Skip to contentSkip to conversation

iio: Silan SC7A20 accelerometer support

PATCH In review Samuel Holland 3 participants 7 messages Jan 30, 2022 – Jan 31, 2022

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

cover iio: Silan SC7A20 accelerometer support

This chip is a clone of an ST part, so it works without any issues when added to the existing driver.

Samuel Holland (3):
dt-bindings: vendor-prefixes: Add silan vendor prefix
dt-bindings: iio: st: Add Silan SC7A20 accelerometer
iio: accel: st_accel: Add support for Silan SC7A20

 .../bindings/iio/st,st-sensors.yaml           |  1 +
 .../devicetree/bindings/vendor-prefixes.yaml  |  2 +
 drivers/iio/accel/st_accel.h                  |  2 +
 drivers/iio/accel/st_accel_core.c             | 79 +++++++++++++++++++
 drivers/iio/accel/st_accel_i2c.c              |  5 ++
 5 files changed, 89 insertions(+)
Signature · 1 line
2.33.1
#1 lore raw

PATCH 1/3 dt-bindings: vendor-prefixes: Add silan vendor prefix

Hangzhou Silan Microelectronics Co., Ltd. (http://www.silan.com.cn/) is a manufacturer of ICs, including MEMS sensors.

---

 Documentation/devicetree/bindings/vendor-prefixes.yaml | 2 ++
 1 file changed, 2 insertions(+)
Documentation/devicetree/bindings/vendor-prefixes.yaml +2 -0
index 294093d45a23..b923be2c8172 100644
@@ -1082,6 +1082,8 @@ patternProperties:
     description: Silicon Image
   "^silabs,.*":
     description: Silicon Laboratories
+  "^silan,.*":
+    description: Hangzhou Silan Microelectronics Co., Ltd.
   "^silead,.*":
     description: Silead Inc.
   "^silergy,.*":
Signature · 1 line
2.33.1
#2 lore raw

PATCH 2/3 dt-bindings: iio: st: Add Silan SC7A20 accelerometer

In reply to Samuel Holland

This chip appears to be a clone of the LIS2DH, as the register bit definitions match exactly.

---

 Documentation/devicetree/bindings/iio/st,st-sensors.yaml | 1 +
 1 file changed, 1 insertion(+)
Documentation/devicetree/bindings/iio/st,st-sensors.yaml +1 -0
index 71de5631ebae..70f755041d8a 100644
@@ -24,6 +24,7 @@ properties:
     oneOf:
       - description: STMicroelectronics Accelerometers
         enum:
+          - silan,sc7a20
           - st,h3lis331dl-accel
           - st,lis2de12
           - st,lis2dw12
Signature · 1 line
2.33.1
maintainer
#3 lore raw

Re: [PATCH 2/3] dt-bindings: iio: st: Add Silan SC7A20 accelerometer

Quoted 14 lines · On Sat, 29 Jan 2022 21:44:40 -0600 Samuel Holland <[email protected]> wrote:

This chip appears to be a clone of the LIS2DH, as the register bit definitions match exactly.

---

 Documentation/devicetree/bindings/iio/st,st-sensors.yaml | 1 +
 1 file changed, 1 insertion(+)
Documentation/devicetree/bindings/iio/st,st-sensors.yaml +0 -0
index 71de5631ebae..70f755041d8a 100644

Documentation/devicetree/bindings/iio/st,st-sensors.yaml properties: · line 25

Quoted 2 lines
Documentation/devicetree/bindings/iio/st,st-sensors.yaml +0 -0
@@ -24,6 +24,7 @@ properties:
     oneOf:
       - description: STMicroelectronics Accelerometers

It's not an STMicroelectronics part, so add a new block for silan ones.

Thanks,

Jonathan

Quoted 5 lines
Documentation/devicetree/bindings/iio/st,st-sensors.yaml +1 -0
         enum:
+          - silan,sc7a20
           - st,h3lis331dl-accel
           - st,lis2de12
           - st,lis2dw12
#4 lore raw

PATCH 3/3 iio: accel: st_accel: Add support for Silan SC7A20

In reply to Samuel Holland

This chip appears to be a clone of the LIS2DH. The new description is a copy of the LIS2DH's description with a different WAI value.

---

 drivers/iio/accel/st_accel.h      |  2 +
 drivers/iio/accel/st_accel_core.c | 79 +++++++++++++++++++++++++++++++
 drivers/iio/accel/st_accel_i2c.c  |  5 ++
 3 files changed, 86 insertions(+)
drivers/iio/accel/st_accel.h +2 -0
index 8750dea56fcb..00e056c21bfc 100644
@@ -36,6 +36,7 @@ enum st_accel_type {
 	LIS3DHH,
 	LIS2DE12,
 	LIS2HH12,
+	SC7A20,
 	ST_ACCEL_MAX,
 };
 
@@ -61,6 +62,7 @@ enum st_accel_type {
 #define LIS3DE_ACCEL_DEV_NAME		"lis3de"
 #define LIS2DE12_ACCEL_DEV_NAME		"lis2de12"
 #define LIS2HH12_ACCEL_DEV_NAME		"lis2hh12"
+#define SC7A20_ACCEL_DEV_NAME		"sc7a20"
 
 #ifdef CONFIG_IIO_BUFFER
 int st_accel_allocate_ring(struct iio_dev *indio_dev);
drivers/iio/accel/st_accel_core.c +79 -0

85 lines not on this page · raw message

drivers/iio/accel/st_accel_i2c.c +5 -0
index c0ce78eebad9..7f5888570e87 100644
@@ -107,6 +107,10 @@ static const struct of_device_id st_accel_of_match[] = {
 		.compatible = "st,lis2hh12",
 		.data = LIS2HH12_ACCEL_DEV_NAME,
 	},
+	{
+		.compatible = "silan,sc7a20",
+		.data = SC7A20_ACCEL_DEV_NAME,
+	},
 	{},
 };
 MODULE_DEVICE_TABLE(of, st_accel_of_match);
@@ -142,6 +146,7 @@ static const struct i2c_device_id st_accel_id_table[] = {
 	{ LIS3DE_ACCEL_DEV_NAME },
 	{ LIS2DE12_ACCEL_DEV_NAME },
 	{ LIS2HH12_ACCEL_DEV_NAME },
+	{ SC7A20_ACCEL_DEV_NAME },
 	{},
 };
 MODULE_DEVICE_TABLE(i2c, st_accel_id_table);
Signature · 1 line
2.33.1
maintainer
#5 lore raw

Re: [PATCH 3/3] iio: accel: st_accel: Add support for Silan SC7A20

Quoted 4 lines · On Sat, 29 Jan 2022 21:44:41 -0600 Samuel Holland <[email protected]> wrote:

This chip appears to be a clone of the LIS2DH. The new description is a copy of the LIS2DH's description with a different WAI value.

Hi Samuel,

This is nasty as 0x11 is a valid who am I for an LIS3DHH which is a very different device.

One request inline for a bit more info in the code about what this device is etc. I've found a Chinese data sheet on one of the datasheet aggregator sites.

If you know of a stable location to add a
Datasheet:
tag to this patch that would also be helpful.

Thanks,

Jonathan

Quoted 32 lines

---

 drivers/iio/accel/st_accel.h      |  2 +
 drivers/iio/accel/st_accel_core.c | 79 +++++++++++++++++++++++++++++++
 drivers/iio/accel/st_accel_i2c.c  |  5 ++
 3 files changed, 86 insertions(+)
drivers/iio/accel/st_accel.h +2 -0
index 8750dea56fcb..00e056c21bfc 100644
@@ -36,6 +36,7 @@ enum st_accel_type {
 	LIS3DHH,
 	LIS2DE12,
 	LIS2HH12,
+	SC7A20,
 	ST_ACCEL_MAX,
 };
 
@@ -61,6 +62,7 @@ enum st_accel_type {
 #define LIS3DE_ACCEL_DEV_NAME		"lis3de"
 #define LIS2DE12_ACCEL_DEV_NAME		"lis2de12"
 #define LIS2HH12_ACCEL_DEV_NAME		"lis2hh12"
+#define SC7A20_ACCEL_DEV_NAME		"sc7a20"
 
 #ifdef CONFIG_IIO_BUFFER
 int st_accel_allocate_ring(struct iio_dev *indio_dev);
drivers/iio/accel/st_accel_core.c +0 -0
index 31ea19d0ba71..d9aa0ff1922b 100644

drivers/iio/accel/st_accel_core.c static const struct st_sensor_settings st_accel_sensors_settings[] = { · line 1090

Quoted 4 lines
drivers/iio/accel/st_accel_core.c +1 -0
@@ -1087,6 +1087,85 @@ static const struct st_sensor_settings st_accel_sensors_settings[] = {
 		.multi_read_bit = true,
 		.bootime = 2,
 	},
+	{

Please add a comment here about the fact it seems to be a clone of the LIS2DH despite the wrong WAI.

It's worth noting that a good part of the maintenance of this driver is done by ST employees who are probably less than happy with seeing
a clone (with bugs) supported. So support is likely to be somewhat best effort / not going to deliberately break support for this part but no means to test it. Hopefully we'll remember to cc you to test anything new added to the driver. Clones aren't always perfect!

Jonathan

Quoted 104 lines
drivers/iio/accel/st_accel_core.c +78 -0
+		.wai = 0x11,
+		.wai_addr = ST_SENSORS_DEFAULT_WAI_ADDRESS,
+		.sensors_supported = {
+			[0] = SC7A20_ACCEL_DEV_NAME,
+		},
+		.ch = (struct iio_chan_spec *)st_accel_12bit_channels,
+		.odr = {
+			.addr = 0x20,
+			.mask = 0xf0,
+			.odr_avl = {
+				{ .hz = 1, .value = 0x01, },
+				{ .hz = 10, .value = 0x02, },
+				{ .hz = 25, .value = 0x03, },
+				{ .hz = 50, .value = 0x04, },
+				{ .hz = 100, .value = 0x05, },
+				{ .hz = 200, .value = 0x06, },
+				{ .hz = 400, .value = 0x07, },
+				{ .hz = 1600, .value = 0x08, },
+			},
+		},
+		.pw = {
+			.addr = 0x20,
+			.mask = 0xf0,
+			.value_off = ST_SENSORS_DEFAULT_POWER_OFF_VALUE,
+		},
+		.enable_axis = {
+			.addr = ST_SENSORS_DEFAULT_AXIS_ADDR,
+			.mask = ST_SENSORS_DEFAULT_AXIS_MASK,
+		},
+		.fs = {
+			.addr = 0x23,
+			.mask = 0x30,
+			.fs_avl = {
+				[0] = {
+					.num = ST_ACCEL_FS_AVL_2G,
+					.value = 0x00,
+					.gain = IIO_G_TO_M_S_2(1000),
+				},
+				[1] = {
+					.num = ST_ACCEL_FS_AVL_4G,
+					.value = 0x01,
+					.gain = IIO_G_TO_M_S_2(2000),
+				},
+				[2] = {
+					.num = ST_ACCEL_FS_AVL_8G,
+					.value = 0x02,
+					.gain = IIO_G_TO_M_S_2(4000),
+				},
+				[3] = {
+					.num = ST_ACCEL_FS_AVL_16G,
+					.value = 0x03,
+					.gain = IIO_G_TO_M_S_2(12000),
+				},
+			},
+		},
+		.bdu = {
+			.addr = 0x23,
+			.mask = 0x80,
+		},
+		.drdy_irq = {
+			.int1 = {
+				.addr = 0x22,
+				.mask = 0x10,
+			},
+			.addr_ihl = 0x25,
+			.mask_ihl = 0x02,
+			.stat_drdy = {
+				.addr = ST_SENSORS_DEFAULT_STAT_ADDR,
+				.mask = 0x07,
+			},
+		},
+		.sim = {
+			.addr = 0x23,
+			.value = BIT(0),
+		},
+		.multi_read_bit = true,
+		.bootime = 2,
+	},
 };
 
 /* Default accel DRDY is available on INT1 pin */
drivers/iio/accel/st_accel_i2c.c +5 -0
index c0ce78eebad9..7f5888570e87 100644
@@ -107,6 +107,10 @@ static const struct of_device_id st_accel_of_match[] = {
 		.compatible = "st,lis2hh12",
 		.data = LIS2HH12_ACCEL_DEV_NAME,
 	},
+	{
+		.compatible = "silan,sc7a20",
+		.data = SC7A20_ACCEL_DEV_NAME,
+	},
 	{},
 };
 MODULE_DEVICE_TABLE(of, st_accel_of_match);
@@ -142,6 +146,7 @@ static const struct i2c_device_id st_accel_id_table[] = {
 	{ LIS3DE_ACCEL_DEV_NAME },
 	{ LIS2DE12_ACCEL_DEV_NAME },
 	{ LIS2HH12_ACCEL_DEV_NAME },
+	{ SC7A20_ACCEL_DEV_NAME },
 	{},
 };
 MODULE_DEVICE_TABLE(i2c, st_accel_id_table);
#6 lore raw

Re: [PATCH 3/3] iio: accel: st_accel: Add support for Silan SC7A20

Quoted 2 lines · Jonathan Cameron

This is nasty as 0x11 is a valid who am I for an LIS3DHH which is a very different device.

The WAI value is just a second line check anyway, the device is probed from the compatible, this is just checked to make sure the device is what it says it is. Annoying, but it will do the job.

Quoted 6 lines

It's worth noting that a good part of the maintenance of this driver is done by ST employees who are probably less than happy with seeing
a clone (with bugs) supported. So support is likely to be somewhat best effort / not going to deliberately break support for this part but no means to test it. Hopefully we'll remember to cc you to test anything new added to the driver. Clones aren't always perfect!

Linux began with people running it on IBM PC clones... ;)

I would assume good faith, especially with the silicon shortage plaguing us it is
perfectly possible that this is a fully licensed copy. I think the two derivatives of the AK8974 magnetometer for example are
legally transferred IP, we just don't know the details behind the curtain.

Yours,
Linus Walleij