[Letux-kernel] gta04a5 + bluetooth + letux 4.19.0.3 = mysterious things ahead

H. Nikolaus Schaller hns at goldelico.com
Mon Nov 5 11:59:04 CET 2018


> Am 05.11.2018 um 07:38 schrieb Andreas Kemnade <andreas at kemnade.info>:
> 
> Hi,
> 
> On Sun, 4 Nov 2018 19:40:48 +0100
> "H. Nikolaus Schaller" <hns at goldelico.com> wrote:
> 
>>> Am 04.11.2018 um 15:18 schrieb Andreas Kemnade <andreas at kemnade.info>:
>>> 
>>> Hi,
>>> 
>>> I am wondering about bluetooth behavior on a5:
>>> 
>>> - Suspend current decrease by 6mA (I have seen 12mA suspend without
>>> modem): my doing the following steps
>>>  1. booting without bluetoothd, module autoloading, no uart idling
>>>  2. hciconfig hci0 up (it is still down)
>>>  3. rmmod hci_uart
>>>  4. suspend
>>> 
>>> To clarify 6mA lower than without modules.
>>> 
> I am wondering what happens here. This sequence changes something.
> But what?
> -External pin state
> - CTS/RTS / PCM pins?

probably yes. AFAIR pins are input by default and pinmux is only set
up if driver is loaded.

> -Internal state of the module (but it should be..off anyways)
> -Internal state of the soc
> -energy harvesting functionality of the bt/wifi combo module ;-)

Yes, that is it. If you find the right sequence you can get to
negative suspend current :)

> 
> But then the interesting question is why I see this only on runtime.
> 
>>> - wlan screws up when hci_uart is removed and a certain set of things
>>> is idled. I have not fully evaluated this.
>>> Required:
>>>   allowing bluetooth uart to suspend
>>> echo 3000
>>>> /sys/bus/platform/drivers/omap_uart/4806a000.serial/power/autosuspend_delay_ms
>>> Definitively not required:
>>>   allowing gps uart to suspend, gps can be active. Their modules
>>>   clocks depend on the same clocks.
>>> 
>>> sdcard access still works.
>>> To recover wifi you have to unbind and rebind omap_hsmmc via sysfs.
>> 
>> Hm. Maybe they share some clock?
>> 
> Or maybe the module gets confused because of some RTS/CTS/whatever
> state. UART1/UART2 clocks have common parents,
> MMC1/2 should also have. So I would expect trouble when both
> UART1/2 are powered down. But not only one.
> 
> Regards,
> Andreas
> _______________________________________________
> http://projects.goldelico.com/p/gta04-kernel/
> Letux-kernel mailing list
> Letux-kernel at openphoenux.org
> http://lists.goldelico.com/mailman/listinfo.cgi/letux-kernel

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 801 bytes
Desc: Message signed with OpenPGP using GPGMail
URL: <http://lists.goldelico.com/pipermail/letux-kernel/attachments/20181105/398c2bd8/attachment.asc>


More information about the Letux-kernel mailing list