[00:46] is multipass basically ubuntu's version of docker? [00:56] is there a way to view my local machine's /var/mail/ in thunderbirD? I tried adding "other account" and it says the spool file can't be found. [01:01] Guest7: there's docker(.io) in ubuntu, too [01:02] I only have a few messages at /var/mail/root. I use thunderbird for regular internet email, but I was hoping to just have a local account side by side those accounts, too [01:03] i see but it seems like they are pretty much the same? [01:03] noarb: make an alias for root pointing email to your user [01:03] Guest7: what are you trying to accomplish? [01:09] leftyfb: do you mean at the MTA or MUA level? [01:09] is it just permissions with thunderbird and root? [01:10] noarb: https://askubuntu.com/a/199453/1151311 [01:12] thanks for the link, I was using that answer in my earlier attempts. I'll try adjusting some fields to see if I can get it to work. Now I'm getting spool file can't be found from my regular user, and the only local mail on the system is roots [01:12] I'm not even interested in sending from this account right now, I would just like to see messages in my regular workflow when they appear [01:18] i am just trying to understand what it is [01:19] another question i have is where is the DNS configuration located on ubuntu? I don [01:19] 't see /etc/bind [01:20] Guest7: multipass is used to spin up vm's, typically kvm vm's, but it does support other hypervisors, and utilize cloud-init to customize their network and such [01:21] so it is an actual virtual machine but happens to work with k8s? [01:21] Guest7: as for DNS, ubuntu does not need a DNS server by default. If you're referring to the client resolver, it uses systemd-resolved for a client resolver [01:22] what directory is that in? [01:22] https://github.com/canonical/multipass [01:22] Guest7: multipass is not a VM, it's a tool used to orchestrate spinning up VM's across many hypervisors [01:22] is there a way to record the terminal output in an animation of the bootup process of ubuntu? [01:22] preferably including the bios screen? [01:23] Guest7: back to DNS, what are you looking to configure exactly? Which nameserver is configured? You typically set that either via your DHCP server (router) or if you want to hard code it, use your network settings [01:23] pycurious: run it in a vm, use your preferred video capture software === M4he is now known as mahe [01:24] i just want to know where dns config is stored, the file and the directory where it is [01:24] pycurious: including the BIOS screen, no, not on bare metal. Not unless you use something out of band like pikvm [01:26] Guest7: systemd-resolve --status , you can add |grep "DNS Servers" to get a list of configured nameservers from you network setting [01:26] yes but in which file and directory can i configure all of that stuff? [01:27] Guest7: you configure your nameserver in /etc/netplan/*.cfg [01:27] ok let me check it out [01:27] Guest7: unless you are on a desktop, then you use the built in GUI network manager [01:28] the only file i have in there is 01-network-manager-all.yml [01:28] Guest7: use the built in network manager [01:28] and unfortunately i am not on an ubuntu OS with a GUI [01:29] Guest7: please past the out of this here: cat /etc/os-release | nc termbin.com 9999 [01:30] https://termbin.com/2xbo [01:30] !20.10 | Guest7 [01:30] Guest7: Ubuntu 20.10 (Groovy Gorilla) was the 33rd release of Ubuntu, support ended July 22nd, 2021. See !eol and https://lists.ubuntu.com/archives/ubuntu-announce/2021-June/000269.html [01:31] Guest7: why do you have a 01-network-manager-all.yml if you're not running a GUI? [01:31] nothing it is blank [01:32] Guest7: that file is only created when you have ubuntu desktop installed [01:32] interetsing [01:32] Guest7: either way, you'll need to upgrade, your release is unsupported [01:33] Guest7: and your network settings should then be in /etc/netplan. [01:35] yes you seem to be rightr [01:44] tomreyn: thanks [01:46] thank you for the answers but i have another question [01:46] is it possible to trigger a kernel panic on purpose? [01:46] if so, why would you do it? [01:46] Guest7: why would you want to do such a thing? [01:47] Guest7: this is more a google question. I see no reason anyone here would help to accomplish such a thing [01:47] just wondering for educational purposes [01:48] Guest7: https://unix.stackexchange.com/questions/66197/how-to-cause-kernel-panic-with-a-single-command [01:49] wow that is wild [01:50] Now you can check if all 3 lights on your keyboard blink on and off repeatedly [01:52] the reason why you would want to do it is to capture a core dump? [01:53] Guest7: the core dump is only used to troubleshoot the issue that caused the kernel panic, which you never want to happen [01:53] I think it's perhaps being too general to say this kind of troubleshooting is the only use of having a core dump in-hand. [01:54] but I can't really think of what others there might be other than "I need to look at a core dump from this kernel right now" ? [01:54] SnoopJ: there's other, less crashy ways to get all the same information from your system for other purposes [01:54] leftyfb, "the same information" being a core dump? [01:54] I guess you could use it as some sort of state capture mechanism [01:55] all of that information should be in lsof and /proc/ [01:55] what if the information you want is the literal layout [01:55] layout? [01:56] of kernel memory [01:57] I mean, I get what you're saying, if you want [some particular information] then there is probably a better way to get it than crashing your system... [01:57] Guest7: there's no reason for you to purposely generate a kernel panic. If, some day you become a kernel developer, then you'll already know how to do such a thing if you have to [01:57] gotcha makes sense [04:26] like never say 'allegiance' again. [04:27] Heron Stiff. [04:32] xb: this is the third time in a few days that you're posting off-topic here. please stop. [04:33] tomreyn: my apologies [04:52] my desktop files have disappeared. when i upgraded from Ubuntu 18.04 to UB 20. Any advice? [04:56] xb: what's in ~/Desktop/ ? [04:56] you may not have apt:gnome-shell-extension-desktop-icons installed === bigpod9 is now known as bigpod [04:57] i'm assuming "UB 20" stands for "Ubuntu 20.04 LTS"? [04:58] xb: ^ [04:58] yes. yes it does. I'm trying the gnome desktop shell install now [05:37] thank you Mr. TomReyn that was the correct diagnosis. [05:39] you're welcome [05:46] * TJ- shares a spoonful of porridge with tomreyn as a reward :) [06:03] TJ-: Yum! Err, I mean, APT! [06:04] * TJ- awards tomreyn the comedy crown [06:25] :) === kn1003855702625 is now known as kn100 [08:58] Hi where can i find the dhcp client config ? [09:01] funabash1: by default ubuntu uses network-manager; what are you using? [09:09] alkisg: its by default i have a fresh installation. but i want to add some lines in dhcp config. cant find it [09:09] alkisg: to be more specific i want to add dhcp options [09:12] funabash1: network-manager has it own internal dhcp client implementation [09:12] It doesn't use isc-dhcp unless you specify it in its configuration file [09:13] Its full documentation is at https://developer-old.gnome.org/NetworkManager/stable/NetworkManager.conf.html [09:23] Can someone please point me on how to create a dkms .deb package in ubuntu? [09:26] Guest11: run `apt source rtl8812au-dkms` and then study! Or download it from https://packages.ubuntu.com/impish/rtl8812au-dkms === EriC^^_ is now known as EriC^^ === EriC^ is now known as EriC^^ [10:02] !paste [10:02] For posting multi-line texts into the channel, please use https://paste.ubuntu.com | To post !screenshots use https://imgur.com/ !pastebinit to paste directly from command line | Make sure you give us the URL for your paste - see also the channel topic. [10:03] !pasteinit [10:03] !pastebinit [10:03] pastebinit is the command-line equivalent of !pastebin - Command output, or other text can be redirected to pastebinit, which then reports an URL containing the output - To use pastebinit, install the « pastebinit » package from a package manager - Simple usage: command | pastebinit [11:23] Hey! How can I set Chrome to use the middle click? On Windows I was closing tabs when clicking on them and scrolling up/down when keeping it pressed. [11:24] Razva: with a mouse? [11:24] G502 [11:24] I already installed Piper [11:24] so yes a mouse [11:24] the button is mapped as "middle click" [11:25] Yes, with a Mouse. [11:25] middle mouse button click? [11:25] yup [11:25] it didn't just work without piper? [11:26] chrome is not really a supported app, test on chromium browser instead [11:26] nope, fresh install [11:26] ok let me try that as well [11:26] Can someone please point me list of commands need to run for creating a dkms .deb package in ubuntu? [11:26] nope, chromium does the same [11:27] firefox does the same [11:27] solved [11:27] somehow Piper unmapped the button [11:27] * Razva facepalms [11:29] Ok, next issue (with any browser) would be the fact that in Windows I was pressing middle click and scrolling down faster. Any way to do the same on Ubuntu? [11:30] Guest11: alkisg already pointed you to a solution [11:31] Razva: so it's doing middle click scrolling, but not as fast? [11:31] it's not doing any scrolling [11:31] lotuspsychje: Sorry I lost the n/w connection.. I might have missed it.. Could you please point me again.. [11:32] Guest11: Guest11: run `apt source rtl8812au-dkms` and then study! Or download it from https://packages.ubuntu.com/impish/rtl8812au-dkms [11:32] I press it and move the mouse up/down but nothing is happening [11:32] lotuspsychje: Thanks [11:33] Razva: in Firefox? [11:33] Firefox and Chromium [11:35] Razva: hrmm, maybe this: https://askubuntu.com/questions/28150/chrome-chromium-middle-mouse-button-for-scroll-linux-mac#answer-868173 [11:38] The Device ID would be 2? https://paste.debian.net/1207984/ [11:40] https://www.irccloud.com/pastebin/7yo4o7rK/ [11:41] Razva: hrmm, think I'd try 9 first [11:42] YEEEEE [11:42] that a good or bad yee? [11:44] a good one [11:44] coo [11:45] I'm noting this down as I have a feeling it'll be necessary for any DE and/or distro [11:46] could be [11:46] I don't remember ever having a problem doing that, "autoscrolling", as it's called, but [11:46] I've been using laptops a lot for the past handful of years [11:46] and Firefox, occasionally chromium, not really chrome [11:47] hi everybody [11:49] i have a problem by connection thorw wireguard on my system [11:49] when i try to connect by this sudo systemctl start wg-quick@London [11:50] i get errors [11:51] i use lubuntu 20.4 [11:52] error: Job for wg-quick@London.service failed because the control process exited with error code. [11:52] See "systemctl status wg-quick@London.service" and "journalctl -xe" for details. [11:53] by using this command journalctl -xe [11:53] the resualts are : [11:55] excuse me [11:56] by using this command : systemctl status wg-quick@London.service [12:03] hi all [12:04] out of around a dozen ubuntu server VMs, i have a single one that doesn't honor external (ACPI from virt-manager) requests to shut down and power off. what might be the reason for that? [12:05] Wow just found out that Tradingview and Think or Swim were available in the Software Repos. [12:06] finally you can exploit the workers on Linux, too =) [12:06] my apologies again that was a line from way way back. [12:07] https://pastebin.com/m13eyRwD [12:08] what should i do to solve it? please help me [12:09] looks the problem is in resolvconf [12:13] perhaps ask in #wireguard? [12:13] ok [12:13] but it think the problem comes with my operating system not the wireguard protocol [12:14] Hi folks [12:14] hi [12:19] well what does `ip a` say? [12:19] and does wireguard even have a tun interface, like openvpn? [12:24] rapha: don't know [12:24] i have no idea about it === Guest42 is now known as ufk [12:25] i think doesn't have because i sat it up like a service [12:25] i used to work [12:25] it used to work* [12:26] but nowadays it doesn't connect [12:27] i just control it by these commands [12:27] sudo systemctl start wg-quick@London [12:27] sudo systemctl stop wg-quick@London [12:28] nothing is changed on my network manager [12:28] what's also strange is that the if should be named "tun.London". i for one have never seen such a thing, but happy to receive some edification here. [12:29] iwhere is the strange? [12:30] tun.london ? [12:31] friends i have problem by connecting to wireguard protocol [12:31] by using this command : sudo systemctl start wg-quick@London [12:31] i get errors [12:32] https://pastebin.com/m13eyRwD [12:32] May you help me please? [12:45] ripper, join #wireguard chat, you might receive answers there much sooner than here [12:47] BluesKaj: ok === hugotty` is now known as hugotty [12:59] Hello all =) [12:59] hi [12:59] 6 [13:01] I have unexpected error when run sudo apt-key add - like gpg: symbol lookup error: gpg: undefined symbol: sqlite3_errstr. Do you know something about it ? libsqlite3-dev and sqlite3 already installed === humberto1 is now known as humbertowoody [13:02] ah yes, ubuntu 20.04 [13:11] how can i install nvidia drivers only for compute and not anything else? [13:11] my main GPU is AMD [13:11] my other GPUs are only for compute [13:11] and its Nvidia === Janvitus_ is now known as Janvitus [13:24] my partitions don't accept edit [13:24] hey [13:24] what should i do please [13:25] Guest97: use the disks program and change mount options [13:25] that is the easiest way [13:25] they're all unmounted by default.. but when i click them they immediately mounted [13:26] but i can't edit or rename ot delete files within [13:26] So use the disks program [13:26] then edit the mount options [13:27] Hi again, I have a problem with Ubuntu 21.04 which occured after a (partially) failed update from Ubuntu 20.10, after running for over 8 hours Gnome-shell starts to consume more than 50% of one of my cpu cores, how can I find out what is causing this? [13:28] use htop to find the program with the most CPU usage [13:29] goddard, thanks, but the program is gnome-shell itself (according to Gnome System Monitor) [13:29] zteam: try to reinstall gnome shell [13:32] i don't know how to do that with my partition program i dont' see options [13:32] its in the hamburger menu [13:32] when you click on the drive [13:32] it should be in the window toolbar [13:32] at the top [13:33] on the right hand side [13:33] KDE partition manager [13:33] nah gnome disks [13:33] dont know about kde partition manager [13:44] goddard, I tried to reconfigure it with dpkg now, hopefully that settles things down a bit [13:46] Guest97, are you sure you have the necesseary permsissions to modify the files ? [13:47] zteam i'm note sure, but each file owner details show my username [13:49] Guest97, right click the file or folder you are trying to modify and check whetever you have permssion to read and write to it [13:50] zteam: owner can view and modify [13:51] I created a simple .deb package as per https://ubuntuforums.org/showthread.php?t=910717 .. I need to print some text as a banner while installing the newly created .deb package, does DEBIAN/control supports this? [13:52] Guest97, Does it show that? (even if it's rare you can actually own files without having write permissions for it) [13:52] Guest11: try #ubuntu-devel [13:54] zteam: yes i right clicked properties: (user- group- other) under permession > Can View & Modify [13:54] all files withing that partition says so [13:58] Guest97, then you need to to alter the mount settings for that partition go to Disks Utitlty -> Select your Disk -> and partition and click on wheel meny-> and select "edit mount settings" [13:59] there's faded (grayed out) menu option (Edit mount point) when i right click on the partition on KDE Manager - [13:59] should i unmount it first? [14:00] or it's not the option we want [14:01] Guest97, it sounds like the right option... Did KDE Parition Manager ask you for any password then you started it? [14:01] yes [14:01] it asked for my username password [14:02] Guest97, okey, I have never used that tool myself, but it sounds like you need to unmount the partition first [14:03] unmounted [14:03] btw it's ntfs if that'sof any importance [14:04] i clicked edit mount point === genii-core is now known as genii [14:04] in the options there's " read only" which is unchecked by default [14:04] run ntfs fix [14:05] ntfsfix is the command [14:05] will it harm my files? [14:05] no [14:05] just fixes ntfs drives that weren't shut down cleanly by windows [14:06] should i do this with the partitions mounted or unmounted? [14:06] it will tell you [14:06] ok [14:06] HELO [14:07] Guest97, No it wont't it it will only do a basic check to make sure the NTFS-filesystem is correct :-) [14:08] https://pastebin.com/Q3z5EV6b [14:08] you gotta tell it which drive [14:08] lsblk [14:08] shows your drives [14:08] it will be /dev/sda or something like that [14:09] unless it is nvme [14:11] https://pastebin.com/Kpd4HXEy [14:11] i tried chkdsk as it says [14:11] Guest97, as goddard says you need to tell it the device name (you can probably view this very easyily from Kde Partion Manager [14:11] command not found. [14:12] Guest97, There is no Linux version of CheckDisk available, so if it wants you to run CheckDisk it means the Windows version from inside Windows [14:13] i have no windows :x [14:15] Guest97, try sudo ntsfix -d /dev/sda3 [14:18] https://pastebin.com/fENiMqWQ [14:18] it looks like it worked [14:18] see if you can use it [14:18] i tried to edit and delete a file.. nothing happened [14:20] something always happens [14:21] T-T [14:21] Guest97, ntfsfix can only do basic checking of a NTFS-volume the only way to work around that is either to 1. get an installation media with windows and run chkdsk /p C: (assuming that is the right letter) or 2. Copy all your files to anotther partition and to recreate the parition that you can't modify the files on [14:22] there's no easy ways to do fixes on linux [14:23] ntfsfix has always worked for me [14:23] the gnome disks to set mount options [14:23] my system also crashed about 5 times in 2 days.. usually when i try to logout from my user or customizing my UI [14:23] Guest97, however I'm confused, ntfsfix is clearly telling you to the filesystem is dirty becuase windows didn't shutdown properly, so the question is what happened to your Windows partition? [14:23] i'm starting to think that something went wrong when i installed kubuntu [14:23] and if you dont have windows installed you should just make the disk an ext4 drive any way [14:23] ntfs always causes problems in my experience [14:24] goddard, this is a good point if he doesn't intend to use that files on Windows in the future [14:24] i couldn't change the partition extension because it'll make me lose the files [14:25] Guest97: ya you gotta move them some place else temporarily [14:25] usb 3 drive maybe [14:25] so they'll remain only readable until i do move them somewhere else? [14:25] even with the free space on the partitions? [14:26] if you can read them now then yeah [14:26] i can red them but i can't modify them [14:26] its your computer and nothing will happen unless you do it [14:26] goddard, "always" is too absolute, i use NTFS on my outboard media drive without any issues [14:27] BluesKaj: ok [14:27] in my experience it is best to do things the Linux way when you move out of Windows land [14:28] is my sudo permssions different from root permssions? [14:28] i want to try something [14:29] kind of but not really it depends [14:29] look at the man page for -i [14:29] it details some differences [14:29] Guest97 sudo just means "perform this command only as root" [14:30] ok i've (rootfs) folder which is owned by root ok? [14:30] inside rootfs foler there's media folder [14:30] (or some other user, if you specify -u) [14:30] media folder has my username inside [14:30] inside my username there's my 2 partitions [14:30] mounted [14:30] and their owner is me [14:31] Guest97, yes, (make sure these has read and write permissions as well [14:31] media? [14:32] Guest97, the folders inside media [14:32] my username folder is owned by root not my username [14:32] media folder is owned by root too [14:33] group root [14:33] Guest97, does other has read and write permissions there ? [14:34] all options in permissions are grayed out [14:34] for my username folder [14:36] sudo chmoud o+rwx /media/directory [14:36] Guest97, sudo chmod o+rwx /media/directory [14:37] Guest97, (replace directory with name of the folder inside media) [14:37] apt-key adv --keyserver keyserver.ubuntu.com --recv-keys 7EA0A9C3F273FCD8 results to gpg: keyserver receive failed: Server indicated a failure [14:38] ivaat, try "sudo apt clean" && apt update [14:38] still can't modify [14:39] zteam: yea the apt update is part of script already [14:40] and how this should help... did not help [14:41] Guest97, nope, Linux is not very happy to touch any NTFS-volume if it finds any defect [14:41] can i run the chkdsk utility from inside linux? [14:42] it it will help? [14:43] ivaat, cleanings apts cache can actually help,since the key can be retrived from a possible corrupt cache (and no I can't now what's inside your scripts) [14:43] i am trying to retrive key from keyserver.. not cache [14:44] "Server indicated a failure" [14:46] ivaat, missed that part sorry [14:46] np [14:48] Guest97, get a usb-drive with Windows-instalation-ISO you can grab that directly from Microsoft, CheckDisk can only be run from Windows there is no way around that [14:49] that or an external hard drive... ? [14:49] those only 2 methods [14:51] Guest97, storing the files tempoarirly on another drive and formating the faulty NTFS-filesystem obviosly works as well yes. [14:51] ok i will try with my usb drive first with windows 10 [14:51] i mean windows 7 [14:52] Guest97, either should be fine, from there just go to the recovery command prompt and run chkdsk /p C: [14:53] do you know to burn a windows 7 iso file in linux? [14:59] Guest97, Try this, the instructions are for Windows 10, but it possibly works for Windows 7 as well https://www.omgubuntu.co.uk/2017/06/create-bootable-windows-10-usb-ubuntu [14:59] ok thank you [14:59] i will try now [15:02] whats the default gnome shell font used? [15:02] Guest97, No problem :-) [15:04] goddard: cantarell? [15:04] Cantarell [15:04] it appears the AMD open driver has font rendering issues in gnome shell at high refresh rates [15:05] makes it look all blurry and "rainbowy".. yes that is a technical term [15:05] the AMD open driver doesn't know a thing about fonts [15:05] knowing a thing about a thing doesn't mean it can't effect a thing [15:05] so it must be whatever software gui toolkit or font toolkit renders them [15:06] easy to find out, so if it's gtk font rendering cause you trouble. try with qt applications if it's the same effect. and can you provide screenshots of the font renderings? [15:07] Guest97: you know Windows 7 isn't supported anymore.... [15:08] burning CDs? :O [15:09] Mekaneck, yes but i'm not installing it.. i will burn it to usb for the sole pupose of running chkdsk command [15:10] anyone can hit how i can get fsck.mode=skip into this solution: https://help.ubuntu.com/community/LiveCDCustomization [15:10] not getting it on how you would do that since windows doesn't do live sessions Guest97... anyway, we're offtopic here [15:12] i can't do live sessions like linux?? [15:12] with windows? No [15:12] do i have to install it alongside ubuntu for this to work?? [15:12] Guest97: everything I have told you should of already be a solution [15:13] hint* [15:13] Guest97: this is not a windows support channel so [15:13] goddard: i will try this from recovery windows 7 to run cmd [15:13] if your drive is failing then you might need custom software to salvage your data [15:13] you shouldn't need windows but if thats what you wanna do [15:14] zteam so far told me i have 2 options, this or buy a new DIsk and move data to it.. === Guest24 is now known as kulukuy [15:17] your problem is how your drive is mounted [15:17] ntfsfix should solve the issue [15:18] goddard, he can't fix an NTFS-drive with serious corruption from Windows, there is no Linux equivalent to CheckDisk ntfsfix is not a full-blown replacement for CheckDisk that' [15:18] i ran ntfsfix command and that's what returned [15:18] Mounting volume... OK [15:18] Processing of $MFT and $MFTMirr completed successfully. [15:18] Checking the alternate boot sector... OK [15:18] NTFS volume version is 3.1. [15:18] NTFS partition /dev/sda3 was processed successfully. [15:18] yeah [15:18] how is the drive mounted? [15:19] fstab [15:19] goddard, that's why I told him to grab a Windows iso [15:22] Guest97, just for curiousity you can try to run sudo dmesg -wH [15:23] I am getting started with containers using systemd-nspawn. I'm working on getting private networking setup. Most guides use sytemd.networkd for this. Ubuntu Server comes with Netplan. What is ubuntu community's recommended configuration (arrangement)? [15:23] no specifc partition? [15:25] Guest97, from terminal, and then to mount the ntfs volume again, this should show any complains Ubuntu has about the NTFS-volume (it can't fix anything, but it let us see, if Ubuntu has any complains about the fileystem [15:26] where i can see if there's a problem in the output? [15:26] My main goal is to have a clone of the running system started up in a container, but with a new static IP address. [15:26] Guest97, it will print all errors in the same window as you enter the command :-) [15:27] there's hella a looot of lines [15:27] colored [15:27] so much lines to check [15:28] Guest97, right.... we could make that easier.... [15:29] guest press ctrl+c instead run sudo dmesg -wH | grep /dev/sda3 === thegodsquirrel is now known as THEGODSQUIRREL [15:29] with sda3 mounted? [15:30] Guest97, unmount and run the command again then remount [15:31] konsole freezes? [15:32] type Xrtil +C? [15:33] Guest97, opens a new console then, and run sudo dmesg -wH | grep /dev/sda3 from there instead [15:34] how long should the command take? [15:35] cursor just sticks to the left of the console [15:35] with no output [15:35] Guest97, even after you mount the faulty drive? :-O [15:36] i haven't mounted the drive yet... [15:36] it's still unounted [15:36] it's a partition* [15:37] i mounted [15:37] and rand the command... [15:37] still the same [15:38] Guest97, dmesg runs until we cancel it, it just prints events from the systemlog, the idea is to see any errors happening, it only shows events made after the dmesg is ran not before [15:41] Guest97, did you run the command after you mounted the drive, if that's the case it cant see it ? :-) [15:41] should i unmount again? rune the command and then remount? [15:42] Guest97, unmount the drive, run the command let it run in the background, and mount the drive, hopefully it will show you something useful [15:42] :-) [15:43] null [15:43] :( [15:43] btw usb is ready [15:44] Guest97,weird... we can try sudo dmesg -wH | grep mount [15:45] ok there's output [15:45] Guest97, paste it on pastebin please [15:45] ok [15:47] https://privatebin.net/?819da867102e3a92#6kMbZjjvpvfkTD49tYuhD2z99NKABhNyoiPqQy7EL5q7 [15:53] Guest97, really weird, it shows no complains about the filesystem... I would still just boot up Windows install and run CheckDisk from that... [15:53] zteam ok i will boot into the usb now [15:54] brb [15:54] Guest97, yes go ahead :-) === humberto1 is now known as humbertowoody === diskin is now known as Guest9075 === diskin_ is now known as diskin [16:12] anyone can hit how i can get fsck.mode=skip into this solution: https://help.ubuntu.com/community/LiveCDCustomization [16:20] i accidentally closed the chat last time, i asked what virtualization system i should use to run windows 10 "without pain" (<--- asking too much?) :D [16:20] i need to run few platforms that require win [16:20] i tried with wine but i had some problems [16:20] what tool can i use? virtualbox? [16:21] !irclogs [16:21] Official channel logs can be found at https://irclogs.ubuntu.com/ . LoCo channels are now logged there too. Meetingology logs at https://ubottu.com/meetingology/logs/ [16:22] omg i think is better to ask the quesiton again.. :D many many many logs [16:22] :D [16:22] you could use virtualbox for that sure [16:22] virt-manager or gnome-boxes or virtualbox (upstream packages, not those in ubuntu) [16:23] found [16:23] virt-manager [16:23] https://irclogs.ubuntu.com/2021/08/14/%23ubuntu.txt [16:24] ravage told me to use that tool [16:24] zteam [16:25] i cant find gnome-boxes [16:25] on software manager [16:26] pycode: virt-manager isn't the greatest user interface, but it does get updated, and builds a proper KVM instance. virtualbox has a more intuitive user interface but fewer options and is from oracle. gnome-boxes has a nice user interface and does kvm as well, but with fewer options and i'm not sure it will get updates (a gnome project) [16:27] the apt package gnome-boxes is present on all supported ubuntu releases in the universe section [16:27] boxes Updated [16:27] June 7, 2021 [16:27] !universe [16:27] The packages in Ubuntu are divided into several sections. More information at https://help.ubuntu.com/community/Repositories - See https://wiki.ubuntu.com/RecommendedSources for the recommended way to set up your repositories. [16:28] yes, gnome-boxes actually does seem to get pretty regular updates upstream [16:29] i will go with it but...one more question, what happen if i install win 10, i suppose i have to buy a license, and i do not know how they can work on a virtual system. i mean, if i remove it and install again i need to buy another key? [16:30] pycode: you may imagine that we don't do windows licensing consultation on #ubuntu as a standard service [16:30] Guest528, yes, I'm here again :-) [16:30] ahaha :) tomreyn yes i can understand [16:30] ahahah [16:31] zteam: thank you sir! it worked! i just ran chkdsk on bot partitions and it solved the problem.. <3 [16:31] thank you [16:32] ivaat: you'd need to integrate it with both syslinux (gfxboot) and grub by editing their respective configuration files. [16:32] ivaat: this guide is old, though, i'm not even sure it's still current [16:32] give it a try, though [16:32] tomreyn: but with 20.04 there seems not to be grub.cfg under extracted iso [16:33] Guest528, Told you ;-) ntfsfix can fix small issues, but it was never meant to be a full replacement to CheckDisk :-) [16:33] ivaat: hmm, i woudn't know why that is [16:34] zteam: yes thank you.. windows indeed has some powerful tools.. [16:34] ivaat: are you extracting the el-torito boot images? [16:36] I just said that i use virt-manager :D [16:36] Guest528, well, Windows has the most powerful tools for managing its own filesystem ;-) Linux has powerul-tools for almost all common filesystems except NTFS :D [16:37] TJ-: el-torito? i have done "Extract the CD .iso contents" and "Extract the Desktop system" according to https://help.ubuntu.com/community/LiveCDCustomization [16:37] Well time to eat :-) [16:37] ivaat: GRUB is the EFI boot loader. There's an EFI-SP stored as an El-torito header [16:37] bon appetite and thank you very much for your time zteam [16:38] TJ-: so i should do the EFI part of instructions as well? [16:40] zteam: interestingly, linux 5.15 or so will have paragon NTFS driver included, much faster and better aiui [16:41] ivaat: looking at that guide, and the section on re-assembling, the "sudo mkisofs ..." command isn't including an EFI-SP bootloader in the el-torito section, so would only boot in BIOS mode [16:44] enyc, isn't that the driver they already use in Linux kernel?, it would be awesome if Paragon did provide there own version of CheckDisk for us as well :D [16:45] zteam: no, ntfs-3g is different project entirely [16:47] Hey, does anyone know how to limit cpu usage of a process to 75%? in ubuntu vps [16:47] !info dumpet | ivaat you can use this tool to extract the El-Torito headers [16:47] ivaat you can use this tool to extract the El-Torito headers: dumpet (2.1-12build1, hirsute): dump information about bootable CDs and other similar formats. In component universe, is optional. Built by dumpet. Size 12 kB / 43 kB [16:47] TJ-: alright i will try [16:49] enyc, that much I actually knew... but ntfs-3g was orginally a fuse-driver, I thought the kernel already did use something different than NTFS-3G :D [16:50] It seems Ubuntu is switching Kernels quite [16:50] ivaat: "dumpedt --dumpdisks --iso file.iso" will create file.iso.0, file.iso.1 - check their sizes. Larger will be the EFI-SP vfat file-system. Then do e.g. "sudo losetup --find --show file.iso.1" and with the resulting loopdev, "sudo mkdir /mnt/efisp; mount /dev/loopX /mnt/efisp; find /mnt/efisp" [16:50] ivaat: typo!! dumpedt == dumpet [16:51] It seems Ubuntu is switching Kernels on 20.04. quite often. I had 5.11 installed, then Ubuntu downgraded me to 5.4, now back to 5.11. I think I was also on 5.8 for some time... Why is this? === TiZ3 is now known as TiZ [16:58] siwica: it depends on the ubuntu release/version [16:58] siwica: they maintain a 5.4LTS kernel [16:58] siwica: there is also a "hwe" hardware-enablement kernel series too [16:59] siwica: the HWe series changed from 5.8 to 5.11 etc Ithink [17:22] enyc: I am just on the normal 20.04 LTS version. And I don't understand why my kernel versions jump so much. Especially why they occasionally get downgraded by a lot. [17:24] How do I know if I am on the hwe series? Is ubuntu performing some sort of hardware check and put me on this series if I am on otherwise unsupported hw? [17:25] siwica: HWE series is automatic from the 20.04.2 installer IIRC. [17:25] If you use the original 20.04(.0) installer you won't be on it. [17:25] There are metapackages that identify what you're on. [17:26] Ok, apparently I am on "Ubuntu 20.04.2 LTS" according to /etc/lsb-release [17:26] https://wiki.ubuntu.com/Kernel/RollingLTSEnablementStack [17:26] That doesn't tell you what installer you used [17:26] Thanks. I will have a read! [17:26] siwica, cat /var/log/installer/media-info [17:27] Look for "hwe" in your installed package list. [17:27] I'm going from that wiki doc [17:27] ioria: "Ubuntu 20.04.1 LTS "Focal Fossa" - Release amd64 (20200731)" [17:28] siwica: you might also be interested in https://wiki.ubuntu.com/Kernel/StableReleaseCadence [17:28] Kernel fixes are so frequent that they are released on a schedule instead of ad-hoc like most other fixes [17:29] rbasak: Thank you, I will have a look. [17:30] What bothers me is that, some kernel updates seem to require me to manually re-install all nvidia-related stuff. [17:30] Otherwise I would not really care all that much. [17:30] I don't have much experience of that side of things (I don't own their hardware). [17:31] how did you install "all nvidia-stuff" initially? [17:31] rbasak: Probably for the best (when it comes to drivers...) :) [17:31] Yeah that's why I prefer hardware that has in-tree drivers :) [17:31] tomreyn: I think I used that ubuntu tool that's dedicated for installing 3rd party drivers. Forgot its name though. [17:32] siwica: if you'd liek to read yet more - i tried to sum up HWE vs GA vs LTSE at https://gist.github.com/tomreyn/8d7675840d7bc7389b32e4d8887ca449 [17:32] okay, i had expected the ubuntu-drivers utility to handle those kernel changes automatically [17:33] tomreyn: Alright, I'll also add that to my reading list. :) Thanks! [17:34] tomreyn: Seems not to always work for me. I find myself doing "sudo apt purge nvidia-*" every two months or so. [17:35] tomreyn: nice writeup! Fancy integrating that into wiki.ubuntu.com? It is a wiki - feel free to edit. You shouldn't feel blocked by "offical-ness". [17:41] rbasak: i was under the impression that the wiki is being deprecated. [17:41] and i've not yet fully convinced myself that community. is the place to put things. i won't mind putting a copy to one of the two === bigpod9 is now known as bigpod [17:42] siwica: hmm, maybe file a bug on this, if you can recreate the situation. i don't think this is how it's meant to be. but then, yes, nvidia drivers... [17:44] tomreyn: I'll pay somewhat closer attention the next time this happens. I've somehow accepted that's the way it sometimes is, but it surely is not ideal. === Ali-Karam__ is now known as Ali-Karam [17:45] If I upgrade the kernel but the nvidia drivers don't update accordingly I sometimes end up in situations where I see the login-screen but cant move the mouse pointer. Then only removing xorg.conf and manually re-installing the nvidia stuff seems to work. [17:46] siwica: The X team may not even be aware this happens, so a bug report will be useful, I think, also if they are aware. [17:46] *If Ubuntu upgrades the kernel [17:46] siwica: nvidia drivers not "updating" translates to "dkms build of nvidia module source failing"... you can easily check for that situation BEFORE rebooting with "dkms status" [17:47] siwica: for each kernel version it should report the module as 'installed' [17:48] I actually just got put on 5.11 again (was on 5.4) before and did not reboot yet. So it might aswell happen today :) [17:49] siwica: check what "dkms status" reports then :) [17:50] It only mentions v4l2loopback (for all recent kernel versions including the new one) right now. [17:51] tomreyn: nvidia still does use dkms doesn't it? I cannot think how else it would manage it [17:51] how in the hell do you enable dhcp on ubuntu-server [17:51] siwica: check the directories where the module sources are put "ls -latr /usr/src/" [17:51] TJ-: i think so. but i don't use it either [17:51] every guide says install dhcp but how can you without internet [17:51] UserUS: with the systemd.network file's [Network] DHCP=yes [17:52] UserUS: or if you're using netplan (/etc/netplan/*.yaml) then adding its dhcp: ... directive [17:52] under netplan it has one labeled "installation" with literally nothing in it [17:52] you have anything on the systemd one? [17:53] is it possible to direct GRUB to prefer one monitor over the other in a dual monitor setup? Currently it displays on a second, rotated monitor [17:53] UserUS: -server uses systemd-networkd which expects configuration files in /etc/systemd/network/ --- or runtime-only configs (as generated by netplan) in /run/systemd/network/ [17:54] UserUS: see https://netplan.io/examples/ [17:54] neither of those exist [17:55] actually the network folder one exists but has nothing in it [17:56] UserUS: right, that is expected if you've not created them! [17:57] I'm on Ubuntu focal. There are some commands which don't have manuals, but are standard, AFAIK. Like `shopt` . How can I install their mans, if there is any? [17:57] I don't like netplan getting in the way so I remove it and write the static configs to /etc/systemd/network/ [17:57] UserUS: Can't you just configure a static IP/Gateway/... if internet access is the problem? [17:57] alzgh: that's a shell-built-in. See "man bash" [17:57] at least temporarily [17:58] alzgh: or do "help shopt" [17:58] got it, thanks TJ- [17:58] UserUS: the first basic DHCP example at netplan.io should get you started [17:59] this is the stupidest shit that it doesn't come with dhcp. I would but nothing's setup right to even add a static to it, it doesn't have any files required [17:59] UserUS: As I said, it DOES have DHCP client waiting, but you've got to configure it. [17:59] It does not, it uses files to do so which are not present [18:00] UserUS: it's a server - you, as the admin, are expected to /configure/ it [18:00] yeah no, no admin thinks having to manually configure something is above and behond [18:01] the fact it asks about it during setup and that doesn't even work but just messes up the config is even better [18:01] UserUS: which installer image did you use? because it should have written a netplan yaml file. It sounds like something broke. or cloud-init is broken [18:02] the one from their website [18:03] I added it and dhcp4 is set to tru but it doesn't grab one [18:04] UserUS: did you apply the netplan config? is there a DHCPv4 server on the link it is connected to? [18:04] I did but it did nothing [18:04] Yes [18:05] UserUS: check if netplan actually generated the systemd-networkd config: "ls -latr /run/systemd/network/" - you should see a .network file there [18:06] UserUS: if you see a file there, do "cat /run/systemd/network/" and pastebin the content so we can check it looks correct [18:06] UserUS: oh, and first, ensure the networkd service is running! "systemctl status systemd-networkd" [18:07] TJ-: /usr/src includes a directory `nvidia-470.57.02`. This however was not updated when I just updated from 5.4 to 5.11 [18:07] siwica: that won't update, but it is what dkms uses to build the nvidia module [18:08] siwica: try this: "sudo dkms install nvidia/470.57.02" [18:08] siwica: oh and that directory does include source files and a dkms.conf doesn't it!? [18:09] TJ-: I haven't rebooted after the kernel update yet. The graphic card works without `dkms status` mentioning antything about nvidia though. [18:09] TJ-: It includes source files but NO dkms.conf [18:10] siwica: which suggests it is using the nouveau module. Try this to check which module: "lspci -nnk -d ::0300" [18:10] TJ-: I explicitely disabled the nouveau module [18:11] nvidia stuff is working (otherwise `nvidia-smi` does not show anything, etc.) [18:11] siwica: well not sure where the existing module is being found then [18:11] yeah, me neither. [18:11] siwica: "modinfo nvidia | head" will show the path [18:12] TJ: it's only got 5 lines, the name, under network dhcp=ipv4, link localaddressing ipv4 and route metric=100 and usemtu=true [18:12] UserUS: good, so it was created, so we know netplan did its job. Is the systemd-networkd service running? [18:13] yes [18:13] UserUS: then DHPC server isn't giving out a lease it seems [18:14] UserUS: we could enable debug-level logging and then restart networkd and see what it logs - that would reveal what networkd is seeing/nnot seeing [18:14] * enyc meows [18:15] siwica TJ- could have been recent 470 bugs issue aswell, we noticed a few passby of those in -announce [18:15] UserUS: are you able to run a tcpdump on the DHCP server to see if it is receiving a DHCP solict from this system? [18:17] UserUS: I'm also wondering if this system has an Intel I219-V Ethernet device? If so, there are known long-standing problems with thpose whereby the power-saving disables the transmit side but not the receive! [18:17] UserUS: use " lspci -nn -d ::0200" to check the device [18:18] lotuspsychje: I'm so glad I know longer allow nvidia devices! [18:18] :p [18:20] siwica: i saw something like bug #1938821 passby [18:20] Bug 1938821 in nvidia-graphics-drivers-470 (Ubuntu) "nvidia-kernel-source-470 470.57.02-0ubuntu0.20.04.1: nvidia kernel module failed to build" [Undecided, Confirmed] https://launchpad.net/bugs/1938821 [18:20] TJ: this is VM in a VCD envrionment, so there's a DHCP server and it doesn't have lscpi [18:21] UserUS: what is 'VCD' ? [18:22] UserUS: is it a cloud image you installed, or a full ubuntu-server install from an ISO? [18:23] UserUS: well, as you have control of the DHCP server, check if it is receiving the DHCP SOLICIT from the VM guest [18:24] @Ubuntu 20.04. how do I find out what's the name of the .desktop file that is associated with the terminal? i.e. gnome terminal. [18:25] hajonnes: dpkg -L gnome-terminal | grep desktop ? [18:26] hajonnes: also works the other way around dpkg -S /some/file [18:27] Thanks a lot!! :) [18:27] tarzeau_, ^ [18:29] tarzeau_, btw how can I figure out that the terminal is called gnome-terminal. In the about section of the menu it just said "Gnome Terminal". [18:29] I think I will use this for other apps and would like a general approach. [18:31] hajonnes: if you look into the desktop file it becomes clear [18:32] hajonnes: if you want to try a different terminal, try cool-retro-term for example [18:32] what is also nice, using your own font, mono fonts work all well for terminals [18:44] tarzeau_ Thanks for the suggestions. :) [18:54] I just noticed that if i boot my machine (20.04) and don't log in I don't establish a wifi connection. It's only when I login that it happens. Is that a feature? [18:55] shinobi, I believe the typical configuration pathway will store that information per-user, so yes I believe that's expected behavior [18:56] turning on autologin would fix it? [18:56] shinobi: if you want to change it, Edit Wi-Fi Network in Network Manager, and check on the first tab "All users can connect to this network" [18:56] for a generous value of "fix" [18:56] Snoop: Thanks, Do you know if I can set that behavior up on boot? My machine is basically a workstation/server. [18:56] shinobi, see rud0lf's remark [18:57] shinobi: i've got the same setup, and i'm loosing internet if the wifi signal drops (due power outage). am i alone? [18:57] SnoopJ, rud0lf: Thanks! [18:57] and should i not be alone, is there a software watchdog checking internet and rebooting if gone? something small for cron @reboot poor mans version? [18:58] tarzeau: Do you just have ... for an indicator? [18:58] shinobi: no idea, it's 1200 km far [18:58] tarzeau: I'm not understanding you. Are you at the machine? [18:59] shinobi: no, the machine is far away, and i can't check what's on the screen right now [19:00] tarzeau: I just solved a similar issue. The signal would get weak, or the router would try and switch to another band and network manager would get stuck. The router would also deauth the PC from time to time causing the same ... network prompt. [19:00] shinobi: how was your fix? [19:00] The solution was to have my modem/router just be a modem and upgrade my router. That solved all my issues. (Except for this on one boot.) [19:01] Now I'm pulling 100Mbs over wifi down to my basement. :) [19:01] (I only have 100Mbs service) [19:02] well i've got an instrument taking measures, and it's too far for cable (another building), so i'm trying with wifi, which is nuts [19:03] works two weeks, and then gone, until reboot [19:03] tarzeau: Do you have wifi in each building? [19:06] one apple airport on 2nd floor, works fairly well on about 900m2 [19:06] shinobi: Because if you don't already have it. You could just get an access point that may have more range to hit the other wifi and penty of range to hit the wnic of the device taking readings. [19:09] Tarzeau: ^^ [19:10] Tarzeau: I don't know how well Apply and linux play together. [19:12] Rud0lf: Do you know what the expected behavior is if I have both my 2.4 and 5 networks set to "available to other users"? Does it connect to the best network at boot? Does it connect to the last one? Both are set to auto connect. [19:12] i think you can set priority [19:12] not sure if -1 is stronger or weaker than 1 tho === polymorp- is now known as polymorphic [19:23] hola [19:24] ??? [19:24] hay alguien aca? [19:24] !es | ceibal [19:24] ceibal: En la mayoría de los canales de Ubuntu, se habla sólo en inglés. Si busca ayuda en español entre al canal #ubuntu-es; escriba " /join #ubuntu-es " (sin comillas) y presione intro. [19:25] thanks [19:39] Hi. I'm having an issue with 20.04 where it seems that there is a failure when trying to use bluetooth firmware. I'm trying to fix the bluetooth support. I'm on a HP ZBook 15 G3. [19:39] here is the dmesg log: https://paste.ubuntu.com/p/FkHSD8qPTV/ [19:39] "Bluetooth: hci0: Failed to send firmware data (-38)" [19:42] ens apt-cache policy linux-firmware | nc termbin.com 9999 [19:43] ioria: https://termbin.com/owx6 [19:43] ens sudo apt update && sudo apt full-upgrade [19:45] ens: show us "ls -l /sys/class/bluetooth/" [19:47] TJ-: lrwxrwxrwx 1 root root 0 Aug 17 20:33 hci0 -> ../../devices/pci0000:00/0000:00:14.0/usb1/1-12/1-12:1.0/bluetooth/hci0 [19:47] ioria: i guess i'll give it a reboot and see if it takes the new firmware [19:47] yep [19:47] ens: thanks, that confirms we're dealing with a USB BT module. You don't need to reboot [19:48] new kernel, he got 25 [19:48] now 27, i assume [19:48] ens: "sudo systemctl stop bluetooth; sudo modprobe -r btusb; sudo modprobe btusb; sudo systemctl start bluetooth" is a way to reload after adding firmware [19:49] ah, too late. already did the deed. thanks for the info though. [19:50] ioria: ens those ACPI errors (WMI) don't look great [19:50] i know === Droid is now known as Mekaneck [19:51] well seems to be no change since the reboot. [19:52] yeah the ACPI stuff seems to flash up early in the boot process. perhaps i need to do a bios update? [19:53] i have it in dual boot config so i can use hp's tool. [19:54] ens: there's a VERY recent firmware update May 2021 for your PC [19:54] ens: https://support.hp.com/us-en/drivers/selfservice/swdetails/hp-zbook-15-g3-mobile-workstation/8693763/swItemId/ob-273500-1 [19:56] ens: WOW, the changelog for that firmware is massive [19:58] ens: there are also a lot of sub-device firmware updates https://support.hp.com/wps/portal/pps/Home/SWDSelfServiceStep/!ut/p/z1/jZDLDoJADEW_hq298o670SgOvqKJgrMxqAgoMmZAvt_XykTRJl20OecmLQkKSRRRnSVRlckiyu_zWtibCTh038PIXfkWGGNTY8AXbeig4An0PDY0nTHgzc0uuDOZjfoL1wC3Sfzj40sx_Oc3AKI5PiDxRJou-JXhk0hyuX29ixVbw01IqPgQq1i1ruq-TqvqUnY0aNhfd6dHJ7K1k2cNn5RUlhWF7yRdzstliKOV12N2AxpR74A!/#panel-10 [20:02] ah nice [20:03] thanks for that. i'll go get the new firmware installed then and give it a shot. [20:27] and so it seems that although the bios update was applied - the dmesg output still contains a littany of acpi complaints and the bluetooth does not work. hrm. i might have to get a cheap usb dongle for it perhaps. [20:28] ens: Have you done a cold power off and restart? [20:29] ens: sometimes warm reboots do not reset hardware correctly (ACPI's role) [20:29] ens: especially you it dual-boots with Windows [20:29] errr you == if [20:29] i'll give that a shot too. [20:40] TJ-: i think my next question should probably be for a recommendation of a good compatible cheap usb dongle. haha. [20:41] ens: I think there will be a solution - just need a little more evidence. Firstly, enable kernel debug at start up. Do you know how to edit the GRUB menu entry at boot time, for a one-time change of the kernel command-line? [20:41] yeah i can do that [20:43] ens: OK, boot to GRUB menu, higlight the default entry, press 'e' to edit it. Navigate to the line starting "linux ..." and remove "quiet" and "splash" replace with "debug systemd.log_level=info nosplash" and press Ctrl+X to boot with those changes. Once booted "pastebinit <( journactl -k )" will show us the kernel log [20:48] ah nice yes. i see there's a lot more info here. https://paste.ubuntu.com/p/q76pk2mVyS/ [20:50] line 1030 "Bluetooth: hci0: Found device firmware: intel/ibt-11-5.sfi" [20:52] ens: let's ensure the firmware file is not corrupt. "dpkg -S intel/ibt-11-5.sfi" then "if dpkg --verify linux-firmware; then echo 'All OK'; fi " [20:53] ens: that error number -38 doesn't make too much sense, unless this is a udev issue. "#define ENOSYS 38 /* Invalid system call number */" [20:54] it echoed All OK [20:54] ens: you can find that with "grep 38 /usr/include/asm-generic/errno.h" [20:55] ens: so let's check the udevd log [20:55] ens: "pastebinit <( journalctl -u systemd-udevd -b 0 )" [20:55] interesting stuff. learning a lot about how to inspect these internals. [20:57] https://paste.ubuntu.com/p/wWwJkrbTzZ/ [20:58] did prime-run get replaced with something else or do I have to install some other package? [21:00] ens: I've found similar reports, and one solution reported is to unload and reload the module,as I suggested earlier, so lets try that [21:01] ens: "sudo systemctl stop bluetooth; sudo modprobe -r btusb; sudo modprobe btusb; sudo systemctl start bluetooth" [21:03] yeah it appears to be 'on' now and the log says it has loaded the firmware. [21:04] goddard: https://answers.launchpad.net/ubuntu/+source/nvidia-prime/+question/692775 [21:05] i was able to pair it with my bluetooth headset and play some audio through it. [21:05] thanks TJ- i think i owe you the drink price equivalent of a cheap bluetooth usb dongle at least! [21:05] The contents of arch's prime-run are basically exactly what's given in that launchpad thread in an executable script, with "$@" trailing the vars. [21:06] doesn't look like Ubuntu ever shipped something like that [21:06] ens: so it is a race condition. do you want to figure out a solution to slow it down? [21:07] TJ-: that would be educational. i'm up for it. [21:08] ens: I'm going to check the kernel source for the exact cause of -38 ... be a few minutes [21:10] Hi! how to limit cpu usage of a process? [21:11] ens: this is the function where that error originates. https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/drivers/bluetooth/btintel.c#n935 [21:12] btintel_download_firmware_payload in btintel.c [21:12] ens: the function btintel_secure_send() https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/drivers/bluetooth/btintel.c#n273 [21:12] HighAdConsultan: please do not cross post [21:16] TJ-: i guess it is not a matter of recompiling the kernel with some sort of sleep() in there? [21:17] ens that leads back to https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/net/bluetooth/hci_request.c#n128 [21:20] TJ-: do you think it's a good idea if i check out the source and try to trace back the function calls? [21:23] ens: I'm looking for the source of that -38 because that should give a clue as to the real cause.I tend to think it is a race between the host chipset (the Intel wifi 8260) and the Bluetooth chipset - they are joined together physically [21:25] ens: found it ... its the default when no other error matches, Grrrr! https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/tree/net/bluetooth/lib.c#n132 [21:27] On this page, https://ubuntu.com/security/CVE-2021-3487, what does "Needed" mean? [21:27] There's a flaw in the BFD library of binutils in versions before 2.36. An attacker who supplies a crafted file to an application linked with BFD, and using the DWARF functionality, could cause an impact to system availability by way of excessive memory consumption. [21:28] ens: there are only two lines with -bt_to_errno() in net/bluetooth/hci_request.c [21:29] ens: lines 155 and 242 [21:29] just checking them now [21:30] xMopx: https://git.launchpad.net/ubuntu-cve-tracker/tree/README#n315 [21:30] ens: and those are working on the variable 'err' (errorno) that in both cases is set just above by "err = wait_event_interruptible_timeout" [21:30] tomreyn: thanks! howd you find that? [21:30] xMopx: there is also #ubuntu-security [21:31] xMopx: i was previously told about its exitence [21:31] ens: so, either the timeout is too low, ir the device isn't ready at this point (maybe due to the iwlwifi device not having set things up on the USB interface fully at that point) [21:31] hah, alright [21:32] ens: so I think the solution is to delay loading of the btusb module for a second or two [21:33] TJ-: so the device is reporting it's ready for the next bit of processing when it's not. i see. [21:33] and then the code tries to load the firmware and gets this issue [21:34] ens: looks that way. I have a delay solution for you [21:34] TJ-: the suggestion i see is to use a systemd definition to load it after the fs gets mounted or such [21:36] ens: " echo "install btusb /usr/bin/sleep 2; /usr/sbin/modprobe --ignore-install btusb" | sudo tee -a /etc/modprobe.d/btusb.conf " [21:36] ens: see "man modprobe.d" for how 'install' works to execute a command in this way [21:41] ah nice so whenever modprobe tries to install btusb it will instead run that sequence. (just read the manual) [21:41] that's handy. [21:43] ens: now you just need to do a reboot test [21:44] well, seems to be working 100% now. [21:44] thanks for the tour! [21:45] ens: the reboot worked? [21:45] yep. it's all chugging along fine now [21:45] fantastic. We'd best remember this one for any others with the same race condition [21:45] tomreyn: ^^^ [21:46] TJ-: hmm? [21:47] remember this race condition solution for bluetooth firmware failure loading [21:47] it's a hard one to diagnose, but a simple solution [21:48] oh, thanks. i can try, but bug reports usually work better [21:48] tomreyn: that's next :) [21:48] :) [21:54] Bug #1940332 [21:54] Bug 1940332 in linux (Ubuntu) "hci0: Failed to send firmware data (-38)" [Undecided, New] https://launchpad.net/bugs/1940332 [21:56] thanks [22:00] really appreciate all the help TJ- . === jwisbell35 is now known as Kathadris [22:08] tomreyn: what is this about? [22:14] goddard: what do you mean? [22:16] goddard: oh you're referring to https://answers.launchpad.net/ubuntu/+source/nvidia-prime/+question/692775 - it explains how you could create your own "prime-run" since this does not exist on Ubuntu [22:16] also alternatives [22:33] i gotcha [22:33] thanks [22:38] okay the prime_run thing did nothing [22:39] but just seting it to on-demand while the AMD card being the default video output card works [22:39] even though xorg is forced [22:39] still i can live with this [22:40] hopefully it is stable though [22:40] a lot of the details are wrong i think though [22:41] so odd how this works [22:44] Buenas alguien en la sala que hable español ? [22:44] is it safe to apply this https://forum.manjaro.org/t/howto-use-wayland-with-propietary-nvidia-drivers/36130 in ubuntu and will it work? [22:44] !es | cuchy [22:44] cuchy: En la mayoría de los canales de Ubuntu, se habla sólo en inglés. Si busca ayuda en español entre al canal #ubuntu-es; escriba " /join #ubuntu-es " (sin comillas) y presione intro. [22:45] ok SnoopJ [22:45] lo siento :/ === pasiz0 is now known as pasiz