The 'reset-gpios' will not work when the switch reset is controlled by a
reset controller.
Although the reset is optional and the driver performs a soft reset
during setup, if the initial reset state was asserted, the driver will
not detect it.
The reset controller will take precedence over the reset GPIO.
Signed-off-by: Luiz Angelo Daros de Luca <[email protected]>
---
drivers/net/dsa/realtek/realtek.h | 2 ++
drivers/net/dsa/realtek/rtl83xx.c | 52 +++++++++++++++++++++++++++++++++++----
drivers/net/dsa/realtek/rtl83xx.h | 2 ++
3 files changed, 51 insertions(+), 5 deletions(-)
diff --git a/drivers/net/dsa/realtek/realtek.h b/drivers/net/dsa/realtek/realtek.h
index b80bfde1ad04..e0b1aa01337b 100644
--- a/drivers/net/dsa/realtek/realtek.h
+++ b/drivers/net/dsa/realtek/realtek.h
@@ -12,6 +12,7 @@
#include <linux/platform_device.h>
#include <linux/gpio/consumer.h>
#include <net/dsa.h>
+#include <linux/reset.h>
#define REALTEK_HW_STOP_DELAY 25 /* msecs */
#define REALTEK_HW_START_DELAY 100 /* msecs */
@@ -48,6 +49,7 @@ struct rtl8366_vlan_4k {
struct realtek_priv {
struct device *dev;
+ struct reset_control *reset_ctl;
struct gpio_desc *reset;
struct gpio_desc *mdc;
struct gpio_desc *mdio;
diff --git a/drivers/net/dsa/realtek/rtl83xx.c b/drivers/net/dsa/realtek/rtl83xx.c
index 801873754df2..3d1f541936b8 100644
--- a/drivers/net/dsa/realtek/rtl83xx.c
+++ b/drivers/net/dsa/realtek/rtl83xx.c
@@ -184,6 +184,13 @@ rtl83xx_probe(struct device *dev,
"realtek,disable-leds");
/* TODO: if power is software controlled, set up any regulators here */
+ priv->reset_ctl = devm_reset_control_get_optional(dev, NULL);
+ if (IS_ERR(priv->reset_ctl)) {
+ ret = PTR_ERR(priv->reset_ctl);
+ dev_err_probe(dev, ret, "failed to get reset control\n");
+ return ERR_CAST(priv->reset_ctl);
+ }
+
priv->reset = devm_gpiod_get_optional(dev, "reset", GPIOD_OUT_LOW);
if (IS_ERR(priv->reset)) {
dev_err(dev, "failed to get RESET GPIO\n");
@@ -192,11 +199,11 @@ rtl83xx_probe(struct device *dev,
dev_set_drvdata(dev, priv);
- if (priv->reset) {
- gpiod_set_value(priv->reset, 1);
+ if (priv->reset_ctl || priv->reset) {
+ rtl83xx_reset_assert(priv);
dev_dbg(dev, "asserted RESET\n");
msleep(REALTEK_HW_STOP_DELAY);
- gpiod_set_value(priv->reset, 0);
+ rtl83xx_reset_deassert(priv);
msleep(REALTEK_HW_START_DELAY);
dev_dbg(dev, "deasserted RESET\n");
}
@@ -292,11 +299,46 @@ EXPORT_SYMBOL_NS_GPL(rtl83xx_shutdown, REALTEK_DSA);
void rtl83xx_remove(struct realtek_priv *priv)
{
/* leave the device reset asserted */
- if (priv->reset)
- gpiod_set_value(priv->reset, 1);
+ rtl83xx_reset_assert(priv);
}
EXPORT_SYMBOL_NS_GPL(rtl83xx_remove, REALTEK_DSA);
+void rtl83xx_reset_assert(struct realtek_priv *priv)
+{
+ int ret;
+
+ if (priv->reset_ctl) {
+ ret = reset_control_assert(priv->reset_ctl);
+ if (!ret)
+ return;
+
+ dev_warn(priv->dev,
+ "Failed to assert the switch reset control: %pe\n",
+ ERR_PTR(ret));
+ }
+
+ if (priv->reset)
+ gpiod_set_value(priv->reset, true);
+}
+
+void rtl83xx_reset_deassert(struct realtek_priv *priv)
+{
+ int ret;
+
+ if (priv->reset_ctl) {
+ ret = reset_control_deassert(priv->reset_ctl);
+ if (!ret)
+ return;
+
+ dev_warn(priv->dev,
+ "Failed to deassert the switch reset control: %pe\n",
+ ERR_PTR(ret));
+ }
+
+ if (priv->reset)
+ gpiod_set_value(priv->reset, false);
+}
+
MODULE_AUTHOR("Luiz Angelo Daros de Luca <[email protected]>");
MODULE_AUTHOR("Linus Walleij <[email protected]>");
MODULE_DESCRIPTION("Realtek DSA switches common module");
diff --git a/drivers/net/dsa/realtek/rtl83xx.h b/drivers/net/dsa/realtek/rtl83xx.h
index 0ddff33df6b0..c8a0ff8fd75e 100644
--- a/drivers/net/dsa/realtek/rtl83xx.h
+++ b/drivers/net/dsa/realtek/rtl83xx.h
@@ -18,5 +18,7 @@ int rtl83xx_register_switch(struct realtek_priv *priv);
void rtl83xx_unregister_switch(struct realtek_priv *priv);
void rtl83xx_shutdown(struct realtek_priv *priv);
void rtl83xx_remove(struct realtek_priv *priv);
+void rtl83xx_reset_assert(struct realtek_priv *priv);
+void rtl83xx_reset_deassert(struct realtek_priv *priv);
#endif /* _RTL83XX_H */
--
2.43.0
Hi Luiz,
thanks for your patch!
On Wed, Feb 14, 2024 at 1:54 AM Luiz Angelo Daros de Luca
<[email protected]> wrote:
> The 'reset-gpios' will not work when the switch reset is controlled by a
> reset controller.
>
> Although the reset is optional and the driver performs a soft reset
> during setup, if the initial reset state was asserted, the driver will
> not detect it.
>
> The reset controller will take precedence over the reset GPIO.
>
> Signed-off-by: Luiz Angelo Daros de Luca <[email protected]>
(...)
> +void rtl83xx_reset_assert(struct realtek_priv *priv)
> +{
> + int ret;
> +
> + if (priv->reset_ctl) {
> + ret = reset_control_assert(priv->reset_ctl);
Actually, reset_control_assert() is NULL-tolerand (as well as the
stubs) so you can just issue this unconditionally. If priv->reset_ctl
is NULL it will just return 0.
> + if (!ret)
> + return;
> +
> + dev_warn(priv->dev,
> + "Failed to assert the switch reset control: %pe\n",
> + ERR_PTR(ret));
> + }
> +
> + if (priv->reset)
> + gpiod_set_value(priv->reset, true);
Same here! Also NULL-tolerant. You do not need to check priv->reset.
Just issue it.
> +void rtl83xx_reset_deassert(struct realtek_priv *priv)
Same comments for deassert.
With this fixed (the rest looks just fine):
Reviewed-by: Linus Walleij <[email protected]>
Yours,
Linus Walleij
> Hi Luiz,
Hi Linus,
> thanks for your patch!
I'm glad to help ;-)
> On Wed, Feb 14, 2024 at 1:54 AM Luiz Angelo Daros de Luca
> <[email protected]> wrote:
>
> > The 'reset-gpios' will not work when the switch reset is controlled by a
> > reset controller.
> >
> > Although the reset is optional and the driver performs a soft reset
> > during setup, if the initial reset state was asserted, the driver will
> > not detect it.
> >
> > The reset controller will take precedence over the reset GPIO.
> >
> > Signed-off-by: Luiz Angelo Daros de Luca <[email protected]>
> (...)
> > +void rtl83xx_reset_assert(struct realtek_priv *priv)
> > +{
> > + int ret;
> > +
> > + if (priv->reset_ctl) {
> > + ret = reset_control_assert(priv->reset_ctl);
>
> Actually, reset_control_assert() is NULL-tolerand (as well as the
> stubs) so you can just issue this unconditionally. If priv->reset_ctl
> is NULL it will just return 0.
The idea was to avoid gpiod_set_value if the reset_control_assert was
configured and worked. However, I don't see a big issue in calling
them both as you don't expect both to be configured.
I'll remove the "ifs not null" and let both be called.
>
> > + if (!ret)
> > + return;
> > +
> > + dev_warn(priv->dev,
> > + "Failed to assert the switch reset control: %pe\n",
> > + ERR_PTR(ret));
> > + }
> > +
> > + if (priv->reset)
> > + gpiod_set_value(priv->reset, true);
>
> Same here! Also NULL-tolerant. You do not need to check priv->reset.
> Just issue it.
>
> > +void rtl83xx_reset_deassert(struct realtek_priv *priv)
>
> Same comments for deassert.
>
> With this fixed (the rest looks just fine):
> Reviewed-by: Linus Walleij <[email protected]>
>
> Yours,
> Linus Walleij
On Wed, Feb 14, 2024 at 07:17:55PM -0300, Luiz Angelo Daros de Luca wrote:
> > On Wed, Feb 14, 2024 at 1:54 AM Luiz Angelo Daros de Luca
> > <[email protected]> wrote:
> >
> > > The 'reset-gpios' will not work when the switch reset is controlled by a
> > > reset controller.
> > >
> > > Although the reset is optional and the driver performs a soft reset
> > > during setup, if the initial reset state was asserted, the driver will
> > > not detect it.
> > >
> > > The reset controller will take precedence over the reset GPIO.
> > >
> > > Signed-off-by: Luiz Angelo Daros de Luca <[email protected]>
> > (...)
> > > +void rtl83xx_reset_assert(struct realtek_priv *priv)
> > > +{
> > > + int ret;
> > > +
> > > + if (priv->reset_ctl) {
> > > + ret = reset_control_assert(priv->reset_ctl);
> >
> > Actually, reset_control_assert() is NULL-tolerand (as well as the
> > stubs) so you can just issue this unconditionally. If priv->reset_ctl
> > is NULL it will just return 0.
>
> The idea was to avoid gpiod_set_value if the reset_control_assert was
> configured and worked. However, I don't see a big issue in calling
> them both as you don't expect both to be configured.
> I'll remove the "ifs not null" and let both be called.
In the defense of Linus' comment, your proposed code did not do what you
seem to have intended anyway. If priv->reset_ctl was non-NULL, it fell
through and potentially ran gpiod_set_value() anyway - there was no
early "return". So, the simplification comment was predictable.
On Wed, Feb 14, 2024 at 07:17:55PM -0300, Luiz Angelo Daros de Luca wrote:
> > > The reset controller will take precedence over the reset GPIO.
>
> I'll remove the "ifs not null" and let both be called.
If you do, be sure to update the commit message about it (see first line).