refactor-the-rule-for-regenerating-utf8data.h

   1 unicode: refactor the rule for regenerating utf8data.h
   2
   3 From: Masahiro Yamada <yamada.masahiro@socionext.com>
   4
   5 scripts/mkutf8data is used only when regenerating utf8data.h,
   6 which never happens in the normal kernel build. However, it is
   7 irrespectively built if CONFIG_UNICODE is enabled.
   8
   9 Moreover, there is no good reason for it to reside in the scripts/
  10 directory since it is only used in fs/unicode/.
  11
  12 Hence, move it from scripts/ to fs/unicode/.
  13
  14 In some cases, we bypass build artifacts in the normal build. The
  15 conventional way to do so is to surround the code with ifdef REGENERATE_*.
  16
  17 For example,
  18
  19  - 7373f4f83c71 ("kbuild: add implicit rules for parser generation")
  20  - 6aaf49b495b4 ("crypto: arm,arm64 - Fix random regeneration of S_shipped")
  21
  22 I rewrote the rule in a more kbuild'ish style.
  23
  24 In the normal build, utf8data.h is just shipped from the check-in file.
  25
  26 $ make
  27   [ snip ]
  28   SHIPPED fs/unicode/utf8data.h
  29   CC      fs/unicode/utf8-norm.o
  30   CC      fs/unicode/utf8-core.o
  31   CC      fs/unicode/utf8-selftest.o
  32   AR      fs/unicode/built-in.a
  33
  34 If you want to generate utf8data.h based on UCD, put *.txt files into
  35 fs/unicode/, then pass REGENERATE_UTF8DATA=1 from the command line.
  36 The mkutf8data tool will be automatically compiled to generate the
  37 utf8data.h from the *.txt files.
  38
  39 $ make REGENERATE_UTF8DATA=1
  40   [ snip ]
  41   HOSTCC  fs/unicode/mkutf8data
  42   GEN     fs/unicode/utf8data.h
  43   CC      fs/unicode/utf8-norm.o
  44   CC      fs/unicode/utf8-core.o
  45   CC      fs/unicode/utf8-selftest.o
  46   AR      fs/unicode/built-in.a
  47
  48 I renamed the check-in utf8data.h to utf8data.h_shipped so that this
  49 will work for the out-of-tree build.
  50
  51 You can update it based on the latest UCD like this:
  52
  53 $ make REGENERATE_UTF8DATA=1 fs/unicode/
  54 $ cp fs/unicode/utf8data.h fs/unicode/utf8data.h_shipped
  55
  56 Also, I added entries to .gitignore and dontdiff.
  57
  58 Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
  59 Signed-off-by: Theodore Ts'o <tytso@mit.edu>
  60 ---
  61
  62 Changes in v2:
  63  - Make this work correctly with O= option
  64
  65  Documentation/dontdiff                        |  2 ++
  66  fs/unicode/.gitignore                         |  2 ++
  67  fs/unicode/Makefile                           | 41 ++++++++++++++++++++-------
  68  fs/unicode/README.utf8data                    |  9 +++---
  69  {scripts => fs/unicode}/mkutf8data.c          |  0
  70  fs/unicode/{utf8data.h => utf8data.h_shipped} |  0
  71  scripts/Makefile                              |  1 -
  72  7 files changed, 38 insertions(+), 17 deletions(-)
  73  create mode 100644 fs/unicode/.gitignore
  74  rename {scripts => fs/unicode}/mkutf8data.c (100%)
  75  rename fs/unicode/{utf8data.h => utf8data.h_shipped} (100%)
  76
  77 diff --git a/Documentation/dontdiff b/Documentation/dontdiff
  78 index ef25a06..9369377 100644
  79 --- a/Documentation/dontdiff
  80 +++ b/Documentation/dontdiff
  81 @@ -176,6 +176,7 @@ mkprep
  82  mkregtable
  83  mktables
  84  mktree
  85 +mkutf8data
  86  modpost
  87  modules.builtin
  88  modules.order
  89 @@ -254,6 +255,7 @@ vsyscall_32.lds
  90  wanxlfw.inc
  91  uImage
  92  unifdef
  93 +utf8data.h
  94  wakeup.bin
  95  wakeup.elf
  96  wakeup.lds
  97 diff --git a/fs/unicode/.gitignore b/fs/unicode/.gitignore
  98 new file mode 100644
  99 index 0000000..0381e22
 100 --- /dev/null
 101 +++ b/fs/unicode/.gitignore
 102 @@ -0,0 +1,2 @@
 103 +mkutf8data
 104 +utf8data.h
 105 diff --git a/fs/unicode/Makefile b/fs/unicode/Makefile
 106 index 671d31f..d46e9ba 100644
 107 --- a/fs/unicode/Makefile
 108 +++ b/fs/unicode/Makefile
 109 @@ -5,15 +5,34 @@ obj-$(CONFIG_UNICODE_NORMALIZATION_SELFTEST) += utf8-selftest.o
 110
 111  unicode-y := utf8-norm.o utf8-core.o
 112
 113 -# This rule is not invoked during the kernel compilation.  It is used to
 114 -# regenerate the utf8data.h header file.
 115 -utf8data.h.new: *.txt $(objdir)/scripts/mkutf8data
 116 -       $(objdir)/scripts/mkutf8data \
 117 -               -a DerivedAge.txt \
 118 -               -c DerivedCombiningClass.txt \
 119 -               -p DerivedCoreProperties.txt \
 120 -               -d UnicodeData.txt \
 121 -               -f CaseFolding.txt \
 122 -               -n NormalizationCorrections.txt \
 123 -               -t NormalizationTest.txt \
 124 +$(obj)/utf8-norm.o: $(obj)/utf8data.h
 125 +
 126 +# In the normal build, the checked-in utf8data.h is just shipped.
 127 +#
 128 +# To generate utf8data.h from UCD, put *.txt files in this directory
 129 +# and pass REGENERATE_UTF8DATA=1 from the command line.
 130 +ifdef REGENERATE_UTF8DATA
 131 +
 132 +quiet_cmd_utf8data = GEN     $@
 133 +      cmd_utf8data = $< \
 134 +               -a $(srctree)/$(src)/DerivedAge.txt \
 135 +               -c $(srctree)/$(src)/DerivedCombiningClass.txt \
 136 +               -p $(srctree)/$(src)/DerivedCoreProperties.txt \
 137 +               -d $(srctree)/$(src)/UnicodeData.txt \
 138 +               -f $(srctree)/$(src)/CaseFolding.txt \
 139 +               -n $(srctree)/$(src)/NormalizationCorrections.txt \
 140 +               -t $(srctree)/$(src)/NormalizationTest.txt \
 141                 -o $@
 142 +
 143 +$(obj)/utf8data.h: $(obj)/mkutf8data $(filter %.txt, $(cmd_utf8data)) FORCE
 144 +       $(call if_changed,utf8data)
 145 +
 146 +else
 147 +
 148 +$(obj)/utf8data.h: $(src)/utf8data.h_shipped FORCE
 149 +       $(call if_changed,shipped)
 150 +
 151 +endif
 152 +
 153 +targets += utf8data.h
 154 +hostprogs-y += mkutf8data
 155 diff --git a/fs/unicode/README.utf8data b/fs/unicode/README.utf8data
 156 index eeb7561..459eebf 100644
 157 --- a/fs/unicode/README.utf8data
 158 +++ b/fs/unicode/README.utf8data
 159 @@ -41,15 +41,14 @@ released version of the UCD can be found here:
 160
 161    http://www.unicode.org/Public/UCD/latest/
 162
 163 -To build the utf8data.h file, from a kernel tree that has been built,
 164 -cd to this directory (fs/unicode) and run this command:
 165 +Then, build under fs/unicode/ with REGENERATE_UTF8DATA=1:
 166
 167 -       make C=../.. objdir=../.. utf8data.h.new
 168 +       make REGENERATE_UTF8DATA=1 fs/unicode/
 169
 170 -After sanity checking the newly generated utf8data.h.new file (the
 171 +After sanity checking the newly generated utf8data.h file (the
 172  version generated from the 12.1.0 UCD should be 4,109 lines long, and
 173  have a total size of 324k) and/or comparing it with the older version
 174 -of utf8data.h, rename it to utf8data.h.
 175 +of utf8data.h_shipped, rename it to utf8data.h_shipped.
 176
 177  If you are a kernel developer updating to a newer version of the
 178  Unicode Character Database, please update this README.utf8data file
 179 diff --git a/scripts/mkutf8data.c b/fs/unicode/mkutf8data.c
 180 similarity index 100%
 181 rename from scripts/mkutf8data.c
 182 rename to fs/unicode/mkutf8data.c
 183 diff --git a/fs/unicode/utf8data.h b/fs/unicode/utf8data.h_shipped
 184 similarity index 100%
 185 rename from fs/unicode/utf8data.h
 186 rename to fs/unicode/utf8data.h_shipped
 187 diff --git a/scripts/Makefile b/scripts/Makefile
 188 index b87e3e0..9d442ee 100644
 189 --- a/scripts/Makefile
 190 +++ b/scripts/Makefile
 191 @@ -20,7 +20,6 @@ hostprogs-$(CONFIG_ASN1)       += asn1_compiler
 192  hostprogs-$(CONFIG_MODULE_SIG)  += sign-file
 193  hostprogs-$(CONFIG_SYSTEM_TRUSTED_KEYRING) += extract-cert
 194  hostprogs-$(CONFIG_SYSTEM_EXTRA_CERTIFICATE) += insert-sys-cert
 195 -hostprogs-$(CONFIG_UNICODE) += mkutf8data
 196
 197  HOSTCFLAGS_sortextable.o = -I$(srctree)/tools/include
 198  HOSTCFLAGS_asn1_compiler.o = -I$(srctree)/include
 199 --
 200 2.7.4
 201
 202