Add support for switching the endianness of regmap accesses via the
"little-endian", "big-endian", and "native-endian" boolean properties in
the device tree.

The default endianness is native endianness.

Signed-off-by: Mario Six <mario.six@gdsys.cc>
Reviewed-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>

It is useful to be able to treat the different ranges of a regmap
separately to be able to use distinct offset for them, but this is
currently not implemented in the regmap API.

To preserve backwards compatibility, add regmap_read_range and
regmap_write_range functions that take an additional parameter
'range_num' that identifies the range to operate on.

Reviewed-by: Anatolij Gustschin <agust@denx.de>
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Mario Six <mario.six@gdsys.cc>

The regmap functions currently assume that all register map accesses
have a data width of 32 bits, but there are maps that have different
widths.

To rectify this, implement the regmap_raw_read and regmap_raw_write
functions from the Linux kernel API that specify the width of a desired
read or write operation on a regmap.

Implement the regmap_read and regmap_write functions using these raw
functions in a backwards-compatible manner.

Reviewed-by: Anatolij Gustschin <agust@denx.de>
Signed-off-by: Mario Six <mario.six@gdsys.cc>
Reviewed-by: Simon Glass <sjg@chromium.org>

Add some debug output in cases where the initialization of a regmap
fails.

Reviewed-by: Anatolij Gustschin <agust@denx.de>
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Mario Six <mario.six@gdsys.cc>

Both fdtdec_get_addr_size_fixed and of_address_to_resource can fail with
an error, which is not currently checked during regmap initialization.

Since the indentation depth is already quite deep, extract a new
'init_range' method to do the initialization.

Reviewed-by: Anatolij Gustschin <agust@denx.de>
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Mario Six <mario.six@gdsys.cc>

ofnode_read_simple_addr_cells may fail and return a negative error code.
Check for this when initializing regmaps.

Also check if both_len is zero, since this is perfectly possible, and
would lead to a division-by-zero further down the line.

Reviewed-by: Anatolij Gustschin <agust@denx.de>
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Mario Six <mario.six@gdsys.cc>

Document the regmap_alloc() function.

Reviewed-by: Anatolij Gustschin <agust@denx.de>
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Mario Six <mario.six@gdsys.cc>

Add the regmap_update_bits() to simply the read/modify/write of registers
in a single command. The function is taken from Linux regmap
implementation.

Signed-off-by: Neil Armstrong <narmstrong@baylibre.com>
Reviewed-by: Simon Glass <sjg@chromium.org>

Currently, regmap_init_mem() takes a udevice. This requires the node
has already been associated with a device. It prevents syscon/regmap
from behaving like those in Linux.

Change the first argumenet to take a device node.

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Acked-by: Neil Armstrong <narmstrong@baylibre.com>
Reviewed-by: Simon Glass <sjg@chromium.org>

Putting zero length array at the end of struct is a common technique
to embed arbitrary length of members.  There is no good reason to let
regmap_alloc_count() branch by "if (count <= 1)".

As far as I understood the code, regmap->base is an alias of
regmap->ranges[0].start, but it is not helpful but make the code
just ugly.

Rename regmap_alloc_count() to regmap_alloc() because the _count
suffix seems pointless.

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
[trini: fixup cpu_info-rcar.c]
Signed-off-by: Tom Rini <trini@konsulko.com>