|
I agree with immersion learning. But also don't be scared of a Linux reinstall. it takes like 15m at most.
|
#
?
Mar 9, 2024 17:55
|
|
|
|
| # ? May 14, 2024 07:52 |
|
|
Reinstalling is also a lot less scary if you keep /home on its own partition or disk. This way, you can blast away the OS without losing files you care about
|
|
#
?
Mar 9, 2024 18:03
|
|
|
spiritual bypass posted:Reinstalling is also a lot less scary if you keep /home on its own partition or disk. This way, you can blast away the OS without losing files you care about yeah definitely!!
|
|
#
?
Mar 9, 2024 19:12
|
|
|
buglord posted:So I unplugged the HD Audio cable from the motherboard header and I got an interesting result. Like Voodoo Cafe said, not only does my computer boot way faster, but the errors are mostly gone? So I was curious enough to see what happened if i unplugged all my motherboard header cables, then plugged them back in (because I would like front panel audio if I need it). I'm left with the 1 weird momentary error, but still have the benefit of *much faster* boot speed. So maybe I had something slightly plugged in the wrong way for some reason? I have no clue, but im considering this a victory because I'm much better than when I started off, before the problem got worse. i can live with this tiny wrinkle in an otherwise stable and gaming-worthy install of Kubuntu. Thanks everyone 
|
|
#
?
Mar 9, 2024 22:01
|
|
|
Klyith posted:So I don't do this myself, I'm just going off the instructions and various reports that it can work. But for every 1 person that it works for there are 5 who have problems. Good luck! Only tried with one game installed in windows and it couldn't see it. I've been able to install a game in Linux and have it work fine. The dream would be that I could boot into either and play without having to download and save elden ring or whatever twice. quote:
Will try this, thanks!
|
|
#
?
Mar 10, 2024 01:12
|
|
|
Shoud I be worried of these sort of errors on my dmesg:code:
|
|
#
?
Mar 10, 2024 15:12
|
|
|
Hm kinda seems like rasdaemon itself is crashing using it's own database; I would probably stop the daemon (systemctl stop rasdaemon) and clear out the db; on Debian it's /var/lib/rasdaemon/ras-mc_event.db and then start it again (systemctl start rasdaemon).
|
|
#
?
Mar 10, 2024 16:17
|
|
|
Less Fat Luke posted:Hm kinda seems like rasdaemon itself is crashing using it's own database; I would probably stop the daemon (systemctl stop rasdaemon) and clear out the db; on Debian it's /var/lib/rasdaemon/ras-mc_event.db and then start it again (systemctl start rasdaemon). code:
|
|
#
?
Mar 10, 2024 16:43
|
|
|
Pretty weird, what distribution? I would probably reboot and force a filesystem check and maybe run memtest for a few passes.
|
|
#
?
Mar 10, 2024 16:58
|
|
|
It’s just regular Ubuntu 23.10 box. I’ll reboot as soon as zfs scrub finishes and have it check the ext4 root partition. E: nope, still errors after clean reboot & fsck; namely when I do service rasdaemon restart code:Kivi fucked around with this message at 18:07 on Mar 10, 2024 |
|
#
?
Mar 10, 2024 17:38
|
|
|
Kivi posted:[ 240.276255] rasdaemon[12145]: segfault at 21 ip 00007e097ea1fa42 sp 00007e08157f89f0 error 6 in libsqlite3.so.0.8.6[7e097e9a4000+f7000] likely on CPU 37 (core 17, socket 0) Do you actually have a bigass CPU in there? One thing that comes up on google is an old (2022) kernel bug about something that would make rasdeamon try to access more CPUs than actually existed. And that would crash the system. And the kernel bug ends with "Success! Now rasdaemon is segfaulting, but that`s hardly a kernel problem." (lmao)
|
|
#
?
Mar 10, 2024 19:51
|
|
|
Klyith posted:Do you actually have a bigass CPU in there? One thing that comes up on google is an old (2022) kernel bug about something that would make rasdeamon try to access more CPUs than actually existed. And that would crash the system. 
|
|
#
?
Mar 10, 2024 20:36
|
|
|
I'm using it on a 32 core Threadripper with no issue so I don't think it's just simply "too many cores"
|
|
#
?
Mar 10, 2024 23:26
|
|
|
Is updating my linux kernel to the latest from the one that came installed with my distro a bad idea? Saw someone out there suggest updating it might give a performance boost with my hardware.
|
|
#
?
Mar 11, 2024 00:13
|
|
|
It's the kind of nerdy thing a bored teenager with infinite free time does, your distro should already be supplying you with a recent kernel unless you're on some kind of LTS release.
|
|
#
?
Mar 11, 2024 00:28
|
|
|
It is so, so rare that kernel performance increases are noticeable in the real world. Are you having specific performance problems? Most of the time, you get a new kernel by upgrading your packages and it's no big deal.
|
|
#
?
Mar 11, 2024 00:53
|
|
|
Most distros are built around the assumption that you're using the kernel they built for you and poo poo can blow up badly if you try to work around it. Assuming you get it working, the next package updates (even non kernel related) are very likely going to cause headache.
|
|
#
?
Mar 11, 2024 01:39
|
|
|
spiritual bypass posted:It is so, so rare that kernel performance increases are noticeable in the real world. Are you having specific performance problems? Battery consumption seems really high so I'm trying to see if it can be better. I've also seen recommendations for tools like TLP, power-profile-daemon and the like - any recommendations for an AMD laptop?
|
|
#
?
Mar 11, 2024 02:00
|
|
|
Updating the kernel if it's offered in the packages updates of your distro is perfectly fine. If it isn't available and you want something newer, take a look at the testing repos of your distro to see if they're available there.
|
|
#
?
Mar 11, 2024 02:09
|
|
|
mila kunis posted:Battery consumption seems really high so I'm trying to see if it can be better. I've also seen recommendations for tools like TLP, power-profile-daemon and the like - any recommendations for an AMD laptop? I used the powercap settings exposed by the kernel for an intel laptop for a while. Current laptops have super high burstable power limits but Imo they just end up burning a lot of that power into heat, especially when the AC configured limits don't expect you to have the laptop charging on your lap.
|
|
#
?
Mar 11, 2024 02:19
|
|
|
mila kunis posted:Battery consumption seems really high so I'm trying to see if it can be better. I've also seen recommendations for tools like TLP, power-profile-daemon and the like - any recommendations for an AMD laptop? I've heard good things about tuned (replaces power-profile-daemon) but usually followed by the caveat that you should use a custom profile, and with that name it is infuriatingly hard to Google useful examples.
|
|
#
?
Mar 11, 2024 06:59
|
|
|
Less Fat Luke posted:I'm using it on a 32 core Threadripper with no issue so I don't think it's just simply "too many cores" On the kernel lore posting, quote:> Just figured out something, the poll() errors starting at fd 29 (which
|
|
#
?
Mar 11, 2024 07:33
|
|
|
64 entries there. What's the processor model you're using?
|
|
#
?
Mar 11, 2024 14:27
|
|
|
Less Fat Luke posted:64 entries there. What's the processor model you're using?
|
|
#
?
Mar 11, 2024 17:43
|
|
|
I bought a system76 pangolin laptop a year ago and it arrived with a defective keyboard - pressing anywhere near the print screen key caused that key to register. And by "anywhere near" I mean -, +, backspace, or just the frame itself. I ended up getting an RMA and the new one worked fine. Until a couple days ago that is. Now the trackpad has suddenly turned into a jittery insensitive piece of junk. It was never great mind you but now it's borderline unusable. I did some searching and it looks like this is a fairly common hardware problem, suggested solutions include blasting it with a hair dryer which I'm not too excited about. I really wanted to support a pro-Linux manufacturer but ehhh this experience hasn't been great. I'm just going to replace it with a macbook (I had one years ago and the hardware was fantastic) and I'd definitely think twice about ordering anything from them again until I hear their track record has improved.
|
|
#
?
Mar 11, 2024 17:59
|
|
|
I thought it was still basically mandatory to buy an old thinkpad if you want to run Linux on a laptop.
|
|
#
?
Mar 11, 2024 18:01
|
|
|
keep punching joe posted:I thought it was still basically mandatory to buy an old thinkpad if you want to run Linux on a laptop. Yeah almost certainly. Before the pangolin I was running Linux on a dell xps 13 and it worked okay... but it never slept properly and would burn through like 6% battery/hour even with the lid closed. I've heard it works fine on thinkpads and MAYBE on frameworks? Regardless I just want a laptop that works and I'm not a hardcore enough Linux zealot to pursue it further. I'll content myself with Linux on my desktop (this is the year!!!) and homelab servers.
|
|
#
?
Mar 11, 2024 18:11
|
|
|
My AMD Framework burns about 1% battery per hour sleeping. Not great, but it's ok if I intend to use it the next day and also since it charges quite quickly. I want to switch over to making it hibernate instead of sleep but it's a big pain for me to repartition and configure the button and lid close triggers.
|
|
#
?
Mar 11, 2024 18:15
|
|
|
acetcx posted:I bought a system76 pangolin laptop a year ago and it arrived with a defective keyboard - pressing anywhere near the print screen key caused that key to register. And by "anywhere near" I mean -, +, backspace, or just the frame itself. I ended up getting an RMA and the new one worked fine. My housemate has a system76 laptop they got 2nd-hand. The keyboard on that is flaky too, and it's a totally different & much older model. OTOH system 76 was incredibly good about answering questions & support long after the warranty expired. One of the fan units died and we contacted their support asking if we could buy a replacement. The model was old enough that they didn't have parts, but they gave us exact part numbers for everything and pointed to places on ali express etc that still carried stuff. So yeah some ups and downs. (My housemate uses Windows on it, and amusingly the hibernation doesn't work properly.) Mantle posted:I want to switch over to making it hibernate instead of sleep but it's a big pain for me to repartition and configure the button and lid close triggers. Apparently you can do hibernate with a swapfile instead of swap partitions if you do some fiddling.
|
|
#
?
Mar 11, 2024 18:50
|
|
|
Do System 76 build their own laptops or are they just rebranded Clevo devices?
|
|
#
?
Mar 11, 2024 18:54
|
|
|
Early modern suspend laptops were all kind of crap in that regard imo, even if you were using windows.
|
|
#
?
Mar 11, 2024 19:45
|
|
|
Macs with their M1,2,3 etc are the only ones that get suspend right.
|
|
#
?
Mar 11, 2024 19:57
|
|
|
keep punching joe posted:I thought it was still basically mandatory to buy an old thinkpad if you want to run Linux on a laptop. I've been using Chromebooks as my "Linux" laptop of choice for the past decade and although there is an entirely different set of janitorial issues with them, you can be sure you'll always have at least a fully functional web browser and terminal available, which is 90% of my work anyways.
|
|
#
?
Mar 11, 2024 22:57
|
|
|
Lenovo offer Linux pre installed also.
|
|
#
?
Mar 11, 2024 23:05
|
|
|
I'm trying to learn how to best write shell scripts. I have the following script for downloading YouTube videos as MP3 and injecting them with metadata + cover art:code:code:pre:music url1 && music url2
|
|
#
?
Mar 12, 2024 01:39
|
|
|
neurotech posted:I'm trying to learn how to best write shell scripts. I have the following script for downloading YouTube videos as MP3 and injecting them with metadata + cover art: Exec is replacing the current process, which in this case is bash, (just like the exec call) with the new one. While I honestly have no idea what happens when you chain multiple exec calls (never occurred to me to try such a thing, I should experiment with this new found idea) this is most definitely the culprit. Just make your script executable (chmod u+x) and remove the exec in the alias and you should be golden. Volguus fucked around with this message at 01:50 on Mar 12, 2024 |
|
#
?
Mar 12, 2024 01:46
|
|
|
If it only runs the first one then it's returning something other than zero; '&&' is a logical AND, so if the first one is false then there's no way the entire statement can be anything other than false. So, is yt-dlp returning an error? why do you have the $SHELL at the end there? etc Also just a tip, if you have a folder full of scripts you can just add it to your path instead of individual aliases for everything
|
|
#
?
Mar 12, 2024 01:49
|
|
|
Also you don't need a shell alias, ~/bin and ~/.local/bin are probably already in your path.
|
|
#
?
Mar 12, 2024 01:49
|
|
|
Thanks for those replies, that is a huge help.hifi posted:So, is yt-dlp returning an error? why do you have the $SHELL at the end there? etc Adding $SHELL to the end is (supposedly) to ensure the script doesn't kill my shell and drop my ssh connection. At least that's what superuser says haha pseudorandom name posted:Also you don't need a shell alias, ~/bin and ~/.local/bin are probably already in your path. Volguus posted:Just make your script executable (chmod u+x) and remove the exec in the alias and you should be golden.
|
|
#
?
Mar 12, 2024 01:52
|
|
|
|
| # ? May 14, 2024 07:52 |
|
|
neurotech posted:What does chmod u+x do? I normally just do chmod +x +x only for the owning user rather than everyone
|
|
#
?
Mar 12, 2024 01:56
|
|