| Age | Commit message (Collapse) | Author |
|---|
|
When the populate-volatile.sh initscript tests if a configured symlink
is already in place, it uses readlink with the '-f' (follow) option:
> [ "$(readlink -f $source)" = "$dest" ]
If the test fails, it proceeds to delete the exisiting folder/file, and
create the configured symlink.
However, the '-f' option to readlink makes it follow symlinks pointing
at symlinks. If the $dest argument is a symlink, the above test fails,
and warnings are printed due to changing a read-only rootfs.
This is the case for /tmp, and /etc/resolv.conf:
> /tmp -> /var/tmp -> /var/volatile/tmp
> /etc/resolv.conf -> /var/run/resolv.conf -> /run/resolv.conf
Fix the warnings by removing the '-f' option, so that the test matches
the configuration.
[ YOCTO #10814 ]
Signed-off-by: Martin Hundebøll <martin@geanix.com>
Signed-off-by: Jacob Kroon <jacob.kroon@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Rationale - excerp from `info chown`
====================================
OWNER‘:’GROUP
If the OWNER is followed by a colon and a GROUP (a group name or
numeric group ID), with no spaces between them, the group ownership
of the files is changed as well (to GROUP).
Some older scripts may still use ‘.’ in place of the ‘:’ separator.
POSIX 1003.1-2001 (*note Standards conformance::) does not require
support for that, but for backward compatibility GNU ‘chown’ supports
‘.’ so long as no ambiguity results. New scripts should avoid the use
of ‘.’ because it is not portable, and because it has undesirable
results if the entire OWNER‘.’GROUP happens to identify a user whose
name contains ‘.’.
Signed-off-by: Kosta Zertsekel <zertsekel@gmail.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Checking the requirements for each volatiles file in the
populate-volatiles script can be very slow when there are a large number
of volatiles files, easily consuming over 80% of the processing time.
These checks don't usually uncover any problems so concatenate all the
volatiles files together and process them as one large file for a "fast
path" option. This ensures that the penalty for checking the
requirements is only incurred once. In the event that checking the
requirements for the unified file fails, fall back to the slow process
of checking each one individually so that the offending one can be
skipped.
The core file is handled separately because it is responsible for
creating the temp directory used by check_requirements and thus must
always run first and without having its requirements checked.
[YOCTO #12949]
Signed-off-by: Joshua Watt <JPEWhacker@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
In some cases, it may be useful to populate a volatile file from an
existing one, e.g. a file in a read-only rootfs that may be edited in a
read-write destination.
To provide this behavior, creation of volatile files has been updated to
copy a file which has been given in the <linksource> field. If set to
none, the current behavior is preserved.
Signed-off-by: David Vincent <freesilicon@gmail.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
rm: can't remove '/etc/resolv.conf': Read-only file system
ln: /etc/resolv.conf: File exists
/etc/default/volatiles contains an entry:
l root root 0644 /etc/resolv.conf /var/run/resolv.conf
which causes populate-volatile.sh to execute the following in link_file():
if [ -L \"$2\" ]; then
[ \"\$(readlink -f \"$2\")\" != \"\$(readlink -f \"$1\")\" ] && { rm -f \"$2\"; ln -sf \"$1\" \"$2\"; };
elif [ -d \"$2\" ]; then
...
At the time the image is created, /etc/resolv.conf is already a
symlink to /var/run/resolv.conf, but at boot time when
populate-volatiles.sh is run, /var/run/resolv.conf doesn't
exist, causing it to try to rm -f and ln -sf which of course
fails due to the read-only filesystem.
[YOCTO #10814]
Signed-off-by: André Draszik <adraszik@tycoint.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
If commands are run asynchronously they may be completed out of
order causing problems if later entries depend on earlier ones.
Signed-off-by: Mans Rullgard <mans@mansr.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
This improves the config file parsing to permit blank lines and
comments following an entry or preceeded by whitespace.
Signed-off-by: Mans Rullgard <mans@mansr.com>
Signed-off-by: Ross Burton <ross.burton@intel.com>
|
|
Add ability to run "mount --bind" to populate-volatiles. Since several
programs use realpath to determine several full qualified file names,
there is no symlink to be resolved. So when speccing /run/lock - that's
the location - not /var/run/lock because of the program is smarter
than the operator/distributor.
See https://github.com/rehsack/meta-jens/blob/jethro/recipes-core/initscripts/initscripts/volatiles
for an example how to use the "b" feature.
Signed-off-by: Jens Rehsack <sno@netbsd.org>
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.
Signed-off-by: Kai Kang <kai.kang@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Small optimisation in /etc/init.d/populate-volatile.sh.
Replace:
cat <file> | sed -e xxx
By:
sed -e xxx <file>
Signed-off-by: Matthieu Crapet <Matthieu.Crapet@ingenico.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
So far the rules of check_requirements() is too strict to lead mismatch
when empty lines exist in volatiles.
Signed-off-by: Ming Liu <ming.liu@windriver.com>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Previously, dead links in target directory will not be copied.
This is incorrect as dead links are not uncommon in our rootfs.
So we use '-a' option instead.
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
If we're building a read-only rootfs, we'll get the following error now
and then.
tar: .: file changed as we read it
The root cause is that we spawn background process at rootfs time.
When the tar command is running, it's possible that files under rootfs
are changed by background processes, thus this error.
[YOCTO #4937]
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
It's possible that a 'No such directory' error occurs when doing
check_requirement in populate-volatile.sh at rootfs time. This is
because the $ROOT_DIR/var/tmp might be a dead link.
Use $ROOT_DIR/var/volatile/tmp as the TMPDIR instead to avoid this
error.
[YOCTO #4883]
[YOCTO #4103]
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
This avoids triple slashes in the generated /etc/volatile.cache to
reduce disk usage and in the output when verbose mode is enabled.
As all the paths for volatiles start with a slash, we can change
TNAME=${ROOT_DIR}/${TNAME} to TNAME=${ROOT_DIR}${TNAME}. To avoid
a double slash when ROOT_DIR is /, we strip the extra slash from
ROOT_DIR.
Signed-off-by: Jonathan Liu <net147@gmail.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
Remove trailing whitespaces.
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
Previously, if there's a link config item in the config file like
l root root 1777 /tmp /var/tmp
and /tmp has existed, the symlink will not be created correctly.
Another example is the /run directory. If /run directory has been
created by some recipe or script before populate-volatile.sh runs,
the symlink of /run to /var/run will not be created correctly.
This patch ensures that the system creates symlinks exactly as the
config file tells it.
[YOCTO #3404]
[YOCTO #3406]
Signed-off-by: Chen Qi <Qi.Chen@windriver.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
For populate-volatile.sh script to run correctly both at rootfs time and
at system boot time, it needs to be aware of which situation it is now in.
We use the ROOT_DIR variable to indicate whether it is run at rootfs time or
not. ROOT_DIR being "/" indicates that this script is run at system boot time,
otherwise, it is run at rootfs time.
Also, we ignore failures when running this script at rootfs time.
For example, if ${ROOT_DIR}/var/dir1 is symlink to /var/volatile/dir1, it's
possible that the link is a dead link. So if we're going to create some file
under ${ROOT_DIR}/var/dir1, it will fail. But the failure does no harm,
because this script will always run at system boot time to set up the correct
files and directories.
[YOCTO #3406]
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>
|
|
If a directory needs to be created to create something else in volatiles,
there's no guarantees on ordering due to the backgrounding. We can't guarantee
with certainty that the create directory commands are complete before the
later ones run. This ensures that we wait for directory creations to complete
before we proceed.
Chris Hallinan hit an actual failure due to this back in March of last year.
Signed-off-by: Christopher Larson <chris_larson@mentor.com>
Signed-off-by: Saul Wold <sgw@linux.intel.com>
|
|
If a machine loses power while building the volatiles cache, it will
continue to operate with an incomplete set of volatiles. Fix this by
updating atomically.
Signed-off-by: James Limbouris <james@digitalmatter.com.au>
Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
|
|
Having one monolithic packages directory makes it hard to find things
and is generally overwhelming. This commit splits it into several
logical sections roughly based on function, recipes.txt gives more
information about the classifications used.
The opportunity is also used to switch from "packages" to "recipes"
as used in OpenEmbedded as the term "packages" can be confusing to
people and has many different meanings.
Not all recipes have been classified yet, this is just a first pass
at separating things out. Some packages are moved to meta-extras as
they're no longer actively used or maintained.
Signed-off-by: Richard Purdie <rpurdie@linux.intel.com>
|
Martin Hundebøll
Kosta Zertsekel
Joshua Watt
David Vincent
André Draszik
Mans Rullgard
Jens Rehsack
Kai Kang
Matthieu Crapet
Ming Liu
Chen Qi
Jonathan Liu
Richard Purdie
Christopher Larson
James Limbouris
Richard Purdie