In the commit 'initscripts: save /etc/timestamp with seconds accuracy',
a bashism was introduced in the bootmisc.sh script in the code to set
the current date from the stored value in /etc/timestamp. This causes
that operation to fail with the following message when /bin/sh is not
bash:
/etc/init.d/rc: /etc/rcS.d/S55bootmisc.sh: line 73: syntax error: bad substitution
Fixed by using pattern matching removal rather than bash-specific
substring expansion.
[YOCTO #6566]
(From OE-Core rev: c8a451adc71e4ce9ab963f61d7830c2d75aaffca)
Signed-off-by: Ben Shelton <ben.shelton@ni.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
bootlogd does not write to /var/log/boot if it does not exist,
so if using the volatiles facility (presumed to mount /var/log under a
tmpfs), ensure that /var/log/boot gets created.
(From OE-Core rev: 283d109098358c06ff10daaf612e5a244691d7ac)
Signed-off-by: Richard Tollerton <rich.tollerton@ni.com>
Signed-off-by: Ben Shelton <ben.shelton@ni.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
bootlogd's default log rotation code on stop requires `savelog`, which
is in debianutils, which may not be installed. If it's not installed,
don't try to perform the log rotation.
That is: in the affected code block, `savelog` is what is responsible
for creating "boot.0". When `savelog` doesn't exist, an error message
gets printed on bootup to the effect of "mv: can't find boot.0".
(From OE-Core rev: 5c22973e1bf76615bcf57984ac7a30cf7d0766df)
Signed-off-by: Richard Tollerton <rich.tollerton@ni.com>
Signed-off-by: Ben Shelton <ben.shelton@ni.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The messages echoed when starting and stopping bootlogd are currently
printed regardless of the setting of VERBOSE. Adjust the initscript so
they're only printed when VERBOSE is enabled.
(From OE-Core rev: 42f107513148aa6e9fd69d51e63a183c613114c0)
Signed-off-by: Richard Tollerton <rich.tollerton@ni.com>
Signed-off-by: Ben Shelton <ben.shelton@ni.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
When building out a systemd based image, the udev-hwdb postinstall
script always fails at rootfs time. This is because that the 'udevadm'
command used in udev-hwdb postinstall script is now in the udev
package and udev recommends udev-hwdb.
Instead of letting udev recommends udev-hwdb, we let systemd do it
and make udev-hwdb rdepend on udev.
(From OE-Core rev: f14ee284da19665f157453021e3fa0e3a0dbf75d)
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
During boot, there is a brief window during which /dev/initctl is
missing, which breaks initscripts that would need to access it. This
occurs because /etc/init.d/mountall.sh (rcS.d/S02...) attempts to ensure
/dev/initctl is present, but /etc/init.d/mdev (rcS.d/S06...) mounts over
/dev and clobbers the work done by mountall, and then does not wait
synchronously until initctl is ready before continuing.
To close this window, in /etc/init.d/mdev, we check whether /dev/initctl
is present, and if not, we remove it and recreate it. This is the same
thing that is done by /etc/init.d/mountall.sh, and we have verified that
any writers of /dev/initctl will wait synchronously until sysvinit
notices the change in fd and does the read, so no race exists.
(From OE-Core rev: 53543363a8ab4424c17ed7aec0e8aefc4df86b3d)
Signed-off-by: Richard Tollerton <rich.tollerton@ni.com>
Signed-off-by: Ben Shelton <ben.shelton@ni.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
/dev/mdev.seq exists to synchronize concurrently running instances of
mdev and to ensure that they execute in the proper order. Without this
synchronization, it is possible to have inconsistent mount points, to
leak device nodes, or to have a node erroneously removed in rapid
hotplug scenarios.
Enable the use of mdev.seq by creating an empty /dev/mdev.seq at boot.
(From OE-Core rev: 3b256653ae9ab2cf56b21cb82a8bd2a562397b27)
Signed-off-by: Richard Tollerton <rich.tollerton@ni.com>
Signed-off-by: Ben Shelton <ben.shelton@ni.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Otherwise, users cannot use named semaphores or shared memory.
(From OE-Core rev: a8f35be467836d8fa0bc8f8e85eba6f89a13f087)
Signed-off-by: Richard Tollerton <rich.tollerton@ni.com>
Signed-off-by: Ben Shelton <ben.shelton@ni.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The problem is that this would cause bash installed to meta-toolchain,
the dependencies chain is:
meta-toolchain -> eglibc-utils -> bash
We should eliminate bash dependencies in anything core to a tiny system.
The eglibc-utils also provides eglibc-utils/usr/bin/zdump,
eglibc-utils/usr/bin/zic which all of them belonged to tzcode, so we
should split these three utils as tzcode package.
[YOCTO #6544]
(From OE-Core rev: f7f3a1fb3855799bb48708f32533e93e336e6995)
Signed-off-by: Hongxu Jia <hongxu.jia@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
It was required by tzselect which is provided by eglibc utils.
The eglibc is a fundamental package, so enable math functions
of awk is reasonable.
Here is the error if we not enable that:
...
root@qemuarm:~# tzselect -c +4852+00220
awk: cmd. line:39: Math support is not compiled in
...
[YOCTO #6544]
(From OE-Core rev: 52a3df6a57c22df9da5ce05134395cf0756c63c6)
Signed-off-by: Hongxu Jia <hongxu.jia@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Source /etc/default/urandom if present. This allows the rootfs to
remain read-only while enabling the user to override the location of the
random seed file.
(From OE-Core rev: 415e1a4ac8120b28118671698459b098c965a4f6)
Signed-off-by: Richard Tollerton <rich.tollerton@ni.com>
Signed-off-by: Ben Shelton <ben.shelton@ni.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
inetutils and busybox hostname utils support `hostname -F`; coreutils
hostname doesn't. So just use `cat` instead.
(From OE-Core rev: acb8674e498468088d867ffae9a458caa08d95d5)
Signed-off-by: Richard Tollerton <rich.tollerton@ni.com>
Signed-off-by: Ben Shelton <ben.shelton@ni.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
If the seed file is empty or does not exist, the date is an extremely
poor backup source of entropy, but it is better than nothing.
(From OE-Core rev: 0a41a7c20316c7d3330233a624d8cf20ea5a81ae)
Signed-off-by: Richard Tollerton <rich.tollerton@ni.com>
Signed-off-by: Ben Shelton <ben.shelton@ni.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Currently, the random seed file location is hardcoded to
/var/lib/urandom/random-seed. Refactor it to a parameter
(RANDOM_SEED_FILE) so the file location is defined in only one place.
(From OE-Core rev: 558ba23cfdd60bf64b9214460a2772be70079796)
Signed-off-by: Richard Tollerton <rich.tollerton@ni.com>
Signed-off-by: Ben Shelton <ben.shelton@ni.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Currently, /etc/timestamp is saved with minutes accuracy. To increase
the accuracy, modify the save-rtc.sh and bootmisc.sh scripts to save and
read /etc/timestamp respectively with seconds accuracy.
(From OE-Core rev: 8fed53e4e72230c61f23cb36eda36c228aede1e0)
Signed-off-by: Richard Tollerton <rich.tollerton@ni.com>
Signed-off-by: Ben Shelton <ben.shelton@ni.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
busybox sysctl may lack the "-q" setting, so simulate it with redirects.
(From OE-Core rev: 5672b5188557b940340bdc3e6a3ac30835b829fb)
Signed-off-by: Richard Tollerton <rich.tollerton@ni.com>
Signed-off-by: Ben Shelton <ben.shelton@ni.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
If users accidently override the DISTRO_FEATURES in local.conf,
then build failures occur for lack of libc functions.
All features in DISTRO_FEATURES_LIBC should be in DISTRO_FEATURES.
That's why this patch let the recipe inherit distro_features_check
class and set REQUIRED_DISTRO_FEATURES to "${DISTRO_FEATURES_LIBC}".
[YOCTO #6381]
(From OE-Core rev: 82d9d12a3364914a5c1f354671adac1f0a71eea3)
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
It let the environment variable XML_CATALOG_FILES could work
which was required by xmlto.
[YOCTO #2416]
(From OE-Core rev: 1ea74a265b3c8f36e07c3cf2c26d8e60518da5a7)
Signed-off-by: Hongxu Jia <hongxu.jia@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
There are cases where software after boot may need to know the
current boot disk. Under the current system, it is not guaranteed
which disk is the boot. While /media/sda is a good guess, it
isn't always right, nor is it a good assumption that only one boot
disk is in the system. This gives a standard path to the original
boot disk mount which can be used to, for instance, update the
syslinux file on the boot media with a newer kernel, or updating
the boot parameters to add user options for future boots. Knowing
which disk is the boot media keeps from updating the non-boot
disk when for instance multiple syslinux boot medias are plugged in
(ie ensure correct syslinux is updated when the booted system is
updated).
(From OE-Core rev: 2be3b2607fd164d18498299dbfc020ff17dd2ca9)
Signed-off-by: Brian Lloyd <blloyd@familyhonor.net>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
(This patch was originally done against init-install.sh in
OE-Core rev 358f0584d779825307eec08c023b5ff14e72cf9e)
Previously, only unremovable hard drives are searched and are treated
as candidates of target disks to intall into.
However, it's possible that we're going to install the live image into
a removable media such as an USB. This patch enables this possibility.
In addition, this patch presents more information about the hard drives
so that user may have more knowledge about which hard drive they are
going to install their image into.
(From OE-Core rev: 7386acf4ab63a5959e4907b29459b767f2bf2fdb)
Signed-off-by: Drew Moseley <drew_moseley@mentor.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
(This patch was originally done against init-install.sh in
OE-Core rev aa67b1333b4774e1845f562085f7048df65a644f)
Previously, the boot partition was created for the target hard drive
but there was no corresponding entry for it in /etc/fstab. Besides,
even if the boot partition was mounted, it would just result in odd
directory hierarchy like /boot/boot/grub. However, what we really need
is /boot/grub. This patch fixes this problem.
Besides, for future maintance work, this patch also renames some of the
intermediate directories. It uses more descriptive names like /tgt_root
and /src_root. The name of /ssd is dropped.
(From OE-Core rev: 3b1bae7ad8d36930aae840175c6a3433c1469772)
Signed-off-by: Drew Moseley <drew_moseley@mentor.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Some mmc cards do not have all the data files in /sys/block
populated. Check for existence before displaying the files
to avoid erroring out of the install process.
(From OE-Core rev: 1d73e3f9d9977382efdb0c111c556c6048bd60b4)
Signed-off-by: Drew Moseley <drew_moseley@mentor.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Fixed the race issue:
util-macros.do_package failed: 'getpwnam(): name not found: root'
The error happens when there is a half etc/passwd, fixed by:
$ install usr/share/base-passwd/passwd.master etc/passwd.master
$ mv etc/passwd.master etc/passwd
The "mv" is atomic which will fix this problem.
[YOCTO #6124]
(From OE-Core rev: cdbe55215e3dd1b82a6c147a31c7c40186a8bf80)
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The test suite is installed into ${libexecdir} and as that may be outside of
${libdir} the QA test needs to be silenced.
(From OE-Core rev: d354d339a38d22f1a5bca14341229d2282647fbb)
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Add a '\n' to the last line of the file to fix:
No newline at end of file
(From OE-Core rev: b3090263ba31702631270643c7a7d7af8f4d9234)
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Add a hotkey for the GRUB 'test' menuentry. This can be used by expect scripts to boot into 'test' when doing runtime hardware tests.
(From OE-Core rev: 17b97fd6c724ba6e506cbadb18facdfd9c472e79)
Signed-off-by: Corneliu Stoicescu <corneliux.stoicescu@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The init-live.sh scripts assumes that the boot label set by
the LABELS variable is either "boot", "install", or
"install-efi". If that variable is overridden to something else
we fall off the end of the case statement and the system locks
up. If the boot label is unknown, at least attempt to boot.
(From OE-Core rev: 98353862c08be2f1724aaad7aa4ed0521e3621f2)
Signed-off-by: Drew Moseley <drew_moseley@mentor.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Now that udev-utils package has been removed and 'udevadm' has been
moved to udev package, the initramfs-framework should also get rid
of the udev-utils package.
(From OE-Core rev: f0b8f08e13f395820172d16c9d4c016a0fbdbfe9)
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The udev package RDEPENDS on the udev-utils package, and the actually
udev-utils package relies on the udev package because if there's no
udev daemon, 'udevadm' command doesn't make much sense.
Now that we have removed udev-utils package and moved 'udevadm' to
the udev package in the older udev recipe, we should do the same
for systemd.
(From OE-Core rev: 393867e8818b54255271bb70bcd7b6ffea137c0a)
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Previously, the udev package only RRECOMMENDS udev-utils package,
but its init script in the udev package really relies on the
'udevadm' command to work properly. As a result, if 'NO_RECOMMENDATIONS'
is set to '1', errors occur at system start-up.
The udev-utils package only contains one file, that is /bin/udevadm.
So it's better that we remove the udev-utils package and move the
'udevadm' command to the udev package.
[YOCTO #6388]
(From OE-Core rev: 4814a7a85ccff74923a554bdc11820fceff84e2e)
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Force probing of "platform" bus by default.
(From OE-Core rev: 6aae37c66fb6e7153b829ad860b7e7f94e804bd4)
Signed-off-by: Denys Dmytriyenko <denys@ti.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Make probing of "platform" bus conditional based on PROBE_PLATFORM_BUS
variable from /etc/default/udev-cache on subsequent boots when udev
cache is used. PROBE_PLATFORM_BUS has to be set to "yes" in order to
force probing on every boot, otherwise it uses the old default behaviour
of probing it just the first time.
This is helpful on modern SoCs where most of the low level peripheral
drivers are registered on the platform bus and need to be probed to load
the necessary modules and enable the connected buses and subsystems.
(From OE-Core rev: 70a695735e0a7d14448f2f5a9986bfe105210a91)
Signed-off-by: Denys Dmytriyenko <denys@ti.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
If a partition such as /tmp mounted on a volatile directory which create
by script populate-volatile.sh from initscripts, it will show errors.
In /var/log/boot, error message:
Thu Jun 19 05:39:09 2014: bootlogd.
Thu Jun 19 05:39:10 2014: rm: cannot remove '/tmp': Device or resource busy
Check volatile directories and if it is be mounted then don't force make
it as a link.
(From OE-Core rev: cc4b0936c7a6a1563dc88d62d8c9020791eaa446)
Signed-off-by: Kai Kang <kai.kang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
_EOF marker was not used properly
(space left before end of line).
(From OE-Core rev: 913fd224499c57c7596bd49e1eec5f570c3edf68)
Signed-off-by: Cristian Iorga <cristian.iorga@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Overwriting of /etc/mtab would fail as below if the /etc/mtab link already
exist during installation phase, this patch fix this problem by checking
existance of the link before try to overwrite it.
Error message during installation if the /etc/mtab exists:
"cat: /proc/mounts: input file is output file"
(From OE-Core rev: 5cd96e28825d345650be878d4b7be4fea2996839)
Signed-off-by: Cristian Iorga <cristian.iorga@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
For automated hardware testing, boot process control
via serial interface is needed. As such, in grub, serial
line support is added upon testmaster image install.
Also add a specific timeout to automatically start
the master image upon start of testing phase.
Tested on multiple hardware targets without issues.
(From OE-Core rev: 219228805a4d5d822894c8f6c2526e1b9a8609ff)
Signed-off-by: Cristian Iorga <cristian.iorga@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
We need use EXTRA_OECONF_append_class-native rather than
EXTRA_OECONF_class-native.
(From OE-Core rev: 5e1560b98680c79c979b8f9325f9f9997f9fd438)
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Found this error on CentOS 5.10, other distros may also has this
problem, I think that we can disable the selinux for dbus-native:
bus/selinux.c:327: error: array type has incomplete element type
[YOCTO #6374]
(From OE-Core rev: 7a2d2d81d8694806060f7bccfbdba42dc356d477)
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
The older kernel's linux/if_link.h doesn't have IFLA_VLAN_PROTOCOL, we need
check whether it has been defined or not.
The maintainer said that he would fix it:
http://thread.gmane.org/gmane.comp.sysutils.systemd.devel/18200
Also we need redefine IFLA_MAX from 34 to 35 when define IFLA_CARRIER,
otherwise there would be error:
| src/libsystemd/sd-rtnl/rtnl-types.c:233:9: error: array index in initializer exceeds array bounds
| [IFLA_CARRIER] = { .type = NLA_U8 },
[YOCTO #6380]
(From OE-Core rev: 0e626d5023fee4dbcc5d94e6b787b4c5fe4b2687)
Signed-off-by: Robert Yang <liezhi.yang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
This adds a binconfig-disabled class which can be used by recipes where
a -config file is installed but we wish to disable it and just rely on
the .pc files instead.
Rather than simply deleting it, we make the script "exit 1" so that it
can be found in PATH and raise a build error rather than something
silently falling back to the build system for example.
Rather than randomly finding -config files, this adds in the
specification of a list of binconfig scripts which is more deterministic
and maintainable moving forward.
This patch converts various users in OE-Core to use this, a world build
of OE-Core tests out ok with this change. There will likely be issues in
other layers however, hence this being a RFT.
(From OE-Core rev: 5870bd272b0b077d0826fb900b251884c1c05061)
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
If CONFIG_SYSLOGD is not enabled, then the related service file should
not be installed as it will always fail at system start-up. The error
message is as following.
[FAILED] Failed to start System Logging Service.
The same logic applies to CONFIG_KLOGD.
So we should first check the configuration before we install these
service files.
[YOCTO #5302]
(From OE-Core rev: b44e291a87539fbb8e6da1a16c56f425a417e7bd)
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
on rebuilds, dbus-native can detect an already existing dbus-glib and
tries to link some tests against it. As files were unstaged and due
to libtool insanities this fails with:
| /usr/bin/grep: .../sysroots/x86_64-oe-linux/usr/lib/libdbus-1.la: No such file or directory
| sed: can't read .../sysroots/x86_64-oe-linux/usr/lib/libdbus-1.la: No such file or directory
| x86_64-oe-linux-libtool: link: `.../sysroots/x86_64-oe-linux/usr/lib/libdbus-1.la' is not a valid libtool archive
| make[3]: *** [libdbus-testutils.la] Error 1
| make[3]: Leaving directory `.../work/x86_64-oe-linux/dbus-native/1.8.2-r0/build/test'
Make builds predicatable (and working) by disabling usage of dbus-glib
explicitly.
(From OE-Core rev: 0eeeb40d2af99448c3b94047f26d33a9983a221b)
Signed-off-by: Enrico Scholz <enrico.scholz@sigma-chemnitz.de>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
