2 Adding a new board to LinuxSH
3 ================================
5 Paul Mundt <lethal@linux-sh.org>
7 This document attempts to outline what steps are necessary to add support
8 for new boards to the LinuxSH port under the new 2.5 and 2.6 kernels. This
9 also attempts to outline some of the noticeable changes between the 2.4
10 and the 2.5/2.6 SH backend.
12 1. New Directory Structure
13 ==========================
15 The first thing to note is the new directory structure. Under 2.4, most
16 of the board-specific code (with the exception of stboards) ended up
17 in arch/sh/kernel/ directly, with board-specific headers ending up in
18 include/asm-sh/. For the new kernel, things are broken out by board type,
19 companion chip type, and CPU type. Looking at a tree view of this directory
20 hierarchy looks like the following:
29 | | `-- board-specific files
31 | | `-- board-specific files
33 | ... more boards here ...
38 | `-- board-specific headers
40 | `-- board-specific headers
42 .. more boards here ...
44 Next, for companion chips:
51 `-- cchip-specific files
53 ... and so on. Headers for the companion chips are treated the same way as
54 board-specific headers. Thus, include/asm-sh/hd64461 is home to all of the
55 hd64461-specific headers.
57 Finally, CPU family support is also abstracted:
64 | | | `-- SH-2 generic files
66 | | | `-- SH-3 generic files
68 | | `-- SH-4 generic files
70 | `-- This is also broken out per CPU family, so each family can
71 | have their own set of cache/tlb functions.
76 | `-- SH-2 specific headers
78 | `-- SH-3 specific headers
80 `-- SH-4 specific headers
82 It should be noted that CPU subtypes are _not_ abstracted. Thus, these still
83 need to be dealt with by the CPU family specific code.
88 The first thing to determine is whether the board you are adding will be
89 isolated, or whether it will be part of a family of boards that can mostly
90 share the same board-specific code with minor differences.
92 In the first case, this is just a matter of making a directory for your
93 board in arch/sh/boards/ and adding rules to hook your board in with the
94 build system (more on this in the next section). However, for board families
95 it makes more sense to have a common top-level arch/sh/boards/ directory
96 and then populate that with sub-directories for each member of the family.
97 Both the Solution Engine and the hp6xx boards are an example of this.
99 After you have setup your new arch/sh/boards/ directory, remember that you
100 should also add a directory in include/asm-sh for headers localized to this
101 board (if there are going to be more than one). In order to interoperate
102 seamlessly with the build system, it's best to have this directory the same
103 as the arch/sh/boards/ directory name, though if your board is again part of
104 a family, the build system has ways of dealing with this (via incdir-y
105 overloading), and you can feel free to name the directory after the family
108 There are a few things that each board is required to have, both in the
109 arch/sh/boards and the include/asm-sh/ hierarchy. In order to better
110 explain this, we use some examples for adding an imaginary board. For
111 setup code, we're required at the very least to provide definitions for
112 get_system_type() and platform_setup(). For our imaginary board, this
113 might look something like:
116 * arch/sh/boards/vapor/setup.c - Setup code for imaginary board
118 #include <linux/init.h>
119 #include <asm/rtc.h> /* for board_time_init() */
121 const char *get_system_type(void)
123 return "FooTech Vaporboard";
126 int __init platform_setup(void)
129 * If our hardware actually existed, we would do real
130 * setup here. Though it's also sane to leave this empty
131 * if there's no real init work that has to be done for
136 * Presume all FooTech boards have the same broken timer,
137 * and also presume that we've defined foo_timer_init to
138 * do something useful.
140 board_time_init = foo_timer_init;
142 /* Start-up imaginary PCI ... */
144 /* And whatever else ... */
149 Our new imaginary board will also have to tie into the machvec in order for it
152 machvec functions fall into a number of categories:
154 - I/O functions to IO memory (inb etc) and PCI/main memory (readb etc).
155 - I/O mapping functions (ioport_map, ioport_unmap, etc).
156 - a 'heartbeat' function.
157 - PCI and IRQ initialization routines.
158 - Consistent allocators (for boards that need special allocators,
159 particularly for allocating out of some board-specific SRAM for DMA
162 There are machvec functions added and removed over time, so always be sure to
163 consult include/asm-sh/machvec.h for the current state of the machvec.
165 The kernel will automatically wrap in generic routines for undefined function
166 pointers in the machvec at boot time, as machvec functions are referenced
167 unconditionally throughout most of the tree. Some boards have incredibly
168 sparse machvecs (such as the dreamcast and sh03), whereas others must define
169 virtually everything (rts7751r2d).
171 Adding a new machine is relatively trivial (using vapor as an example):
173 If the board-specific definitions are quite minimalistic, as is the case for
174 the vast majority of boards, simply having a single board-specific header is
177 - add a new file include/asm-sh/vapor.h which contains prototypes for
178 any machine specific IO functions prefixed with the machine name, for
179 example vapor_inb. These will be needed when filling out the machine
182 Note that these prototypes are generated automatically by setting
183 __IO_PREFIX to something sensible. A typical example would be:
185 #define __IO_PREFIX vapor
186 #include <asm/io_generic.h>
188 somewhere in the board-specific header. Any boards being ported that still
189 have a legacy io.h should remove it entirely and switch to the new model.
191 - Add machine vector definitions to the board's setup.c. At a bare minimum,
192 this must be defined as something like:
194 struct sh_machine_vector mv_vapor __initmv = {
199 - finally add a file arch/sh/boards/vapor/io.c, which contains definitions of
200 the machine specific io functions (if there are enough to warrant it).
202 3. Hooking into the Build System
203 ================================
205 Now that we have the corresponding directories setup, and all of the
206 board-specific code is in place, it's time to look at how to get the
207 whole mess to fit into the build system.
209 Large portions of the build system are now entirely dynamic, and merely
210 require the proper entry here and there in order to get things done.
212 The first thing to do is to add an entry to arch/sh/Kconfig, under the
218 select Vapor if configuring for a FooTech Vaporboard.
220 next, this has to be added into arch/sh/Makefile. All boards require a
221 machdir-y entry in order to be built. This entry needs to be the name of
222 the board directory as it appears in arch/sh/boards, even if it is in a
223 sub-directory (in which case, all parent directories below arch/sh/boards/
224 need to be listed). For our new board, this entry can look like:
226 machdir-$(CONFIG_SH_VAPOR) += vapor
228 provided that we've placed everything in the arch/sh/boards/vapor/ directory.
230 Next, the build system assumes that your include/asm-sh directory will also
231 be named the same. If this is not the case (as is the case with multiple
232 boards belonging to a common family), then the directory name needs to be
233 implicitly appended to incdir-y. The existing code manages this for the
234 Solution Engine and hp6xx boards, so see these for an example.
236 Once that is taken care of, it's time to add an entry for the mach type.
237 This is done by adding an entry to the end of the arch/sh/tools/mach-types
238 list. The method for doing this is self explanatory, and so we won't waste
239 space restating it here. After this is done, you will be able to use
240 implicit checks for your board if you need this somewhere throughout the
241 common code, such as:
243 /* Make sure we're on the FooTech Vaporboard */
244 if (!mach_is_vapor())
247 also note that the mach_is_boardname() check will be implicitly forced to
248 lowercase, regardless of the fact that the mach-types entries are all
249 uppercase. You can read the script if you really care, but it's pretty ugly,
250 so you probably don't want to do that.
252 Now all that's left to do is providing a defconfig for your new board. This
253 way, other people who end up with this board can simply use this config
254 for reference instead of trying to guess what settings are supposed to be
257 Also, as soon as you have copied over a sample .config for your new board
258 (assume arch/sh/configs/vapor_defconfig), you can also use this directly as a
259 build target, and it will be implicitly listed as such in the help text.
261 Looking at the 'make help' output, you should now see something like:
263 Architecture specific targets (sh):
264 zImage - Compressed kernel image (arch/sh/boot/zImage)
265 adx_defconfig - Build for adx
266 cqreek_defconfig - Build for cqreek
267 dreamcast_defconfig - Build for dreamcast
269 vapor_defconfig - Build for vapor
271 which then allows you to do:
273 $ make ARCH=sh CROSS_COMPILE=sh4-linux- vapor_defconfig vmlinux
275 which will in turn copy the defconfig for this board, run it through
276 oldconfig (prompting you for any new options since the time of creation),
277 and start you on your way to having a functional kernel for your new