From: Patrick Rudolph <[email protected]>
Maxim Max7357 has a configuration register to enable additional
features. These features aren't enabled by default & its up to
board designer to enable the same as it may have unexpected side effects.
These should be validated for proper functioning & detection of devices
in secondary bus as sometimes it can cause secondary bus being disabled.
Add booleans for:
- maxim,isolate-stuck-channel
- maxim,send-flush-out-sequence
- maxim,preconnection-wiggle-test-enable
Signed-off-by: Patrick Rudolph <[email protected]>
Signed-off-by: Naresh Solanki <[email protected]>
Reviewed-by: Rob Herring <[email protected]>
---
Changes in V4:
- Drop max7358.
Changes in V3:
- Update commit message
Changes in V2:
- Update properties.
---
.../bindings/i2c/i2c-mux-pca954x.yaml | 30 +++++++++++++++++++
1 file changed, 30 insertions(+)
diff --git a/Documentation/devicetree/bindings/i2c/i2c-mux-pca954x.yaml b/Documentation/devicetree/bindings/i2c/i2c-mux-pca954x.yaml
index 2d7bb998b0e9..9aa0585200c9 100644
--- a/Documentation/devicetree/bindings/i2c/i2c-mux-pca954x.yaml
+++ b/Documentation/devicetree/bindings/i2c/i2c-mux-pca954x.yaml
@@ -71,6 +71,23 @@ properties:
description: A voltage regulator supplying power to the chip. On PCA9846
the regulator supplies power to VDD2 (core logic) and optionally to VDD1.
+ maxim,isolate-stuck-channel:
+ type: boolean
+ description: Allows to use non faulty channels while a stuck channel is
+ isolated from the upstream bus. If not set all channels are isolated from
+ the upstream bus until the fault is cleared.
+
+ maxim,send-flush-out-sequence:
+ type: boolean
+ description: Send a flush-out sequence to stuck auxiliary buses
+ automatically after a stuck channel is being detected.
+
+ maxim,preconnection-wiggle-test-enable:
+ type: boolean
+ description: Send a STOP condition to the auxiliary buses when the switch
+ register activates a channel to detect a stuck high fault. On fault the
+ channel is isolated from the upstream bus.
+
required:
- compatible
- reg
@@ -95,6 +112,19 @@ allOf:
"#interrupt-cells": false
interrupt-controller: false
+ - if:
+ not:
+ properties:
+ compatible:
+ contains:
+ enum:
+ - maxim,max7357
+ then:
+ properties:
+ maxim,isolate-stuck-channel: false
+ maxim,send-flush-out-sequence: false
+ maxim,preconnection-wiggle-test-enable: false
+
unevaluatedProperties: false
examples:
base-commit: 76998e5bcdf155b36c7066808a0a65b2ee13cb2a
--
2.41.0
From: Patrick Rudolph <[email protected]>
Enable additional features based on DT settings and unconditionally
release the shared interrupt pin after 1.6 seconds and allow to use
it as reset.
These features aren't enabled by default and it's up to board designer
to validate for proper functioning and detection of devices in secondary
bus as sometimes it can cause secondary bus being disabled.
Signed-off-by: Patrick Rudolph <[email protected]>
Signed-off-by: Naresh Solanki <[email protected]>
Reviewed-by: Andi Shyti <[email protected]>
---
Changes in V5:
- Fix typos
- Update comment
- Add newline in dev_warn
Changes in V4:
- Drop max7358
- Update #define
- Move conf variable
- Print warning when I2C_FUNC_SMBUS_WRITE_BYTE_DATA isn't supported
Changes in V3:
- Delete unused #define
- Update pca954x_init
- Update commit message
Changes in V2:
- Update comments
- Update check for DT properties
---
drivers/i2c/muxes/i2c-mux-pca954x.c | 43 ++++++++++++++++++++++++++++-
1 file changed, 42 insertions(+), 1 deletion(-)
diff --git a/drivers/i2c/muxes/i2c-mux-pca954x.c b/drivers/i2c/muxes/i2c-mux-pca954x.c
index 2219062104fb..f5dfc33b97c0 100644
--- a/drivers/i2c/muxes/i2c-mux-pca954x.c
+++ b/drivers/i2c/muxes/i2c-mux-pca954x.c
@@ -57,6 +57,20 @@
#define PCA954X_IRQ_OFFSET 4
+/*
+ * MAX7357's configuration register is writeable after POR, but
+ * can be locked by setting the basic mode bit. MAX7358 configuration
+ * register is locked by default and needs to be unlocked first.
+ * The configuration register holds the following settings:
+ */
+#define MAX7357_CONF_INT_ENABLE BIT(0)
+#define MAX7357_CONF_FLUSH_OUT BIT(1)
+#define MAX7357_CONF_RELEASE_INT BIT(2)
+#define MAX7357_CONF_DISCON_SINGLE_CHAN BIT(4)
+#define MAX7357_CONF_PRECONNECT_TEST BIT(7)
+
+#define MAX7357_POR_DEFAULT_CONF MAX7357_CONF_INT_ENABLE
+
enum pca_type {
max_7356,
max_7357,
@@ -470,7 +484,34 @@ static int pca954x_init(struct i2c_client *client, struct pca954x *data)
else
data->last_chan = 0; /* Disconnect multiplexer */
- ret = i2c_smbus_write_byte(client, data->last_chan);
+ if (device_is_compatible(&client->dev, "maxim,max7357")) {
+ if (i2c_check_functionality(client->adapter, I2C_FUNC_SMBUS_WRITE_BYTE_DATA)) {
+ u8 conf = MAX7357_POR_DEFAULT_CONF;
+ /*
+ * The interrupt signal is shared with the reset pin. Release the
+ * interrupt after 1.6 seconds to allow using the pin as reset.
+ */
+ conf |= MAX7357_CONF_RELEASE_INT;
+
+ if (device_property_read_bool(&client->dev, "maxim,isolate-stuck-channel"))
+ conf |= MAX7357_CONF_DISCON_SINGLE_CHAN;
+ if (device_property_read_bool(&client->dev,
+ "maxim,send-flush-out-sequence"))
+ conf |= MAX7357_CONF_FLUSH_OUT;
+ if (device_property_read_bool(&client->dev,
+ "maxim,preconnection-wiggle-test-enable"))
+ conf |= MAX7357_CONF_PRECONNECT_TEST;
+
+ ret = i2c_smbus_write_byte_data(client, data->last_chan, conf);
+ } else {
+ dev_warn(&client->dev, "Write byte data not supported."
+ "Cannot enable enhanced mode features\n");
+ ret = i2c_smbus_write_byte(client, data->last_chan);
+ }
+ } else {
+ ret = i2c_smbus_write_byte(client, data->last_chan);
+ }
+
if (ret < 0)
data->last_chan = 0;
--
2.41.0
On Wed, Dec 20, 2023 at 01:58:01PM +0530, Naresh Solanki wrote:
> From: Patrick Rudolph <[email protected]>
>
> Maxim Max7357 has a configuration register to enable additional
> features. These features aren't enabled by default & its up to
> board designer to enable the same as it may have unexpected side effects.
>
> These should be validated for proper functioning & detection of devices
> in secondary bus as sometimes it can cause secondary bus being disabled.
>
> Add booleans for:
> - maxim,isolate-stuck-channel
> - maxim,send-flush-out-sequence
> - maxim,preconnection-wiggle-test-enable
>
> Signed-off-by: Patrick Rudolph <[email protected]>
> Signed-off-by: Naresh Solanki <[email protected]>
> Reviewed-by: Rob Herring <[email protected]>
Rob, are you really OK with these bindings? They look more like
configuration instead of HW description to me.
> ---
> Changes in V4:
> - Drop max7358.
> Changes in V3:
> - Update commit message
> Changes in V2:
> - Update properties.
> ---
> .../bindings/i2c/i2c-mux-pca954x.yaml | 30 +++++++++++++++++++
> 1 file changed, 30 insertions(+)
>
> diff --git a/Documentation/devicetree/bindings/i2c/i2c-mux-pca954x.yaml b/Documentation/devicetree/bindings/i2c/i2c-mux-pca954x.yaml
> index 2d7bb998b0e9..9aa0585200c9 100644
> --- a/Documentation/devicetree/bindings/i2c/i2c-mux-pca954x.yaml
> +++ b/Documentation/devicetree/bindings/i2c/i2c-mux-pca954x.yaml
> @@ -71,6 +71,23 @@ properties:
> description: A voltage regulator supplying power to the chip. On PCA9846
> the regulator supplies power to VDD2 (core logic) and optionally to VDD1.
>
> + maxim,isolate-stuck-channel:
> + type: boolean
> + description: Allows to use non faulty channels while a stuck channel is
> + isolated from the upstream bus. If not set all channels are isolated from
> + the upstream bus until the fault is cleared.
> +
> + maxim,send-flush-out-sequence:
> + type: boolean
> + description: Send a flush-out sequence to stuck auxiliary buses
> + automatically after a stuck channel is being detected.
> +
> + maxim,preconnection-wiggle-test-enable:
> + type: boolean
> + description: Send a STOP condition to the auxiliary buses when the switch
> + register activates a channel to detect a stuck high fault. On fault the
> + channel is isolated from the upstream bus.
> +
> required:
> - compatible
> - reg
> @@ -95,6 +112,19 @@ allOf:
> "#interrupt-cells": false
> interrupt-controller: false
>
> + - if:
> + not:
> + properties:
> + compatible:
> + contains:
> + enum:
> + - maxim,max7357
> + then:
> + properties:
> + maxim,isolate-stuck-channel: false
> + maxim,send-flush-out-sequence: false
> + maxim,preconnection-wiggle-test-enable: false
> +
> unevaluatedProperties: false
>
> examples:
>
> base-commit: 76998e5bcdf155b36c7066808a0a65b2ee13cb2a
> --
> 2.41.0
>
>
On 20/12/2023 21:50, Wolfram Sang wrote:
> On Wed, Dec 20, 2023 at 01:58:01PM +0530, Naresh Solanki wrote:
>> From: Patrick Rudolph <[email protected]>
>>
>> Maxim Max7357 has a configuration register to enable additional
>> features. These features aren't enabled by default & its up to
>> board designer to enable the same as it may have unexpected side effects.
>>
>> These should be validated for proper functioning & detection of devices
>> in secondary bus as sometimes it can cause secondary bus being disabled.
>>
>> Add booleans for:
>> - maxim,isolate-stuck-channel
>> - maxim,send-flush-out-sequence
>> - maxim,preconnection-wiggle-test-enable
>>
>> Signed-off-by: Patrick Rudolph <[email protected]>
>> Signed-off-by: Naresh Solanki <[email protected]>
>> Reviewed-by: Rob Herring <[email protected]>
>
> Rob, are you really OK with these bindings? They look more like
> configuration instead of HW description to me.
Some explanation was provided here:
https://lore.kernel.org/all/CABqG17g8QOgU7cObe=4EMLbEC1PeZWxdPXt7zzFs35JGqpRbfg@mail.gmail.com/
AFAIU, these properties are board-design choice.
Best regards,
Krzysztof
> Some explanation was provided here:
> https://lore.kernel.org/all/CABqG17g8QOgU7cObe=4EMLbEC1PeZWxdPXt7zzFs35JGqpRbfg@mail.gmail.com/
>
> AFAIU, these properties are board-design choice.
Hey, thanks for the heads up. I agree that these options should not be
"on" by default. I am still not fully convinced they serve as hardware
description, though. Need to think about it some more...
Hi Wolfram,
On Fri, 22 Dec 2023 at 16:05, Wolfram Sang <[email protected]> wrote:
>
>
> > Some explanation was provided here:
> > https://lore.kernel.org/all/CABqG17g8QOgU7cObe=4EMLbEC1PeZWxdPXt7zzFs35JGqpRbfg@mail.gmail.com/
> >
> > AFAIU, these properties are board-design choice.
>
> Hey, thanks for the heads up. I agree that these options should not be
> "on" by default. I am still not fully convinced they serve as hardware
> description, though. Need to think about it some more...
Any update on this ? Let me know if I can help with additional information.
Regards,
Naresh
>