RE: [PATCH v2 RFC 1/9] wifi: rtw88: Clear RTW_FLAG_POWERON early in rtw_mac_power_switch()

[Ping-Ke Shih <pkshih@realtek.com>]

> -----Original Message-----
> From: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
> Sent: Saturday, March 11, 2023 4:29 AM
> To: linux-wireless@vger.kernel.org
> Cc: Yan-Hsuan Chuang <tony0620emma@gmail.com>; Kalle Valo <kvalo@kernel.org>; Ulf Hansson
> <ulf.hansson@linaro.org>; linux-kernel@vger.kernel.org; netdev@vger.kernel.org;
> linux-mmc@vger.kernel.org; Chris Morgan <macroalpha82@gmail.com>; Nitin Gupta <nitin.gupta981@gmail.com>;
> Neo Jou <neojou@gmail.com>; Ping-Ke Shih <pkshih@realtek.com>; Jernej Skrabec <jernej.skrabec@gmail.com>;
> Martin Blumenstingl <martin.blumenstingl@googlemail.com>
> Subject: [PATCH v2 RFC 1/9] wifi: rtw88: Clear RTW_FLAG_POWERON early in rtw_mac_power_switch()
> 
> The SDIO HCI implementation needs to know when the MAC is powered on.
> This is needed because 32-bit register access has to be split into 4x
> 8-bit register access when the MAC is not fully powered on or while
> powering off. When the MAC is powered on 32-bit register access can be
> used to reduce the number of transfers but splitting into 4x 8-bit
> register access still works in that case.
> 
> During the power on sequence is how RTW_FLAG_POWERON is only set when
> the power on sequence has completed successfully. During power off
> however RTW_FLAG_POWERON is set. This means that the upcoming SDIO HCI
> implementation does not know that it has to use 4x 8-bit register
> accessors. Clear the RTW_FLAG_POWERON flag early when powering off the
> MAC so the whole power off sequence is processed with RTW_FLAG_POWERON
> unset. This will make it possible to use the RTW_FLAG_POWERON flag in
> the upcoming SDIO HCI implementation.
> 
> Signed-off-by: Martin Blumenstingl <martin.blumenstingl@googlemail.com>
> ---
> Changes since v1:
> - This replaces a previous patch called "rtw88: hci: Add an optional
>   power_switch() callback to rtw_hci_ops" which added a new callback
>   to the HCI ops.
> 
> 
>  drivers/net/wireless/realtek/rtw88/mac.c | 5 +++--
>  1 file changed, 3 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/net/wireless/realtek/rtw88/mac.c b/drivers/net/wireless/realtek/rtw88/mac.c
> index f3a566cf979b..cfdfc8a2c836 100644
> --- a/drivers/net/wireless/realtek/rtw88/mac.c
> +++ b/drivers/net/wireless/realtek/rtw88/mac.c
> @@ -273,6 +273,9 @@ static int rtw_mac_power_switch(struct rtw_dev *rtwdev, bool pwr_on)
>         if (pwr_on == cur_pwr)
>                 return -EALREADY;
> 
> +       if (!pwr_on)
> +               clear_bit(RTW_FLAG_POWERON, rtwdev->flags);
> +
>         pwr_seq = pwr_on ? chip->pwr_on_seq : chip->pwr_off_seq;
>         ret = rtw_pwr_seq_parser(rtwdev, pwr_seq);
>         if (ret)

This patch changes the behavior if rtw_pwr_seq_parser() returns error while
doing power-off, but I dig and think further about this case hardware stays in
abnormal state. I think it would be fine to see this state as POWER_OFF.
Do you agree this as well?


> @@ -280,8 +283,6 @@ static int rtw_mac_power_switch(struct rtw_dev *rtwdev, bool pwr_on)
> 
>         if (pwr_on)
>                 set_bit(RTW_FLAG_POWERON, rtwdev->flags);
> -       else
> -               clear_bit(RTW_FLAG_POWERON, rtwdev->flags);
> 
>         return 0;
>  }
> --
> 2.39.2