[PATCH v1] Bluetooth: hci_qca: Optimizations in init sequence for WCN6750.

[PATCH v1] Bluetooth: hci_qca: Optimizations in init sequence for WCN6750.

[pharish]

This change adds optimazation in init sequence for WCN6750.

Signed-off-by: pharish <pharish@codeaurora.org>
---
 drivers/bluetooth/hci_qca.c | 18 +++++++++++-------
 1 file changed, 11 insertions(+), 7 deletions(-)

diff --git a/drivers/bluetooth/hci_qca.c b/drivers/bluetooth/hci_qca.c
index 6f44b26..4dedaaa 100644
--- a/drivers/bluetooth/hci_qca.c
+++ b/drivers/bluetooth/hci_qca.c
@@ -1603,15 +1603,19 @@ static int qca_regulator_init(struct hci_uart *hu)
 	 */
 	qcadev = serdev_device_get_drvdata(hu->serdev);
 	if (!qcadev->bt_power->vregs_on) {
-		serdev_device_close(hu->serdev);
+		if (qca_is_wcn399x(soc_type))
+			serdev_device_close(hu->serdev);
+
 		ret = qca_regulator_enable(qcadev);
 		if (ret)
 			return ret;
 
-		ret = serdev_device_open(hu->serdev);
-		if (ret) {
-			bt_dev_err(hu->hdev, "failed to open port");
-			return ret;
+		if (qca_is_wcn399x(soc_type)) {
+			ret = serdev_device_open(hu->serdev);
+			if (ret) {
+				bt_dev_err(hu->hdev, "failed to open port");
+				return ret;
+			}
 		}
 	}
 
@@ -1635,9 +1639,8 @@ static int qca_regulator_init(struct hci_uart *hu)
 		}
 	}
 
-	qca_set_speed(hu, QCA_INIT_SPEED);
-
 	if (qca_is_wcn399x(soc_type)) {
+		qca_set_speed(hu, QCA_INIT_SPEED);
 		ret = qca_send_power_pulse(hu, true);
 		if (ret)
 			return ret;
@@ -1648,6 +1651,7 @@ static int qca_regulator_init(struct hci_uart *hu)
 	 * Without this, we will have RTS and CTS synchronization
 	 * issues.
 	 */
+
 	serdev_device_close(hu->serdev);
 	ret = serdev_device_open(hu->serdev);
 	if (ret) {
-- 
QUALCOMM INDIA, on behalf of Qualcomm Innovation Center, Inc. is a member 
of Code Aurora Forum, hosted by The Linux Foundation

[PATCH v1] Bluetooth: hci_qca: Stop IBS timer during BT OFF

[pharish]

This change stops IBS timers during BT OFF.

Signed-off-by: pharish <pharish@codeaurora.org>
---
 drivers/bluetooth/hci_qca.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/bluetooth/hci_qca.c b/drivers/bluetooth/hci_qca.c
index dd768a8..6f44b26 100644
--- a/drivers/bluetooth/hci_qca.c
+++ b/drivers/bluetooth/hci_qca.c
@@ -1928,6 +1928,9 @@ static int qca_power_off(struct hci_dev *hdev)
 	hu->hdev->hw_error = NULL;
 	hu->hdev->cmd_timeout = NULL;
 
+	mod_timer(&qca->tx_idle_timer, 0);
+	mod_timer(&qca->wake_retrans_timer, 0);
+
 	/* Stop sending shutdown command if soc crashes. */
 	if (soc_type != QCA_ROME
 		&& qca->memdump_state == QCA_MEMDUMP_IDLE) {
-- 
QUALCOMM INDIA, on behalf of Qualcomm Innovation Center, Inc. is a member 
of Code Aurora Forum, hosted by The Linux Foundation

Re: [PATCH v1] Bluetooth: hci_qca: Stop IBS timer during BT OFF

[Marcel Holtmann]

Hi,

> This change stops IBS timers during BT OFF.
> 
> Signed-off-by: pharish <pharish@codeaurora.org>

clear name please.

> ---
> drivers/bluetooth/hci_qca.c | 3 +++
> 1 file changed, 3 insertions(+)
> 
> diff --git a/drivers/bluetooth/hci_qca.c b/drivers/bluetooth/hci_qca.c
> index dd768a8..6f44b26 100644
> --- a/drivers/bluetooth/hci_qca.c
> +++ b/drivers/bluetooth/hci_qca.c
> @@ -1928,6 +1928,9 @@ static int qca_power_off(struct hci_dev *hdev)
> 	hu->hdev->hw_error = NULL;
> 	hu->hdev->cmd_timeout = NULL;
> 
> +	mod_timer(&qca->tx_idle_timer, 0);
> +	mod_timer(&qca->wake_retrans_timer, 0);
> +

And I would really prefer if this gets changed to use a workqueue instead of a timer.

Regards

Marcel

Re: [PATCH v1] Bluetooth: hci_qca: Optimizations in init sequence for WCN6750.

[Marcel Holtmann]

Hi,

> This change adds optimazation in init sequence for WCN6750.

you need to explain what you are doing.
> 
> Signed-off-by: pharish <pharish@codeaurora.org>
> ---
> drivers/bluetooth/hci_qca.c | 18 +++++++++++-------
> 1 file changed, 11 insertions(+), 7 deletions(-)
> 
> diff --git a/drivers/bluetooth/hci_qca.c b/drivers/bluetooth/hci_qca.c
> index 6f44b26..4dedaaa 100644
> --- a/drivers/bluetooth/hci_qca.c
> +++ b/drivers/bluetooth/hci_qca.c
> @@ -1603,15 +1603,19 @@ static int qca_regulator_init(struct hci_uart *hu)
> 	 */
> 	qcadev = serdev_device_get_drvdata(hu->serdev);
> 	if (!qcadev->bt_power->vregs_on) {
> -		serdev_device_close(hu->serdev);
> +		if (qca_is_wcn399x(soc_type))
> +			serdev_device_close(hu->serdev);
> +
> 		ret = qca_regulator_enable(qcadev);
> 		if (ret)
> 			return ret;
> 
> -		ret = serdev_device_open(hu->serdev);
> -		if (ret) {
> -			bt_dev_err(hu->hdev, "failed to open port");
> -			return ret;
> +		if (qca_is_wcn399x(soc_type)) {
> +			ret = serdev_device_open(hu->serdev);
> +			if (ret) {
> +				bt_dev_err(hu->hdev, "failed to open port");
> +				return ret;
> +			}

I am really not doing this. You need to stop doing this SoC X or SoC Y crap. I said this before and I am saying this again, start from scratch and write a clean serdev only driver. Adding things to hci_qca is just a hack.

> 		}
> 	}
> 
> @@ -1635,9 +1639,8 @@ static int qca_regulator_init(struct hci_uart *hu)
> 		}
> 	}
> 
> -	qca_set_speed(hu, QCA_INIT_SPEED);
> -
> 	if (qca_is_wcn399x(soc_type)) {
> +		qca_set_speed(hu, QCA_INIT_SPEED);
> 		ret = qca_send_power_pulse(hu, true);
> 		if (ret)
> 			return ret;
> @@ -1648,6 +1651,7 @@ static int qca_regulator_init(struct hci_uart *hu)
> 	 * Without this, we will have RTS and CTS synchronization
> 	 * issues.
> 	 */
> +

Unrelated change.

> 	serdev_device_close(hu->serdev);
> 	ret = serdev_device_open(hu->serdev);
> 	if (ret) {

Regards

Marcel

Re: [PATCH v1] Bluetooth: hci_qca: Stop IBS timer during BT OFF

[pharish]

On 2021-11-26 00:12, Marcel Holtmann wrote:
> Hi,
> 
>> This change stops IBS timers during BT OFF.
>> 
>> Signed-off-by: pharish <pharish@codeaurora.org>
> 
> clear name please.

[Harish]: will update

> 
>> ---
>> drivers/bluetooth/hci_qca.c | 3 +++
>> 1 file changed, 3 insertions(+)
>> 
>> diff --git a/drivers/bluetooth/hci_qca.c b/drivers/bluetooth/hci_qca.c
>> index dd768a8..6f44b26 100644
>> --- a/drivers/bluetooth/hci_qca.c
>> +++ b/drivers/bluetooth/hci_qca.c
>> @@ -1928,6 +1928,9 @@ static int qca_power_off(struct hci_dev *hdev)
>> 	hu->hdev->hw_error = NULL;
>> 	hu->hdev->cmd_timeout = NULL;
>> 
>> +	mod_timer(&qca->tx_idle_timer, 0);
>> +	mod_timer(&qca->wake_retrans_timer, 0);
>> +
> 
> And I would really prefer if this gets changed to use a workqueue
> instead of a timer.

[Harish]: The full implementation of IBS is based on timers
to that reason I have switched to timers.

> 
> Regards
> 
> Marcel