On Thu, Nov 7, 2024 at 11:00 AM Pali Rohár <pali@xxxxxxxxxx> wrote: > > On Thursday 07 November 2024 09:32:55 Rob Herring (Arm) wrote: > > The mvebu "ranges" is a bit unusual with its own encoding of addresses, > > but it's still just normal "ranges" as far as parsing is concerned. > > Convert mvebu_get_tgt_attr() to use the for_each_of_range() iterator > > instead of open coding the parsing. > > > > Signed-off-by: Rob Herring (Arm) <robh@xxxxxxxxxx> > > --- > > Compile tested only. > > I see no reason for such change, which was even non tested at all and > does not fix any issue. Maintenance of the kernel is an issue. Maybe not for you, but for some of us. > There are more important issues in the driver, > it was decided that bug fixes are not going to be included (yet). The larger reason is to get rid of custom parsing of address properties throughout the kernel. As well as remove and consolidate uses of_n_addr_cells() and remove its support of long since deprecated behavior. Also, I intend to make of_get_property/of_find_property() warn about leaking data since it just returns a pointer into the raw DT data and we have no control over the lifetime. Then we can't free those properties. Why do we care? Overlays and Rust. Rob
