Regarding the error message: * /etc/init.d/net.eth0 uses runscript, please convert to openrc-run. a search on the gentoo forums suggests that those should be reported here. On my system, all the following scripts still use the deprecated /sbin/runscript: /etc/init.d/ $ grep runscript * alsasound:#!/sbin/runscript anacron:#!/sbin/runscript busybox-ntpd:#!/sbin/runscript busybox-watchdog:#!/sbin/runscript consolekit:#!/sbin/runscript fancontrol:#!/sbin/runscript ip6tables:#!/sbin/runscript iptables:#!/sbin/runscript lm_sensors:#!/sbin/runscript net.lo:#!/sbin/runscript nullmailer:#!/sbin/runscript pydoc-2.7:#!/sbin/runscript pydoc-3.4:#!/sbin/runscript rsyncd:#!/sbin/runscript sysklogd:#!/sbin/runscript udev:#!/sbin/runscript vixie-cron:#!/sbin/runscript
if I don't misremember, there was a news item suggesting to re-emerge the affected packages Something like "emerge -1 /etc/init.d" should do the job
Created attachment 451973 [details] emerge -1 sys-fs/udev-init-scripts It is the latest amd64 stable of sys-fs/udev-init-scripts using runscript.
(In reply to augustin from comment #0) > Regarding the error message: > * /etc/init.d/net.eth0 uses runscript, please convert to openrc-run. > a search on the gentoo forums suggests that those should be reported here. > > On my system, all the following scripts still use the deprecated > /sbin/runscript: > > /etc/init.d/ $ grep runscript * > > alsasound:#!/sbin/runscript > anacron:#!/sbin/runscript > busybox-ntpd:#!/sbin/runscript > busybox-watchdog:#!/sbin/runscript > consolekit:#!/sbin/runscript > fancontrol:#!/sbin/runscript > ip6tables:#!/sbin/runscript > iptables:#!/sbin/runscript > lm_sensors:#!/sbin/runscript > net.lo:#!/sbin/runscript > nullmailer:#!/sbin/runscript > pydoc-2.7:#!/sbin/runscript > pydoc-3.4:#!/sbin/runscript > rsyncd:#!/sbin/runscript > sysklogd:#!/sbin/runscript > udev:#!/sbin/runscript > vixie-cron:#!/sbin/runscript At least the following need to be added to that list: /etc/init.d/atieventsd:#!/sbin/runscript /etc/init.d/cgconfig:#!/sbin/runscript /etc/init.d/cgred:#!/sbin/runscript /etc/init.d/cgroups:#!/sbin/runscript /etc/init.d/deluged:#!/sbin/runscript /etc/init.d/dgc:#!/sbin/runscript /etc/init.d/eposd:#!/sbin/runscript /etc/init.d/fbcondecor:#!/sbin/runscript /etc/init.d/gpm:#!/sbin/runscript /etc/init.d/gpsd:#!/sbin/runscript /etc/init.d/mdadm:#!/sbin/runscript /etc/init.d/mdraid:#!/sbin/runscript /etc/init.d/nas:#!/sbin/runscript /etc/init.d/NetworkManagerDispatcher:#!/sbin/runscript /etc/init.d/slpd:#!/sbin/runscript /etc/init.d/syslog-ng:#!/sbin/runscript /etc/init.d/system-tools-backends:#!/sbin/runscript /etc/init.d/timidity:#!/sbin/runscript /etc/init.d/ufw:#!/sbin/runscript /etc/init.d/xdm:#!/sbin/runscript /etc/init.d/xdm-setup:#!/sbin/runscript
@Pacho I couldn't find any relevant news. The ones about openrc are about different issues. I may have missed something, though. $ eselect news list News items: [1] 2009-04-18 (2009-04-18-java-config-wrapper-0.16 - removed?) [2] 2012-05-21 (2012-05-21-portage-config-protect-if-modified - removed?) [3] 2012-11-06 (2012-11-06-PYTHON_TARGETS-deployment - removed?) [4] 2013-06-01 PBXT now unsupported in MySQL/MariaDB [5] 2013-06-07 (2013-06-07-portage-preserve-libs-default - removed?) [6] 2013-08-23 (2013-08-23-emerge-language - removed?) [7] 2013-09-27 Separate /usr on Linux requires initramfs [8] 2014-03-16 Ruby 1.8 removal; Ruby 1.9/2.0 default [9] 2014-06-15 GCC 4.8.3 defaults to -fstack-protector [10] 2014-10-26 GCC 4.7 Introduced the New C++11 ABI [11] 2014-11-07 Upgrade to udev >= 217 or eudev >= 2.1 [12] 2014-11-25 bash-completion-2.1-r90 [13] 2015-01-28 CPU_FLAGS_X86 introduction [14] 2015-02-01 ffmpeg/libav conflict management: USE=libav [15] 2015-02-02 New portage plug-in sync system [16] 2015-03-28 True multilib support on amd64 [17] 2015-04-06 Apache AddHandler/AddType exploit protection [18] 2015-04-16 FFmpeg default [19] 2015-06-08 udev-init-scripts-29 important changes [20] 2015-07-25 Python 3.4 enabled by default [21] 2015-07-28 (2015-07-28-mysql-packaging-changes - removed?) [22] 2015-08-13 OpenSSH 7.0 disables ssh-dss keys by default [23] 2015-08-26 Ruby 1.9 removal; Ruby 2.0/2.1 default [24] 2015-10-07 OpenRC-0.18 localmount and netmount changes [25] 2016-01-08 Some dhcpcd hooks are now examples [26] 2016-01-27 Upgrading Apache from 2.2 to 2.4 [27] 2016-04-02 KDE Plasma 5 Upgrade [28] 2016-06-19 L10N USE_EXPAND variable replacing LINGUAS [29] 2016-08-11 Grub2 multislot default setting is changing [30] 2016-09-27 OpenRC 0.22 updates [31] 2016-10-25 LLVM 3.9 with LLVM_TARGETS
https://bugs.gentoo.org/show_bug.cgi?id=593604
By the way, I did what was suggested in the forum: sed -i -- '1s/runscript/openrc-run/' /etc/init.d/* however, I noticed that it transformed a symlink net.eth0 => net.lo into a file, a copy of the original. I deleted it and restored the symlink.
Besides, I don't understand the general advice, above and also in the forums, to re-emerge the affected packages. We all do emerge -uDNat @world or something similar. Wouldn't that eventually update the packages, provided they have been fixed in the gentoo repositories? Aren't the scripts in /etc/init.d/ updated as well?
(In reply to augustin from comment #7) > Besides, I don't understand the general advice, above and also in the > forums, to re-emerge the affected packages. > We all do emerge -uDNat @world or something similar. Wouldn't that > eventually update the packages, provided they have been fixed in the gentoo > repositories? Aren't the scripts in /etc/init.d/ updated as well? Eventually, yes. The affected packages should have been revisioned (given a new -r suffix) to force everyone to rebuild them, since without one, you have to figure out what to re-emerge yourself or just wait a while.
(In reply to Michael Orlitzky from comment #8) > (In reply to augustin from comment #7) > > Besides, I don't understand the general advice, above and also in the > > forums, to re-emerge the affected packages. > > We all do emerge -uDNat @world or something similar. Wouldn't that > > eventually update the packages, provided they have been fixed in the gentoo > > repositories? Aren't the scripts in /etc/init.d/ updated as well? > > Eventually, yes. The affected packages should have been revisioned (given a > new -r suffix) to force everyone to rebuild them, since without one, you > have to figure out what to re-emerge yourself or just wait a while. I revbumped all packages that were maintainer-needed, and sent a message to gentoo-dev-announce@g.o. telling maintainers to revbump their packages. If you're still seeing runscript, you should file bugs against the relevant packages, and tell them to revbump.
Per comment #9, this is not an OpenRC bug; the bugs should be filed against the individual packages requesting a revbump. Thanks, William
(In reply to Pacho Ramos from comment #1) > > Something like "emerge -1 /etc/init.d" should do the job This approach was less of a brute-force method: grep -l runscript /etc/init.d/* |xargs emerge -1 Halved the number of packages being built. I get that a revbump is the best solution, I'm just a bit impatient with my stable system spitting out so many of these errors. The sed method probably would have been fine too, not everything updated like the aforementioned udev and net.lo scripts.
Be careful when using sed! You are probably going to destroy your symlinks...