RE: [PATCH 2/4] cdx: add sysfs for reset_all

[Gangurde, Abhijit]

[AMD Official Use Only - General]

> > Add sysfs for reset_all entry which resets all the
> > devices on the CDX bus.
>
> On all CDX busses, right?  Why all?  Why not per-bus?
>

This was on similar line to bus enable/disable. Would extend this to support per bus in v2.

>
> >
> > Co-developed-by: Puneet Gupta <puneet.gupta@xxxxxxx>
> > Signed-off-by: Puneet Gupta <puneet.gupta@xxxxxxx>
> > Co-developed-by: Nipun Gupta <nipun.gupta@xxxxxxx>
> > Signed-off-by: Nipun Gupta <nipun.gupta@xxxxxxx>
> > Signed-off-by: Abhijit Gangurde <abhijit.gangurde@xxxxxxx>
> > Reviewed-by: Pieter Jansen van Vuuren <pieter.jansen-van-
> vuuren@xxxxxxx>
> > Tested-by: Nikhil Agarwal <nikhil.agarwal@xxxxxxx>
> > ---
> >  Documentation/ABI/testing/sysfs-bus-cdx | 11 ++++++++++
> >  drivers/cdx/cdx.c                       | 29 +++++++++++++++++++++++++
> >  2 files changed, 40 insertions(+)
> >
> > diff --git a/Documentation/ABI/testing/sysfs-bus-cdx
> b/Documentation/ABI/testing/sysfs-bus-cdx
> > index 0afa85b3c63b..d9e00058471d 100644
> > --- a/Documentation/ABI/testing/sysfs-bus-cdx
> > +++ b/Documentation/ABI/testing/sysfs-bus-cdx
> > @@ -22,6 +22,17 @@ Description:
> >
> >               # echo 0 > /sys/bus/cdx/enable
> >
> > +What:              /sys/bus/cdx/reset_all
> > +Date:              July 2023
> > +Contact:   puneet.gupta@xxxxxxx
> > +Description:
> > +           Writing y/1/on to this file resets all the devices
> > +           present on the CDX bus
> > +
> > +           For example::
> > +
> > +             # echo 1 > /sys/bus/cdx/reset_all
>
> What does resetting a device mean will happen?

It would be same of pcie flr to the device. Would add more description in v2.

>
> > +
> >  What:              /sys/bus/cdx/devices/.../vendor
> >  Date:              March 2023
> >  Contact:   nipun.gupta@xxxxxxx
> > diff --git a/drivers/cdx/cdx.c b/drivers/cdx/cdx.c
> > index 48c493a43491..4d20047b55bb 100644
> > --- a/drivers/cdx/cdx.c
> > +++ b/drivers/cdx/cdx.c
> > @@ -106,6 +106,11 @@ int cdx_dev_reset(struct device *dev)
> >  }
> >  EXPORT_SYMBOL_GPL(cdx_dev_reset);
> >
> > +static int reset_cdx_device(struct device *dev, void *data)
> > +{
> > +   return cdx_dev_reset(dev);
> > +}
> > +
> >  /**
> >   * cdx_unregister_device - Unregister a CDX device
> >   * @dev: CDX device
> > @@ -433,9 +438,33 @@ static ssize_t rescan_store(const struct bus_type
> *bus,
> >  }
> >  static BUS_ATTR_WO(rescan);
> >
> > +static ssize_t reset_all_store(const struct bus_type *bus,
> > +                          const char *buf, size_t count)
> > +{
> > +   bool val;
> > +   int ret;
> > +
> > +   if (kstrtobool(buf, &val) < 0)
> > +           return -EINVAL;
> > +
> > +   if (!val)
> > +           return -EINVAL;
> > +
> > +   /* Reset all the devices attached to cdx bus */
> > +   ret = bus_for_each_dev(bus, NULL, NULL, reset_cdx_device);
>
> No locking needed?

Would protect this with  a lock.

>
> > +   if (ret) {
> > +           pr_err("error in CDX bus reset\n");
>
> What error?  For what device?  Put the error message in the reset
> callback, not here, as you have no idea what device failed.

Would correct in v2.

Thanks,
Abhijit