Finding the correct ttySx

[lauro4]

Hi, I'm working on a CherryTrail with rlt8723bs on SDIO.

I read I should run dmesg | grep -i "ttyS.* at MMIO" to locate the correct ttySx. Well, initially I got no matches, but reading other issues I learned I need to enable CONFIG_SERIAL_8250_DW on my kernel.

With that module loaded I see some ttySx on dmesg.

[    9.404177] 8086228A:00: ttyS1 at MMIO 0x9181b000 (irq = 39, base_baud = 2764800) is a 16550A
[    9.406147] 8086228A:01: ttyS2 at MMIO 0x91819000 (irq = 40, base_baud = 2764800) is a 16550A

I figured 8086228A:00: ttyS1 is the correct one looking at ls /sys/bus/platform/devices/8086228A\:00/ and seeing it lists OBDA8723:00/ under it.

Running start_bt.sh the only new message I see on dmesg is:
[ 154.996398] ttyS1 - failed to request DMA

After that I learned I should enable CONFIG_RFKILL_GPIO on my kernel and add OBDA8723 to the rfkill-gpio acpi list.

+++ net/rfkill/rfkill-gpio.c
@@ -163,7 +163,6 @@ static int rfkill_gpio_remove(struct platform_device *pdev)
 static const struct acpi_device_id rfkill_acpi_match[] = {
        { "BCM4752", RFKILL_TYPE_GPS },
        { "LNV4752", RFKILL_TYPE_GPS },
+       { "OBDA8723", RFKILL_TYPE_BLUETOOTH },
        { },
 };
 MODULE_DEVICE_TABLE(acpi, rfkill_acpi_match);

With those changes I got the driver working fine. Sorry if this issue looks more like a forum post, I just wanted to report bluetooth is working well so far in our tests.
That's it, thank you.