Re: [PATCH v1] Bluetooth: hci_qca: Add qca_power_on() API to support both wcn399x and Rome power up

From: Rocky Liao
Date: Tue Jan 07 2020 - 22:02:51 EST

Hi Matt,

å 2020-01-08 01:31ïMatthias Kaehlcke åéï
Hi Rocky,

On Tue, Jan 07, 2020 at 01:26:01PM +0800, Rocky Liao wrote:
This patch adds a unified API qca_power_on() to support both wcn399x and
Rome power on. For wcn399x it calls the qca_wcn3990_init() to init the
regulators, and for Rome it pulls up the bt_en GPIO to power up the btsoc.

Signed-off-by: Rocky Liao <rjliao@xxxxxxxxxxxxxx>
drivers/bluetooth/hci_qca.c | 21 +++++++++++++++++++++
1 file changed, 21 insertions(+)

diff --git a/drivers/bluetooth/hci_qca.c b/drivers/bluetooth/hci_qca.c
index 9392cc7f9908..f6555bd1adbc 100644
--- a/drivers/bluetooth/hci_qca.c
+++ b/drivers/bluetooth/hci_qca.c
@@ -1532,6 +1532,27 @@ static int qca_wcn3990_init(struct hci_uart *hu)
return 0;

+static int qca_power_on(struct hci_dev *hdev)
+ struct hci_uart *hu = hci_get_drvdata(hdev);
+ enum qca_btsoc_type soc_type = qca_soc_type(hu);
+ struct qca_serdev *qcadev;
+ int ret = 0;

another option would be to return directly from the if/else branches,
but either way is fine.

+ if (qca_is_wcn399x(soc_type)) {
+ ret = qca_wcn3990_init(hu);
+ } else {
+ if (hu->serdev) {
+ qcadev = serdev_device_get_drvdata(hu->serdev);
+ gpiod_set_value_cansleep(qcadev->bt_en, 1);
+ /* Controller needs time to bootup. */
+ msleep(150);
+ }
+ }
+ return ret;

I expected qca_power_on() would be called from qca_open(), but as is
this would only work for ROME, and not WCN399x, which only enables
the regulators in qca_open(), qca_wcn3990_init() is called from
qca_setup(). Is there a particular reason for this assymmetry between
the ROME and WCN399x initialization (i.e. one is fully powered up after
open(), the other not)?

I prefer to move the power on call from qca_open() to qca_setup() for Rome,
I don't see any reason for this difference with wcn399x.I will send patch for
this if you have no concern.