| sparr | brightness hotkeys (Fn+F11 and Fn+F12) are not working on my dell xps laptop after upgrading to eoan. xev reports no key events when I press them. | 02:31 |
|---|---|---|
| sparr | I am able to adjust the brightness by writing to /sys/class/backlight/intel_backlight/brightness directly | 02:35 |
| lotuspsychje | another crash bug tomreyn ? | 09:16 |
| tomreyn | yes, yesterdays' daily desktop amd64 crashed | 09:33 |
| tomreyn | reading the log i'm not exactly sure why: "cannot resolve path '/dev/sda4'" | 09:34 |
| tomreyn | it may have to do with the existing installation on the target. | 09:35 |
| tomreyn | hmm, must be some race. i am installing with the same options and again over an existing installation and its succeeding this time. | 09:41 |
| tomreyn | hmm no luck reproducing it. | 09:50 |
| johnjbogle | Hi may I get some assistance with updating my bios please? | 09:58 |
| lotuspsychje | !biosupdate | johnjbogle | 10:37 |
| ubottu | johnjbogle: To see how to update your bios on Ubuntu visit the community collected methods here: https://help.ubuntu.com/community/BIOSUpdate | 10:37 |
| lotuspsychje | johnjbogle: see also fwupdate -l | 10:37 |
| tomreyn | looks like my bug report is a dupe of bug 1847826 | 11:05 |
| ubottu | bug 1847826 in ubiquity (Ubuntu) "ZFS installer crashes (Xubuntu)" [Undecided,New] https://launchpad.net/bugs/1847826 | 11:05 |
| tomreyn | oof crap now it even went OOM | 11:12 |
| tomreyn | and reproduced | 11:27 |
| Gargoyle | Hello fellow +1 ers. When waking from screen blank, I am getting a long delay. The monitors wake up, and the mouse cursor shows but it's frozen. It stays like this for 15-30 seconds then normally carries on OK. | 11:34 |
| tomreyn | hi Gargoyle. what are the hardware sepcifications? | 11:35 |
| Gargoyle | Ryzen 3800X + Radeon VII | 11:35 |
| tomreyn | and plenty of ram, i assume? | 11:35 |
| Gargoyle | 32GB | 11:36 |
| tomreyn | which grpahics driver? | 11:36 |
| Gargoyle | NVMe system disk. | 11:36 |
| Gargoyle | The one that's built in to the kernel. | 11:36 |
| tomreyn | so not the -pro overlay? | 11:36 |
| Gargoyle | Nope | 11:37 |
| Gargoyle | Pretty sure the only thing I had to do when I switched from nvidia cards back in January was to upgrade the kernel and manually add some missing firmware files. | 11:38 |
| tomreyn | what's in "journalctl -b 0" (or "journalctl -b -1" if you rebooted since) around the time you last had this happen? | 11:38 |
| tomreyn | i didn't mean to suggest you should use the -pro overlay, i would not personally recommend it. | 11:38 |
| Gargoyle | Errr. wow... a LOT! Let me pastebin some of this! | 11:39 |
| tomreyn | much is probably just gnome log spill | 11:40 |
| tomreyn | maybe add -k to only get kernel messages | 11:40 |
| Gargoyle | https://paste.ubuntu.com/p/kmVVC624Yx/ | 11:42 |
| Gargoyle | I am guessing 10:15:34 is the last message before display sleep, and 12:28:08 is when I just tried to wake it up?> | 11:43 |
| Gargoyle | https://paste.ubuntu.com/p/CghGkQwDyB/ <-- same timeframe but kernel only messages. | 11:45 |
| tomreyn | did you post the same log twice? | 11:46 |
| Gargoyle | Second paste is with the "-k" flag added to journalctl | 11:47 |
| tomreyn | you have a kernel oops on the second log, line 60 and following | 11:47 |
| tomreyn | and then several page faults triggered by amdgpu when trying to work with chrome, i guess | 11:48 |
| Gargoyle | So what's the best thing to do with that? Send it to kernel list somewhere? AMD? | 11:48 |
| tomreyn | can i see a full boot log? | 11:49 |
| Gargoyle | The one from where I just forced powered off and booted back up again? | 11:49 |
| tomreyn | yes, that'd work, just journalctl -b | nc termbin.com 9999 | 11:50 |
| Gargoyle | https://termbin.com/t6xy | 11:50 |
| Gargoyle | I think this was a clean 18:04 install -> upgraded to 18:10 -> then to 19:04 -> then to 19:10. If that helps. | 11:54 |
| tomreyn | oh, upgraded, good to know. i'm still reading, give me 5 more minutes | 11:54 |
| tomreyn | there's some issue with the iommu grouping, causing some pci devices to operate slowert han needed, i think, but this seems to be platform specific currently, will probably get better in later kernel versions. | 11:55 |
| tomreyn | Oct 12 12:31:02 ryzen1910 kernel: pci 0000:01:00.0: 63.012 Gb/s available PCIe bandwidth, limited by 16 GT/s x4 link at 0000:00:01.2 (capable of 126.024 Gb/s with 16 GT/s x8 link) | 11:55 |
| tomreyn | ^ indicated by this. but we can ignore this for now, i think | 11:56 |
| Gargoyle | Oh! | 11:57 |
| tomreyn | Oct 12 12:31:02 ryzen1910 kernel: Unstable clock detected, switching default tracing clock to "global" | 11:58 |
| tomreyn | (and following lines) indicate hardware clock issues, or failure of this linux version to use it properly. it worked around it, though. | 11:58 |
| Gargoyle | I've had that same message on both motherboards. Previously this system was Ryzen 1800X and a crosshair hero vi. (I upgraded mobo and CPU then just re-used the same OS without re-installing anything). | 11:59 |
| Gargoyle | I think previous googling for that clock error is what prompted me to add tsc=unstable to the kernel boot flags | 12:03 |
| tomreyn | i'm on a ryzen 1800x with asrock x370 taichi board, don't have "Unstable clock detected" in any past log, and have bene running various kernel versions incl. 5.3.0 there. | 12:04 |
| Gargoyle | Perhaps I should remove that flag now? | 12:04 |
| tomreyn | ^ i wa sjuist saying this to point out it's not generic to ryzen | 12:05 |
| tomreyn | oh i missed that you have tsc=unstable | 12:05 |
| tomreyn | yes, you should remove it if you don't strictly need it | 12:05 |
| Gargoyle | The ASUS mobo was a bit of a pig! | 12:05 |
| tomreyn | these messages i have also not seen before: amdgpu 0000:0a:00.0: remove_conflicting_pci_framebuffers: bar 0: 0xe0000000 -> 0xefffffff | 12:07 |
| tomreyn | but i suggest you reboot without tsc=unstable now (can just remove it via grub menu for this boot if you're not sure it's safe to remove) | 12:07 |
| tomreyn | and then post another log | 12:07 |
| Gargoyle | np. I've already edited my refind config... will reboot. | 12:08 |
| Gargoyle | brb | 12:08 |
| tomreyn | amdgpu 0000:0a:00.0: Direct firmware load for amdgpu/vega20_ta.bin failed with error -2 | 12:08 |
| tomreyn | amdgpu 0000:0a:00.0: psp v11.0: Failed to load firmware "amdgpu/vega20_ta.bin" | 12:08 |
| tomreyn | fix this first | 12:08 |
| Gargoyle | oh | 12:08 |
| Gargoyle | I don't have a vega20_ta.bin! | 12:09 |
| Gargoyle | I originally grabbed my vega20* files from here:- https://people.freedesktop.org/~agd5f/radeon_ucode/vg20/ | 12:10 |
| tomreyn | hmm yes, this firmware is not available, yet. ask in #amdgpu. maybe they don't want to release it. | 12:11 |
| tomreyn | and maybe it's not actually needed. or amd just forgot ;) | 12:11 |
| lotuspsychje | ryzen might have issues with kernel 5.3 | 12:13 |
| tomreyn | that's news to me, got any reference? | 12:14 |
| tomreyn | Gargoyle: https://bugs.freedesktop.org/show_bug.cgi?id=109955 has some similar logs to the first ones you posted | 12:14 |
| ubottu | Error: Could not parse XML returned by Freedesktop: ('The read operation timed out',) (http://bugs.freedesktop.org/show_bug.cgi?id=109955&ctype=xml) | 12:14 |
| lotuspsychje | im not the ryzen expert myself tomreyn but thats the reason lordievader went gentoo/kernel 4.19 | 12:15 |
| tomreyn | thanks, i'll talk to him then | 12:15 |
| tomreyn | Gargoyle: so the bug i posted suggests this can be powerplay related. | 12:16 |
| tomreyn | so disabling this (for now) may help then, i guess. | 12:16 |
| Gargoyle | Just reading it now. What's powerplay? | 12:17 |
| tomreyn | https://en.wikipedia.org/wiki/AMD_PowerPlay | 12:17 |
| tomreyn | power management / saving | 12:17 |
| lotuspsychje | tomreyn: kinda related recall: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1842785 | 12:17 |
| ubottu | Launchpad bug 1842785 in linux (Ubuntu) "Freeze after waking up from suspend on a Ryzen 3000 based system" [Undecided,Confirmed] | 12:17 |
| Gargoyle | ^^ I didn't search for suspend because I have auto suspend off - just screen blanking. | 12:19 |
| tomreyn | Gargoyle: the firmware is in one of AMD's proprietary packages: https://www.reddit.com/r/linuxquestions/comments/b3enfw/amdgpu_radeon_vii_direct_firmware_load_for/ | 12:23 |
| tomreyn | i guess this can mean it is not allowed for redistribution, not sure. | 12:23 |
| tomreyn | the exact link would now be http://repo.radeon.com/rocm/apt/debian/pool/main/r/rock-dkms/rock-dkms_2.9-6_all.deb | 12:24 |
| tomreyn | it's there: data.tar.xz:/./usr/src/amdgpu-2.9-6/firmware/amdgpu/vega20_ta.bin | 12:26 |
| Gargoyle | Should I take all .bin's from there to keep them matched up? | 12:28 |
| tomreyn | normally you should only install linux-firmware updates, which should provide all legally installable firmwares | 12:30 |
| tomreyn | that's a package name | 12:30 |
| tomreyn | https://people.freedesktop.org/~agd5f/radeon_ucode/vg20/ are what an amd employee pushes to the public for linux. | 12:30 |
| tomreyn | if they look fine, they end up at https://git.kernel.org/pub/scm/linux/kernel/git/firmware/linux-firmware.git/tree/amdgpu | 12:31 |
| tomreyn | and from there are packaged into ubuntu | 12:31 |
| tomreyn | all of these steps introduces delay, though, so YMMMV | 12:31 |
| tomreyn | but note that (at least supposedly, haven't seen this myself) there have been situations where people just copied all of ~agd5f/radeon_ucode and it created problems which had not emerged if they had just used linux-firmware.git | 12:33 |
| tomreyn | by the way there is also #radeon - which may be better than #amdgpu - i'm not sure how to tell them apart. | 12:35 |
| johnjbogle | lotuspsychje: is that fwupdate -l a terminal command? | 12:36 |
| lotuspsychje | johnjbogle: yes | 12:36 |
| lotuspsychje | johnjbogle: are you using ubuntu 19.10? | 12:37 |
| Gargoyle | thanks for all the help so far tomreyn. Going for the reboot now! (fingers crossed) | 12:38 |
| johnjbogle | im not sure what version. Terminal says that command not found | 12:39 |
| Gargoyle | Well. I'm back - so that's a good start! :D | 12:40 |
| Gargoyle | Fresh boot log: https://termbin.com/edz7 | 12:42 |
| tomreyn | lotuspsychje: i would recommend using fwupdmgr rather than fwupdate these days. see also the notes on fwupd at https://github.com/rhboot/fwupdate | 12:43 |
| tomreyn | "Most users should apply UEFI capsule updates with a higher level tool such as fwupd." | 12:44 |
| lotuspsychje | ok tnx tomreyn | 12:45 |
| tomreyn | yw :) | 12:46 |
| tomreyn | Gargoyle: on a first glance, this looks better | 12:46 |
| tomreyn | Oct 12 13:39:03 ryzen1910 kernel: sched_clock: Marking stable (527835196, -23995170)->(529166236, -25326210) | 12:48 |
| Gargoyle | \o/ | 12:49 |
| * Gargoyle goes to stomp all over the old ASUS mobo! | 12:49 | |
| Gargoyle | A a final FU from that mobo, it bent some of the pins on my old 1800X during removal.... OK. I can't really blame that on the mobo! ;-) | 12:51 |
| tomreyn | hmm "AMD-Vi: AMD IOMMUv2 functionality not available on this system" - is iommu disabled in bios? | 12:51 |
| Gargoyle | If it is, it's been like that by default from ASRock - I have literally only changed fan speed settings. | 12:53 |
| tomreyn | yes, that's well possible, i've seen ryzen board where it defaults to disabled | 12:53 |
| Gargoyle | will double check. brb... | 12:53 |
| tomreyn | also check for AMD-V | 12:53 |
| Gargoyle | Back again. | 13:01 |
| Gargoyle | Couldn't find anything related to IOMMU in BIOS. | 13:02 |
| Gargoyle | brb again... it might have a silly name (Advanced>AMD CBS>NBIO Common Options) :/ | 13:05 |
| Gargoyle | tomreyn: Something is not happy! | 13:13 |
| Gargoyle | https://termbin.com/m1u5 | 13:16 |
| Gargoyle | Had a corrupt GDM login screen, but could still type password and login - main session started but it was slloooooow to begin with. | 13:17 |
| Gargoyle | gnome shell is not happy. I now have a functioning terminal window which I cannot move or resize! | 13:20 |
| Gargoyle | ha ha.. this is freaky. gnome-shell (or the compositor or whatever) things the window is smaller than it's displayed. The grab handle is about an inch from the top of the window! | 13:22 |
| lotuspsychje | oO | 13:30 |
| lotuspsychje | Starting GNOME Shell on Wayland... | 13:30 |
| lotuspsychje | thats a lot of errors after wow | 13:31 |
| lotuspsychje | Gargoyle: are you logging into wayland on gnome? | 13:32 |
| lotuspsychje | Gargoyle: your dmesg spits out a lot of errors after that | 13:34 |
| tomreyn | the first graphics related error (?) i see is kernel: [drm] psp command failed and response status is (256) | 13:34 |
| tomreyn | the "AMD IOMMUv2 functionality not available on this system" message from earlier is now gone, though, maybe due to the clock now being stable. | 13:35 |
| Gargoyle | Current boot with IMMOU disabled again:- https://termbin.com/4utmm | 13:36 |
| tomreyn | so you did find an iommu option in bios? | 13:36 |
| Gargoyle | Yeah. I't was burried. Turning it on caused a l;ot of issues! | 13:37 |
| lotuspsychje | Gargoyle: using alot of gnome extensions too i see? | 13:37 |
| Gargoyle | 3 | 13:37 |
| Gargoyle | Is that a lot? | 13:37 |
| lotuspsychje | its reccomended to disable those when testing things on gnome | 13:38 |
| lotuspsychje | saves the developers much work debugging | 13:38 |
| lotuspsychje | Gargoyle: did you clean install a 19.10 daily or beta? or was this an upgrade from 19.04? | 13:40 |
| Gargoyle | lotuspsychje: upgrade | 13:40 |
| lotuspsychje | thats what i thought | 13:40 |
| lotuspsychje | also reccomended to test devel versions clean install to avoid things like you experience | 13:41 |
| Gargoyle | I'll be getting some new drives soon, so will be due for a clean install. | 13:41 |
| Gargoyle | lotuspsychje: Gotta test the upgrade path too though! ;-) | 13:42 |
| lotuspsychje | yeah i understand that needs to be tested at some point too | 13:42 |
| lotuspsychje | but still understand also, if creating bugs this situation makes it real hard for yourself & the devs to find where it comes from | 13:43 |
| lotuspsychje | at least comparing with a clean daily, to see if you can reproduce things the same | 13:44 |
| tomreyn | Gargoyle: amd powerplay can be controlled via a feature mask. amdgpu.ppfeaturemask= kernel parameter. but you'll need to look up the values yourself in the source code (or maybe on some websites) | 13:44 |
| tomreyn | Gargoyle: personally, i prefer gnome-shell on X for now (but i'm also on 18.04 LTS), ubuntu's primary target is also X for now. | 13:45 |
| tomreyn | 20.04 LTS wont default to wayland | 13:45 |
| Gargoyle | I'm not on wayland as far as I know. | 13:47 |
| tomreyn | maybe it was just gdm, then it's ok, i didn't look for it in your logs, yet | 13:47 |
| lotuspsychje | Gargoyle: dmesg said you were | 13:48 |
| lotuspsychje | Starting GNOME Shell on Wayland... | 13:48 |
| Gargoyle | It starts it and stops it again!? | 13:50 |
| lotuspsychje | its your system Gargoyle i dont know whats all hapenning to it | 13:51 |
| lotuspsychje | but your dmesg is drowning in errors | 13:52 |
| Gargoyle | Possibly I experimented with wayland on 18:04 or 18:10... deffo not tried it in a while. | 13:53 |
| tomreyn | yes, you ran gnome on wayland. you choose the session upon login after selecting the user to login with, by clicking the little cog | 13:54 |
| tomreyn | from what i can tell, powerplay is the main issue. iommuv2 looked fine to me. | 13:54 |
| Gargoyle | Don't remember doing that. I thought "env" output made some reference if you were running on wayland. | 13:54 |
| tomreyn | echo $XDG_CURRENT_DESKTOP $XDG_SESSION_TYPE | 13:55 |
| tomreyn | that's in env, yes | 13:55 |
| Gargoyle | tomreyn: So you think if I find the correct code to dissable powerplay, I should be ok to try and re-enable iommu in the bios? | 13:55 |
| tomreyn | yes, that's what i think | 13:56 |
| Gargoyle | https://paste.ubuntu.com/p/54bmqhxzCS/ | 13:56 |
| Gargoyle | x11 | 13:56 |
| Gargoyle | I'm pretty sure I only every tried wayland once - quite a while ago | 13:57 |
| Gargoyle | *ever | 13:57 |
| tomreyn | hmm i don't have "Starting GNOME Shell on Wayland" on my log, though | 13:58 |
| tomreyn | also using gnome-shell here | 13:58 |
| tomreyn | let me see on 19.10 | 13:58 |
| Gargoyle | Is there something wayland related I should be removing? | 13:58 |
| lotuspsychje | Gargoyle: when you say messign with wayland, did it envolve editing things? | 13:59 |
| Gargoyle | I thought it was currently something that just lives in the background while its still being developed. | 13:59 |
| Gargoyle | maybe I have some packages I need to remove! :/ | 14:01 |
| tomreyn | it seems to be normal that you first have "Starting GNOME Shell on Wayland" and later "Starting GNOME Shell on X11" on your logs on 19.10. | 14:01 |
| tomreyn | Gargoyle: can you journalctl -b | grep 'Starting GNOME Shell on' | 14:02 |
| Gargoyle | Yeah. The only non-library package I have installed for wayland is "xwayland" and if you try to remove that it wants to take ubuntu-desktop ubuntu-desktop-minimal and ubuntu-session with it. | 14:03 |
| Gargoyle | https://paste.ubuntu.com/p/zrr85WZGwr/ | 14:04 |
| Gargoyle | Looks like it starts both... | 14:04 |
| Gargoyle | But only stops the wayland one 13 seconds later | 14:05 |
| Gargoyle | https://paste.ubuntu.com/p/DRVkv4Snj8/ | 14:06 |
| Gargoyle | Right... I've got to go for a bit. I'm gonna let this thing "sleep" and then come back and see what happens! | 14:06 |
| Gargoyle | Thanks for all the help! :-) | 14:06 |
| tomreyn | i guess it ust starts those to test whether they're available to build the available options for the gdm3 session menu | 14:06 |
| cybercrypto | Hi there, I see that 'zfs option' is missing from the installer GUI during 19.10 installation. I am using freebsd-host virtualbox installing guest-ubuntu-19.10. Does anyone knows something that may prevent zfs option to be presented? | 15:13 |
| kinghat | if you install kubuntu desktop on top of ubuntu, both DE will have their own updates, correct? | 16:08 |
| bittin | Ubuntu 19.10 RC2 on Monday and release on Thursday | 16:14 |
| Gargoyle | tomreyn: Just "woke" pc up again. No pauses or anything. \o/ | 17:00 |
| Gargoyle | Plenty of errors in the journal, but I don't think they are related to the earlier problem! :-) (https://termbin.com/s1iy) | 17:00 |
| tomreyn | bittin: So, this and your previous announcement make it sound (at least to me, so it could be my fault) like the RC was a separate release, but, quoting someone who certainly knows well, "RCs are just the dailies leading up to final release, there's no formal publishing for them". also, RC2 is not this monday. | 17:39 |
| tomreyn | well at leats not according to https://wiki.ubuntu.com/EoanErmine/ReleaseSchedule | 17:40 |
| tomreyn | Gargoyle: so which changes do you think made the wake from dpms work well this time? | 17:45 |
| tomreyn | Gargoyle: also, do you have one of these packages installed? mariadb-server-10.3, mysql-server-8.0? the reason i'm asking is because you have a broken apparmor rule at /etc/apparmor.d/usr.sbin.mysqld (search this on yuor log) | 17:48 |
| bittin | Happy 15th Birthday Ubuntu | 18:08 |
| tomreyn | Gargoyle: "Starting MySQL Community Server..." - so it appears to be installed, but I suspect that your copy of /etc/apparmor.d/usr.sbin.mysqld can be an old one. | 18:10 |
| cybercrypto | Hi there, I see that 'zfs option' is missing from the installer GUI during 19.10 installation. I am using freebsd-host virtualbox installing guest-ubuntu-19.10. Does anyone knows something that may prevent zfs option to be presented? | 18:22 |
| Sbur3 | I'm trying to upgrade to 19.10. Most is good. But I get 'errors were encountered while processing" ... linux-firmware, gosa, linux-image-lowlatency, linux-lowlatency, initramfs-tools, as well as linux-image-5.3.0-13-lowlatency and 5.3.0-17-lowlatency. That doesn't prevent 19.10 from working. I imagine it will work starting on Oct 17? | 18:23 |
| tomreyn | cybercrypto: i don't know what "guest-ubuntu-19.10" is, but which installer image are you using? | 18:24 |
| cybercrypto | tomreyn: I figure it out. The beta 19.10 image does not allow zfs installation yet. I had to download the daily snapshot from current builds. It is now showing the zfs experimental option during install. | 19:12 |
| tomreyn | cybercrypto: right ;) | 19:13 |
| cybercrypto | tomreyn: guest-ubuntu-10.10, means that I am installing ubuntu, version 19.10 as a guest operating system, within virtual box host-freebsd. | 19:13 |
| cybercrypto | tomreyn: hope it is clear now. | 19:14 |
| tomreyn | cybercrypto: it is, it just looked like a guest template / VDI name or something you might have downloaded elsewhere (maybe inoffocial ubuntu build), so i asked for clarification. | 19:15 |
| cybercrypto | tomreyn: I had two crashes during zfs install, and it is finally running. I am going to test the storage pool tools and play around with new devices. I reported back the crash errors to cannonical btw. | 19:17 |
| tomreyn | very well. | 19:17 |
| cybercrypto | tomreyn: sure, np. thank you. I am using official build snapshot from US ubuntu server. | 19:17 |
| valorie | kinghat: if you want kubuntu, just install it | 19:34 |
| valorie | if you want to test other desktops, sudo apt install kubuntu-desktop | 19:35 |
| kinghat | ya thats what i meant, installed kubuntu-desktop on top of ubuntu. | 19:35 |
| kinghat | wasnt sure how updating packages worked between both environments. are all packages updated regardless of environment youre on? | 19:36 |
| tomreyn | yes, during their support cycles, which can difer from those of plain ubuntu. | 19:38 |
| valorie | also you can add various PPAs to make it even more complicated | 19:53 |
| Gargoyle | booo.. Just had the hang again. :-( | 20:21 |
| Gargoyle | Yeah. I noticed that apparmor a few weeks ago, tomreyn. It's on my list to look into. I think I had mysql 5.7 - then removed it and just yesterday re-installed mysql 8.0 to help a friend with some SQL. | 20:22 |
| Gargoyle | Fresh log. I attempted to wake the machine at 21:18 - the screen wakes instantly but then hangs until giving me the login prompt. Finger is deffo pointing to powerplay (https://termbin.com/vkrw) | 20:24 |
| Gargoyle | But this makes no sense to me! - https://github.com/torvalds/linux/blob/master/drivers/gpu/drm/amd/include/amd_shared.h#L121 | 20:24 |
| tomreyn | start by just setting it to 0 | 20:25 |
| Gargoyle | Oh. so just disable everything. OK. | 20:26 |
| Gargoyle | This is going to be one of those annoying things to test though... as the wakeup a few hours ago worked fine! :/ | 20:28 |
| tomreyn | at least that's what i'd do, yes (and i'd enable iommuv2 as well) | 20:29 |
| tomreyn | note that setting it to 0 may actually make things worse, and could make it run hot, so keep an eye on the temps | 20:30 |
| Gargoyle | I can't!!! <lotuspsychje> made me turn off freon! ;-) | 20:31 |
| Gargoyle | what's the best thing to do with apparmor? just delete the mysql profile and to a forced re-install of mysql-server ? | 20:35 |
| Gargoyle | Is there any way to see the contents of the files listed here:- https://packages.ubuntu.com/eoan/amd64/mysql-server-8.0/filelist ? | 20:38 |
| Gargoyle | Looks like a force re-install should but the /etc/apparmor.d/usr.sbin.mysqld back in place, but nothing about abstractions. | 20:39 |
| tomreyn | is there not another file with similar name next to the existing apparmor profile? | 21:00 |
| Gargoyle | Nope. There is a local/usr.sbin.mysqld which is empty. | 21:02 |
| tomreyn | --force-confall or something similar should put a missing configuration file in place (but may also overwrite existing ones, i don't remember) | 21:02 |
| tomreyn | another option is to apt purge and apt install, but then you certianly loose configurations | 21:03 |
| Gargoyle | I do normally purge whenever I remove something. | 21:03 |
| Gargoyle | Let me try again. | 21:03 |
| erle- | Ctrl+Alt+T opens two terminals in Activities view | 21:03 |
| Gargoyle | OK. an autoremove --purge got rid of them. | 21:06 |
| Gargoyle | OK. And a re-install of mysql-client and mysql-server has put it back, but still with an #include <abstractions/mysql> reference which doesn't exist! | 21:08 |
| Gargoyle | OK. So that abstraction file should be provided by the apparmor package itself - can't see how it would be removed - I try to avoid manual tinkering with things like that. I'll try a reinstall of apparmour | 21:14 |
| Gargoyle | Nope. A re-install does not put the file back. Any other suggestions from anyone? | 21:16 |
| Gargoyle | I've unpacked the .deb, and the file is there - it just not being put in place when I re-install the package with apt. | 21:32 |
| valorie | sounds like a packaging problem to me | 22:02 |
| tomreyn | i don't think it is, re-installing a package that's already installed would not re-create a missing configuration file unless it's run with --force-confmiss. | 22:45 |
| tomreyn | ^ from memory, hope it's correct | 22:46 |
| Gargoyle | tomreyn: Yup. I asked in snappy (because I know snaps rely on apparmor) Someone confirmed that. At some point I must have removed the file by mistake and apt was preserving my "choice" :-) | 22:54 |
| valorie | oh good | 23:10 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!