core: sanitize RPATH in target tree before copying the overlay
We sanitize the RPATH of ELF files in the target tree to deal
with stupid packages that don't correctly use --prefix/DESTDIR
and that end up putting the full absolute build-time directory
in the RPATH.
We do it before copying the overlay and calling the post-build
script. The user is completely responsible for what happens
in the last two steps, and it should never be touched by us.
Signed-off-by: Wolfgang Grandegger <wg@grandegger.com> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Samuel Martin [Thu, 20 Jul 2017 14:35:14 +0000 (16:35 +0200)]
support/scripts: add fix-rpath script to sanitize the rpath
This commit introduces the script "fix-rpath" able to scan a tree,
detect ELF files, check their RPATH and fix it in a proper way.
The RPATH fixup is done by the patchelf utility using the option
"--make-rpath-relative <root-directory>".
Signed-off-by: Samuel Martin <s.martin49@gmail.com> Signed-off-by: Wolfgang Grandegger <wg@grandegger.com> Acked-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
package/patchelf: add patch for rpath sanitization under a root directory
The patch allows to use patchelf to sanitize the rpath of the buildroot
libraries and binaries using the option "--make-rpath-relative <rootdir>".
Recent versions of patchelf will not built on old Debian and RHEL systems
due to C++11 constructs. Therefore we stick with v0.9 for the time being.
Signed-off-by: Wolfgang Grandegger <wg@grandegger.com> Acked-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
utils/scanpypi: don't pass any arguments to main()
'if __name__ == "__main__"' idiom typically calls main function that
doesn't take any arguments in most cases. We shouldn't pass any tuple to
it.
I've tested the script with python-idna-2.5 and now it works with this
little change.
Qt 5.9 is a Long Term Support version [1]. Some license file changes are
involved, a new dependency on pcre2, and some filename changes. qt5webkit
popped back into the picture for 5.9.1 after being absent for 5.9.0
"officially." QT5_SNAPSHOTS_SITE is removed with this commit as it is not used
for qtwebkit anymore.
This removes the following patches that have been upstreamed for 5.9.1:
Markus Mayer [Tue, 18 Jul 2017 22:10:31 +0000 (15:10 -0700)]
package/linux-tools: add support for building tmon
Signed-off-by: Markus Mayer <mmayer@broadcom.com>
[Thomas:
- tweak Config.in help text as provided by Markus
- adjust commit log
- add missing select BR2_PACKAGE_NCURSES
- add missing dependency on host-pkgconf, used by the tmon Makefile
to find ncurses
- add logic to disable -fstack-protector when the toolchain doesn't
have SSP support.] Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Markus Mayer [Tue, 18 Jul 2017 18:11:37 +0000 (11:11 -0700)]
package/linux-tools: change method for including linux-tool sub-makefiles
Make inclusion ordering of all linux-tool-*.mk sub-makefiles explicit
instead of relying on alphabetical sort order. This is done by
renaming the Linux tools sub-makefiles to the format linux-tool-*.mk.in.
This causes the top-level Makefile to ignore the Linux tools
sub-makefiles.
Until now, the main Makefile included all linux-tool-*.mk files, as
well as linux-tools.mk, and it relied on alphabetical sorting to
include them in the proper order (linux-tool-*.mk before
linux-tools.mk).
Signed-off-by: Markus Mayer <mmayer@broadcom.com> Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
[Thomas: improve comment in the code as suggested by Yann.] Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
nvme-cli had many bugfixes and improvements over the releases.
Buildroot still has version 0.3, so this patch bumps it to the
latest upstream release, version 1.3.
Signed-off-by: Guilherme G. Piccoli <gpiccoli@linux.vnet.ibm.com> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Since commit 654de6512554c (httping: use the new gettext logic), host-gettext
is not a dependency of httping. This breaks the build because the httping
Makefile calls msgfmt unconditionally. Add a patch that allows build without
translation files generation.
support/tests: allow properly indented config fragment
Currently, defining a config fragment in the runtime test infra requires
that the fragment not to be indented. This is beark, and causes grievance
when looking at the code (e.g. to fix it).
Just strip out all leading spaces/tabs when writing the configuration
lines into the config file, allowing in-line indented config fragments,
like so:
class TestFoo(bla):
config = bla.config + \
"""
FOO=y
# BAR is not set
"""
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>
expat: fix build on and for kernel older than 3.17
The expat build system now fails when the getrandom() system call is not
supported. This affect both host and target builds. Define XML_POOR_ENTROPY
for target kernels older than 3.17 to fix the build. For the host package
define XML_POOR_ENTROPY unconditionally since we have no easy way to know the
host kernel version. Note that expat will still use getrandom() on the host
when it is available, we don't make security any worse.
Peter Korsgaard [Sun, 16 Jul 2017 14:23:09 +0000 (16:23 +0200)]
libosip2: add upstream security fix
Fixes CVE-2016-10324 - In libosip2 in GNU oSIP 4.1.0, a malformed SIP
message can lead to a heap buffer overflow in the osip_clrncpy() function
defined in osipparser2/osip_port.c.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
MIPS32 support different FP modes (32,xx,64), so give the user the
opportunity to choose between them. That will cause host-gcc to be built
using the --with-fp-32=[32|xx|64] configure option. Also the
-mfp[32|xx|64] gcc option will be added to TARGET_CFLAGS and to the
toolchain wrapper.
FP mode option shouldn't be used for soft-float, so we add logic in the
toolchain wrapper if -msoft-float is among the arguments in order to not
append the -fp[[32|xx|64] option, otherwise the compilation may fail.
NaN legacy is the only option available for R2 cores and older.
NaN 2008 is the only option available for R6 cores.
R5 cores can have either NaN legacy or NaN 2008, depending on the
implementation. So, if the user selects a generic R5 target architecture
variant, we show a choice menu with both options available. For well
known R5 cores we directly select the NaN enconding they use.
Signed-off-by: Vicente Olivert Riera <Vincent.Riera@imgtec.com> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
As discussed in the mailing list, grub2 usage notes were growing too big
for a Config.in documentation, and so it was agreed that a readme.txt in
the package directory is a better place to put them.
This commit simply moves the documentation as-is to preserve the
original contents as they were in Config.in which can be worked on in
further commits.
Signed-off-by: Erico Nunes <nunes.erico@gmail.com> Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
After many years since the last release and a long time with grub 2.02
in beta, there is finally a release and it brings many bug fixes and
interesting features such as support for ARM.
Patch boot/grub2/0001-remove-gets.patch doesn't seem to be required
anymore as grub-core/gnulib/stdio.in.h has changed significantly since
"053cfcd Import new gnulib." and has another treatment for gets.
Patch
boot/grub2/0002-grub-core-gettext-gettext.c-main_context-secondary_c.patch
was a backport which is present after the bump and therefore is also no
longer necessary.
Since we're adding a Config.in comment, we also introduce a
BR2_TARGET_GRUB2_ARCH_SUPPORTS hidden boolean, in order to avoid
repeating the architecture dependencies.
Signed-off-by: Erico Nunes <nunes.erico@gmail.com> Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
[Thomas: add BR2_TARGET_GRUB2_ARCH_SUPPORTS, remove bogus dependencies
on ARM and AArch64, since enabling Grub2 on those architectures is
done in another commit.] Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
/home/buildroot/buildroot/output/build/kodi-17.3-Krypton/xbmc/filesystem/FTPParse.cpp:37:1:
error: ‘string’ does not name a type
string CFTPParse::getName()
^~~~~~
/home/buildroot/buildroot/output/build/kodi-17.3-Krypton/xbmc/filesystem/FTPParse.cpp:62:25:
error: variable or field ‘setTime’ declared void
void CFTPParse::setTime(string str)
^~~~~~
/home/buildroot/buildroot/output/build/kodi-17.3-Krypton/xbmc/filesystem/FTPParse.cpp:62:25:
error: ‘string’ was not declared in this scope
Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Since commit bbfb02598bf75 (git: use the new gettext logic) host-gettext is no
longer an unconditional dependency of git. When NLS is disabled host-gettext
is not built. This breaks the build of git, because the git Makefile runs
msgfmt unless NO_GETTEXT is defined.
Define NO_GETTEXT when NLS is disabled to fix the build.
Drop the board specific patches which has been applied upstream.
[Peter: also drop now unused BR2_GLOBAL_PATCH_DIR] Signed-off-by: Jörg Krause <joerg.krause@embedded.rocks> Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
package/lshw: don't use NONLS when BR2_SYSTEM_ENABLE_NLS is not set
Since [1], NONLS is added to LSHW_CFLAGS even for glibc based toolchain.
Previously it was not the case since BR2_ENABLE_LOCALE is always
selected for glibc based toolchain.
With NONLS, config.h try to redefine two Intl function (textdomain and
bindtextdomain) that are used by glibc internally.
This break the build with the following error:
sysroot/usr/include/libintl.h:82:52: error: expected unqualified-id before 'throw'
extern char *textdomain (const char *__domainname) __THROW;
The NONLS mode is really for cases where the C library does not provide
*any* gettext implementation, as would be the case with uClibc without
intl stubs enabled.
But in the context of Buildroot, all C libraries provide a gettext
implementation. It might be full-featured or minimal, but it always
exists.
So, remove NONLS from CFLAGS to avoid the build issue with glibc
toolchains.
Build tested with a uClibc toolchain without locale enabled.
pulseaudio: add optional dependency on bluez5_utils
The pulseaudio configure script autodetects the presence of
bluez 4.x and 5.x packages on the system and will exclude the
bluetooth-related modules in their absence.
This commit ensures that bluez5_utils, if selected, are installed
before pulseaudio. The same already happens for bluez_utils (4.x).
Signed-off-by: Calin Crisan <ccrisan at gmail dot com> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Matt Weber [Thu, 13 Jul 2017 20:00:05 +0000 (15:00 -0500)]
gcc: fix build of libsanitizer in gcc 4.9 and 5.x on PowerPC
libsanitizer in gcc fails to build on PowerPC with gcc versions 4.9
and 5.x used in conjunction with glibc 2.25, with the following error:
../../../../gcc-host/libsanitizer/asan/asan_linux.cc: In function 'bool __asan::AsanInterceptsSignal(int)':
../../../../gcc-host/libsanitizer/asan/asan_linux.cc:222:20: error: 'SIGSEGV' was not declared in this scope
return signum == SIGSEGV && common_flags()->handle_segv;
This commit adds a patch that has been submitted to upstream gcc
(https://patchwork.ozlabs.org/patch/725596/) but not merged. The patch
is no longer needed with gcc 6.x and later because the code has been
reworked.
Fixes Buildroot bug #10061
Signed-off-by: Matthew Weber <matthew.weber@rockwellcollins.com>
[Thomas: rework commit log.] Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
board/arcturus/ppc-ucp1020: add patch to fix build with gcc 6.x.
This commit adds a Linux kernel patch to solve a build failure with
the arcturus_ucp1020_defconfig with gcc 6.x:
arch/powerpc/kernel/ptrace.c:407:24: warning: index 32 denotes an offset greater than size of 'u64[32][1] {aka long long unsigned int[32][1]}' [-Warray-bounds]
offsetof(struct thread_fp_state, fpr[32][0]));
^
The patch is upstream in Linux, and can be dropped when
arcturus_ucp1020_defconfig is updated to use a new Linux kernel
version.
Signed-off-by: Oleksandr Zhadan <oleks@arcturusnetworks.com> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Peter Korsgaard [Fri, 14 Jul 2017 13:04:17 +0000 (15:04 +0200)]
package/Makefile.in: export O= to post-build/image scripts for out-of-tree builds
Sometimes it can be interesting to call back into buildroot from a
post-build/image script (E.G. make printvars or similar). For this to work
correctly with out-of-tree builds we need to pass O= to make, but this is
currently not available in the environment of post-build/image scripts.
In concept, O could be derrived from BUILD_DIR (E.G. by stripping /build),
but directly exporting O is cleaner.
O= cannot be exported globally as it interferes with various build systems,
so instead add it to EXTRA_ENV.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com> Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The vcdbg utility is linked to a few libraries, which so far were all
provided by the rpi-userland package.
But a not-so-recent bump of rpi-firmware pulled in a vcdbg that is
linked to an additional library, which is not privided by rpi-userland,
so we must install it.
Reported-by: cluelessperson on #buildroot Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr> Cc: Maxime Hadjinlian <maxime.hadjinlian@gmail.com> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
The current test fails because of a legacy option, renamed during the
recent ext overhaul.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr> Cc: Sébastien Szymanski <sebastien.szymanski@armadeus.com> Cc: Samuel Martin <s.martin49@gmail.com> Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Cc: Arnout Vandecappelle <arnout@mind.be> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Peter Korsgaard [Fri, 14 Jul 2017 14:24:26 +0000 (16:24 +0200)]
tiff: add upstream security fix for CVE-2017-10688
Fixes CVE-2017-10688 - n LibTIFF 4.0.8, there is a assertion abort in the
TIFFWriteDirectoryTagCheckedLong8Array function in tif_dirwrite.c. A
crafted input will lead to a remote denial of service attack.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Peter Korsgaard [Fri, 14 Jul 2017 09:08:12 +0000 (11:08 +0200)]
nginx: security bump to version 1.12.1
Fixes CVE-2017-7529 - Nginx versions since 0.5.6 up to and including 1.13.2
are vulnerable to integer overflow vulnerability in nginx range filter
module resulting into leak of potentially sensitive information triggered by
specially crafted request.
For more details, see:
http://mailman.nginx.org/pipermail/nginx-announce/2017/000200.html
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Valery Kholodkov (5):
Added tag for version 2.0.8
Recreated tag for version 2.0.8
Backported to nginx 0.5.37 by Anthony Kholodkov
Updated Changelog
Merge pull request #88 from antonbarinov/2.255
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Martin Bark [Thu, 13 Jul 2017 12:24:31 +0000 (13:24 +0100)]
package/nodejs: security bump to version 8.1.4
Fixes CVE-2017-1000381 - The c-ares function ares_parse_naptr_reply(), which
is used for parsing NAPTR responses, could be triggered to read memory
outside of the given input buffer if the passed in DNS response packet was
crafted in a particular way. This patch checks that there is enough data
for the required elements of an NAPTR record (2 int16, 3 bytes for string
lengths) before processing a record.
See https://nodejs.org/en/blog/release/v8.1.4/
[Peter: add CVE info] Signed-off-by: Martin Bark <martin@barkynet.com> Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
All versions of Samba from 4.0.0 onwards using embedded Heimdal
Kerberos are vulnerable to a man-in-the-middle attack impersonating
a trusted server, who may gain elevated access to the domain by
returning malicious replication or authorization data.
Samba binaries built against MIT Kerberos are not vulnerable.
CVE-2017-7244 - The _pcre32_xclass function in pcre_xclass.c in libpcre1 in
PCRE 8.40 allows remote attackers to cause a denial of service (invalid
memory read) via a crafted file.
CVE-2017-7245 - Stack-based buffer overflow in the pcre32_copy_substring
function in pcre_get.c in libpcre1 in PCRE 8.40 allows remote attackers to
cause a denial of service (WRITE of size 4) or possibly have unspecified
other impact via a crafted file.
CVE-2017-7246 - Stack-based buffer overflow in the pcre32_copy_substring
function in pcre_get.c in libpcre1 in PCRE 8.40 allows remote attackers to
cause a denial of service (WRITE of size 268) or possibly have unspecified
other impact via a crafted file.
Thomas Petazzoni [Thu, 13 Jul 2017 17:27:27 +0000 (19:27 +0200)]
python-twisted: add missing dependency on host-python-incremental
The recent change on PYTHONPATH for Python 2.x has revealed a missing
dependency in the python-twisted package. The incremental Python
module is listed in both setup_requires and install_requires, so we
must depend on both its target *and* host variants.
Thomas Petazzoni [Wed, 12 Jul 2017 16:28:01 +0000 (18:28 +0200)]
python: remove target Python packages from PYTHONPATH
We currently have
$(TARGET_DIR)/usr/lib/python$(PYTHON_VERSION_MAJOR)/site-packages/
inside the PYTHON_PATH variable, which gets used to define PYTHONPATH,
passed to the host Python interpreter when building/installing target
packages.
However, this is terribly wrong, as it causes the host interpreter to
potentially import target Python packages. This is wrong for several
reasons:
- Some Python packages might need some Python modules to be installed
on the host (described in setup_requires in setup.py), but their
installation currently works because by luck the corresponding
Python module is installed for the target. Some of those cases were
happening for real, and fixed by previous patches.
- Some Python packages include some native code, therefore built for
a specific CPU architecture. When you point the host Python
interpreter to native libraries built for the target, you get nice
build failures, such as the one affecting the python-cffi related
packages.
Making this change allows to fix the python-cffi related build
failures:
Thomas Petazzoni [Wed, 12 Jul 2017 16:28:00 +0000 (18:28 +0200)]
python-treq: needs host-python-incremental
The python-treq package lists the incremental Python module as part of
its setup_requires variable in setup.py, so it must be added as a host
dependency of the python-treq package to avoid build failures.
So far, this issue wasn't visible because python-treq selects
python-twisted, which itself selects the target python-incremental
package. Because python-incremental was before python-treq in the
alphabetic ordering, it was always built before python-treq. And due
to the fact that PYTHONPATH currently contains the directory with
target Python modules, the host Python interpreter was happily using
the target python-incremental while running on the host. But as we are
going to clean up PYTHONPATH, this will no longer be the case, and
hence python-treq needs to be fixed.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Reviewed-by: Yegor Yefremov <yegorslists@googlemail.com> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Thomas Petazzoni [Wed, 12 Jul 2017 16:27:58 +0000 (18:27 +0200)]
python-json-schema-validator: needs versiontools on the host
python-json-schema-validator does not need versiontools on the target,
but only on the host, as it's listed in setup_requires in setup.py.
This was not noticed so far because host Python interpreter is started
with a PYTHONPATH that contains a directory with target Python
packages, so versiontools was found there. But as we are about to fix
PYTHONPATH to no longer include such a directory,
python-json-schema-validator would fail due to versiontools being
missed on the host.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com> Reviewed-by: Yegor Yefremov <yegorslists@googlemail.com> Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Thomas Petazzoni [Wed, 12 Jul 2017 16:27:56 +0000 (18:27 +0200)]
python-u-msgpack: switch to setuptools instead of distutils
python-u-msgpack can use setuptools instead of distutils, and
using setuptools is generally preferred.
In addition, using setuptools allows to make sure the package will
continue to build when we will adjust the PYTHONPATH variable to no
longer point to target Python modules. Without such a change to
setuptools, the build would fail with:
=====================================================================
running install
Checking .pth file support in /home/test/buildroot/output/target/usr/lib/python2.7/site-packages/
/home/test/buildroot/output/host/bin/python -E -c pass
TEST FAILED: /home/test/buildroot/output/target/usr/lib/python2.7/site-packages/ does NOT support .pth files
error: bad install directory or PYTHONPATH
You are attempting to install a package to a directory that is not
on PYTHONPATH and which Python does not read ".pth" files from. The
installation directory you specified (via --install-dir, --prefix, or
the distutils default setting) was:
Here are some of your options for correcting the problem:
* You can choose a different installation directory, i.e., one that is
on PYTHONPATH or supports .pth files
* You can add the installation directory to the PYTHONPATH environment
variable. (It must then also be on PYTHONPATH whenever you run
Python and want to use the package(s) you are installing.)
* You can set up the installation directory to support ".pth" files by
using one of the approaches described here:
Thomas Petazzoni [Wed, 12 Jul 2017 16:27:55 +0000 (18:27 +0200)]
python-pyro: switch to setuptools instead of distutils
python-pyro can use setuptools instead of distutils, and using
setuptools is generally preferred.
In addition, using setuptools allows to make sure the package will
continue to build when we will adjust the PYTHONPATH variable to no
longer point to target Python modules. Without such a change to
setuptools, the build would fail with:
=====================================================================
running install
Checking .pth file support in /home/test/buildroot/output/target/usr/lib/python2.7/site-packages/
/home/test/buildroot/output/host/bin/python -E -c pass
TEST FAILED: /home/test/buildroot/output/target/usr/lib/python2.7/site-packages/ does NOT support .pth files
error: bad install directory or PYTHONPATH
You are attempting to install a package to a directory that is not
on PYTHONPATH and which Python does not read ".pth" files from. The
installation directory you specified (via --install-dir, --prefix, or
the distutils default setting) was:
Here are some of your options for correcting the problem:
* You can choose a different installation directory, i.e., one that is
on PYTHONPATH or supports .pth files
* You can add the installation directory to the PYTHONPATH environment
variable. (It must then also be on PYTHONPATH whenever you run
Python and want to use the package(s) you are installing.)
* You can set up the installation directory to support ".pth" files by
using one of the approaches described here:
Thomas Petazzoni [Wed, 12 Jul 2017 16:27:54 +0000 (18:27 +0200)]
python-pyasn: switch to setuptools instead of distutils
python-pyasn can use setuptools instead of distutils, and using
setuptools is generally preferred.
In addition, using setuptools allows to make sure the package will
continue to build when we will adjust the PYTHONPATH variable to no
longer point to target Python modules. Without such a change to
setuptools, the build would fail with:
=====================================================================
running install
Checking .pth file support in /home/test/buildroot/output/target/usr/lib/python2.7/site-packages/
/home/test/buildroot/output/host/bin/python -E -c pass
TEST FAILED: /home/test/buildroot/output/target/usr/lib/python2.7/site-packages/ does NOT support .pth files
error: bad install directory or PYTHONPATH
You are attempting to install a package to a directory that is not
on PYTHONPATH and which Python does not read ".pth" files from. The
installation directory you specified (via --install-dir, --prefix, or
the distutils default setting) was:
Here are some of your options for correcting the problem:
* You can choose a different installation directory, i.e., one that is
on PYTHONPATH or supports .pth files
* You can add the installation directory to the PYTHONPATH environment
variable. (It must then also be on PYTHONPATH whenever you run
Python and want to use the package(s) you are installing.)
* You can set up the installation directory to support ".pth" files by
using one of the approaches described here:
projects / coffee / buildroot.git / log