Re: [PATCH V3 3/4] coresight: trbe: Add a representative coresight_platform_data for TRBE

[Suzuki K Poulose]

On 04/08/2023 10:18, Anshuman Khandual wrote:
> On 8/3/23 19:25, Suzuki K Poulose wrote:
> > On 03/08/2023 06:56, Anshuman Khandual wrote:
> > > TRBE coresight devices do not need regular connections information, as the
> > > paths get built between all percpu source and their respective percpu sink
> > > devices. Please refer 'commit 2cd87a7b293d ("coresight: core: Add support
> > > for dedicated percpu sinks")' which added support for percpu sink devices.
> > >
> > > coresight_register() expect device connections via the platform_data. TRBE
> > > devices do not have any graph connections and thus is empty. With upcoming
> > > ACPI support for TRBE, we do not get a real acpi_device and thus
> > > coresight_get_platform_dat() will end up in failures. Hence this allocates
> > > a zeroed coresight_platform_data structure and assigns that back into the
> > > device.
> > >
> > > Cc: Suzuki K Poulose <suzuki.poulose@xxxxxxx>
> > > Cc: Mike Leach <mike.leach@xxxxxxxxxx>
> > > Cc: Leo Yan <leo.yan@xxxxxxxxxx>
> > > Cc: Alexander Shishkin <alexander.shishkin@xxxxxxxxxxxxxxx>
> > > Cc: coresight@xxxxxxxxxxxxxxxx
> > > Cc: linux-arm-kernel@xxxxxxxxxxxxxxxxxxx
> > > Cc: linux-kernel@xxxxxxxxxxxxxxx
> > > Signed-off-by: Anshuman Khandual <anshuman.khandual@xxxxxxx>
> > > ---
> > >    drivers/hwtracing/coresight/coresight-trbe.c | 6 +++---
> > >    1 file changed, 3 insertions(+), 3 deletions(-)
> > >
> > > diff --git a/drivers/hwtracing/coresight/coresight-trbe.c b/drivers/hwtracing/coresight/coresight-trbe.c
> > > index 7720619909d6..e1d9d06e7725 100644
> > > --- a/drivers/hwtracing/coresight/coresight-trbe.c
> > > +++ b/drivers/hwtracing/coresight/coresight-trbe.c
> > > @@ -1494,9 +1494,9 @@ static int arm_trbe_device_probe(struct platform_device *pdev)
> > >        if (!drvdata)
> > >            return -ENOMEM;
> > >    -    pdata = coresight_get_platform_data(dev);
> > > -    if (IS_ERR(pdata))
> > > -        return PTR_ERR(pdata);
> > > +    pdata = devm_kzalloc(dev, sizeof(*pdata), GFP_KERNEL);
> > > +    if (!pdata)
> > > +        return -ENOMEM;
> >
> > Please could you add a comment in here, on why we use a dummy platform
> > data ? It is good to have documented it in the code too.
>
> Sure, will add the following in-code documentation.
>
> +       /*
> +        * TRBE coresight devices do not need regular connections
> +        * information, as the paths get built between all percpu
> +        * source and their respective percpu sink devices. Though
> +        * coresight_register() expect device connections via the
> +        * platform_data, which TRBE devices do not have. As they
> +        * are not real ACPI devices, coresight_get_platform_dat()

minor nit: s/coresight_get_platform_dat/coresight_get_platform_data/
here and above in the description.

Otherwise, looks good.

Suzuki