Fixes CVE-2018-12910: The get_cookies function in soup-cookie-jar.c in
libsoup 2.63.2 allows attackers to have unspecified impact via an empty
hostname.
Cc: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit 8a4536b2dd)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fixes:
http://autobuild.buildroot.net/results/d6d/d6dc9a640aa1f6650a3e7b9397f2fe2ae3433f4d/http://autobuild.buildroot.net/results/ab5/ab5a58ea7845f9f378454ee1aa7e872448618ba9/
ebx was recently added to the x86 inline asm MULADDC_STOP clobber list to
fix#1550, but this causes the build to fail with GCC < 5 when building in
PIC mode with errors like:
include/mbedtls/bn_mul.h:46:13: error: PIC register clobbered by ‘ebx’ in ‘asm’
This is because older GCC versions treated the x86 ebx register (which is
used for the GOT) as a fixed reserved register when building as PIC.
This is fixed by an improved register allocator in GCC 5+. From the release
notes:
Register allocation improvements: Reuse of the PIC hard register, instead of
using a fixed register, was implemented on x86/x86-64 targets. This
improves generated PIC code performance as more hard registers can be used.
https://www.gnu.org/software/gcc/gcc-5/changes.html
As a workaround, add a patch to detect this situation and disable the inline
assembly, similar to the MULADDC_CANNOT_USE_R7 logic.
Patch submitted upstream: https://github.com/ARMmbed/mbedtls/pull/1986
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit 11241ac656)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
This fixes autobuild failures like
http://autobuild.buildroot.net/results/3288b742cee650ee47a41c5b4d6aaef1fe67bff1
php compile breaks with:
ext/mysqlnd/mysqlnd_auth.o: In function `php_mysqlnd_scramble_sha2':
mysqlnd_auth.c:(.text+0x1054): undefined reference to `PHP_SHA256Init'
mysqlnd_auth.c:(.text+0x1064): undefined reference to `PHP_SHA256Update'
mysqlnd_auth.c:(.text+0x1070): undefined reference to `PHP_SHA256Final'
mysqlnd_auth.c:(.text+0x1078): undefined reference to `PHP_SHA256Init'
mysqlnd_auth.c:(.text+0x1088): undefined reference to `PHP_SHA256Update'
mysqlnd_auth.c:(.text+0x1094): undefined reference to `PHP_SHA256Final'
mysqlnd_auth.c:(.text+0x109c): undefined reference to `PHP_SHA256Init'
mysqlnd_auth.c:(.text+0x10ac): undefined reference to `PHP_SHA256Update'
mysqlnd_auth.c:(.text+0x10bc): undefined reference to `PHP_SHA256Update'
mysqlnd_auth.c:(.text+0x10c8): undefined reference to `PHP_SHA256Final'
It looks like the php mysqli extension needs the hash extension to work. This
seems to be a php Make dependany bug. This patch works around it until the
upstream maintainers can fix it.
Signed-off-by: Christopher McCrory <chrismcc@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit 65f9645263)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Dropbear 2018.76 now uses the --enable-static option to indicate that a static
binary should be built. This will incorrectly pick up the generic buildroot
option intended for building static libraries, causing an unwanted static
binary build with BR2_SHARED_STATIC_LIBS.
Fix by appending an --disable-static configure flag, overriding the buildroot
default.
Signed-off-by: Stefan Sørensen <stefan.sorensen@spectralink.com>
Reviewed-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit c9922a4d2f)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Dropbear 2018.76 changed the default ecdsa host key size form 521 to 256
bits, but this breaks systems with an existing 521 bit key, blocking ssh
logins.
Apply the upstream fix from https://secure.ucc.asn.au/hg/dropbear/rev/0dc3103a5900 :
Only advertise a single server ecdsa key when -R (generate as required) is
specified. Fixes -R now that default ecdsa key size has changed.
[Peter: apply-patches.sh does not like suffix-less filename, so include
patch in Buildroot]
Signed-off-by: Stefan Sørensen <stefan.sorensen@spectralink.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit 8a38d7cc9a)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
By default, Dropbear's configure script enables hardening
flags. Unfortunately, the check for SSP only uses AC_COMPILE_IFELSE(),
and therefore doesn't properly test for the availability of libssp,
visible only at link time.
In addition, Buildroot passes its own hardening flags, depending on
various global options. So, we simply disable hardening flags in
Dropbear.
This fixes a build failure with non-SSP capable toolchains happening
since the bump to 2018.76.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit b0aafef619)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
with this new version:
- "configure --enable-static" should now be used instead of
"make STATIC=1"
- any customised options should be put in localoptions.h
Signed-off-by: Francois Perrad <francois.perrad@gadz.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 2e035a9aec)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
brltty has a very inventive buildsystem, where it internall runs
./configure for the build machine In doing so, it generates a list
of make variables to define what the build machine supports, like
it does for the target.
However, the build variables are generated with a convoluted sed
script that scans the target list, and appends _FOR_BUILD to each
target variables. Then, both lists are included from the Makefile,
on the assumption that the build variables will not clash with the
target variables.
Where it gets interesting, is that that sed script considers the
variables names to match '[A-Za-z][A-Za-z0-9_]*'
And there we see why ATSPI2_PACKAGE does not match: it contains a
digit.
So, some build variables will inevitably override target ones.
Fix that by simply expanding the matching regexp to allow digits
in variable names.
Fixes:
http://autobuild.buildroot.org/results/a37/a37782b3cfc1a96cc129db8fade20a36a7b2d470/http://autobuild.buildroot.org/results/97e/97edc6a47d2140968e84b409cdc960604e5896f2/
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Mario Lang <mlang@blind.guru>
Cc: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 8909897ab3)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
eigen generates a pkg-config file which has a broken prefix
(@CMAKE_INSTALL_PREFIX@). This broken prefix causes an incorrect path when
other packages call pkg-config --cflags eigen.
This patch fixes the prefix in the generated eigen pc file, so projects
which depend on this pc file can now correctly find the eigen include
directory at build time.
Also correct the Cflags output to use the runtime prefix instead of the
build time STAGING_DIR, like we do elsewhere.
[Peter: drop backslashes, tweak commit message]
Signed-off-by: Matt Flax <flatmax@flatmax.org>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit e477dc19b5)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Bump to the latest release of the 1.11.x LTS series as 1.10.x is no longer
supported upstream:
https://www.djangoproject.com/download/
Fixes the following security issues:
- CVE-2017-12794: Possible XSS in traceback section of technical 500 debug
page (1.11.5)
- CVE-2018-6188: Information leakage in AuthenticationForm (1.11.10)
- CVE-2018-7536: Denial-of-service possibility in urlize and urlizetrunc
template filters (1.11.11)
- CVE-2018-7537: Denial-of-service possibility in truncatechars_html and
truncatewords_html template filters (1.11.11)
- CVE-2018-14574: Open redirect possibility in CommonMiddleware (1.11.15)
Also add a hash for the license file.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit ae977e9428)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
glibc-2.28 did quite some lifting in their headers, which breaks the
way some packages were detecting glibc, like gnulib.
However, packages do bundle gnulib (it was meant to be bundled),
and so does m4.
Since m4 hasn't seen the slightest commit since 2017-01-09, it is
bundling an old gnulib version, that predates glibc-2.28, and thus
breaks. It also means that upstream hasn't already fixed the issue.
Furthermore, as upstream is using a git submodule for gnulib, and
that the paths are not the same in the release tarball (in lib/)
and in the git tree (in gnulib/), we can't do a plain backport.
So, we selectively backport the two patches from gnulib upstream,
restricted to only the files that happen to be used in m4.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Reported-by: c32 on IRC
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit c48f8a6462)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The threads dependency comment is currently shown even though the
toolchain supports threads, only because kernel build is disabled.
Merge the kernel and threads comments. This is similar to what we have
in other packages that need the kernel.
Cc: Yann E. MORIN <yann.morin.1998@free.fr>
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit f70b13a7cb)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fixes CVE-2018-10906 - In fuse before versions 2.9.8 and 3.x before 3.2.5,
fusermount is vulnerable to a restriction bypass when SELinux is active.
This allows non-root users to mount a FUSE file system with the
'allow_other' mount option regardless of whether 'user_allow_other' is set
in the fuse configuration. An attacker may use this flaw to mount a FUSE
file system, accessible by other users, and trick them into accessing files
on that file system, possibly causing Denial of Service or other unspecified
effects.
And additionally:
- libfuse no longer segfaults when fuse_interrupted() is called outside the
event loop.
- The fusermount binary has been hardened in several ways to reduce
potential attack surface. Most importantly, mountpoints and mount options
must now match a hard-coded whitelist. It is expected that this whitelist
covers all regular use-cases.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit 9c2bbc3fc9)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Commit 6288409642 ("libselinux: add
patch to fix build with gcc < 4.7") introduced a patch, but its file
name was incorrect, so it was never applied. In addition, the patch
was generated against the Git repository of SELinux, which includes
all projects, and therefore it doesn't apply to the libselinux source
code extracted from the tarball: the "libselinux/" component path
needs to be removed from the patch.
This commit fixes both problems, which should finally and really fix:
http://autobuild.buildroot.net/results/c3272566bb808e43bb77ec59cfe596f7e0fe9a64/
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit cb8d1743ca)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
When using uclibc libdevmapper.so was calling dm_task_get_info_base()
function recursively, leading to segmentation fault. This was
happening because uclibc linker loader just takes first existing
'dm_task_get_info' (which is 'dm_task_get_info_base') symbol in elf
binary, instead of default version.
Add upstreamable lvm2 patch [1], which introduces
--enable-symvers[=STYLE] switch. Use that switch to disable symbol
versions, as we do not plan to support binaries compiled against
old libdevmapper library.
Fixes bug #10781.
[1] https://www.redhat.com/archives/dm-devel/2018-July/msg00187.html
Signed-off-by: Marcin Niestroj <m.niestroj@grinn-global.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 5313c50aa3)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Problem starting lighttpd application with systemd.
/usr/sbin/lighttpd -D -f /etc/lighttpd/lighttpd.conf
2018-06-22 11:21:34: (server.c.733) opening errorlog '/var/log/lighttpd-error.log' failed: Permission denied
2018-06-22 11:21:34: (server.c.1420) Opening errorlog failed. Going down.
Lighttpd can not write the 'lighttpd-access.log' and 'lighttpd-error.log' files
to the directory '/var/log/'.
When using systemd the directory '/var/log' does not allow the user www-data to
write.
To correct the problem, we add /usr/lib/tmpfiles.d/lighttpd.conf.
This file create the 'lighttpd-access.log' and 'lighttpd-error.log' files with
the permission
Signed-off-by: Laurent Hartanerot <laurent.hartanerot@atos.net>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 7c2eb68cac)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
ATF in version 1.2 fails to build with:
./build/juno/release/bl1/context_mgmt.o: In function `cm_prepare_el3_exit':
context_mgmt.c:(.text.cm_prepare_el3_exit+0x54): undefined reference to `cm_set_next_context'
context_mgmt.c:(.text.cm_prepare_el3_exit+0x54): relocation truncated to fit: R_AARCH64_JUMP26 against undefined symbol `cm_set_next_context'
This has been fixed in ATF v1.3. Even though there are even newer
versions of ATF available, we take a conservative approach, and bump
to the first version that has the build issue fixed.
Fixes:
https://gitlab.com/buildroot.org/buildroot/-/jobs/88314771
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 395bc11dde)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
512B is not a correct size to express "512 bytes", and causes a
genimage failure:
ERROR: Invalid size suffix 'B' in '512B'
To express "512 bytes", using just "512" is sufficient. With this
commit, genimage works fine, and we indeed have a 512 bytes unused
partition:
$ fdisk -l output/images/sdcard.img
Disk output/images/sdcard.img: 60 MiB, 62915584 bytes, 122882 sectors
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disklabel type: dos
Disk identifier: 0x00000000
Device Boot Start End Sectors Size Id Type
output/images/sdcard.img1 1 1 1 512B 0 Empty
output/images/sdcard.img2 2 524289 524288 256M 83 Linux
Fixes:
https://gitlab.com/buildroot.org/buildroot/-/jobs/88314963
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit f1bdb63ff4)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
As noted by Arnout in [1], the logic in mesa3d-headers.mk generates a
bogus dri.pc file, which looks like this:
prefix=/usr
exec_prefix=/usr
libdir=/lib
includedir=/include
dridriverdir=/dri
Indeed, the ${...} are expanded as shell variables when the sed
command is executed, while the intention is that those ${...} should
go in the .pc file. By escaping those using $${...}, we get the
expected .pc file:
prefix=/usr
exec_prefix=/usr
libdir=${exec_prefix}/lib
includedir=${prefix}/include
dridriverdir=${libdir}/dri
This was detected by the not yet committed check-package improvement
from Ricardo that detects bogus ${...} usage to reference make
variables.
[1] http://lists.busybox.net/pipermail/buildroot/2018-July/225402.html
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 98e3c1eee5)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Since commit b35ad5d0b4 (ncurses: make host-ncurses use host terminfo), we
are now pointing host-ncurses to the host terminfo (typically) located in
/usr/share/terminfo.
With this change we are reusing the existing host terminfo database, so
there is no point in trying to install our own on top. The user running
buildroot typically will have no write access to /usr/share/terminfo, but
tic in that case falls back to writing the database to $HOME/.terminfo.
Neither of which are desirable.
In case $HOME/.terminfo also isn't writable, tic fails, breaking the install
step for host-ncurses:
** Building terminfo database, please wait...
Running sh ./shlib tic to install /usr/share/terminfo ...
You may see messages regarding extended capabilities, e.g., AX.
These are extended terminal capabilities which are compiled
using
tic -x
If you have ncurses 4.2 applications, you should read the INSTALL
document, and install the terminfo without the -x option.
"terminfo.tmp", line 21272, terminal 'v3220': /home/peko/.terminfo: permission denied (errno 30)
To fix all of this, simply disable the terminfo database install.
Suggested-by: Arnout Vandecappelle <arnout@mind.be>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Acked-by: Hollis Blanchard <hollis_blanchard@mentor.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit b42fb29048)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
chrony calls getrandom() at startup if available, so it needs a workaround
for the blocking behaviour on recent (4.14.39+), similar to what was done
for util-linux in commit c4d86707cd (util-linux: add two upstream patches
to fix blocking on getrandom() with recent kernels).
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit d9937b62e0)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Release notes: https://www.samba.org/samba/history/samba-4.8.4.html
Fixes
o CVE-2018-1139 (Weak authentication protocol allowed.)
o CVE-2018-1140 (Denial of Service Attack on DNS and LDAP server.)
o CVE-2018-10858 (Insufficient input validation on client directory
listing in libsmbclient.)
o CVE-2018-10918 (Denial of Service Attack on AD DC DRSUAPI server.)
o CVE-2018-10919 (Confidential attribute disclosure from the AD LDAP
server.)
Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit 3d7ce0124a)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
This commit reformats
0002-patch-source3__libads__kerberos_keytab.c.patch as a Git-formatted
patch.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 90b26f8764)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Use the correct SPDX identifier for the 3 clause BSD license, like we do
elsewhere.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit f2d748e13b)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Add patches for the following security issues:
CVE-2017-14501 - An out-of-bounds read flaw exists in parse_file_info in
archive_read_support_format_iso9660.c when extracting a specially
crafted iso9660 iso file.
CVE-2017-14502 - Off-by-one error for UTF-16 names in RAR archives,
leading to an out-of-bounds read in archive_read_format_rar_read_header.
CVE-2017-14503 - Out-of-bounds read within lha_read_data_none() in
archive_read_support_format_lha.c when extracting a specially crafted
lha archive.
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit 760fbe789c)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
linuxptp missing.h header implements a replacement clock_nanosleep()
function, which was used when the thread implementation was not NPTL,
because uClibc failed to provide clock_nanosleep() in such
configurations.
However, uClibc-ng has fixed this problem upstream, and has backported
this change in Buildroot in patch
package/uclibc/0002-librt-declare-clock_nanosleep-independent-of-thread-.patch
(the code is upstream and will be part of uClibc-ng 1.0.31). Due to
this, there is now a conflicting definition of clock_nanosleep()
between the C library and the linuxptp missing.h code, which manifests
itself by the following build failure:
missing.h:117:19: error: static declaration of 'clock_nanosleep' follows non-static declaration
static inline int clock_nanosleep(clockid_t clock_id, int flags,
^~~~~~~~~~~~~~~
In file included from clockadj.h:24:0,
from clockadj.c:24:
This commit fixes that by adding a patch that removes the replacement
clock_nanosleep() implementation from the linuxptp code base.
Fixes:
http://autobuild.buildroot.net/results/bf400095a853f5beb28c77a50fcffefe52c3d769/
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit ae2cae70b8)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Prevent creating a dangling symlink when vim is not present on the host
machine. With BR2_ROOTFS_MERGED_USR, just link to "vim", since they are
on the same directory, otherwise link to "../usr/bin/vim".
Signed-off-by: Carlos Santos <casantos@datacom.com.br>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 52385d789a)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The BR2_PACKAGE_PULSEAUDIO_HAS_ATOMIC dependency of
BR2_PACKAGE_PULSEAUDIO was not properly propagated to reverse
dependencies, causing the following kconfig warning:
warning: (BR2_PACKAGE_ESPEAK_AUDIO_BACKEND_PULSEAUDIO && BR2_PACKAGE_GST_PLUGINS_GOOD_PLUGIN_PULSE && BR2_PACKAGE_GST1_PLUGINS_GOOD_PLUGIN_PULSE && BR2_PACKAGE_KODI_PULSEAUDIO && BR2_PACKAGE_MPD_PULSEAUDIO && BR2_PACKAGE_EFL_PULSEAUDIO) selects BR2_PACKAGE_PULSEAUDIO which has unmet direct dependencies (BR2_PACKAGE_PULSEAUDIO_HAS_ATOMIC && BR2_USE_WCHAR && BR2_TOOLCHAIN_HAS_THREADS && !BR2_STATIC_LIBS && BR2_USE_MMU)
This commit fixes that by propagating the dependency as it should have
been done.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 68161802eb)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
When building on a ppc64le host we see this error:
build/genmatch --gimple ../../gcc/match.pd \
> tmp-gimple-match.c
../../gcc/match.pd:120:1 error: expected (, got NAME
negative value by 0 gives -0, not +0. */
This was reported upstream[1] and fixed on the GCC 6 [2] and GCC 7 [3]
branches:
Backport from mainline
2018-01-10 Kelvin Nilsen <kelvin@gcc.gnu.org>
* lex.c (search_line_fast): Remove illegal coercion of an
unaligned pointer value to vector pointer type and replace with
use of __builtin_vec_vsx_ld () built-in function, which operates
on unaligned pointer values.
The patches included in Buildroot contain just the code changes, and not
the changelog, to make it easer to manage backporting.
Tested on Ubuntu Cosmic ppc64le.
[1] https://gcc.gnu.org/bugzilla/show_bug.cgi?id=86162
[2] https://gcc.gnu.org/viewcvs/gcc?view=revision&revision=261621
[3] https://gcc.gnu.org/viewcvs/gcc?view=revision&revision=262243
Signed-off-by: Joel Stanley <joel@jms.id.au>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 88a161b388)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
GCC 4.4 and before do not support a message associated to the
"deprecated" gcc attribute. Since such messages are not that useful in
the context of Buildroot, this commit adds a patch that removes them
in the attr source code.
Signed-off-by: Hollis Blanchard <hollis_blanchard@mentor.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit f910320143)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The current version of attr uses a canned custom buildsystem, that
is borked in quite a few ways (no support for static, overwrites
destination files without unlinking...)
There has been a release recently-ish, with a complete overhaul of
the buildsystem. We can now drop all our patches.
The option to disable NLS has changed, so update accordingly.
Fixes: #10986
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Yegor Yefremov <yegorslists@googlemail.com>
[Arnout: use the original SITE instead of a specific mirror]
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
(cherry picked from commit b02616cff4)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The acl source code uses pragmas inside functions, which is not
supported by gcc 4.4, still in use in older distributions. This commit
adds a patch that drops such pragma. Indeed, this acl is not built
with -Werror, the warning removals are not that important.
Signed-off-by: Hollis Blanchard <hollis_blanchard@mentor.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit a957cdfbb6)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Of most interest is the build system overhaul, and the drop of the
dependency on attr-provided <attr/xattr.h> in favour of the
system-provided <sys/xattr.h>.
That last bit meaning that we will be able to bump attr.
We can drop our patches: static is now natively supported thanks to
the use of libtool, and the internal symbols patch was a backport,
and finally, our install patch is superseded by the use of autotools.
The option to disable NLS has changed, so update accordingly.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Yegor Yefremov <yegorslists@googlemail.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
(cherry picked from commit a37657f44e)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
We don't carry a git-formatted patch, because upstream is in fact a
collection of git trees, while the release tarball is an aggregate
of those repositories. Thus, the layout is different between the
tarball and the SCM...
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Steve Kenton <skenton@ou.edu>
[Thomas:
- add host-pkgconf as a dependency of host-xorriso, since it's needed
for autoreconf to work
- drop HOST_XORRISO_AUTORECONF = YES, since it's implied by
XORRISO_AUTORECONF = YES.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 798d1ec309)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
In a static-only configuration, mariadb fails to build because it
tries to build a shared library:
[ 18%] Linking CXX shared module ha_spider.so
output/host/opt/ext-toolchain/bin/../lib/gcc/arm-buildroot-linux-uclibcgnueabi/6.4.0/../../../../arm-buildroot-linux-uclibcgnueabi/bin/ld: output/host/opt/ext-toolchain/bin/../lib/gcc/arm-buildroot-linux-uclibcgnueabi/6.4.0/../../../../arm-buildroot-linux-uclibcgnueabi/lib/libstdc++.a(eh_globals.o)(.text.__cxa_get_globals_fast+0x14): R_ARM_TLS_LE32 relocation not permitted in shared object
This is not detected by the autobuilders, because mariadb is part of a
Config.in choice, and such choices are not randomized by our current
testing infrastructure.
Signed-off-by: Ryan Coe <bluemrp9@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 58cb24d883)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
There are two comments inside the post install define block that show up in
the build. Fix this by moving the comments outside the block.
Signed-off-by: Ryan Coe <bluemrp9@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit d0a0d49c20)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
GHOSTSCRIPT_FONTS_TARGET_DIR is set to $(TARGET_DIR)/usr/share/fonts/gs
in ghostscript-fonts.mk. If we pass this full path to ghostscript, it
will look for fonts in $(TARGET_DIR), which doesn't exist on the
target.
Instead of /usr/share/fonts/gs, use /usr/share/fonts so ghostscript can
also access other fonts than the ones installed by ghostscript-fonts.
Signed-off-by: Thomas Ehrhardt <tehrhardt@innovaphone.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
(cherry picked from commit a76eab228f)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The help text says that BR2_DEFCONFIG will be used as input, but a
BR2_DEFCONFIG specified in the existing .config file will *not* be
used. So say explicitly that it must be specified on the command line.
Note that both "BR2_DEFCONFIG=... make defconfig" and
"make defconfig BR2_DEFCONFIG=..." will work.
While we're at it, add a semicolon to separate the two statements.
Note that this overflows the help text beyond 80 characters, but that
is already the case in many other lines.
Reported-by: Anisse Astier <anisse@astier.eu>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 27aa7ae618)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Install the LTTng control library headers and shared objects
to staging.
The C interface to LTTng described here:
https://lttng.org/docs/v2.10/#doc-liblttng-ctl-lttng
requires including <lttng/lttng.h> and linking against liblttng-ctl,
but those parts are not available unless this package does a staging
install.
Signed-off-by: John Faith <jfaith@impinj.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit 3466298b0f)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Added license hash, switched _UPSTREAM to an upstream directory which
also contains older release tarballs.
Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit 621363f5ac)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
arm64 systems support SMBIOS, so update the dmidecode config
to allow building dmidecode for arm64.
Signed-off-by: Geoff Levand <geoff@infradead.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 6e00a671bc)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Currently makedevs silently ignores extended attributes with leading
whitespace, for example those added to a <PACKAGE>_PERMISSIONS following
the recommended style from check-package.
Makedevs already ignores leading whitespace for normal entries (file
permission changes and device files creation). Do the same for extended
attributes.
Fixes: #11191.
Reported-by: Jean-pierre Cartal <jpcartal@free.fr>
Signed-off-by: Ricardo Martincoski <ricardo.martincoski@gmail.com>
Cc: Arnout Vandecappelle <arnout@mind.be>
Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit 2d8d5ced10)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fixes:
*) SECURITY: CVE-2018-8011 (cve.mitre.org)
mod_md: DoS via Coredumps on specially crafted requests
*) SECURITY: CVE-2018-1333 (cve.mitre.org)
mod_http2: DoS for HTTP/2 connections by specially crafted requests
Changelog: http://www.apache.org/dist/httpd/CHANGES_2.4.34
Signed-off-by: Bernd Kuhls <bernd.kuhls@t-online.de>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit 8ef1aaa084)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
When the output directory is nested under a long path name, calling the
gdbus-codegen, glib-genmarshal, or glib-mkenums scripts throws the
error:
> '/usr/bin/env: /...: No such file or directory'
This is because libglib2 uses the absolute path to the python
interpreter for its shebang statements, and shebangs have a max length
of 127 chars[1]. A long absolute path will be cut off and appear as a
missing file.
Since the host Python is present in the $PATH passed to all packages, we
can copy the workaround from Yocto and just use the python interpreter
in $PATH[2]. However, 'python' is used instead of 'python3' as the
scripts are compatible with both.
[1] https://linux.die.net/man/2/execve
[2] http://cgit.openembedded.org/openembedded-core/commit/?id=eef7883587acc933d6f34b559ec03ff84d18573b
Signed-off-by: Brandon Maier <brandon.maier@rockwellcollins.com>
Signed-off-by: David Owens <david.owens@rockwellcollins.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 3ab5dfb548)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Starting from kernel version 4.15, Linux supports loading
regulatory.db and regulatory.db.p7s files directly from /lib/firmware,
without the need of crda. If the kernel can't load those files it'll
fallback to using crda, but the direct way is preferred.
This patch copies those those files under /lib/firmware. The use of
crda shouldn't be necessary with kernels of 4.15 or later.
Signed-off-by: Christoph Engelbert <me@noctarius.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 39e4b46748)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The bind configure.in now checks for "${with_zlib}/include/zlib.h".
Remove the redundant "include/".
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit ba3c7e806d)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fixes CVE-2018-5738: When recursion is enabled but the allow-recursion
and allow-query-cache ACLs are not specified, they should be limited to
local networks, but they were inadvertently set to match the default
allow-query, thus allowing remote queries.
Update license file hash; copyright year update.
Add reference to tarball signature key.
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit b36577a266)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Host GDB suffers a serious problem: pressing backspace (or ^W ^U or any other
"delete" key) results in a plain space being printed instead, making the
command prompt almost completely unusable.
That's because it's using host-ncurses, which embeds a path for the terminfo
database into the library itself. That path ends up being something like
/home/hollisb/buildroot.git/output/host/share/terminfo, which obviously doesn't
generally exist other hosts. ('relocate-sdk.sh' cannot and does not edit
binaries like libncurses.so.6, so doesn't resolve this problem.)
/usr/share/terminfo is a far better path to use, since it almost certainly
exists on the host. Theoretically, it could be from a different ncurses version
with incompatible terminfo database format, but this doesn't seem to be a
problem in practice. (Future patches could address the theoretical problem if
it actually appears in real life.)
This change allows buildroot's host gdb, which uses ncurses 6.x, to work on
RHEL5, RHEL6, and RHEL7, which all provide terminfo from ncurses 5.x.
Signed-off-by: Hollis Blanchard <hollis_blanchard@mentor.com>
Reviewed-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit b35ad5d0b4)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The 2017 MinnowBoard Turbot skews no longer use the Realtek 8169
Ethernet chip. The Turbot D0 uses The Intel I211, the Turbot Dual
Ethernet has an I210 (quad-core) or I211 (dual-core).
Add CONFIG_IGB to the Linux config but keep the Realtek 8169 support
to avoid breaking older boards
Tested on Turbot D0, lspci -k shows the igb driver is used:
03:00.0 Class 0200: 8086:1539 igb
Signed-off-by: Pierre-Louis Bossart <pierre-louis.bossart@linux.intel.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 02b7fbf8fc)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fixes CVE-2018-0500: curl might overflow a heap based memory buffer when
sending data over SMTP and using a reduced read buffer.
Drop upstream patch.
Add reference to tarball signature key.
Drop CRYPTO_lock seed. Removed from configure script since 7.45.
Cc: Matt Weber <matthew.weber@rockwellcollins.com>
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit bf79731153)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Currently, the stdout is consigned to oblivion, while the stderr is not
redirected at all. So, when the configuration under test redefines a
symbol, like is the case when testing busybox for example, there is an
ugly warning.
So, just redirect both stdout and stderr to the logfile.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Tested-by: Matt Weber <matthew.weber@rockwellcollins.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
(cherry picked from commit 2664022787)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
* chacha20poly1305: use slow crypto on -rt kernels on arm too
Leftover from the last commit of the previous snapshot that we forgot to
handle.
* tools: getentropy requires macOS 10.12
Small build time fixup for old versions of macOS.
* queueing: remove useless spinlocks on sc
* queueing: re-enable preemption periodically to lower latency
* simd: encapsulate fpu amortization into nice functions
* simd: no need to restore fpu state when no preemption
This will improve general system latency on preempt-enabled systems, like
desktops.
* dns-hatchet: apply resolv.conf's selinux context to new resolv.conf
Fixes wg-quick's dns hatchet on CentOS.
* qemu: bump default kernel
By bumping to 4.17.2, we actually uncovered a bug in the SLUB allocator, which
upstream is now fixing: https://lkml.org/lkml/2018/6/18/1407
* noise: take locks for ss precomputation
* netlink: maintain static_identity lock over entire private key update
Minor locking correctness fixes and optimizations.
* noise: wait for crng before taking locks
We now make sure that an outgoing packet which needs a potentially unseeded
rng won't block a call to wg(8), which takes similar locks for retrieving
data.
* receive: drop handshake packets if rng is not initialized
If the rng is unseeded, we drop incoming handshake packets, so that it's not
possible for an attacker to fill the handshake queue thereby provoking
cookies.
* ratelimiter: mitigate reference underflow
* ratelimiter: do not allow concurrent init and uninit
Minor correctness and hardening fixes, which don't fix anything particular in
WireGuard, but might be useful if our ratelimiter is ever used elsewhere.
* compat: use stabler lkml links
* poly1305: add missing string.h header
Minor fixups.
* receive: don't toggle bh
The last snapshot caused a big performance regression, which we partially
revert here. This general matter, though, will be revisited in the future,
perhaps by switching to NAPI.
* main: test poly1305 before chacha20poly1305
* poly1305: give linker the correct constant data section size
While the default bfd linker did the right thing, gold would sometimes merge
section incorrectly because of an incorrect section length field, resulting in
wrong calculations.
* simd: add missing header
Fixes a compile error on a few odd kernels.
* global: fix a few typos
* manpages: eliminate whitespace at the end of the line
* tools: fix misspelling of strchrnul in comment
Cosmetic fixups.
* global: use ktime boottime instead of jiffies
* global: use fast boottime instead of normal boottime
* compat: more robust ktime backport
We now use the equivalent of clock_gettime(CLOCK_BOOTTIME) for doing age
checks on time-limited objects, such as ephemeral keys, so that on systems
where we don't clear before sleep (like Android), we make sure to invalidate
the objects after the proper amount of time, taking into account time spent
asleep.
* wg-quick: android: prevent outgoing handshake packets from being dropped
Recent android phones block outgoing packets using iptables while the system
is asleep. This makes sense for most services, but not for a tunnel device
itself, so we work around this by inserting our own iptables rule.
* device: print daddr not saddr in missing peer error
* receive: style
Debug messages now make sense again.
* wg-quick: android: support excluding applications
Android now supports excluding certain apps (uids) from the tunnel.
* selftest: ratelimiter: improve chance of success via retry
* qemu: bump default kernel version
* qemu: decide debug kernel based on KERNEL_VERSION
Some improvements to our testing infrastructure.
* receive: use NAPI on the receive path
This is a big change that should both improve preemption latency (by not
disabling it unconditionally) and vastly improve rx performance on most
systems by using NAPI. The main purpose of this snapshot is to test out this
technique.
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit 150fa57ed0)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
When json-c is enabled but libcurl is disabled, clamav tries to build
the clamsubmit program, which fails with:
CC clamsubmit.o
clamsubmit.c:6:23: fatal error: curl/curl.h: No such file or directory
#include <curl/curl.h>
This is due to an incorrect curl-config detection logic, leading to
/bin/curl-config being present making the configure script believe
that curl is available, even when --without-libcurl is explicitly
passed.
This commit adds a patch, submitted upstream, which fixes this
problem.
Fixes:
http://autobuild.buildroot.net/results/c43d2ebd8ab30016969d642dbd71c297dc5f6bab/
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 037572ee56)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Increase the filesystem size for raspberrypi3 by setting
BR2_TARGET_ROOTFS_EXT2_SIZE to "120M". The filesystem for
raspberrypi3_64 has the same size. The size is so large because the
kernel config enables many many modules.
Fixes:
Copying files into the device: __populate_fs: Could not allocate block in ext2 filesystem while writing file "services"
mkfs.ext4: Could not allocate block in ext2 filesystem while populating file system
This error is not reproducible, it seems to depend on e.g. directory
ordering.
Signed-off-by: Leon Anavi <leon.anavi@konsulko.com>
Reviewed-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
(cherry picked from commit e26a71d1a9)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
We need to install polkit before systemd if we want to enable
polkit integration, because systemd will not install polkit
policy files unless it finds polkit-gobject-1.
Signed-off-by: Asgeir Bjarni Ingvarsson <asgeir@appdynamic.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: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
(cherry picked from commit 0fe0c50d36)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The original site www.liblogging.org has been taken down because of the
new European data protection regulation (GDPR), to which it was not
compliant.
Signed-off-by: Carlos Santos <casantos@datacom.com.br>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
(cherry picked from commit 9d1d513c65)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Perf profiling cannot be used if CONFIG_PERF_EVENTS is not enabled in the
kernel configuration. Similar to other tools, like ktap, we can enable the
right options automatically.
Signed-off-by: Jan Heylen <jan.heylen@nokia.com>
Signed-off-by: Thomas De Schampheleire <thomas.de_schampheleire@nokia.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 63d76c6f2f)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Add the i686 package list to install when using pre-built 32 bits
binaries with a redhat/fedora host distribution (glibc.i686 and
zlib.i686).
Signed-off-by: David De Grave (Essensium/Mind) <david.degrave@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit d0505b849f)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
In commit 52282828f0
("package/network-manager: add patches to build with kernel headers
3.2"), network-manager was changed to build with kernel headers as old
as 3.2, instead of 3.7 before. The dependency was changed accordingly,
but the Config.in comment was not.
This commit fixes this inconsistency.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 760dc48e15)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Fixes CVE-2018-9918: mishandle certain "expected dictionary key but
found non-name object" cases, allowing remote attackers to cause a
denial of service (stack exhaustion)
https://github.com/qpdf/qpdf/issues/202
Drop local SHA256 hash since we use upstream provided SHA512.
Signed-off-by: Baruch Siach <baruch@tkos.co.il>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 473390a30c)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
When the host has a DB library installed, namely LMDB, host-heimdal
will detect it and try to use, resulting in a build failure due to
missing symbols.
Really disable all the DB backends, we don't need them.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 07d4d97171)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The compiler g++ reports an error when the header arm_acle.h is
included from version 7.
This patch backports the bugfix upstreamed [1] for gcc-7 and gcc-8.
Fixes:
In file included from ../../include/QtCore/5.10.1/QtCore/private/../../../../../src/corelib/tools/qsimd_p.h:333,
from ../../include/QtCore/5.10.1/QtCore/private/qsimd_p.h:1,
from global/qlogging.cpp:58:
/home/gportay/src/buildroot/output/host/lib/gcc/arm-buildroot-linux-gnueabihf/8.1.0/include/arm_acle.h: In function ‘void __arm_ldc(unsigned int, unsigned int, const void*)’:
/home/gportay/src/buildroot/output/host/lib/gcc/arm-buildroot-linux-gnueabihf/8.1.0/include/arm_acle.h:48:46: error: invalid conversion from ‘const void*’ to ‘const int*’ [-fpermissive]
return __builtin_arm_ldc (__coproc, __CRd, __p);
^~~
<built-in>: note: initializing argument 3 of ‘void __builtin_arm_ldc(unsigned int, unsigned int, const int*)’
...
Makefile:196: recipe for target 'sub-corelib-make_first' failed
make[3]: *** [sub-corelib-make_first] Error 2
Makefile:48: recipe for target 'sub-src-make_first' failed
make[2]: *** [sub-src-make_first] Error 2
package/pkg-generic.mk:229: recipe for target '/home/gportay/src/buildroot/output/build/qt5base-5.10.1/.stamp_built' failed
make[1]: *** [/home/gportay/src/buildroot/output/build/qt5base-5.10.1/.stamp_built] Error 2
Makefile:16: recipe for target '_all' failed
make: *** [_all] Error 2
[1]: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81497
[Peter: drop gcc-8.x patch]
Signed-off-by: Gaël PORTAY <gael.portay@savoirfairelinux.com>
Reviewed-by: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 78af2a6362)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Hidapi source uses iconv_open UTF-16 conversion [1] which requires
gconv libararies when buiding with glibc in order to successfully
find serial number, manufacturer and product strings [2].
Use select in the config file to autmatically select gconv libararies
for inclusion on the target when building with glibc
Tested with ./utils/test-pkg -c hidapi.cfg -p hidapi and hidapi.cfg file
contents:
BR2_PACKAGE_HIDAPI=y
BR2_PACKAGE_EUDEV=y
BR2_ROOTFS_DEVICE_CREATION_DYNAMIC_EUDEV=y
BR2_USE_MMU=y
BR2_USE_WCHAR=y
Also tested on x86-64 build target with application using hidapi
[1]: a6a622ffb6/libusb/hid.c (L446)
[2]: https://www.silabs.com/community/interface/forum.topic.html/cp2114_linux_sta-RtWg
Signed-off-by: Dan Walkes <danwalkes@trellis-logic.com>
Tested-by: Dan Walkes <danwalkes@trellis-logic.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit a7404a79c6)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
This commit adds a patch to the host package patchelf to fix an issue
with endianness when the --make-rpath-relative option is
specified. Currently, patchelf does not take into account the
architecture of the target ELF when it performs the fixup on the
RPATH. Thus, if it differs from the host in endianness, the RUNPATH
field will be invalid. The patch fixes the bug reported here:
https://bugs.busybox.net/show_bug.cgi?id=11101
Signed-off-by: Bryce Ferguson <bryce.ferguson@rockwellcollins.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 17c9810fe7)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
So the host ncurses includes and library are used instead of a mix of both,
causing corrupted characters. Similar to the linux fix in commit
6d3d09e232 (linux: don't override HOSTCC for kconfig), except that we
pass the linker flags in HOSTLDFLAGS.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Acked-by: Jagan Teki <jagan@amarulasolutions.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit 9255fd9fc7)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
- Since version 5.5, daemon is licensed under MIT:
889e5e6bd5
- Plugins are licensed under MIT, GPL-2.0 or LGPL-2.1
- Add libltdl/COPYING.LIB to license files to have a copy of LGPL-2.1 as
it is not in COPYING
- Add hash for both license files
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit d5fa317f2d)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
As for strace [1], when <sys/reg.h> is included after <linux/ptrace.h>,
the build fails on m68k with the following diagnostics:
In file included from ./../nat/linux-ptrace.h:28:0,
from linux-low.h:27,
from linux-m68k-low.c:20:
[...]/usr/include/sys/reg.h:26:3: error: expected identifier before numeric constant
PT_D1 = 0,
^
[...]usr/include/sys/reg.h:26:3: error: expected « } » before numeric constant
[...]usr/include/sys/reg.h:26:3: error: expected unqualified-id before numeric constant
In file included from linux-m68k-low.c:27:0:
[...]usr/include/sys/reg.h:99:1: error: expected declaration before « } » token
};
^
Fix this by moving <sys/reg.h> on top of "linux-low.h".
Fixes:
https://gitlab.com/free-electrons/toolchains-builder/-/jobs/72006385
[1] 6ebf6c4f9e
Signed-off-by: Romain Naour <romain.naour@gmail.com>
Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Cc: Waldemar Brodkorb <wbx@openadk.org>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit 8646a7fae9)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Contains a fix for a crash when forwarding packets from devices that use
flow offloading and a x86 FPU issue on -rt kernels.
Also add a post-release upstream patch fixing a compilation issue related to
memcpy().
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit 630fd8785b)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Contains a number of fixes, including fixes for building with GCC 8.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit f6b9d59f61)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
gcc-8.1 for xtensa miscompiles uClibc dynamic linker due to gcc PR
target/65416. The build completes successfully, but the binary is
non-functional because the following fragment in the _dl_get_ready_to_run
in ld-uClibc.so overwrites register spill area on stack causing register
corruption in the previous call frame and a subsequent crash:
419f: f0c1b2 addi a11, a1, -16
41a2: 1ba9 s32i.n a10, a11, 4
41a4: 0bc9 s32i.n a12, a11, 0
41a6: 5127f2 l32i a15, a7, 0x144
41a9: 1765b2 s32i a11, a5, 92
41ac: 4e2782 l32i a8, a7, 0x138
41af: 146af2 s32i a15, a10, 80
41b2: 001b10 movsp a1, a11
The crash terminates the init process and causes kernel panic.
The fix prevents reordering of movsp opcode and any access to the stack
frame memory and is applicable to all existing gcc versions.
[Peter: drop gcc-8.x patch]
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
(cherry picked from commit 91e0fc0bf4)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
On some architectures, atomic binutils are provided by the libatomic
library from gcc. Linking with libatomic is therefore necessary,
otherwise the build fails with:
sparc-buildroot-linux-uclibc/sysroot/lib/libatomic.so.1: error adding symbols: DSO missing from command line
This is often for example the case on sparcv8 32 bit.
Fixes:
http://autobuild.buildroot.net/results/fd6bee70ff20bee9607a9f6f557a3a793d00cd9f
Signed-off-by: Fabrice Fontaine <fontaine.fabrice@gmail.com>
[Thomas: use LIBS instead of LDFLAGS.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 265e25834f)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Both plugins `geometryloaders' and `renderplugins' are available since
Qt 5.9.
Fixes:
>>> qt53d 5.6.3 Installing to target
cp -dpf /home/gportay/src/buildroot/output/host/arm-buildroot-linux-gnueabihf/sysroot/usr/lib/libQt53D*.so.*
/home/gportay/src/buildroot/output/target/usr/lib
cp -dpfr /home/gportay/src/buildroot/output/host/arm-buildroot-linux-gnueabihf/sysroot/usr/lib/qt/plugins/geometryloaders
/home/gportay/src/buildroot/output/target/usr/lib/qt/plugins
cp: cannot stat '/home/gportay/src/buildroot/output/host/arm-buildroot-linux-gnueabihf/sysroot/usr/lib/qt/plugins/geometryloaders': No such file or directory
package/pkg-generic.mk:310: recipe for target '/home/gportay/src/buildroot/output/build/qt53d-5.6.3/.stamp_target_installed' failed
make[2]: *** [/home/gportay/src/buildroot/output/build/qt53d-5.6.3/.stamp_target_installed] Error 1
Signed-off-by: Gaël PORTAY <gael.portay@savoirfairelinux.com>
Signed-off-by: Gaël PORTAY <<a href="mailto:gael.portay@savoirfairelinux.com" target="_blank" rel="noreferrer">gael.portay@savoirfairelinux.com</a>><br>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit e922bbc7ae)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Adds WTF platform support for the 32-bits armv8-a architectures.
Fixes:
In file included from ../3rdparty/javascriptcore/JavaScriptCore/config.h:26,
from ../3rdparty/javascriptcore/JavaScriptCore/pcre/pcre_compile.cpp:44:
../3rdparty/javascriptcore/JavaScriptCore/wtf/Platform.h:370:6: error: #error "Not supported ARM architecture"
# error "Not supported ARM architecture"
^~~~~
The patch is based on two upstream fixes in WebKit[1][2].
See also commit cea7aa873a.
[1]: 313d9fc4bd
[2]: 98f0de0709
Signed-off-by: Gaël PORTAY <gael.portay@savoirfairelinux.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
(cherry picked from commit 15706f86ba)
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Some files were not shown because too many files have changed in this diff
Show More
Reference in New Issue
Block a user
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
