|
I've got a problem on my System76 Pop OS laptop. My wifi module is not finding any networks and intermittently the OS says that the card doesn't exist. The PCI device is installed, the driver is installed, but ifconfig/ip/networkctl say that the device is down and when I try to bring it up I get a timeout error. Not sure what to do here, I've googled quite a bit and haven't found anything that works. Has anyone seen this? One weird thing is before I log in on the OS I can see the networks, but then it disappears a few seconds after I log in. E: I forgot to mention that this has also caused my Bluetooth to stop working, but it isn't saying I don't have bluetooth. 22 Eargesplitten fucked around with this message at 22:25 on Aug 27, 2022 |
#
?
Aug 27, 2022 22:19
|
|
|
|
| # ? May 28, 2024 10:30 |
|
|
Nitrousoxide posted:That's a bummer, I hope you caught it early and it's treatable. Got here from the lepers colony but I agree with this post, good luck to you
|
|
#
?
Aug 28, 2022 00:36
|
|
|
I for one appreciate BSDs unique form of posting, best of luck friend  Also gently caress off whoever Aware is
|
|
#
?
Aug 28, 2022 01:25
|
|
|
Mr. Crow posted:I for one appreciate BSDs unique form of posting, best of luck friend BSD is the funnest troll in SH/SC, and I love seeing them get the linux thread riled up like they were
|
|
#
?
Aug 28, 2022 04:44
|
|
|
Who ITT hasn’t learned more than they care to know about BSD thanks to them 
|
|
#
?
Aug 28, 2022 08:07
|
|
|
Me: Hmm, I don't see the power profiles features in KDE, but they should have been available and default since Fedora 34... odd. Let's check the terminal:quote:$ sudo powerprofilesctl Since it's not urgent, I'd like to figure this out on my own... Where should I start looking, in general, if a package claims to be installed but a file or daemon it should provide is missing?
|
|
#
?
Aug 28, 2022 09:38
|
|
|
I would try to find that package on the web version of the package repository and see if it has any further descriptions, links, or other information about it. E: it might also be worth seeing if your package manager offers a "repair" function or something similar.
|
|
#
?
Aug 28, 2022 11:36
|
|
|
NihilCredo posted:Me: Hmm, I don't see the power profiles features in KDE, but they should have been available and default since Fedora 34... odd. Let's check the terminal: It looks like Fedora has possible conflicts with other power services and they mask power-profiles? https://gitlab.freedesktop.org/hadess/power-profiles-daemon#conflicts https://bugzilla.redhat.com/show_bug.cgi?id=2028701#c17
|
|
#
?
Aug 28, 2022 13:42
|
|
|
RFC2324 posted:I really wish my buttons worked here I know, right?
|
|
#
?
Aug 29, 2022 08:57
|
|
|
22 Eargesplitten posted:I've got a problem on my System76 Pop OS laptop. My wifi module is not finding any networks and intermittently the OS says that the card doesn't exist. The PCI device is installed, the driver is installed, but ifconfig/ip/networkctl say that the device is down and when I try to bring it up I get a timeout error. Not sure what to do here, I've googled quite a bit and haven't found anything that works. Has anyone seen this? I'm not familiar with Pop OS, but I've had wifi chips be a pain on Linux in the past. I'd suggest pinning down exactly what chipset your wifi uses, and then double-check the drivers that are installed. The system might have installed an incorrect package for your specific card - I think I remember Broadcom doing this more than any others, but I switched over to Intel for my wifi and Bluetooth a couple years ago and haven't had problems since. What does lspci bring up for your wireless card, and how does that compare to what drivers are installed?
|
|
#
?
Aug 29, 2022 09:02
|
|
|
Klyith posted:It looks like Fedora has possible conflicts with other power services and they mask power-profiles? I'll try when I get back home, but the symptoms don't fit. If the service were simply masked, the error should be "Failed to start butts.service: Unit butts.service is masked", and `powerprofilesctl` would still exist. e: I might go down a different rabbit hole and try installing System76-power instead, it has a RPM version and people speak very well of it. NihilCredo fucked around with this message at 10:29 on Aug 29, 2022 |
|
#
?
Aug 29, 2022 10:23
|
|
|
CaptainSarcastic posted:I'm not familiar with Pop OS, but I've had wifi chips be a pain on Linux in the past. I'd suggest pinning down exactly what chipset your wifi uses, and then double-check the drivers that are installed. The system might have installed an incorrect package for your specific card - I think I remember Broadcom doing this more than any others, but I switched over to Intel for my wifi and Bluetooth a couple years ago and haven't had problems since. lspci brings up "00:14.3 Network controller: Intel Corporation Comet Lake PCH-LP CNVi WiFi" and lsmod | grep iwlwifi brings up: "iwlwifi 495616 1 iwlmvm iwlmei 53248 2 iwlmvm,iwlwifi cfg80211 1003520 4 iwlmvm,iwlmei,iwlwifi,mac80211" My understanding is that iwlwifi is the Intel wifi driver from doing some google searching. I don't think any updates installed between it working and not working, I was using the laptop, left the house, and when I came back it wasn't working.
|
|
#
?
Aug 29, 2022 13:12
|
|
|
22 Eargesplitten posted:lspci brings up "00:14.3 Network controller: Intel Corporation Comet Lake PCH-LP CNVi WiFi" and lsmod | grep iwlwifi brings up: Do system76 have any support or user forums? Since it's the shipped installation (right?) on official hardware, it should either be something that's already been noticed and maybe there's a guide, or it's a hardware problem in which case it's a warranty issue. Edit: also found this bug (https://bugs.launchpad.net/ubuntu/+source/linux-firmware/+bug/1867026) via Reddit: https://www.reddit.com/r/System76/comments/fj698z/solved_darter_pro_no_wifi_after_update_ac9560/ Same hwid (00:14.3) so it's worth checking if you're on the same kernel version, or at least not on the one they say fixed it. Phosphine fucked around with this message at 13:24 on Aug 29, 2022 |
|
#
?
Aug 29, 2022 13:19
|
|
|
I have a LinuxCNC machine which has its wifi stop working in a similar intermittent way unless I boot it with iommu=soft on the kernel command line. The same hardware configuration worked fine in Windows so I assume something in Linux doesn't agree with it. However, LinuxCNC is based on an older version of Debian so I'm not sure if it was fixed in later kernels. Maybe worth a try just for troubleshooting purposes, although I expect that it could harm performance in games or other really intensive I/O heavy programs.
|
|
#
?
Aug 29, 2022 13:25
|
|
|
So I found out what was wrong with my Samba, and the part where I thought the top-level share was the only thing impacted was wrong. After I wiped & remade all my samba configs, I also wasn't able to create usershares via the dophin GUI. But those error messages got me the google results I needed of other people talking about what's happening. Turns out I was right about this:Klyith posted:So I'm writing this off to bleeding edge distro problems. The root problem is that AppArmor needs updates for Samba 4.16, but AppArmor is a slower update cycle so Arch is just semi-broken for the moment unless you fix it manually. So I learned about AppArmor and the basics of writing apparmor configs. And also a new thing to look for when something mysteriously doesn't work -- journalctl | grep "DENIED" is everything blocked by apparmor.
|
|
#
?
Aug 29, 2022 17:42
|
|
|
Klyith posted:So I found out what was wrong with my Samba, and the part where I thought the top-level share was the only thing impacted was wrong. After I wiped & remade all my samba configs, I also wasn't able to create usershares via the dophin GUI. But those error messages got me the google results I needed of other people talking about what's happening. Turns out I was right about this: BlankSystemDaemon fucked around with this message at 19:28 on Aug 29, 2022 |
|
#
?
Aug 29, 2022 19:26
|
|
|
22 Eargesplitten posted:lspci brings up "00:14.3 Network controller: Intel Corporation Comet Lake PCH-LP CNVi WiFi" and lsmod | grep iwlwifi brings up: Huh, it looks like Intel is doing something I don't completely understand there. https://www.intel.com/content/www/us/en/support/articles/000026155/wireless.html
|
|
#
?
Aug 29, 2022 19:53
|
|
|
BlankSystemDaemon posted:I assume you follow what Chris writes but just in case you aren't No, I'm new enough to linux that I don't have much of a reading bookmark list other than Phoronix and other super-obvious stuff. Which I guess is a thing I should ask for: favorite linux / unix blogs and writers? Especially anyone who writes well / is entertaining, since I'm still learning for fun and not as a sysadmin job. BlankSystemDaemon posted:he has a very recent article about AppArmor Persistence on Ubuntu which may or may not be relevant to Arch, since it seems to relate to the third-party software repos? Either way, it might be worth checking up on if you got spare cycles, to avoid a potential headache in the future. I actually got it to work without disabling AppArmor -- there are systemd-type dropins to add local exceptions & rules. /etc/apparmor.d/local/samba-dcerpcd posted:/var/cache/samba/{,**} wrc /etc/apparmor.d/local/usr.sbin.smbd posted:/home/afolderwithusershares/** lrk, I don't know how often apparmor will cause me difficulty in the future, but I'd rather not turn off a potentially useful security feature unless I really have to.
|
|
#
?
Aug 29, 2022 20:11
|
|
|
Klyith posted:No, I'm new enough to linux that I don't have much of a reading bookmark list other than Phoronix and other super-obvious stuff. Chris' blog is one of the big exceptions that I'll recommend to just about anyone working with the care and feeding of computers, because while he uses Linux now he was a longtime Solaris and later Illumos-derived sysadmin, and knows his way around a fair few Unix-likes, their histories, and will often go into quite a bit of detail. In addition, he also writes a lot about ZFS internals. Another one that's fairly cross-sectional is Dan Langilles blog, and the most obvious one is probably Jim Salter over at Ars.
|
|
#
?
Aug 29, 2022 21:34
|
|
|
BlankSystemDaemon posted:I assume you follow what Chris writes but just in case you aren't, he has a very recent article about AppArmor Persistence on Ubuntu which may or may not be relevant to Arch, since it seems to relate to the third-party software repos? Either way, it might be worth checking up on if you got spare cycles, to avoid a potential headache in the future. Just wanted to say thanks for this; like many other things I lost with the shutdown of google reader, Chris’s blog is an essential resource(IMHO).
|
|
#
?
Aug 30, 2022 00:33
|
|
|
NihilCredo posted:Me: Hmm, I don't see the power profiles features in KDE, but they should have been available and default since Fedora 34... odd. Let's check the terminal: Try running an rpm -ql on the package - it should at least allow you to see where it's meant to be installed and if the files have gone missing.
|
|
#
?
Aug 30, 2022 16:42
|
|
|
Any advice on how to track down mystery kernel panics? I have two ryzen systems, a 2700 and a 3600, both doing lots of VM stuff (qemu, xen at different times). The 3600 has no problems with any kernel, but the 2700 has started mystery reboots whenever the host or dom0 (for xen) is running newer kernels. I think some of the earlier 5.x kernels were okay... ish... though they would still crash very occasionally, but newer stuff 5.11+ crashes like crazy. Memtest still runs just fine and the system is absolutely rock solid on a 4.x kernel. The machine freezes for a few seconds and then reboots into the BIOS startup. I tried disabling reboot on panic but it still resets, and the /sys/fs/pstore stuff doesn't seem to work even though kernel config says it's enabled. Pretty sure modern PC motherboards don't have a JTAG connector to hook up to so I'm at a loss. Just stay on a 4. kernel literally forever? My only other thought would be to try to add a serial console to the kernel cmdline and monitor it from another computer to see if it spits anything out when it happens, but I don't think that motherboard has a real serial port, unless I add a PCIe serial port card maybe.
|
|
#
?
Sep 6, 2022 01:02
|
|
|
You could also try netconsole, but if it still rebooting even with reboot on panic turned off then you're probably out of luck.
|
|
#
?
Sep 6, 2022 01:09
|
|
|
Rescue Toaster posted:Any advice on how to track down mystery kernel panics? I have two ryzen systems, a 2700 and a 3600, both doing lots of VM stuff (qemu, xen at different times). The 3600 has no problems with any kernel, but the 2700 has started mystery reboots whenever the host or dom0 (for xen) is running newer kernels. I think some of the earlier 5.x kernels were okay... ish... though they would still crash very occasionally, but newer stuff 5.11+ crashes like crazy. Memtest still runs just fine and the system is absolutely rock solid on a 4.x kernel. Is the BIOS appropriately updated on the problem machine? Microcode up to date?
|
|
#
?
Sep 6, 2022 01:17
|
|
|
CaptainSarcastic posted:Is the BIOS appropriately updated on the problem machine? Microcode up to date? It looked like the newer BIOS updates were only for adding newer Ryzen support, there was no mention of anything for the older 2000 series. But there is newer AGESA obviously so maybe I'll give it a shot. Who knows, maybe that'll fix the pstore not working since it's tied in to the UEFI variables. Just to be sure, if /etc/sysctl.conf has the panic.reboot = 0, and then /proc/sys/kernel/panic is 0, that should do it as far as auto-reboot?
|
|
#
?
Sep 6, 2022 01:29
|
|
|
Rescue Toaster posted:The 3600 has no problems with any kernel, but the 2700 has started mystery reboots whenever the host or dom0 (for xen) is running newer kernels. Complete speculation warning: one thing that's different between Zen+ (the 2700) and Zen 2 & higher is hardware support of Mode-Based Execution Control, which is part of the virtualization feature set. I can't imagine that a default kernel would try to use MBEC on a CPU that doesn't have it. The kernel should know the capabilities of a CPU it's running on. But if you've doing something like imaged the same set-up between both PCs complete with custom kernel flags, that might do it? Or especially if you're compiling your own kernel on the 3600 and then copying to the 2700. Anyways if you're doing the same thing on both CPUs, that's the one big thing I know is not the same between the two.
|
|
#
?
Sep 6, 2022 01:35
|
|
|
Klyith posted:Complete speculation warning: one thing that's different between Zen+ (the 2700) and Zen 2 & higher is hardware support of Mode-Based Execution Control, which is part of the virtualization feature set. Interesting. I haven't messed with anything so I'm assuming the kernel is detecting it right. But it does seem to specifically be related to VM launching/stopping, so that's a clue for a difference anyway.
|
|
#
?
Sep 6, 2022 01:46
|
|
|
Rescue Toaster posted:It looked like the newer BIOS updates were only for adding newer Ryzen support, there was no mention of anything for the older 2000 series. But there is newer AGESA obviously so maybe I'll give it a shot. Who knows, maybe that'll fix the pstore not working since it's tied in to the UEFI variables. It's just something I'd take a look at - you might not need the newest BIOS (or it could be problematic) but checking it seems worth a shot. What GPU are you running? Usually when I've seen a mismatch between kernel and Nvidia drivers it results in not booting to a graphical desktop, but that might be worth looking at, too.
|
|
#
?
Sep 6, 2022 01:56
|
|
|
I'm getting an unstable virtualized install of Windows 10. Its so bad the windows installer is bluescreening. Sometimes I can get it to install but marginal. I do have a FreeBSD and Linux install working as guests on this platform, they're working fine, even under load. Its just windows that is upset. This is a dedicated server without IMPI, I sent in a support ticket to flash the latest bios. What else should I be looking at? Here is what I have done to diagnose this:
Ryzen 5 3600, 64GB ram. Ubuntu 20.04.5 LTS libvirt version: 6.0.0, package: 0ubuntu8.16 qemu version: 4.2.1Debian 1:4.2-3ubuntu6.23 kernel: 5.4.0-125-generic pre:virt-install --name=win10-vm \ --cpu host \ --vcpus=6,sockets=1,cores=6 \ --memory=16384 \ --network bridge=br1,model=virtio \ --cdrom=/tmp/Win10_21H2_English_x64.iso \ --disk=/tmp/virtio-win.iso,device=cdrom \ --graphics vnc,listen=127.0.0.1 --noautoconsole \ --disk size=64,bus=virtio,format=raw,cache=none,io=native \ --os-variant=win10
|
|
#
?
Sep 6, 2022 01:57
|
|
|
Rescue Toaster posted:Any advice on how to track down mystery kernel panics? I have two ryzen systems, a 2700 and a 3600, both doing lots of VM stuff (qemu, xen at different times). The 3600 has no problems with any kernel, but the 2700 has started mystery reboots whenever the host or dom0 (for xen) is running newer kernels. I think some of the earlier 5.x kernels were okay... ish... though they would still crash very occasionally, but newer stuff 5.11+ crashes like crazy. Memtest still runs just fine and the system is absolutely rock solid on a 4.x kernel. Try kdump.
|
|
#
?
Sep 6, 2022 07:10
|
|
Associate Christ
|
I'm looking into remotely unlocking a Fedora home server's LUKS drives after a hard reboot. For Debian/Ubuntu, it seems very straightforward: the `dropbear-initramfs` package will add a minimal SSH server to your initialization, so you can connect and decrypt the boot partition. For RHEL/Fedora, the package does not exist. Some people have managed to manually install dropbear into initramfs using `dracut`, but we're talking "some random comment on serverfault" level of unreliability. What is supported in Red Hat land is using the Network Bound Disk Encryption feature to run the "Clevis and Tang" combo. Tang is basically an auth server that must run on a separate machine, and Clevis is a LUKS management tool that lets you bind a LUKS drive to a Tang server (in addition to your regular password). So you reboot, Clevis calls the Tang server, and the Tang server provides the unlock key. https://www.golinuxcloud.com/network-bound-disk-encryption-tang-clevis/ (Clevis can also use a TPM 2.0 chip to unlock the drive, but the documentation warns that after certain package upgrades the unlock might not work any more, requiring you to temporarily fall back to the passphrase. Since I'm going to be in a different country, I can't solely rely on this.) I actually have a Raspberry Pi in the house so I could go the Clevis and Tang route, but it's not a great fit for my needs. C&T is obviously way more robust and scalable if you have a fleet of servers to manage, but I have just one that should only reboot either manually or if the power temporarily goes out in my home. #1 issue: If my Pi's SD card gives up the ghost I'm SOL (passphrase still works but I need to be physically there) #2 issue: LUKS protects my data in case a burglar grabs my PC. But if they grab my Pi in addition to the server, they can potentially bypass the encryption since Tang starts automatically on boot. It's very unlikely because they'd need to boot up both devices while having them connected to the same network, but it still rubs my security instincts the wrong way. Dropbear would actually be perfect for me, but I'm not gonna switch to Debian just for that :P I guess I don't really have a specific question, I just wanted to know if someone has faced a similar issue and if there are other options I can look into.
|
|
#
?
Sep 6, 2022 11:05
|
|
|
Make a PiKVM.
|
|
#
?
Sep 6, 2022 11:33
|
|
|
NihilCredo posted:I'm looking into remotely unlocking a Fedora home server's LUKS drives after a hard reboot. I'm using vaultlocker for this. Authentication done using the AppRole method with SecretID credential, with CIDR based ACL so that it won't work if the host is moved. I don't host the Vault instance in my home so there's no risk of it being stolen, but since it can't be manipulated without authentication and boots in a locked state, it wouldn't really be a (security) problem. This would be even more overkill for a single host, but the Tang based solution doesn't sound very appealing if it doesn't make any effort to protect the keys.
|
|
#
?
Sep 6, 2022 11:41
|
|
|
It seems like someone made an easy access version of the Dracut system for this kind of situation https://github.com/dracut-crypt-ssh/dracut-crypt-ssh It seems to be recommended by this hosting company for exactly your scenario (step 6) https://www.vultr.com/docs/install-and-setup-centos-7-to-remotely-unlock-lvm-on-luks-disk-encryption-using-ssh/ Not to say necessarily trust it but it looks more legitimate than just a random comment on serverfault.
|
|
#
?
Sep 6, 2022 11:46
|
|
|
Pablo Bluth posted:Make a PiKVM. Interesting. Seems like a bit of effort and/or money for something that doesn't solve #1. However, the ability to handle BIOS issues and control ATX power is neat. Keito posted:I'm using vaultlocker for this. Authentication done using the AppRole method with SecretID credential, with CIDR based ACL so that it won't work if the host is moved. I don't host the Vault instance in my home so there's no risk of it being stolen, but since it can't be manipulated without authentication and boots in a locked state, it wouldn't really be a (security) problem. Very cool. Vault is a tool I wanted to learn for professional use anyway, and I have my own domain so I could easily make a backup copy of the locked-down Vault database and run it at any publicly-accessible machine I control (the aforementioned Pi, a cheap cloud VM, even my laptop - just update the domain). The one issue I see is that, if I'm reading the README correctly, this does not allow for traditional LUKS password decryption alongside vault unlocking, does it? So if I screw something up with the configuration, or I hit a bug, I cannot recover my data by physically typing in an emergency decryption key. Related to the above, is it still maintained by the OpenStack guys? Both the documentation and the source code links are 404ing. e: Tesseraction posted:It seems like someone made an easy access version of the Dracut system for this kind of situation https://github.com/dracut-crypt-ssh/dracut-crypt-ssh Ok, I still need to read the whole README but this should be exactly what I was initially looking for. Thank you! e: I also found this repo linked from the issues in the one you linked. It's slightly more recent, but the most interesting thing for me is that it's been tested with Silverblue. e2: that one's readme has a pretty good overview of the available options, too 
NihilCredo fucked around with this message at 16:04 on Sep 6, 2022 |
|
#
?
Sep 6, 2022 12:10
|
|
|
other people posted:Try kdump. I think I've heard of Netflix using it on their Linux based front-end in AWS, but I'm not having much finding definitive proof it exists. EDIT: As soon as I wrote that, I noticed it's in the SEE ALSO section of the crash(8) section, but somehow doesn't have a working manual page of its own, or isn't generated properly? Oh well, at least it definitely exists - because it's useful as heck. BlankSystemDaemon fucked around with this message at 13:12 on Sep 6, 2022 |
|
#
?
Sep 6, 2022 13:09
|
|
|
NihilCredo posted:Very cool. Vault is a tool I wanted to learn for professional use anyway, and I have my own domain so I could easily make a backup copy of the locked-down Vault database and run it at any publicly-accessible machine I control (the aforementioned Pi, a cheap cloud VM, even my laptop - just update the domain). vaultlocker just generates a random key which is passed to LUKS when formatting the device, so you could use this key to manually unlock, or add yourself a different passphrase in another key slot. It's not a tool for FDE though. I don't think this was an official OpenStack project in the first place, but not sure why their links 404. Maybe I never checked them. The maintainer indeed doesn't seem very active, but it's packaged in Ubuntu and the codebase is simple so I figure I'd just fork it if ever needed. Been using it for two years and haven't had an issue yet.
|
|
#
?
Sep 6, 2022 14:47
|
|
|
I have some folks that are killing docker commands that see some strace output about futexes hanging around. It looks like there was a bunch of stuff waiting on futexes with an indefinite timeout. Their attitude was that this is a kernel bug and it has to clean these up. I'm not convinced. You see stuff like this in all kinds of google searches:code:
|
|
#
?
Sep 6, 2022 19:45
|
|
|
It is the application's problem unless it is explicitly using robust mutexes. edit: It is still the application's problem if you're using robust mutexes, but the kernel will at least tell you there is a problem. pseudorandom name fucked around with this message at 22:30 on Sep 6, 2022 |
|
#
?
Sep 6, 2022 21:51
|
|
|
|
| # ? May 28, 2024 10:30 |
|
|
What am I missing there, FUTEX_WAKE doesn't block, and the fact that strace of the call has a return value means it's complete. And if you're talking about _PRIVATE stuff, once the process is gone, all its private futexes are too? These are in the docker cli?
|
|
#
?
Sep 6, 2022 23:11
|
|