Fix several issues discovered while debugging UCSI implementation on Qualcomm platforms (ucsi_glink). With these patches I was able to get a working Type-C port managament implementation. Tested on SC8280XP (Lenovo X13s laptop) and SM8350-HDK.
Signed-off-by: Dmitry Baryshkov dmitry.baryshkov@linaro.org --- Dmitry Baryshkov (7): usb: typec: ucsi: fix race condition in connection change ACK'ing usb: typec: ucsi: acknowledge the UCSI_CCI_NOT_SUPPORTED usb: typec: ucsi: make ACK_CC_CI rules more obvious usb: typec: ucsi: allow non-partner GET_PDOS for Qualcomm devices usb: typec: ucsi: limit the UCSI_NO_PARTNER_PDOS even further usb: typec: ucsi: properly register partner's PD device soc: qcom: pmic_glink: reenable UCSI on sc8280xp
drivers/soc/qcom/pmic_glink.c | 1 - drivers/usb/typec/ucsi/ucsi.c | 51 +++++++++++++++++++++++++++++++++---------- 2 files changed, 39 insertions(+), 13 deletions(-) --- base-commit: ea86f16f605361af3779af0e0b4be18614282091 change-id: 20240312-qcom-ucsi-fixes-6578d236b60b
Best regards,
The code to handle connection change events contains a race: there is an open window for notifications to arrive between clearing EVENT_PENDING bit and sending the ACK_CC_CI command to acknowledge the connection change. This is mostly not an issue, but on Qualcomm platforms when the PPM receives ACK_CC_CI with the ConnectorChange bit set if there is no pending reported Connector Change, it responds with the CommandCompleted + NotSupported notifications, completely breaking UCSI state machine.
Fix this by reading out CCI after ACK_CC_CI and scheduling the work if there is a connector change reported.
Fixes: bdc62f2bae8f ("usb: typec: ucsi: Simplified registration and I/O API") Cc: stable@vger.kernel.org Signed-off-by: Dmitry Baryshkov dmitry.baryshkov@linaro.org --- drivers/usb/typec/ucsi/ucsi.c | 20 +++++++++++++++++--- 1 file changed, 17 insertions(+), 3 deletions(-)
diff --git a/drivers/usb/typec/ucsi/ucsi.c b/drivers/usb/typec/ucsi/ucsi.c index cf52cb34d285..4abb752c6806 100644 --- a/drivers/usb/typec/ucsi/ucsi.c +++ b/drivers/usb/typec/ucsi/ucsi.c @@ -61,12 +61,28 @@ static int ucsi_acknowledge_command(struct ucsi *ucsi)
static int ucsi_acknowledge_connector_change(struct ucsi *ucsi) { + unsigned int con_num; u64 ctrl; + u32 cci; + int ret;
ctrl = UCSI_ACK_CC_CI; ctrl |= UCSI_ACK_CONNECTOR_CHANGE;
- return ucsi->ops->sync_write(ucsi, UCSI_CONTROL, &ctrl, sizeof(ctrl)); + ret = ucsi->ops->sync_write(ucsi, UCSI_CONTROL, &ctrl, sizeof(ctrl)); + if (ret) + return ret; + + clear_bit(EVENT_PENDING, &ucsi->flags); + ret = ucsi->ops->read(ucsi, UCSI_CCI, &cci, sizeof(cci)); + if (ret) + return ret; + + con_num = UCSI_CCI_CONNECTOR(cci); + if (con_num) + ucsi_connector_change(ucsi, con_num); + + return 0; }
static int ucsi_exec_command(struct ucsi *ucsi, u64 command); @@ -1215,8 +1231,6 @@ static void ucsi_handle_connector_change(struct work_struct *work) if (con->status.change & UCSI_CONSTAT_CAM_CHANGE) ucsi_partner_task(con, ucsi_check_altmodes, 1, 0);
- clear_bit(EVENT_PENDING, &con->ucsi->flags); - mutex_lock(&ucsi->ppm_lock); ret = ucsi_acknowledge_connector_change(ucsi); mutex_unlock(&ucsi->ppm_lock);
Hi Dmitry,
On Wed, Mar 13, 2024 at 05:54:11AM +0200, Dmitry Baryshkov wrote:
The code to handle connection change events contains a race: there is an open window for notifications to arrive between clearing EVENT_PENDING bit and sending the ACK_CC_CI command to acknowledge the connection change. This is mostly not an issue, but on Qualcomm platforms when the PPM receives ACK_CC_CI with the ConnectorChange bit set if there is no pending reported Connector Change, it responds with the CommandCompleted
- NotSupported notifications, completely breaking UCSI state machine.
Fix this by reading out CCI after ACK_CC_CI and scheduling the work if there is a connector change reported.
Fixes: bdc62f2bae8f ("usb: typec: ucsi: Simplified registration and I/O API") Cc: stable@vger.kernel.org Signed-off-by: Dmitry Baryshkov dmitry.baryshkov@linaro.org
UCSI specification quite clearly states that the PPM must wait until OPM has acknowledged the notification before sending the next notification, so this looks like a workaround for Qualcomm specific issue. Ideally it would have been isolated - now this is done on every platform.
I'm a little bit uncomfortable with the unconditional reading of the CCI field. On most systems reading the field will clear it completely. Hopefully that will not cause more problems.
Reviewed-by: Heikki Krogerus heikki.krogerus@linux.intel.com
drivers/usb/typec/ucsi/ucsi.c | 20 +++++++++++++++++--- 1 file changed, 17 insertions(+), 3 deletions(-)
diff --git a/drivers/usb/typec/ucsi/ucsi.c b/drivers/usb/typec/ucsi/ucsi.c index cf52cb34d285..4abb752c6806 100644 --- a/drivers/usb/typec/ucsi/ucsi.c +++ b/drivers/usb/typec/ucsi/ucsi.c @@ -61,12 +61,28 @@ static int ucsi_acknowledge_command(struct ucsi *ucsi) static int ucsi_acknowledge_connector_change(struct ucsi *ucsi) {
- unsigned int con_num; u64 ctrl;
- u32 cci;
- int ret;
ctrl = UCSI_ACK_CC_CI; ctrl |= UCSI_ACK_CONNECTOR_CHANGE;
- return ucsi->ops->sync_write(ucsi, UCSI_CONTROL, &ctrl, sizeof(ctrl));
- ret = ucsi->ops->sync_write(ucsi, UCSI_CONTROL, &ctrl, sizeof(ctrl));
- if (ret)
return ret;- clear_bit(EVENT_PENDING, &ucsi->flags);
- ret = ucsi->ops->read(ucsi, UCSI_CCI, &cci, sizeof(cci));
- if (ret)
return ret;- con_num = UCSI_CCI_CONNECTOR(cci);
- if (con_num)
ucsi_connector_change(ucsi, con_num);- return 0;
} static int ucsi_exec_command(struct ucsi *ucsi, u64 command); @@ -1215,8 +1231,6 @@ static void ucsi_handle_connector_change(struct work_struct *work) if (con->status.change & UCSI_CONSTAT_CAM_CHANGE) ucsi_partner_task(con, ucsi_check_altmodes, 1, 0);
- clear_bit(EVENT_PENDING, &con->ucsi->flags);
- mutex_lock(&ucsi->ppm_lock); ret = ucsi_acknowledge_connector_change(ucsi); mutex_unlock(&ucsi->ppm_lock);
-- 2.39.2
When the PPM reports UCSI_CCI_NOT_SUPPORTED for the command, the flag remains set and no further commands are allowed to be processed until OPM acknowledges failed command completion using ACK_CC_CI. Add missing call to ucsi_acknowledge_command().
Fixes: bdc62f2bae8f ("usb: typec: ucsi: Simplified registration and I/O API") Cc: stable@vger.kernel.org Signed-off-by: Dmitry Baryshkov dmitry.baryshkov@linaro.org --- drivers/usb/typec/ucsi/ucsi.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/drivers/usb/typec/ucsi/ucsi.c b/drivers/usb/typec/ucsi/ucsi.c index 4abb752c6806..bde4f03b9aa2 100644 --- a/drivers/usb/typec/ucsi/ucsi.c +++ b/drivers/usb/typec/ucsi/ucsi.c @@ -167,8 +167,10 @@ static int ucsi_exec_command(struct ucsi *ucsi, u64 cmd) if (!(cci & UCSI_CCI_COMMAND_COMPLETE)) return -EIO;
- if (cci & UCSI_CCI_NOT_SUPPORTED) - return -EOPNOTSUPP; + if (cci & UCSI_CCI_NOT_SUPPORTED) { + ret = ucsi_acknowledge_command(ucsi); + return ret ? ret : -EOPNOTSUPP; + }
if (cci & UCSI_CCI_ERROR) { if (cmd == UCSI_GET_ERROR_STATUS)
On Wed, Mar 13, 2024 at 05:54:12AM +0200, Dmitry Baryshkov wrote:
When the PPM reports UCSI_CCI_NOT_SUPPORTED for the command, the flag remains set and no further commands are allowed to be processed until OPM acknowledges failed command completion using ACK_CC_CI. Add missing call to ucsi_acknowledge_command().
Fixes: bdc62f2bae8f ("usb: typec: ucsi: Simplified registration and I/O API") Cc: stable@vger.kernel.org Signed-off-by: Dmitry Baryshkov dmitry.baryshkov@linaro.org
Reviewed-by: Heikki Krogerus heikki.krogerus@linux.intel.com
drivers/usb/typec/ucsi/ucsi.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/drivers/usb/typec/ucsi/ucsi.c b/drivers/usb/typec/ucsi/ucsi.c index 4abb752c6806..bde4f03b9aa2 100644 --- a/drivers/usb/typec/ucsi/ucsi.c +++ b/drivers/usb/typec/ucsi/ucsi.c @@ -167,8 +167,10 @@ static int ucsi_exec_command(struct ucsi *ucsi, u64 cmd) if (!(cci & UCSI_CCI_COMMAND_COMPLETE)) return -EIO;
- if (cci & UCSI_CCI_NOT_SUPPORTED)
return -EOPNOTSUPP;
- if (cci & UCSI_CCI_NOT_SUPPORTED) {
ret = ucsi_acknowledge_command(ucsi);return ret ? ret : -EOPNOTSUPP;- }
if (cci & UCSI_CCI_ERROR) { if (cmd == UCSI_GET_ERROR_STATUS)
-- 2.39.2
On Wed, Mar 13, 2024 at 05:54:10AM +0200, Dmitry Baryshkov wrote:
Fix several issues discovered while debugging UCSI implementation on Qualcomm platforms (ucsi_glink). With these patches I was able to get a working Type-C port managament implementation. Tested on SC8280XP (Lenovo X13s laptop) and SM8350-HDK.
Dmitry Baryshkov (7): usb: typec: ucsi: fix race condition in connection change ACK'ing usb: typec: ucsi: acknowledge the UCSI_CCI_NOT_SUPPORTED usb: typec: ucsi: make ACK_CC_CI rules more obvious usb: typec: ucsi: allow non-partner GET_PDOS for Qualcomm devices usb: typec: ucsi: limit the UCSI_NO_PARTNER_PDOS even further usb: typec: ucsi: properly register partner's PD device
soc: qcom: pmic_glink: reenable UCSI on sc8280xp
I just gave this series a quick spin on my X13s and it seems there are still some issues that needs to be resolved before merging at least the final patch in this series:
[ 7.786167] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: unknown error 0 [ 7.786445] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: UCSI_GET_PDOS failed (-5) [ 7.883493] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: unknown error 0 [ 7.883614] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: UCSI_GET_PDOS failed (-5) [ 7.905194] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: unknown error 0 [ 7.905295] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: UCSI_GET_PDOS failed (-5) [ 7.913340] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: unknown error 0 [ 7.913409] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: UCSI_GET_PDOS failed (-5)
I see these errors on boot both with and without my charger and ethernet device connected.
I'm afraid I won't have to time to help debug this myself at least for another week.
Johan
On Fri, 22 Mar 2024 at 14:16, Johan Hovold johan@kernel.org wrote:
On Wed, Mar 13, 2024 at 05:54:10AM +0200, Dmitry Baryshkov wrote:
Fix several issues discovered while debugging UCSI implementation on Qualcomm platforms (ucsi_glink). With these patches I was able to get a working Type-C port managament implementation. Tested on SC8280XP (Lenovo X13s laptop) and SM8350-HDK.
Dmitry Baryshkov (7): usb: typec: ucsi: fix race condition in connection change ACK'ing usb: typec: ucsi: acknowledge the UCSI_CCI_NOT_SUPPORTED usb: typec: ucsi: make ACK_CC_CI rules more obvious usb: typec: ucsi: allow non-partner GET_PDOS for Qualcomm devices usb: typec: ucsi: limit the UCSI_NO_PARTNER_PDOS even further usb: typec: ucsi: properly register partner's PD device
soc: qcom: pmic_glink: reenable UCSI on sc8280xpI just gave this series a quick spin on my X13s and it seems there are still some issues that needs to be resolved before merging at least the final patch in this series:
[ 7.786167] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: unknown error 0 [ 7.786445] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: UCSI_GET_PDOS failed (-5) [ 7.883493] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: unknown error 0 [ 7.883614] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: UCSI_GET_PDOS failed (-5) [ 7.905194] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: unknown error 0 [ 7.905295] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: UCSI_GET_PDOS failed (-5) [ 7.913340] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: unknown error 0 [ 7.913409] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: UCSI_GET_PDOS failed (-5)
I see these errors on boot both with and without my charger and ethernet device connected.
Just to doublecheck: do you have latest adsp installed? Do you have your bootloaders updated?
If you back up the patch #5 ("limit the UCSI_NO_PARTNER_PDOS even further"), does it still break for you?
I'm afraid I won't have to time to help debug this myself at least for another week.
Johan
-- With best wishes Dmitry
On Fri, Mar 22, 2024 at 03:39:36PM +0200, Dmitry Baryshkov wrote:
On Fri, 22 Mar 2024 at 14:16, Johan Hovold johan@kernel.org wrote:
On Wed, Mar 13, 2024 at 05:54:10AM +0200, Dmitry Baryshkov wrote:
Dmitry Baryshkov (7): usb: typec: ucsi: fix race condition in connection change ACK'ing usb: typec: ucsi: acknowledge the UCSI_CCI_NOT_SUPPORTED usb: typec: ucsi: make ACK_CC_CI rules more obvious usb: typec: ucsi: allow non-partner GET_PDOS for Qualcomm devices usb: typec: ucsi: limit the UCSI_NO_PARTNER_PDOS even further usb: typec: ucsi: properly register partner's PD device
soc: qcom: pmic_glink: reenable UCSI on sc8280xpI just gave this series a quick spin on my X13s and it seems there are still some issues that needs to be resolved before merging at least the final patch in this series:
[ 7.786167] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: unknown error 0 [ 7.786445] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: UCSI_GET_PDOS failed (-5)
I see these errors on boot both with and without my charger and ethernet device connected.
Just to doublecheck: do you have latest adsp installed?
Yes, there has only been one adsp fw released to linux-firmware and that's the one I'm using.
If this depends on anything newer that should have been mentioned in the commit message and we obviously can't enable UCSI until that firmware has been released.
Do you have your bootloaders updated?
Yes.
If you back up the patch #5 ("limit the UCSI_NO_PARTNER_PDOS even further"), does it still break for you?
I don't understand what you mean by "back up the patch #5". Revert (drop) patch 5?
The errors are still there with patch 5 reverted.
Johan
On Fri, 22 Mar 2024 at 14:16, Johan Hovold johan@kernel.org wrote:
On Wed, Mar 13, 2024 at 05:54:10AM +0200, Dmitry Baryshkov wrote:
Fix several issues discovered while debugging UCSI implementation on Qualcomm platforms (ucsi_glink). With these patches I was able to get a working Type-C port managament implementation. Tested on SC8280XP (Lenovo X13s laptop) and SM8350-HDK.
Dmitry Baryshkov (7): usb: typec: ucsi: fix race condition in connection change ACK'ing usb: typec: ucsi: acknowledge the UCSI_CCI_NOT_SUPPORTED usb: typec: ucsi: make ACK_CC_CI rules more obvious usb: typec: ucsi: allow non-partner GET_PDOS for Qualcomm devices usb: typec: ucsi: limit the UCSI_NO_PARTNER_PDOS even further usb: typec: ucsi: properly register partner's PD device
soc: qcom: pmic_glink: reenable UCSI on sc8280xpI just gave this series a quick spin on my X13s and it seems there are still some issues that needs to be resolved before merging at least the final patch in this series:
[ 7.786167] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: unknown error 0 [ 7.786445] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: UCSI_GET_PDOS failed (-5) [ 7.883493] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: unknown error 0 [ 7.883614] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: UCSI_GET_PDOS failed (-5) [ 7.905194] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: unknown error 0 [ 7.905295] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: UCSI_GET_PDOS failed (-5) [ 7.913340] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: unknown error 0 [ 7.913409] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: UCSI_GET_PDOS failed (-5)
I have traced what is causing these messages. During UCSI startup the ucsi_register_port() function queries for PDOs associated with the on-board USB-C port. This is allowed by the spec. Qualcomm firmware detects that there is no PD-device connected and instead of returning corresponding set of PDOs returns Eerror Indicator set to 1b but then it returns zero error status in response to GET_ERROR_STATUS, causing "unknown error 0" code. I have checked the PPM, it doesn't even have the code to set the error status properly in this case (not to mention that asking for device's PDOs should not be an error, unless the command is inappropriate for the target.
Thus said, I think the driver is behaving correctly. Granted that these messages are harmless, we can ignore them for now. I'll later check if we can update PD information for the device's ports when PD device is attached. I have verified that once the PD device is attached, corresponding GET_PDOS command returns correct set of PD objects. Ccurrently the driver registers usb_power_delivery devices, but with neither source nor sink set of capabilities.
An alternative option is to drop patches 4 and 5, keeping 'NO_PARTNER_PDOS' quirk equivalent to 'don't send GET_PDOS at all'. However I'd like to abstain from this option, since it doesn't allow us to check PD capabilities of the attached device.
Heikki, Johan, WDYT?
For reference, here is a trace of relevant messages exchanged over the GLINK interface during UCSI bootstrap:
[ 11.030838] write: 00000000: 10 00 01 00 07 00 00 00
GET_PDOS(connection 1, Source, 3 PDOs)
[ 11.044171] write ack: 0 [ 11.044263] notify: 00000000: 00 00 00 c0 00 00 00 00
Command Complete, Error
[ 11.044458] read: 00000000: 00 01 00 00 00 00 00 c0 00 00 00 00 00 00 00 00 [ 11.044460] read: 00000010: e7 3f 00 00 00 00 00 00 02 00 20 01 00 03 30 01 [ 11.044462] read: 00000020: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 [ 11.059790] read: 00000000: 00 01 00 00 00 00 00 c0 00 00 00 00 00 00 00 00 [ 11.059797] read: 00000010: e7 3f 00 00 00 00 00 00 02 00 20 01 00 03 30 01 [ 11.059801] read: 00000020: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 [ 11.059814] write: 00000000: 04 00 02 00 00 00 00 00
Ack_CC command
[ 11.075509] write ack: 0 [ 11.075544] notify: 00000000: 00 00 00 20 00 00 00 00
Ack for Ack_CC
[ 11.091828] read: 00000000: 00 01 00 00 00 00 00 20 00 00 00 00 00 00 00 00 [ 11.091864] read: 00000010: e7 3f 00 00 00 00 00 00 02 00 20 01 00 03 30 01 [ 11.091879] read: 00000020: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 [ 11.092339] write: 00000000: 13 00 00 00 00 00 00 00
GET_ERROR_STATUS
[ 11.106398] write ack: 0 [ 11.106435] notify: 00000000: 00 10 00 80 00 00 00 00
command complete, 0x10 bytes of response
[ 11.122729] read: 00000000: 00 01 00 00 00 10 00 80 00 00 00 00 00 00 00 00 [ 11.122758] read: 00000010: 00 00 00 00 00 00 00 00 02 00 20 01 00 03 30 01 [ 11.122770] read: 00000020: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
Zero response data
[ 11.137523] read: 00000000: 00 01 00 00 00 10 00 80 00 00 00 00 00 00 00 00 [ 11.137548] read: 00000010: 00 00 00 00 00 00 00 00 02 00 20 01 00 03 30 01 [ 11.137559] read: 00000020: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 [ 11.153028] read: 00000000: 00 01 00 00 00 10 00 80 00 00 00 00 00 00 00 00 [ 11.153064] read: 00000010: 00 00 00 00 00 00 00 00 02 00 20 01 00 03 30 01 [ 11.153080] read: 00000020: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 [ 11.153492] write: 00000000: 04 00 02 00 00 00 00 00
Ack_CC for the GET_ERROR_STATUS command
[ 11.169060] write ack: 0 [ 11.169108] notify: 00000000: 00 00 00 20 00 00 00 00
Ack for ACK_CC
[ 11.184114] read: 00000000: 00 01 00 00 00 00 00 20 00 00 00 00 00 00 00 00 [ 11.184140] read: 00000010: 00 00 00 00 00 00 00 00 02 00 20 01 00 03 30 01 [ 11.184152] read: 00000020: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 [ 11.184548] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: unknown error 0 [ 11.184695] ------------[ cut here ]------------ [ 11.184703] WARNING: CPU: 2 PID: 28 at drivers/usb/typec/ucsi/ucsi.c:140 ucsi_exec_command+0x284/0x328 [typec_ucsi] [ 11.185488] Call trace: [ 11.185494] ucsi_exec_command+0x284/0x328 [typec_ucsi] [ 11.185519] ucsi_send_command+0x54/0x118 [typec_ucsi] [ 11.185543] ucsi_read_pdos+0x5c/0xdc [typec_ucsi] [ 11.185567] ucsi_get_pdos+0x30/0xa4 [typec_ucsi] [ 11.185590] ucsi_init_work+0x3bc/0x82c [typec_ucsi] [ 11.185614] process_one_work+0x148/0x2a0 [ 11.185638] worker_thread+0x2fc/0x40c [ 11.185655] kthread+0x110/0x114 [ 11.185668] ret_from_fork+0x10/0x20
Then comes the same log for the Connector=1, Sink PDOs, Connector=2 Source PDOs and finally Connector=2 Sink PDOs.
On Mon, Mar 25, 2024 at 10:56:21PM +0200, Dmitry Baryshkov wrote:
On Fri, 22 Mar 2024 at 14:16, Johan Hovold johan@kernel.org wrote:
I just gave this series a quick spin on my X13s and it seems there are still some issues that needs to be resolved before merging at least the final patch in this series:
[ 7.786167] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: unknown error 0 [ 7.786445] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: UCSI_GET_PDOS failed (-5) [ 7.883493] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: unknown error 0 [ 7.883614] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: UCSI_GET_PDOS failed (-5) [ 7.905194] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: unknown error 0 [ 7.905295] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: UCSI_GET_PDOS failed (-5) [ 7.913340] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: unknown error 0 [ 7.913409] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: UCSI_GET_PDOS failed (-5)
I have traced what is causing these messages. During UCSI startup the ucsi_register_port() function queries for PDOs associated with the on-board USB-C port. This is allowed by the spec. Qualcomm firmware detects that there is no PD-device connected and instead of returning corresponding set of PDOs returns Eerror Indicator set to 1b but then it returns zero error status in response to GET_ERROR_STATUS, causing "unknown error 0" code. I have checked the PPM, it doesn't even have the code to set the error status properly in this case (not to mention that asking for device's PDOs should not be an error, unless the command is inappropriate for the target.
Thus said, I think the driver is behaving correctly. Granted that these messages are harmless, we can ignore them for now. I'll later check if we can update PD information for the device's ports when PD device is attached. I have verified that once the PD device is attached, corresponding GET_PDOS command returns correct set of PD objects. Ccurrently the driver registers usb_power_delivery devices, but with neither source nor sink set of capabilities.
An alternative option is to drop patches 4 and 5, keeping 'NO_PARTNER_PDOS' quirk equivalent to 'don't send GET_PDOS at all'. However I'd like to abstain from this option, since it doesn't allow us to check PD capabilities of the attached device.
Heikki, Johan, WDYT?
Whatever you do, you need to suppress those errors above before enabling anything more on sc8280xp (e.g. even if it means adding a quirk to the driver).
Johan
On Tue, 26 Mar 2024 at 10:41, Johan Hovold johan@kernel.org wrote:
On Mon, Mar 25, 2024 at 10:56:21PM +0200, Dmitry Baryshkov wrote:
On Fri, 22 Mar 2024 at 14:16, Johan Hovold johan@kernel.org wrote:
I just gave this series a quick spin on my X13s and it seems there are still some issues that needs to be resolved before merging at least the final patch in this series:
[ 7.786167] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: unknown error 0 [ 7.786445] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: UCSI_GET_PDOS failed (-5) [ 7.883493] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: unknown error 0 [ 7.883614] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: UCSI_GET_PDOS failed (-5) [ 7.905194] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: unknown error 0 [ 7.905295] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: UCSI_GET_PDOS failed (-5) [ 7.913340] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: unknown error 0 [ 7.913409] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: UCSI_GET_PDOS failed (-5)
I have traced what is causing these messages. During UCSI startup the ucsi_register_port() function queries for PDOs associated with the on-board USB-C port. This is allowed by the spec. Qualcomm firmware detects that there is no PD-device connected and instead of returning corresponding set of PDOs returns Eerror Indicator set to 1b but then it returns zero error status in response to GET_ERROR_STATUS, causing "unknown error 0" code. I have checked the PPM, it doesn't even have the code to set the error status properly in this case (not to mention that asking for device's PDOs should not be an error, unless the command is inappropriate for the target.
Thus said, I think the driver is behaving correctly. Granted that these messages are harmless, we can ignore them for now. I'll later check if we can update PD information for the device's ports when PD device is attached. I have verified that once the PD device is attached, corresponding GET_PDOS command returns correct set of PD objects. Ccurrently the driver registers usb_power_delivery devices, but with neither source nor sink set of capabilities.
An alternative option is to drop patches 4 and 5, keeping 'NO_PARTNER_PDOS' quirk equivalent to 'don't send GET_PDOS at all'. However I'd like to abstain from this option, since it doesn't allow us to check PD capabilities of the attached device.
Heikki, Johan, WDYT?
Whatever you do, you need to suppress those errors above before enabling anything more on sc8280xp (e.g. even if it means adding a quirk to the driver).
Why? The errors are legitimate.
On Tue, 26 Mar 2024 at 12:22, Dmitry Baryshkov dmitry.baryshkov@linaro.org wrote:
On Tue, 26 Mar 2024 at 10:41, Johan Hovold johan@kernel.org wrote:
On Mon, Mar 25, 2024 at 10:56:21PM +0200, Dmitry Baryshkov wrote:
On Fri, 22 Mar 2024 at 14:16, Johan Hovold johan@kernel.org wrote:
I just gave this series a quick spin on my X13s and it seems there are still some issues that needs to be resolved before merging at least the final patch in this series:
[ 7.786167] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: unknown error 0 [ 7.786445] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: UCSI_GET_PDOS failed (-5) [ 7.883493] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: unknown error 0 [ 7.883614] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: UCSI_GET_PDOS failed (-5) [ 7.905194] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: unknown error 0 [ 7.905295] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: UCSI_GET_PDOS failed (-5) [ 7.913340] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: unknown error 0 [ 7.913409] ucsi_glink.pmic_glink_ucsi pmic_glink.ucsi.0: UCSI_GET_PDOS failed (-5)
I have traced what is causing these messages. During UCSI startup the ucsi_register_port() function queries for PDOs associated with the on-board USB-C port. This is allowed by the spec. Qualcomm firmware detects that there is no PD-device connected and instead of returning corresponding set of PDOs returns Eerror Indicator set to 1b but then it returns zero error status in response to GET_ERROR_STATUS, causing "unknown error 0" code. I have checked the PPM, it doesn't even have the code to set the error status properly in this case (not to mention that asking for device's PDOs should not be an error, unless the command is inappropriate for the target.
Thus said, I think the driver is behaving correctly. Granted that these messages are harmless, we can ignore them for now. I'll later check if we can update PD information for the device's ports when PD device is attached. I have verified that once the PD device is attached, corresponding GET_PDOS command returns correct set of PD objects. Ccurrently the driver registers usb_power_delivery devices, but with neither source nor sink set of capabilities.
An alternative option is to drop patches 4 and 5, keeping 'NO_PARTNER_PDOS' quirk equivalent to 'don't send GET_PDOS at all'. However I'd like to abstain from this option, since it doesn't allow us to check PD capabilities of the attached device.
Heikki, Johan, WDYT?
Whatever you do, you need to suppress those errors above before enabling anything more on sc8280xp (e.g. even if it means adding a quirk to the driver).
Why? The errors are legitimate.
Just to expand my answer. We have the 'driver should be quiet by default' policy. Which usually means that there should be no 'foo registered' or 'foo bound' messages if everything goes smooth. We do not have 'don't warn about manufacturer issues' or 'don't barf if firmware returns an error' kind of requirements. We can be nicer and skip something if we know that it may return an error. But we don't have to.
-- With best wishes Dmitry
linux-stable-mirror@lists.linaro.org
-
Dmitry Baryshkov -
Heikki Krogerus -
Johan Hovold