1
0
mirror of https://github.com/systemd/systemd.git synced 2024-12-23 21:35:11 +03:00
systemd/test/units/testsuite-23.Upholds.sh
Zbigniew Jędrzejewski-Szmek 1c5d54b2df test: use 'until' instead of 'while !'
In general, it's better to avoid a negation. And "!" is special, because it is
used for history expansion, i.e. the same command would behave differently if
pasted on the command line.

Inspired by 4a899c5a23.
2023-09-06 19:54:29 +01:00

100 lines
3.1 KiB
Bash
Executable File

#!/usr/bin/env bash
# SPDX-License-Identifier: LGPL-2.1-or-later
set -eux
set -o pipefail
# Test OnSuccess= + Uphold= + PropagatesStopTo= + BindsTo=
systemd-analyze log-level debug
systemd-analyze log-target journal
# Idea is this:
# 1. we start testsuite-23-success.service
# 2. which through OnSuccess= starts testsuite-23-fail.service,
# 3. which through OnFailure= starts testsuite-23-uphold.service,
# 4. which through Uphold= starts/keeps testsuite-23-short-lived.service running,
# 5. which will sleep 1s when invoked, and on the 5th invocation send us a SIGUSR1
# 6. once we got that we finish cleanly
sigusr1=0
trap sigusr1=1 SIGUSR1
trap -p SIGUSR1
systemctl start testsuite-23-success.service
while [ "$sigusr1" -eq 0 ] ; do
sleep .5
done
systemctl stop testsuite-23-uphold.service
systemctl enable testsuite-23-upheldby-install.service
# Idea is this:
# 1. we start testsuite-23-retry-uphold.service
# 2. which through Uphold= starts testsuite-23-retry-upheld.service
# 3. which through Requires= starts testsuite-23-retry-fail.service
# 4. which fails as /tmp/testsuite-23-retry-fail does not exist, so testsuite-23-retry-upheld.service
# is no longer restarted
# 5. we create /tmp/testsuite-23-retry-fail
# 6. now testsuite-23-retry-upheld.service will be restarted since upheld, and its dependency will
# be satisfied
rm -f /tmp/testsuite-23-retry-fail
systemctl start testsuite-23-retry-uphold.service
systemctl is-active testsuite-23-upheldby-install.service
until systemctl is-failed testsuite-23-retry-fail.service ; do
sleep .5
done
(! systemctl is-active testsuite-23-retry-upheld.service)
touch /tmp/testsuite-23-retry-fail
until systemctl is-active testsuite-23-retry-upheld.service ; do
sleep .5
done
systemctl stop testsuite-23-retry-uphold.service testsuite-23-retry-fail.service testsuite-23-retry-upheld.service
# Idea is this:
# 1. we start testsuite-23-prop-stop-one.service
# 2. which through Wants=/After= pulls in testsuite-23-prop-stop-two.service as well
# 3. testsuite-23-prop-stop-one.service then sleeps indefinitely
# 4. testsuite-23-prop-stop-two.service sleeps a short time and exits
# 5. the StopPropagatedFrom= dependency between the two should ensure *both* will exit as result
# 6. an ExecStopPost= line on testsuite-23-prop-stop-one.service will send us a SIGUSR2
# 7. once we got that we finish cleanly
sigusr2=0
trap sigusr2=1 SIGUSR2
systemctl start testsuite-23-prop-stop-one.service
while [ "$sigusr2" -eq 0 ] ; do
sleep .5
done
# Idea is this:
# 1. we start testsuite-23-binds-to.service
# 2. which through BindsTo=/After= pulls in testsuite-23-bound-by.service as well
# 3. testsuite-23-bound-by.service suddenly dies
# 4. testsuite-23-binds-to.service should then also be pulled down (it otherwise just hangs)
# 6. an ExecStopPost= line on testsuite-23-binds-to.service will send us a SIGRTMIN1+1
# 7. once we got that we finish cleanly
sigrtmin1=0
trap sigrtmin1=1 SIGRTMIN+1
systemctl start testsuite-23-binds-to.service
while [ "$sigrtmin1" -eq 0 ] ; do
sleep .5
done
systemd-analyze log-level info