Bug#1021391: Possible solution with kernel 6.0.11

[Salvatore Bonaccorso <carnil@debian.org>]

Hi Bernhard,

On Mon, Dec 05, 2022 at 12:21:48PM +0000, Bernhard wrote:
> Hello Salvatore
> 
> It is possible, that kernel 6.0.11 solves the Bug #1021391 because of
> 
> >>>
> - [arm64,armhf] bus: sunxi-rsb: Remove the shutdown callback
>     - [arm64,armhf] bus: sunxi-rsb: Support atomic transfers
> <<<
> 
> If i have a look at the ARM kernel mailinglist, there are bugfixes for this I2C:
> 
> >>>
> Shutting down the RSB controller prevents communicating with a PMIC
> inside pm_power_off(), so it breaks system poweroff on some boards.
> <<<
> 
> >>>
> This series fixes a couple of issues that occur when powering off a
> board using a PMIC attached to the RSB bus.
> 
> These issues only affected 32-bit platforms, because 64-bit platforms
> use PSCI for pm_power_off(), and the PSCI firmware reinitializes the
> RSB controller.
> <<<
> 
> >>>
> When communicating with a PMIC during system poweroff (pm_power_off()),
> IRQs are disabled and we are in a RCU read-side critical section, so we
> cannot use wait_for_completion_io_timeout(). Instead, poll the status
> register for transfer completion.
> <<<
> 
> Is it planned, to release this kernel in the next days?
> I'm really very interested in testing.

I was pondering it actually, and the changes are waiting in the sid
branch, but wanted to include the changes of upcoming 6.0.12 as well
in the next update.

Regards,
Salvatore