Re: [PATCH v8 0/5] regulator: Add X-Powers AXP318W PMIC support

[Chen-Yu Tsai]

Hi Lee,

On Thu, Sep 10, 2026 at 6:44 PM Jerome Brunet <jbrunet@xxxxxxxxxxxx> wrote:
>
> This patch series adds support for the X-Powers AXP318W PMIC, which is
> used recently on new boards with Allwinner SoCs (the A733).
>
> The PMIC features 9 DCDC buck converters and 28 LDOs, plus the usual ADC,
> interrupts, and power key components.
> A datasheet can be found linked in this Wiki table:
> https://linux-sunxi.org/AXP_PMICs
>
> There was an unclosed topic regarding the handling of LDOs output
> voltage being limited to input voltage being provided by the supply.
> Chen-Yu noted the min_dropout_uV would be a good fit for that but
> the actual minimal dropout value was unknown. I've arbitrarily picked
> 1uV because it is the minimum the framework supports. Realistically a low
> dropout regulator is bound to have some dropout. This choice does the
> trick for now and is certain be less restrictive than what it is in
> reality.
>
> If anyone has problems because the dropout is actually higher
> * it is arguably an HW design error
> * it is a good time to let us know what the actual dropout is
>
> Changes in v8:
> - Fix interrupt optional in the binding (mis-understanding on my end)
>   Resent quickly to avoid wasting time on v7 over this
> - Link to v7: https://patch.msgid.link/20260910-axp318-regulator-v7-0-de720ceb9bac@xxxxxxxxxxxx
>
> Changes in v7:
> - Rebased on v7.3-rc1.
> - Applied Chen-Yu change requests.
>   - More complete register definition all in the MFD change.
>   - Add writeable region holes.
>   - Fix invertion between short and long press IRQ.
>   - Drop confusing comment about DVM.
>   - Use MFD_CELL_NAME() for the regulator device.
> - Link to v6: https://patch.msgid.link/20260812-axp318-regulator-v6-0-d6b84b55c000@xxxxxxxxxxxx
>
> Changes in v6:
> - Apply wording suggestions from Chen-Yu
> - Link to v5: https://patch.msgid.link/20260720-axp318-regulator-v5-0-a1afa7ee24f9@xxxxxxxxxxxx
>
> Changes in v5:
> - Fix shutdown register offset
> - Re-order patches to put fixes first
> - Tighten DT regexp around swout
> - Other points raised by Sashiko on v4 are unrelated to the addition
>   of the AXP318 support and pre-exist the series. They are not
>   addressed ATM.
> - Link to v4: https://patch.msgid.link/20260717-axp318-regulator-v4-0-50aff97084dd@xxxxxxxxxxxx
>
> Changes in v4:
> - Capitalize patch 3 title
> - regulator: add support for workmode and poly-phase.
> - i2c: fix probe through i2c sysfs
> - Link to v3: https://patch.msgid.link/20260717-axp318-regulator-v3-0-779dc1b97dd4@xxxxxxxxxxxx
>
> Changes in v3:
> - Correct mess up v2 errors regarding ops and ids.
> - Correctly propagate errors when checking axp318 threshold.
> - Fix DT binding regex regarding dcdc max ids
> - Add missing RTCLDO.
> - Remove axp318 PEK MFD cell:
>   As pointed out by sashiko, the driver name did not match anything in the
>   pek driver. While the device does support some handling of the power
>   button, it does not seem to match the existing driver. Register 0x36
>   relates to DLDO3, not the power key, as expected by the misc driver.
>   Leave this out for now.
> - Constify axp313 and axp717 mfd_cells array for consistency.
> - Link to v2: https://patch.msgid.link/20260710-axp318-regulator-v2-0-ee5f1c56b49f@xxxxxxxxxxxx
>
> Changes in v2:
> - Dropped _NUM_VOLTAGES macros
> - Use min_dropout_uV to limit LDO voltage to supplies
> - Add bypass mode for EDLO4/5
> - Prevent crossing 1.54V threshold for DCDC6/7/8/9
> - Link to v1: https://lore.kernel.org/r/20251021112013.2710903-1-andre.przywara@xxxxxxx
>
> ---
> Andre Przywara (3):
>       dt-bindings: mfd: x-powers,axp152: Document AXP318W
>       mfd: axp20x: Add support for AXP318W PMIC
>       regulator: axp20x: add support for the AXP318W
>
> Jerome Brunet (2):
>       mfd: axp20x: Constify axp313 and axp717 cells
>       mfd: axp20x-i2c: Fix i2c instantiation

I think this series is ready to be merged. Mark gave his reviewed-by for
the regulator patch, which IIRC he said acks the patch to go through
another tree.

So please merge the whole series through the mfd tree.


Thanks
ChenYu