[PATCH v2] Bluetooth: btmtksdio: Add in-band wakeup support

[PATCH v2] Bluetooth: btmtksdio: Add in-band wakeup support

[sean.wang]

From: Sean Wang <sean.wang@mediatek.com>

Commit ce64b3e94919 ("Bluetooth: mt7921s: Support wake on bluetooth")
adds the wake on bluethooth via a dedicated GPIO.

Extend the wake-on-bluetooth to use the SDIO DAT1 pin (in-band wakeup),
when supported by the SDIO host driver.

Co-developed-by: Yake Yang <yake.yang@mediatek.com>
Signed-off-by: Yake Yang <yake.yang@mediatek.com>
Signed-off-by: Sean Wang <sean.wang@mediatek.com>
---
v2: enhance the patch description and comments
---
 drivers/bluetooth/btmtksdio.c | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/drivers/bluetooth/btmtksdio.c b/drivers/bluetooth/btmtksdio.c
index d6700efcfe8c..9d79c9107b3a 100644
--- a/drivers/bluetooth/btmtksdio.c
+++ b/drivers/bluetooth/btmtksdio.c
@@ -118,6 +118,7 @@ MODULE_DEVICE_TABLE(sdio, btmtksdio_table);
 #define BTMTKSDIO_FUNC_ENABLED		3
 #define BTMTKSDIO_PATCH_ENABLED		4
 #define BTMTKSDIO_HW_RESET_ACTIVE	5
+#define BTMTKSDIO_INBAND_WAKEUP		6
 
 struct mtkbtsdio_hdr {
 	__le16	len;
@@ -1294,6 +1295,9 @@ static bool btmtksdio_sdio_wakeup(struct hci_dev *hdev)
 		.wakeup_delay = cpu_to_le16(0x20),
 	};
 
+	if (test_bit(BTMTKSDIO_INBAND_WAKEUP, &bdev->tx_state))
+		return may_wakeup;
+
 	if (may_wakeup && bdev->data->chipid == 0x7921) {
 		struct sk_buff *skb;
 
@@ -1384,6 +1388,10 @@ static int btmtksdio_probe(struct sdio_func *func,
 	 */
 	pm_runtime_put_noidle(bdev->dev);
 
+	/* Mark if MMC host supports wake on bluetooth by SDIO */
+	if (device_can_wakeup(func->card->host->parent))
+		set_bit(BTMTKSDIO_INBAND_WAKEUP, &bdev->tx_state);
+
 	err = device_init_wakeup(bdev->dev, true);
 	if (err)
 		bt_dev_err(hdev, "failed to initialize device wakeup");
-- 
2.25.1

RE: [v2] Bluetooth: btmtksdio: Add in-band wakeup support

[bluez.test.bot]

[-- Attachment #1: Type: text/plain, Size: 1096 bytes --]

This is automated email and please do not reply to this email!

Dear submitter,

Thank you for submitting the patches to the linux bluetooth mailing list.
This is a CI test results with your patch series:
PW Link:https://patchwork.kernel.org/project/bluetooth/list/?series=652944

---Test result---

Test Summary:
CheckPatch                    PASS      0.90 seconds
GitLint                       PASS      0.56 seconds
SubjectPrefix                 PASS      0.42 seconds
BuildKernel                   PASS      30.71 seconds
BuildKernel32                 PASS      27.17 seconds
Incremental Build with patchesPASS      37.53 seconds
TestRunner: Setup             PASS      464.66 seconds
TestRunner: l2cap-tester      PASS      16.55 seconds
TestRunner: bnep-tester       PASS      5.60 seconds
TestRunner: mgmt-tester       PASS      97.86 seconds
TestRunner: rfcomm-tester     PASS      8.98 seconds
TestRunner: sco-tester        PASS      8.81 seconds
TestRunner: smp-tester        PASS      8.87 seconds
TestRunner: userchan-tester   PASS      5.78 seconds



---
Regards,
Linux Bluetooth

Re: [PATCH v2] Bluetooth: btmtksdio: Add in-band wakeup support

[AngeloGioacchino Del Regno]

Il 22/06/22 23:56, sean.wang@mediatek.com ha scritto:
> From: Sean Wang <sean.wang@mediatek.com>
> 
> Commit ce64b3e94919 ("Bluetooth: mt7921s: Support wake on bluetooth")
> adds the wake on bluethooth via a dedicated GPIO.
> 
> Extend the wake-on-bluetooth to use the SDIO DAT1 pin (in-band wakeup),
> when supported by the SDIO host driver.
> 
> Co-developed-by: Yake Yang <yake.yang@mediatek.com>
> Signed-off-by: Yake Yang <yake.yang@mediatek.com>
> Signed-off-by: Sean Wang <sean.wang@mediatek.com>
> ---
> v2: enhance the patch description and comments
> ---
>   drivers/bluetooth/btmtksdio.c | 8 ++++++++
>   1 file changed, 8 insertions(+)
> 
> diff --git a/drivers/bluetooth/btmtksdio.c b/drivers/bluetooth/btmtksdio.c
> index d6700efcfe8c..9d79c9107b3a 100644
> --- a/drivers/bluetooth/btmtksdio.c
> +++ b/drivers/bluetooth/btmtksdio.c
> @@ -118,6 +118,7 @@ MODULE_DEVICE_TABLE(sdio, btmtksdio_table);
>   #define BTMTKSDIO_FUNC_ENABLED		3
>   #define BTMTKSDIO_PATCH_ENABLED		4
>   #define BTMTKSDIO_HW_RESET_ACTIVE	5
> +#define BTMTKSDIO_INBAND_WAKEUP		6
>   
>   struct mtkbtsdio_hdr {
>   	__le16	len;
> @@ -1294,6 +1295,9 @@ static bool btmtksdio_sdio_wakeup(struct hci_dev *hdev)
>   		.wakeup_delay = cpu_to_le16(0x20),
>   	};
>   
> +	if (test_bit(BTMTKSDIO_INBAND_WAKEUP, &bdev->tx_state))
> +		return may_wakeup;
> +

Uhm... this flag is either *always* set, or *always not set*... and we decide that
during probe time... and we use it in just one function as well.

At this point, I would just avoid the addition of the BTMTKSDIO_INBAND_WAKEUP flag
and add a new function for handling the .wakeup() callback - something like:

static bool btmtksdio_sdio_inband_wakeup(struct hci_dev *hdev)
{
	struct btmtksdio_dev *bdev = hci_get_drvdata(hdev);

	return device_may_wakeup(bdev->dev);
}

static int btmtksdio_probe(...)
{
	.... code ....

	/*
	 * If SDIO controller supports wake on Bluetooth, sending a wakeon
	 * command is not necessary.
	 */
	if (device_can_wakeup(func->card->host->parent))
		hdev->wakeup = btmtksdio_sdio_inband_wakeup;
	else
		hdev->wakeup = btmtksdio_sdio_wakeup;

	.... etc ....
}

Regards,
Angelo