Re: [PATCH RFC 2/2] interconnect: qcom: sdm845: Own the ALC BCM to drop its boot vote

[Ard Biesheuvel]

On Tue, 13 Oct 2026, at 22:26, yoxcu wrote:
> From: Michael Auer <mick@xxxxxxxx>
>
> On the OnePlus 6 the bootloader leaves a vote on the ALC BCM
> (rpmh_read() on the APPS DRV returns 0x2000000f). No driver owns this
> BCM, so the vote is never removed and the SoC never reaches AOSS sleep
> or CX collapse (qcom_stats aosd and cxsd stay at 0).
>
> Add a node without consumers for the ALC BCM on mem_noc. The provider
> then owns the BCM and drops the vote at sync_state.
>
> With this change ALC reads 0 after boot and suspend reaches CX
> collapse. The average current in an idle push standby test (Wi-Fi on)
> was 35.6 mA, against 43.7 mA with the ALC vote kept in the sleep set.
>
> Assisted-by: LLM
> Signed-off-by: Michael Auer <mick@xxxxxxxx>
> ---
>  drivers/interconnect/qcom/sdm845.c | 17 +++++++++++++++++
>  1 file changed, 17 insertions(+)
>

Tested-by: Ard Biesheuvel <ardb@xxxxxxxxxx> # Lenovo C630 Yoga

With this change (and a few others), my Yoga now suspends properly,
to the point where it shows the pulsating LED in the power button.
Battery seems much happier too.