sed(1) in busybox does not support this functionality:
https://git.savannah.gnu.org/cgit/sed.git/tree/sed/execute.c#n598
This causes /etc/group to become mislabeled when a package requests
that a uid/gid be added on OpenWrt with SELinux
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
[move restorecon inside lock]
Signed-off-by: Dominick Grift <dominick.grift@defensec.nl>
Commit ecbcc0b595 bricks devices on which the raw kernel and UBI mtd
partitions overlap.
This is the case of the ZyXEL NR7101 for example. Its OEM bootloader has
no UBI support. OpenWrt splits the stock kernel mtd partition into a raw
kernel part used by the bootloader and a UBI part used to store rootfs
and rootfs_data. Running mtd erase on the complete partition during
sysupgrade erases the UBI part and results in a soft brick.
Arguably the best solution would be to fix the partition layouts so that
kernel and UBI partitions do not overlap, also including a stock_kernel
partition to help reverting to stock firmware. This would have the added
benefit of protecting UBI from kernel images that are excessively large.
Fixes: ecbcc0b595 ("base-files: safer sysupgrade.tar for kernel-out-of-UBI")
Reported-by: Bjørn Mork <bjorn@mork.no>
Signed-off-by: Rodrigo Balerdi <lanchon@gmail.com>
Attempt to minimize the time during which an interrupted nand sysupgrade
can lead to a non-functional device by flushing caches before starting
the upgrade procedure.
Signed-off-by: Rodrigo Balerdi <lanchon@gmail.com>
Fix issues while retaining configuration during nand sysupgrade:
- abort configuration saving if data partition is not found
- generate diagnostics if saving fails (eg, because of lack of space)
- do not output "sysupgrade successful" in case of errors
Signed-off-by: Rodrigo Balerdi <lanchon@gmail.com>
Remove redundant check from nand ubinized sysupgrade code. This check
has already been done in the only caller of the affected function:
nand_do_upgrade.
Signed-off-by: Rodrigo Balerdi <lanchon@gmail.com>
Prepares code for ubirename-based safe sysupgrade implementation.
Fixes several issues:
- the special CI_KERNPART value "none" is ignored if an MTD partition
named "none" exists
- misleading variable names (such as has_kernel to mean "tar has kernel
and it should not be written to an MTD partition but a UBI volume")
- inconsistent treatment of zero-length tar member files
- inconsistent meaning of "0" and "" variable values
- redundant operations (unneeded untaring, repeated untaring, unneeded
partition lookups)
- inconsistent variable quoting
Signed-off-by: Rodrigo Balerdi <lanchon@gmail.com>
Ensure that the kernel CRC is invalidated while rootfs is being updated.
This allows the bootloader to detect an interrupted sysupgrade and fall
back to an alternate booting method, such as TFTP, instead of just going
ahead with normal boot and effectively bricking the device.
Signed-off-by: Rodrigo Balerdi <lanchon@gmail.com>
Ensure that the kernel CRC is invalidated while rootfs is being updated.
This allows the bootloader to detect an interrupted sysupgrade and fall
back to an alternate booting method, instead of just going ahead with
normal boot and effectively bricking the device.
Possible fallbacks include a recovery initramfs partition or UBI volume
and TFTP. See here for an example U-Boot configuration with fallbacks:
https://shorturl.at/befsA (https://github.com/Lanchon/openwrt-tr4400-v2/
blob/e7d707d6bd7839fbd0b8d0bd180fce451df77e47/install-recovery.sh#L52-L63)
Signed-off-by: Rodrigo Balerdi <lanchon@gmail.com>
Emit diagnostics if nand sysupgrade is aborted because UBI partition
cannot be attached. Also avoid redudndant checks.
Signed-off-by: Rodrigo Balerdi <lanchon@gmail.com>
Make sure sysupgrade on NAND also works in case of UBI volumes having
index >9. While at it, also make sure UBI device is detected and abort
in case it isn't. Use Shell built-in shorthand ':' instead of 'true'.
Fixes#9708
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
See firmware-utils.git commits [1], which implemented the cros-vbutil
verified-boot payload-packing tool, and extended ptgen for the CrOS
kernel partition type. With these, it's now possible to package kernel +
rootfs to make disk images that can boot a Chrome OS-based system (e.g.,
Chromebooks, or even a few AP models).
Regarding PARTUUID= changes: Chromium bootloaders work well with a
partition number offset (i.e., relative to the kernel partition), so
we'll be using a slightly different root UUID line.
NB: I've made this support specific to ip40xx for now, because I only
plan to support an IPQ4019-based AP that uses a Chromium-based
bootloader, but this image format can be used for essentially any
Chromebook, as well as the Google OnHub, a prior Chromium-based AP using
an IPQ8064 chipset.
[1]
ptgen: add Chromium OS kernel partition support
https://git.openwrt.org/?p=project/firmware-utils.git;a=commit;h=6c95945b5de973026dc6f52eb088d0943efa96bb
cros-vbutil: add Chrome OS vboot kernel-signing utility
https://git.openwrt.org/?p=project/firmware-utils.git;a=commit;h=8e7274e02fdc6f2cb61b415d6e5b2e1c7e977aa1
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
A service managed by procd does have a json object with usefull information.
This information could by dumped with the following command.
ubus call service list "{ 'verbose':true, 'name': '<service-name>)'". }"
This line is long and complicated to enter. This commit adds a wrapper
call to the procd service section tool to simplify the input and get the
output faster.
We could now enter the command /etc/initd/<service> info to get the info
faster.
Signed-off-by: Florian Eckert <fe@dev.tdt.de>
The service command belongs to the procd and does not belong in the
shinit. In the course of the move, the script was also checked with
shellcheck and cleaned up.
Signed-off-by: Florian Eckert <fe@dev.tdt.de>
Rootfs overlays get created at a ROOTDEV_OVERLAY_ALIGN (64KiB)
alignment after the rootfs, but emmc_do_upgrade() is assuming
it comes at the very next 512-byte sector.
Suggested-by: Christian Lamparter <chunkeey@gmail.com>
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
(move spaces around, mention fstools' libtoolfs)
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
OpenWrt uses a lot of (b)ash scripts for initial setup. This isn't the
best solution as they almost never consider syncing files / data. Still
this is what we have and we need to try living with it.
Without proper syncing OpenWrt can easily get into an inconsistent state
on power cut. It's because:
1. Actual (flash) inode and data writes are not synchronized
2. Data writeback can take up to 30 seconds (dirty_expire_centisecs)
3. ubifs adds extra 5 seconds (dirty_writeback_centisecs) "delay"
Some possible cases (examples) for new files:
1. Power cut during 5 seconds after write() can result in all data loss
2. Power cut happening between 5 and 35 seconds after write() can result
in empty file (inode flushed after 5 seconds, data flush queued)
Above affects e.g. uci-defaults. After executing some migration script
it may get deleted (whited out) without generated data getting actually
written. Power cut will result in missing data and deleted file.
There are three ways of dealing with that:
1. Rewriting all user-space init to proper C with syncs
2. Trying bash hacks (like creating tmp files & moving them)
3. Adding sync and hoping for no power cut during critical section
This change introduces the last solution that is the simplest. It
reduces time during which things may go wrong from ~35 seconds to
probably less than a second. Of course it applies only to IO operations
performed before /etc/init.d/boot . It's probably the stage when the
most new files get created.
All later changes are usually done using smarter C apps (e.g. busybox or
uci) that creates tmp files and uses rename() that is expected to be
atomic.
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
Acked-by: Hauke Mehrtens <hauke@hauke-m.de>
Acked-by: Sergey Ryazanov <ryazanov.s.a@gmail.com>
In the default shadow file, as visible in the failsafe mode, the user
root has value of `0` set in the 3rd field, the date of last password
change. This setting means that the password needs to be changed the
next time the user will log in the system. `dropbear` server is ignoring
this setting but `openssh-server` tries to enforce it and fails in the
failsafe mode because the rootfs is R/O.
Disable the password aging feature for user root by setting the 3rd
filed empty.
Signed-off-by: Rucke Teg <rucketeg@protonmail.com>
Not all targets create /var/lock or touch /var/lock/fw_printenv.lock in
their platform.sh. This is problematic as fw_printenv then fails in
case /var/lock/fw_printenv.lock has not been created by previous calls
to fw_printenv/fw_setenv before sysupgrade is run.
Targets using fw_printenv/fw_setenv during sysupgrade:
* ath79/*
* ipq40xx/*
* ipq806x/*
* kirkwood/*
* layerscape/*
* mediatek/mt7622
* mvebu/*
* ramips/*
* realtek/*
Targets currently using additional steps in /lib/upgrade/platform.sh
to make sure /var/lock/fw_printenv.lock (or at least /var/lock)
actually exists:
* ath79/* (openmesh devices)
* ipq40xx/* (linksys devices)
* ipq806x/* (linksys devices)
* kirkwood/* (linksys devices)
* layerscape/*
* mvebu/cortexa9 (linksys devices)
Given that accessing the U-Boot environment during sysupgrade is not
uncommon and the situation across targets is currently quite diverse,
just make sure both tools as well fw_env.config are always copied to
the ramdisk used for sysupgrade. Also make sure /var/lock always
exists.
This now allows to remove copying of fw_printenv/fw_setenv as well as
fw_env.config, creation of /var/lock or even /var/lock/fw_printenv.lock
from lib/upgrade/platform.sh or files included there.
As the same applies also to 'fwtool' which is used by generic eMMC
sysupgrade, also always copy that to ramdisk.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
This patch adds support for creation heartbeat led trigger with,
for example, this command:
ucidef_set_led_heartbeat "..." "..." "..."
from /etc/board.d/01_leds.
Signed-off-by: Alexey Smirnov <s.alexey@gmail.com>
For sysupgrade on NAND/UBI devices there is the U-Boot environment
variable rootfs_data_max which can be used to limit the size of the
rootfs_data volume created on sysupgrade.
This stopped working reliable with recent kernels, probably due to a
race condition when reading the number of free erase blocks from sysfs
just after removing a volume.
Change the script to just try creating rootfs_data with the desired
size and retry with maximum size in case that fails. Hence calculating
the available size in the script can be dropped which works around the
problem.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
fgrep is deprecated and replaced by grep -F. The latter is used
throughout the tree whereas this is the only usage of the former.
Signed-off-by: Rosen Penev <rosenp@gmail.com>
We were missing (not using) the last sector of each partition,
compared with the output of gparted.
Signed-off-by: Javier Marcet <javier@marcet.info>
[moved the dot]
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
The following command checks if a instance of a service is running.
/etc/init.d/<service> running <instance>
In the variable `$@`, which is passed to the function
`service_running`, the first argument is always the `instance` which
should be checked. Because all other variables where removed from `$@`
with `shift`.
Before this change the first argument of `$@` was set to the `$service`
Variable. So the function does not work as expected. The `$service`
variable was always the instance which should be checked. This is not
what we want.
Signed-off-by: Florian Eckert <fe@dev.tdt.de>
Reviewed-by: Sungbo Eo <mans0n@gorani.run>
Adds generic support for sysupgrading on eMMC-based devices.
Provide function emmc_do_upgrade and emmc_copy_config to be used in
/lib/upgrade/platform.sh instead of redundantly implementing the same
logic over and over again.
Similar to generic sysupgrade on NAND, use environment variables
CI_KERNPART, CI_ROOTPART and newly introduce CI_DATAPART to indicate
GPT partition names to be used. On devices with more than one MMC
block device, CI_ROOTDEV can be used to specify the MMC device for
partition name lookups.
Also allow to select block devices directly using EMMC_KERN_DEV,
EMMC_ROOT_DEV and EMMC_DATA_DEV, as using GPT partition names is not
always an option (e.g. when forced to use MBR).
To easily handle writing kernel and rootfs make use of sysupgrade.tar
format convention which is also already used for generic NAND support.
Signed-off-by: Enrico Mioso <mrkiko.rs@gmail.com>
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
CC: Li Zhang <li.zhang@gl-inet.com>
CC: TruongSinh Tran-Nguyen <i@truongsinh.pro>
Currently nand_upgrade_tar() will pass the kernel length
to nand_upgrade_prepare_ubi() in all cases except for when
the kernel is to be installed in a separate partition as a
binary with the MTD tool.
While this is fine for almost all cases newer MikroTik NAND
devices like hAP ac3 require the kernel to be installed as a
UBIFS packed UBI volume in its own partition.
So, since we have a custom recipe to use ubiformat to flash
the kernel in its partition it makes no sense for sysupgrade
to also install the kernel as a UBI volume in the "ubi"
partition as it only wastes space and will never be used.
So, simply check whether CI_KERNPART is set to "none" and
if so unset the "has_kernel" variable which will in turn
prevent the kernel length from being passed on and then
the kernel UBI volume wont be created for no usefull purpose.
The ath79 MikroTik NAND target has been setting CI_KERNPART
to "none" for a while now altough that was not preventing
the kernel to be installed as UBI volume as well.
Signed-off-by: Robert Marko <robimarko@gmail.com>
Simply reading /proc/*/stat as a space-separated string will not work
as the process name may itself contain spaces. Hence we must match on
the '(' and ')' characters around the process name and can then handle
the remaining string as space-separated values.
This fixes shell error messages which have been popping up the console
due to spaces in process names being interpreted as field separators.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
find_mmc_part provides a better alternative and all users of
get_partition_by_name have been removed.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Some devices got more than one mmc device.
Allow specifying the root device as 2nd parameter of find_mmc_part so
scripts can avoid matching irrelevant partitions on wrong mmc device.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Added minimal mmc support for helper functions:
- find_mmc_part: Look for a given partition name. Returns the
coresponding partition path
- caldata_extract_mmc: Look for a given partition name and then
extracts the calibration data
- mmc_get_mac_binary: Returns the mac address from a given partition
name and offset
Signed-off-by: Davide Fioravanti <pantanastyle@gmail.com>
Signed-off-by: Robert Marko <robimarko@gmail.com>
[replace dd with caldata_dd, moved sysupgrade mmc to orbi]
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
Some packages may require additional group membership for the system
user added by that package. Allow defining additional groups as third
member of the ':'-separated tuple, allowing to specify multiple
','-separated groups with optional GID.
Example:
USERID:=foouser=1000:foogroup=1000:addg1=1001,addg2=1002,addg3
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Per FHS 3.0, /var/lock is the location for lock files [1].
However its current permissions (755) are too restrictive
for use by unprivileged processes.
Debian and Ubuntu set them to 1777, and now so do we.
[1] <https://refspecs.linuxfoundation.org/FHS_3.0/fhs-3.0.html#varlockLockFiles>
Signed-off-by: Deomid Ryabkov <rojer@rojer.me>
[fixed typo in commit message, had to remove "rojer" due to git hooks]
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
The `mkdir` commands supports passing multiple arguments to batch create
multiple folders, instead of calling the tool every single time.
If the creation of one of the folders fails, all other folder are still
created and therefore doesn't change the error handling.
Also stop creating `/etc/` explicitly after subfolders of `/etc/` were
already created.
Signed-off-by: Paul Spooren <mail@aparcar.org>
The `sed`-script shouldn't be called multiple times, especially not with
the same files.
This commit merges all files together in a single `sed`-script call.
Signed-off-by: Paul Spooren <mail@aparcar.org>
The option was initially named TARGET_ROOTFS_LN_VAR_TMP, and the check
was correct. When renaming the option to something more suitable, the
check was changed to check for n, but when an option is not set, it's
not n but empty. This results in the check always evaluating to false.
Fix the check by checking for y with ifneq.
Fixes: 57807f50de ("base-files: add option to make /var persistent")
Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>
In OpenWrt, /var is symlinked to /tmp by default. This is done to reduce
the amount of writes to the flash chip, which often have not the
greatest durability. As a result, things like DHCP or UPnP lease files,
are not persistent across reboots.
Since OpenWrt can run on devices with more durable storage, it makes
sense to have an option for a persistent /var. Add an option to make
/var persistent. When enabled, /var will no longer be symlinked to /tmp,
but /var/run will be symlink to /tmp/run, as it should contains only
files that should not be kept during reboot. The option is off by
default, to maintain the current behaviour.
Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>
While an image layout based on MBR and 'bootfs' partition may be easy
to understand for users who are very used to the IBM PC and always have
the option to access the SD card outside of the device (and hence don't
really depend on other recovery methods or dual-boot), in my opinion
it's a dead end for many desirable features on embedded systems,
especially when managed remotely (and hence without an easy option to
access the SD card using another device in case things go wrong, for
example).
Let me explain:
* using a MSDOS/VFAT filesystem to store kernel(s) is problematic, as a
single corruption of the bootfs can render the system into a state
that it no longer boots at all. This makes dual-boot useless, or at
least very tedious to setup with then 2 independent boot partitions
to avoid the single point of failure on a "hot" block (the FAT index
of the boot partition, written every time a file is changed in
bootfs). And well: most targets even store the bootloader environment
in a file in that very same FAT filesystem, hence it cannot be used
to script a reliable dual-boot method (as loading the environment
itself will already fail if the filesystem is corrupted).
* loading the kernel uImage from bootfs and using rootfs inside an
additional partition means the bootloader can only validate the
kernel -- if rootfs is broken or corrupted, this can lead to a reboot
loop, which is often a quite costly thing to happen in terms of
hardware lifetime.
* imitating MBR-boot behavior with a FAT-formatted bootfs partition
(like IBM PC in the 80s and 90s) is just one of many choices on
embedded targets. There are much better options with modern U-Boot
(which is what we use and build from source for all targets booting
off SD cards), see examples in mediatek/mt7622 and mediatek/mt7623.
Hence rename the 'sdcard' feature to 'legacy-sdcard', and prefix
functions with 'legacy_sdcard_' instead of 'sdcard_'.
Tested-by: Stijn Tintel <stijn@linux-ipv6.be>
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Add a generic sdcard upgrade method instead of duplicating code in yet
another target, and add a feature flag to only install this upgrade
method in targets that set this flag. Copied from mvebu.
Signed-off-by: Stijn Tintel <stijn@linux-ipv6.be>
Try umount on device mapper and loop devices still mounted, so the
subsequent call to disactivate all physical volumes and delete all
loop devices is more likely to succeed.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
If the busybox applet losetup was selected, `command -v` selects that
during sysupgrade. As this applet is in another path and doesn't cover
the '-D' option which is used to make sure user-defined loop devices
are no longer active during sysupgrade.
Detect losetup at the path of the full utility to avoid error messages
in case of the busybox applet being selected.
Reported-by: fda77 <fda77@users.noreply.github.com>
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
commit 5edbd390d321532d9a697d6895a1a7c71c40bd5d rearranged the
"wifi up" code.
This commit tidies up the "wifi reconf" code so as to
keep it aligned with the "wifi up" code.
branches affected: trunk, 21.02
Signed-off-by: Bob Cantor <coxede6557@w3boats.com>
"/sbin/wifi up" makes three ubus calls:
1. ubus call network reload
2. ubus call network.wireless down
3. ubus call network.wireless up
The first and third ubus calls call drv_mac80211_setup,
while the second ubus call triggers wireless_device_setup_cancel,
so the call sequence becomes,
1. drv_mac80211_setup
2. wireless_device_setup_cancel
3. drv_mac80211_setup
This commit swaps the order of the first two ubus calls,
1. ubus call network.wireless down
2. ubus call network reload
3. ubus call network.wireless up
Consequently drv_mac80211_setup is only called once,
and two related bugs (#FS3784 and #FS3902) are no longer triggered
by /sbin/wifi.
branches affected: trunk, 21.02
Signed-off-by: Bob Cantor <coxede6557@w3boats.com>
The zoneinfo packages are not installed per default so neither
/tmp/localtime nor /tmp/TZ is generated.
This patch mostly reverts the previous fix and instead incooperates a
solution suggested by Jo.
Fixes "base-files: fix zoneinfo support " 8af62ed
Signed-off-by: Paul Spooren <mail@aparcar.org>
The system init script currently sets /tmp/localinfo when zoneinfo is
populated. However, zoneinfo has spaces in it whereas the actual files
have _ instead of spaces. This made the if condition never return true.
Example failure when removing the if condition:
/tmp/localtime -> /usr/share/zoneinfo/America/Los Angeles
This file does not exist. America/Los_Angeles does.
Ran through shfmt -w -ci -bn -sr -s
Signed-off-by: Rosen Penev <rosenp@gmail.com>
Vlan subinterface was never brought up when using vlan-based preinit network.
Tested forcing ifname="" before preinit_ip() on a Tp-Link Archer C5v4.
Signed-off-by: Luiz Angelo Daros de Luca <luizluca@gmail.com>
Some interfaces have a VLAN modifier like :t in lan1:t, this modifier
should be removed from the interface before calling preinit_ip_config().
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Adapt the preinit_config_board() to the board.json network changes. It
now looks for the device and the ports variables to configure the LAN
network.
This works with swconfig configurations.
Fixes: FS#3866
Fixes: d42640e389 ("base-files: use "ports" array in board.json network for bridges")
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de>
Reviewed-by: Rafał Miłecki <rafal@milecki.pl>
There are services that have only STOP value set. They are executed only
on shutdown and it is common to use them for system cleanup. There is
one such service shipped directly with base-files, it is 'umount'. Those
work the same way as those with START but enabled does not report them
as enabled although it should have as they can be enabled and disabled
as any other service.
This also changes check from check for executable to check for symbolic
link. The implementation depends on those being links to service file
and it is much cleaner and direct to check for them being links.
Signed-off-by: Karel Kočí <karel.koci@nic.cz>
If one of the programmes is not running, then we see the following
output in the logs.
`killall: telnetd: no process killed`
To ensure that the log is clean, redirect the output to /dev/null
Signed-off-by: Florian Eckert <fe@dev.tdt.de>
The remaining vn calls have been ported to v.
Therefore, these functions are no longer needed and will be removed.
Signed-off-by: Florian Eckert <fe@dev.tdt.de>
The logging output should not only be displayed in the calling shell
session but also in the syslog. A sysupgrade and a configuration
import, export can thus be traced in the syslog.
Signed-off-by: Florian Eckert <fe@dev.tdt.de>
Calling `switch_to_ramfs()` will not copy the gzip executable
(/bin/gzip) to ramfs, but `/bin/zcat` will call `/bin/gzip` when
package gzip is installed, instead of the busybox-supplied zcat.
This will cause `zcat` to fail to find `gzip`, then cause the
sysupgrade to fail. Adding the `busybox` prefix here will solve
the problem.
Signed-off-by: Chuck Fan <fanck0605@qq.com>
bzip2 adds about 8kb of size. For tiny builds it's often disabled.
It's not directly used by stock OpenWrt programs.
Kernel images compressed with bzip2 are also not fully supported.
Signed-off-by: Sergey Ponomarev <stokito@gmail.com>
[fix \ indention]
Signed-off-by: Paul Spooren <mail@aparcar.org>
This restores the original config_generate behaviour. With MAC set for
bridged devices the bridge automatically gets its MAC adjusted (it picks
the lowest MAC of bridged devices).
This fixes confusing interfaces setup (bridge ports not having custom
MAC assigned).
Reported-by: Koen Vandeputte <koen.vandeputte@citymesh.com>
Fixes: e002179a6d ("base-files: simplify setting device MAC")
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
These processes are managed by procd and set to start again when killed
via the procd instance parameter "respawn" being set during init.
Example:
procd_set_param respawn 3600 1 0
When they are killed manually during sysupgrade,
they are started again in 5 seconds or less, depending on
how the "respawn" parameter is set.
Use procd through ubus to disable the instances that respawn them,
however, allow dnsmasq, netifd, and logd to restart for remote logging.
Properly closing all these processes increases free memory by about 3 MB,
which should help low memory devices upgrade without crashing.
For very low memory devices (set to 32 MB for now)
also kill dnsmasq, netifd, and logd for an additional 3 MB of free memory.
Also, bump sleep values to allow at least 10 seconds
for network interfaces and daemons
to come up after they are killed and restarted
before caches are dropped.
Signed-off-by: Michael Pratt <mcpratt@pm.me>
1. Move code above interface generation
It results in more logical order. Device gets its config section
above interface section.
2. Drop the loop
We have separated code handling bridges now so $device should be
guaranteed to contain a single device name.
3. Drop section name
It's not required by netifd or LuCI & it's not needed by this script
as $device contains a single device name now.
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
Replace "ifname" with "device" as netifd has been recently patches to
used the later one. It's more clear and accurate.
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
After the commit 43fc720657
("base-files: generate "device UCI type section for bridge"), the wrong
network configuration is generated for the devices that already have the
bridge device section for VLAN, such as the devices in realtek target.
As a result, the bridge device by additional "device" section is
specified to the "ports" option in the "bridge-vlan" section and netifd
shuts down the switch and the ethernet when the network service started.
Fixes: 43fc720657 ("base-files: generate "device" UCI type section for bridge")
Signed-off-by: INAGAKI Hiroshi <musashino.open@gmail.com>
[rmilecki: use $ports for generate_bridge_vlan argument]
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
This reverts commit f716c30241.
Migrating everyone to the new syntax could break downgrades. We may
reintroduce it way later if needed.
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
Missing br- prefix could result in name conflict between DSA port
interface and bridge interface. Some devices with just one LAN port use
"lan" interface name for DSA port. Trying to create bridge with the same
"lan" name was failing.
Reported-by: David Bauer <mail@david-bauer.net>
Fixes: 43fc720657 ("base-files: generate "device" UCI type section for bridge")
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
/etc/syslog.conf is used by sysklogd, and /etc/crontabs is used
by crond, both features of busybox. Given this, ownership for
these files should be bound to busybox, especially if one day
there's a way to do an in-place opkg update of busybox.
There's also the busybox provided syslogd which uses this file
if CONFIG_BUSYBOX_FEATURE_SYSLOGD_CFG is set.
Signed-off-by: Philip Prindeville <philipp@redfish-solutions.com>
netifd has been recently patched to use more accurate "ports" option
instead of "ifname". This is a simple translation between two UCI
options.
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
This switches from the old way of defining bridges in an "interface" UCI
section type (that should be used for layer 3 only). From now a defualt
board switch will have its own "device" UCI section type. It's a new &
preferred way of defining L2 devices.
Before:
config interface 'lan'
option type 'bridge'
option ifname 'lan1 lan2 lan3 lan4'
option proto 'static'
option ipaddr '192.168.1.1'
option netmask '255.255.255.0'
After:
config device
option name 'lan'
option type 'bridge'
list ports 'lan1'
list ports 'lan2'
list ports 'lan3'
list ports 'lan4'
config interface 'lan'
option ifname 'lan'
option proto 'static'
option ipaddr '192.168.1.1'
option netmask '255.255.255.0'
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
Bridge aggregates multiple ports so use a more accurate name ("ports")
and format (array) for storing them in board.json.
Example:
"network": {
"lan": {
"ports": [
"lan1",
"lan2",
"lan3",
"lan4"
],
"protocol": "static"
},
"wan": {
"ifname": "wan",
"protocol": "dhcp"
}
}
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
Remove vn call in favour of v call. This commit serves as preparation
for removing the v function call.
Signed-off-by: Florian Eckert <fe@dev.tdt.de>
[alter slightly to prevent double space after colon]
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
Before this commit, it was assumed that mkhash is in the PATH. While
this was fine for the normal build workflow, this led to some issues if
make TOPDIR="$(pwd)" -C "$pkgdir" compile
was called manually. In most of the cases, I just saw warnings like this:
make: Entering directory '/home/.../package/gluon-status-page'
bash: line 1: mkhash: command not found
bash: line 1: mkhash: command not found
bash: line 1: mkhash: command not found
bash: line 1: mkhash: command not found
bash: line 1: mkhash: command not found
bash: line 1: mkhash: command not found
bash: line 1: mkhash: command not found
bash: line 1: mkhash: command not found
[...]
While these were only warnings and the package still compiled sucessfully,
I also observed that some package even fail to build because of this.
After applying this commit, the variable $(MKHASH) is introduced. This
variable points to $(STAGING_DIR_HOST)/bin/mkhash, which is always the
correct path.
Signed-off-by: Leonardo Mörlein <me@irrelefant.net>
Users of devices with large block storage may choose to have an LVM
partition on the same device which is used for booting OpenWrt.
The presents a problem during sysupgrade as the root device is then
still busy and changing partitions will not work as desired,
leading to data corruption in case the newly flashed image is larger
than the currently installed one.
Having loop devices setup causes similar havoc.
Make sure all volume groups are offline and all loop devices have been
released before sysupgrade.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
If service() is called w/o parameter then the status display for services
with multiple instances is incorrect. E.g. samba4 or wpad have 2 instances.
root@OpenWrt:~# /etc/init.d/samba4 status
running
root@OpenWrt:~# /etc/init.d/wpad status
running
Before change:
/etc/init.d/samba4 enabled stopped
/etc/init.d/wpad enabled stopped
After change:
/etc/init.d/samba4 enabled running
/etc/init.d/wpad enabled running
Signed-off-by: Alexander Egorenkov <egorenar-dev@posteo.net>
Introduce cmdline_get_var() to /lib/function.sh and make use of it in
export_rootdev() in /lib/upgrade/common.sh, making the code more
simple and removing one level of indentation.
Introduce get_partition_by_name() to /lib/upgrade/common.sh which is
useful on non-EFI GPT platforms like mt7622.
Remove some dead-code while at it.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
So far, board.d files were having execute bit set and contained a
shebang. However, they are just sourced in board_detect, with an
apparantly unnecessary check for execute permission beforehand.
Replace this check by one for existance and make the board.d files
"normal" files, as would be expected in /etc anyway.
Note:
This removes an apparantly unused '#!/bin/sh /etc/rc.common' in
target/linux/bcm47xx/base-files/etc/board.d/01_network
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
For now we have only kernel LED trigger support. With this change it is now
possible to use application triggers.
If we configure a LED with a non kernel trigger, then we check on every
restart and boot of the LED service if we have this trigger as an application
in "/usr/libexec/led-trigger". If this file with the name is found, then we
execute this to init the LED.
Possible use cases are:
- Start/Stop/Restart an application led trigger service for this led
- Init a LED that is configured by a hotplug script (VPN tunnel established)
Signed-off-by: Florian Eckert <fe@dev.tdt.de>
Instead of only relying in /sysupgrade.tgz being present in rootfs to
restore configuration, also grab /tmp/sysupgrade.tar which may have
magically gotten there during preinit...
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
When using Shell arithmetric evaluation via $((..)) the variables in
the expression do not need to be prefixed by the '$' sign.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Check if firmware environment variable 'rootfs_data_max' exists and is
set to a numerical value greater than 0. If so, limit rootfs_data
volume to that size instead of using the maximum available size.
This is useful on devices with lots of flash where users may want to
have eg. a volume for persistent logs and statistics or for external
applications/containers. Persistence on rootfs overlay is limited by
the size of memory available during the sysugprade process as that
data needs to be copied to RAM while the volume is being recreated
during sysupgrade. Hence it is unsuitable for keeping larger amounts
of data accross upgrade which makes additional volume(s) for
application data desirable.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Allow for single (external-data) FIT image to hold kernel, dtb and
squashfs. In that way, the bootloader verifies the system integrity
including the rootfs, because what's the point of checking that the
hash of the kernel is correct if it won't boot in case of squashfs
being corrupted? Better allow bootloader to check everything needed
to make it at least up to failsafe mode. As a positive side effect
this change also makes the sysupgrade process on nand potentially
much easier as it is now.
In short: mkimage has a parameter '-E' which allows generating FIT
images with 'external' data rather than embedding the data into the
device-tree blob itself. In this way, the FIT structure itself remains
small and can be parsed easily (rather than having to page around
megabytes of image content). This patch makes use of that and adds
support for adding sub-images of type 'filesystem' which are used to
store the squashfs. Now U-Boot can verify the whole OS and the new
partition parsers added in the Linux kernel can detect the filesystem
sub-images, create partitions for them, and select the active rootfs
volume based on the configuration in FIT (passing configuration via
device tree could be implemented easily at a later stage).
This new FIT partition parser works for NOR flash (on top of mtdblock),
NAND flash (on top of ubiblock) as well as classic block devices
(ie. eMMC, SDcard, SATA, NVME, ...).
It could even be used to mount such FIT images via `losetup -P` on a
user PC if this patch gets included in Linux upstream one day ;)
Signed-off-by: John Crispin <john@phrozen.org>
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
The date -k patch is non standard and will be removed in the next
commit.
Tested behavior to be identical with a simple C program:
#define _GNU_SOURCE
#include <unistd.h>
#include <stdio.h>
#include <sys/time.h>
#include <sys/syscall.h>
int main()
{
struct timezone tt;
struct timezone tz;
int a = syscall(SYS_gettimeofday, NULL, &tt);
int b = gettimeofday(NULL, &tz);
printf("%d - %d, %d\n", a, tt.tz_minuteswest, tt.tz_dsttime);
printf("%d - %d, %d\n", b, tz.tz_minuteswest, tz.tz_dsttime);
}
Signed-off-by: Rosen Penev <rosenp@gmail.com>
Pstore (persistent store) can be used to stash debug information (kernel
console, panics, ftrace) across reboots or crashes. If the filesystem is
present, mount it.
Signed-off-by: Brian Norris <computersforpeace@gmail.com>
/lib/functions.sh was executable for no obvious reason and its
execute property was even checked in package-ipkg.mk just to
source it afterwards.
Remove the execute bit and shebang as this is clearly a library.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
Reviewed-by: Philip Prindeville <philipp@redfish-solutions.com>
This drops the shebang from another bunch of files in various /lib
folders, as these are sourced and the shebang is useless.
Fix execute bit in one case, too.
This should cover almost all trivial cases now, i.e. where /lib is
actually used for library files.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
This commit is only added to keep the PKG_RELEASE correct after fixing
the $(COMMITCOUNT) logic in the previous commit.
This way the PKG_RELEASE stays the same while the compiled packages
content isn't changed.
Signed-off-by: Paul Spooren <mail@aparcar.org>
The newly added `$(COMMITCOUNT)` varialbe allows automatic versioning
based on the number of Git commits of a package. Replace *tedious to
bump* and *merge conflict causing* `PKG_RELEASE` and replace it with
`$(COMMITCOUNT)`.
Signed-off-by: Paul Spooren <mail@aparcar.org>
While the speed improvement might be negligible, there is still no
reason to read individual bytes.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
Adds a new function get_magic_fat32() in base-files to read FAT32 magic.
Now FAT32 EFI system partition can be handled in the same way as FAT12/FAT16.
Signed-off-by: Kagurazaka Kotori <kagurazakakotori@gmail.com>
[replace '-o' with '] || [' to satisfy shellsheck]
Signed-off-by: Paul Spooren <mail@aparcar.org>
The `functions.sh` script has `config_get_bool()` function, which is
usable when using UCI config direct access API, but there is no
equivalent for the callback API. Introduce `get_bool()` function to
allow reusing it from init scripts.
Example:
```sh
option_cb() {
local option="$1"
local value="$(get_bool "$2")"
...
}
```
Signed-off-by: Oldřich Jedlička <oldium.pro@gmail.com>
The find command to retrieve files from /etc/sysupgrade.conf and
/lib/upgrade/keep.d/* is used twice in almost the same way.
Move it into a function to consolidate, enhance readability and make
future adjustments easier.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
Acked-by: Paul Spooren <mail@aparcar.org>
Reviewed-by: Philip Prindeville <philipp@redfish-solutions.com>
Flush kernel memory caches during sysupgrade in order
to mitigate the impact from memory consumption spikes
in low-RAM devices.
This may help to prevent sysupgrade causing a reboot
before the actual flashing starts.
Signed-off-by: Hannu Nyman <hannu.nyman@iki.fi>
Commit "initramfs: switch to tmpfs to fix ujail" switched initramfs to
now use tmpfs, it causes $(rootfs_type) to now return tmpfs when
running initramfs image instead of being empty.
This broke initramfs detection which prevents config files from
being saved as it does not work from initramfs.
So, lets test for $(rootfs_type) returning "tmpfs" instead.
Fixes: 7fd3c68 ("initramfs: switch to tmpfs to fix ujail)
Signed-off-by: Robert Marko <robimarko@gmail.com>
This dd flag ensures that the requested size
is retrieved from pipes or special filesystems (if available).
Without this flag, on multi-core systems,
Piped or special filesystem data can be truncated
when a size greater than PIPE_BUF is requested.
Fixes: FS#3494
Fixes: 7557e7f ("package/base-files: caldata: work around dd's
limitation")
Cc: Thibaut VARÈNE <hacks@slashdirt.org>
Signed-off-by: John Thomson <git@johnthomson.fastmail.com.au>
Add code for setting mac addresses inside board.json and rendering
them out to uci. On switches we want to have a unique MAC on each port.
With 48 port switches that would require 48 device sections in
/etc/config/network. Doing so via board.json is easier.
Signed-off-by: John Crispin <john@phrozen.org>
Support installations without root-overlayfs (and hence without /rom)
when migrating user accounts.
Signed-off-by: Imran Khan <gururug@gmail.com>
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
[simplified patch, bumped PKG_RELEASE, cleaned message]
Strictly speaking, ash does not support it.
From https://wiki.ubuntu.com/DashAsBinSh#A.5B.5E.5D
Not to be confused by sed's and other program's regular expression
syntax. Uses of [^...] in case (parameter/word expansion in general) need
to be replaced with [!...].
Found with shellcheck: https://github.com/koalaman/shellcheck/wiki/SC2169
Signed-off-by: Rosen Penev <rosenp@gmail.com>
[minor commit title/message adjustments]
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
Avoid needlessly breaking old initscripts that set EXTRA_COMMANDS. This
will aid in debugging (as it simplifies reverting to an older version of
a package) and unbreaks third-party feeds (and packages that maintain
their OpenWrt initscripts as part of the software's repo instead of the
OpenWrt feed like fastd).
Without this, initscripts that set EXTRA_COMMANDS become completely
unusable, as all default commands like start/stop cease working.
Fixes: 1a69f50dc6 ("base-files: fix rc.common help alignment")
Signed-off-by: Matthias Schiffer <mschiffer@universe-factory.net>
The intent is to make it sound more like info level message, not some
error like "404 not found". x86 target at the moment makes image with
only signature but no metadata (ref commit f8141216 "x86: append
metadata to combined images").
Signed-off-by: Yousong Zhou <yszhou4tech@gmail.com>
Reviewed-By: Philip Prindeville <philipp@redfish-solutions.com>
This will have at least the following effects
- Log lines will have common prefix
- They will be output to stderr instead of stdout
Signed-off-by: Yousong Zhou <yszhou4tech@gmail.com>
This is mainly to handle stderr message "Broken pipe", "F+P records
in/out" by common pattern "xcat | dd .."
Ref: https://bugs.openwrt.org/index.php?do=details&task_id=3140
Reported-by: Philip Prindeville <philipp@redfish-solutions.com>
Signed-off-by: Yousong Zhou <yszhou4tech@gmail.com>
Reviewed-By: Philip Prindeville <philipp@redfish-solutions.com>
Latest netifd allows us to setup network bridges with implicit vlan
tagging. For this to work, we need to setup several additional uci
sections. This feature is particularly usefull for DSA tupe devices.
Add board.d and uci-defaults support for generating the sections.
Signed-off-by: John Crispin <john@phrozen.org>
This commit introduces a new function `extra_command` to better format
the help text without having to calculate the indentation in every startup
script that wants to add a new command. So far it looks weird and is not
formatted correctly on some startup scripts.
After using the new `extra_command` wrapper the alignement looks correctly.
And if the indentation is not sufficient in the future, this can be
changed in the function extra_command at a central location.
Signed-off-by: Florian Eckert <fe@dev.tdt.de>
Rather than unconditionally adding busybox and procd to the set of
default packages, add busybox-selinux and procd-selinux in case
CONFIG_SELINUX is set.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
Without the model-based devicename for LEDs, there are still cases
where a third component is required, typically when it refers to
internal "devices" like phys etc. An example are the following two
found on ramips:
- rt2800soc-phy0::radio
- rt2800pci-phy0::radio
So far, the rt2800*-phy: prefixes would be removed by the devicename
removal ("migration") script, and the configuration for these LEDs
would be broken.
To address this, this patch allows to add arguments to a call of
remove_devicename_leds, which will be compared against the first
part of the LED names/labels, and then be ignored by the routine,
and thus not removed:
remove_devicename_leds "rt2800soc-phy0" "rt2800pci-phy0"
This mechanism is supposed to be used when a "devicename" applies
to several devices. If only a single device is affected, it might
be more effective to use a case statement and exclude the device
from migration by that entirely.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
Currently, we request LED labels in OpenWrt to follow the scheme
modelname:color:function
However, specifying the modelname at the beginning is actually
entirely useless for the devices we support in OpenWrt. In patches
subsequent to this one, we will thus remove the modelname from
the label definitions on various targets.
To migrate the existing definitions from older installations,
a migration script needs to be deployed that does
modelname:color:function -> color:function
e.g.
dir-789:green:status -> green:status
This patch introduces two functions that do exactly that:
For each entry in /etc/config/system, the routine will check whether
two (or more) colons are present, and then remove everything up to
(and including) the first colon.
For now, this will be applied unconditionally, i.e. if the function
is called for a device, all labels will be cut like this.
However, for a future case of mixed three-part and two-part labels,
it should not be too hard to provide a function argument with
exceptions to the removal.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
vconfig has been disabled by default since 2015 [1] and there are
no remaining uses in entire OpenWrt trunk. However, we still set up
a specific name_type for it during boot.
While this setup is properly implemented to be only triggered when
vconfig is present, it still seems anachronistic and unnecessary
to set up a standard for a tool that is not used anymore.
Therefore, this removes the set_name_type initialization and leaves
it for those people actually using the tool to configure it as needed.
[1] 899a23227e ("busybox: improve applets & deprecate ifconfig, route")
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
Set the default state for LEDs to off. When a trigger is set, the
trigger will turn the LED automatically on.
Currently LEDs might stay on, e.g. when the LED trigger is set to a
netdev trigger and the interface is never activated or the 'none'
trigger is selected without setting the 'default' option to 0 and it's
set for the LED indicating the system running state.
Using off as a default value is also consistent with the documentation
in the OpenWrt wiki.
Signed-off-by: David Bauer <mail@david-bauer.net>
The LED's "label" property has been deprecated in upstream by:
|commit c5d18dd6b64e09dd6984bda9bdd55160af537a8c
|Author: Jacek Anaszewski <jacek.anaszewski@gmail.com>
|Date: Sun Jun 9 20:19:04 2019 +0200
|
| dt-bindings: leds: Add properties for LED name construction
|
| Introduce dedicated properties for conveying information about
| LED function and color. Mark old "label" property as deprecated.
|
| Additionally function-enumerator property is being provided
| for the cases when neither function nor color can be used
| for LED differentiation.
in order to be somewhat prepared, this patch adds a fallback
as a last resort to make the current led code work by falling
back to the node-name as the "label".
Signed-off-by: Christian Lamparter <chunkeey@gmail.com>
Fix typo in comment.
Signed-off-by: Walter Sonius <walterav1984@gmail.com>
[commit title/message facelift]
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
The variable VERSION_REPO is used by opkg to download package(list)s.
Now that the default installation support encrypted HTTP opkg should
make use of it.
Suggested-by: Petr Štetiar <ynezz@true.cz>
Suggested-by: Baptiste Jonglez <baptiste@bitsofnetworks.org>
Signed-off-by: Paul Spooren <mail@aparcar.org>
Acked-by: Baptiste Jonglez <baptiste@bitsofnetworks.org>
Bump PKG_RELEASE for the affected packages as replacing "which" by
"command -v" represents a content change.
Fixes: 1fdf6b745c ("treewide: replace `which` with `command -v`")
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
Fix shellcheck SC2230
> which is non-standard. Use builtin 'command -v' instead.
Using `command -v` is POSIX compliant while `which` is not. Also to
mention, `command -v` is a shell builtin whereas `which` is a separate
busybox applet.
Once applied to everything concerning OpenWrt we can disable the busybox
feature `which` and save 3.8kB.
Acked-by: Stijn Tintel <stijn@linux-ipv6.be>
Signed-off-by: Paul Spooren <mail@aparcar.org>
[also replace cases in zram-swap]
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
When passing a section or option value to config_get() which contains
characters that happen to be valid variable interpolation expressions,
the function returns a nonsensical expression result instead of the
expected empty string.
When the passed section or option name contains other characters which
are not valid within a shell variable name, a substitution error is
occuring instead.
The issue can be easily reproduced by one of the following examples:
root@OpenWrt:~# . /lib/functions.sh
root@OpenWrt:~# config load system
root@OpenWrt:~# config_get variable invalid-section option
root@OpenWrt:~# echo "$variable"
section_option:-
root@OpenWrt:~# . /lib/functions.sh
root@OpenWrt:~# config load system
root@OpenWrt:~# config_get variable section invalid-option
root@OpenWrt:~# echo "$variable"
option:-
root@OpenWrt:~# . /lib/functions.sh
root@OpenWrt:~# config load system
root@OpenWrt:~# config_get variable section invalid@option
-ash: eval: syntax error: bad substitution
Fix this issue by only performing interpolations when the given section
and option arguments are free of illegal characters.
Signed-off-by: Jo-Philipp Wich <jo@mein.io>
Explicitly mount the BPF filesystem if available. This is used for pinning
eBPF programs and maps, making them accessible to other eBPF programs or
from userspace with the help of libbpf or bpftool.
Signed-off-by: Tony Ambardar <itugrok@yahoo.com>
[daniel@makrotopia.org: bumped PKG_RELEASE]
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
So far, the compatibility mechanism only works if both device and
image are already updated to the new routines. This patch extends
the sysupgrade metadata and fwtool_check_image() to account for
"older" images as well:
The basic mechanism for older devices to check for image compatibility
is the supported_devices entry. This can be exploited by putting
a custom message into this variable of the metadata, so older FW
will produce a mismatch and print the message as it thinks it's the
list of supported devices. So, we have two cases:
device 1.0, image 1.0:
The metadata will just contain supported_devices as before.
device 1.0, image 1.1:
The metadata will contain:
"new_supported_devices":["device_string1", "device_string2", ...],
"supported_devices":["Image version 1.1 incompatible to device: ..."]
If the device is "legacy", i.e. does not have the updated fwtool.sh,
it will just fail with image check and print the content of
supported_devices. If DEVICE_COMPAT_MESSAGE is set, this will be
printed on old devices as well through the same mechanism. Otherwise
a generic "Please check documentation ..." is appended.
Upgrade can still be performed with -F like when
SUPPORTED_DEVICES has been removed to prevent bricking.
If the device has updated fwtool.sh (but is 1.0), it will just use
the new_supported_devices instead, and work as intended (flashing
with -n will work, flashing without will print the appropriate
warning).
This mechanism should provide a fair tradeoff between simplicity
and functionality.
Since we touched a lot of fields in metadata, this also bumps
metadata_version to 1.1.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
We regularly encounter the situation that devices are subject to
changes that will make them incompatible to previous versions.
Removing SUPPORTED_DEVICES will not really be helpful in most of these
cases, as this only helps after a rename.
To solve this situation, this patchset introduces a compatibility
version for devices. In this patch, the actual checks are implemented
into fwtool_check_image():
If an incompatible change is introduced, one can increase either
the minor version (1.0->1.1) or the major version (1.0->2.0).
Minor version increment:
This will still allow sysupgrade, but require to reset config
(-n or SAVE_CONFIG=0). If sysupgrade is called without -n, a
corresponding message will be printed. If sysupgrade is called
with -n, it will just pass, with supported devices being checked
as usual. (Which will allow us to add back SUPPORTED_DEVICES for
many cases.)
Major version increment:
This is meant for potential (rare) cases where sysupgrade is
not possible at all, because it would break the device.
In this case, a warning will be printed, and -n won't help.
If image check fails because of one of the versions parts not
matching, the content of DEVICE_COMPAT_MESSAGE is printed in
addition to the generic message (if set).
For both cases, upgrade can still be forced with -F as usual.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
We regularly encounter the situation that devices are subject to
changes that will make them incompatible to previous versions.
Removing SUPPORTED_DEVICES will not really be helpful in most of these
cases, as this only helps after a rename.
To solve this situation, this patchset introduces a compatibility
version for devices. To complement the DEVICE_COMPAT_VERSION set
for the image to be flashed, this implements a compat_version on
the device, so it will have something to compare with the image.
The only viable way to achieve this seems to be via board.d files,
i.e. this is technically adding a compat version for the device's
config.
Like for the network setup, this will set up a command
ucidef_set_compat_version to set the compat_version in board.d.
This will then add a string to /etc/board.json, which will be
translated into uci system config by bin/config_generate.
By this, the compat_version, being a version of the config, will
also be exposed to the user.
As with DEVICE_COMPAT_VERSION, missing uci entry will be assumed
as compat_version "1.0", so we only need to add this if a device
needs to be bumped, e.g.
ucidef_set_compat_version "1.1"
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
This adds a function for generating a valid random MAC address (unset MC
bit / set locally administered bit).
It is necessary for devices which do not have a MAC address programmed
by the manufacturer.
Signed-off-by: David Bauer <mail@david-bauer.net>
This replaces deprecated backticks by more versatile $(...) syntax.
This does not touch lib/upgrade/nand.sh, as there replacement is
not trivial.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
Some devices (e.g. Arduino Yun) need bitwise operations during MAC address
setup. This commit adds generalized versions of macaddr_setbit_la(), which
are helpful when manipulating a single bit in a MAC address.
Signed-off-by: Sungbo Eo <mans0n@gorani.run>
With package "coreutil-sha256sum" installed "sysupgrade" fails to perform 'sha256sum -s' and instead returns 'invalid option -- 's''.
This is caused due to:
different syntax for a sha256sum status check ('sha256sum --status' with "coreutil-sha256sum")
'/usr/bin/sha256sum' being symlinked to '/usr/bin/gnu-sha256sum' (after installation of "coreutil-sha256sum")
"coreutil-sha256sum" package from the packages feed replaces the Busybox sha256sum
This patch restores for 'sysupgrade' the busybox call to its sha256sum applet.
Signed-off-by: Huangbin Zhan <zhanhb88@gmail.com>
urandom-seed has a separate Makefile, we can safely remove the definition here.
Fixes: 27bfde9c9f ("base-files: move urandom seed bits into separate package")
Signed-off-by: Sungbo Eo <mans0n@gorani.run>
Previously, gpio_switch only accepts GPIO pin number as input. Once a
GPIO pin is exported and named by device tree, its pin state cannot be
configured and saved across reboots by UCI.
This patch adds support for named GPIO pins. Thus GPIO pin can be
exported by device tree with active high/low correctly configured,
having human-readable name in /sys/class/gpio/ is also now possible.
More importantly, GPIO pins which are referenced by name will be immune
from pin mapping breakage while unintentional pin number changes are
introduced by kernel or driver updates.
Signed-off-by: Kuan-Yi Li <kyli@abysm.org>
Due to filesystem write caching the old configuration data could stay
out of flash for a long time during a first boot after the sysupgrade.
Power loss during this period could damage the overlay data and even
make device inaccessable via the network.
Fix this by syncing data to a flash as soon as the previous
configuration will be unpacked after the sysupgrade. Also sync the FS
state after the sysupgrade.tgz archive removing to prevent duplicative
extraction of a previous configuration.
Tested with AMD Geode based board.
Signed-off-by: Sergey Ryazanov <ryazanov.s.a@gmail.com>
This changes the ide-disk LED trigger to the generic disk-activity as
ide-disk trigger was removed in upstream commit eb25cb9956cc ("leds:
convert IDE trigger to common disk trigger").
Signed-off-by: Thomas Albers <thomas.gameiro@googlemail.com>
[split into separate commit, commit description facelift]
Signed-off-by: Petr Štetiar <ynezz@true.cz>
tl;dr: dd will silently truncate the output if reading from special
files (e.g. sysfs attributes) with a too large bs parameter.
This problem was exposed on some RouterBOARD ipq40xx devices which use a
caldata payload which is larger than PAGE_SIZE, contrary to all other
currently supported RouterBOARD devices: the caldata would fail to
properly load with the current scripts.
Background: dd doesn't seem to correctly handle read() results that
return less than requested data. sysfs attributes have a kernel exchange
buffer which is at most PAGE_SIZE big, so only 1 page can be read() at a
time. In this case, if bs is larger than PAGE_SIZE, dd will silently
truncate blocks to PAGE_SIZE. With the current scripts using bs=<size>
count=1, the data is truncated to PAGE_SIZE as soon as the requested
<size> exceeds this value.
This commit works around this problem by using `cat` in the caldata
routines that can read from a file (routines that read from mtd devices
are untouched). cat correctly handles partial read requests. The output
is then piped to dd with the same parameters as before, to ensure that
the resulting file remains exactly the same.
This is a simple workaround, the downside is that it uses a pipe and one
more executable, and therefore has a larger memory footprint and is
slower. This is deemed acceptable considering these routines are only
used at boot time.
Tested-by: Robert Marko <robimarko@gmail.com>
Signed-off-by: Thibaut VARÈNE <hacks@slashdirt.org>
Use same indent as for the rest of the file.
Signed-off-by: Javier Marcet <javier@marcet.info>
[add commit description]
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
This will enable platforms to extract caldata to an arbitrary file,
or patch mac in an abitrary file.
Signed-off-by: Thibaut VARÈNE <hacks@slashdirt.org>
As touch creates files with permission 0644 use umask to create
config files with permission 0600 to be inline with INSTALL_CONF
Signed-off-by: Hans Dedecker <dedeckeh@gmail.com>
Including the local build key in /etc/opkg/keys isn't feasible when
building on the buildbot: The included key collides with its copy
already in openwrt-keyring which breaks the ImageBuilder.
Not including a locally generated key also makes the base-files package
more reproducible.
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
current preinit code in base-files doesn't config switch when there are
no port roles defined. But this kind of configuration exists on single
port devices where switch vlan is simply disabled.
configure reset and enable_vlan property when a switch node exist.
Signed-off-by: Chuanhong Guo <gch981213@gmail.com>
The file /lib/functions/system.sh depends on find_mtd_index() and
find_mtd_part() located in /lib/function.sh, so let's source that
file.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
The default_postinst() function in /lib/functions.sh sources
/lib/functions/system.sh before cycling through uci-defaults files.
This creates a pseudo-cyclic dependency as system.sh also uses
functions that are located in functions.sh. Despite that, there
is actually only one uci-defaults file in the entire repo that needs
system.sh, and this one contains an explicit source for system.sh
anyway.
Consequently, this patch removes the sourcing of system.sh in
functions.sh. There are no relevant uses in packages, routing and
luci repositories.
This may require adjustments for downstream, though.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
Add EFI platform bootable images for x86 platforms. These images can
also boot from legacy BIOS platform.
EFI System Partition need to be fat12/fat16/fat32 (not need to load
filesystem drivers), so the first partition of EFI images are not ext4
filesystem any more.
GPT partition table has an alternate partition table, we did not
generate it. This may cause problems when use these images as qemu disk
(kernel can not find rootfs), we pad enough sectors will be ok.
Signed-off-by: 李国 <uxgood.org@gmail.com>
[part_magic_* refactoring, removed genisoimage checks]
Signed-off-by: Petr Štetiar <ynezz@true.cz>
Because /etc/profile (and ~/.profile) are read by login shells only,
aliases and functions defined there are not available to non-login
shells, e.g. when using screen or tmux.
If the ENV environment variable exists (exported by /etc/profile or
~/.profile) and references an existing file, then all interactive shells
(login or non-login) will read that file as well.
This sets the ENV environment variable in /etc/profile, pointing to
/etc/shinit.
This also adds /etc/shinit, which:
* Contains alias and function definitions originally in /etc/profile
* Sources /etc/mkshrc if the user is using mksh (also originally in
/etc/profile), as /etc/mkshrc is meant for all interactive shells
* Sources ~/.mkshrc if the user is using mksh, to compensate for the
fact that mksh will not read ~/.mkshrc if ENV is set
* Sources ~/.shinit if the user is not using mksh
This also removes the shebang from /etc/profile, as the file is sourced,
not executed.
Signed-off-by: Jeffery To <jeffery.to@gmail.com>
CONFIG_INCLUDE_CONFIG option is helpful for being able to rebuild the
exact same firmware as you see on a live OpenWRT instance, but it's
crucially missing feeds information, so we can't rebuild the exact same
package versions. This commit fixes this by adding the remaining feeds
(and version) buildinfo files to the image.
Signed-off-by: Xu Wang <xwang1498@gmx.com>
For devices without a dedicated 'diag' LED, we use sometimes one of
other LEDs for indicating at least 'boot', 'failsafe' and 'upgrade'
stages. In some cases, at the same time these LEDs have defined default
triggers in DTS using 'linux,default-trigger' property. Current 'diag'
setup removes the trigger and turns off 'boot' LED after bootup.
One of the examples of such device is TP-Link TL-WR841N v14 (ramips)
which uses 'wlan' LED with defined 'linux,default-trigger' for 'diag':
aliases {
led-boot = &led_wlan;
led-failsafe = &led_wlan;
led-upgrade = &led_wlan;
};
[...]
led_wlan: wlan {
label = "tl-wr841n-v14:green:wlan";
gpios = <&gpio1 9 GPIO_ACTIVE_LOW>;
linux,default-trigger = "phy0tpt";
};
This patch extends 'diag.sh' and 'leds.sh' scripts to make sure default
trigger defined in DTS is restored for 'diag' LED which isn't used for
indicating 'running' stage.
Acked-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
Signed-off-by: Piotr Dymacz <pepe2k@gmail.com>
This change makes the names of Broadcom targets consistent by using
the common notation based on SoC/CPU ID (which is used internally
anyway), bcmXXXX instead of brcmXXXX.
This is even used for target TITLE in make menuconfig already,
only the short target name used brcm so far.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
"[[" is a bash extension for test. As the ash-implementation is
not fully compatible we drop its usage.
Also change to "=" for simple test, which is sufficient. (see d6ac8ca76c)
Signed-off-by: Sven Roederer <devel-sven@geroedel.de>
[split patch, removed shebang]
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
It's cleaner and faster as it does not need to do extra work.
Also removed $() to avoid executing the output. The shell can handle it.
https://github.com/koalaman/shellcheck/wiki/SC2143
Signed-off-by: Rosen Penev <rosenp@gmail.com>
[correct || to && for one conversion]
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
The sender domain has a DMARC Reject/Quarantine policy which disallows
sending mailing list messages using the original "From" header.
To mitigate this problem, the original message has been wrapped
automatically by the mailing list software.
Failsafe code of dropbear should be in the dropbear package not the
base-files package.
Signed-off-by: Kyle Copperfield <kmcopper@danwin1210.me>
Setting CONFIG_IPK_FILES_CHECKSUMS=y causes sha256 checksum files to be
included with the packages to check for corruption. This commit fixes two
issues:
- /sbin/pkg_check was being removed incorrectly if IPK_FILES_CHECKSUMS=y
- checksums were being saved in the wrong file
Signed-off-by: Xu Wang <xwang1498@gmx.com>
With this change the well known jshn library will be used, to build the
json arguments for the ubus sysupgrade method. This is also used in all
other shell program that uses JSON. This commit unifies that.
Signed-off-by: Florian Eckert <fe@dev.tdt.de>
The preinit network initialisation and failsafe informational message
are inherently racy as the interface takes some time to become
functional after "ip link set $pi_ifname up" command.
Consider this timing:
[ 12.002713] IPv6: ADDRCONF(NETDEV_UP): eth1: link is not ready
[ 12.008819] IPv6: ADDRCONF(NETDEV_UP): eth1.1: link is not ready
[ 12.118877] random: procd: uninitialized urandom read (4 bytes read)
[ 13.068614] eth1: link up (1000Mbps/Full duplex)
[ 13.073309] IPv6: ADDRCONF(NETDEV_CHANGE): eth1: link becomes ready
[ 13.080445] IPv6: ADDRCONF(NETDEV_CHANGE): eth1.1: link becomes ready
Since the UDP message was sent prior to link becoming ready, it was
never seen on the wire.
The default failsafe timeout is set to 2 seconds, so with this patch
there are two attempts to send the message, one spent in vain, and the
other visible in tcpdump on an attached host. Of course, in cases when
the interface is brought up faster it leads to two messages, however it
should be harmless. This patch (almost) doesn't affect normal boot time
while still allowing to enter failsafe reliably with a single button
press, matching the official "generic failsafe" documentation.
Signed-off-by: Paul Fertser <fercerpav@gmail.com>
netifd does not handle network.@device[x].name properly if it
contains multiple ifaces separated by spaces. Due to this, board.d
lan_mac setup does not work if multiple ifaces are set to LAN by
ucidef_set_interface_lan.
To fix this, create a device node for each member iface when
running config_generate instead. Those are named based on the
member ifname:
ucidef_set_interface_lan "eth0 eth1.1"
ucidef_set_interface_macaddr "lan" "yy:yy:yy:yy:yy:01"
will return
config device 'lan_eth0_dev'
option name 'eth0'
option macaddr 'yy:yy:yy:yy:yy:01'
config device 'lan_eth1_1_dev'
option name 'eth1.1'
option macaddr 'yy:yy:yy:yy:yy:01'
ref: https://github.com/openwrt/openwrt/pull/2542
Signed-off-by: Sungbo Eo <mans0n@gorani.run>
[always use new scheme, extend description, change commit title]
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
Commit ed5b9129d7 ("base-files: implement generic service_running")
has added EXTRA_HELP variable, thus overriding already available
EXTRA_HELP text available in other init scripts, resulting in the
missing help text from services like dropbear for example.
So fix this regression by appending EXTRA_HELP text provided by the
other init scripts into the one provided by the script itself.
Fixes: ed5b9129d7 ("base-files: implement generic service_running")
Signed-off-by: Peter Stadler <peter.stadler@student.uibk.ac.at>
[commit title/description facelift, fixes tag, fixed From:, pkg bump]
Signed-off-by: Petr Štetiar <ynezz@true.cz>
Discovered recent changes had broken sysupgrade for ar71xx mikrotik
rb-493g, traced the problem to missing /usr/share/libubox/jshn.sh after
switching to tmpfs.
Signed-off-by: Russell Senior <russell@personaltelco.net>
uci-defaults are sourced and non-executable, so they do not require
a shebang.
While at it, apply consistent naming scheme.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
Now that netifd and hostapd allow dynamic reconfiguration, add a
command to trigger it.
Signed-off-by: John Crispin <john@phrozen.org>
Signed-off-by: Daniel Golle <daniel@makrotopia.org>
This reverts commit 6170c46b47.
There has been demand for further evaluation of the impact of a
changed hostname, so this is reverted for now. The default hostname
will be "OpenWrt" again after this commit.
The macaddr_geteui() function is not removed by this revert.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
If a label MAC address is provided for device, system
will rename the hostname with OpenWrt_{eui mac address}.
This helps to distinguish between different devices.
Since it's no good idea to nest json_* functions, this code does
not use get_mac_label directly, but only get_mac_label_dt as
external resource.
Signed-off-by: Rosy Song <rosysong@rosinson.com>
[merged with commit introducing macaddr_geteui, rebased on updated
label MAC address storage, extended commit message]
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
If set, label MAC address is available from one of two sources,
device tree or board.json. So far, the function get_mac_label
was meant for retrieving the address, while an option in uci
system config was specified only for case 2 (board.json).
The uci config option has several drawbacks:
- it is only used for a fraction of devices (those not in DT)
- label MAC address is a device property, while config implies
user interaction
- label_macaddr option will only be set if /etc/config/system
does not exist (i.e. only for new installations)
Thus, this patch changes the behavior of get_mac_label:
Instead of writing the value in board.json to uci system config
and reading from this location afterwards, get_mac_label now
extracts data from board.json directly. The uci config option
won't be used anymore.
In addition, two utility functions for extraction only from DT
or from board.json are introduced.
Since this is only changing the access to the label MAC address, it
won't interfere with the addresses stored in the code base so far.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
"block mount" invokes "hotplug-call mount". It emits the following
error when mount is not present
hotplug-call call failed
Signed-off-by: Yousong Zhou <yszhou4tech@gmail.com>
Moving a file between tmpfs and other fs is neither
faster nor safer, thus no point in doing it in two steps.
Use new jshn option to write output directly to file.
Originally discussed here:
http://lists.openwrt.org/pipermail/openwrt-devel/2017-December/010127.html
Signed-off-by: Roman Yeryomin <roman@advem.lv>
This separates the options for signature creation and verification
* SIGNED_PACKAGES create Packages.sig
* SIGNED_IMAGES add ucert signature to created images
* CHECK_SIGNATURE add verification capabilities to images
* INSTALL_LOCAL_KEY add local key-build to /etc/opkg/keys
Right now the buildbot.git contains some hacks to create images that
have signature verification capabilities while not storing private keys
on buildbot slaves. This commit allows to disable these steps for the
buildbots and only perform signing on the master.
Signed-off-by: Paul Spooren <mail@aparcar.org>
While all ath10k eeproms have a checksum field, so far two
functions for patching ath10k MAC address have been present (and
been used).
This merges code to provide a single function ath10k_patch_mac
in caldata.sh, having its name in accordance with ath9k functions.
By doing so, correct MAC patching for current and future ath10k
devices should be ensured.
This patch adds checksum adjustments for several targets on
ath79 and lantiq.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
This unifies MAC address patch functions and moves them to a
common script. While those were implemented differently for
different targets, they all seem to do the same. The number of
different variants is significantly reduced by this patch.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
This moves the almost identical calibration data extraction
functions present multiple times in several targets to a single
library file /lib/functions/caldata.sh.
Functions are renamed with more generic names to merge different
variants that only differ in their names.
Most of the targets used find_mtd_chardev, while some used
find_mtd_part inside the extraction code. To merge them, the more
abundant version with find_mtd_chardev is used in the common code.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
[rebase on latest master; add mpc85xx]
Signed-off-by: David Bauer <mail@david-bauer.net>
The xor() function is defined in each of the caldata extraction
scripts for several targets. Move it to functions.sh to reduce
duplicate code.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
"coreutil-date" package from the packages feed replaces the Busybox date
applet by symlinking /usr/bin/gnu-date to /bin/date. This prevents the system
init script from setting kernel timezone because the GNU date utility does not
provide such functionality:
root@OpenWrt:~# date -k
date: invalid option -- 'k'
Try 'date --help' for more information.
A specific reference to the Busybox date applet prevents alternative date
utilities from breaking the system init script.
Signed-off-by: Val Kulkov <val.kulkov@gmail.com>
The factory uboot of the Turris Omnia boots with "root=b301", and we
instruct new users to sysupgrade from there (e.g. method 1, step 7).
Currently, this will fail with "Unable to determine upgrade device".
Add a new case to export_bootdevice, which parses the hex argument.
Fixes commit 2e5a0b81 ("mvebu: sysupgrade: sdcard: keep user added ...")
Signed-off-by: Klaus Kudielka <klaus.kudielka@gmail.com>
For many devices, MAC addresses cannot be retrieved via the
device tree alias.
To still provide the label MAC address for those, this implements
a second mechanism that will put the address into uci config.
Note that this stores the actual MAC address, whereas in DTS
we reference the bearing device.
This is based on the work of Rosy Song <rosysong@rosinson.com>
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
To refer to the MAC address on a device's label, one can
specify the alias label-mac-device in the DTS which should
point to the bearer of the corresponding MAC address.
With the function get_mac_label, the user can retrieve then
retrieve this address and use it as a value that uniquely
identifies his device.
This is severely helpful for several downstream functionalities,
e.g. define MAC addresses of custom netifs or change the SSID to
be easily recognizable.
Signed-off-by: Adrian Schmutzler <freifunk@adrianschmutzler.de>
Adds a default status action for init.d scripts.
procd "service status" will return:
0) for loaded services (even if disabled by conf or dead)
3) for inactive services
4) when filtering a non-existing instance
Signed-off-by: Luiz Angelo Daros de Luca <luizluca@gmail.com>
[rebased, cleaned up]
Signed-off-by: Petr Štetiar <ynezz@true.cz>
Now that $UPGRADE_BACKUP is set conditionally there is no need to check
the $UPGRADE_OPT_SAVE_CONFIG anymore. All conditions can be simplified.
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
This explicitly tells procd what backup file should be used during
sysupgrade (if any). It's much more generic this way compared to the
magic /tmp/sysupgrade.tgz file that had to be created before a call.
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
This allows platform code to check if firmware image can be used with
preserving a backup. It may be used e.g. when installing vendor
firmwares that won't restore appended backup archive.
Suggested-by: Luis Araneda <luaraneda@gmail.com>
Signed-off-by: Rafał Miłecki <rafal@milecki.pl>
