- Add support for mips32r6 and mips64r6 target architecture variants
- Disable unsupported gcc versions
- Disable unsupported binutils versions
- Disable unsupported external toolchains
- Disable unsuported C libraries
- Add a hook in order to make glibc compile for MIPS R6.
[Thomas: slightly tweak the glibc hack explanation, to make it
hopefully clearer.]
Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Fixes#8386
We should check if BR_CROSS_PATH_ABS is defined, not if it evalutates to
true for the pre processor.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Those two comments:
- are exactly the same
- have the same dependencies (except for arm/armeb)
So, make it a common comment. It will be useful to have that comment
when we introduce new Linaro toolchain versions.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
When building in a different output directory than the original build,
there will currently be a lot of ccache misses because in many cases
there is some -I/... absolute path in the compilation. Ccache has an
option CCACHE_BASEDIR to substitute absolute paths with relative paths,
so they wil be the same in the hash (and in the output).
Since there are some disadvantages to this path rewriting, it is made
optional as BR2_CCACHE_USE_BASEDIR. It defaults to y because the
usefulness of ccache is severely reduced without this option.
In addition to CCACHE_BASEDIR, we also substitute away the occurences
of $(HOST_DIR) in the calculation of the compiler hash. This is done
regardless of the setting of BR2_CCACHE_USE_BASEDIR because it's
quite harmless.
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Our current ccache disables hashing of the compiler executable itself,
because using the default 'mtime' doesn't work in buildroot: we always
rebuild the compiler, so the mtime is always different, so the cache
always misses.
However, in the current situation, if a user changes the compiler
configuration (which would result in the compiler generating different
object files than before) and does 'make clean all', ccache may in fact
reuse object files from the previous run. This rarely gives problems,
because
(1) the cache expires quite quickly (it's only 1GB by default),
(2) radically changing compiler options will cause cache misses because
different header files are used,
(3) many compiler changes (e.g. changing -mtune) have little practical
effect because the resulting code is usually still compatible,
(4) we currently don't use CCACHE_BASEDIR, and almost all object files
will contain an absolute path (e.g. in debug info), so when
building in a different directory, most of it will miss,
(5) we do mostly build test, and many of the potential problems only
appear at runtime.
Still, when ccache _does_ use the wrong cached object files, the
effects are really weird and hard to debug. Also, we want reproducible
builds and obviously the above makes builds non-reproducible. So we
have a FAQ entry that warns against using ccache and tells the user to
clear the cache in case of problems.
Now that ccache is called from the toolchain wrapper, it is in fact
possible to at least use the 'mtime' compiler hash for the external
toolchain and for the host-gcc. Indeed, in this case, the compiler
executable comes from a tarball so the mtime will be a good reference
for its state. Therefore, the patch (sed script) that changes the
default from 'mtime' to 'none' is removed.
For the internal toolchain, we can do better by providing a hash of
the relevant toolchain options. We are only interested in things that
affect the compiler itself, because ccache also processes the header
files and it doesn't look at libraries because it doesn't cache the
link step, just compilation. Everything that affects the compiler
itself can nicely be summarised in $(HOST_GCC_FINAL_CONF_OPTS). Of
course, also the compiler source itself is relevant, so the source
tarball and all the patches are included in the hash. For this purpose,
a new HOST_GCC_XTENSA_OVERLAY_TAR is introduced.
The following procedure tests the ccache behaviour:
Use this defconfig:
BR2_arm=y
BR2_CCACHE=y
make
readelf -A output/build/uclibc-1.0.6/libc/signal/signal.os
-> Tag_CPU_name: "ARM926EJ-S"
Now make menuconfig, change variant into BR2_cortex_a9
make clean; make
readelf -A output/build/uclibc-1.0.6/libc/signal/signal.os
-> Tag_CPU_name: "ARM926EJ-S"
should be "Cortex-A9"
After this commit, it is "Cortex-A9".
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Cc: Danomi Manchego <danomimanchego123@gmail.com>
Cc: Károly Kasza <kaszak@gmail.com>
Cc: Samuel Martin <s.martin49@gmail.com>
Cc: Romain Naour <romain.naour@openwide.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
By moving the ccache call to the toolchain wrapper, the following
scenario no longer works:
make foo-dirclean all BR2_CCACHE=
That's a sometimes useful call to check if some failure is perhaps
caused by ccache.
We can enable this scenario again by exporting BR_NO_CCACHE when
BR2_CCACHE is not set, and by handling this in the toolchain wrapper.
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Tested-by: Romain Naour <romain.naour@openwide.fr>
Reviewed-by: Romain Naour <romain.naour@openwide.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Since we always have a toolchain wrapper now, we can move the ccache
call to the toolchain wrapper.
The hostcc ccache handling obviously stays.
The global addition of ccache to TARGET_CC/CXX is removed, but many
individual packages and infras still add it. This means we have a
chain like this: ccache -> toolchain-wrapper -> ccache -> gcc
However, this is fairly harmless: for cache misses, the inner ccache
just adds overhead and for cache hits, the inner ccache is never
called. Later patches will remove these redundant ccache calls.
As a side effect, perl now supports ccache as well.
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Cc: Danomi Manchego <danomimanchego123@gmail.com>
Cc: Károly Kasza <kaszak@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
We have a toolchain wrapper for external toolchain, but it is also
beneficial for internal toolchains, for the following reasons:
1. It can make sure that BR2_TARGET_OPTIMIZATION is passed to the
compiler even if a package's build system doesn't honor CFLAGS.
2. It allows us to do the unsafe path check (i.e. -I/usr/include)
without patching gcc.
3. It makes it simpler to implement building each package with a
separate staging directory (per-package staging).
4. It makes it simpler to implement a compiler hash check for ccache.
The wrapper is reused from the external toolchain. A third CROSS_PATH_
option is added to the wrapper: in this case, the real executable is in
the same directory, with the extension .real.
The creation of the simple symlinks is merged with the creation of the
wrapper symlinks, otherwise part of the -gcc-ar handling logic would
have to be repeated.
The complex case-condition could be refactored with the one for the
external toolchain, but then it becomes even more complex because
they each have special corner cases. For example, the internal
toolchain has to handle *.real to avoid creating an extra indirection
after host-gcc-{final,initial}-rebuild.
Instead of creating the .real files, it would also have been possible
to install the internal toolchain in $(HOST_DIR)/opt, similar to what
we do for the external toolchain. However, then we would also have to
copy things to the sysroot and do more of the magic that the external
toolchain is doing. So keeping it in $(HOST_DIR)/usr/bin is much
simpler.
Note that gcc-initial has to be wrapped as well, because it is used for
building libc and we want to apply the same magic when building libc.
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Cc: Fabio Porcedda <fabio.porcedda@gmail.com>
Cc: Jérôme Oufella <jerome.oufella@savoirfairelinux.com>
Reviewed-by: Romain Naour <romain.naour@openwide.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The toolchain wrapper will be reused for the internal toolchain, so it
belongs in the toolchain directory. Also, the ext- prefix is removed
from it. The build commands are moved to a new toolchain-wrapper.mk.
The wrapper arguments that are also relevant for the internal toolchain
wrapper are moved to toolchain-wrapper.mk, the rest stays in
toolchain-external.mk.
While we're at it, move the building of the toolchain wrapper to the
build step of toolchain-external. There is no specific reason to do
this, other than that it fits better semantically. Also remove the
MESSAGE call, otherwise we'd see:
>>> toolchain-external undefined Building
>>> toolchain-external undefined Building toolchain wrapper
/usr/bin/gcc ...
Having an extra "Building toolchain wrapper' message is pointless.
The useless condition on $(BR2_TARGET_OPTIMIZATION) is removed. It was
always true because it wasn't qstrip'ped first, so clearly it works
without that condition as well.
Also rewrapped some comments and removed the 'external' reference.
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Cc: Fabio Porcedda <fabio.porcedda@gmail.com>
Cc: Jérôme Oufella <jerome.oufella@savoirfairelinux.com>
Reviewed-by: Romain Naour <romain.naour@openwide.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
For some external toolchain vendors the actual source code URL can be simply
derived from the binary file URL.
Here we obtain TOOLCHAIN_EXTERNAL_ACTUAL_SOURCE_TARBALL for all Mentor and
Linaro toolchains with a few $(subst) calls.
Signed-off-by: Luca Ceresoli <luca@lucaceresoli.net>
Cc: Thomas De Schampheleire <patrickdepinguin@gmail.com>
Cc: Arnout Vandecappelle <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Trailing slashes are going to be declared illegal from FOO_SITE
variables.
But Buildroot internally generates such a variable when using a custom
external toolchain (i.e. BR2_TOOLCHAIN_EXTERNAL_CUSTOM). This is
because TOOLCHAIN_EXTERNAL_SITE is set to
$(dir $(call qstrip,$(BR2_TOOLCHAIN_EXTERNAL_URL))), and $(dir)
leaves a trailing slash.
Fix it using patsubst, just like linux and the bootloaders do.
Signed-off-by: Luca Ceresoli <luca@lucaceresoli.net>
Reported-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Arnout Vandecappelle <arnout@mind.be>
Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Introduced by previous patch 0f75b2635e,
this printf would break the build of glibc, because there is no format
to printf:
printf: usage: printf [-v var] format [arguments]
Signed-off-by Maxime Hadjinlian <maxime.hadjinlian@gmail.com>
Reported-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
'echo -n' is not a POSIX construct (no flag support), we shoud use
'printf', especially in init script.
This patch was generated by the following command line:
git grep -l 'echo -n' -- `git ls-files | grep -v 'patch'` | xargs sed -i 's/echo -n/printf/'
Signed-off-by: Maxime Hadjinlian <maxime.hadjinlian@gmail.com>
Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
When ARCH is arm and the hard-floating-point option is on executables
expect to find the dynamic linker at /lib/ld-musl-armhf.so.1 and not
/lib/ld-musl-arm.so.1.
This patch adjusts the logic that creates the symbolic link from the
dynamic linker path to the musl C library (since musl has everything
built into a single file).
[Thomas: tweak the commit log.]
Signed-off-by: Vicente Bergas <vicencb@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Before this commit, the output of the toolchain-external build steps
looked like this (abbreviated for clarity):
>>> toolchain-external undefined Building
>>> toolchain-external undefined Installing to staging directory
>>> toolchain-external undefined Copying external toolchain sysroot to staging...
>>> toolchain-external undefined Building ext-toolchain wrapper
mkdir -p output/host/usr/bin; cd output/host/usr/bin; for i in ...
/usr/bin/gcc -O2 -Ioutput/host/usr/include -DBR_SYSROOT='...
if test -f output/host/usr/bin/i686-pc-linux-gnu-gdb ; then mkdir -p ...
>>> toolchain-external undefined Fixing libtool files
>>> toolchain-external undefined Installing to target
>>> toolchain-external undefined Copying external toolchain libraries to target...
if test -e output/target/lib/ld-uClibc.so.1; then ln -sf ld-uClibc.so.1 output/target/lib/ld-uClibc.so.0 ; fi
if test -e output/target/lib/ld64-uClibc.so.1; then ln -sf ld64-uClibc.so.1 output/target/lib/ld64-uClibc.so.0 ; fi
All the long lines with conditions and loops in them are not usefull,
so put $(Q) in front of them. The line with mkdir can better be split
on a separate line so the cd stands out more. There are two redundant
semicolons that can be removed. The installation of gdbinit could
use an extra message so the user can see what is going on.
After this commit, the toolchain-external build steps look like this:
>>> toolchain-external undefined Building
>>> toolchain-external undefined Installing to staging directory
>>> toolchain-external undefined Copying external toolchain sysroot to staging...
>>> toolchain-external undefined Building ext-toolchain wrapper
/usr/bin/gcc -O2 -Ioutput/host/usr/include -DBR_SYSROOT='...
>>> toolchain-external undefined Installing gdbinit
>>> toolchain-external undefined Fixing libtool files
>>> toolchain-external undefined Installing to target
>>> toolchain-external undefined Copying external toolchain libraries to target...
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Commit c68c365d29 ("toolchain-external:
remove CS sh2 toolchains") removed the definitions of the
BR2_TOOLCHAIN_EXTERNAL_CODESOURCERY_SH2A_201103 and
BR2_TOOLCHAIN_EXTERNAL_CODESOURCERY_SH2A_201009, but did not actually
remove the code that was using those options.
So this commit removes the parts of the code that are currently dead
due to this: the definition of the prefix of those toolchains, the
hashes, and the URLs.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Currently, when a preconfigured prebuilt toolchain forgets to specify
its gcc version, the error message is a bit misleading, like:
Incorrect selection of gcc version: expected .x, got 4.9.2
Add a an explicit check for the gcc version being set, that reports a
better error message.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
gcc will always report a three-digit version sting, like 4.9.3 or 5.1.0.
For gcc before 5, we want to check the first two digits, while starting
with gcc 5, we are only concerned about the first digit.
So, change our matching code to test for the leading part of the version
string, up to the first dot after as-many version digit we're interested
in.
Note: we're adding the dot in the .mk code rather than in the Kconfig
symbol, because it seemed cleaner to do so.
Reported-by: Jörg Krause <joerg.krause@embedded.rocks>
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This commit wires up the gcc version dependency mechanism in the
external toolchain backend. To do so, it:
* Changes the definition of all pre-defined external toolchain
profiles to select the appropriate BR2_TOOLCHAIN_GCC_AT_LEAST_*
option.
* For custom external toolchains, provides a visible Config.in
"choice" to select the gcc version used in the external toolchain.
* Adds a new check_gcc_version function, that verifies that the real
gcc version found in the external toolchain matches the one
declared in the Buildroot configuration.
[Thomas: use better sed expression proposed by Yann E. Morin, which
works with more cases.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This commit adds a number of hidden Config.in options, that will be
used to handle dependencies on the gcc version. We mimic the model
that was used for the kernel headers dependency mechanism.
These hidden options will be selected by the internal and external
toolchain backend logic respectively, in follow-up commits.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Before commit 5715d2dcf4, the external
toolchain wrapper would not pass its own march/mcpu/mtune flags to the real
compiler if at least one of them was passed on the wrapper command-line.
The mentioned commit intended to remove the passing of an mtune parameter
coming from Buildroot, which was always empty after some other refactoring,
but the changes have the side-effect that march/mcpu is now also passed when
mtune is already given on the command-line. In that case, only mtune should
be passed to the real compiler.
Restore part of the original toolchain wrapper code to check the presence of
mtune on the command-line.
Signed-off-by: Thomas De Schampheleire <thomas.de.schampheleire@gmail.com>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Commit 34f95bf9db (toolchain-external: fix support of uClibc-ng toolchains,
2015-07-13) added the missing ld-uClibc.so.1 dynamic linker symlink that
binaries expect when linked with uClibc-ng. However on 64bit targets the
linker is called ld64-uClibc.so.1. Handle that case as well.
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The uClibc-ng dynamic loader is called ld-uClibc.so.1, but gcc is not
patched specifically for uClibc-ng, so it continues to generate
binaries that expect the dynamic loader to be named ld-uClibc.so.0,
like with the original uClibc.
Therefore, when a uClibc-ng toolchain is used as an external
toolchain, we need to create an additional symbolic link to make
uClibc-ng systems work properly.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The toolchain from the Cavium Networks Octeon SDK provides a sysroot
with library directories lib32, lib32-fp, lib64 and lib64-fp. The -fp
variants are used for processors with hardware floating point unit, such
as the Octeon III variants.
When specifying -march=octeon3 in BR2_TARGET_OPTIMIZATION, the toolchain
will use lib32-fp, but currently Buildroot does not accept that pattern.
This patch improves the matching by accepting lib(32|64)?([^/]*)? as lib
directory name.
Signed-off-by: Bai Yingjie <byj.tea@gmail.com>
[ThomasDS: update commit message]
Signed-off-by: Thomas De Schampheleire <thomas.de.schampheleire@gmail.com>
[Thomas: add comment above the function being modified to illustrate
the various cases we try to handle.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This makes sure we don't have any weird permissions on the staging dir,
which could affect the target.
Signed-off-by: Guido Martínez <guido@vanguardiasur.com.ar>
Acked-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Add aarch64_be support. Note that CONFIG_CPU_BIG_ENDIAN should be
defined in kernel config when building a big endian kernel.
Signed-off-by: Zhang Jian(Bamvor) <bamvor.zhangjian@huawei.com>
Acked-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
check_arm_abi builds a test C file to check that the toolchain is
working correctly, with the output redirected to /dev/null.
However, some toolchains (OSELAS 2014.12.0, for instance) foolishly
append ".gdb" to the output filename for an intermediate file, causing
an attempt to write to /dev/null.gdb, which obviously fails.
Fix this by adding changing the output to a temporary file, which is
later removed along with any other "suffixed" files.
Suggested-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Guido Martínez <guido@vanguardiasur.com.ar>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
eglibc is a dead project and has not been making any release since a
while, now that glibc is back and kicking. So let's deprecated our
eglibc support.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Instead of blacklisting which architectures support MMUs (mandatorily
or optionally), introduce two Kconfig options that are selected by each
architecture in each case.
This simplifies the logic in BR2_USE_MMU.
Signed-off-by: Guido Martínez <guido@vanguardiasur.com.ar>
Acked-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Currently, we instruct users to enable/disable BR2_TOOLCHAIN_HAS_NATIVE_RPC
but that is a blind option. The only option users can set/unset is
BR2_TOOLCHAIN_EXTERNAL_INET_RPC.
Use that in the error message.
Notes: the only way for this message to appear is for a custom external
toolchain, either downloaded or pre-installed, so even though we check
the validity of the toolchain with BR2_TOOLCHAIN_HAS_NATIVE_RPC, we do
report on BR2_TOOLCHAIN_EXTERNAL_INET_RPC.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Acked-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Like we do for the internal musl backend. We still see a large number of
build failures with musl, so warn users about it.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The Linux kernel does force compile with -m4-nofpu, which is only
available when building a multilib toolchain.
The interesting part here is, that buildroot use --disable-multilib for
gcc configure, but enables --with-multilib-list=m4,m4-nofpu in
the default configuration for Qemu targeting r2d emulation.
This results in a toolchain, which can be used for the kernel and
for userland without creating a multilib toolchain with different
kinds of libgcc version. In the multilib case there would be
subdirectories created (!m4 and m4-nofpu). As buildroot uses a
short version of toolchain creation, a multilib enabled gcc build
fails when creating libgcc.
So the best solution is to just keep multilib disabled, but always
add --with-multilib-list when sh4/sh4eb/sh4a/sh4aeb is choosen.
Tested with sh4/sh4a toolchain build and qemu defconfig with
gcc 4.8.x/4.9.x (with and without C++ enabled), uClibc and glibc.
Disable sh4a/sh4aeb for uClibc, as it does not implemented, yet.
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Tested-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
(ARM and SH4 uClibc toolchain builds)
Since 1.1.6, the mips softfloat toolchains are merged into the mips
toolchain using multilib. Our external toolchain infrastructure copies
the correct version to the target depending on the BR2_SOFT_FLOAT
option.
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Add hashes for all musl toolchains, including the ones that we
currently don't support (arm hf, sh4, x86_64-x32).
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
For Linaro toolchains, a special post install staging hook is used to
create two symlinks needed for the dynamic loader to find the
libraries. However, the way the link is created prevents a 'make
toolchain-external-reinstall' from succeeding, because the symlink
already exists and points to a directory:
ln -sf . /home/thomas/projets/outputs/training/target/lib/arm-linux-gnueabihf
ln: '/home/thomas/projets/outputs/training/target/lib/arm-linux-gnueabihf/.': cannot overwrite directory
This commit adjust the hook to pass the '-n' option so that the link
name is treated as a normal file if it is a symbolic link to a
directory.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
This uClibc toolchain does not provide an appropriate uClibc
configuration for Buildroot: missing IPv6, missing nsl stub, missing
program invocation, etc. Therefore, we mark it as broken, waiting for
a new upstream release of a new toolchain.
We keep around the toolchain-external Synopsys code anyway, since it
will most likely be identical for the new toolchain version. However,
we remove all the quirks that were introduced to start work around
issues related to this toolchain.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
We will *always* be missing a hash file for custom external toolchains
that are downloaded.
So, just ignore that failure.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Gustavo Zacarias <gustavo@zacarias.com.ar>
Acked-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Acked-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The 1.1.6 version of musl-cross fixes the two issues that had been
preventing versions after 1.1.1 being used by buildroot, namely:
- sysroot is enabled again
- kernel headers are included again
Signed-off-by: Will Wagner <will_wagner@carallon.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
It's no longer used so farewell.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Remove BR2_INET_IPV6 select for predefined external toolchains.
Remove the (non)IPv6 option prompt since it's now mandatory.
And force the toolchain check now that internal uclibc is always built
with IPv6 support and external non-IPv6 toolchains are disallowed.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Normally we'd deprecate them, but:
1) They don't support IPv6 and it's being removed so it makes no sense.
2) They're based on uClibc 0.9.30-ish which is very old and surely has
package build breakage all over it.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
For a multi-arch toolchain, gconv modules are in a sub-directory named
after the machine gcc targets. This is the case, for example, for the
Linaro ARM 2014.09 toolchain, which has the gconv modules in (relative
to the sysroot):
/usr/lib/arm-linux-gnueabihf/gconv
while the Sourcery CodeBench ARM 2014.05 (non-multi-arch) has them in:
/usr/lib/gconv
So, to catter for both cases, search both paths. We want to favour the
machine-specific gconv modules over potentially existing "generic" ones,
so we first search that (if it exists) and fallback to looking in the
generic location.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
It's required in some 32-bit architectures for the extended (64-bit)
atomic operations, like __sync_add_and_fetch_8.
These arches are at least: i386, mips & mipsel.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Distro toolchains, i.ie. toolchains coing with distributions, will
almost invariably be unsuitable for use with Buildroot:
- they are mostly non-relocatable;
- their sysroot is tainted with a lot of extra libraries.
Especially, the toolchains coming with Ubuntu (really, all the Debian
familly of distros) are configured with --sysroot=/ which makes them
non-relocatable, and they already contain quite some libraries that
conflict (in any combination of version, API or ABI) with what Buildroot
wants to build (i.e. extra libraries, some not even present in
Buildroot...) but also their mere preence when Buildroot does not expect
them to be already built (so that a package would enable features when
it should not).
So, try to detect those toolchains and black-list them; inform the user
that the toolchain is unusable for the reasons mentioned above.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Arnout Vandecappelle <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
It's now unused so remove it.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
There's no need for toolchains or the user to declare largefile support
since it's now mandatory.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This will allow us to remove largefile handling in the tree without
breaking things while doing so.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
As discussed on the mailing list drop the non-largefile option for
toolchains.
The size delta is minimal and it just complicates package dependencies.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
We currently only check that the Buildroot configuration matches what is
available in the toolchain.
Since we're going to remove the check for LFS and make it a mandatory
feature, we will lose the corresponding buildroot option, so we won't be
able to use check_uclibc_feature as-is.
Introduce a magic value passed as the buildroot option name to recognise
checks for mandatory uclibc options that do not have a corresponding
option in buildroot.
If the buildroot option name is empty then the check is against a
mandatory uclibc option.
If a mandatory uclibc option is missing we reject the toolchain as being
unusable by buildroot.
[Thomas: minor tweaks in comment, remove space instead of tab.]
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The SuperH architecture is supported by the musl libc since some time
now, so let's enable it.
Tested via qemu_sh4_r2d_defconfig.
Signed-off-by: Lionel Orry <lionel.orry@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
musl 1.1.7 brings in experimental aarch64 support so enable it.
Tested via qemu_aarch64_virt_defconfig.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The internal toolchain was a "best effort" approach - we strived to make
it build properly and all but it's mostly untested.
Since it's got issues disable it until it's properly fixed and tested
and leave the official ADI toolchain instead.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
With this change we add pre-built external toolachins for DesignWare ARC
cores. All currently existed flavours are supported:
* ARC 700 and ARC HS cores
* Little- and big-endian configurations
These pre-built tools are built with build scripts available here
(https://github.com/foss-for-synopsys-dwc-arc-processors/toolchain/tree/arc-2014.12)
and correspond to arc-2014.12 release of sources.
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Anton Kolesov <akolesov@synopsys.com>
Cc: "Yann E. MORIN" <yann.morin.1998@free.fr>
The computation of TOOLCHAIN_EXTERNAL_BIN has a special case for
Blackfin, where it's set to
$(TOOLCHAIN_EXTERNAL_INSTALL_DIR)/$(TOOLCHAIN_EXTERNAL_PREFIX)/bin
instead of $(TOOLCHAIN_EXTERNAL_INSTALL_DIR)/bin for other
architectures.
However, this is actually only true for Analog Devices pre-built
toolchains. Other Blackfin external toolchains (such as ones built by
Buildroot) do not have this special organization.
Therefore, in order to make those non-ADI Blackfin toolchains work, we
need to change the condition from BR2_bfin to testing specifically for
the ADI toolchains.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The recommended form is without the trailing slash. Buildroot will add a slash
between FOO_SITE and FOO_SOURCE as appropriate.
Reported-by: Arnout Vandecappelle <arnout@mind.be>
Signed-off-by: Luca Ceresoli <luca@lucaceresoli.net>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Add a new option BR2_GCC_ENABLE_LTO that builds gcc and binutils with
LTO support.
Individual packages still have to enable LTO explicitly by passing '-flto' to
GCC, which passes it on to the linker. This option does not add that flag
globally. Some packages detect if the compiler supports LTO and enable the flag
if it does.
To support LTO, ar and ranlib must be called with an argument which triggers the
usage of the LTO plugin. Since GCC doesn't call these tools itself, it instead
provides wrappers for ar and ranlib that pass the LTO arguments. This way
existing Makefiles don't need to be changed for LTO support. However, these
wrappers are called <tuple>-gcc-ar which matches the pattern to link to the
buildroot wrapper in the external toolchain logic. So the external toolchain
logic is updated to provide the correct symlink.
[Thomas:
- Add a separate BR2_BINUTILS_ENABLE_LTO option to enable LTO
support in binutils. This is a blind option, selected by
BR2_GCC_ENABLE_LTO. It just avoids having binutils.mk poke
directly into gcc Config.in options.
- Remove the check on the AVR32 special gcc version, which we don't
support anymore.
- Adapt the help text of the LTO Config.in option to no longer
mention "Since version 4.5", since we only support gcc >= 4.5 in
Buildroot anyway.
- Fix typo in toolchain-external.mk comment.]
Signed-off-by: Peter Kümmel <syntheticpp@gmx.net>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
This toolchain is AMD64-only so restrict it accordingly.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Add support for AMD steamroller optimizations, available in gcc 4.8+ as
bdver3.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
These have been deprecated since 2014.02
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Acked-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Default to glibc over eglibc where it's possible and/or convenient.
Since the eglibc project is basically gone and merged with glibc it
doesn't make sense to keep defaulting to it for architectures that
aren't uClibc-capable.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
External toolchain can also have been generated by Buildroot previously, as
the list that follows demonstrates. Rephrase the paragraph describing what an
external toolchain is as suggested by Thomas Petazzoni, to make it clearer.
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
For now we can only support glibc.
Signed-off-by: Ezequiel Garcia <ezequiel@vanguardiasur.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Slightly reword a comment to no longer mention avr32.
This part dealing with sysroot detection will have to be reworked, now
that we got rid of avr32: we can now require a fully sysroot-aware
toolchain, i.e. at least gcc-4.4.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
CC: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Currently, all the installation work of the toolchain-external package
is done during the install-staging step. However, in order to be able
to properly collect the size added by each package to the target
filesystem, we need to make sure that toolchain-external installs its
files to $(TARGET_DIR) during the install-target step.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Tested-by: Jérôme Pouiller <jezz@sysmic.org>
Getting the hashes from upstream is not always possible:
- Mentor's Sourcery: seems to require an account
- TI's Arago: not able to locate the upstream.
- Linaro: only signatures
- Misc other toolchains.
So, all hashes were locally computed.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Peter Korsgaard <jacmet@uclibc.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Those toolchains are downloaded from Sourceforge, and are therefore
affected by the Sourcefoge download issues. Therefore, this commit
adds the hashes for those toolchain tarballs.
Fixes:
http://autobuild.buildroot.org/results/fa5/fa5e38246dddd661f1d674f3521d21297796bce3/
(and other similar issues)
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
-pipe is causing some build failures in Linux kernel >= 3.17.
Also, nowadays, using -pipe does not gain as much as it used to back in
the days:
Measurements made with a 3.16.7 Linux kernel:
make linux-depends
time sh -c 'make linux-build >/dev/null 2>&1'
Without -pipe:
716.32user 54.44system 3:42.12elapsed 346%CPU
721.22user 54.47system 3:41.81elapsed 349%CPU
722.44user 54.00system 3:42.13elapsed 349%CPU
721.03user 53.81system 3:41.92elapsed 349%CPU
713.21user 53.63system 3:40.51elapsed 347%CPU
706.67user 52.42system 3:38.40elapsed 347%CPU
714.40user 53.18system 3:40.16elapsed 348%CPU
706.01user 53.09system 3:37.87elapsed 348%CPU
705.98user 53.01system 3:38.03elapsed 348%CPU
714.17user 53.55system 3:39.98elapsed 348%CPU
Average: 3:40.29elapsed
With -pipe:
720.13user 53.90system 3:41.98elapsed 348%CPU
713.38user 53.69system 3:40.44elapsed 347%CPU
711.60user 52.81system 3:39.06elapsed 348%CPU
708.66user 53.09system 3:38.59elapsed 348%CPU
711.76user 53.00system 3:38.48elapsed 350%CPU
717.85user 53.97system 3:41.77elapsed 348%CPU
716.77user 53.77system 3:40.91elapsed 348%CPU
717.48user 53.65system 3:41.24elapsed 348%CPU
721.44user 55.67system 3:43.45elapsed 347%CPU
724.61user 55.63system 3:43.35elapsed 349%CPU
Average: 3:40.93elapsed
The delta is well in the measurement noise.
Just get rid of it.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Romain Naour <romain.naour@openwide.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Since a while, the semantic of BR2_PREFER_STATIC_LIB has been changed
from "prefer static libraries when possible" to "use only static
libraries". The former semantic didn't make much sense, since the user
had absolutely no control/idea of which package would use static
libraries, and which packages would not. Therefore, for quite some
time, we have been starting to enforce that BR2_PREFER_STATIC_LIB
should really build everything with static libraries.
As a consequence, this patch renames BR2_PREFER_STATIC_LIB to
BR2_STATIC_LIBS, and adjust the Config.in option accordingly.
This also helps preparing the addition of other options to select
shared, shared+static or just static.
Note that we have verified that this commit can be reproduced by
simply doing a global rename of BR2_PREFER_STATIC_LIB to
BR2_STATIC_LIBS plus adding BR2_PREFER_STATIC_LIB to Config.in.legacy.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
The CodeSourcery toolchains have a very interesting feature: they warn
the user when an unsafe header or library path is used, i.e a path
that will lead host headers or libraries to leak into the build.
This commit adds a similar functionality into our external toolchain
wrapper, so that it can be used with all external toolchains, and can
also be tuned as needed. By default, the external toolchain wrapper
now gives warnings such as:
arm-linux-gcc: WARNING: unsafe header/library path used in cross-compilation: '-I /usr/foo'
arm-linux-gcc: WARNING: unsafe header/library path used in cross-compilation: '-L /usr/bleh'
but the compilation continues successfully. One can then easily grep
in his build log to search for occurences of this message.
Optionally, if BR_COMPILER_PARANOID_UNSAFE_PATH is defined in the
environment to a non empty value, the external wrapper will instead
error out and abort the compilation.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: Romain Naour <romain.naour@openwide.fr>
Tested-by: Romain Naour <romain.naour@openwide.fr>
- Fix the help message for CodeSourcery MIPS toolchains
- Add a hash file
Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Some crazy folks use MIPS machines as build machines. ;-)
On MIPS, the only acceptable hash-style is 'sysv', because the MIPS ABI
defines that the GOT ordering to be the same as the symbols ordering,
while GNU hash requires symbols to be sorted by their hash.
Looking at binutils' code, it seems that only MIPS suffers from that
limitation.
Currently, we force the toolchain wrapper to be linked with both hash
styles, which breaks on MIPS.
So, fix that by singling out MIPS, and use sysv in that case, and both
otherwise.
Reported-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Reviewed-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Tested-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
When the C library being used is uClibc, the locale support can be
disabled. In this case, it does not make sense to show the "Generate
locales" option.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Generating locales is possible in two situations:
- With the internal toolchain backend, when the uClibc library is
used. With uClibc, locales are generated at build time of the C
library, so with uClibc it's only possible with the internal
toolchain backend.
- With either the internal or external toolchain backend when the
glibc library is used. With glibc, locales can be generated
afterwards, using the host-localedef utility.
Until we had the musl C library supported in the internal toolchain
backend, the condition: BR2_TOOLCHAIN_BUILDROOT ||
BR2_TOOLCHAIN_USES_GLIBC was correct to capture the above two
situations. Now that we have musl support in the internal toolchain
backend, then BR2_TOOLCHAIN_BUILDROOT is incorrect, and we should use
BR2_TOOLCHAIN_BUILDROOT_UCLIBC instead.
Basic locale support in musl has appeared in musl 1.1.4, but we are
not yet capable of generating the locale files for musl.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Bump the ARM, ARMeb and AArch64 Linaro toolchains from 14.08 to
14.09. We can't bump to 14.10, because they completely changed the
toolchains and they are now completely broken: they switched from
Crosstool-NG to a new build tool to generate the toolchain, and now
the sysroot handling is completely borked.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Since the BR2_GCC_TARGET_TUNE value is always empty now, there is no
longer a point in using it in the external toolchain logic.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Refactor the toolchain-external Config.in file to use the
BR2_ARM_CPU_ARM*. All of the changes are purely mechanical, except for
the Arago ARMv5 toolchain: it had a 'depends on BR2_GCC_TARGET_ARCH !=
"armv5t"', but armv5t was not a possible value for
BR2_GCC_TARGET_ARCH. Since the toolchain is ARMv5TE, the only ARM
architectures we need to exclude are ARMv4 and ARMv4T.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Don't blindly install the /etc/nsswitch.conf file, it's useless for
toolchains that aren't (e)glibc-based and misleading.
Make the installation conditional on a (e)glibc toolchain.
[Thomas: use $(INSTALL) instead of cp.]
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Commit 2d312b7b61 had a typo
"." instead of "_" in BR2_TOOLCHAIN_HEADERS_AT_LEAST_3.17.
This made selecting 3.17 as custom external headers version
impossible.
Signed-off-by: Karoly Kasza <kaszak@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The Buildroot coding style defines one space around make assignments and
does not align the assignment symbols.
This patch does a bulk fix of offending packages. The package
infrastructures (or more in general assignments to calculated variable
names, like $(2)_FOO) are not touched.
Alignment of line continuation characters (\) is kept as-is.
The sed command used to do this replacement is:
find * -name "*.mk" | xargs sed -i \
-e 's#^\([A-Z0-9a-z_]\+\)\s*\([?:+]\?=\)\s*$#\1 \2#'
-e 's#^\([A-Z0-9a-z_]\+\)\s*\([?:+]\?=\)\s*\([^\\]\+\)$#\1 \2 \3#'
-e 's#^\([A-Z0-9a-z_]\+\)\s*\([?:+]\?=\)\s*\([^\\ \t]\+\s*\\\)\s*$#\1 \2 \3#'
-e 's#^\([A-Z0-9a-z_]\+\)\s*\([?:+]\?=\)\(\s*\\\)#\1 \2\3#'
Brief explanation of this command:
^\([A-Z0-9a-z_]\+\) a regular variable at the beginning of the line
\([?:+]\?=\) any assignment character =, :=, ?=, +=
\([^\\]\+\) any string not containing a line continuation
\([^\\ \t]\+\s*\\\) string, optional whitespace, followed by a
line continuation character
\(\s*\\\) optional whitespace, followed by a line
continuation character
Hence, the first subexpression handles empty assignments, the second
handles regular assignments, the third handles regular assignments with
line continuation, and the fourth empty assignments with line
continuation.
This expression was tested on following test text: (initial tab not
included)
FOO = spaces before
FOO = spaces before and after
FOO = tab before
FOO = tab and spaces before
FOO = tab after
FOO = tab and spaces after
FOO = spaces and tab after
FOO = \
FOO = bar \
FOO = bar space \
FOO = \
GENIMAGE_DEPENDENCIES = host-pkgconf libconfuse
FOO += spaces before
FOO ?= spaces before and after
FOO :=
FOO =
FOO =
FOO =
FOO =
$(MAKE1) CROSS_COMPILE=$(TARGET_CROSS) -C
AT91BOOTSTRAP3_DEFCONFIG = \
AXEL_DISABLE_I18N=--i18n=0
After this bulk change, following manual fixups were done:
- fix line continuation alignment in cegui06 and spice (the sed
expression leaves the number of whitespace between the value and line
continuation character intact, but the whitespace before that could have
changed, causing misalignment.
- qt5base was reverted, as this package uses extensive alignment which
actually makes the code more readable.
Finally, the end result was manually reviewed.
Signed-off-by: Thomas De Schampheleire <thomas.de.schampheleire@gmail.com>
Cc: Yann E. Morin <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
It's not supported and the build breaks.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Acked-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
This new toolchain release fixes a number of issues found with the previous
one. In particular, the issues with fallocate64, prlimit64 and the
ill-installed linux headers seem to be fixed now.
Therefore, there's not need to sanitize the headers for this toolchain and some
packages should now build fine (e.g. fio).
Signed-off-by: Ezequiel Garcia <ezequiel@vanguardiasur.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
It's now been replaced with BR2_ARCH_HAS_ATOMICS, annd all packages have
been changed to use that instead.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Anton Kolesov <Anton.Kolesov@synopsys.com>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The fact that atomic operations are available is not really a
specificity of the toolchain, but rather of the architecture.
So, add a new option that architectures that have atomic operations
can select. This in turn selects the current toolchain atomic option,
until all packages have been converted, at which point the old
toolchain option can be removed.
Reported-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Anton Kolesov <Anton.Kolesov@synopsys.com>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
GCC has several builtin functions that implement atomic operations. Those
functions are architecture specific and may not be implemented by the
specific toolchain. In case of GCC for ARC those functions rely on
LLOCK/SCOND instructions which are optional in ARC CPU's. If ARC CPU doesn't
support those instructions but software tries to use them, then application
will be aborted with Illegal instruction exception. To avoid confusion user
should first specify that their CPU supports atomic extension, which will
allow selection of packages that use builtin atomic functions.
Signed-off-by: Anton Kolesov <Anton.Kolesov@synopsys.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
As reported in bug #7172 [0], setting BR2_TARGET_LDFLAGS to a value
containing a $ sign can lead to unexpected results.
This is because it is very hard to know when the $ sign gets evaluated:
- in the Buildroot-level make
- in the shell called by the Buildroot-level make
- in the package's own build-system, either at configure time, in the
Makefile, in a shell in the Makefile...
So, it is very difficult to know how much escaping that would need.
A proposal is to use a shell variable to pass such values unmolested.
But it is not that simple either, since it still contains a $ sign, and
there is not much certainty as to when it would be evaluated.
Instead, just document this limitation, both in the help text for
BR2_TARGET_LDFLAGS, and in the known-issues section in the manual.
Does not really fix#7172, but at least the limitation is documented.
[0] https://bugs.buildroot.org/show_bug.cgi?id=7172
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Mike Zick <minimod@morethan.org>
Reviewed-by: Thomas De Schampheleire <thomas.de.schampheleire@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The gconv libraries are used to translate between different character sets
('charsets', even 'csets' sometimes). Some packages need them to present
text to the user (eg. XBMC Gotham).
In (e)glibc they are implemented by the internal implemenation of iconv,
called gconv, and are provided as dlopen-able libraries.
Note that some gconv modules need extra libraries (shared by more than
one gconv module), so we must, when adding a subset of modules, scan the
installed modules in search of the missing libraries.
[Thomas: add general explanation in expunge-gconv-modules and fix
coding style.]
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Maxime Hadjinlian <maxime.hadjinlian@gmail.com>
Cc: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Cc: Eric Limpens <limpens@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The 2014R1 toolchain is provided in two flavours by Analog Devices:
one based on gcc 4.3.x, which is considered stable, and one based on
gcc 4.5.x, considered experimental. In commit
5a65b8e185 ("toolchain-external: add ADI
Blackfin 2014R1 toolchain, remove 2012R1") both variants were added.
However, after some testing in the autobuilders, and discussion with
the Analog Devices folks, it turns out that the experimental version
of the toolchain is too experimental. It causes numerous build
failures, and the Analog Devices folks clearly say that it's an early
release and that they expect quite a few problems to show up.
Therefore, this commit removes the experimental flavor and keeps only
the stable variant. Note that the removal/renaming of the Config.in
options is not a problem, since those options were added after the
2014.05 release.
Fixes:
http://autobuild.buildroot.org/results/2a9/2a9d9c332a206fdb46bc8ba022c74d23082a6312/http://autobuild.buildroot.org/results/e1c/e1ce0c1cdd0139208dddaa8f2441ab0e3ab2385e/http://autobuild.buildroot.org/results/993/993aca3f4719afaa4b37524f9136fb8cdc53a066/
and more.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Since commit b2e88073db (toolchain: check ARM EABI vs. EABIhf for external
toolchains), check_arm_abi accepts a second parameter. Update the comment
accordingly.
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Acked-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Since we only need to know whether the compiler runs successfully, and
stderr is empty for the success case, there is no need to redirect stderr to
/dev/null. Moreover, stderr output of the failing case reveals valuable
information on the real failure reason.
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Acked-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
This patch migrates the toolchain and toolchain-buildroot packages to the
virtual package infrastructure, causing the log messages to change from:
>>> toolchain undefined Downloading
>>> toolchain undefined Extracting
...
to
>>> toolchain virtual Downloading
>>> toolchain virtual Extracting
...
and similar for 'toolchain-buildroot', simply because it looks nicer.
At the same time, the directory names also become toolchain-virtual,
toolchain-buildroot-virtual instead of the corresponding 'undefined'
variants.
Signed-off-by: Thomas De Schampheleire <thomas.de.schampheleire@gmail.com>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Acked-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
However, this toolchain is only usable for e500v2 with the SPE ABI.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Tested-by: Ryan Barnett <ryan.barnett@rockwellcollins.com>
And while we're at it, factorize the definition of the musl version,
since it's common to the definition of the tarball names for the
various supported architectures.
[Peter: Adjust Config.in info to match new version]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
This commit adds support for two ADI Blackfin toolchains: the 'stable'
2014R1 based on gcc 4.3, and the 'experimental' 2014R1 based on gcc
4.5.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
For some reason, there is no ARMeb toolchain available in the 2014.05
Linaro release.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
The major changes are: switch to gcc 4.9 instead of 4.8, and switch to
glibc 2.19 instead of glibc 2.18.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Linaro toolchains are released so frequently (every month) that it
doesn't make much sense to support 3 consecutive versions. So, like we
do for ARM big-endian, let's support only one version at a time.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Linaro toolchains are released so frequently (every month) that it
doesn't make much sense to support 3 consecutive versions. So, like we
do for ARM big-endian, let's support only one version at a time.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
When getting the sysroot used for the kernel headers version check,
passing TOOLCHAIN_EXTERNAL_CFLAGS causes a problem when used with
multilib toolchains, where only the main sysroot has the header files,
and the other sysroots only have the libraries.
Since the kernel headers version used is normally the same for all
sysroots, this commit solves this problem by removing the
TOOLCHAIN_EXTERNAL_CFLAGS argument when calling
toolchain_find_sysroot, so that it returns the main sysroot, in which
<linux/version.h> can be found for the kernel headers version check.
Cc: Yann E. MORIN <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Since the introduction of the kernel headers Config.in options, the
external toolchain logic had a check for custom external toolchains to
verify that the kernel headers version entered by the user matches the
one of the toolchain. However, this check was not made for non-custom
external toolchains (i.e the built-in profiles, such as Linaro,
CodeSourcery and al.), making the assumption that the Buildroot
developers will do the right selection.
However, it is quite nice when bumping external toolchains to have
this automatic kernel headers version check, to ensure we select the
appropriate kernel headers version.
Therefore, this commit makes the kernel headers version check
applicable to non-custom external toolchains.
Cc: Yann E. MORIN <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
[Thomas: fix BR2_TOOLCHAIN_HEADERS_AT_LEAST_3_15 to select
BR2_TOOLCHAIN_HEADERS_AT_LEAST_3_14 and not itself.]
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The assumption that musl libc does not support microblaze
little endian mode is wrong. See
http://git.musl-libc.org/cgit/musl/tree/configure line
number 447-448.
Tested with qemu. Just revert previous commit as suggested by
Thomas Petazzoni.
Signed-off-by: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The Musl C library only supports Microblaze BE, not Microblaze LE, so
this commit adjusts the dependencies of the toolchain-buildroot
package to not allow the selection of Musl on Microblaze LE.
Cc: William Welch <bvwelch@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
As reported by William Welch <bvwelch@gmail.com>, the Musl external
toolchain provided by the musl-cross project is only Microblaze
big-endian. In fact, Musl seems to only support the big endian variant
of the Microblaze architecture, with the microblaze-* tuple.
This commit makes sure the Musl pre-built external toolchain provided
by musl-cross can only be selected for a big-endian Microblaze
configuration.
Reported-by: William Welch <bvwelch@gmail.com>
Cc: William Welch <bvwelch@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
This enables powerpc64 and powerpc64le. Currently, le needs at least
glibc 2.19 and gcc 4.9.0. For gdb, 7.7.1 works (added in an earlier
patch).
[Peter: also disallow gcc 4.8 for ppc64le]
Signed-off-by: Cody P Schafer <cody@linux.vnet.ibm.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
This patch adds support for powerpc64le-linux-gnu. This includes
needed patches to fakeroot and gmp.
gmp patch is from upstream HG tree.
fakeroot patch is from Ubuntu written by Adam Conrad.
Signed-off-by: Jeff Bailey <jeffbailey@google.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
In commit cd32da8f79
("toolchain-external: add Linaro ARM big endian toolchain") a mistake
was made, probably due to a rebase conflict that was incorrectly solved:
the call to the post install staging hook that creates the necessary
symbolic links for a root filesystem based on Linaro 2014.02 to work
was removed.
This commit reinstates this call, which should fix the problem
observed by Maxime Hadjinlian while using Linaro 2014.02.
Reported-by: Maxime Hadjinlian <maxime.hadjinlian@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Tested-by: Maxime Hadjinlian <maxime.hadjinlian@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Currently the check for EABI/EABIhf toolchains looks for the
Tag_ABI_VFP_args attribute in the crt1.o file which gcc adds in an
EABIhf toolchain.
In uClibc, however, crt1.o is not compiled from c but assembly, so the
Tag_ABI_VFP_args attribute is not added in the object file. This causes
the EABIhf check in the external toolchain logic to fail for
uClibc-based toolchains.
Fix by compiling a dummy .c file and trying to link the object against the
C library. Since it is impossible to mix EABI and EABIhf code, a mismatch
between the buildroot and toolchain ABI settings will be detected during
this link step.
Fixes bug #6842: https://bugs.busybox.net/show_bug.cgi?id=6842
[Peter: fix final 'fi']
Signed-off-by: Stefan Sørensen <stefan.sorensen@spectralink.com>
[ThomasDS: do full link iso readelf test, update commit message]
Signed-off-by: Thomas De Schampheleire <thomas.de.schampheleire@gmail.com>
Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
MIPS nommu never took off, the only MIPS processors without MMU are
microcontrollers and the only uclinux effort for them has non-upstream
patches against very very old versions.
See http://www.xiptech.com/uclinuxformips.htm
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Otherwise we're getting musl for everything except those who have a
default (mips64*).
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
This commit adds ten toolchains based on the musl C library that are
publicly available from the musl-cross project.
[Peter: fix ppc prefix, only for classic ABI]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
This commit slightly improves the external toolchain backend, and the
gdb build logic to create a file named
$(STAGING_DIR)/usr/share/buildroot/gdbinit which can be used as a
gdbinit file using gdb -x option. This allows gdb to automatically use
the proper sysroot to find libraries.
The initial insight for this patch comes from the report of Oded
Hanson <OHanson@xsightsys.com>, who found an issue with the Eclipse
Buildroot plugin, which was setting a solib-path in gdb, but not a
sysroot. Setting a solib-path was enough to find shared libraries, but
not the dynamic linker. And since Eclipse doesn't allow to set the
sysroot in any other way than giving a gdbinit file, it makes sense to
have Buildroot generate a gdbinit file (which can be used in other
situations than Eclipse).
To achieve this, this commit introduces a gen_gdbinit_file helper in
toolchain/helpers.mk, and uses it for the internal toolchain and
external toolchain backends.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
[ThomasDS: minor updates in commit message]
Signed-off-by: Thomas De Schampheleire <thomas.de.schampheleire@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Enable the internal toolchain backend for aarch64.
Tested with arm_foundationv8_defconfig and ARMs foundation v8 emulator.
Both glibc & eglibc work.
Signed-off-by: Gustavo Zacarias <gustavo@zacarias.com.ar>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The usual way to enable a package using the package infrastructure is to
use a config option so instead to add the toolchain package to the
TARGETS variable in the Makefile add a config option like all the other
toolchain packages.
[Thomas: remove comment that no longer made sense in the main
Makefile, and add a comment above the new hidden Config.in option to
explain what it is useful for.]
Signed-off-by: Fabio Porcedda <fabio.porcedda@gmail.com>
Cc: Arnout Vandecappelle <arnout@mind.be>
Cc: Yann E. MORIN <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Linaro has started to release ARM big endian toolchains, so we
integrate this toolchain in the external toolchain logic of
Buildroot. Since ARM big endian is probably going to be a lot more
uncommon than ARM little endian, we will only support one version at a
time of this toolchain.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The information "To use this toolchain, you must disable soft float
usage." which was visible in the help text of Linaro toolchains is no
longer useful, since those toolchains are only visible when the ARM
EABIhf ABI is selected, which by design is not compatible with
soft-float.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
The Nios-II Sourcery external toolchain (the only Nios-II we currently
support) exports broken kernel headers. In particular, these kernels should
be exported using the "headers_install" rule which applies a set of fixes
on the kernel headers so they are suitable for userspace usage.
In order to fix this, add a post-install hook to perform the header fixes
ourselves. The result is equivalent to apply the "headers_install" rule.
Fixes:
http://autobuild.buildroot.net/results/c32/c32ad4bac5f651502e551f7733f702afaa0e742a/
Signed-off-by: Ezequiel Garcia <ezequiel@vanguardiasur.com.ar>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Old toolchains, with old gcc that do not support -print-sysroot, break the
kernel-headers version check script: it fails to find the sysroot of the
toolchain, and thus ends up including the host's linux/version.h.
Most of the time, this will break early, since the host's kernel headers
will not match the toolchain settings.
But it can happen that the check is succesful, although the configuration
of the toolchain is wrong:
- the custom toolchain has kernel headers vX.Y
- the user selected vX.Z (Z!=Y)
- the host has headers vX.Y
In this case, the check passes OK, but the build of some packages later on
will break (which is exactly what those _AT_LEAST_XXX options were added to
avoid).
Fix that by passing the sysroot to the check script, instead of the cross
compiler.
We get the sysroot as thus:
- for custom toolchains, we use the macro toolchain_find_sysroot. We can
do that, because we already have a complete sysroot with libc.a at that
time.
- for internal toolchain using a custom kernel headers version, we just
use $(STAGING_DIR). We can't use the macro as for custom toolchains
above, because at the time we install the kernel headers, we do not yet
have a complete sysroot with a libc.a. But we can just use
$(STAGING_DIR), since we're only interested in the kernel headers.
For all other types of toolchains, we already have the _AT_LEAST_XXX options
properly set, so we need not add a check in this case.
Fixes:
http://autobuild.buildroot.net/results/f33/f331a6eff0b0b93c73af52db3a6b43e4e598577e/http://autobuild.buildroot.net/results/a57/a5797c025bec50c10efdcff74945aab4021d05e4/
[...]
[Thanks to Thomas for pointing out the toolchain_find_sysroot macro!]
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Signed-off-by: Samuel Martin <s.martin49@gmail.com>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
