Recommended Posts

So I had a pretty much fine arch linux 64 bit install with KDE, I had pulseaudio and using alsamixer I could see the full range of my card's audio devices, master, front, left, side, mic, line, etc. and all was great...

I've been looking just now though and alsamixer's sound has been reduced to what I will consider utter garbage that makes it pretty much unusable. Master, headphone (there is no headphone slot), speaker (there is no internal speaker), PCM, Mic Boost, Beep, Auto+Mute Mode, Independant HP...

Now, I've no idea what auto+mute is nor what independant HP is, but what's really ****ed me off is I can no longer get mic or line or play either out through my speakers. I frantically removed pulseaudio and reinstalled it a few times thinking I'd somehow messed that up but it wasn't that, looking through the pacman upgrade log the only thing I see that has anything to do with sound is upgrading from kernel 3.8 to 3.9 earlier...

So does this mean it's the doing of the 3.9 kernel or what? Has the kernel been bastardised now so that it's just a pile of garbage? And most importantly, how the hell do I go about getting back a usable sound system instead of this unworkable system?

Link to comment
https://www.neowin.net/forum/topic/1151822-kernel-39-messed-up-my-sound/
Share on other sites

Oh and would you look at that, downgraded back to 3.8 and all my sound options comes back.

I'm not even blaming the arch team for this, now it seems the kernel team are incompetant.

Getting so tired or actual idiots 'fixing' or 'improving' things when they haven't got a clue and they just break everything, gonna format this and install debian stable later I think.

For anyone else on arch that needs a fix, go to the arch linux release mirror thing and find linux-3.8.11, pacman -U <package name>, reboot and all will be fine again.

Oh and would you look at that, downgraded back to 3.8 and all my sound options comes back.

I'm not even blaming the arch team for this, now it seems the kernel team are incompetant.

Getting so tired or actual idiots 'fixing' or 'improving' things when they haven't got a clue and they just break everything, gonna format this and install debian stable later I think.

For anyone else on arch that needs a fix, go to the arch linux release mirror thing and find linux-3.8.11, pacman -U <package name>, reboot and all will be fine again.

Debian Stable is a blast. I would highly recommend it. :)

  • Like 3

*sigh* don't blame the distro or the kernel maintainers. Blame yourself, what chipset is the audio card you have, have you read the changelog for kernel 3.9 have you read Arch documentation before upgrading to 3.9? Did you try loading the needed kernel modules after the upgrade?

The kernel maintainers are strict with their releases, its not their fault. As for Arch im an Arch user and have had no problems with any upgrades because i don't upgrade blind, with rolling release distros you need to read things first before you upgrade to ensure that things wont break and if they do you will then know how to reconfigure them correctly.

did you try recompiling a custom 3.9 kernel to ensure you have the correct modules loaded?

All modules are loaded correctly, it picks up the exact same audio card on 3.8 and 3.9, the only difference is that 3.9's audio channels are missing basically everything.

I don't have the time nor effort to be bothered to compile a kernel and test it on this pretty slow PC.

If you've got arch, and have upgraded it without ever having problems then you are doing something wrong unless you do a brand new install every so often.

If you've got arch, and have upgraded it without ever having problems then you are doing something wrong unless you do a brand new install every so often.

Wait.... What!?

I would say they are actually reading the change notes and not indiscriminately applying all updates, also known as doing it right.

  • Like 3

*sigh* don't blame the distro or the kernel maintainers. Blame yourself, what chipset is the audio card you have, have you read the changelog for kernel 3.9 have you read Arch documentation before upgrading to 3.9? Did you try loading the needed kernel modules after the upgrade?

The kernel maintainers are strict with their releases, its not their fault. As for Arch im an Arch user and have had no problems with any upgrades because i don't upgrade blind, with rolling release distros you need to read things first before you upgrade to ensure that things wont break and if they do you will then know how to reconfigure them correctly.

did you try recompiling a custom 3.9 kernel to ensure you have the correct modules loaded?

He should blame himself for something that should work but doesn't? Yeah, somehow that doesn't make sense. I only run Linux in a VM and this is an example why that makes more sense. I feel for some of you troopers that have to go all through this pain to get stuff to work.

Should I waste my life away and my time away on just getting basic things to work or should I just get work done that needs to get done? hmmmm.

Wait.... What!?

I would say they are actually reading the change notes and not indiscriminately applying all updates, also known as doing it right.

Yeah that's right they fully documented that when they first switched to systemd, it wasn't actually complete and lots of things didn't work... OH NO WAIT, they didn't!

Then there was the boost problem which still hasn't been fixed, oh and who reported that? People that use arch and had their installs **** up because of it.

More recently was the hash of GCC 4.8, it was posted all over the net (yes, even on gcc's site) about there being problems with files not being included... Did arch apply the patch from GCC? No. Did arch revert back to the 4.7 version that works? No. Does the arch team tell you that this bug is present? No.

So yes you go and read the change notes and spot where it says 'Oh and btw applying this update will make you unable to compile any packages'.

So I had a pretty much fine arch linux 64 bit install with KDE, I had pulseaudio and using alsamixer I could see the full range of my card's audio devices, master, front, left, side, mic, line, etc. and all was great...

I've been looking just now though and alsamixer's sound has been reduced to what I will consider utter garbage that makes it pretty much unusable. Master, headphone (there is no headphone slot), speaker (there is no internal speaker), PCM, Mic Boost, Beep, Auto+Mute Mode, Independant HP...

Now, I've no idea what auto+mute is nor what independant HP is, but what's really ****ed me off is I can no longer get mic or line or play either out through my speakers. I frantically removed pulseaudio and reinstalled it a few times thinking I'd somehow messed that up but it wasn't that, looking through the pacman upgrade log the only thing I see that has anything to do with sound is upgrading from kernel 3.8 to 3.9 earlier...

So does this mean it's the doing of the 3.9 kernel or what? Has the kernel been bastardised now so that it's just a pile of garbage? And most importantly, how the hell do I go about getting back a usable sound system instead of this unworkable system?

I've upgraded my Arch install (with Pulseaudio) to 3.9 without issue, and I don't see any of these other issues that you're having either. My Arch install has been running perfectly fine for months, and I upgrade my packages daily. My last full re-install was in 2011, and that's because I shafted my system by doing stuff I shouldn't have and didn't have the ability to recover it. Are you doing something special with your packages? Custom compiles, that kind of thing? My understanding is that there are no changes to the sound system in 3.9 (there are however a few changes coming in 3.10).

For what it's worth, my understanding of the Arch packages is that they don't really do much to them. They simply compile them to the Arch configuration. If the GCC/Boost patches didn't come from the source as an official release, I doubt that Arch would add them.

To be frank, I don't think a bleeding edge distro is what you're looking for in a Linux flavour. If you're wanting reliability and stability over having the latest 'n' greatest, your time would probably be better served using the likes of Debian and CentOS, which place a much higher emphasis on stability than the likes of Arch.

Nope everything on this arch pc is standard off their mirrors.

Boost in arch is broken, I don't know if it's not compiled right or what but it's missing a lot of stuff.

The GCC patch for 4.8 is used on gentoo and other distros and is accepted by the GCC team.

I don't need stability on this PC, heck I run arch servers and they're very stable, apart from the messups with arch packages like changing cairo to require nvidia VDPAU libraries which have a tonne of dependancies for no reason what so ever (on a server).

It sounds like an upstream kernel/alsa regression, a quick google search shows similar issues for people using kernel 3.9 from kernel:stable on opensuse 12.3, and people using ubuntu 13.04 that manually upgraded to kernel 3.9.

He should blame himself for something that should work but doesn't? Yeah, somehow that doesn't make sense. I only run Linux in a VM and this is an example why that makes more sense. I feel for some of you troopers that have to go all through this pain to get stuff to work.

Should I waste my life away and my time away on just getting basic things to work or should I just get work done that needs to get done? hmmmm.

He's using a bleeding edge rolling-release power-user focused distro, if you use a bleeding edge distro you will run into regressions from time to time after updates, its simply the nature of running cutting-edge software. And if you run cutting edge software you should be prepared to report bugs instead of freaking out every time you encounter one. Most distros do not do major kernel updates in between releases (only smaller security and bugfix updates), so this type of thing is much more rare.

From the threads I've seen from him lately, he should blame himself for choosing a distro that clearly does not fit his needs.

He's using a bleeding edge rolling-release power-user focused distro, if you use a bleeding edge distro you will run into regressions from time to time after updates, its simply the nature of running cutting-edge software. And if you run cutting edge software you should be prepared to report bugs instead of freaking out every time you encounter one. Most distros do not do major kernel updates in between releases (only smaller security and bugfix updates), so this type of thing is much more rare.

From the threads I've seen from him lately, he should blame himself for choosing a distro that clearly does not fit his needs.

I think you're getting a bit confused between using the normal arch linux repos and the TESTING arch linux repos.

Yes, arch IS a rolling release system, but that's for the testing repos, the whole idea of arch is brand spanking new packages are put into testing and tested to see if anyone finds any problems with them or whatnot and then once they're verified to be usable, they're moved into the normal repos. Moving kernel 3.9 when sound is completely messed up from testing to normal signifies to me that either the arch team don't do enough testing/are too eager to move things over or the hardware this affects is relatively small (I don't know what it does this on other than my system).

I've never used the testing repos and don't see why it's supposedly my fault if something like this comes from a package in the normal repo. The biggest difference between testing things and normal things is testing are meant to have untested features that might have stability problems or be wrongly implemented, they're not meant to break rather large features that worked absolutely fine in all previous versions.

No.. you lost your hearing... your sounds are fine as we can hear it from here... :-P

Joking..

Anyway, Some systems are different based on system configurations... If it doesn't work, you can rollback to previous version or try to install a different distro such as Ubuntu, RedHat, etc.

Never use testing apps or OS (beta or public preview release) on your primary system... you will need a backup system for testing to make sure they work... if so, you can upgrade or fresh install on your primary.

Never use testing apps or OS (beta or public preview release) on your primary system... you will need a backup system for testing to make sure they work... if so, you can upgrade or fresh install on your primary.

Unless his backup system is exactly the same, that would be irrelevant.

I find this funny really, this problem is clearly related to your configuration or your system. Not the kernel nor the arch team. I have the 3.9 kernel (see below) and my sound hasn't took a hit in the slightest using both my Logitech USB headset and my Onboard Intel sound.

All you appear to be doing in this thread is distro bashing with no interest in finding out why your system didnt like the 3.9 update.

qSHEfRu.png

  • Like 1

have you tried, disabling alsa, disabling p/a then re-enabling them both and removing the kernel module and re modprobing? (rmmod snd_hda_intel && modprobe snd_hda_intel) Also may be required to re-set your default sound card.

It's unrelated to pulseaudio, logging in to fluxbox or console without starting pulseaudio shows the same problem. Removing and reloading the kernel module gives the same result. Resetting default sound card doesn't matter as there's only one sound card in the system.

And I just installed Ubuntu 13.04 no problems... it even detected my creative x-fi usb soundcard and runs steam, man...

I'm planning to hit debian stable up and give that a go.

I think you're getting a bit confused between using the normal arch linux repos and the TESTING arch linux repos.

Yes, arch IS a rolling release system, but that's for the testing repos, the whole idea of arch is brand spanking new packages are put into testing and tested to see if anyone finds any problems with them or whatnot and then once they're verified to be usable, they're moved into the normal repos. Moving kernel 3.9 when sound is completely messed up from testing to normal signifies to me that either the arch team don't do enough testing/are too eager to move things over or the hardware this affects is relatively small (I don't know what it does this on other than my system).

I've never used the testing repos and don't see why it's supposedly my fault if something like this comes from a package in the normal repo. The biggest difference between testing things and normal things is testing are meant to have untested features that might have stability problems or be wrongly implemented, they're not meant to break rather large features that worked absolutely fine in all previous versions.

I am not confused, arch is a self-proclaimed bleeding edge distro. Even the packages in stable are more cutting edge than most other distros, and add the fact that its an uncompromising rolling release. A lot of people like this about arch and that's why they use it, but you shouldn't expect absolute stability out of it.

With more stable non rolling release distros, the packages they are using generally have way more testing than anything in arch's stable repos, and normal updates generally don't break things because its more limited to smaller bugfix and security updates (although when you do a dist-upgrade to a new major versions thats when things can certainly break). With arch you are pretty much always doing a "dist-upgrade", but in smaller increments if you update often. Any arch update can update core parts of your system to totally new major versions, and this can introduce major regressions.

I'm not saying its 'your fault' that kernel 3.9 broke your sound, thats probably some sort of upstream kernel bug, what I am saying is that your expectations of a cutting edge rolling release distribution to be regression-free is unrealistic.

With arch you get the latest versions of upstream software, and when you use the latest versions of upsteam software, you get the latest bugs of upstream software.

Well yes I agree arch is bleeding edge and that's nice, but I've always been under the assumption that testing was for the bleeding edge packages and the normal repos were for packages that were tested and are confirmed working and not borked hence why gnome 3.6 stayed in testing for a long time?

Maybe it's not like that then. I've got some free time anyway now so I'll be giving debian stable a try and see if it's worth moving my servers over to that.

Has anyone ever bothered with LFS (Linux from scratch)? I know it'd be a pain in the arse in terms of old files left all over the place but starting to think it's worth compiling everything from scratch and that'll show all the errors and whatnot better than things like this and the GCC bug.

Unless his backup system is exactly the same, that would be irrelevant.

So? That does not matter if it is same system or not. He needs a backup/test system so he could test any apps or OS to make sure it works properly...

If not working, you can always rollback to previous version or install an alternative distro...

I have primary system that works fine... I have backup which is for testing so I could make sure it runs/works well.

I am not that stupid to install unstable or buggy app/OS on my primary system...

If my friends come by to borrow my system... I let them to use my testing system for a bit. I wouldn't let anyone to use my primary system..

This topic is now closed to further replies.
  • Recently Browsing   0 members

    • No registered users viewing this page.
  • Posts

    • This makes me think of Dune for some reason.
    • I'm too old to return to the "good old days" when I was installing custom ROMs and tinkering with my devices - now I just want to turn it on and use it. I've read that banking and payment apps work on Murena /e/OS (I'll have to check the ones I use) and I also really want to support Fairphone 😉
    • Time to start going to the local church and play Bingo for a while.
    • NVIDIA announces 35 new AI HPC supercomputers across Europe by Fiza Ali NVIDIA has announced that 35 AI high-performance computing (HPC) supercomputers are planned to open throughout Europe this year. This marks what the company describes as the largest single-year expansion of AI infrastructure in the history of the continent. These new systems, unveiled at ISC High Performance 2026, will be placed at a number of national supercomputing centres, AI factories, and research institutes to provide advanced computing resources to more than three million researchers. Describing AI, NVIDIA founder and CEO Jensen Huang stated, "AI is the new instrument of science, and Europe is building the infrastructure to put it in the hands of millions of researchers." Built on NVIDIA's Blackwell and Hopper architectures, the new systems will support research in climate science, healthcare, clean energy, quantum computing, and other scientific fields. Among the major projects are the Barcelona Supercomputing Center's MareNostrum 5 AI upgrade, BavariaAI's Blue Swan platform in Germany, Italy's IT4LIA AI factory, Germany's HammerHAI project, and Sweden's Mimer AI Factory. The Barcelona Supercomputing Center plans to expand MareNostrum 5 with NVIDIA GB300 NVL72 and GB200 NVL4 systems. In total, the BSC expects to deliver up to 20 exaflops of AI training performance and 33 exaflops of AI inference performance. This increased computational capability will support research efforts related to climate modelling, biotechnology, energy systems, etc. Furthermore, as part of the IT4LIA project, more than 8,000 GPUs, each based on NVIDIA’s GB200 NVL4 architecture, will be used in Italy. This represents one of the largest AI factory initiatives announced to date. Additionally, the Blue Swan platform from BavariaAI will include about 1,000 GPUs to help develop multimodal AI models for use in the medical field, robotics, and various areas of scientific research. NVIDIA also emphasized in the announcement how rapidly growth of accelerated computing usage is taking place within both energy and climate-related research. The company said Siemens Energy uses NVIDIA-powered technologies to significantly accelerate the process of designing and simulating hydrogen-capable gas turbines. Using those same acceleration technologies, Siemens was able to reduce simulation time by up to 77 percent. The company also highlighted several quantum computing initiatives across Europe. CINECA, EuroHPC, and Pasqal are integrating a quantum processing unit into Italy's CINECA supercomputing centre using NVIDIA's CUDA-Q platform. Meanwhile, researchers at Germany's Julich Supercomputing Centre recently simulated a universal 50-qubit quantum computer on the JUPITER supercomputer. The announcement demonstrates Europe's continued commitment to building out its infrastructure supporting AI and supercomputing as governments, research organizations, and technology companies compete to build out their respective computing capacities and secure their positions in advanced scientific research.
  • Recent Achievements

    • First Post
      mike_rumble earned a badge
      First Post
    • Dedicated
      tuben earned a badge
      Dedicated
    • Week One Done
      mnsgroup earned a badge
      Week One Done
    • Conversation Starter
      sumytbe earned a badge
      Conversation Starter
    • One Year In
      B4dM1k3 earned a badge
      One Year In
  • Popular Contributors

    1. 1
      +primortal
      508
    2. 2
      +Edouard
      203
    3. 3
      PsYcHoKiLLa
      98
    4. 4
      Michael Scrip
      82
    5. 5
      neufuse
      67
  • Tell a friend

    Love Neowin? Tell a friend!