It's confusing to set the status of a node to disabled and afterwards
it's invisible. Similarly it's confusing to be able to enable invisible
nodes.
Also print disabled nodes.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
We used to populate the devices from the devicetree based on the
presence of the 'reg' property. This is incorrect since this only
allows us to probe devices with resources.
Instead use the 'simple-bus' property to see if we have iterate
deeper. This also registers devices with their buses as parents.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
memory is always in /mem, so call of_add_memory from of_probe once
instead of in the recursive tree iteration. This makes it possible
to limit the device population to nodes with the "simple-bus" property
set.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
When registering devices from the devicetree we check if these
devices already exist. When doing this not only check the resource
start but also the resource end.
This helps with the probing of simple buses for which child nodes
often begin at the very same address as the parents.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
This patch separates out the "generic" memory
segment registration function (of_add_memory_bank())
from of_add_memory().
The MIPS architecture has different view on memory
resources than the ARM and PPC architectures
so the "generic" of_add_memory_bank() is
unusable for the MIPS architecture.
We can add MIPS-specific of_add_memory_bank()
into arch/mips code.
Signed-off-by: Antony Pavlov <antonynpavlov@gmail.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Linux expects linux,initrd-end to contain the first unused address. As
this doesn't match the end semantic used by barebox (i.e. end contains
the last used address) adding one is necessary.
Without this change Linux fails for me to correctly extract a gzipped
cpio archive provided as initrd.
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
The phandles have to be parsed completely before registering the devices
from the devicetree. Otherwise drivers can't rely on of_find_node_by_phandle
in their probe.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
In order to be able to handle multiple devicetrees, do not assume
the tree to be unflattened is the barebox internal one. Instead,
just return a pointer to it and assign the barebox internal root_node
external to the unflatten function.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
The allnodes list makes it hard to handle multiple devicetrees. Having
a list to iterate over all nodes of a tree is still good to have though.
This patch uses the list_head of the root node as the head of the list.
This way the root node is no longer part of the list, but when iterating
over a tree the root node is not interesting anyway.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
of_find_node_by_path iterates over the allnodes list. Depending on
where the node we look for is, this can be significantly slower than
using the tree structure to look for a node, so iterate over the tree
instead.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
so when the first stage booloader of firmware provide the dtb
we can use it to probe the memory
also allow to print what we probe
Cc: Rob Herring <rob.herring@calxeda.com>
Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
We calculate the full path of the current node using fdt_get_path which
takes a lot of time since libfdt has to iterate over the dtb several times.
Speed this up by creating and using a of_find_child function which does
not have to iterate over the whole tree but only over its direct children.
On an i.MX51 board this speeds up unflatten the tree from 60ms to 3.7ms.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
of_new_node now takes the parent node and the name as argument and
creates a new node. This simplifies the caller and also makes the
function useful for others, so export it.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
This adds a function to delete a property from the currently loaded
devicetree. Also export new_property as of_new_property.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
This implements overlaying a currently loaded dtb with another dtb.
We used to return -EBUSY when a oftree is currently loaded. Instead
of doing this, check if a node already exists before creating a new
one. Similarly, if a property already exists, just overwrite the
value instead of creating a new property.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
We already have support for unflattening the devicetree. This patch
adds support for converting it back to a dtb.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
The process of unflatten the device tree is known from the kernel,
so rename the function, because that's what it does.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
- Add missing prototype for of_free()
- When root node is freed set root_node to NULL
- (re)scan aliases after freeing a node. This is necessary because
the freed node may be an alias or point to an alias
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
When of_find_node_by_path is called with a non existing path we
return just some node. return NULL instead.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
The console layer calls of_device_is_stdout_path for a new console. When
we are booting without devicetree then of_chosen is NULL which makes barebox
crash. Check for a NULL pointer in of_find_property to prevent this.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
This adds code to probe devices from a devicetree. Most helper
functions are directly imported from Linux.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>