1 .. SPDX-License-Identifier: GPL-2.0
3 =============================
4 Adding a new board to LinuxSH
5 =============================
7 Paul Mundt <lethal@linux-sh.org>
9 This document attempts to outline what steps are necessary to add support
10 for new boards to the LinuxSH port under the new 2.5 and 2.6 kernels. This
11 also attempts to outline some of the noticeable changes between the 2.4
12 and the 2.5/2.6 SH backend.
14 1. New Directory Structure
15 ==========================
17 The first thing to note is the new directory structure. Under 2.4, most
18 of the board-specific code (with the exception of stboards) ended up
19 in arch/sh/kernel/ directly, with board-specific headers ending up in
20 include/asm-sh/. For the new kernel, things are broken out by board type,
21 companion chip type, and CPU type. Looking at a tree view of this directory
22 hierarchy looks like the following:
31 | | `-- board-specific files
33 | | `-- board-specific files
35 | ... more boards here ...
40 | `-- board-specific headers
42 | `-- board-specific headers
44 .. more boards here ...
46 Next, for companion chips::
54 `-- cchip-specific files
56 ... and so on. Headers for the companion chips are treated the same way as
57 board-specific headers. Thus, include/asm-sh/hd64461 is home to all of the
58 hd64461-specific headers.
60 Finally, CPU family support is also abstracted::
68 | | | `-- SH-2 generic files
70 | | | `-- SH-3 generic files
72 | | `-- SH-4 generic files
74 | `-- This is also broken out per CPU family, so each family can
75 | have their own set of cache/tlb functions.
80 | `-- SH-2 specific headers
82 | `-- SH-3 specific headers
84 `-- SH-4 specific headers
86 It should be noted that CPU subtypes are _not_ abstracted. Thus, these still
87 need to be dealt with by the CPU family specific code.
92 The first thing to determine is whether the board you are adding will be
93 isolated, or whether it will be part of a family of boards that can mostly
94 share the same board-specific code with minor differences.
96 In the first case, this is just a matter of making a directory for your
97 board in arch/sh/boards/ and adding rules to hook your board in with the
98 build system (more on this in the next section). However, for board families
99 it makes more sense to have a common top-level arch/sh/boards/ directory
100 and then populate that with sub-directories for each member of the family.
101 Both the Solution Engine and the hp6xx boards are an example of this.
103 After you have setup your new arch/sh/boards/ directory, remember that you
104 should also add a directory in include/asm-sh for headers localized to this
105 board (if there are going to be more than one). In order to interoperate
106 seamlessly with the build system, it's best to have this directory the same
107 as the arch/sh/boards/ directory name, though if your board is again part of
108 a family, the build system has ways of dealing with this (via incdir-y
109 overloading), and you can feel free to name the directory after the family
112 There are a few things that each board is required to have, both in the
113 arch/sh/boards and the include/asm-sh/ hierarchy. In order to better
114 explain this, we use some examples for adding an imaginary board. For
115 setup code, we're required at the very least to provide definitions for
116 get_system_type() and platform_setup(). For our imaginary board, this
117 might look something like::
120 * arch/sh/boards/vapor/setup.c - Setup code for imaginary board
122 #include <linux/init.h>
124 const char *get_system_type(void)
126 return "FooTech Vaporboard";
129 int __init platform_setup(void)
132 * If our hardware actually existed, we would do real
133 * setup here. Though it's also sane to leave this empty
134 * if there's no real init work that has to be done for
138 /* Start-up imaginary PCI ... */
140 /* And whatever else ... */
145 Our new imaginary board will also have to tie into the machvec in order for it
148 machvec functions fall into a number of categories:
150 - I/O functions to IO memory (inb etc) and PCI/main memory (readb etc).
151 - I/O mapping functions (ioport_map, ioport_unmap, etc).
152 - a 'heartbeat' function.
153 - PCI and IRQ initialization routines.
154 - Consistent allocators (for boards that need special allocators,
155 particularly for allocating out of some board-specific SRAM for DMA
158 There are machvec functions added and removed over time, so always be sure to
159 consult include/asm-sh/machvec.h for the current state of the machvec.
161 The kernel will automatically wrap in generic routines for undefined function
162 pointers in the machvec at boot time, as machvec functions are referenced
163 unconditionally throughout most of the tree. Some boards have incredibly
164 sparse machvecs (such as the dreamcast and sh03), whereas others must define
165 virtually everything (rts7751r2d).
167 Adding a new machine is relatively trivial (using vapor as an example):
169 If the board-specific definitions are quite minimalistic, as is the case for
170 the vast majority of boards, simply having a single board-specific header is
173 - add a new file include/asm-sh/vapor.h which contains prototypes for
174 any machine specific IO functions prefixed with the machine name, for
175 example vapor_inb. These will be needed when filling out the machine
178 Note that these prototypes are generated automatically by setting
179 __IO_PREFIX to something sensible. A typical example would be::
181 #define __IO_PREFIX vapor
182 #include <asm/io_generic.h>
184 somewhere in the board-specific header. Any boards being ported that still
185 have a legacy io.h should remove it entirely and switch to the new model.
187 - Add machine vector definitions to the board's setup.c. At a bare minimum,
188 this must be defined as something like::
190 struct sh_machine_vector mv_vapor __initmv = {
195 - finally add a file arch/sh/boards/vapor/io.c, which contains definitions of
196 the machine specific io functions (if there are enough to warrant it).
198 3. Hooking into the Build System
199 ================================
201 Now that we have the corresponding directories setup, and all of the
202 board-specific code is in place, it's time to look at how to get the
203 whole mess to fit into the build system.
205 Large portions of the build system are now entirely dynamic, and merely
206 require the proper entry here and there in order to get things done.
208 The first thing to do is to add an entry to arch/sh/Kconfig, under the
214 select Vapor if configuring for a FooTech Vaporboard.
216 next, this has to be added into arch/sh/Makefile. All boards require a
217 machdir-y entry in order to be built. This entry needs to be the name of
218 the board directory as it appears in arch/sh/boards, even if it is in a
219 sub-directory (in which case, all parent directories below arch/sh/boards/
220 need to be listed). For our new board, this entry can look like::
222 machdir-$(CONFIG_SH_VAPOR) += vapor
224 provided that we've placed everything in the arch/sh/boards/vapor/ directory.
226 Next, the build system assumes that your include/asm-sh directory will also
227 be named the same. If this is not the case (as is the case with multiple
228 boards belonging to a common family), then the directory name needs to be
229 implicitly appended to incdir-y. The existing code manages this for the
230 Solution Engine and hp6xx boards, so see these for an example.
232 Once that is taken care of, it's time to add an entry for the mach type.
233 This is done by adding an entry to the end of the arch/sh/tools/mach-types
234 list. The method for doing this is self explanatory, and so we won't waste
235 space restating it here. After this is done, you will be able to use
236 implicit checks for your board if you need this somewhere throughout the
237 common code, such as::
239 /* Make sure we're on the FooTech Vaporboard */
240 if (!mach_is_vapor())
243 also note that the mach_is_boardname() check will be implicitly forced to
244 lowercase, regardless of the fact that the mach-types entries are all
245 uppercase. You can read the script if you really care, but it's pretty ugly,
246 so you probably don't want to do that.
248 Now all that's left to do is providing a defconfig for your new board. This
249 way, other people who end up with this board can simply use this config
250 for reference instead of trying to guess what settings are supposed to be
253 Also, as soon as you have copied over a sample .config for your new board
254 (assume arch/sh/configs/vapor_defconfig), you can also use this directly as a
255 build target, and it will be implicitly listed as such in the help text.
257 Looking at the 'make help' output, you should now see something like:
259 Architecture specific targets (sh):
261 ======================= =============================================
262 zImage Compressed kernel image (arch/sh/boot/zImage)
263 adx_defconfig Build for adx
264 cqreek_defconfig Build for cqreek
265 dreamcast_defconfig Build for dreamcast
267 vapor_defconfig Build for vapor
268 ======================= =============================================
270 which then allows you to do::
272 $ make ARCH=sh CROSS_COMPILE=sh4-linux- vapor_defconfig vmlinux
274 which will in turn copy the defconfig for this board, run it through
275 oldconfig (prompting you for any new options since the time of creation),
276 and start you on your way to having a functional kernel for your new