There are several packages which all provide /usr/lib/sendmail
when lsb is enabled. So use alternative to manage it.
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Bump up SRCREV_drbd-utils to includes two more commits to fix the following
netlink failure with nested attributes.
$ drbdsetup new-resource r0
Invalid argument
92ade5989027 ("netlink: prepare for kernel v5.2")
859151b228d3 ("netlink: Add NLA_F_NESTED flag to nested attribute")
Signed-off-by: He Zhe <zhe.he@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Rename systemd service units since it uses strongswan-starter.service
for the legacy unit and strongswan.service for the modern one.
Signed-off-by: Yi Zhao <yi.zhao@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Refresh all patches.
Remove openipmigui-not-compile-pyc-pyo.patch and
do-not-install-pyc-and-pyo.patch as the issues had been fixed upstream.
Add ipmi_serial_bmc_emu.c-include-readline.h-from-readli.patch to fix
the build error.
Signed-off-by: Yi Zhao <yi.zhao@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
* so that -Wno-address-of-packed-member is used for all the builds not
only for target builds
* with native it was overwritten with BUILD_CFLAGS and for nativesdk with
BUILDSDK_CFLAGS
* this issue isn't specific for gcc-9 as the original commit said, I see
nativesdk-spice failing on older hosts with gcc-7 with the same issue
* also use -Wno-error=address-of-packed-member instead of
-Wno-address-of-packed-member so that there is still a warning, just
not fatal for the build in combination with -Werror.
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
There are folowing compile errors on Linux 32bit system:
red-channel.c:207:73: error: format '%x' expects argument of type
'unsigned int', but argument 7 has type 'long unsigned int' [-Werror=format=]
|207| red_channel_debug(self, "thread_id 0x%" G_GSIZE_MODIFIER "x",
~~~~~~~~~~~~~~~~~~~~~^
self->priv->thread_id);
~~~~~~~~~~~~~~~~~~~~~^
On 32bit system, #define G_GSIZE_MODIFIER "". But the type of
'self->priv->thread_id' is 'unsigned long int' which should match '%lx'
not '%x'.
So we should recovery the <0x%" G_GSIZE_MODIFIER "x"> to <0x%lx">.
And others files modification are similar to G_GSIZE_MODIFIER.
Signed-off-by: Hongzhi.Song <hongzhi.song@windriver.com>
Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
samba have bundled libldb, so when both samba and libldb, or both pyldb \
and samba-python is installed, below error will appear:
file /usr/bin/ldbadd conflicts between attempted installs of samba-4.8.11-r0.i586 and libldb-1.4.1-r0.i586
file /usr/bin/ldbdel conflicts between attempted installs of samba-4.8.11-r0.i586 and libldb-1.4.1-r0.i586
...
file /usr/lib/python2.7/site-packages/ldb.so conflicts between attempted installs of libpyldb-util1-1.4.1-r0.i586 and samba-python-4.8.11-r0.i586
so add rconflicts for both packages
Signed-off-by: Changqing Li <changqing.li@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Drop the patch for fixing printf formatting, its fixed differently
upstream
Forward port pthread_t printf patch
Add -Wno-address-of-packed-member needed with gcc9
replace _append += with just _append syntax
Signed-off-by: Khem Raj <raj.khem@gmail.com>
drbd-utils installs directory /var/run but is a link file installed by
package base-files. Remove /var/run to fix the file conflict issue.
Signed-off-by: Kai Kang <kai.kang@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
It fails to build drbd with errors:
| LD [M] .../tmp/work/qemux86_64-poky-linux/drbd/9.0.17-1-r0/drbd-9.0.17-1/drbd/drbd.o
| x86_64-poky-linux-ld.bfd: cannot find
| .../tmp/work/qemux86_64-poky-linux/drbd/9.0.17-1-r0/drbd-9.0.17-1/drbd/drbd_bitmap.o: No such file or directory
Backport patch from upstream to fix the issue.
Signed-off-by: Kai Kang <kai.kang@windriver.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
Functions registered by atexit and on_exit have different signatures,
if registered with atexit no parameters are passed.
The function only prints a trace on nonzero exit(),
so can safely be disabled in musl builds.
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Khem Raj <raj.khem@gmail.com>
