test-suite: Added new test suite feature (experimental)

This patch adds support for installing the gcc test suite. A helper
Makefile is provided for building and running the gcc tests.

The default configuration runs all gcc tests and requires automatic
ssh/scp login access to a networked target board. See README for
more details.

Note: Current feature is tested with the powerpc-unknown-linux-gnu
sample but it should work with others as well.

Signed-off-by: Martin Lund <mgl@doredevelopment.dk>

scripts/wrapper.c: Under MacOS set DYLD_LIBRARY_PATH

Depending on (predefined) macro __APPLE__, use DYLD_LIBRARY_PATH
instead of LD_LIBRARY_PATH.

docs/overview.txt: Provide short mercurial HOWTO

The usage of hg mq is imho not very well documented.
Give a short intro for the most important use cases
for contributions to ct-ng.

yann.morin.1998@anciens.enib.fr:
Slightly rewrote the explanations for the introductory message.

ct-ng: Add ability to configure and use some GNU tools

Make the paths to libtool, objcopy, objdump, readelf, patch configurable.

scripts/build/internals.sh: Always use binary wrapper under BSD/MacOS

The shell wrapper script uses a nonportable call to readlink.
Thus, always use the binary wrapper under BSD/MacOS.

yann.morin.1998@anciens.enib.fr:
Use 'case' instead of 'if'.

scripts/wrapper.in: Add a comment about readlink not being portable.

Documented that the call to readlink will not work on BSD systems.

configure: make call to readlink portable to non-GNU (BSD) systems

The argument '-e' for readlink is a GNU extension.
In setting the variable 'where' the argument '-e' is not necessary, thus eliminated.

scripts/crosstool-NG.sh.in: On MacOS/BSD use the output of CT_DoConfigGuess for CT_BUILD.

On 64bit MacOS `gcc -dumpmachine` gives i686 for the host machine.
This conflicts with the expectations of some following configure scripts
that a 64bit x86 is given as x86_64; i686 is understood as a 32 bit machine.
config.guess sets the host machine in CT_BUILD correctly.

yann.morin.1998@anciens.enib.fr:
As suggested by Khem RAJ on the ML, always use config.guess.

configure: Do not require stat to be of GNU coreutils

calls to stat have been made portable, thus it is not necessary
that the GNU version is present on the user's system.

scripts/functions: Use stat correctly on non-GNU (BSD/Darwin) systems.

Call to get the directory mode depending on $CT_SYS_OS

yann.morin.1998@anciens.enib.fr:
CT_SYS_OS has changed on Linuxsystem, it only gets the kernel name "Linux",
and not the system name, 'GNU/'.

build/internals.sh: Replace nonportable call to stat

The call to stat to find out if a file is a symlink works only on GNU systems,
and the replacing portable call to readlink is also shorter and more concise code.

yann.morin.1998@anciens.enib.fr:
Apply simpler test, after discussion with author and Arnaud LACOMBE on the ML.

crosstool-NG.sh: suppress failure message from uname -o

On non-GNU systems (BSD/MacOS) there is no uname -o.
Suppress the failure message on these systems in the
call to set CT_SYS_OS (uname -s actually sets this variable).

yann.morin.1998@anciens.enib.fr: remove 'uname -o' altogether.

cc/gcc: add latest version

Add 4.4.4 and forward-port patchset from 4.4.3

kernel/linux: update versions list

Merge.

Revert changeset e7f7220b5ba2

Arnaud LACOMBE rightly pointed out that this was done on the wrong branch.

Revert changeset d341be24dc68

Arnaud LACOMBE rightly pointed out that this was done on the wrong branch.

cc/gcc: add latest version

Add 4.4.4 and forward-port patchset from 4.4.3

kernel/linux: update versions list

1.7: update version to 1.7.0+hg

Tagging release 1.7.0

1.7: create maintenance branch, update version to 1.7.0

scripts/mk-release: pretty-up the per-author stats

Reverse sort first by number of changesets, then sort alphabetically
on the author's names.

docs: update the known issues

samples: update

debug/gdb: install tic in the buildtools directory

'tic' is a build tool, its place is in the buildtools
directory, not in the toolchain's bin directory.

debug/gdb: be less verbose during the ncurses build

scripts: don't set LD_PRELOAD with static companion libraries

When cpmpanion libraries are built statically, it is useless to set
LD_PRELOAD. Worse, it might break the build in some corner cases.

docs: update the section about the wrapper

The wrapper is no longer needed when building
non-shared companion libraries. Document that.

arch/sh: fix type in variant handling

scripts/savesample: be silent when checking for .config file

1.6: update version to 1.6.2+hg

Tagging release 1.6.2

1.6: update version to 1.6.2

cc/gcc: fix cc symlink when executables have an extension

When building a cross-compiler for a host which depends
on file extensions the symlink for cc was not installed correctly

Signed-off-by: Bart vdr. Meulen <bartvdrmeulen@gmail.com>
[Yann E. MORIN: style fixes, enhancements, code prettying]
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@anciens.enib.fr>
(transplanted from 214c64ba8b21db15f610eca886fb194a6675e40c)

libc/newlib: don't do // downloads

The newlib FTP site does not like // downloads, and quickly
blocks the IP from the downloader.

Override the number of chunks when downloading newlib.

libc/glibc: fix kernel version check

Since we log the test strings, it is no longer possible to call
the CT_Test* functions from sub-functions that print results to
their stdout.

scripts: also prepare include symlink to sysroot/include

scripts: check g++ only for host

g++ is only needed when building additonal libs on the HOST,
so check wheter g++ is available for the HOST compiler only

Signed-off-by: Bart vdr. Meulen <bartvdrmeulen@gmail.com>
[Yann E. MORIN: fix space damage]
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@anciens.enib.fr>

cc/gcc: use proper host compiler for canadian bare-metal

While compiling a canadian toolchain for host=mingw32, build=linux,
target=m68k-elf the build fails because in this step of the gcc build
the Host compiler is used in this stage with the build-flags for the
build system. This results in an error where the header <sys/wait.h>
cannot be found.
This problem happens at least in the GCC-4.3.x and GCC-4.4.x range.

This is solved by passing the proper compilers on the Make cmd-line

Signed-off-by: Remy Bohmer <linux@bohmer.net>

cc/gcc: allow building a canadian bare-metal compiler

Previous addition of the canadian cross compiler did not allow
to build a baremetal only variant, no reason why this is not
allowed

Signed-off-by: Bart vdr. Meulen <bartvdrmeulen@gmail.com>

cc/gcc: fix cc symlink when executables have an extension

When building a cross-compiler for a host which depends
on file extensions the symlink for cc was not installed correctly

Signed-off-by: Bart vdr. Meulen <bartvdrmeulen@gmail.com>
[Yann E. MORIN: style fixes, enhancements, code prettying]
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@anciens.enib.fr>

libc/newlib: the m68k processor does not like unaligned

Disable unaligned access at least for mcpu32, m68010 and m68020.
These processors certainly do not like unaligned accesses.

Signed-off-by: Remy Bohmer <linux@bohmer.net>
(transplanted from faade25219ed565993b7b26357991d3130697515)

complibs/mpfr: fix extract

With MPFR 2.4.2, autoreconf needs to be run as well.
(transplanted from b376baa43a50e178915f8d5dac40d9408bf2e962)

cc/gcc: add 4.5.0

cc/gcc: add support for 4.5.0 or later versions

starting with 4.5.0, gcc requires libelf.

complibs/libelf: remove old versions

Those old versions are not suitable for use by gcc.

complibs/libelf: enable the host build

Host libelf is required by gcc-4.5.0

complibs/libelf: build target libelf in its own directory

debug/gdb: remove insight

Insight seems to be very slow to follow up on mainstreram gdb.
Latest snapshots are more than 6 months old.

Moreover, I don't have time to maintain insight support in crosstool-NG;
and, because I don't use it, I am unable to find any breakage.

debug/gdb: reorder options in the menuconfig

debug/gdb: allow static build with complibs

It is now possible to built static versions of the companion libraries,
so we can now build a static cross-gdb using the complibs.

scripts: don't print the build system guess warning

complibs/gmp: mark 5.0.1 as being experimental

MPFR can't cope with GMP > 4.x.y so hide it behind EXPERIMENTAL
until we fix MPFR

samples: add Blackfin sample

Add basic support for the Blackfin architecture

For uClibc, the name of the Blackfin architecture is 'bfin'. Actually,
the naming of the architecture is quite messy: for toolchain tuples
and uClibc, it's bfin, but for the kernel, it's blackfin. We've
arbitraly choosen to name it "blackfin" in Crosstool-NG.

Add Blackfin-related uClibc patch to fix a build failure related to
fork() being used in unistd/daemon.c.

Yann E. MORIN:
Apply the patch to the kernel/linux build script to use 'linux'
in the noMMU tuples. See:
http://sourceware.org/ml/crossgcc/2010-04/msg00010.html

scripts/functions: log the test strings

scripts/functions: compress with gzip level 3, as stated in the documentation

Thanks to Bruno Tarquini <btarquini@gmail.com> for the report.

scripts/functions: add a save/restore handler

Saving and restoring the steps requires saving/restoring multiple
directories. Depending on the configuration, some may not exist.

Add a wrapper that checks before creating/extracting the tarballs.

scripts/functions: fix CT_Patch again

Merge.

Revert 3ea4e6f7b85f: Force autoreconf for mpfr

It breaks on my machine. Revert until we find the exact cause leading to
the initial patch, and we come up with a patch that properly fixes it.

scripts/functions: fix CT_Patch

libc/newlib: the m68k processor does not like unaligned

Disable unaligned access at least for mcpu32, m68010 and m68020.
These processors certainly do not like unaligned accesses.

Signed-off-by: Remy Bohmer <linux@bohmer.net>

scripts/functions: make CT_Patch dumber

It is the responsibility of the caller to split the package name from
its version. It already knows that.

libc/newlib: only compute version string, not full filename

debug/gdb: remove snapshot version

cc/gcc: add bugurl and compiler version to core gcc compiler

When building for bare-metal the core-gcc compiler is delivered
as final compiler, so the version info and bugurl is useful
in the core compiler as well.

Signed-off-by: Remy Bohmer <linux@bohmer.net>

binutils: allow choice for elf format when building bare-metal cross compiler

When building a baremetal cross compiler I want to be able to select
the elf format and not be forced to build the elf2flt package.

Signed-off-by: Bart vdr. Meulen <bartvdrmeulen@gmail.com>
[Yann E. MORIN: tweak the commit message]
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@anciens.enib.fr>

scripts: prevent trailing - in CT_TARGET

Not all target tuples consist of an VENDOR, KERNEL and SYSTEM part, build up the
tuple in such a way to no extra or trailing dashes are added to CT_TARGET

Signed-off-by: Bart vdr Meulen <bartvdrmeulen@gmail.com>

[CT-NG:patch 05/19] Force autoreconf for mpfr

In some exotic case the autoreconf step of mpfr is not executed (correctly)
leaving an incorrect version number for libtool in the configure script.

After extracting the sources files, force autoreconf to be executed.

Signed-off-by: Bart vdr. Meulen <bartvdrmeulen@gmail.com>

complibs: save directory with static complibs for stop/restart

complibs: don't install in toolchain dir when built as static

complibs: allow either static or shared build

complibs: build static libraries

config: replace COMP_LIBS with COMPLIBS

complibs/mpfr: fix extract

With MPFR 2.4.2, autoreconf needs to be run as well.

scripts: add script to publish patches

This script creates the patches for an existing version.

Re-enable the FD_PIC ELF configuration option

Blackfin uses FD_PIC ELF format to provide shared libraries on a
non-MMU architecture.

Fix the uClibc clean-up-O_CLOEXEC-handling patch

In the Blackfin case (untested until now), this patch was lacking a
leading / at the end of a comment.

libc/glibc: add patch to ports-2.9 to fix sjlj from ARM FPU

As discussed there: http://sourceware.org/bugzilla/show_bug.cgi?id=9678#c4

Reported by: Sasha Sirotkin <buildroot@browserseal.com>
(transplanted from dfc2b3e07ea6ea0761fb91df7c3c7fcdc718cc5f)

scripts/populate: fix handling the forced libraries list-file

If a list-file is used, then each library in the file will be handled
twice (not a real issue, as the second iteration will find the library
already present, just avoid doing the job twice).
(transplanted from 7fec5db8933cb296c665b8526f27a779aa0c0c7a)

cc/gcc: fix building for powerpc for the 4.4.x series

Reported by: Thomas Petazzoni
http://sourceware.org/ml/crossgcc/2010-03/msg00057.html

Discussions:
http://lists.uclibc.org/pipermail/uclibc/2010-March/043697.html
http://lists.uclibc.org/pipermail/buildroot/2010-March/032790.html
http://ibot.rikers.org/%23uclibc/20100316.html.gz
(search for "copysignl", by kos_tom, blindvt and y_morin)
(transplanted from 76b1923da63d2e3b8e9010b6138e33f6a99720a8)

libc/uclibc: add uClibc 0.9.31

Signed-off-by: Arnaud Lacombe <lacombar@gmail.com>
[Yann E. MORIN: mark it EXPERIMENTAL]
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@anciens.enib.fr>

cc/gcc: fix thread-less final gcc build

If threads are disabled in libc, we don't want to enable them in the
final compiler. Doing so pass the configure stage, but fails latter on
a missing <pthread.h>.

Moreover, we don't want to build libgomp if threads are disabled; its
configure script would fails anyway.

Signed-off-by: Arnaud Lacombe <lacombar@gmail.com>

cc/gcc: fix recent gcc build

This fix missed conversion of CT_GCC_USE_* to CT_CC_GCC_USE_*.

Signed-off-by: Arnaud Lacombe <lacombar@gmail.com>

complibs/gmp: fix patch against 5.0.1

config: allow libc selection in backend mode

In backend mode, allow the upper-layer build system to force the
C library selection.

config: do not force setting kernel and arch

In backend mode, only enforce the arch and/or kernel selection if the
upper-layer build system effectively forces the selection.

scripts: misc help and auto-complete fixes

- don't list samples in the main help screen
- improve the samples listing in list-samples
- don't document the 'config' action, it's long dead
- document the 'V' environment variable
- improve on START, STOP and PREFIX environment variables
- add PREFIX and V to autocomplete
- advertise auto-complete at install time

kernel/linux: add latest versions: 2.6.33.2 2.6.32.11 2.6.31.13 2.6.27.46

debug/gdb: update version to 7.1

complibs: update versions

GMP : 4.3.2, 5.0.1*
CLooG/PPL: 0.15.8 0.15.9
libelf : 0.8.13

*: patch series added

scripts/addToolsVersion: fix handling libelf

libelf has moved to the companion libraries section,
and is no longer in the tools section.

libc/glibc: add patch to ports-2.9 to fix sjlj from ARM FPU

As discussed there: http://sourceware.org/bugzilla/show_bug.cgi?id=9678#c4

Reported by: Sasha Sirotkin <buildroot@browserseal.com>

config: fix style issues when source-ing files

Always enclose the sourced file between double-quotes (purely for
consistency; enclosing or not are both allowed by the kconfig language).

config: move backend-related options to their own file

config: hide arch and kernel selection when used as a backend

When used as a backend, it is the responsibility of the upper-layer build
system to set the target architecture and kernel.

config: hide options to debug crosstool-NG when a backend

When used as a backend, it does not make sense to debug crosstool-NG,
so hide the option.

config: hide more config options when a backend

Hide the prompts for some behavioral options, for which the upper-layer build
system is responsible for:
- parallel jobs and maximum load
- use pipes
- use custom shell

config: hide paths when used as a backend

When crosstool-NG is used as a backend, it is the responsibility of the
upper-layer build-system to properly set paths, so we just hide the
prompts in this case.

scripts/populate: optimise search loop

Curently, populate will iterate over all ELF (shared objects|executables)
to look for missing NEEDED DSOs, adding to the list at every iterations
of the search loop.

Instead of looking again at previously handled ELF files, recursively
resolve every ELf files.

Also, in case there are a whole lot of files (more than the shell can
accept as arguments list, or creating a command line longer than the
shell can cope with), use a temporary file with the list of files
to search for missing dependencies.

scripts/populate: cleanups and misc fixes

- it's a POSIX compliant shell script: drop bash, use /bin/sh
- fix help text
- use an absolute path for sysroot
- replace "echo" with "printf"
- replace "stat -c '%i'" with "ls -1id"
- replace "pushd / popd" with "cd / cd -"
- remove superfluous break
- bail out if required lib not found, except if forced

scripts/populate: fix handling the forced libraries list-file

If a list-file is used, then each library in the file will be handled
twice (not a real issue, as the second iteration will find the library
already present, just avoid doing the job twice).

scripts/populate: properly locate the sysroot and required tools

This fixes two problems:
- the sysroot might be in a sub-directory (think SYSROOT_DIR_PREFIX)
- it is not needed to have the target tuple to properly detect the sysroot
and the required tools

As a side effect, this script is now no longer dependent on the target
tuple, and in the future, we might be able to share it across many
toolchains (when/if we can install all of them in the same place).

scripts/populate: add -m option, to merge source and dest

Merge the source rootfs into the (pre-existing) destination rootfs,
and populate the result accordingly.

scripts/populate: add option to use an alternate sysroot

Add a new command line option, "-r", which allows the user to specify
an alternate sysroot location to copy libraries from. This is useful
when using the toolchain in combination with a separate root filesystem,
or when working with multiple different root filesystems.

Signed-off-by: Nate Case <ncase@xes-inc.com>

binutils/sstrip: fix when sstrip is disabled

cc/gcc: fix building for powerpc for the 4.4.x series

Reported by: Thomas Petazzoni
http://sourceware.org/ml/crossgcc/2010-03/msg00057.html

Discussions:
http://lists.uclibc.org/pipermail/uclibc/2010-March/043697.html
http://lists.uclibc.org/pipermail/buildroot/2010-March/032790.html
http://ibot.rikers.org/%23uclibc/20100316.html.gz
(search for "copysignl", by kos_tom, blindvt and y_morin)

binutils/sstrip: get rid of ELFkickers

ELFkickers are looong dead and unmaintained, and the sstrip
from buildroot is working fine *and* is maintained.

binutils/sstrip: depends on ELF

No point building sstrip is the target is not ELF.

tools: move sstrip to the binary utilities menu

sstrip is now alone in its 'tools' menu, and we will probably never gain
any other 'tool'. Besides, sstrip is just strip, but a little bit more
agressive, so it deserves going to the 'binary utilities' menu.

debug/gdb: gdb>=7.0 can use MPC, enable for cross-gdb

debug/gdb: add handling of 7_0_or_later

We'll need to differentiate gdb 7.0 or later, from gdb prior to 7.0

debug/gdb: add option to use GMP and MPFR

GMP and MPFR are optional for gdb, so offer a config option.

debug/gdb: can't build a static native gdb if using GMP/MPFR

It is currently impossible to reliably build static GMP/MPFR libs,
so we can't build a static native gdb using them, obviously.

debug/gdb: split menuconfig for easier maintenance

cc/gcc: fix use of companion libraries

complibs: hide prompts

scripts: fix dumping the config for companion libraries for target

Companion libraries can be disable for the host, but still used for
the target (eg. binutils or gdb). Dump the version also in this case.