2 # (C) Copyright 2014 Google, Inc
3 # Simon Glass <sjg@chromium.org>
5 # SPDX-License-Identifier: GPL-2.0+
8 DEPRECATION NOTICE FOR arch/<arch>/lib/board.c
10 For board maintainers: Please submit patches for boards you maintain before
11 July 2014, to make them use generic board.
13 For architecture maintainers: Please submit patches to remove your
14 architecture-specific board.c file before October 2014.
20 U-Boot has traditionally had a board.c file for each architecture. This has
21 introduced quite a lot of duplication, with each architecture tending to do
22 initialisation slightly differently. To address this, a new 'generic board
23 init' feature was introduced a year ago in March 2013 (further motivation is
24 provided in the cover letter below).
30 The main change is that the arch/<arch>/lib/board.c file is being removed in
31 favour of common/board_f.c (for pre-relocation init) and common/board_r.c
32 (for post-relocation init).
34 Related to this, the global_data and bd_t structures now have a core set of
35 fields which are common to all architectures. Architecture-specific fields
36 have been moved to separate structures.
39 Supported Arcthitectures
40 ------------------------
42 If you are unlucky then your architecture may not support generic board.
43 The following architectures are supported at the time of writing:
51 If your architecture is not supported, you need to adjust your
52 arch/<arch>/config.mk file to include:
54 __HAVE_ARCH_GENERIC_BOARD := y
56 and test it with a suitable board, as follows.
59 Adding Support for your Board
60 -----------------------------
62 To enable generic board for your board, define CONFIG_SYS_GENERIC_BOARD in
63 your board config header file.
65 Test that U-Boot still functions correctly on your board, and fix any
66 problems you find. Don't be surprised if there are no problems - generic
67 board has had a reasonable amount of testing with common boards.
73 Please don't take this the wrong way - there is no intent to make your life
74 miserable, and we have the greatest respect and admiration for U-Boot users.
75 However, with any migration there has to be a period where the old way is
76 deprecated and removed. Every patch to the deprecated code introduces a
77 potential breakage in the new unused code. Therefore:
79 Boards or architectures not converted over to general board by the
80 end of 2014 may be forcibly changed over (potentially causing run-time
88 The full text of the original generic board series is reproduced below.
92 This series creates a generic board.c implementation which contains
93 the essential functions of the major arch/xxx/lib/board.c files.
95 What is the motivation for this change?
97 1. There is a lot of repeated code in the board.c files. Any change to
98 things like setting up the baud rate requires a change in 10 separate
101 2. Since there are 10 separate files, adding a new feature which requires
102 initialisation is painful since it must be independently added in 10
105 3. As time goes by the architectures naturely diverge since there is limited
106 pressure to compare features or even CONFIG options against simiilar things
107 in other board.c files.
109 4. New architectures must implement all the features all over again, and
110 sometimes in subtley different ways. This places an unfair burden on getting
111 a new architecture fully functional and running with U-Boot.
113 5. While it is a bit of a tricky change, I believe it is worthwhile and
114 achievable. There is no requirement that all code be common, only that
115 the code that is common should be located in common/board.c rather than
116 arch/xxx/lib/board.c.
118 All the functions of board_init_f() and board_init_r() are broken into
119 separate function calls so that they can easily be included or excluded
120 for a particular architecture. It also makes it easier to adopt Graeme's
121 initcall proposal when it is ready.
123 http://lists.denx.de/pipermail/u-boot/2012-January/114499.html
125 This series removes the dependency on generic relocation. So relocation
126 happens as one big chunk and is still completely arch-specific. See the
127 relocation series for a proposed solution to this for ARM:
129 http://lists.denx.de/pipermail/u-boot/2011-December/112928.html
131 or Graeme's recent x86 series v2:
133 http://lists.denx.de/pipermail/u-boot/2012-January/114467.html
135 Instead of moving over a whole architecture, this series takes the approach
136 of simply enabling generic board support for an architecture. It is then up
137 to each board to opt in by defining CONFIG_SYS_GENERIC_BOARD in the board
138 config file. If this is not done, then the code will be generated as
139 before. This allows both sets of code to co-exist until we are comfortable
140 with the generic approach, and enough boards run.
142 ARM is a relatively large board.c file and one which I can test, therefore
143 I think it is a good target for this series. On the other hand, x86 is
144 relatively small and simple, but different enough that it introduces a
145 few issues to be solved. So I have chosen both ARM and x86 for this series.
146 After a suggestion from Wolfgang I have added PPC also. This is the
147 largest and most feature-full board, so hopefully we have all bases
150 A generic global_data structure is also required. This might upset a few
151 people. Here is my basic reasoning: most fields are the same, all
152 architectures include and need it, most global_data.h files already have
153 #ifdefs to select fields for a particular SOC, so it is hard to
154 see why architecures are different in this area. We can perhaps add a
155 way to put architecture-specific fields into a separate header file, but
156 for now I have judged that to be counter-productive.
158 Similarly we need a generic bd_info structure, since generic code will
159 be accessing it. I have done this in the same way as global_data and the
162 There was dicussion on the list about passing gd_t around as a parameter
163 to pre-relocation init functions. I think this makes sense, but it can
164 be done as a separate change, and this series does not require it.
166 While this series needs to stand on its own (as with the link script
167 cleanup series and the generic relocation series) the goal is the
168 unification of the board init code. So I hope we can address issues with
169 this in mind, rather than focusing too narrowly on particular ARM, x86 or
172 I have run-tested ARM on Tegra Seaboard only. To try it out, define
173 CONFIG_SYS_GENERIC_BOARD in your board file and rebuild. Most likely on
174 x86 and PPC at least it will hang, but if you are lucky it will print
177 I have run this though MAKEALL with CONFIG_SYS_GENERIC_BOARD on for all
178 ARM, PPC and x86 boards. There are a few failures due to errors in
179 the board config, which I have sent patches for. The main issue is
180 just the difference between __bss_end and __bss_end__.
182 Note: the first group of commits are required for this series to build,
183 but could be separated out if required. I have included them here for
188 Simon Glass, sjg@chromium.org