Where communities thrive


  • Join over 1.5M+ people
  • Join over 100K+ communities
  • Free without limits
  • Create your own community
People
Activity
  • Dec 07 21:30
    johannbg commented #1664
  • Dec 07 20:32
    mwilck commented #1664
  • Dec 07 20:32
    mwilck synchronize #1664
  • Dec 07 20:26
    johannbg commented #1521
  • Dec 07 20:26
    johannbg commented #1521
  • Dec 07 20:16
    mwilck commented #1664
  • Dec 07 20:13
    github-actions[bot] labeled #1664
  • Dec 07 20:13
    github-actions[bot] labeled #1664
  • Dec 07 20:13
    mwilck review_requested #1664
  • Dec 07 20:13
    mwilck review_requested #1664
  • Dec 07 20:13
    mwilck review_requested #1664
  • Dec 07 20:13
    mwilck opened #1664
  • Dec 07 19:42
    AdamWill commented #1521
  • Dec 07 19:39
    johannbg closed #1521
  • Dec 07 19:39
    johannbg commented #1521
  • Dec 07 15:54
    ddiss commented #1662
  • Dec 07 15:46
    mwilck commented #1662
  • Dec 07 14:08
    rmetrich closed #1630
  • Dec 07 14:08
    rmetrich commented #1630
  • Dec 07 12:58
    ddiss commented #1662
Harald Hoyer
@haraldh
so, if anything urgent has to be done, I can be reached
Jóhann B. Guðmundsson
@johannbg
Thanks for the heads up I'm also unsure on my available in the next weeks as well since I might be doing contracting job far east of the country, should know more in the coming days
Enjoy your vacation
Harald Hoyer
@haraldh
thanks
Ash Berlin-Taylor
@ashb
Hey -- does any know of a (reusable) module for authorizing thunderbolt modules in the boot env? I've found https://github.com/bmwiedemann/openSUSE/blob/ed475a1e032bc18d734ed7facb7767432e03e52c/packages/t/thunderbolt-user-space/0002-initramfs-support.patch -- but I'd like something I can install in Arch
yesi
@yesi1_gitlab
Hi, i have a boring message from dracut before and after the boot - i use a crypt disk with lvm and btrfs ; please, have a look from my poste : https://forums.gentoo.org/viewtopic-t-1137738.html
Harald Hoyer
@haraldh
re
Daniel Molkentin
@danimo
Hey everyone
@haraldh @johannbg @Conan-Kudo: around?
Daniel Molkentin
@danimo
Anyway: I'll be leaving SUSE at the end of the month. Thus, I am stepping down as the (open)SUSE maintainer of dracut. @tblume will be taking over my responsibilities effective next week (I have some vacation left :)
I am not vaninshing off the earth, but I won't be able to spend much time on dracut in the forseeable future.
I'll be going over my issues with Thomas and either close or reassign them
Harald Hoyer
@haraldh
I'll be leaving Red Hat at the end of the month and my maintainer time of dracut will be very limited, too.
@danimo :point_up:
@danimo good luck with your new job!
Daniel Molkentin
@danimo
@haraldh and with yours. Who is going to stand in for you?
Harald Hoyer
@haraldh
nobody yet
Daniel Molkentin
@danimo
:/
Harald Hoyer
@haraldh
yeah
Daniel Molkentin
@danimo
Ok, seems we need to make sure that submissions do not get stuck...
Jóhann B. Guðmundsson
@johannbg
Everyone is heading for new adventures ;)
I'm doing contracting work until December so I'm more or less just available during lunch and evenings
Neal Gompa (ニール・ゴンパ)
@Conan-Kudo
holy crap
everyone is leaving? :(
@haraldh, @danimo: I guess good luck on your next adventures...
John Schaeffer
@jnschaeffer
hello! I'm running into an issue trying to get dracut 055 working on Arch Linux ARM with an iSCSI root device. setting netroot=iscsi:... and root=PARTUUID=... results in initqueue timing out, and setting root=iscsi:... directly (just as a test) fails immediately with an error saying a special device with a name matching the root iSCSI device specifier cannot be found. in both cases the debug output is showing no iSCSI devices in /dev/disk/by-path or anywhere else. is there a known issue with dracut 055 and iSCSI root paths, or is this a package-specific/user (as in me) issue?
Mrfai
@Mrfai
@danimo it would be nice if my PR #896 would be included before you leave. I'm not sure why it was not included. What is still missing?
dngray
@dngray:polarbear.army
[m]
is this the official room now for dracut?
I couldn't login because my IP was apparently restricted, couldn't login via Tor even when connecting with SASL and creating an account via their SSO domain. I couldn't use the web client either as apparently my IP of my VPN and ISP were both on dronebl. On top of that email cloaking services seemed to be blocked as well.
recently tried using freenode and god that is awful
not #dracut on freenode?
i think someone said that #dracut on freenode was going to move when redhat or something moved channels
Carl Lei
@XeCycle
Hi; I wonder if it is beneficial to decompress .ko.zst to .ko? The kernel has code to decompress initramfs in zstd, and .ko.zst requires a copy of libzstd for insmod to load it; libzstd occupies 1.1MiB, and zstd -3 compresses it to 440KiB. Looks like there may be room for minor space savings.
Neal Gompa (ニール・ゴンパ)
@Conan-Kudo
@dngray:polarbear.army at the moment, this is the official channel
dngray
@dngray:polarbear.army
[m]
ah okay, thanks for that
quite glad i don't have to go anywhere near freenode
smile330803
@smile330803
Hi All, I am new to dracut, just wondering which modules will be included when run 'dracut' without any other parameters?
xevious
@xevious:matrix.org
[m]
Can anyone clarify how /run gets mounted? I assumed it was happening here, but it gets mounted with a size parameter that isn't present here: https://github.com/dracutdevs/dracut/blob/master/modules.d/99base/init.sh#L63-L74
I've added rd.break=cmdline rd.debug to my kernel command line parameters and /run is already mounted with the size parameter when I reach the command line. I checked the journal and I don't see any debug log entries about it being mounted.
Debian's initramfs-tools has a parameter, initramfs.runsize, for setting the size of /run. Is there an equivalent setting in Dracut?
1 reply
xevious
@xevious:matrix.org
[m]
Yeah, I've traced through everything and realized that /run gets mounted by systemd, not Dracut.
Kairui Song
@ryncsn
Hi, @haraldh, what is the status of dash support? There are dash test cases, but I see here: https://github.com/dracutdevs/dracut/blob/049973b708298ea0ce1ac9c869b404f4c718eff3/pkgbuild/dracut.spec#L214, dash was considered not-supported a very long time ago (RHEL7.2?), any chance to bring it back? I tried a few run with dash and everything seems OK, especially after we have ShellCheck.
Jóhann B. Guðmundsson
@johannbg
@ryncsn doubt we want to risk (re)introducing any dash bugs, why do you want it back?
Kairui Song
@ryncsn
@johannbg dash does help save some memory and speed up the system in some use cases, and so far I didn't see any issue yet. I'm not sure about the risk, just an idea though.
Jóhann B. Guðmundsson
@johannbg
Well there are bugs and supporting multiple shells come at an additional maintenance price and not all modules are posix compliant ( not that it matters since posix does not work in practise nor is a standard in unix which many people think it is )
Jóhann B. Guðmundsson
@johannbg
If we are looking at performance improvements I'm sure somebody from the rust community has written a shell that is more secure and outperforms dash ( and bash and all the others ) at this point
yup ion-shell
xevious
@xevious:matrix.org
[m]
The advantage of sticking to Bourne shell syntax would be that it decouples the script syntax from the interpreter implementation. So, you can switch to a faster or more secure interpreter without having to rewrite all the modules.
Jóhann B. Guðmundsson
@johannbg
Well I do think we have to rewrite the modules more or less regardless but yeah.