Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
 
  • Post
  • Reply
Subjunctive
Sep 12, 2006

✨sparkle and shine✨

Volguus posted:

Maybe setxkbmap is being called by someone? User login (.profile, .bash_profile, whoever)? https://superuser.com/questions/1147320/how-to-change-keyboard-layout-in-i3

Yeah I have a setxbdmap in my i3 startup now, but that still leaves the login manager with the wrong keyboard, and I think the Lock Screen?

I couldn’t find anything else that called it, either in my config or in /etc.

I just want to change the thing that the installer sets because it’s out there somewhere messing with me…

Adbot
ADBOT LOVES YOU

VictualSquid
Feb 29, 2012

Gently enveloping the target with indiscriminate love.
Looks like you have been granted an opportunity to find out how good endeavour's forums/discord/irc are at answering endeavour specific questions.

Subjunctive
Sep 12, 2006

✨sparkle and shine✨

Yeah in the past it’s been OK, I’ll give it a whirl.

Subjunctive
Sep 12, 2006

✨sparkle and shine✨

The installer only sets 00-keyboard.conf and /etc/default/keyboard and those are correct now. But when slick-greeter starts up it sees the us-intl keyboard as being current. I wonder what is getting cached from first-run or whatever.

Volguus
Mar 3, 2009

Subjunctive posted:

The installer only sets 00-keyboard.conf and /etc/default/keyboard and those are correct now. But when slick-greeter starts up it sees the us-intl keyboard as being current. I wonder what is getting cached from first-run or whatever.

That must be something that could be changed in /etc/lightdm/lightdm.conf (or /etc/lightdm/lightdm.conf.d/...). Maybe if it's not set that's what it sets, and setting something will make it accept the new keyboard.

Subjunctive
Sep 12, 2006

✨sparkle and shine✨

Volguus posted:

That must be something that could be changed in /etc/lightdm/lightdm.conf (or /etc/lightdm/lightdm.conf.d/...). Maybe if it's not set that's what it sets, and setting something will make it accept the new keyboard.

Yeah I don’t see anything about keyboard in lightdm’s config, and it’s not logging the echo from its Xsession that shows it’s loading a modmap.

Might be time for some dtrace!

Volguus
Mar 3, 2009

Subjunctive posted:

Yeah I don’t see anything about keyboard in lightdm’s config, and it’s not logging the echo from its Xsession that shows it’s loading a modmap.

Might be time for some dtrace!

Or just add the property(ies) and see what happens.

Kivi
Aug 1, 2006
I care
Why does snap try to delete files in my home folder? Journal is full of these:

[ 891.802051] audit: type=1400 audit(1690713641.411:498): apparmor="DENIED" operation="rmdir" class="file" profile="snap.snapd-desktop-integration.snapd-desktop-integration" name="/home/kivi/tmp/" pid=91662 comm="rmdir" requested_mask="d" denied_mask="d" fsuid=1000 ouid=1000

Yes, the folder is called tmp and houses my browsers’ downloads and other misc data like that. Why is it trying to delete it?

AlexDeGruven
Jun 29, 2007

Watch me pull my dongle out of this tiny box


Maybe a legacy operation from the app when they created the snap?

SadBag
Jun 24, 2012

Something has gone very wrong for us to get to the point where Hot Dog is the admiral.
Question about dealing with snap eating all of my disk space
Computer itself is a chromebook, uname -r returns 5.19.0-46-generic, os-release returns Ubuntu 22.04.2 LTS. Ubuntu flavor is Lubuntu, ~16GB of drive space

After my computer unexpectedly shut down today, and my computer to hung whenever I entered password to login, I used alt-f2 to go to terminal, was able to login, and after attempting to update and seeing that I was out of drive space, ran df -h and saw
code:
Filesystem     Size   Used  Avail Use%  Mounted on
tmpfs          178M    17M  161M   10%  /run
/dev/sda2       15G    15G     0  100%  /
tmpfs          888M      0  888M    0%  dev/shm
tmpfs          5.0M   4.0K  5.0M    1%  /run/lock
tmpfs          888M      0  888M    0%  /tmp
/dev/sda1      511M    11M  501M    3%  /boot/efi
tmpfs          178M    72K  178M    1%  /run/user/1000
Then I tried running du -s /*
code:
0        /bin
329444   /boot
0        /dev
10780    /etc
287892   /home
0        /lib
0        /lib32
0        /lib64
0        /libx32
16       /lost+found
16       /media
4        /mnt
4        /opt
du: cannot access '/proc/1559/task/1559/fd/4': No such file or directory
du: cannot access '/proc/1559/task/1559/fdinfo/4': No such file or directory
du: cannot access '/proc/1559/fd/3': No such file or directory
du: cannot access '/proc/1559/fdinfo/3': No such file or directory
0       /proc
11556   /root
19660   /run
0       /sbin
7279650 /snap
4       /srv
524292  /swapfile
0       /sys
0       /tmp
6809168 /usr
6790296 /var
So I guess at some point /snap got really large? I tried running sudo apt purge snapd which is recommended elsewhere when snap gets large, but apparently I can't run that without any space. Running "snap set system refresh.retain=2" in hopes of deleting older snaps times out. Am I at the point where I need to start rm'ing things in /snap in order to get space back?

BrainDance
May 8, 2007

Disco all night long!

I'm just collecting my thoughts on this so sorry if this is vague, I'm planning out what I'm gonna do next.

I got my NAS/a bunch of things server on my local network. It's running Debian testing.

It has a jellyfin server on it, working on a nextcloud server, etc. Gonna eventually have them all behind nginx and I got a domain for it it's gonna be great, but that comes later.
Since I live in China to do most things on the Internet I have to use a VPN. Even the stuff that's not blocked, it's often excruciatingly slow if not through the VPN for reasons that are somewhat mysterious to me. I don't want the jellyfin server to be behind the VPN though because I want to access it from within China easily, and also port forwarding becomes a major hassle when it's through the VPN (I get one port unless I pay them a bunch of extra money, and then it's still through a VPN so it'd be slow as hell from a normal Chinese connection.)

I run the VPN through my router so I can give uncensored usable Internet to all the devices on my network. I have to use some garbage proprietary router app for my vpn because normal vpn protocols don't really work in China, they get blocked right away (it's astrill, the one 90% of people in China use) so I'm limited with what I can actually do here. They do have an "exceptions" tab where I can make exceptions for ports but it just doesn't seem to work like that and I have no idea why. I have port forwarding working and when the VPN is down I can get to my Jellyfin server from anywhere, no problem. I make an exception for port 8096, turn on the VPN, it dies. I make an exception for the whole device, it works.

Most things this computer does I want behind the VPN, like all my torrents and stuff, it's just the Jellyfin server, nextcloud setup, etc that I dont. So if I cant get those port exceptions working here is my plan.

2 ethernet cards. gently caress it, I got a switch in there, one ethernet card goes through the VPN, the other doesnt. The servers all serve through the non-VPN one, everything else just defaults to the VPN one. That's the idea and I think it will work, I cant think of a reason it wouldn't and I cant think of another way to make this work. I know Linux can do this, I mean routers run Linux, of course it can.

Problem is, networking stuff has always terrified me. I don't know where to even start with this. How do I tell certain things to use one of the cards and not the other? Is this a massive task? Where do I begin and what do I do? An arch wiki link to what I gotta do will probably be good enough. Thanks

VictualSquid
Feb 29, 2012

Gently enveloping the target with indiscriminate love.


run "du -h | sort -h" instead, to find out more. And 16Gb is pretty tight for a modern linux, needs lots of juggling for many uses. Your /var also looks pretty large for such a tight system.
Just find snap's and some other cachefiles and rm them to get the space. Or your journals in case those are what is filling up var.

Tesseraction
Apr 5, 2009

BrainDance posted:

Most things this computer does I want behind the VPN, like all my torrents and stuff, it's just the Jellyfin server, nextcloud setup, etc that I dont. So if I cant get those port exceptions working here is my plan.

I think what you're asking about is how to get Jellyfin to work even when the VPN is up. I'm suspecting that you're getting issues related to ephemeral ports not being allowed by the firewall https://en.wikipedia.org/wiki/Ephemeral_port#Range

Keito
Jul 21, 2005

WHAT DO I CHOOSE ?

SadBag posted:

Question about dealing with snap eating all of my disk space
Computer itself is a chromebook, uname -r returns 5.19.0-46-generic, os-release returns Ubuntu 22.04.2 LTS. Ubuntu flavor is Lubuntu, ~16GB of drive space

After my computer unexpectedly shut down today, and my computer to hung whenever I entered password to login, I used alt-f2 to go to terminal, was able to login, and after attempting to update and seeing that I was out of drive space, ran df -h and saw
code:
Filesystem     Size   Used  Avail Use%  Mounted on
tmpfs          178M    17M  161M   10%  /run
/dev/sda2       15G    15G     0  100%  /
tmpfs          888M      0  888M    0%  dev/shm
tmpfs          5.0M   4.0K  5.0M    1%  /run/lock
tmpfs          888M      0  888M    0%  /tmp
/dev/sda1      511M    11M  501M    3%  /boot/efi
tmpfs          178M    72K  178M    1%  /run/user/1000
Then I tried running du -s /*
code:
0        /bin
329444   /boot
0        /dev
10780    /etc
287892   /home
0        /lib
0        /lib32
0        /lib64
0        /libx32
16       /lost+found
16       /media
4        /mnt
4        /opt
du: cannot access '/proc/1559/task/1559/fd/4': No such file or directory
du: cannot access '/proc/1559/task/1559/fdinfo/4': No such file or directory
du: cannot access '/proc/1559/fd/3': No such file or directory
du: cannot access '/proc/1559/fdinfo/3': No such file or directory
0       /proc
11556   /root
19660   /run
0       /sbin
7279650 /snap
4       /srv
524292  /swapfile
0       /sys
0       /tmp
6809168 /usr
6790296 /var
So I guess at some point /snap got really large? I tried running sudo apt purge snapd which is recommended elsewhere when snap gets large, but apparently I can't run that without any space. Running "snap set system refresh.retain=2" in hopes of deleting older snaps times out. Am I at the point where I need to start rm'ing things in /snap in order to get space back?

What about running "snap list" and then "snap remove --purge" components you don't need?

https://manpages.ubuntu.com/manpages/jammy/en/man8/snap.8.html

I'd also recommend using ncdu when trying to figure out where space is being used.

BrainDance
May 8, 2007

Disco all night long!

Tesseraction posted:

I think what you're asking about is how to get Jellyfin to work even when the VPN is up. I'm suspecting that you're getting issues related to ephemeral ports not being allowed by the firewall https://en.wikipedia.org/wiki/Ephemeral_port#Range

Maybe that's what I want, I'm not too sure. But I think that would end up with my Jellyfin server going through the VPN and reaching the Internet in America (or wherever the VPN server is), right?

The VPN client runs on the router. It is not a standard VPN but some proprietary garbage specifically designed to get around the Great Firewall, which limits my options. I'm unable to use my router's (an asus router running openwrt-merlin) vpn client or any other vpn software because of this.

The Jellyfin server is running on the NAS/Everything-Server Debian Testing machine.

I want most of the traffic on that computer to go through the VPN, but not the Jellyfin server. I dont want it to just work with the VPN, but not go through it at all, because I'm going to be using it in China and if it's going through a VPN in America then it's going to be very slow from China.

So, my plan was to just put 2 NICs in the Debian machine. I send one through the VPN in the router, and I exclude one from the VPN. The Jellyfin server goes through the non-vpn NIC, my torrents and stuff go through the VPN NIC. But I don't know how to be like "ok qbittorrent you get the Internet through nic1 jellyfin you're going through nic2, Linux just default to nic1 unless I tell you otherwise" so I need to know where to start to be doing that.

The software included in the garbage proprietary VPN router applet has a feature for excluding certain ports from the VPN, but it seems to just.... not work? So I'm gonna message them about it, but I'm really not optimistic.

Tesseraction
Apr 5, 2009

Does your NAS have multiple ethernet ports? If so you could have a plain router as your default gateway, have that connect to your always-on-VPN-router which then connects to your home network, and have the NAS plug in to both of those routers - say 192.168.1.0/24 on one and 10.0.0.0/24 on the other, have your gateway router connect to the NAS-only IP for external connectivity and bypass the VPN gateway, and use its normal IP for everything on your internal.

SadBag
Jun 24, 2012

Something has gone very wrong for us to get to the point where Hot Dog is the admiral.

VictualSquid posted:

run "du -h | sort -h" instead, to find out more. And 16Gb is pretty tight for a modern linux, needs lots of juggling for many uses. Your /var also looks pretty large for such a tight system.
Just find snap's and some other cachefiles and rm them to get the space. Or your journals in case those are what is filling up var.

Yeah, emptying out journals gave me enough space to get back into gui.

Keito posted:

What about running "snap list" and then "snap remove --purge" components you don't need?

https://manpages.ubuntu.com/manpages/jammy/en/man8/snap.8.html

I'd also recommend using ncdu when trying to figure out where space is being used.

Wasn't able to run anything with snap because everything timed out, but will look into this now

Nitrousoxide
May 30, 2011

do not buy a oneplus phone



SadBag posted:

Yeah, emptying out journals gave me enough space to get back into gui.

Wasn't able to run anything with snap because everything timed out, but will look into this now

Might want to setup a systemd service and timer to do some maintenance pruning once a week or so since you've got a tight storage limit. The normal linux cleanup services probably won't be aggressive enough for you and this will happen again.

pseudorandom name
May 6, 2007

You don’t have to create any units to do that, journald has size and rotation interval options that control its behavior.

Klyith
Aug 3, 2007

GBS Pledge Week

SadBag posted:

Wasn't able to run anything with snap because everything timed out, but will look into this now

On a 16GB system I'd really consider removing snap entirely (and also flatpak if you have it). Snap & flatpak are not conservative over drive space, they're kinda designed presuming a modern environment where that is not a problem.

On a minimal chromebook you probably aren't running a ton of stuff that particularly benefits from snap/flatpak instead of a standard install.


If you do keep snaps installed, you can run this to make snap not keep old versions of a snap around when they update:
sudo snap set system refresh.retain=1
(The default is to retain 3 versions.)

Nitrousoxide
May 30, 2011

do not buy a oneplus phone



pseudorandom name posted:

You don’t have to create any units to do that, journald has size and rotation interval options that control its behavior.

I don't just mean journald. That was certainly part of their storage usage issue, but I'm sure it's not the only thing gradually filling up their limited space.

SadBag
Jun 24, 2012

Something has gone very wrong for us to get to the point where Hot Dog is the admiral.

Klyith posted:

If you do keep snaps installed, you can run this to make snap not keep old versions of a snap around when they update:
sudo snap set system refresh.retain=1
(The default is to retain 3 versions.)

Turns out the retain number must be between 2-20.
Did reduce it to 2.
With that and a couple of other things, currently working with 1.8 Gigs of space. Thanks people.

Nitrousoxide posted:

Might want to setup a systemd service and timer to do some maintenance pruning once a week or so since you've got a tight storage limit. The normal linux cleanup services probably won't be aggressive enough for you and this will happen again.

Will look into, thanks!

bsaber
Jul 27, 2007
I have a question regarding Wireguard but not sure where is the right place to ask. So let me know if there's a better place to ask. This is how I have it setup: client1 <-> wireguard server <-> client2

Where traffic between client1 and client2 are routed through wireguard server. Is client1's traffic that is routed through the wireguard server to client2 encrypted such that wireguard server is unable to see/decrypt? I'm assuming yes since wireguard server is the one that is the "endpoint" for both client1 and client2.

Nitrousoxide
May 30, 2011

do not buy a oneplus phone



The server should have the public/private keys needed to decrypt any connection with any client, yes.

Computer viking
May 30, 2011
Now with less breakage.

Yup, the incoming packets from client1 show up in the clear on the server, which then looks at their destination and routes them to client2, and then they are encrypted again on the way out.

It is in a general sense possible to make a system where the contents are encrypted one level deeper than the routing information, to get something that's like a VPN through untrusted hosts. Connections inside the Tor network work that way, for instance. Or you could just use an encrypted protocol through your wireguard links - ssh or https can be at least as well encrypted as the wireguard links themselves, and those are end-to-end between client1 and client2.

bsaber
Jul 27, 2007
That’s what I thought. Just wanted to be sure I understood properly. Thanks for the clarification.

wash bucket
Feb 21, 2006

I recently upgraded to the latest Nvidia drivers (535.86.05) and now the top half of my screen seems to randomly flicker every 15 seconds or so. Just enough to be irritating.

Any suggestions on what thread to tug on first? My video card is officially supported according to the docs. No such problems when I boot into windows so I don't think it's a hardware issue.

Klyith
Aug 3, 2007

GBS Pledge Week

wash bucket posted:

I recently upgraded to the latest Nvidia drivers (535.86.05) and now the top half of my screen seems to randomly flicker every 15 seconds or so. Just enough to be irritating.

Any suggestions on what thread to tug on first?

Kernel parameter nvidia-drm.modeset=1 is a common fix for nvidia flicker problems. Try it the first time by manually typing it into GRUB so it's a temporary change (apparently on some configs this gives you no screen at all).

OTOH this seems like a common issue with 535 that modeset doesn't fix anymore, so if that doesn't work I'd just roll back to whatever you were using before.

Subjunctive
Sep 12, 2006

✨sparkle and shine✨

Computer viking posted:

It is in a general sense possible to make a system where the contents are encrypted one level deeper than the routing information, to get something that's like a VPN through untrusted hosts.

Tailscale provides this because it builds a full wireguard mesh between nodes.

wash bucket
Feb 21, 2006

Klyith posted:

Kernel parameter nvidia-drm.modeset=1 is a common fix for nvidia flicker problems. Try it the first time by manually typing it into GRUB so it's a temporary change (apparently on some configs this gives you no screen at all).

OTOH this seems like a common issue with 535 that modeset doesn't fix anymore, so if that doesn't work I'd just roll back to whatever you were using before.

Thanks for the info. Good to know it's at least a known issue. Now to decide which will be less painful, putting up with the flicker or trying to roll back my distro's custom nvidia drivers.

VictualSquid
Feb 29, 2012

Gently enveloping the target with indiscriminate love.
Does anybody know how to adjust player choice for the kde media player widget? Or for kde-connect.

I use kde connect. I like to listen to podcasts on my android. Then I pause the podcast, and start my music player on my desktop. And kde connect connects those.
Now the media control widget on my desktop controls the (paused) podcast app on my phone, and the media control widget on my phone controls mpd on my desktop. Anybody know a way to switch default controlled apps?

VictualSquid fucked around with this message at 13:43 on Aug 5, 2023

Klyith
Aug 3, 2007

GBS Pledge Week

VictualSquid posted:

Does anybody know how to adjust player choice for the kde media player widget? Or for kde-connect.

I use kde connect. I like to listen to podcasts on my android. Then I pause the podcast, and start my music player on my desktop. And kde connect connects those.
Now the media control widget on my desktop controls the (paused) podcast app on my phone, and the media control widget on my phone controls mpd on my desktop. Anybody know a way to switch controlled apps?

Do you ever want your phone to be a media remote control for your desktop, or vice versa? Because the easiest way to fix this would be to turn off the relevant plugins on both sides.

On the phone:
• Media Player Control
• Multimedia Controls
On the desktop:
• Multimedia Control Receiver
(Maybe also MprisRemote? I had this turned on and my phone wasn't butting in to my media playback. But Mpris is KDE's general multimedia control service.)

If you wanted controls in only one direction, I feel like you could probably still make that work? The plugins each seem to control one direction of sending & receiving. But I dunno, I just turned off all three.



Aside from that, the KDE desktop media player widget seems to pick actively playing -> first to start playing -> most recently playing for the priority of who gets the star. Generally seems sensible, I hardly ever feel like it's made the "wrong" choice when I press the play/pause key on my keyboard.

But I could easily see KDE Connect messing that up the first / most recent order.

bsaber
Jul 27, 2007
I was transferring some large files over wireless from my laptop running Arch Linux to desktop running Windows 10 via SMB. Twice now during the transfers the laptop screen would go black and system would completely lock up and require a forced power off. I did a search of journalctl (used: journalctl -p err -b) and the only thing that looks related is this:

code:
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: Microcode SW error detected.  Restarting 0x82000000.
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: Start IWL Error Log Dump:
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: Transport status: 0x0000004A, valid: 6
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: Loaded firmware version: 36.ca7b901d.0 8265-36.ucode
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x00001043 | ADVANCED_SYSASSERT          
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x059002A0 | trm_hw_status0
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x00000000 | trm_hw_status1
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x00024A8C | branchlink2
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x0003AC1E | interruptlink1
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x00000000 | interruptlink2
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x00002600 | data1
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x00002C48 | data2
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0xDEADBEEF | data3
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0xE180468B | beacon time
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x278FCB43 | tsf low
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x0000068C | tsf hi
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x00004625 | time gp1
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x1EB890B4 | time gp2
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x00000001 | uCode revision type
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x00000024 | uCode version major
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0xCA7B901D | uCode version minor
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x00000230 | hw version
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x00489000 | board version
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x0B4A001C | hcmd
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0xA7FA3802 | isr0
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x11C50000 | isr1
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x0800191A | isr2
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x4041B9C7 | isr3
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x00000000 | isr4
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x0B36001C | last cmd Id
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x00000000 | wait_event
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x000000C4 | l2p_control
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x00018020 | l2p_duration
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x00000007 | l2p_mhvalid
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x00000081 | l2p_addr_match
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x0000000D | lmpm_pmg_sel
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x10032209 | timestamp
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x00001018 | flow_handler
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: Start IWL Error Log Dump:
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: Transport status: 0x0000004A, valid: 7
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x00000070 | NMI_INTERRUPT_LMAC_FATAL
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x00000000 | umac branchlink1
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0xC008694C | umac branchlink2
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0xC0083B0C | umac interruptlink1
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0xC0083B0C | umac interruptlink2
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x00000800 | umac data1
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0xC0083B0C | umac data2
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0xDEADBEEF | umac data3
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x00000024 | umac major
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0xCA7B901D | umac minor
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0xC088628C | frame pointer
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0xC088628C | stack pointer
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x0052014E | last host cmd
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x00000000 | isr status reg
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: IML/ROM dump:
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x000028A7 | IML/ROM error/state
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x00000003 | IML/ROM data1
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: Fseq Registers:
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x08D1D060 | FSEQ_ERROR_CODE
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x51810B4A | FSEQ_TOP_INIT_VERSION
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x63E942AE | FSEQ_CNVIO_INIT_VERSION
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x0000A10B | FSEQ_OTP_VERSION
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x3BA64560 | FSEQ_TOP_CONTENT_VERSION
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x24E76973 | FSEQ_ALIVE_TOKEN
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0xDDE88F45 | FSEQ_CNVI_ID
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x9B4F00E1 | FSEQ_CNVR_ID
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x00000010 | CNVI_AUX_MISC_CHIP
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x0BADCAFE | CNVR_AUX_MISC_CHIP
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x0BADCAFE | CNVR_SCU_SD_REGS_SD_REG_DIG_DCDC_VTRIM
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: 0x0BADCAFE | CNVR_SCU_SD_REGS_SD_REG_ACTIVE_VDIG_MIRROR
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: Collecting data: trigger 2 fired.
Aug 06 23:38:08 hostname kernel: ieee80211 phy0: Hardware restart was requested
Aug 06 23:38:08 hostname kernel: iwlwifi 0000:02:00.0: Failing on timeout while stopping DMA channel 8 [0x07fd0001]
Doing some quick Googling I found a few forum posts that show it is related to a bug in firmware or something to do with power saving. The firmware bug looks to have been fixed years ago. Checked power saving by running this: cat /sys/module/iwlwifi/parameters/power_save which returns N. Which I believe means power saving is disabled?

As for this line: Microcode SW error detected. Restarting 0x82000000. I can't find anything that has that exact code with the Googling I did.

Before continuing, would someone with better knowledge take a look and make sure I'm not going down the wrong road? Is the above showing a problem that I need to worry about or is it unrelated to my issue?

Klyith
Aug 3, 2007

GBS Pledge Week

bsaber posted:

Twice now during the transfers the laptop screen would go black and system would completely lock up and require a forced power off.
...
Before continuing, would someone with better knowledge take a look and make sure I'm not going down the wrong road? Is the above showing a problem that I need to worry about or is it unrelated to my issue?

The wifi adapter crashing shouldn't necessarily take down the whole system, but it may be down to your laptop's particular hardware. It seems like the most relevant thread to pull on.


bsaber posted:

As for this line: Microcode SW error detected. Restarting 0x82000000. I can't find anything that has that exact code with the Googling I did.

googling iwlwifi + "Microcode SW error" was much more helpful. You get stuff like this which says it's power management problems in the intel microcode.
https://bbs.archlinux.org/viewtopic.php?id=254766

This link:
https://askubuntu.com/questions/1283313/unstable-wifi-connection-on-ubuntu-20-04
has the most clearly written instructions for setting the options to fix it.


Aside from that you can use fwupd to see if there's better firmware for the wifi card.

AlexDeGruven
Jun 29, 2007

Watch me pull my dongle out of this tiny box


What model laptop is this? There's one from several years ago that used a notoriously bad WiFi chipset. Caused me a lot of pain dealing with my wife's IT department trying to shove it off on my network when literally every other device on the WiFi worked just fine.

bsaber
Jul 27, 2007

Klyith posted:

This link:
https://askubuntu.com/questions/1283313/unstable-wifi-connection-on-ubuntu-20-04
has the most clearly written instructions for setting the options to fix it.

Thanks for the help. This link seems to have fixed the issue. Specifically adding this: options iwlwifi power_save=0. So looks to be power saving related.

AlexDeGruven posted:

What model laptop is this? There's one from several years ago that used a notoriously bad WiFi chipset. Caused me a lot of pain dealing with my wife's IT department trying to shove it off on my network when literally every other device on the WiFi worked just fine.

It's a Thinkpad X1 Carbon 6th Gen. The wireless card is an Intel AC 8265 according to lspci.

waffle iron
Jan 16, 2004
Honestly if it's not too difficult to take the laptop apart, I would look into buying a better supported m.2 wifi card.

bsaber
Jul 27, 2007
It's a Thinkpad so should be really easy to open and replace. Is there a recommended card?

Bozza
Mar 5, 2004

"I'm a really useful engine!"
Potentially an absolutely stupid question but looking for some advice wrt firewalls

I've got a relatively fresh arch install which I've been dicking about with but working through the tips on the wiki I've neglected to install a firewall. tbh most of the time I'm in my house so sat behind my router firewall and therefore not really that bothered. however I do occasionally use my laptop when I go away for work (for personal use, have a work laptop so there's nothing "sensitive" on it as such) so thought I'd beef it up a bit.

In classic arch-wiki style, the nftables / iptables pages are absolutely impenetrable to a newbie and beyond just bumbling my way through a badly configured copy/paste install I would just like to know the following questions:

- is it actually worth installing a linux firewall in 2023? there were some posts on reddit/stackoverflow/assorted blogs saying that they are fundamentally security theatre if you have disabled/not installed anything that can connect externally
- same question again but I have wireguard configured to connect back to my home router via VPN so theoretically same level of protection I have sat on my sofa. just leave it in VPN and all hunky dory?
- if the to the above is still yes, it is worth installing: is there a total fuckwits guide to setting up either iptables/nftables, or which one is better, and how to understand what ports etc I want open/closed?

Arch wiki recommends the following for nftables but I can vaguely follow it. Is this all I need really and good to go? I mostly just use my laptop for watching YouTube and chatting on Discord so don't have massive need for loads of weird ports open but would be useful to see if I try and run something in the future and it bumps off it where I can find out who/what/why it did that and how to add info.

code:
/etc/nftables.conf

flush ruleset

table inet my_table {
	set LANv4 {
		type ipv4_addr
		flags interval

		elements = { 10.0.0.0/8, 172.16.0.0/12, 192.168.0.0/16, 169.254.0.0/16 }
	}
	set LANv6 {
		type ipv6_addr
		flags interval

		elements = { fd00::/8, fe80::/10 }
	}

	chain my_input_lan {
		udp sport 1900 udp dport >= 1024 meta pkttype unicast limit rate 4/second burst 20 packets accept comment "Accept UPnP IGD port mapping reply"

		udp sport netbios-ns udp dport >= 1024 meta pkttype unicast accept comment "Accept Samba Workgroup browsing replies"

	}

	chain my_input {
		type filter hook input priority filter; policy drop;

		iif lo accept comment "Accept any localhost traffic"
		ct state invalid drop comment "Drop invalid connections"
		ct state established,related accept comment "Accept traffic originated from us"

		meta l4proto ipv6-icmp accept comment "Accept ICMPv6"
		meta l4proto icmp accept comment "Accept ICMP"
		ip protocol igmp accept comment "Accept IGMP"

		udp dport mdns ip6 daddr ff02::fb accept comment "Accept mDNS"
		udp dport mdns ip daddr 224.0.0.251 accept comment "Accept mDNS"

		ip6 saddr @LANv6 jump my_input_lan comment "Connections from private IP address ranges"
		ip saddr @LANv4 jump my_input_lan comment "Connections from private IP address ranges"

		counter comment "Count any other traffic"
	}

	chain my_forward {
		type filter hook forward priority filter; policy drop;
		# Drop everything forwarded to us. We do not forward. That is routers job.
	}

	chain my_output {
		type filter hook output priority filter; policy accept;
		# Accept every outbound connection
	}

}
thanks goons

Adbot
ADBOT LOVES YOU

unruly
May 12, 2002

YES!!!

Bozza posted:

Potentially an absolutely stupid question but looking for some advice wrt firewalls

I've got a relatively fresh arch install which I've been dicking about with but working through the tips on the wiki I've neglected to install a firewall. tbh most of the time I'm in my house so sat behind my router firewall and therefore not really that bothered. however I do occasionally use my laptop when I go away for work (for personal use, have a work laptop so there's nothing "sensitive" on it as such) so thought I'd beef it up a bit.

In classic arch-wiki style, the nftables / iptables pages are absolutely impenetrable to a newbie and beyond just bumbling my way through a badly configured copy/paste install I would just like to know the following questions:

- is it actually worth installing a linux firewall in 2023? there were some posts on reddit/stackoverflow/assorted blogs saying that they are fundamentally security theatre if you have disabled/not installed anything that can connect externally
- same question again but I have wireguard configured to connect back to my home router via VPN so theoretically same level of protection I have sat on my sofa. just leave it in VPN and all hunky dory?
- if the to the above is still yes, it is worth installing: is there a total fuckwits guide to setting up either iptables/nftables, or which one is better, and how to understand what ports etc I want open/closed?

Arch wiki recommends the following for nftables but I can vaguely follow it. Is this all I need really and good to go? I mostly just use my laptop for watching YouTube and chatting on Discord so don't have massive need for loads of weird ports open but would be useful to see if I try and run something in the future and it bumps off it where I can find out who/what/why it did that and how to add info.

[snip]

thanks goons
Install ufw or firewalld. loving around with nftables or iptables is for greybeards. On a machine that travels, it's not worth the headache.

  • 1
  • 2
  • 3
  • 4
  • 5
  • Post
  • Reply