radiusd.service references a legacy path for its PIDFile, which
results in a warning at boot:
systemd[1]: /lib/systemd/system/radiusd.service:7: PIDFile= references a path
below legacy directory /var/run/, updating /var/run/radiusd/radiusd.pid →
/run/radiusd/radiusd.pid; please update the unit file accordingly.
Modify the recipe's radiusd.service file to use the correct path.
Signed-off-by: Trevor Gamblin <trevor.gamblin@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Multiple quagga service files are causing the following type of message to
appear during boot:
/lib/systemd/system/zebra.service:10: PIDFile= references a path below legacy
directory /var/run/, updating /var/run/quagga/zebra.pid → /run/quagga/zebra.pid;
please update the unit file accordingly.
Update the service files included as part of the recipe to use /run instead of
/var/run as the PIDFile path.
Signed-off-by: Trevor Gamblin <trevor.gamblin@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Strongswan installs a signal handler for SIGSEGV, SIGILL, and SIGBUS
which attempts to print a stack trace of the crash. For producing line
numbers in the stack trace, it can use libbfd from binutils, or
libunwind, or else it falls back to a slower method using
/usr/bin/addr2line.
Currently the addr2line method is unlikely to actually work, since there
is no RDEPENDS to pull that command into the image.
This patch adds a PACKAGECONFIG to enable the libbfd-based stack traces,
which is likely the best alternative since binutils is already required
for building everything, and it will be faster than the addr2line method
(which requires addr2line and libbfd anyway).
Signed-off-by: Khem Raj <raj.khem@gmail.com>
After upgrade to 1.7.0, path of database file changed from
/var/kea to /var/lib/kea, correct the path to fix service
start failed problem
DHCP6_CONFIG_LOAD_FAIL configuration error using file:
/etc/kea/kea-dhcp6.conf, reason: Unable to open database:
unable to open '/var/lib/kea/kea-leases6.csv'
Signed-off-by: Changqing Li <changqing.li@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Use the systemd class to correctly plug the package into the systemd
infrastructure.
Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
We already depend on systemd in DISTRO_FEATURES so adding it to
RDEPENDS is redundant. We also rdepend on two python packages, so
there's no need to explicitly depend on python3.
Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Use distro_features_check to check for systemd in DISTRO_FEATURES
instead of a hand-crafted python function.
Signed-off-by: Bartosz Golaszewski <bgolaszewski@baylibre.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Delete patch "0001-CVE-2017-16808-AoE-Add-a-missing-bounds-check.patch"
since it is not used in the tcpdump recipe anymore.
Signed-off-by: Peiran Hong <peiran.hong@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
-License-Update: Copyright year updated to 2019.
-fetchmail/02_remove_SSLv3.patch
Removed since this is included in 6.4.1.
Signed-off-by: Zang Ruochen <zangrc.fnst@cn.fujitsu.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
I have no idea if this is the right thing to do, but without the patch I
can't actually buil OE because none of these layers are compatible
with the change in openembedded-core to move to zeus.
Fixes: a5c9709b8d ("layer.conf: Update for zeus series") # openembedded-core
Signed-off-by: Palmer Dabbelt <palmer@dabbelt.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
freeradius fails to build for armv5:
| *** Warning: Linking the executable build/bin/local/radeapclient against the loadable module
| *** libfreeradius-server.so is not portable!
|
| *** Warning: Linking the executable build/bin/local/radeapclient against the loadable module
| *** libfreeradius-eap.so is not portable!
path -Wl,/yow-lpggp31/tgamblin/freeradius.build/tmp-glibc/work/armv5e-oe-linux-gnueabi/freeradius/3.0.19-r0/git/build/lib/local//.libs
/arm-oe-linux-gnueabi/9.2.0/ld: build/lib/local/.libs/libfreeradius-radius.so: undefined reference to `__atomic_compare_exchange_8'
/arm-oe-linux-gnueabi/9.2.0/ld: build/lib/local/.libs/libfreeradius-radius.so: undefined reference to `__atomic_load_8'
/arm-oe-linux-gnueabi/9.2.0/ld: build/lib/local/.libs/libfreeradius-radius.so: undefined reference to `__atomic_store_8'
| collect2: error: ld returned 1 exit status
| scripts/boiler.mk:630: recipe for target 'build/bin/local/radeapclient' failed
Explicitly link libatomic to fix the issue.
Signed-off-by: Trevor Gamblin <trevor.gamblin@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
upb dependency needs to fed as source, since it lacks the CMake based
external module builds like some other deps
Forward port the cross lib installation patch
Drop gettid patch as it was a backport which is in this revision
Link with libatomic on mips
Signed-off-by: Khem Raj <raj.khem@gmail.com>
This upgrade adds some new features and fixes numerous bugs including
the following CVEs:
CVE: CVE-2017-16808 (AoE)
CVE: CVE-2018-14468 (FrameRelay)
CVE: CVE-2018-14469 (IKEv1)
CVE: CVE-2018-14470 (BABEL)
CVE: CVE-2018-14466 (AFS/RX)
CVE: CVE-2018-14461 (LDP)
CVE: CVE-2018-14462 (ICMP)
CVE: CVE-2018-14465 (RSVP)
CVE: CVE-2018-14881 (BGP)
CVE: CVE-2018-14464 (LMP)
CVE: CVE-2018-14463 (VRRP)
CVE: CVE-2018-14467 (BGP)
CVE: CVE-2018-10103 (SMB - partially fixed, but SMB printing disabled)
CVE: CVE-2018-10105 (SMB - too unreliably reproduced,
SMB printing disabled)
CVE: CVE-2018-14880 (OSPF6)
CVE: CVE-2018-16451 (SMB)
CVE: CVE-2018-14882 (RPL)
CVE: CVE-2018-16227 (802.11)
CVE: CVE-2018-16229 (DCCP)
CVE: CVE-2018-16301 (was fixed in libpcap)
CVE: CVE-2018-16230 (BGP)
CVE: CVE-2018-16452 (SMB)
CVE: CVE-2018-16300 (BGP)
CVE: CVE-2018-16228 (HNCP)
CVE: CVE-2019-15166 (LMP)
CVE: CVE-2019-15167 (VRRP)
CVE: CVE-2018-14879 (tcpdump -V)
Deleted patch "0001-CVE-2017-16808-AoE-Add-a-missing-bounds-check.patch"
since the fix is included in the upgrade.
Modified patches "avoid-absolute-path-when-searching-for-libdlpi.patch",
"unnecessary-to-check-libpcap.patch", and "add-ptest.path" since
the upgrade renamed configure.in to configure.ac and made changes
to the file.
Added PACKAGECONFIG for smb. It is disabled by default in
the upgraded version in both the package's configure script and this
bitbake recipe since it is insecure.
Modified the parsing of ptest result to align with the new output
format.
With core-image-minimal on qemux86-64/kvm:
Recipe | Passed | Failed | Skipped | Time(s)
Before | 408 | 0 | 2 | 4
After | 431 | 11 | 2 | 10
11 test failed after the upgrade since libpcap is not upgraded
alongside with tcpdump.
Signed-off-by: Peiran Hong <peiran.hong@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
According to configure.ac, make install might fail when run with multiple jobs:
$ tail -15 log.do_configure
...
When running "make install" do not use any form of parallel or job
server options (such as GNU make's -j option). Doing so may cause
errors.
...
Signed-off-by: Ovidiu Panait <ovidiu.panait@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
These are needed for other packages which want to link against
libstrongswan or other libraries included with Strongswan.
By default, no headers are installed.
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Add qttools-native to PACKAGECONFIG[qt5] DEPENDS to resolve missing
Qt5LinguistTools build error.
Add qtmultimedia to PACKAGECONFIG[qt5] DEPENDS to resolve missing
Qt5Multimedia build error.
Add qtsvg to PACKAGECONFIG[qt5] DEPENDS to resolve missing Qt5Svg build
error.
Inherit cmake_qt5 when qt5 is in PACKAGECONFIG to resolve
get_target_property() called with non-existent target "Qt5::qmake"
build error.
Automatically add qt5 to PACKAGECONFIG when meta-qt5 is in the build
since adding qt5 via a .bbappend won't satisfy the conditional inherit
cmake_qt5. The poppler recipe does exactly this.
Signed-off-by: George McCollister <george.mccollister@gmail.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
This reverts commit 5f32fd6b08.
* fixed by restricting -Wno-error=address-of-packed-member only for
target, spice-native is still useful for qemu-native when spice
PACKAGECONFIG is enabled
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
* with older native gcc on host this will break spice-native with:
cc1: error: -Werror=address-of-packed-member: no option -Waddress-of-packed-member
because older gcc doesn't recognize address-of-packed-member warning
to work around this ignore them all
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
some plugins are installed into libdir/opensaf but we were making a copy
into libdir as well, this patch changes that so the packaged files are
appearing only once
create_empty_library should be using cross compiler with linker flags,
existig code in this area is not cross compile friendly
skip dev-so QA test since some .so are stubs which are packages in PN
Signed-off-by: Khem Raj <raj.khem@gmail.com>
* Remove perl-lib since it had been removed by oe-core:
commit 68552c353255188de3d5b42135360a30e7eac535
Author: Alexander Kanavin <alex.kanavin@gmail.com>
Date: Sun Dec 2 12:46:37 2018 +0100
perl: remove the previous version of the recipe
Now the files are in perl pacakge.
* Fix perl paths when perl is enabled.
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
libOpenIPMI.so.0 is SONAME for openIPMI.so in openipmi-perl package
which means the shlibs code will automatically add it as a provider for
this shared library but actual public library is provided by openipmi
package, and it results in
ERROR: openipmi-2.0.27-r0 do_package: openipmi: Multiple shlib providers for libOpenIPMI.so.0: openipmi-perl, openipmi (used by files: /mnt/jenkins/workspace/Yocto-world-musl/build/tmp/work/aarch64-yoe-linux/openipmi/2.0.27-r0/packages-split/openipmi/usr/bin/openipmi_eventd)
The library in perl package is actually not required to compete to
provide for public interfaces
Signed-off-by: Khem Raj <raj.khem@gmail.com>
$ bitbake spice-native
checking whether the C compiler works... no
configure: error: in `/path/to/spice-native/0.14.2+gitAUTOINC+7cbd70b931_4fc4c2db36-r0/build':
configure: error: C compiler cannot create executables
It's a broken native recipe which means no ones need it any more, so remove it.
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
The restriction to ARM instruction set came in the original
wireshark recipe, which was 2 major versions ago (and also
a few toolchains ago). Wireshark 3.x seems to be building
fine allowing thumb instructions, at least on cortexa9t2hf.
Signed-off-by: Khem Raj <raj.khem@gmail.com>
A lot of changes upstream since the last upgrade, the highlights:
* MQTT v5 protocol support
* Performance improvements
* New mosquitto_rr (request/response) client
* TLS enhancements: OCSP stapling, TLS Engine support, explicit TLS v1.3
support, removed TLS v1.0 support
For further details, see https://mosquitto.org/blog/
Additionally, SUMMARY and DESCRIPTION were tidied up as needed.
Based upon an earlier upgrade patch by Oleksandr Kravchuk <open.source@oleksandr-kravchuk.com>
Signed-off-by: Paul Eggleton <paul.eggleton@linux.intel.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
* when usrmerge is enabled, ${libdir} is /usr/lib, and
${systemd_unitdir} is /usr/lib/systemd, sine PACKAGE
ntpdate is after ntp in variable PACKAGES, so file
${systemd_unitdir}/system/ntpdate.service will be populated
into PACKAGE ntp, but actually we have add it into FILES_ntpdate
when usrmerge is disabled, ${libdir} is empty, and usrmerge is
enabled, files under ${libdir} have been covered by other FILES
config, so fix by remove ${libdir}
* libexecdir is empty, so remove it FILES_${PN}
Signed-off-by: Changqing Li <changqing.li@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Backport selected parts of three upstream commits to fix
CVE-2017-16808 where tcpdump 4.9.2 has a heap-based buffer over-read.
Upstream-Status: Backport
[ several ]
Upstream commits fully backported:
46aead6 [CVE-2017-16808/AoE: Add a missing bounds check]
Upstream commits partially backported:
7068209 [Use nd_ types in 802.x and FDDI headers.]
84ef17a [Replace ND_TTEST2()/ND_TCHECK2() macros by macros using
pointers (1/n)]
46aead6 fixes the vulnerability and requires two macros defined in
7068209 and 84ef17a, which are committed after the release of 4.9.2.
Only the definition of the macros are taken from the two commits
as they impact a wide range of code and are difficult to integrate.
CVE: CVE-2017-16808
Signed-off-by: Peiran Hong <peiran.hong@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Do not try to compile ptests for snmp-bc if it is not in PACKGECONFIG.
Signed-off-by: Joe Slater <joe.slater@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Christophe PRIOUZEAU
Trevor Gamblin
Yi Zhao
Callaghan, Dan
Changqing Li
Adrian Bunk
Bartosz Golaszewski
Paul Eggleton
Peiran Hong
Zang Ruochen
Palmer Dabbelt
Khem Raj
Ovidiu Panait
George McCollister
Martin Jansa
Robert Yang
Otavio Salvador
Leon Anavi
S. Lockwood-Childs
Yuan Chao
Oleksandr Kravchuk
Scott Ellis
Slater, Joseph