[PATCH v3 0/4] Bug fixes for Qualcomm BT chip wcn3990

[PATCH v3 0/4] Bug fixes for Qualcomm BT chip wcn3990

[Balakrishna Godavarthi]

The below issues are found in the recent testing.

1. Observed device is not going into off state or not responding.
    As wcn3990 require a power pulses to turn on the irrespctive of
    igniting regulators, it was observed that power on or power off
    pulses are not in sync with respective to chip.
    The below patch will help us to wait until byte is pushed on to wires.          

    * Bluetooth: hci_qca: use wait_until_sent() for power pulses

2. Observed Chip responding when we are in sleep.
   This is due to turn on flow control during change baudrate request.
   The below patch will only pull the RTS line high instead of turning off
   the flow.

   * Bluetooth: hci_qca: Pull RTS line high for baudrate change command.

3. Frame reassembly errors splashing on console.
   wcn3990 requires will use multiple baudrates during booting stage.
   i.e. 2400 bps while sending power off pulse
    115200 bps while sending power on pulse
    port close
    port open
    set baudrate to 115200 request the chip version.

  during above process, we are seeing some stray bytes coming up
  on the UART Rx FIFO it could be due to frequent baudrate change.

  This patch will stop the frame reassembly errors.

  * Bluetooth: hci_qca: Fix frame reassembly errors for wcn3990

4. Disable IBS state machine and flush Tx buffer
   We are disabling IBS and flushing the Tx buffer before turning off the chip.
  
   This is due to IBS state machine is active when we turn off the chip.
   This will stop queuing IBS protocol bytes.

   * Bluetooth: hci_qca: Disable IBS state machine and flush Tx buffer

Changes in v3:
 * moved IBS & qca_flush to different patch
 * updated comments in code fo Deassert RTS patch

Changes in v2:
 * Removed Bluetooth: hci_qca: clear HCI_QUIRK_RESET_ON_CLOSE flag patch
   will be sending as separate series.
Balakrishna Godavarthi (4):
  Bluetooth: hci_qca: use wait_until_sent() for power pulses
  Bluetooth: hci_qca: Deassert RTS while baudrate change command
  Bluetooth: hci_qca: Fix frame reassembly errors for wcn3990
  Bluetooth: hci_qca: Disable IBS state machine and flush Tx buffer

 drivers/bluetooth/hci_qca.c | 77 +++++++++++++++++++++----------------
 1 file changed, 43 insertions(+), 34 deletions(-)

-- 
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
a Linux Foundation Collaborative Project

[PATCH v3 1/4] Bluetooth: hci_qca: use wait_until_sent() for power pulses

[Balakrishna Godavarthi]

wcn3990 requires a power pulse to turn ON/OFF along with
regulators. Sometimes we are observing the power pulses are sent
out with some time delay, due to queuing these commands. This is
causing synchronization issues with chip, which intern delay the
chip setup or may end up with communication issues.

Signed-off-by: Balakrishna Godavarthi <bgodavar@codeaurora.org>
---
v3:
  * no change.
v2: 
  * Updated function qca_send_power_pulse()
  * addressed reviewer comments.

v1:
 * initial patch
---
 drivers/bluetooth/hci_qca.c | 37 +++++++++++++------------------------
 1 file changed, 13 insertions(+), 24 deletions(-)

diff --git a/drivers/bluetooth/hci_qca.c b/drivers/bluetooth/hci_qca.c
index f036c8f98ea3..f5dd323c1967 100644
--- a/drivers/bluetooth/hci_qca.c
+++ b/drivers/bluetooth/hci_qca.c
@@ -1013,11 +1013,9 @@ static inline void host_set_baudrate(struct hci_uart *hu, unsigned int speed)
 		hci_uart_set_baudrate(hu, speed);
 }
 
-static int qca_send_power_pulse(struct hci_dev *hdev, u8 cmd)
+static int qca_send_power_pulse(struct hci_uart *hu, u8 cmd)
 {
-	struct hci_uart *hu = hci_get_drvdata(hdev);
-	struct qca_data *qca = hu->priv;
-	struct sk_buff *skb;
+	int ret;
 
 	/* These power pulses are single byte command which are sent
 	 * at required baudrate to wcn3990. On wcn3990, we have an external
@@ -1029,19 +1027,16 @@ static int qca_send_power_pulse(struct hci_dev *hdev, u8 cmd)
 	 * save power. Disabling hardware flow control is mandatory while
 	 * sending power pulses to SoC.
 	 */
-	bt_dev_dbg(hdev, "sending power pulse %02x to SoC", cmd);
-
-	skb = bt_skb_alloc(sizeof(cmd), GFP_KERNEL);
-	if (!skb)
-		return -ENOMEM;
-
+	bt_dev_dbg(hu->hdev, "sending power pulse %02x to SoC", cmd);
 	hci_uart_set_flow_control(hu, true);
+	ret = serdev_device_write(hu->serdev, &cmd, sizeof(cmd), 0);
+	if (ret < 0) {
+		bt_dev_err(hu->hdev, "failed to send power pulse %02x to SoC",
+			   cmd);
+		return ret;
+	}
 
-	skb_put_u8(skb, cmd);
-	hci_skb_pkt_type(skb) = HCI_COMMAND_PKT;
-
-	skb_queue_tail(&qca->txq, skb);
-	hci_uart_tx_wakeup(hu);
+	serdev_device_wait_until_sent(hu->serdev, 0);
 
 	/* Wait for 100 uS for SoC to settle down */
 	usleep_range(100, 200);
@@ -1116,7 +1111,6 @@ static int qca_set_speed(struct hci_uart *hu, enum qca_speed_type speed_type)
 
 static int qca_wcn3990_init(struct hci_uart *hu)
 {
-	struct hci_dev *hdev = hu->hdev;
 	struct qca_serdev *qcadev;
 	int ret;
 
@@ -1139,12 +1133,12 @@ static int qca_wcn3990_init(struct hci_uart *hu)
 
 	/* Forcefully enable wcn3990 to enter in to boot mode. */
 	host_set_baudrate(hu, 2400);
-	ret = qca_send_power_pulse(hdev, QCA_WCN3990_POWEROFF_PULSE);
+	ret = qca_send_power_pulse(hu, QCA_WCN3990_POWEROFF_PULSE);
 	if (ret)
 		return ret;
 
 	qca_set_speed(hu, QCA_INIT_SPEED);
-	ret = qca_send_power_pulse(hdev, QCA_WCN3990_POWERON_PULSE);
+	ret = qca_send_power_pulse(hu, QCA_WCN3990_POWERON_PULSE);
 	if (ret)
 		return ret;
 
@@ -1274,13 +1268,8 @@ static const struct qca_vreg_data qca_soc_data = {
 
 static void qca_power_shutdown(struct hci_uart *hu)
 {
-	struct serdev_device *serdev = hu->serdev;
-	unsigned char cmd = QCA_WCN3990_POWEROFF_PULSE;
-
 	host_set_baudrate(hu, 2400);
-	hci_uart_set_flow_control(hu, true);
-	serdev_device_write_buf(serdev, &cmd, sizeof(cmd));
-	hci_uart_set_flow_control(hu, false);
+	qca_send_power_pulse(hu, QCA_WCN3990_POWEROFF_PULSE);
 	qca_power_setup(hu, false);
 }
 
-- 
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
a Linux Foundation Collaborative Project

[PATCH v3 2/4] Bluetooth: hci_qca: Deassert RTS while baudrate change command

[Balakrishna Godavarthi]

This patch will help to stop frame reassembly errors while changing
the baudrate. This is because host send a change baudrate request
command to the chip with 115200 bps, Whereas chip will change their
UART clocks to the enable for new baudrate and sends the response
for the change request command with newer baudrate, On host side
we are still operating in 115200 bps which results of reading garbage
data. Here we are pulling RTS line, so that chip we will wait to send data
to host until host change its baudrate.

Signed-off-by: Balakrishna Godavarthi <bgodavar@codeaurora.org>
Tested-by: Matthias Kaehlcke <mka@chromium.org>
Reviewed-by: Matthias Kaehlcke <mka@chromium.org>
---
v3:
 * Updated comments in code to be more specific.
v2:
 * updated commit text and comments.
v1:
 * initial patch
---
 drivers/bluetooth/hci_qca.c | 24 +++++++++++++-----------
 1 file changed, 13 insertions(+), 11 deletions(-)

diff --git a/drivers/bluetooth/hci_qca.c b/drivers/bluetooth/hci_qca.c
index f5dd323c1967..253b5bae30c2 100644
--- a/drivers/bluetooth/hci_qca.c
+++ b/drivers/bluetooth/hci_qca.c
@@ -963,7 +963,6 @@ static int qca_set_baudrate(struct hci_dev *hdev, uint8_t baudrate)
 	struct hci_uart *hu = hci_get_drvdata(hdev);
 	struct qca_data *qca = hu->priv;
 	struct sk_buff *skb;
-	struct qca_serdev *qcadev;
 	u8 cmd[] = { 0x01, 0x48, 0xFC, 0x01, 0x00 };
 
 	if (baudrate > QCA_BAUDRATE_3200000)
@@ -977,13 +976,6 @@ static int qca_set_baudrate(struct hci_dev *hdev, uint8_t baudrate)
 		return -ENOMEM;
 	}
 
-	/* Disabling hardware flow control is mandatory while
-	 * sending change baudrate request to wcn3990 SoC.
-	 */
-	qcadev = serdev_device_get_drvdata(hu->serdev);
-	if (qcadev->btsoc_type == QCA_WCN3990)
-		hci_uart_set_flow_control(hu, true);
-
 	/* Assign commands to change baudrate and packet type. */
 	skb_put_data(skb, cmd, sizeof(cmd));
 	hci_skb_pkt_type(skb) = HCI_COMMAND_PKT;
@@ -999,9 +991,6 @@ static int qca_set_baudrate(struct hci_dev *hdev, uint8_t baudrate)
 	schedule_timeout(msecs_to_jiffies(BAUDRATE_SETTLE_TIMEOUT_MS));
 	set_current_state(TASK_RUNNING);
 
-	if (qcadev->btsoc_type == QCA_WCN3990)
-		hci_uart_set_flow_control(hu, false);
-
 	return 0;
 }
 
@@ -1086,6 +1075,7 @@ static int qca_check_speeds(struct hci_uart *hu)
 static int qca_set_speed(struct hci_uart *hu, enum qca_speed_type speed_type)
 {
 	unsigned int speed, qca_baudrate;
+	struct qca_serdev *qcadev;
 	int ret;
 
 	if (speed_type == QCA_INIT_SPEED) {
@@ -1097,6 +1087,15 @@ static int qca_set_speed(struct hci_uart *hu, enum qca_speed_type speed_type)
 		if (!speed)
 			return 0;
 
+		/* Deassert RTS while changing the baudrate of chip and host.
+		 * This will prevent chip from transmitting its response with
+		 * the new baudrate while the host port is still operating at
+		 * the old speed.
+		 */
+		qcadev = serdev_device_get_drvdata(hu->serdev);
+		if (qcadev->btsoc_type == QCA_WCN3990)
+			serdev_device_set_rts(hu->serdev, false);
+
 		qca_baudrate = qca_get_baudrate_value(speed);
 		bt_dev_dbg(hu->hdev, "Set UART speed to %d", speed);
 		ret = qca_set_baudrate(hu->hdev, qca_baudrate);
@@ -1104,6 +1103,9 @@ static int qca_set_speed(struct hci_uart *hu, enum qca_speed_type speed_type)
 			return ret;
 
 		host_set_baudrate(hu, speed);
+
+		if (qcadev->btsoc_type == QCA_WCN3990)
+			serdev_device_set_rts(hu->serdev, true);
 	}
 
 	return 0;
-- 
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
a Linux Foundation Collaborative Project

[PATCH v3 3/4] Bluetooth: hci_qca: Fix frame reassembly errors for wcn3990

[Balakrishna Godavarthi]

During initalization of wcn3990, we observed UART is reading some
stray bytes on the Rx line. This is logging Frame reassembly errors
on the serial console. This could be because of tristate of Tx line
of wcn3990 during boot up.

[  176.929612] Bluetooth: hci_qca.c:qca_recv() hci0: Frame reassembly failed (-84)
[  176.945734] Bluetooth: hci_qca.c:qca_recv() hci0: Frame reassembly failed (-84)
[  176.953298] Bluetooth: hci_qca.c:qca_recv() hci0: Frame reassembly failed (-84)
[  177.010660] Bluetooth: hci_qca.c:qca_recv() hci0: Frame reassembly failed (-84)
[  177.067633] Bluetooth: hci_qca.c:qca_recv() hci0: Frame reassembly failed (-84)

Now we enable a flag during bootup to stop executing proto receive
function and clear it back once the initialization is done.

Signed-off-by: Balakrishna Godavarthi <bgodavar@codeaurora.org>
Tested-by: Matthias Kaehlcke <mka@chromium.org>
---
v3:
 * moved ibs disable and qca_flush to different patch
 * updated commnents in code.
v2:
 * Updated commit text & comments.
v1: 
 * initial patch
---
 drivers/bluetooth/hci_qca.c | 16 ++++++++++++++++
 1 file changed, 16 insertions(+)

diff --git a/drivers/bluetooth/hci_qca.c b/drivers/bluetooth/hci_qca.c
index 253b5bae30c2..d595a82cb4b9 100644
--- a/drivers/bluetooth/hci_qca.c
+++ b/drivers/bluetooth/hci_qca.c
@@ -56,6 +56,7 @@
 
 /* Controller states */
 #define STATE_IN_BAND_SLEEP_ENABLED	1
+#define STATE_DISCARD_RX		2
 
 #define IBS_WAKE_RETRANS_TIMEOUT_MS	100
 #define IBS_TX_IDLE_TIMEOUT_MS		2000
@@ -511,6 +512,7 @@ static int qca_open(struct hci_uart *hu)
 		} else {
 			hu->init_speed = qcadev->init_speed;
 			hu->oper_speed = qcadev->oper_speed;
+			set_bit(STATE_DISCARD_RX, &qca->flags);
 			ret = qca_power_setup(hu, true);
 			if (ret) {
 				destroy_workqueue(qca->workqueue);
@@ -903,6 +905,13 @@ static int qca_recv(struct hci_uart *hu, const void *data, int count)
 	if (!test_bit(HCI_UART_REGISTERED, &hu->flags))
 		return -EUNATCH;
 
+	/* We discard Rx data received while device is in booting
+	 * stage, This is because of BT chip Tx line is in tristate.
+	 * Due to this we read some garbage data on UART Rx.
+	 */
+	if (test_bit(STATE_DISCARD_RX, &qca->flags))
+		return 0;
+
 	qca->rx_skb = h4_recv_buf(hu->hdev, qca->rx_skb, data, count,
 				  qca_recv_pkts, ARRAY_SIZE(qca_recv_pkts));
 	if (IS_ERR(qca->rx_skb)) {
@@ -1194,6 +1203,7 @@ static int qca_setup(struct hci_uart *hu)
 		if (ret)
 			return ret;
 
+		clear_bit(STATE_DISCARD_RX, &qca->flags);
 		ret = qca_read_soc_version(hdev, &soc_ver);
 		if (ret)
 			return ret;
@@ -1270,6 +1280,12 @@ static const struct qca_vreg_data qca_soc_data = {
 
 static void qca_power_shutdown(struct hci_uart *hu)
 {
+	struct qca_data *qca = hu->priv;
+
+	/* From this point we go into power off state. But serial port is
+	 * still open, discard all the garbage data received on the Rx line.
+	 */
+	set_bit(STATE_DISCARD_RX, &qca->flags);
 	host_set_baudrate(hu, 2400);
 	qca_send_power_pulse(hu, QCA_WCN3990_POWEROFF_PULSE);
 	qca_power_setup(hu, false);
-- 
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
a Linux Foundation Collaborative Project

[PATCH v3 4/4] Bluetooth: hci_qca: Disable IBS state machine and flush Tx buffer

[Balakrishna Godavarthi]

During hci down we observed IBS sleep commands are queued in the Tx
buffer and hci_uart_write_work is sending data to the chip which is
not required as the chip is powered off. This patch will disable IBS
and flush the Tx buffer before we turn off the chip.

Signed-off-by: Balakrishna Godavarthi <bgodavar@codeaurora.org>
---
v3: 
  * initial patch
---
 drivers/bluetooth/hci_qca.c | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/drivers/bluetooth/hci_qca.c b/drivers/bluetooth/hci_qca.c
index d595a82cb4b9..6f2a4bf7ba17 100644
--- a/drivers/bluetooth/hci_qca.c
+++ b/drivers/bluetooth/hci_qca.c
@@ -1286,6 +1286,8 @@ static void qca_power_shutdown(struct hci_uart *hu)
 	 * still open, discard all the garbage data received on the Rx line.
 	 */
 	set_bit(STATE_DISCARD_RX, &qca->flags);
+	clear_bit(STATE_IN_BAND_SLEEP_ENABLED, &qca->flags);
+	qca_flush(hu);
 	host_set_baudrate(hu, 2400);
 	qca_send_power_pulse(hu, QCA_WCN3990_POWEROFF_PULSE);
 	qca_power_setup(hu, false);
-- 
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum,
a Linux Foundation Collaborative Project

Re: [PATCH v3 1/4] Bluetooth: hci_qca: use wait_until_sent() for power pulses

[Johan Hovold]

On Fri, Nov 30, 2018 at 08:32:44PM +0530, Balakrishna Godavarthi wrote:
> wcn3990 requires a power pulse to turn ON/OFF along with
> regulators. Sometimes we are observing the power pulses are sent
> out with some time delay, due to queuing these commands. This is
> causing synchronization issues with chip, which intern delay the
> chip setup or may end up with communication issues.
> 
> Signed-off-by: Balakrishna Godavarthi <bgodavar@codeaurora.org>
> ---
> v3:
>   * no change.
> v2: 
>   * Updated function qca_send_power_pulse()
>   * addressed reviewer comments.

Please make sure to include reviewers on CC when resending, and as
someone else already mentioned, be a bit more specific about what
changes you actually made in response to the review feedback you
received.

> v1:
>  * initial patch
> ---
>  drivers/bluetooth/hci_qca.c | 37 +++++++++++++------------------------
>  1 file changed, 13 insertions(+), 24 deletions(-)
> 
> diff --git a/drivers/bluetooth/hci_qca.c b/drivers/bluetooth/hci_qca.c
> index f036c8f98ea3..f5dd323c1967 100644
> --- a/drivers/bluetooth/hci_qca.c
> +++ b/drivers/bluetooth/hci_qca.c
> @@ -1013,11 +1013,9 @@ static inline void host_set_baudrate(struct hci_uart *hu, unsigned int speed)
>  		hci_uart_set_baudrate(hu, speed);
>  }
>  
> -static int qca_send_power_pulse(struct hci_dev *hdev, u8 cmd)
> +static int qca_send_power_pulse(struct hci_uart *hu, u8 cmd)
>  {
> -	struct hci_uart *hu = hci_get_drvdata(hdev);
> -	struct qca_data *qca = hu->priv;
> -	struct sk_buff *skb;
> +	int ret;
>  
>  	/* These power pulses are single byte command which are sent
>  	 * at required baudrate to wcn3990. On wcn3990, we have an external
> @@ -1029,19 +1027,16 @@ static int qca_send_power_pulse(struct hci_dev *hdev, u8 cmd)
>  	 * save power. Disabling hardware flow control is mandatory while
>  	 * sending power pulses to SoC.
>  	 */
> -	bt_dev_dbg(hdev, "sending power pulse %02x to SoC", cmd);
> -
> -	skb = bt_skb_alloc(sizeof(cmd), GFP_KERNEL);
> -	if (!skb)
> -		return -ENOMEM;
> -
> +	bt_dev_dbg(hu->hdev, "sending power pulse %02x to SoC", cmd);
>  	hci_uart_set_flow_control(hu, true);
> +	ret = serdev_device_write(hu->serdev, &cmd, sizeof(cmd), 0);

You're still using 0 as a timeout here which is broken, as I already
told you.

From 4.21 this will result in an indefinite timeout, but currently
implies not to wait for a full write buffer to drain at all.

As I also mentioned, you need to to make sure to call
serdev_device_write_wakeup() in the write_wakup() path if you are going
to use serdev_device_write() at all.

Johan

Re: [PATCH v3 1/4] Bluetooth: hci_qca: use wait_until_sent() for power pulses

[Balakrishna Godavarthi]

Hi Johan,

On 2018-12-05 11:55, Johan Hovold wrote:
> On Fri, Nov 30, 2018 at 08:32:44PM +0530, Balakrishna Godavarthi wrote:
>> wcn3990 requires a power pulse to turn ON/OFF along with
>> regulators. Sometimes we are observing the power pulses are sent
>> out with some time delay, due to queuing these commands. This is
>> causing synchronization issues with chip, which intern delay the
>> chip setup or may end up with communication issues.
>> 
>> Signed-off-by: Balakrishna Godavarthi <bgodavar@codeaurora.org>
>> ---
>> v3:
>>   * no change.
>> v2:
>>   * Updated function qca_send_power_pulse()
>>   * addressed reviewer comments.
> 
> Please make sure to include reviewers on CC when resending, and as
> someone else already mentioned, be a bit more specific about what
> changes you actually made in response to the review feedback you
> received.
> 

[Bala]: sure will add and provide more info in version change history.

>> v1:
>>  * initial patch
>> ---
>>  drivers/bluetooth/hci_qca.c | 37 
>> +++++++++++++------------------------
>>  1 file changed, 13 insertions(+), 24 deletions(-)
>> 
>> diff --git a/drivers/bluetooth/hci_qca.c b/drivers/bluetooth/hci_qca.c
>> index f036c8f98ea3..f5dd323c1967 100644
>> --- a/drivers/bluetooth/hci_qca.c
>> +++ b/drivers/bluetooth/hci_qca.c
>> @@ -1013,11 +1013,9 @@ static inline void host_set_baudrate(struct 
>> hci_uart *hu, unsigned int speed)
>>  		hci_uart_set_baudrate(hu, speed);
>>  }
>> 
>> -static int qca_send_power_pulse(struct hci_dev *hdev, u8 cmd)
>> +static int qca_send_power_pulse(struct hci_uart *hu, u8 cmd)
>>  {
>> -	struct hci_uart *hu = hci_get_drvdata(hdev);
>> -	struct qca_data *qca = hu->priv;
>> -	struct sk_buff *skb;
>> +	int ret;
>> 
>>  	/* These power pulses are single byte command which are sent
>>  	 * at required baudrate to wcn3990. On wcn3990, we have an external
>> @@ -1029,19 +1027,16 @@ static int qca_send_power_pulse(struct hci_dev 
>> *hdev, u8 cmd)
>>  	 * save power. Disabling hardware flow control is mandatory while
>>  	 * sending power pulses to SoC.
>>  	 */
>> -	bt_dev_dbg(hdev, "sending power pulse %02x to SoC", cmd);
>> -
>> -	skb = bt_skb_alloc(sizeof(cmd), GFP_KERNEL);
>> -	if (!skb)
>> -		return -ENOMEM;
>> -
>> +	bt_dev_dbg(hu->hdev, "sending power pulse %02x to SoC", cmd);
>>  	hci_uart_set_flow_control(hu, true);
>> +	ret = serdev_device_write(hu->serdev, &cmd, sizeof(cmd), 0);
> 
> You're still using 0 as a timeout here which is broken, as I already
> told you.
> 

[Bala]: got the change now will update to timeout to non zero value.

> From 4.21 this will result in an indefinite timeout, but currently
> implies not to wait for a full write buffer to drain at all.
> 
> As I also mentioned, you need to to make sure to call
> serdev_device_write_wakeup() in the write_wakup() path if you are going
> to use serdev_device_write() at all.
> 

[Bala]: this where i am confused.
         calling serdev_device_write is calling an wakeup internally. 
below is the flow

         ttyport_write_buf:
               * calling serdev_device_write() will call write_buf() in 
this call we are enabling bit "TTY_DO_WRITE_WAKEUP" and calling write()
                 i.e. uart_write() where we call in start_tx. this will 
go to the vendor specific write where in isr we call uart_write_wakeup()
                 
https://elixir.bootlin.com/linux/latest/source/drivers/tty/serial/qcom_geni_serial.c#L756

         
uart_write_wakeup()->ttyport_write_wakeup()->serdev_controller_write_wakeup()->hci_uart_write_wakeup()->hci_uart_tx_wakeup()

         the above is flow when serdev_device_write() is called, it is 
indirectly calling serdev_write_wakeup().

         Why actual we need to call an serdev_write_wakeup() is this 
wakeup related to the UART port or for the BT chip.

> Johan

-- 
Regards
Balakrishna.

Re: [PATCH v3 1/4] Bluetooth: hci_qca: use wait_until_sent() for power pulses

[Balakrishna Godavarthi]

Hi Johan,

On 2018-12-06 16:10, Balakrishna Godavarthi wrote:
> Hi Johan,
> 
> On 2018-12-05 11:55, Johan Hovold wrote:
>> On Fri, Nov 30, 2018 at 08:32:44PM +0530, Balakrishna Godavarthi 
>> wrote:
>>> wcn3990 requires a power pulse to turn ON/OFF along with
>>> regulators. Sometimes we are observing the power pulses are sent
>>> out with some time delay, due to queuing these commands. This is
>>> causing synchronization issues with chip, which intern delay the
>>> chip setup or may end up with communication issues.
>>> 
>>> Signed-off-by: Balakrishna Godavarthi <bgodavar@codeaurora.org>
>>> ---
>>> v3:
>>>   * no change.
>>> v2:
>>>   * Updated function qca_send_power_pulse()
>>>   * addressed reviewer comments.
>> 
>> Please make sure to include reviewers on CC when resending, and as
>> someone else already mentioned, be a bit more specific about what
>> changes you actually made in response to the review feedback you
>> received.
>> 
> 
> [Bala]: sure will add and provide more info in version change history.
> 
>>> v1:
>>>  * initial patch
>>> ---
>>>  drivers/bluetooth/hci_qca.c | 37 
>>> +++++++++++++------------------------
>>>  1 file changed, 13 insertions(+), 24 deletions(-)
>>> 
>>> diff --git a/drivers/bluetooth/hci_qca.c 
>>> b/drivers/bluetooth/hci_qca.c
>>> index f036c8f98ea3..f5dd323c1967 100644
>>> --- a/drivers/bluetooth/hci_qca.c
>>> +++ b/drivers/bluetooth/hci_qca.c
>>> @@ -1013,11 +1013,9 @@ static inline void host_set_baudrate(struct 
>>> hci_uart *hu, unsigned int speed)
>>>  		hci_uart_set_baudrate(hu, speed);
>>>  }
>>> 
>>> -static int qca_send_power_pulse(struct hci_dev *hdev, u8 cmd)
>>> +static int qca_send_power_pulse(struct hci_uart *hu, u8 cmd)
>>>  {
>>> -	struct hci_uart *hu = hci_get_drvdata(hdev);
>>> -	struct qca_data *qca = hu->priv;
>>> -	struct sk_buff *skb;
>>> +	int ret;
>>> 
>>>  	/* These power pulses are single byte command which are sent
>>>  	 * at required baudrate to wcn3990. On wcn3990, we have an external
>>> @@ -1029,19 +1027,16 @@ static int qca_send_power_pulse(struct 
>>> hci_dev *hdev, u8 cmd)
>>>  	 * save power. Disabling hardware flow control is mandatory while
>>>  	 * sending power pulses to SoC.
>>>  	 */
>>> -	bt_dev_dbg(hdev, "sending power pulse %02x to SoC", cmd);
>>> -
>>> -	skb = bt_skb_alloc(sizeof(cmd), GFP_KERNEL);
>>> -	if (!skb)
>>> -		return -ENOMEM;
>>> -
>>> +	bt_dev_dbg(hu->hdev, "sending power pulse %02x to SoC", cmd);
>>>  	hci_uart_set_flow_control(hu, true);
>>> +	ret = serdev_device_write(hu->serdev, &cmd, sizeof(cmd), 0);
>> 
>> You're still using 0 as a timeout here which is broken, as I already
>> told you.
>> 
> 
> [Bala]: got the change now will update to timeout to non zero value.
> 
>> From 4.21 this will result in an indefinite timeout, but currently
>> implies not to wait for a full write buffer to drain at all.
>> 
>> As I also mentioned, you need to to make sure to call
>> serdev_device_write_wakeup() in the write_wakup() path if you are 
>> going
>> to use serdev_device_write() at all.
>> 
> 
> [Bala]: this where i am confused.
>         calling serdev_device_write is calling an wakeup internally.
> below is the flow
> 
>         ttyport_write_buf:
>               * calling serdev_device_write() will call write_buf() in
> this call we are enabling bit "TTY_DO_WRITE_WAKEUP" and calling
> write()
>                 i.e. uart_write() where we call in start_tx. this will
> go to the vendor specific write where in isr we call
> uart_write_wakeup()
> 
> https://elixir.bootlin.com/linux/latest/source/drivers/tty/serial/qcom_geni_serial.c#L756
> 
> 
> uart_write_wakeup()->ttyport_write_wakeup()->serdev_controller_write_wakeup()->hci_uart_write_wakeup()->hci_uart_tx_wakeup()
> 
>         the above is flow when serdev_device_write() is called, it is
> indirectly calling serdev_write_wakeup().
> 
>         Why actual we need to call an serdev_write_wakeup() is this
> wakeup related to the UART port or for the BT chip.
> 
>> Johan

Can you help me to understand, whether my understating is correct wrt 
serdev_wakeup().

-- 
Regards
Balakrishna.

Re: [PATCH v3 1/4] Bluetooth: hci_qca: use wait_until_sent() for power pulses

[Johan Hovold]

On Thu, Dec 06, 2018 at 04:10:07PM +0530, Balakrishna Godavarthi wrote:
> Hi Johan,
> 
> On 2018-12-05 11:55, Johan Hovold wrote:
> > On Fri, Nov 30, 2018 at 08:32:44PM +0530, Balakrishna Godavarthi wrote:

> >> +	ret = serdev_device_write(hu->serdev, &cmd, sizeof(cmd), 0);
> > 
> > You're still using 0 as a timeout here which is broken, as I already
> > told you.
> 
> [Bala]: got the change now will update to timeout to non zero value.
> 
> > From 4.21 this will result in an indefinite timeout, but currently
> > implies not to wait for a full write buffer to drain at all.
> > 
> > As I also mentioned, you need to to make sure to call
> > serdev_device_write_wakeup() in the write_wakup() path if you are going
> > to use serdev_device_write() at all.
> 
> [Bala]: this where i am confused.
>          calling serdev_device_write is calling an wakeup internally. 
> below is the flow
> 
>          ttyport_write_buf:
>                * calling serdev_device_write() will call write_buf() in 
> this call we are enabling bit "TTY_DO_WRITE_WAKEUP" and calling write()
>                  i.e. uart_write() where we call in start_tx. this will 
> go to the vendor specific write where in isr we call uart_write_wakeup()
>                  
> https://elixir.bootlin.com/linux/latest/source/drivers/tty/serial/qcom_geni_serial.c#L756
> 
>          
> uart_write_wakeup()->ttyport_write_wakeup()->serdev_controller_write_wakeup()->hci_uart_write_wakeup()->hci_uart_tx_wakeup()
> 
>          the above is flow when serdev_device_write() is called, it is 
> indirectly calling serdev_write_wakeup().

No, serdev_device_write_wakeup() is currently not called in this path,
which means you cannot use serdev_device_write().

>          Why actual we need to call an serdev_write_wakeup() is this 
> wakeup related to the UART port or for the BT chip.

serdev_device_write_wakeup() is where a writer blocked on a full write
buffer in serdev_device_write() is woken up. 

Johan

Re: [PATCH v3 1/4] Bluetooth: hci_qca: use wait_until_sent() for power pulses

[Balakrishna Godavarthi]

Hi Johan,

On 2018-12-12 22:12, Johan Hovold wrote:
> On Thu, Dec 06, 2018 at 04:10:07PM +0530, Balakrishna Godavarthi wrote:
>> Hi Johan,
>> 
>> On 2018-12-05 11:55, Johan Hovold wrote:
>> > On Fri, Nov 30, 2018 at 08:32:44PM +0530, Balakrishna Godavarthi wrote:
> 
>> >> +	ret = serdev_device_write(hu->serdev, &cmd, sizeof(cmd), 0);
>> >
>> > You're still using 0 as a timeout here which is broken, as I already
>> > told you.
>> 
>> [Bala]: got the change now will update to timeout to non zero value.
>> 
>> > From 4.21 this will result in an indefinite timeout, but currently
>> > implies not to wait for a full write buffer to drain at all.
>> >
>> > As I also mentioned, you need to to make sure to call
>> > serdev_device_write_wakeup() in the write_wakup() path if you are going
>> > to use serdev_device_write() at all.
>> 
>> [Bala]: this where i am confused.
>>          calling serdev_device_write is calling an wakeup internally.
>> below is the flow
>> 
>>          ttyport_write_buf:
>>                * calling serdev_device_write() will call write_buf() 
>> in
>> this call we are enabling bit "TTY_DO_WRITE_WAKEUP" and calling 
>> write()
>>                  i.e. uart_write() where we call in start_tx. this 
>> will
>> go to the vendor specific write where in isr we call 
>> uart_write_wakeup()
>> 
>> https://elixir.bootlin.com/linux/latest/source/drivers/tty/serial/qcom_geni_serial.c#L756
>> 
>> 
>> uart_write_wakeup()->ttyport_write_wakeup()->serdev_controller_write_wakeup()->hci_uart_write_wakeup()->hci_uart_tx_wakeup()
>> 
>>          the above is flow when serdev_device_write() is called, it is
>> indirectly calling serdev_write_wakeup().
> 
> No, serdev_device_write_wakeup() is currently not called in this path,
> which means you cannot use serdev_device_write().
> 
>>          Why actual we need to call an serdev_write_wakeup() is this
>> wakeup related to the UART port or for the BT chip.
> 
> serdev_device_write_wakeup() is where a writer blocked on a full write
> buffer in serdev_device_write() is woken up.
> 
> Johan

Is it preferred to use and serdev_device_write_buf() followed by 
serdev_device_wait_until_sent()
or do we required an write_wakeup() called before writing into 
serdev_device_write_buf()

-- 
Regards
Balakrishna.

Re: [PATCH v3 1/4] Bluetooth: hci_qca: use wait_until_sent() for power pulses

[Johan Hovold]

On Fri, Dec 14, 2018 at 06:02:40PM +0530, Balakrishna Godavarthi wrote:
> Hi Johan,
> 
> On 2018-12-12 22:12, Johan Hovold wrote:
> > On Thu, Dec 06, 2018 at 04:10:07PM +0530, Balakrishna Godavarthi wrote:

> >> uart_write_wakeup()->ttyport_write_wakeup()->serdev_controller_write_wakeup()->hci_uart_write_wakeup()->hci_uart_tx_wakeup()
> >> 
> >>          the above is flow when serdev_device_write() is called, it is
> >> indirectly calling serdev_write_wakeup().
> > 
> > No, serdev_device_write_wakeup() is currently not called in this path,
> > which means you cannot use serdev_device_write().
> > 
> >>          Why actual we need to call an serdev_write_wakeup() is this
> >> wakeup related to the UART port or for the BT chip.
> > 
> > serdev_device_write_wakeup() is where a writer blocked on a full write
> > buffer in serdev_device_write() is woken up.
> > 
> > Johan
> 
> Is it preferred to use and serdev_device_write_buf() followed by 
> serdev_device_wait_until_sent()

That's up to the driver author and depends on what you want to achieve.

If you want to do something synchronously, you always have to call
serdev_device_wait_until_sent() after, as both serdev_device_write_buf()
and serdev_device_write() only buffer the data.

If you use serdev_device_write_buf() you also have to make sure that you
can handle incomplete buffering yourself (e.g. when the underlying tty
driver buffer is getting full).

serdev_device_write() was added as a convenience helper for this, but
depends on serdev_device_write_wakeup() being called in the write_wakeup
path to make forward progress.

> or do we required an write_wakeup() called before writing into 
> serdev_device_write_buf()

No, that doesn't make any sense. serdev_device_write_wakeup() needs to
be called in the write-wakeup path, but only if you use
serdev_device_write().

Please see the documentation of these function that I added to
linux-next (and that I linked to earlier).

Johan

Re: [PATCH v3 1/4] Bluetooth: hci_qca: use wait_until_sent() for power pulses

[Balakrishna Godavarthi]

Hi Johan,

On 2018-12-14 18:46, Johan Hovold wrote:
> On Fri, Dec 14, 2018 at 06:02:40PM +0530, Balakrishna Godavarthi wrote:
>> Hi Johan,
>> 
>> On 2018-12-12 22:12, Johan Hovold wrote:
>> > On Thu, Dec 06, 2018 at 04:10:07PM +0530, Balakrishna Godavarthi wrote:
> 
>> >> uart_write_wakeup()->ttyport_write_wakeup()->serdev_controller_write_wakeup()->hci_uart_write_wakeup()->hci_uart_tx_wakeup()
>> >>
>> >>          the above is flow when serdev_device_write() is called, it is
>> >> indirectly calling serdev_write_wakeup().
>> >
>> > No, serdev_device_write_wakeup() is currently not called in this path,
>> > which means you cannot use serdev_device_write().
>> >
>> >>          Why actual we need to call an serdev_write_wakeup() is this
>> >> wakeup related to the UART port or for the BT chip.
>> >
>> > serdev_device_write_wakeup() is where a writer blocked on a full write
>> > buffer in serdev_device_write() is woken up.
>> >
>> > Johan
>> 
>> Is it preferred to use and serdev_device_write_buf() followed by
>> serdev_device_wait_until_sent()
> 
> That's up to the driver author and depends on what you want to achieve.
> 

[Bala]: During the BT on & of, we need to send the commands to the 
controller
at a specific baudrate. as the data is getting queued in the buffer we 
see some times
the data is transmitting on the newer baudrate instead of the actual 
baudrate.
we want to achieve an synchronous data transfer for some commands to 
over come byte corruptions.

> If you want to do something synchronously, you always have to call
> serdev_device_wait_until_sent() after, as both 
> serdev_device_write_buf()
> and serdev_device_write() only buffer the data.
> 
> If you use serdev_device_write_buf() you also have to make sure that 
> you
> can handle incomplete buffering yourself (e.g. when the underlying tty
> driver buffer is getting full).
> 
> serdev_device_write() was added as a convenience helper for this, but
> depends on serdev_device_write_wakeup() being called in the 
> write_wakeup
> path to make forward progress.
> 

[Bala]: Thanks for the info, now i understood the usage of the both 
write() and write_buf().

>> or do we required an write_wakeup() called before writing into
>> serdev_device_write_buf()
> 
> No, that doesn't make any sense. serdev_device_write_wakeup() needs to
> be called in the write-wakeup path, but only if you use
> serdev_device_write().
> 
> Please see the documentation of these function that I added to
> linux-next (and that I linked to earlier).
> 

> Johan

Thanks Johan for the detailed info.

-- 
Regards
Balakrishna.