docs/overview.txt
author "Yann E. MORIN" <yann.morin.1998@anciens.enib.fr>
Sat Jan 23 16:49:56 2010 +0100 (2010-01-23)
changeset 1745 41c2a6716511
parent 1714 5d693a13c84a
child 1803 2c0ed9ec9a8c
permissions -rw-r--r--
docs: do not document CT_KERNEL_ARCH, it's going away...

Since we merged the 32/64-bit archs, and we explicitly used
the Linux kernel arch name, we can get rid of CT_KERNEL_ARCH.
yann@1
     1
File.........: overview.txt
yann@197
     2
Content......: Overview of how crosstool-NG works.
yann@92
     3
Copyrigth....: (C) 2007 Yann E. MORIN <yann.morin.1998@anciens.enib.fr>
yann@192
     4
License......: Creative Commons Attribution Share Alike (CC-by-sa), v2.5
yann@92
     5
yann@628
     6
____________________
yann@628
     7
                   /
yann@628
     8
Table Of Content  /
yann@628
     9
_________________/
yann@628
    10
yann@628
    11
yann@628
    12
Introduction
yann@628
    13
History
yann@1581
    14
Referring to crosstool-NG
yann@628
    15
Installing crosstool-NG
yann@628
    16
  Install method
yann@628
    17
  The hacker's way
yann@1048
    18
  Preparing for packaging
yann@837
    19
  Shell completion
yann@628
    20
  Contributed code
yann@628
    21
Configuring crosstool-NG
yann@628
    22
  Interesting config options
yann@628
    23
  Re-building an existing toolchain
yann@628
    24
Running crosstool-NG
yann@628
    25
  Stopping and restarting a build
yann@628
    26
  Testing all toolchains at once
yann@628
    27
  Overriding the number of // jobs
yann@1491
    28
  Note on // jobs
yann@1493
    29
  Tools wrapper
yann@628
    30
Using the toolchain
yann@1554
    31
  The 'populate' script
yann@628
    32
Toolchain types
yann@1551
    33
  Seemingly-native toolchains
yann@1575
    34
Contributing
yann@1575
    35
  Sending a bug report
yann@1575
    36
  Sending patches
yann@628
    37
Internals
yann@628
    38
  Makefile front-end
yann@628
    39
  Kconfig parser
yann@628
    40
  Architecture-specific
yann@628
    41
  Adding a new version of a component
yann@628
    42
  Build scripts
yann@628
    43
yann@1580
    44
yann@1
    45
________________
yann@1
    46
               /
yann@1
    47
Introduction  /
yann@1
    48
_____________/
yann@1
    49
yann@1
    50
crosstool-NG aims at building toolchains. Toolchains are an essential component
yann@1
    51
in a software development project. It will compile, assemble and link the code
rpjday@436
    52
that is being developed. Some pieces of the toolchain will eventually end up
yann@1
    53
in the resulting binary/ies: static libraries are but an example.
yann@1
    54
yann@1
    55
So, a toolchain is a very sensitive piece of software, as any bug in one of the
yann@1
    56
components, or a poorly configured component, can lead to execution problems,
yann@1
    57
ranging from poor performance, to applications ending unexpectedly, to
yann@1
    58
mis-behaving software (which more than often is hard to detect), to hardware
rpjday@436
    59
damage, or even to human risks (which is more than regrettable).
yann@1
    60
yann@1
    61
Toolchains are made of different piece of software, each being quite complex
yann@1
    62
and requiring specially crafted options to build and work seamlessly. This
yann@1
    63
is usually not that easy, even in the not-so-trivial case of native toolchains.
yann@1
    64
The work reaches a higher degree of complexity when it comes to cross-
yann@40
    65
compilation, where it can become quite a nightmare...
yann@1
    66
yann@40
    67
Some cross-toolchains exist on the internet, and can be used for general
yann@1
    68
development, but they have a number of limitations:
yann@1
    69
  - they can be general purpose, in that they are configured for the majority:
yann@1
    70
    no optimisation for your specific target,
yann@1
    71
  - they can be prepared for a specific target and thus are not easy to use,
yann@1
    72
    nor optimised for, or even supporting your target,
rpjday@436
    73
  - they often are using aging components (compiler, C library, etc...) not
yann@1
    74
    supporting special features of your shiny new processor;
yann@1
    75
On the other side, these toolchain offer some advantages:
yann@1
    76
  - they are ready to use and quite easy to install and setup,
yann@1
    77
  - they are proven if used by a wide community.
yann@1
    78
yann@1
    79
But once you want to get all the juice out of your specific hardware, you will
yann@197
    80
want to build your own toolchain. This is where crosstool-NG comes into play.
yann@1
    81
rpjday@436
    82
There are also a number of tools that build toolchains for specific needs,
rpjday@436
    83
which are not really scalable. Examples are:
rpjday@436
    84
  - buildroot (buildroot.uclibc.org) whose main purpose is to build root file
yann@1
    85
    systems, hence the name. But once you have your toolchain with buildroot,
yann@1
    86
    part of it is installed in the root-to-be, so if you want to build a whole
yann@1
    87
    new root, you either have to save the existing one as a template and
yann@1
    88
    restore it later, or restart again from scratch. This is not convenient,
yann@1
    89
  - ptxdist (www.pengutronix.de/software/ptxdist), whose purpose is very
yann@1
    90
    similar to buildroot,
yann@702
    91
  - other projects (openembedded.org for example), which are again used to
yann@1
    92
    build root file systems.
yann@1
    93
rpjday@436
    94
crosstool-NG is really targeted at building toolchains, and only toolchains.
yann@1
    95
It is then up to you to use it the way you want.
yann@1
    96
yann@1580
    97
yann@1
    98
___________
yann@1
    99
          /
yann@1
   100
History  /
yann@1
   101
________/
yann@1
   102
rpjday@436
   103
crosstool was first 'conceived' by Dan Kegel, who offered it to the community
yann@1
   104
as a set of scripts, a repository of patches, and some pre-configured, general
yann@1
   105
purpose setup files to be used to configure crosstool. This is available at
yann@203
   106
http://www.kegel.com/crosstool, and the subversion repository is hosted on
yann@203
   107
google at http://code.google.com/p/crosstool/.
yann@1
   108
yann@1
   109
I once managed to add support for uClibc-based toolchains, but it did not make
yann@437
   110
into mainline, mostly because I didn't have time to port the patch forward to
yann@1
   111
the new versions, due in part to the big effort it was taking.
yann@1
   112
yann@1
   113
So I decided to clean up crosstool in the state it was, re-order the things
yann@437
   114
in place, add appropriate support for what I needed, that is uClibc support
yann@437
   115
and a menu-driven configuration, named the new implementation crosstool-NG,
yann@437
   116
(standing for crosstool Next Generation, as many other comunity projects do,
yann@437
   117
and as a wink at the TV series "Star Trek: The Next Generation" ;-) ) and
yann@437
   118
made it available to the community, in case it was of interest to any one.
yann@1
   119
yann@1580
   120
yann@1581
   121
_____________________________
yann@1581
   122
                            /
yann@1581
   123
Referring to crosstool-NG  /
yann@1581
   124
__________________________/
yann@1581
   125
yann@1581
   126
yann@1581
   127
The long name of the project is crosstool-NG:
yann@1581
   128
  * no leading uppercase (except as first word in a sentence)
yann@1581
   129
  * crosstool and NG separated with a hyphen (dash)
yann@1581
   130
  * NG in uppercase
yann@1581
   131
yann@1581
   132
Crosstool-NG can also be referred to by its short name CT-NG:
yann@1581
   133
  * all in uppercase
yann@1581
   134
  * CT and NG separated with a hyphen (dash)
yann@1581
   135
yann@1581
   136
The long name is preferred over the short name, except in mail subjects, where
yann@1581
   137
the short name is a better fit.
yann@1581
   138
yann@1581
   139
When referring to a specific version of crosstool-NG, append the version number
yann@1581
   140
either as:
yann@1581
   141
  * crosstool-NG X.Y.Z
yann@1581
   142
    - the long name, a space, and the version string
yann@1581
   143
  * crosstool-ng-X.Y.Z
yann@1581
   144
    - the long name in lowercase, a hyphen (dash), and the version string
yann@1581
   145
    - this is used to name the release tarballs
yann@1581
   146
  * crosstool-ng-X.Y.Z+hg_id
yann@1581
   147
    - the long name in lowercase, a hyphen, the version string, and the Hg id
yann@1581
   148
      (as returned by: ct-ng version)
yann@1581
   149
    - this is used to differentiate between releases and snapshots
yann@1581
   150
yann@1581
   151
The frontend to crosstool-NG is the command ct-ng:
yann@1581
   152
  * all in lowercase
yann@1581
   153
  * ct and ng separated by a hyphen (dash)
yann@1581
   154
yann@1581
   155
yann@294
   156
___________________________
yann@294
   157
                          /
yann@294
   158
Installing crosstool-NG  /
yann@294
   159
________________________/
yann@294
   160
yann@294
   161
There are two ways you can use crosstool-NG:
yann@294
   162
 - build and install it, then get rid of the sources like you'd do for most
yann@294
   163
   programs,
yann@294
   164
 - or only build it and run from the source directory.
yann@294
   165
yann@294
   166
The former should be used if you got crosstool-NG from a packaged tarball, see
rpjday@436
   167
"Install method", below, while the latter is most useful for developpers that
yann@1576
   168
use a clone of the repository, and want to submit patches, see "The Hacker's
yann@294
   169
way", below.
yann@294
   170
yann@294
   171
Install method |
yann@294
   172
---------------+
yann@294
   173
yann@294
   174
If you go for the install, then you just follow the classical, but yet easy
yann@294
   175
./configure way:
yann@294
   176
  ./configure --prefix=/some/place
yann@294
   177
  make
yann@294
   178
  make install
yann@294
   179
  export PATH="${PATH}:/some/place/bin"
yann@294
   180
yann@294
   181
You can then get rid of crosstool-NG source. Next create a directory to serve
yann@294
   182
as a working place, cd in there and run:
yann@294
   183
  ct-ng help
yann@294
   184
yann@294
   185
See below for complete usage.
yann@294
   186
yann@294
   187
The Hacker's way |
yann@294
   188
-----------------+
yann@294
   189
yann@294
   190
If you go the hacker's way, then the usage is a bit different, although very
yann@294
   191
simple:
yann@294
   192
  ./configure --local
yann@294
   193
  make
yann@294
   194
yann@294
   195
Now, *do not* remove crosstool-NG sources. They are needed to run crosstool-NG!
yann@294
   196
Stay in the directory holding the sources, and run:
yann@294
   197
  ./ct-ng help
yann@294
   198
yann@294
   199
See below for complete usage.
yann@294
   200
yann@1554
   201
Now, provided you used a clone of the repository, you can send me your changes.
yann@1575
   202
See the section titled CONTRIBUTING, below, for how to submit changees.
yann@294
   203
yann@1048
   204
Preparing for packaging |
yann@1048
   205
------------------------+
yann@1048
   206
yann@1048
   207
If you plan on packaging crosstool-NG, you surely don't want to install it
yann@1048
   208
in your root file system. The install procedure of crosstool-NG honors the
yann@1048
   209
DESTDIR variable:
yann@1048
   210
yann@1048
   211
  ./configure --prefix=/usr
yann@1048
   212
  make
rpjday@1291
   213
  make DESTDIR=/packaging/place install
yann@1048
   214
yann@837
   215
Shell completion |
yann@837
   216
-----------------+
yann@837
   217
yann@837
   218
crosstool-NG comes with a shell script fragment that defines bash-compatible
yann@837
   219
completion. That shell fragment is currently not installed automatically, but
yann@837
   220
this is planned.
yann@837
   221
yann@837
   222
To install the shell script fragment, you have two options:
yann@837
   223
 - install system-wide, most probably by copying ct-ng.comp into
yann@837
   224
   /etc/bash_completion.d/
yann@837
   225
 - install for a single user, by copying ct-ng.comp into ${HOME}/ and
yann@837
   226
   sourcing this file from your ${HOME}/.bashrc
yann@837
   227
yann@456
   228
Contributed code |
yann@456
   229
-----------------+
yann@456
   230
yann@456
   231
Some people contibuted code that couldn't get merged for various reasons. This
yann@1574
   232
code is available as lzma-compressed patches, in the contrib/ sub-directory.
yann@1574
   233
These patches are to be applied to the source of crosstool-NG, prior to
yann@1574
   234
installing, using something like the following:
yann@1574
   235
  lzcat contrib/foobar.patch.lzma |patch -p1
yann@620
   236
yann@620
   237
There is no guarantee that a particuliar contribution applies to the current
yann@620
   238
version of crosstool-ng, or that it will work at all. Use contributions at
yann@620
   239
your own risk.
yann@620
   240
yann@1580
   241
yann@168
   242
____________________________
yann@168
   243
                           /
yann@168
   244
Configuring crosstool-NG  /
yann@168
   245
_________________________/
yann@168
   246
yann@620
   247
crosstool-NG is configured with a configurator presenting a menu-stuctured set
yann@620
   248
of options. These options let you specify the way you want your toolchain
yann@620
   249
built, where you want it installed, what architecture and specific processor it
yann@277
   250
will support, the version of the components you want to use, etc... The
yann@277
   251
value for those options are then stored in a configuration file.
yann@277
   252
yann@476
   253
The configurator works the same way you configure your Linux kernel. It is
yann@277
   254
assumed you now how to handle this.
yann@168
   255
yann@168
   256
To enter the menu, type:
yann@192
   257
  ct-ng menuconfig
yann@168
   258
yann@203
   259
Almost every config item has a help entry. Read them carefully.
yann@168
   260
yann@168
   261
String and number options can refer to environment variables. In such a case,
yann@192
   262
you must use the shell syntax: ${VAR}. You shall neither single- nor double-
yann@294
   263
quote the string/number options.
yann@168
   264
yann@192
   265
There are three environment variables that are computed by crosstool-NG, and
yann@168
   266
that you can use:
yann@168
   267
yann@168
   268
CT_TARGET:
yann@335
   269
  It represents the target tuple you are building for. You can use it for
yann@168
   270
  example in the installation/prefix directory, such as:
yann@168
   271
    /opt/x-tools/${CT_TARGET}
yann@168
   272
yann@168
   273
CT_TOP_DIR:
yann@182
   274
  The top directory where crosstool-NG is running. You shouldn't need it in
yann@182
   275
  most cases. There is one case where you may need it: if you have local
yann@182
   276
  patches and you store them in your running directory, you can refer to them
yann@168
   277
  by using CT_TOP_DIR, such as:
yann@168
   278
    ${CT_TOP_DIR}/patches.myproject
yann@168
   279
yann@168
   280
CT_VERSION:
yann@192
   281
  The version of crosstool-NG you are using. Not much use for you, but it's
yann@168
   282
  there if you need it.
yann@168
   283
yann@168
   284
Interesting config options |
yann@476
   285
---------------------------+
yann@168
   286
yann@168
   287
CT_LOCAL_TARBALLS_DIR:
yann@277
   288
  If you already have some tarballs in a direcotry, enter it here. That will
yann@197
   289
  speed up the retrieving phase, where crosstool-NG would otherwise download
yann@168
   290
  those tarballs.
yann@168
   291
yann@168
   292
CT_PREFIX_DIR:
yann@168
   293
  This is where the toolchain will be installed in (and for now, where it
yann@437
   294
  will run from). Common use is to add the target tuple in the directory
yann@277
   295
  path, such as (see above):
yann@277
   296
    /opt/x-tools/${CT_TARGET}
yann@168
   297
yann@168
   298
CT_TARGET_VENDOR:
yann@168
   299
  An identifier for your toolchain, will take place in the vendor part of the
yann@335
   300
  target tuple. It shall *not* contain spaces or dashes. Usually, keep it
yann@168
   301
  to a one-word string, or use underscores to separate words if you need.
yann@168
   302
  Avoid dots, commas, and special characters.
yann@168
   303
yann@168
   304
CT_TARGET_ALIAS:
yann@168
   305
  An alias for the toolchian. It will be used as a prefix to the toolchain
yann@168
   306
  tools. For example, you will have ${CT_TARGET_ALIAS}-gcc
yann@168
   307
yann@246
   308
Also, if you think you don't see enough versions, you can try to enable one of
yann@246
   309
those:
yann@246
   310
yann@246
   311
CT_OBSOLETE:
yann@246
   312
  Show obsolete versions or tools. Most of the time, you don't want to base
yann@246
   313
  your toolchain on too old a version (of gcc, for example). But at times, it
yann@246
   314
  can come handy to use such an old version for regression tests. Those old
yann@1300
   315
  versions are hidden behind CT_OBSOLETE. Those versions (or features) are so
yann@1300
   316
  marked because maintaining support for those in crosstool-NG would be too
yann@1300
   317
  costly, time-wise, and time is dear.
yann@246
   318
yann@246
   319
CT_EXPERIMENTAL:
yann@246
   320
  Show experimental versions or tools. Again, you might not want to base your
yann@246
   321
  toolchain on too recent tools (eg. gcc) for production. But if you need a
yann@246
   322
  feature present only in a recent version, or a new tool, you can find them
yann@1300
   323
  hidden behind CT_EXPERIMENTAL. Those versions (or features) did not (yet)
yann@1300
   324
  receive thorough testing in crosstool-NG, and/or are not mature enough to
yann@1300
   325
  be blindly trusted.
yann@246
   326
yann@276
   327
Re-building an existing toolchain |
yann@276
   328
----------------------------------+
yann@276
   329
yann@276
   330
If you have an existing toolchain, you can re-use the options used to build it
yann@276
   331
to create a new toolchain. That needs a very little bit of effort on your side
yann@894
   332
but is quite easy. The options to build a toolchain are saved with the
yann@894
   333
toolchain, and you can retrieve this configuration by running:
yann@894
   334
  ${CT_TARGET}-config
yann@276
   335
yann@894
   336
This will dump the configuration to stdout, so to rebuild a toolchain with this
yann@894
   337
configuration, the following is all you need to do:
yann@894
   338
  ${CT_TARGET}-config >.config
yann@1098
   339
  ct-ng oldconfig
yann@276
   340
yann@894
   341
Then, you can review and change the configuration by running:
yann@894
   342
  ct-ng menuconfig
yann@276
   343
yann@1580
   344
yann@168
   345
________________________
yann@168
   346
                       /
yann@168
   347
Running crosstool-NG  /
yann@168
   348
_____________________/
yann@1
   349
yann@168
   350
To build the toolchain, simply type:
yann@203
   351
  ct-ng build
yann@135
   352
yann@135
   353
This will use the above configuration to retrieve, extract and patch the
yann@135
   354
components, build, install and eventually test your newly built toolchain.
yann@1
   355
yann@1
   356
You are then free to add the toolchain /bin directory in your PATH to use
yann@1
   357
it at will.
yann@1
   358
yann@135
   359
In any case, you can get some terse help. Just type:
yann@192
   360
  ct-ng help
yann@203
   361
or:
yann@203
   362
  man 1 ct-ng
yann@135
   363
rpjday@436
   364
Stopping and restarting a build |
yann@476
   365
--------------------------------+
yann@135
   366
yann@135
   367
If you want to stop the build after a step you are debugging, you can pass the
yann@135
   368
variable STOP to make:
fr@1643
   369
  ct-ng build STOP=some_step
yann@135
   370
yann@135
   371
Conversely, if you want to restart a build at a specific step you are
yann@135
   372
debugging, you can pass the RESTART variable to make:
fr@1643
   373
  ct-ng build RESTART=some_step
yann@135
   374
yann@136
   375
Alternatively, you can call make with the name of a step to just do that step:
yann@192
   376
  ct-ng libc_headers
yann@136
   377
is equivalent to:
fr@1643
   378
  ct-ng build RESTART=libc_headers STOP=libc_headers
yann@136
   379
yann@304
   380
The shortcuts +step_name and step_name+ allow to respectively stop or restart
yann@136
   381
at that step. Thus:
fr@1643
   382
  ct-ng +libc_headers              and:    ct-ng libc_headers+
yann@136
   383
are equivalent to:
fr@1643
   384
  ct-ng build STOP=libc_headers    and:    ct-ng build RESTART=libc_headers
yann@136
   385
yann@181
   386
To obtain the list of acceptable steps, please call:
yann@544
   387
  ct-ng list-steps
yann@181
   388
yann@168
   389
Note that in order to restart a build, you'll have to say 'Y' to the config
yann@168
   390
option CT_DEBUG_CT_SAVE_STEPS, and that the previous build effectively went
yann@168
   391
that far.
yann@92
   392
yann@1025
   393
Building all toolchains at once |
yann@1025
   394
--------------------------------+
yann@92
   395
yann@1025
   396
You can build all samples; simply call:
yann@1025
   397
  ct-ng build-all
yann@40
   398
yann@335
   399
Overriding the number of // jobs |
yann@476
   400
---------------------------------+
yann@335
   401
yann@335
   402
If you want to override the number of jobs to run in // (the -j option to
yann@335
   403
make), you can either re-enter the menuconfig, or simply add it on the command
yann@335
   404
line, as such:
yann@335
   405
  ct-ng build.4
yann@335
   406
yann@335
   407
which tells crosstool-NG to override the number of // jobs to 4.
yann@335
   408
yann@335
   409
You can see the actions that support overriding the number of // jobs in
yann@335
   410
the help menu. Those are the ones with [.#] after them (eg. build[.#] or
yann@1025
   411
build-all[.#], and so on...).
yann@1025
   412
yann@1025
   413
Note on // jobs |
yann@1025
   414
----------------+
yann@1025
   415
yann@1025
   416
The crosstool-NG script 'ct-ng' is a Makefile-script. It does *not* execute
yann@1025
   417
in parallel (there is not much to gain). When speaking of // jobs, we are
yann@1025
   418
refering to the number of // jobs when making the *components*. That is, we
yann@1025
   419
speak of the number of // jobs used to build gcc, glibc, and so on...
yann@1025
   420
yann@1493
   421
Tools wrapper |
yann@1493
   422
--------------+
yann@1493
   423
yann@1493
   424
Starting with gcc-4.3 come two new dependencies: GMP and MPFR. With gcc-4.4,
yann@1493
   425
come three new ones: GMP, PPL and CLooG/ppl. These are libraries that enable
yann@1493
   426
advanced features to gcc. Additionally, some of the libraries can be used by
yann@1493
   427
binutils and gdb. Unfortunately, not all systems on which crosstool-NG runs
yann@1493
   428
have all of those libraries. And for those that do, the versions of those
yann@1493
   429
libraries may be older than the version required by gcc.
yann@1493
   430
yann@1493
   431
This is why crosstool-NG builds its own set of libraries as part of the
yann@1493
   432
toolchain.
yann@1493
   433
yann@1493
   434
The libraries are built as shared libraries, because building them as static
yann@1493
   435
libraries has some short-comings. This poses no problem at build time, as
yann@1513
   436
crosstool-NG correctly points gcc (and binutils and gdb) to the correct
yann@1493
   437
place where our own version of the libraries are installed. But it poses
yann@1493
   438
a problem when gcc et al. are run: the place where the libraries are is most
yann@1493
   439
probably not known to the host dynamic linker. Still worse, if the host system
yann@1493
   440
has its own versions, then ld.so would load the wrong library!
yann@1493
   441
yann@1493
   442
So we have to force the dynamic linker to load the correct version. We do this
yann@1493
   443
by using the LD_LIBRARY_PATH variable, that informs the dynamic linker where
yann@1493
   444
to look for shared libraries prior to searching its standard places. But we
yann@1493
   445
can't impose that burden on all the system (because it'd be a nightmare to
yann@1513
   446
configure, and because two toolchains on the same system may use different
yann@1493
   447
versions of the libraries); so we have to do it on a per-toolchain basis.
yann@1493
   448
yann@1493
   449
So we rename all binaries of the toolchain (by adding a dot '.' as their first
yann@1493
   450
character), and add a small program, the so-called "tools wrapper", that
yann@1493
   451
correctly sets LD_LIBRARY_PATH prior to running the real tool.
yann@1493
   452
yann@1493
   453
First, the wrapper was written as a POSIX-compliant shell script. That shell
yann@1493
   454
script is very simple, if not trivial, and works great. The only drawback is
yann@1493
   455
that it does not work on host systems that lack a shell, for example the
yann@1493
   456
MingW32 environment. To solve the issue, the wrapper has been re-written in C,
yann@1493
   457
and compiled at build time. This C wrapper is much more complex than the shell
yann@1493
   458
script, and although it sems to be working, it's been only lightly tested.
yann@1493
   459
Some of the expected short-comings with this C wrapper are;
yann@1493
   460
 - multi-byte file names may not be handled correctly
yann@1493
   461
 - it's really big for what it does
yann@1493
   462
yann@1493
   463
So, the default wrapper installed with your toolchain is the shell script.
yann@1493
   464
If you know that your system is missing a shell, then you shall use the C
yann@1493
   465
wrapper (and report back whether it works, or does not work, for you).
yann@1493
   466
yann@335
   467
yann@227
   468
_______________________
yann@227
   469
                      /
yann@227
   470
Using the toolchain  /
yann@227
   471
____________________/
yann@227
   472
yann@227
   473
Using the toolchain is as simple as adding the toolchain's bin directory in
yann@227
   474
your PATH, such as:
yann@227
   475
  export PATH="${PATH}:/your/toolchain/path/bin"
yann@227
   476
yann@335
   477
and then using the target tuple to tell the build systems to use your
yann@227
   478
toolchain:
yann@335
   479
  ./configure --target=your-target-tuple
yann@294
   480
or
yann@335
   481
  make CC=your-target-tuple-gcc
yann@294
   482
or
yann@335
   483
  make CROSS_COMPILE=your-target-tuple-
yann@294
   484
and so on...
yann@227
   485
yann@476
   486
It is strongly advised not to use the toolchain sys-root directory as an
yann@620
   487
install directory for your programs/packages. If you do so, you will not be
yann@476
   488
able to use your toolchain for another project. It is even strongly advised
yann@476
   489
that your toolchain is chmod-ed to read-only once successfully build, so that
yann@620
   490
you don't go polluting your toolchain with your programs/packages' files.
yann@476
   491
yann@476
   492
Thus, when you build a program/package, install it in a separate directory,
yann@476
   493
eg. /your/root. This directory is the /image/ of what would be in the root file
yann@620
   494
system of your target, and will contain all that your programs/packages have
yann@476
   495
installed.
yann@476
   496
Yann@1405
   497
The 'populate' script |
Yann@1405
   498
----------------------+
Yann@1405
   499
yann@227
   500
When your root directory is ready, it is still missing some important bits: the
yann@227
   501
toolchain's libraries. To populate your root directory with those libs, just
yann@227
   502
run:
yann@335
   503
  your-target-tuple-populate -s /your/root -d /your/root-populated
yann@227
   504
yann@227
   505
This will copy /your/root into /your/root-populated, and put the needed and only
yann@227
   506
the needed libraries there. Thus you don't polute /your/root with any cruft that
yann@227
   507
would no longer be needed should you have to remove stuff. /your/root always
yann@227
   508
contains only those things you install in it.
yann@227
   509
yann@227
   510
You can then use /your/root-populated to build up your file system image, a
yann@227
   511
tarball, or to NFS-mount it from your target, or whatever you need.
yann@227
   512
Yann@1405
   513
The populate script accepts the following options:
yann@294
   514
Yann@1405
   515
 -s src_dir
Yann@1405
   516
    Use 'src_dir' as the un-populated root directory.
yann@294
   517
Yann@1405
   518
 -d dst_dir
Yann@1405
   519
    Put the populated root directory in 'dst_dir'.
Yann@1405
   520
Yann@1405
   521
 -l lib1 [...]
Yann@1405
   522
    Always add specified libraries.
Yann@1405
   523
Yann@1405
   524
 -L file
Yann@1405
   525
    Always add libraries listed in 'file'.
yann@294
   526
yann@294
   527
 -f
Yann@1405
   528
    Remove 'dst_dir' if it previously existed; continue even if any library
Yann@1405
   529
    specified with -l or -L is missing.
yann@294
   530
yann@294
   531
 -v
yann@294
   532
    Be verbose, and tell what's going on (you can see exactly where libs are
yann@294
   533
    coming from).
yann@294
   534
yann@294
   535
 -h
Yann@1405
   536
    Print the help.
Yann@1405
   537
Yann@1405
   538
See 'your-target-tuple-populate -h' for more information on the options.
yann@294
   539
Yann@1406
   540
Here is how populate works:
Yann@1406
   541
Yann@1406
   542
  1) performs some sanity checks:
Yann@1406
   543
     - src_dir and dst_dir are specified
Yann@1406
   544
     - src_dir exists
Yann@1406
   545
     - unless forced, dst_dir does not exist
Yann@1406
   546
     - src_dir != dst_dir
Yann@1406
   547
Yann@1406
   548
  2) copy src_dir to dst_dir
Yann@1406
   549
Yann@1406
   550
  3) add forced libraries to dst_dir
Yann@1406
   551
     - build the list from -l and -L options
Yann@1406
   552
     - get forced libraries from the sysroot (see below for heuristics)
Yann@1406
   553
       - abort on the first missing library, unless -f is specified
Yann@1406
   554
Yann@1406
   555
  4) add all missing libraries to dst_dir
Yann@1406
   556
     - scan dst_dir for every ELF files that are 'executable' or
Yann@1406
   557
       'shared object'
Yann@1406
   558
     - list the "NEEDED Shared library" fields
Yann@1406
   559
       - check if the library is already in dst_dir/lib or dst_dir/usr/lib
Yann@1406
   560
       - if not, get the library from the sysroot
Yann@1406
   561
         - if it's in sysroot/lib, copy it to dst_dir/lib
Yann@1406
   562
         - if it's in sysroot/usr/lib, copy it to dst_dir/usr/lib
Yann@1406
   563
         - in both cases, use the SONAME of the library to create the file
Yann@1406
   564
           in dst_dir
Yann@1406
   565
         - if it was not found in the sysroot, this is an error.
Yann@1406
   566
yann@1580
   567
yann@40
   568
___________________
yann@40
   569
                  /
yann@40
   570
Toolchain types  /
yann@40
   571
________________/
yann@40
   572
yann@40
   573
There are four kinds of toolchains you could encounter.
yann@40
   574
yann@40
   575
First off, you must understand the following: when it comes to compilers there
yann@40
   576
are up to four machines involved:
yann@40
   577
  1) the machine configuring the toolchain components: the config machine
yann@40
   578
  2) the machine building the toolchain components:    the build machine
yann@40
   579
  3) the machine running the toolchain:                the host machine
yann@203
   580
  4) the machine the toolchain is generating code for: the target machine
yann@40
   581
yann@40
   582
We can most of the time assume that the config machine and the build machine
yann@40
   583
are the same. Most of the time, this will be true. The only time it isn't
yann@40
   584
is if you're using distributed compilation (such as distcc). Let's forget
yann@40
   585
this for the sake of simplicity.
yann@40
   586
yann@40
   587
So we're left with three machines:
yann@40
   588
 - build
yann@40
   589
 - host
yann@40
   590
 - target
yann@40
   591
yann@40
   592
Any toolchain will involve those three machines. You can be as pretty sure of
yann@40
   593
this as "2 and 2 are 4". Here is how they come into play:
yann@40
   594
yann@40
   595
1) build == host == target
yann@40
   596
    This is a plain native toolchain, targetting the exact same machine as the
yann@40
   597
    one it is built on, and running again on this exact same machine. You have
yann@40
   598
    to build such a toolchain when you want to use an updated component, such
yann@40
   599
    as a newer gcc for example.
yann@197
   600
    crosstool-NG calls it "native".
yann@40
   601
yann@40
   602
2) build == host != target
yann@40
   603
    This is a classic cross-toolchain, which is expected to be run on the same
yann@40
   604
    machine it is compiled on, and generate code to run on a second machine,
yann@40
   605
    the target.
yann@197
   606
    crosstool-NG calls it "cross".
yann@40
   607
yann@40
   608
3) build != host == target
yann@40
   609
    Such a toolchain is also a native toolchain, as it targets the same machine
yann@40
   610
    as it runs on. But it is build on another machine. You want such a
yann@40
   611
    toolchain when porting to a new architecture, or if the build machine is
yann@40
   612
    much faster than the host machine.
yann@197
   613
    crosstool-NG calls it "cross-native".
yann@40
   614
yann@40
   615
4) build != host != target
yann@92
   616
    This one is called a canadian-toolchain (*), and is tricky. The three
yann@40
   617
    machines in play are different. You might want such a toolchain if you
yann@40
   618
    have a fast build machine, but the users will use it on another machine,
yann@40
   619
    and will produce code to run on a third machine.
yann@197
   620
    crosstool-NG calls it "canadian".
yann@40
   621
yann@197
   622
crosstool-NG can build all these kinds of toolchains (or is aiming at it,
yann@197
   623
anyway!)
yann@40
   624
yann@40
   625
(*) The term Canadian Cross came about because at the time that these issues
yann@40
   626
    were all being hashed out, Canada had three national political parties.
yann@40
   627
    http://en.wikipedia.org/wiki/Cross_compiler
yann@40
   628
yann@1551
   629
yann@1575
   630
________________
yann@1575
   631
               /
yann@1575
   632
Contributing  /
yann@1575
   633
_____________/
yann@1575
   634
yann@1575
   635
Sending a bug report |
yann@1575
   636
---------------------+
yann@1575
   637
yann@1575
   638
If you need to send a bug report, please send a mail with subject
yann@1575
   639
prefixed with "[CT_NG]" with to following destinations:
yann@1575
   640
    TO: yann.morin.1998 (at) anciens.enib.fr
yann@1575
   641
    CC: crossgcc (at) sourceware.org
yann@1575
   642
yann@1575
   643
Sending patches |
yann@1575
   644
----------------+
yann@1575
   645
yann@1575
   646
If you want to enhance crosstool-NG, there's a to-do list in the TODO file.
yann@1575
   647
yann@1575
   648
Patches should come with the appropriate SoB line. A SoB line is typically
yann@1575
   649
something like:
yann@1575
   650
   Signed-off-by: John DOE <john.doe@somewhere.net>
yann@1575
   651
yann@1575
   652
The SoB line is clearly described in Documentation/SubmittingPatches , section
yann@1575
   653
12, of your favourite Linux kernel source tree.
yann@1575
   654
yann@1575
   655
Then you'll need to correctly configure Mercurial. There are two extensions
yann@1575
   656
that you may find usefull:
yann@1575
   657
  - mq        : http://mercurial.selenic.com/wiki/MqExtension
yann@1575
   658
  - patchbomb : http://mercurial.selenic.com/wiki/PatchbombExtension
yann@1575
   659
yann@1575
   660
Commit messages should look like (without leading pipes):
yann@1575
   661
 |component: short, one-line description
yann@1575
   662
 |
yann@1575
   663
 |optional longer description
yann@1575
   664
 |on multiple lines if needed
yann@1575
   665
yann@1575
   666
Here is an example commit message (see revision a53a5e1d61db):
yann@1575
   667
 |comp-libs/cloog: fix building
yann@1575
   668
 |
yann@1575
   669
 |For CLooG/PPL 0.15.3, the directory name was simply cloog-ppl.
yann@1575
   670
 |For any later versions, the directory name does have the version, such as
yann@1575
   671
 |cloog-ppl-0.15.4.
yann@1575
   672
yann@1575
   673
Here's a typical hacking session:
yann@1575
   674
  hg clone http://ymorin.is-a-geek.org/hg/crosstool-ng crosstool-ng
yann@1575
   675
  cd crosstool-ng
yann@1575
   676
  hg qinit
yann@1575
   677
  hg qnew -D -U -e my_first_patch
yann@1575
   678
  *edit patch description*
yann@1575
   679
  *hack* *hack* *check* *fails* *hack* *hack* *check* *works*
yann@1575
   680
  hg qref -D -e
yann@1575
   681
  *edit patch description, serving as commit message*
yann@1575
   682
  hg qnew -D -U -e my_second_patch
yann@1575
   683
  *edit patch description*
yann@1575
   684
  *hack* *hack* *check* *fails* *hack* *hack* *check* *works*
yann@1575
   685
  hg qref -D -e
yann@1575
   686
  *edit patch description, serving as commit message*
yann@1575
   687
  hg email --outgoing --intro   \
yann@1575
   688
           --from '"Your Full NAME" <your.email (at) your.domain>'   \
yann@1575
   689
           --to '"Yann E. MORIN" <yann.morin.1998 (at) anciens.enib.fr>'    \
yann@1575
   690
           --cc 'crossgcc (at) sourceware.org'
yann@1575
   691
  *edit introductory message*
yann@1575
   692
  *wait for feedback*
yann@1575
   693
  *re-send if no answer for a few days*
yann@1575
   694
yann@1575
   695
Note: replace '(at)' above with a plain '@'.
yann@1575
   696
yann@1575
   697
yann@1
   698
_____________
yann@1
   699
            /
yann@1
   700
Internals  /
yann@1
   701
__________/
yann@1
   702
yann@92
   703
Internally, crosstool-NG is script-based. To ease usage, the frontend is
yann@92
   704
Makefile-based.
yann@92
   705
yann@92
   706
Makefile front-end |
yann@476
   707
-------------------+
yann@92
   708
yann@203
   709
The entry point to crosstool-NG is the Makefile script "ct-ng". Calling this
yann@203
   710
script with an action will act exactly as if the Makefile was in the current
yann@203
   711
working directory and make was called with the action as rule. Thus:
yann@203
   712
  ct-ng menuconfig
yann@294
   713
yann@203
   714
is equivalent to having the Makefile in CWD, and calling:
yann@203
   715
  make menuconfig
yann@203
   716
yann@203
   717
Having ct-ng as it is avoids copying the Makefile everywhere, and acts as a
yann@203
   718
traditional command.
yann@203
   719
yann@203
   720
ct-ng loads sub- Makefiles from the library directory $(CT_LIB_DIR), as set up
yann@203
   721
at configuration time with ./configure.
yann@203
   722
yann@437
   723
ct-ng also searches for config files, sub-tools, samples, scripts and patches in
yann@203
   724
that library directory.
yann@92
   725
yann@294
   726
Because of a stupid make behavior/bug I was unable to track down, implicit make
yann@294
   727
rules are disabled: installing with --local would triger those rules, and mconf
yann@294
   728
was unbuildable.
yann@294
   729
yann@182
   730
Kconfig parser |
yann@476
   731
---------------+
yann@92
   732
yann@965
   733
The kconfig language is a hacked version, vampirised from the Linux kernel
yann@965
   734
(http://www.kernel.org/), and (heavily) adapted to my needs.
yann@92
   735
yann@1040
   736
The list of the most notable changes (at least the ones I remember) follows:
yann@1040
   737
- the CONFIG_ prefix has been replaced with CT_
yann@1040
   738
- a leading | in prompts is skipped, and subsequent leading spaces are not
yann@1040
   739
  trimmed
yann@1040
   740
- otherwise leading spaces are silently trimmed
yann@1040
   741
yann@203
   742
The kconfig parsers (conf and mconf) are not installed pre-built, but as
yann@203
   743
source files. Thus you can have the directory where crosstool-NG is installed,
yann@203
   744
exported (via NFS or whatever) and have clients with different architectures
yann@203
   745
use the same crosstool-NG installation, and most notably, the same set of
yann@203
   746
patches.
yann@203
   747
yann@381
   748
Architecture-specific |
yann@476
   749
----------------------+
yann@381
   750
yann@628
   751
Note: this chapter is not really well written, and might thus be a little bit
yann@628
   752
complex to understand. To get a better grasp of what an architecture is, the
yann@628
   753
reader is kindly encouraged to look at the "arch/" sub-directory, and to the
yann@628
   754
existing architectures to see how things are laid out.
yann@628
   755
yann@381
   756
An architecture is defined by:
yann@381
   757
yann@381
   758
 - a human-readable name, in lower case letters, with numbers as appropriate.
yann@628
   759
   The underscore is allowed; space and special characters are not.
yann@628
   760
     Eg.: arm, x86_64
yann@903
   761
 - a file in "config/arch/", named after the architecture's name, and suffixed
yann@903
   762
   with ".in".
yann@903
   763
     Eg.: config/arch/arm.in
yann@903
   764
 - a file in "scripts/build/arch/", named after the architecture's name, and
yann@903
   765
   suffixed with ".sh".
yann@903
   766
     Eg.: scripts/build/arch/arm.sh
yann@628
   767
yann@903
   768
The architecture's ".in" file API:
yann@628
   769
 > the config option "ARCH_%arch%" (where %arch% is to be replaced with the
yann@628
   770
   actual architecture name).
yann@628
   771
   That config option must have *neither* a type, *nor* a prompt! Also, it can
yann@628
   772
   *not* depend on any other config option (EXPERIMENTAL is managed as above).
yann@628
   773
     Eg.:
yann@628
   774
       config ARCH_arm
yann@630
   775
   + mandatory:
yann@702
   776
       defines a (terse) help entry for this architecture:
yann@630
   777
       Eg.:
yann@630
   778
         config ARCH_arm
yann@630
   779
           help
yann@630
   780
             The ARM architecture.
yann@628
   781
   + optional:
yann@628
   782
       selects adequate associated config options.
yann@1038
   783
       Note: 64-bit architectures *shall* select ARCH_64
yann@628
   784
       Eg.:
yann@628
   785
         config ARCH_arm
yann@628
   786
           select ARCH_SUPPORTS_BOTH_ENDIAN
yann@628
   787
           select ARCH_DEFAULT_LE
yann@630
   788
           help
yann@630
   789
             The ARM architecture.
yann@1038
   790
       Eg.:
yann@1038
   791
         config ARCH_x86_64
yann@1038
   792
            select ARCH_64
yann@1038
   793
            help
yann@1038
   794
              The x86_64 architecture.
yann@628
   795
yann@628
   796
 > other target-specific options, at your discretion. Note however that to
yann@628
   797
   avoid name-clashing, such options shall be prefixed with "ARCH_%arch%",
yann@628
   798
   where %arch% is again replaced by the actual architecture name.
yann@628
   799
   (Note: due to historical reasons, and lack of time to clean up the code,
yann@628
   800
    I may have left some config options that do not completely conform to
yann@628
   801
    this, as the architecture name was written all upper case. However, the
yann@628
   802
    prefix is unique among architectures, and does not cause harm).
yann@381
   803
yann@903
   804
The architecture's ".sh" file API:
yann@965
   805
 > the function "CT_DoArchTupleValues"
yann@381
   806
   + parameters: none
yann@381
   807
   + environment:
yann@901
   808
     - all variables from the ".config" file,
yann@901
   809
     - the two variables "target_endian_eb" and "target_endian_el" which are
yann@901
   810
       the endianness suffixes
yann@381
   811
   + return value: 0 upon success, !0 upon failure
yann@381
   812
   + provides:
yann@391
   813
     - mandatory
yann@383
   814
     - the environment variable CT_TARGET_ARCH
yann@389
   815
     - contains:
yann@389
   816
       the architecture part of the target tuple.
yann@389
   817
       Eg.: "armeb" for big endian ARM
yann@389
   818
            "i386" for an i386
yann@389
   819
   + provides:
yann@391
   820
     - optional
yann@389
   821
     - the environment variable CT_TARGET_SYS
yann@456
   822
     - contains:
yann@383
   823
       the sytem part of the target tuple.
yann@383
   824
       Eg.: "gnu" for glibc on most architectures
yann@383
   825
            "gnueabi" for glibc on an ARM EABI
yann@383
   826
     - defaults to:
yann@383
   827
       - for glibc-based toolchain: "gnu"
yann@383
   828
       - for uClibc-based toolchain: "uclibc"
yann@383
   829
   + provides:
yann@383
   830
     - optional
yann@767
   831
     - the environment variables to configure the cross-gcc (defaults)
yann@767
   832
       - CT_ARCH_WITH_ARCH    : the gcc ./configure switch to select architecture level         ( "--with-arch=${CT_ARCH_ARCH}"   )
yann@767
   833
       - CT_ARCH_WITH_ABI     : the gcc ./configure switch to select ABI level                  ( "--with-abi=${CT_ARCH_ABI}"     )
yann@767
   834
       - CT_ARCH_WITH_CPU     : the gcc ./configure switch to select CPU instruction set        ( "--with-cpu=${CT_ARCH_CPU}"     )
yann@767
   835
       - CT_ARCH_WITH_TUNE    : the gcc ./configure switch to select scheduling                 ( "--with-tune=${CT_ARCH_TUNE}"   )
yann@767
   836
       - CT_ARCH_WITH_FPU     : the gcc ./configure switch to select FPU type                   ( "--with-fpu=${CT_ARCH_FPU}"     )
yann@767
   837
       - CT_ARCH_WITH_FLOAT   : the gcc ./configure switch to select floating point arithmetics ( "--with-float=soft" or /empty/  )
yann@391
   838
   + provides:
yann@391
   839
     - optional
yann@767
   840
     - the environment variables to pass to the cross-gcc to build target binaries (defaults)
yann@391
   841
       - CT_ARCH_ARCH_CFLAG   : the gcc switch to select architecture level                     ( "-march=${CT_ARCH_ARCH}"            )
yann@456
   842
       - CT_ARCH_ABI_CFLAG    : the gcc switch to select ABI level                              ( "-mabi=${CT_ARCH_ABI}"              )
yann@391
   843
       - CT_ARCH_CPU_CFLAG    : the gcc switch to select CPU instruction set                    ( "-mcpu=${CT_ARCH_CPU}"              )
yann@391
   844
       - CT_ARCH_TUNE_CFLAG   : the gcc switch to select scheduling                             ( "-mtune=${CT_ARCH_TUNE}"            )
yann@391
   845
       - CT_ARCH_FPU_CFLAG    : the gcc switch to select FPU type                               ( "-mfpu=${CT_ARCH_FPU}"              )
yann@391
   846
       - CT_ARCH_FLOAT_CFLAG  : the gcc switch to choose floating point arithmetics             ( "-msoft-float" or /empty/           )
yann@391
   847
       - CT_ARCH_ENDIAN_CFLAG : the gcc switch to choose big or little endian                   ( "-mbig-endian" or "-mlittle-endian" )
yann@391
   848
     - default to:
yann@391
   849
       see above.
yann@767
   850
   + provides:
yann@767
   851
     - optional
yann@767
   852
     - the environement variables to configure the core and final compiler, specific to this architecture:
yann@767
   853
       - CT_ARCH_CC_CORE_EXTRA_CONFIG   : additional, architecture specific core gcc ./configure flags
yann@767
   854
       - CT_ARCH_CC_EXTRA_CONFIG        : additional, architecture specific final gcc ./configure flags
yann@767
   855
     - default to:
yann@767
   856
       - all empty
yann@767
   857
   + provides:
yann@767
   858
     - optional
yann@767
   859
     - the architecture-specific CFLAGS and LDFLAGS:
yann@767
   860
       - CT_ARCH_TARGET_CLFAGS
yann@767
   861
       - CT_ARCH_TARGET_LDFLAGS
yann@767
   862
     - default to:
yann@767
   863
       - all empty
yann@628
   864
yann@903
   865
You can have a look at "config/arch/arm.in" and "scripts/build/arch/arm.sh" for
yann@903
   866
a quite complete example of what an actual architecture description looks like.
yann@901
   867
yann@890
   868
Kernel specific |
yann@890
   869
----------------+
yann@890
   870
yann@890
   871
A kernel is defined by:
yann@890
   872
yann@890
   873
 - a human-readable name, in lower case letters, with numbers as appropriate.
yann@890
   874
   The underscore is allowed; space and special characters are not (although
yann@890
   875
   they are internally replaced with underscores.
yann@890
   876
     Eg.: linux, bare-metal
yann@890
   877
 - a file in "config/kernel/", named after the kernel name, and suffixed with
yann@890
   878
   ".in".
yann@890
   879
     Eg.: config/kernel/linux.in, config/kernel/bare-metal.in
yann@901
   880
 - a file in "scripts/build/kernel/", named after the kernel name, and suffixed
yann@901
   881
   with ".sh".
yann@901
   882
     Eg.: scripts/build/kernel/linux.sh, scripts/build/kernel/bare-metal.sh
yann@890
   883
yann@890
   884
The kernel's ".in" file must contain:
yann@890
   885
 > an optional lines containing exactly "# EXPERIMENTAL", starting on the
yann@890
   886
   first column, and without any following space or other character.
yann@890
   887
   If this line is present, then this kernel is considered EXPERIMENTAL,
yann@890
   888
   and correct dependency on EXPERIMENTAL will be set.
yann@901
   889
yann@890
   890
 > the config option "KERNEL_%kernel_name%" (where %kernel_name% is to be
yann@890
   891
   replaced with the actual kernel name, with all special characters and
yann@890
   892
   spaces replaced by underscores).
yann@890
   893
   That config option must have *neither* a type, *nor* a prompt! Also, it can
yann@890
   894
   *not* depends on EXPERIMENTAL.
yann@890
   895
     Eg.: KERNEL_linux, KERNEL_bare_metal
yann@890
   896
   + mandatory:
yann@890
   897
       defines a (terse) help entry for this kernel.
yann@890
   898
       Eg.:
yann@890
   899
         config KERNEL_bare_metal
yann@890
   900
           help
yann@890
   901
             Build a compiler for use without any kernel.
yann@890
   902
   + optional:
yann@890
   903
       selects adequate associated config options.
yann@890
   904
       Eg.:
yann@890
   905
         config KERNEL_bare_metal
yann@890
   906
           select BARE_METAL
yann@890
   907
           help
yann@890
   908
             Build a compiler for use without any kernel.
yann@890
   909
yann@890
   910
 > other kernel specific options, at your discretion. Note however that, to
yann@890
   911
   avoid name-clashing, such options should be prefixed with
yann@890
   912
   "KERNEL_%kernel_name%", where %kernel_name% is again tp be replaced with
yann@890
   913
   the actual kernel name.
yann@890
   914
   (Note: due to historical reasons, and lack of time to clean up the code,
yann@890
   915
    I may have left some config options that do not completely conform to
yann@890
   916
    this, as the kernel name was written all upper case. However, the prefix
yann@890
   917
    is unique among kernels, and does not cause harm).
yann@890
   918
yann@901
   919
The kernel's ".sh" file API:
yann@901
   920
 > is a bash script fragment
yann@901
   921
yann@965
   922
 > defines the function CT_DoKernelTupleValues
yann@965
   923
   + see the architecture's CT_DoArchTupleValues, except for:
yann@965
   924
   + set the environment variable CT_TARGET_KERNEL, the kernel part of the
yann@965
   925
     target tuple
yann@965
   926
   + return value: ignored
yann@965
   927
yann@901
   928
 > defines the function "do_kernel_get":
yann@901
   929
   + parameters: none
yann@901
   930
   + environment:
yann@901
   931
      - all variables from the ".config" file.
yann@901
   932
   + return value: 0 for success, !0 for failure.
yann@901
   933
   + behavior: download the kernel's sources, and store the tarball into
yann@901
   934
     "${CT_TARBALLS_DIR}". To this end, a functions is available, that
yann@901
   935
     abstracts downloading tarballs:
yann@901
   936
     - CT_DoGet <tarball_base_name> <URL1 [URL...]>
yann@901
   937
       Eg.: CT_DoGet linux-2.6.26.5 ftp://ftp.kernel.org/pub/linux/kernel/v2.6
yann@901
   938
     Note: retrieving sources from svn, cvs, git and the likes is not supported
yann@901
   939
     by CT_DoGet. You'll have to do this by hand, as it is done for eglibc in
yann@901
   940
     "scripts/build/libc/eglibc.sh"
yann@901
   941
yann@901
   942
 > defines the function "do_kernel_extract":
yann@901
   943
   + parameters: none
yann@901
   944
   + environment:
yann@901
   945
      - all variables from the ".config" file,
yann@901
   946
   + return value: 0 for success, !0 for failure.
yann@901
   947
   + behavior: extract the kernel's tarball into "${CT_SRC_DIR}", and apply
yann@901
   948
     required patches. To this end, a function is available, that abstracts
yann@901
   949
     extracting tarballs:
yann@901
   950
     - CT_ExtractAndPatch <tarball_base_name>
yann@901
   951
       Eg.: CT_ExtractAndPatch linux-2.6.26.5
yann@901
   952
yann@901
   953
 > defines the function "do_kernel_headers":
yann@901
   954
   + parameters: none
yann@901
   955
   + environment:
yann@901
   956
      - all variables from the ".config" file,
yann@901
   957
   + return value: 0 for success, !0 for failure.
yann@901
   958
   + behavior: install the kernel headers (if any) in "${CT_SYSROOT_DIR}/usr/include"
yann@901
   959
yann@901
   960
 > defines any kernel-specific helper functions
yann@901
   961
   These functions, if any, must be prefixed with "do_kernel_%CT_KERNEL%_",
yann@901
   962
   where '%CT_KERNEL%' is to be replaced with the actual kernel name, to avoid
yann@901
   963
   any name-clashing.
yann@901
   964
yann@901
   965
You can have a look at "config/kernel/linux.in" and "scripts/build/kernel/linux.sh"
yann@903
   966
as an example of what a complex kernel description looks like.
yann@901
   967
yann@620
   968
Adding a new version of a component |
yann@476
   969
------------------------------------+
yann@476
   970
yann@476
   971
When a new component, such as the Linux kernel, gcc or any other is released,
yann@476
   972
adding the new version to crosstool-NG is quite easy. There is a script that
yann@476
   973
will do all that for you:
yann@1095
   974
  scripts/addToolVersion.sh
yann@476
   975
yann@476
   976
Run it with no option to get some help.
yann@381
   977
yann@203
   978
Build scripts |
yann@476
   979
--------------+
yann@203
   980
yann@203
   981
To Be Written later...