.h>
 #include <linux/watchdog.h>
-#include <linux/delay.h>

 #define WDT_MAX_TIMEOUT		31
 #define WDT_MIN_TIMEOUT		1

@@ -44,6 +49,9 @@
 #define WDT_SWRST		0x14
 #define WDT_SWRST_KEY		0x1209
 
+#define WDT_SWSYSRST		0x18U
+#define WDT_SWSYS_RST_KEY	0x88000000
+
 #define DRV_NAME		"mtk-wdt"
 #define DRV_VERSION		"1.0"
 

@@ -53,8 +61,97 @@ static unsigned int timeout;
 struct mtk_wdt_dev {
 	struct watchdog_device wdt_dev;
 	void __iomem *wdt_base;
+	spinlock_t lock; /* protects WDT_SWSYSRST reg */
+	struct reset_controller_dev rcdev;
+};
+
+struct mtk_wdt_data {
+	int infracfg_sw_rst_num;

This is not used at all, better remove it.

  I will remove it.

quoted

+	int toprgu_sw_rst_num;
+};
+
+static const struct mtk_wdt_data mt2712_data = {
+	.toprgu_sw_rst_num = MT2712_TOPRGU_SW_RST_NUM,
+};
+
+static const struct mtk_wdt_data mt8183_data = {
+	.infracfg_sw_rst_num = MT8183_INFRACFG_SW_RST_NUM,

Same as above.

  I will remove it.

quoted

+	.toprgu_sw_rst_num = MT8183_TOPRGU_SW_RST_NUM,
+};
+
+static int toprgu_reset_update(struct reset_controller_dev *rcdev,
+			       unsigned long id, bool assert)
+{
+	unsigned int tmp;
+	unsigned long flags;
+

This empty line can be removed.

  I will remove it.

quoted

+	struct mtk_wdt_dev *data =
+		 container_of(rcdev, struct mtk_wdt_dev, rcdev);
+
+	spin_lock_irqsave(&data->lock, flags);
+
+	tmp = readl(data->wdt_base + WDT_SWSYSRST);
+	if (assert)
+		tmp |= BIT(id);
+	else
+		tmp &= ~BIT(id);
+	tmp |= WDT_SWSYS_RST_KEY;
+	writel(tmp, data->wdt_base + WDT_SWSYSRST);
+
+	spin_unlock_irqrestore(&data->lock, flags);
+
+	return 0;
+}
+
+static int toprgu_reset_assert(struct reset_controller_dev *rcdev,
+			       unsigned long id)
+{
+	return toprgu_reset_update(rcdev, id, true);
+}
+
+static int toprgu_reset_deassert(struct reset_controller_dev *rcdev,
+				 unsigned long id)
+{
+	return toprgu_reset_update(rcdev, id, false);
+}
+
+static int toprgu_reset(struct reset_controller_dev *rcdev,
+			unsigned long id)
+{
+	int ret;
+
+	ret = toprgu_reset_assert(rcdev, id);
+	if (ret)
+		return ret;
+
+	return toprgu_reset_deassert(rcdev, id);
+}

As mentioned above, is this needed? Does this work for all modules?
Only implement this if you are sure both are true.

  I think it's safe. Many modules has used it in MTK inside.

regards
Philipp

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel