(I accidentally sent this originally with HTML formatting, sorry if
you are getting this twice.)
Hi,
I only just started doing anything with kernel development, this is my
first email on this list, so I'm sorry if I ask a dumb question or
mess up on any rules of etiquette, I'm trying to follow them.
I'm sure there is a reason for this, but of_n_{addr,size}_cells() will
only return a parent's #address-cells or #size-cells values. I realize
of_property_read_u32(np, "#address-cells", &cells), for example, could
be called on the current node to get that value, but is seems to be
that of_n_{addr,size}_cells() is a little bit of a misleading name; I
expected it to return the above mentioned values of the node passed
in, and if that was not explicitly defined it would look to the
parents. I ask because I ran into a problem when working on a driver
and trying to include an .dtsi that contained a node with different
values for #address-cells and #size-cells; I could never get the
correct values returned, I only got the parents'. So my questions are
1) If the methods are not meant to return the values for
#address-cells and #size-cells should it be named something like
of_n_parent_{addr,size}_cells?
2) If they are meant to return the values for the current node could
there be a check that looks something like this?:
int of_n_addr_cells(struct device_node *np)
{
u32 cells;
if (!of_property_read_u32(np, "#address-cells", &cells))
return cells;
do {
if (np->parent)
np = np->parent;
if (!of_property_read_u32(np, "#address-cells", &cells))
return cells;
} while (np->parent);
/* No #address-cells property for the root node */
return OF_ROOT_NODE_ADDR_CELLS_DEFAULT;
}
I realize I could be missing something and that this isn't an
important change, but I just thought I'd ask. Thanks so much!
Ray
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
