Update this powerpc sample.
/trunk/samples/powerpc-unknown-linux-gnu/crosstool.config | 161 84 77 0 +++++++++++----------
/trunk/samples/powerpc-unknown-linux-gnu/reported.by | 1 1 0 0 +
2 files changed, 85 insertions(+), 77 deletions(-)
1 File.........: overview.txt
2 Content......: Overview of how crosstool-NG works.
3 Copyrigth....: (C) 2007 Yann E. MORIN <yann.morin.1998@anciens.enib.fr>
4 License......: Creative Commons Attribution Share Alike (CC-by-sa), v2.5
14 Installing crosstool-NG
19 Configuring crosstool-NG
20 Interesting config options
21 Re-building an existing toolchain
23 Stopping and restarting a build
24 Testing all toolchains at once
25 Overriding the number of // jobs
32 Adding a new version of a component
40 crosstool-NG aims at building toolchains. Toolchains are an essential component
41 in a software development project. It will compile, assemble and link the code
42 that is being developed. Some pieces of the toolchain will eventually end up
43 in the resulting binary/ies: static libraries are but an example.
45 So, a toolchain is a very sensitive piece of software, as any bug in one of the
46 components, or a poorly configured component, can lead to execution problems,
47 ranging from poor performance, to applications ending unexpectedly, to
48 mis-behaving software (which more than often is hard to detect), to hardware
49 damage, or even to human risks (which is more than regrettable).
51 Toolchains are made of different piece of software, each being quite complex
52 and requiring specially crafted options to build and work seamlessly. This
53 is usually not that easy, even in the not-so-trivial case of native toolchains.
54 The work reaches a higher degree of complexity when it comes to cross-
55 compilation, where it can become quite a nightmare...
57 Some cross-toolchains exist on the internet, and can be used for general
58 development, but they have a number of limitations:
59 - they can be general purpose, in that they are configured for the majority:
60 no optimisation for your specific target,
61 - they can be prepared for a specific target and thus are not easy to use,
62 nor optimised for, or even supporting your target,
63 - they often are using aging components (compiler, C library, etc...) not
64 supporting special features of your shiny new processor;
65 On the other side, these toolchain offer some advantages:
66 - they are ready to use and quite easy to install and setup,
67 - they are proven if used by a wide community.
69 But once you want to get all the juice out of your specific hardware, you will
70 want to build your own toolchain. This is where crosstool-NG comes into play.
72 There are also a number of tools that build toolchains for specific needs,
73 which are not really scalable. Examples are:
74 - buildroot (buildroot.uclibc.org) whose main purpose is to build root file
75 systems, hence the name. But once you have your toolchain with buildroot,
76 part of it is installed in the root-to-be, so if you want to build a whole
77 new root, you either have to save the existing one as a template and
78 restore it later, or restart again from scratch. This is not convenient,
79 - ptxdist (www.pengutronix.de/software/ptxdist), whose purpose is very
81 - other projects (openembedded.org for example), which are again used to
82 build root file systems.
84 crosstool-NG is really targeted at building toolchains, and only toolchains.
85 It is then up to you to use it the way you want.
92 crosstool was first 'conceived' by Dan Kegel, who offered it to the community
93 as a set of scripts, a repository of patches, and some pre-configured, general
94 purpose setup files to be used to configure crosstool. This is available at
95 http://www.kegel.com/crosstool, and the subversion repository is hosted on
96 google at http://code.google.com/p/crosstool/.
98 I once managed to add support for uClibc-based toolchains, but it did not make
99 into mainline, mostly because I didn't have time to port the patch forward to
100 the new versions, due in part to the big effort it was taking.
102 So I decided to clean up crosstool in the state it was, re-order the things
103 in place, add appropriate support for what I needed, that is uClibc support
104 and a menu-driven configuration, named the new implementation crosstool-NG,
105 (standing for crosstool Next Generation, as many other comunity projects do,
106 and as a wink at the TV series "Star Trek: The Next Generation" ;-) ) and
107 made it available to the community, in case it was of interest to any one.
109 ___________________________
111 Installing crosstool-NG /
112 ________________________/
114 There are two ways you can use crosstool-NG:
115 - build and install it, then get rid of the sources like you'd do for most
117 - or only build it and run from the source directory.
119 The former should be used if you got crosstool-NG from a packaged tarball, see
120 "Install method", below, while the latter is most useful for developpers that
121 checked the code out from SVN, and want to submit patches, see "The Hacker's
127 If you go for the install, then you just follow the classical, but yet easy
129 ./configure --prefix=/some/place
132 export PATH="${PATH}:/some/place/bin"
134 You can then get rid of crosstool-NG source. Next create a directory to serve
135 as a working place, cd in there and run:
138 See below for complete usage.
143 If you go the hacker's way, then the usage is a bit different, although very
148 Now, *do not* remove crosstool-NG sources. They are needed to run crosstool-NG!
149 Stay in the directory holding the sources, and run:
152 See below for complete usage.
154 Now, provided you checked-out the code, you can send me your interesting changes
158 and mailing me the result! :-P
163 crosstool-NG comes with a shell script fragment that defines bash-compatible
164 completion. That shell fragment is currently not installed automatically, but
167 To install the shell script fragment, you have two options:
168 - install system-wide, most probably by copying ct-ng.comp into
169 /etc/bash_completion.d/
170 - install for a single user, by copying ct-ng.comp into ${HOME}/ and
171 sourcing this file from your ${HOME}/.bashrc
176 Some people contibuted code that couldn't get merged for various reasons. This
177 code is available as patches in the contrib/ sub-directory. These patches are
178 to be applied to the source of crosstool-NG, prior to installing.
180 An easy way to use contributed code is to pass the --with-contrib= option to
181 ./configure. The possible values depend upon which contributions are packaged
182 with your version, but you can get with it with passing one of those two
185 will list all available contributions
188 will select all avalaible contributions
190 There is no guarantee that a particuliar contribution applies to the current
191 version of crosstool-ng, or that it will work at all. Use contributions at
194 ____________________________
196 Configuring crosstool-NG /
197 _________________________/
199 crosstool-NG is configured with a configurator presenting a menu-stuctured set
200 of options. These options let you specify the way you want your toolchain
201 built, where you want it installed, what architecture and specific processor it
202 will support, the version of the components you want to use, etc... The
203 value for those options are then stored in a configuration file.
205 The configurator works the same way you configure your Linux kernel. It is
206 assumed you now how to handle this.
208 To enter the menu, type:
211 Almost every config item has a help entry. Read them carefully.
213 String and number options can refer to environment variables. In such a case,
214 you must use the shell syntax: ${VAR}. You shall neither single- nor double-
215 quote the string/number options.
217 There are three environment variables that are computed by crosstool-NG, and
221 It represents the target tuple you are building for. You can use it for
222 example in the installation/prefix directory, such as:
223 /opt/x-tools/${CT_TARGET}
226 The top directory where crosstool-NG is running. You shouldn't need it in
227 most cases. There is one case where you may need it: if you have local
228 patches and you store them in your running directory, you can refer to them
229 by using CT_TOP_DIR, such as:
230 ${CT_TOP_DIR}/patches.myproject
233 The version of crosstool-NG you are using. Not much use for you, but it's
234 there if you need it.
236 Interesting config options |
237 ---------------------------+
239 CT_LOCAL_TARBALLS_DIR:
240 If you already have some tarballs in a direcotry, enter it here. That will
241 speed up the retrieving phase, where crosstool-NG would otherwise download
245 This is where the toolchain will be installed in (and for now, where it
246 will run from). Common use is to add the target tuple in the directory
247 path, such as (see above):
248 /opt/x-tools/${CT_TARGET}
251 An identifier for your toolchain, will take place in the vendor part of the
252 target tuple. It shall *not* contain spaces or dashes. Usually, keep it
253 to a one-word string, or use underscores to separate words if you need.
254 Avoid dots, commas, and special characters.
257 An alias for the toolchian. It will be used as a prefix to the toolchain
258 tools. For example, you will have ${CT_TARGET_ALIAS}-gcc
260 Also, if you think you don't see enough versions, you can try to enable one of
264 Show obsolete versions or tools. Most of the time, you don't want to base
265 your toolchain on too old a version (of gcc, for example). But at times, it
266 can come handy to use such an old version for regression tests. Those old
267 versions are hidden behind CT_OBSOLETE.
270 Show experimental versions or tools. Again, you might not want to base your
271 toolchain on too recent tools (eg. gcc) for production. But if you need a
272 feature present only in a recent version, or a new tool, you can find them
273 hidden behind CT_EXPERIMENTAL.
275 Re-building an existing toolchain |
276 ----------------------------------+
278 If you have an existing toolchain, you can re-use the options used to build it
279 to create a new toolchain. That needs a very little bit of effort on your side
280 but is quite easy. The options to build a toolchain are saved with the
281 toolchain, and you can retrieve this configuration by running:
284 This will dump the configuration to stdout, so to rebuild a toolchain with this
285 configuration, the following is all you need to do:
286 ${CT_TARGET}-config >.config
288 Then, you can review and change the configuration by running:
291 ________________________
293 Running crosstool-NG /
294 _____________________/
296 To build the toolchain, simply type:
299 This will use the above configuration to retrieve, extract and patch the
300 components, build, install and eventually test your newly built toolchain.
302 You are then free to add the toolchain /bin directory in your PATH to use
305 In any case, you can get some terse help. Just type:
310 Stopping and restarting a build |
311 --------------------------------+
313 If you want to stop the build after a step you are debugging, you can pass the
314 variable STOP to make:
317 Conversely, if you want to restart a build at a specific step you are
318 debugging, you can pass the RESTART variable to make:
319 ct-ng RESTART=some_step
321 Alternatively, you can call make with the name of a step to just do that step:
324 ct-ng RESTART=libc_headers STOP=libc_headers
326 The shortcuts +step_name and step_name+ allow to respectively stop or restart
328 ct-ng +libc_headers and: ct-ng libc_headers+
330 ct-ng STOP=libc_headers and: ct-ng RESTART=libc_headers
332 To obtain the list of acceptable steps, please call:
335 Note that in order to restart a build, you'll have to say 'Y' to the config
336 option CT_DEBUG_CT_SAVE_STEPS, and that the previous build effectively went
339 Testing all toolchains at once |
340 -------------------------------+
342 You can test-build all samples; simply call:
345 Overriding the number of // jobs |
346 ---------------------------------+
348 If you want to override the number of jobs to run in // (the -j option to
349 make), you can either re-enter the menuconfig, or simply add it on the command
353 which tells crosstool-NG to override the number of // jobs to 4.
355 You can see the actions that support overriding the number of // jobs in
356 the help menu. Those are the ones with [.#] after them (eg. build[.#] or
357 regtest[.#], and so on...).
359 _______________________
361 Using the toolchain /
362 ____________________/
364 Using the toolchain is as simple as adding the toolchain's bin directory in
366 export PATH="${PATH}:/your/toolchain/path/bin"
368 and then using the target tuple to tell the build systems to use your
370 ./configure --target=your-target-tuple
372 make CC=your-target-tuple-gcc
374 make CROSS_COMPILE=your-target-tuple-
377 It is strongly advised not to use the toolchain sys-root directory as an
378 install directory for your programs/packages. If you do so, you will not be
379 able to use your toolchain for another project. It is even strongly advised
380 that your toolchain is chmod-ed to read-only once successfully build, so that
381 you don't go polluting your toolchain with your programs/packages' files.
383 Thus, when you build a program/package, install it in a separate directory,
384 eg. /your/root. This directory is the /image/ of what would be in the root file
385 system of your target, and will contain all that your programs/packages have
388 When your root directory is ready, it is still missing some important bits: the
389 toolchain's libraries. To populate your root directory with those libs, just
391 your-target-tuple-populate -s /your/root -d /your/root-populated
393 This will copy /your/root into /your/root-populated, and put the needed and only
394 the needed libraries there. Thus you don't polute /your/root with any cruft that
395 would no longer be needed should you have to remove stuff. /your/root always
396 contains only those things you install in it.
398 You can then use /your/root-populated to build up your file system image, a
399 tarball, or to NFS-mount it from your target, or whatever you need.
401 populate accepts the following options:
404 Use 'src_dir' as the 'source', un-populated root directory
407 Put the 'destination', populated root directory in 'dst_dir'
410 Remove 'dst_dir' if it previously existed
413 Be verbose, and tell what's going on (you can see exactly where libs are
424 There are four kinds of toolchains you could encounter.
426 First off, you must understand the following: when it comes to compilers there
427 are up to four machines involved:
428 1) the machine configuring the toolchain components: the config machine
429 2) the machine building the toolchain components: the build machine
430 3) the machine running the toolchain: the host machine
431 4) the machine the toolchain is generating code for: the target machine
433 We can most of the time assume that the config machine and the build machine
434 are the same. Most of the time, this will be true. The only time it isn't
435 is if you're using distributed compilation (such as distcc). Let's forget
436 this for the sake of simplicity.
438 So we're left with three machines:
443 Any toolchain will involve those three machines. You can be as pretty sure of
444 this as "2 and 2 are 4". Here is how they come into play:
446 1) build == host == target
447 This is a plain native toolchain, targetting the exact same machine as the
448 one it is built on, and running again on this exact same machine. You have
449 to build such a toolchain when you want to use an updated component, such
450 as a newer gcc for example.
451 crosstool-NG calls it "native".
453 2) build == host != target
454 This is a classic cross-toolchain, which is expected to be run on the same
455 machine it is compiled on, and generate code to run on a second machine,
457 crosstool-NG calls it "cross".
459 3) build != host == target
460 Such a toolchain is also a native toolchain, as it targets the same machine
461 as it runs on. But it is build on another machine. You want such a
462 toolchain when porting to a new architecture, or if the build machine is
463 much faster than the host machine.
464 crosstool-NG calls it "cross-native".
466 4) build != host != target
467 This one is called a canadian-toolchain (*), and is tricky. The three
468 machines in play are different. You might want such a toolchain if you
469 have a fast build machine, but the users will use it on another machine,
470 and will produce code to run on a third machine.
471 crosstool-NG calls it "canadian".
473 crosstool-NG can build all these kinds of toolchains (or is aiming at it,
476 (*) The term Canadian Cross came about because at the time that these issues
477 were all being hashed out, Canada had three national political parties.
478 http://en.wikipedia.org/wiki/Cross_compiler
485 Internally, crosstool-NG is script-based. To ease usage, the frontend is
491 The entry point to crosstool-NG is the Makefile script "ct-ng". Calling this
492 script with an action will act exactly as if the Makefile was in the current
493 working directory and make was called with the action as rule. Thus:
496 is equivalent to having the Makefile in CWD, and calling:
499 Having ct-ng as it is avoids copying the Makefile everywhere, and acts as a
502 ct-ng loads sub- Makefiles from the library directory $(CT_LIB_DIR), as set up
503 at configuration time with ./configure.
505 ct-ng also searches for config files, sub-tools, samples, scripts and patches in
506 that library directory.
508 Because of a stupid make behavior/bug I was unable to track down, implicit make
509 rules are disabled: installing with --local would triger those rules, and mconf
515 The kconfig language is a hacked version, vampirised from the Linux kernel
516 (http://www.kernel.org/), and (heavily) adapted to my needs.
518 The kconfig parsers (conf and mconf) are not installed pre-built, but as
519 source files. Thus you can have the directory where crosstool-NG is installed,
520 exported (via NFS or whatever) and have clients with different architectures
521 use the same crosstool-NG installation, and most notably, the same set of
524 Architecture-specific |
525 ----------------------+
527 Note: this chapter is not really well written, and might thus be a little bit
528 complex to understand. To get a better grasp of what an architecture is, the
529 reader is kindly encouraged to look at the "arch/" sub-directory, and to the
530 existing architectures to see how things are laid out.
532 An architecture is defined by:
534 - a human-readable name, in lower case letters, with numbers as appropriate.
535 The underscore is allowed; space and special characters are not.
537 - a file in "config/arch/", named after the architecture's name, and suffixed
539 Eg.: config/arch/arm.in
540 - a file in "scripts/build/arch/", named after the architecture's name, and
542 Eg.: scripts/build/arch/arm.sh
544 The architecture's ".in" file API:
545 > the config option "ARCH_%arch%" (where %arch% is to be replaced with the
546 actual architecture name).
547 That config option must have *neither* a type, *nor* a prompt! Also, it can
548 *not* depend on any other config option (EXPERIMENTAL is managed as above).
552 defines a (terse) help entry for this architecture:
556 The ARM architecture.
558 selects adequate associated config options.
561 select ARCH_SUPPORTS_BOTH_ENDIAN
562 select ARCH_DEFAULT_LE
564 The ARM architecture.
566 > other target-specific options, at your discretion. Note however that to
567 avoid name-clashing, such options shall be prefixed with "ARCH_%arch%",
568 where %arch% is again replaced by the actual architecture name.
569 (Note: due to historical reasons, and lack of time to clean up the code,
570 I may have left some config options that do not completely conform to
571 this, as the architecture name was written all upper case. However, the
572 prefix is unique among architectures, and does not cause harm).
574 The architecture's ".sh" file API:
575 > the function "CT_DoArchTupleValues"
578 - all variables from the ".config" file,
579 - the two variables "target_endian_eb" and "target_endian_el" which are
580 the endianness suffixes
581 + return value: 0 upon success, !0 upon failure
584 - the environment variable CT_TARGET_ARCH
586 the architecture part of the target tuple.
587 Eg.: "armeb" for big endian ARM
591 - the environment variable CT_TARGET_SYS
593 the sytem part of the target tuple.
594 Eg.: "gnu" for glibc on most architectures
595 "gnueabi" for glibc on an ARM EABI
597 - for glibc-based toolchain: "gnu"
598 - for uClibc-based toolchain: "uclibc"
601 - the environment variable CT_KERNEL_ARCH
603 the architecture name as understandable by the Linux kernel build
605 Eg.: "arm" for an ARM
606 "powerpc" for a PowerPC
612 - the environment variables to configure the cross-gcc (defaults)
613 - CT_ARCH_WITH_ARCH : the gcc ./configure switch to select architecture level ( "--with-arch=${CT_ARCH_ARCH}" )
614 - CT_ARCH_WITH_ABI : the gcc ./configure switch to select ABI level ( "--with-abi=${CT_ARCH_ABI}" )
615 - CT_ARCH_WITH_CPU : the gcc ./configure switch to select CPU instruction set ( "--with-cpu=${CT_ARCH_CPU}" )
616 - CT_ARCH_WITH_TUNE : the gcc ./configure switch to select scheduling ( "--with-tune=${CT_ARCH_TUNE}" )
617 - CT_ARCH_WITH_FPU : the gcc ./configure switch to select FPU type ( "--with-fpu=${CT_ARCH_FPU}" )
618 - CT_ARCH_WITH_FLOAT : the gcc ./configure switch to select floating point arithmetics ( "--with-float=soft" or /empty/ )
621 - the environment variables to pass to the cross-gcc to build target binaries (defaults)
622 - CT_ARCH_ARCH_CFLAG : the gcc switch to select architecture level ( "-march=${CT_ARCH_ARCH}" )
623 - CT_ARCH_ABI_CFLAG : the gcc switch to select ABI level ( "-mabi=${CT_ARCH_ABI}" )
624 - CT_ARCH_CPU_CFLAG : the gcc switch to select CPU instruction set ( "-mcpu=${CT_ARCH_CPU}" )
625 - CT_ARCH_TUNE_CFLAG : the gcc switch to select scheduling ( "-mtune=${CT_ARCH_TUNE}" )
626 - CT_ARCH_FPU_CFLAG : the gcc switch to select FPU type ( "-mfpu=${CT_ARCH_FPU}" )
627 - CT_ARCH_FLOAT_CFLAG : the gcc switch to choose floating point arithmetics ( "-msoft-float" or /empty/ )
628 - CT_ARCH_ENDIAN_CFLAG : the gcc switch to choose big or little endian ( "-mbig-endian" or "-mlittle-endian" )
633 - the environement variables to configure the core and final compiler, specific to this architecture:
634 - CT_ARCH_CC_CORE_EXTRA_CONFIG : additional, architecture specific core gcc ./configure flags
635 - CT_ARCH_CC_EXTRA_CONFIG : additional, architecture specific final gcc ./configure flags
640 - the architecture-specific CFLAGS and LDFLAGS:
641 - CT_ARCH_TARGET_CLFAGS
642 - CT_ARCH_TARGET_LDFLAGS
646 You can have a look at "config/arch/arm.in" and "scripts/build/arch/arm.sh" for
647 a quite complete example of what an actual architecture description looks like.
652 A kernel is defined by:
654 - a human-readable name, in lower case letters, with numbers as appropriate.
655 The underscore is allowed; space and special characters are not (although
656 they are internally replaced with underscores.
657 Eg.: linux, bare-metal
658 - a file in "config/kernel/", named after the kernel name, and suffixed with
660 Eg.: config/kernel/linux.in, config/kernel/bare-metal.in
661 - a file in "scripts/build/kernel/", named after the kernel name, and suffixed
663 Eg.: scripts/build/kernel/linux.sh, scripts/build/kernel/bare-metal.sh
665 The kernel's ".in" file must contain:
666 > an optional lines containing exactly "# EXPERIMENTAL", starting on the
667 first column, and without any following space or other character.
668 If this line is present, then this kernel is considered EXPERIMENTAL,
669 and correct dependency on EXPERIMENTAL will be set.
671 > the config option "KERNEL_%kernel_name%" (where %kernel_name% is to be
672 replaced with the actual kernel name, with all special characters and
673 spaces replaced by underscores).
674 That config option must have *neither* a type, *nor* a prompt! Also, it can
675 *not* depends on EXPERIMENTAL.
676 Eg.: KERNEL_linux, KERNEL_bare_metal
678 defines a (terse) help entry for this kernel.
680 config KERNEL_bare_metal
682 Build a compiler for use without any kernel.
684 selects adequate associated config options.
686 config KERNEL_bare_metal
689 Build a compiler for use without any kernel.
691 > other kernel specific options, at your discretion. Note however that, to
692 avoid name-clashing, such options should be prefixed with
693 "KERNEL_%kernel_name%", where %kernel_name% is again tp be replaced with
694 the actual kernel name.
695 (Note: due to historical reasons, and lack of time to clean up the code,
696 I may have left some config options that do not completely conform to
697 this, as the kernel name was written all upper case. However, the prefix
698 is unique among kernels, and does not cause harm).
700 The kernel's ".sh" file API:
701 > is a bash script fragment
703 > defines the function CT_DoKernelTupleValues
704 + see the architecture's CT_DoArchTupleValues, except for:
705 + set the environment variable CT_TARGET_KERNEL, the kernel part of the
707 + return value: ignored
709 > defines the function "do_print_filename":
712 - all variables from the ".config" file,
713 + return value: ignored
714 + behavior: output the kernel's tarball filename, with adequate suffix,
716 Eg.: linux-2.6.26.5.tar.bz2
718 > defines the function "do_kernel_get":
721 - all variables from the ".config" file.
722 + return value: 0 for success, !0 for failure.
723 + behavior: download the kernel's sources, and store the tarball into
724 "${CT_TARBALLS_DIR}". To this end, a functions is available, that
725 abstracts downloading tarballs:
726 - CT_DoGet <tarball_base_name> <URL1 [URL...]>
727 Eg.: CT_DoGet linux-2.6.26.5 ftp://ftp.kernel.org/pub/linux/kernel/v2.6
728 Note: retrieving sources from svn, cvs, git and the likes is not supported
729 by CT_DoGet. You'll have to do this by hand, as it is done for eglibc in
730 "scripts/build/libc/eglibc.sh"
732 > defines the function "do_kernel_extract":
735 - all variables from the ".config" file,
736 + return value: 0 for success, !0 for failure.
737 + behavior: extract the kernel's tarball into "${CT_SRC_DIR}", and apply
738 required patches. To this end, a function is available, that abstracts
740 - CT_ExtractAndPatch <tarball_base_name>
741 Eg.: CT_ExtractAndPatch linux-2.6.26.5
743 > defines the function "do_kernel_headers":
746 - all variables from the ".config" file,
747 + return value: 0 for success, !0 for failure.
748 + behavior: install the kernel headers (if any) in "${CT_SYSROOT_DIR}/usr/include"
750 > defines any kernel-specific helper functions
751 These functions, if any, must be prefixed with "do_kernel_%CT_KERNEL%_",
752 where '%CT_KERNEL%' is to be replaced with the actual kernel name, to avoid
755 You can have a look at "config/kernel/linux.in" and "scripts/build/kernel/linux.sh"
756 as an example of what a complex kernel description looks like.
758 Adding a new version of a component |
759 ------------------------------------+
761 When a new component, such as the Linux kernel, gcc or any other is released,
762 adding the new version to crosstool-NG is quite easy. There is a script that
763 will do all that for you:
764 tools/addToolVersion.sh
766 Run it with no option to get some help.
771 To Be Written later...