Skip to contentSkip to conversation

power: supply: qcom_battmgr: report the USB type from USB_ADAP_TYPE

PATCH Superseded Joonhoe Kim 2 participants 2 messages Oct 7 – Oct 12

Subsystems (2)

v1 of 2 Superseded Latest is v2 →

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 power: supply: qcom_battmgr: report the USB type from USB_ADAP_TYPE

On SM8350 and SM8550 class firmware the USB power supply reads its usb_type from the USB_TYPE property. On a Kaanapali (SM8850) tablet that property is 1 (SDP) whatever is plugged in, while USB_ADAP_TYPE reports the adapter, using the same numbering as POWER_SUPPLY_USB_TYPE_*:

  PC port                  USB_TYPE 1   USB_ADAP_TYPE 1 (SDP)
  USB-C dock, 5 V PD       USB_TYPE 1   USB_ADAP_TYPE 6 (PD)
  65 W PPS charger         USB_TYPE 1   USB_ADAP_TYPE 8 (PD_PPS)

Qualcomm's downstream qti_battery_charger driver also maps USB_ADAP_TYPE to POWER_SUPPLY_PROP_USB_TYPE.

Request USB_ADAP_TYPE for POWER_SUPPLY_PROP_USB_TYPE instead.

Tested on a Lenovo Legion Tab Y700 gen 5 (qcom,kaanapali-pmic-glink, which uses the SM8550 path). Only tested on this device; the SM8350 variant shares the property map and was not tested.

Assisted-by LLM

---
The downstream qti_battery_charger in SM8350, SM8450, SM8550, SM8650 and SM7635 vendor kernels uses the same property ids, maps USB_ADAP_TYPE to POWER_SUPPLY_PROP_USB_TYPE and reads it on every USB change, so the firmware of this family should implement it. Like downstream, values outside the enum (QC chargers, 0x80 and up there) are passed through.

Not tested with a plain DCP or a QC charger. Once, when the USB gadget did not connect to the PC after plugging, USB_ADAP_TYPE read 4 (ACA) until the cable was plugged again; when the gadget connects, the PC port reads 1 (SDP).

Kamal Wadhwa's multi-port series makes the same switch for X1E80100 (a new x1e80100_usb_prop_map) [1]; this patch is for the shared SM8350/ SM8550 map, which that series leaves on USB_TYPE.

[1] https://lore.kernel.org/all/20260806-b4-battmgr-multiport-usb-v2-1-1a6dd1e06cc2@oss.qualcomm.com/

 drivers/power/supply/qcom_battmgr.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/power/supply/qcom_battmgr.c +2 -2
index 7716eb9e1aff..b24860d2ded2 100644
@@ -975,7 +975,7 @@ static const u8 sm8350_usb_prop_map[] = {
 	[POWER_SUPPLY_PROP_CURRENT_NOW] = USB_CURR_NOW,
 	[POWER_SUPPLY_PROP_CURRENT_MAX] = USB_CURR_MAX,
 	[POWER_SUPPLY_PROP_INPUT_CURRENT_LIMIT] = USB_INPUT_CURR_LIMIT,
-	[POWER_SUPPLY_PROP_USB_TYPE] = USB_TYPE,
+	[POWER_SUPPLY_PROP_USB_TYPE] = USB_ADAP_TYPE,
 };
 
 static int qcom_battmgr_usb_sm8350_update(struct qcom_battmgr *battmgr,
@@ -1513,7 +1513,7 @@ static void qcom_battmgr_sm8350_callback(struct qcom_battmgr *battmgr,
 		case USB_INPUT_CURR_LIMIT:
 			battmgr->usb.current_limit = le32_to_cpu(resp->intval.value);
 			break;
-		case USB_TYPE:
+		case USB_ADAP_TYPE:
 			battmgr->usb.usb_type = le32_to_cpu(resp->intval.value);
 			break;
 		default:
Signature · 1 line
2.43.0
Reviewed-by on v1 1/1
#1 lore raw
Quoted 12 lines · Joonhoe Kim

On SM8350 and SM8550 class firmware the USB power supply reads its usb_type from the USB_TYPE property. On a Kaanapali (SM8850) tablet that property is 1 (SDP) whatever is plugged in, while USB_ADAP_TYPE reports the adapter, using the same numbering as POWER_SUPPLY_USB_TYPE_*:

PC port USB_TYPE 1 USB_ADAP_TYPE 1 (SDP) USB-C dock, 5 V
PD USB_TYPE 1 USB_ADAP_TYPE 6 (PD) 65 W PPS charger
USB_TYPE 1 USB_ADAP_TYPE 8 (PD_PPS)

Qualcomm's downstream qti_battery_charger driver also maps USB_ADAP_TYPE to POWER_SUPPLY_PROP_USB_TYPE.

Quoted 8 lines

Request USB_ADAP_TYPE for POWER_SUPPLY_PROP_USB_TYPE instead.

Tested on a Lenovo Legion Tab Y700 gen 5 (qcom,kaanapali-pmic-glink, which uses the SM8550 path). Only tested on this device; the SM8350 variant shares the property map and was not tested.

Assisted-by LLM

Add fixes tag, with that i think it should be good to go.

Quoted 45 lines

--- The downstream qti_battery_charger in SM8350, SM8450, SM8550, SM8650 and SM7635 vendor kernels uses the same property ids, maps USB_ADAP_TYPE to POWER_SUPPLY_PROP_USB_TYPE and reads it on every USB change, so the firmware of this family should implement it. Like downstream, values outside the enum (QC chargers, 0x80 and up there) are passed through.

Not tested with a plain DCP or a QC charger. Once, when the USB gadget did not connect to the PC after plugging, USB_ADAP_TYPE read 4 (ACA) until the cable was plugged again; when the gadget connects, the PC port reads 1 (SDP).

Kamal Wadhwa's multi-port series makes the same switch for X1E80100 (a new x1e80100_usb_prop_map) [1]; this patch is for the shared SM8350/ SM8550 map, which that series leaves on USB_TYPE.

[1] https://lore.kernel.org/all/20260806-b4-battmgr-multiport-usb-v2-1-1a6dd1e06cc2@oss.qualcomm.com/

 drivers/power/supply/qcom_battmgr.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/power/supply/qcom_battmgr.c +2 -2
index 7716eb9e1aff..b24860d2ded2 100644
@@ -975,7 +975,7 @@ static const u8 sm8350_usb_prop_map[] = {
 	[POWER_SUPPLY_PROP_CURRENT_NOW] = USB_CURR_NOW,
 	[POWER_SUPPLY_PROP_CURRENT_MAX] = USB_CURR_MAX,
 	[POWER_SUPPLY_PROP_INPUT_CURRENT_LIMIT] = USB_INPUT_CURR_LIMIT,
-	[POWER_SUPPLY_PROP_USB_TYPE] = USB_TYPE,
+	[POWER_SUPPLY_PROP_USB_TYPE] = USB_ADAP_TYPE,
 };
 
 static int qcom_battmgr_usb_sm8350_update(struct qcom_battmgr *battmgr,
@@ -1513,7 +1513,7 @@ static void qcom_battmgr_sm8350_callback(struct qcom_battmgr *battmgr,
 		case USB_INPUT_CURR_LIMIT:
 			battmgr->usb.current_limit = le32_to_cpu(resp->intval.value);
 			break;
-		case USB_TYPE:
+		case USB_ADAP_TYPE:
 			battmgr->usb.usb_type = le32_to_cpu(resp->intval.value);
 			break;
 		default:
Signature · 1 line
2.43.0