Re: [PATCH] leds: qcom-lpg: allocate channels with main struct

[Rosen Penev]

On Thu, Apr 9, 2026 at 8:12 AM Lee Jones <lee@xxxxxxxxxx> wrote:
>
> On Mon, 30 Mar 2026, Rosen Penev wrote:
>
> > Use a flexible array member to combine kzalloc and kcalloc. This
> > required moving the struct lpg_channel definition up as flexible array
> > members require a full definition.
> >
> > Add __counted_by for extra runtime analysis.
> >
> > Signed-off-by: Rosen Penev <rosenp@xxxxxxxxx>
> > ---
> >  drivers/leds/rgb/leds-qcom-lpg.c | 117 +++++++++++++++----------------
> >  1 file changed, 56 insertions(+), 61 deletions(-)
> >
> > diff --git a/drivers/leds/rgb/leds-qcom-lpg.c b/drivers/leds/rgb/leds-qcom-lpg.c
> > index f6061c47f863..83cedf4a0cbf 100644
> > --- a/drivers/leds/rgb/leds-qcom-lpg.c
> > +++ b/drivers/leds/rgb/leds-qcom-lpg.c
> > @@ -80,58 +80,8 @@
> >  #define SDAM_PAUSE_HI_MULTIPLIER_OFFSET      0x8
> >  #define SDAM_PAUSE_LO_MULTIPLIER_OFFSET      0x9
> >
> > -struct lpg_channel;
> >  struct lpg_data;
> >
> > -/**
> > - * struct lpg - LPG device context
> > - * @dev:     pointer to LPG device
> > - * @map:     regmap for register access
> > - * @lock:    used to synchronize LED and pwm callback requests
> > - * @pwm:     PWM-chip object, if operating in PWM mode
> > - * @data:    reference to version specific data
> > - * @lut_base:        base address of the LUT block (optional)
> > - * @lut_size:        number of entries in the LUT block
> > - * @lut_bitmap:      allocation bitmap for LUT entries
> > - * @pbs_dev: PBS device
> > - * @lpg_chan_sdam:   LPG SDAM peripheral device
> > - * @lut_sdam:        LUT SDAM peripheral device
> > - * @pbs_en_bitmap:   bitmap for tracking PBS triggers
> > - * @triled_base: base address of the TRILED block (optional)
> > - * @triled_src:      power-source for the TRILED
> > - * @triled_has_atc_ctl:      true if there is TRI_LED_ATC_CTL register
> > - * @triled_has_src_sel:      true if there is TRI_LED_SRC_SEL register
> > - * @channels:        list of PWM channels
> > - * @num_channels: number of @channels
> > - */
> > -struct lpg {
> > -     struct device *dev;
> > -     struct regmap *map;
> > -
> > -     struct mutex lock;
> > -
> > -     struct pwm_chip *pwm;
> > -
> > -     const struct lpg_data *data;
> > -
> > -     u32 lut_base;
> > -     u32 lut_size;
> > -     unsigned long *lut_bitmap;
> > -
> > -     struct pbs_dev *pbs_dev;
> > -     struct nvmem_device *lpg_chan_sdam;
> > -     struct nvmem_device *lut_sdam;
> > -     unsigned long pbs_en_bitmap;
> > -
> > -     u32 triled_base;
> > -     u32 triled_src;
> > -     bool triled_has_atc_ctl;
> > -     bool triled_has_src_sel;
> > -
> > -     struct lpg_channel *channels;
> > -     unsigned int num_channels;
> > -};
> > -
> >  /**
> >   * struct lpg_channel - per channel data
> >   * @lpg:     reference to parent lpg
> > @@ -216,6 +166,55 @@ struct lpg_led {
> >       struct lpg_channel *channels[] __counted_by(num_channels);
> >  };
> >
> > +/**
> > + * struct lpg - LPG device context
> > + * @dev:     pointer to LPG device
> > + * @map:     regmap for register access
> > + * @lock:    used to synchronize LED and pwm callback requests
> > + * @pwm:     PWM-chip object, if operating in PWM mode
> > + * @data:    reference to version specific data
> > + * @lut_base:        base address of the LUT block (optional)
> > + * @lut_size:        number of entries in the LUT block
> > + * @lut_bitmap:      allocation bitmap for LUT entries
> > + * @pbs_dev: PBS device
> > + * @lpg_chan_sdam:   LPG SDAM peripheral device
> > + * @lut_sdam:        LUT SDAM peripheral device
> > + * @pbs_en_bitmap:   bitmap for tracking PBS triggers
> > + * @triled_base: base address of the TRILED block (optional)
> > + * @triled_src:      power-source for the TRILED
> > + * @triled_has_atc_ctl:      true if there is TRI_LED_ATC_CTL register
> > + * @triled_has_src_sel:      true if there is TRI_LED_SRC_SEL register
> > + * @channels:        list of PWM channels
> > + * @num_channels: number of @channels
> > + */
>
> Should we be reordering the kerneldoc descriptions for '@channels' and
> '@num_channels' here to correctly match the updated order in the struct below?
Yeah probably.
>
> > +struct lpg {
> > +     struct device *dev;
> > +     struct regmap *map;
> > +
> > +     struct mutex lock;
> > +
> > +     struct pwm_chip *pwm;
> > +
> > +     const struct lpg_data *data;
> > +
> > +     u32 lut_base;
> > +     u32 lut_size;
> > +     unsigned long *lut_bitmap;
> > +
> > +     struct pbs_dev *pbs_dev;
> > +     struct nvmem_device *lpg_chan_sdam;
> > +     struct nvmem_device *lut_sdam;
> > +     unsigned long pbs_en_bitmap;
> > +
> > +     u32 triled_base;
> > +     u32 triled_src;
> > +     bool triled_has_atc_ctl;
> > +     bool triled_has_src_sel;
> > +
> > +     unsigned int num_channels;
> > +     struct lpg_channel channels[] __counted_by(num_channels);
> > +};
> > +
> >  /**
> >   * struct lpg_channel_data - per channel initialization data
> >   * @sdam_offset:     Channel offset in LPG SDAM
> > @@ -1475,12 +1474,6 @@ static int lpg_init_channels(struct lpg *lpg)
> >       struct lpg_channel *chan;
> >       int i;
> >
> > -     lpg->num_channels = data->num_channels;
> > -     lpg->channels = devm_kcalloc(lpg->dev, data->num_channels,
> > -                                  sizeof(struct lpg_channel), GFP_KERNEL);
> > -     if (!lpg->channels)
> > -             return -ENOMEM;
> > -
> >       for (i = 0; i < data->num_channels; i++) {
> >               chan = &lpg->channels[i];
> >
> > @@ -1603,18 +1596,20 @@ static int lpg_init_sdam(struct lpg *lpg)
> >
> >  static int lpg_probe(struct platform_device *pdev)
> >  {
> > +     const struct lpg_data *data;
> >       struct lpg *lpg;
> >       int ret;
> >       int i;
> >
> > -     lpg = devm_kzalloc(&pdev->dev, sizeof(*lpg), GFP_KERNEL);
> > +     data = of_device_get_match_data(&pdev->dev);
> > +     if (!data)
> > +             return -EINVAL;
> > +
> > +     lpg = devm_kzalloc(&pdev->dev, struct_size(lpg, channels, data->num_channels), GFP_KERNEL);
> >       if (!lpg)
> >               return -ENOMEM;
> >
> > -     lpg->data = of_device_get_match_data(&pdev->dev);
>
> You just NULL-ptr-derefed yourself.  Did you test this?
This can never return NULL. All compatible entries have a
corresponding data value pointing to a statically allocated struct.
>
> > -     if (!lpg->data)
> > -             return -EINVAL;
> > -
> > +     lpg->num_channels = data->num_channels;
> >       lpg->dev = &pdev->dev;
> >       mutex_init(&lpg->lock);
> >
> > --
> > 2.53.0
> >
> >
>
> --
> Lee Jones [李琼斯]