libera/#devuan/ Monday, 2019-10-07

mn3mHello People... Is anyone recently faced with VirtualBox kernel module load issue on Devuan ASCII on latest kernel? (vbox=5.2.24-dfsg-4~bpo9+1 , kernel=4.19.37-5+deb10u2~bpo9+1 ) Error when do modprobe vboxdrv - "module: x86/modules: Skipping invalid relocation target" found this bug with no solution yet. :(12:54
mn3mmodule build finished without any error, but I can't load it :(12:55
mn3m+all worked ok before 'dist-upgrade' with kernel upgrade to linux-image-4.19.0-0.bpo.6-amd64 and due to missing "-common" package for bpo.6 looks like not possible to build kernel module and return to old bpo.5 don't fix the issue, same as switching to 6-insigned as recommended in SO13:02
inklrdevuan works so well, i weep13:09
mn3m@inkdir, according to bug it looks like Debian+Virtualbox issue in general, not Devuan specific...13:15
mn3m@inklr, sorry)13:15
inklrthere's breakage everywhere but that's part of the game13:16
inklri manage to fix a few things now and then13:16
inklrdoin' my part!13:16
mn3minklr, not really understand you well, but thanks)13:20
jaromilmn3m: if its a debian specific problem with a third-party software vendor (vbox) I can imagine it won't be us solving it14:00
jaromili'm sorry to hear this happens, old-stable should be stable14:00
mn3mjaromil, Thank you for your reply :) Huge honor to talk with you. :) Probably issue can be related with missing kernel headers dependency "-common" ( -> 0 results) with existing kernel bpo.6 where fix should exist. Debian package is on place
mn3mstable is not an option for me as for BTRFS user14:10
HumanG33kEvilham, are you here ?14:39
HumanG33ki read the debian bug and try to downgrade && upgrade the kernel but nothing change14:48
HumanG33kmake some paste14:48
cehtehHumanG33k: eh look familar, does that run in a container/lxc?14:52
HumanG33kit a server i rent14:53
HumanG33kso i guess no14:53
cehtehtry that :D14:53
HumanG33kcehteh, can you explain ?14:57
HumanG33k(let me rescue mode the server )14:57
cehtehi think its a deficiency in the virtualized/hardenened /proc filesystem which doesnt allow the guest to set the loginuid14:58
HumanG33kif it s the solution do you know where i can post/email to prevent this kind of issues, know about if your diff not open a security issue ?15:06
cehtehgood question, i picked that up elsewhere while i stumbled over that problem as well, iirc the loginuid is only important for auditing and that bug is already reported/well known you prolly dont need to post it somewhere15:07
cehtehi wont mind when you research this further and give a more conclusive explanation and maybe a fix :)15:08
cehtehfix as in patch to lxcfs or where the cause is15:10
cehtehmhm maybe lxcfs could be copnfigured, but that would need access to the hosting platform15:11
Evilhamugh it'd be nasty if they are using lxc and promising actual VMs15:46
Evilhamor catual servers15:46
cehtehi host some LXC's for myself (and friends) and i also rented a cheap (5Eur month) cloud thingy which states its 'some shared virtualization' ... such is pretty common15:49
Evilhamthat's different15:50
cehtehthere youi get such15:50
Evilhamyeah, if it's clear what you're getting it's fine15:50
EvilhamHumanG33k: anyway, did you check the bug again? apparently the affected person's issue went away after kernel was upgraded15:51
cehtehwell dunno what HumanG33k uses15:51
Evilhamwhich would match why I can't reproduce your issue15:51
EvilhamHumanG33k: since I don't recall having seen your uname -v, and the error is exactly the same, it can't be discarded :-D15:52
Evilhamwould be great if you can check that15:52
cehtehmhm .. i upgraded the kernel but havent rebooted yet :D15:52
cehteh 13:53:20 up 383 days, 17:04,  3 users,  load average: 0,03, 0,11, 0,1515:53
HumanG33kEvilham, i use a v 4 kernel15:55
HumanG33kas say in the report bug ;)15:55
inklr5.x was too problematic for me15:55
Evilhamoh :-D you actually posted there15:55
Evilhamwasn't sure it was the same person15:55
inklr5.x was well into the 'i don't have time for this' zone15:55
cehtehi heared some problems abut 5.x  but generally works for me, wiht some instablilities when trying to match firmwares15:56
mn3mBTW, I fixed by issue with VirtualBox. Looks like there are bug in Debian package name as from source package for  4.19.67-2+deb10u1~bpo9+1 there are build for 4.19.67-2~bpo9+1. When I forced by manual patch to use correct name and install with dpkg after reboot everything works fine. Module is on place and vbox works :)16:05
HumanG33kcehteh, in your solution i have something to do in the  /etc/security/access.conf16:05
mn3ms/by issue/my issue/16:06
cehtehHumanG33k: please keep m noted when you find something more proper16:07
inklrmn3m: just a generic suggestion: posting it to a forum or git issues could help others find the solution16:09
james1138FYI: Newest stable kernels 5.3.4, 5.2.19, 4.19.77, 4.14.147, 4.9.195, and 4.4.195 are out - along with their bug fixes.
HumanG33kyes if that work ^^, and i will have an electricity shutdown near and after i have to go for two days without internet …16:09
HumanG33knear = soon16:12
HumanG33kcehteh, not working16:21
HumanG33ki think i have an error in some other place16:22
HumanG33ki have to disconnect16:22
jamin50devuan ascii rocks, it's excellent.  that is all.17:20

Generated by 2.17.0 by Marius Gedminas - find it at!