Announcement

Collapse
No announcement yet.

The Most Interesting Highlights To The Linux 5.0 Kernel

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • Compartmentalisation
    replied
    Thank you everyone for the NVMe experiences! Don't want to ping you.

    Originally posted by xxmitsu View Post

    If you want to live bleeding edge on fedora, you might be interested in:
    I've been using those for more than a year.

    Code:
    [[email protected] ~]$ glxinfo | grep 'OpenGL renderer'
    OpenGL renderer string: AMD RAVEN (DRM 3.27.0, 5.0.0-0.rc7.git0.1.fc30.x86_64, LLVM 9.0.0)
    [[email protected] ~]$ uname -a
    Linux localhost.localdomain 5.0.0-0.rc7.git0.1.fc30.x86_64 #1 SMP Mon Feb 18 04:32:16 UTC 2019 x86_64 x86_64 x86_64 GNU/Linux
    Thanks for the links. It's actually Fedora using mainline kernel that makes it incredibly buggy even on "blessed" ThinkPad hardware with Intel graphics. Most of the rest of the software is bug-free. But you have things like a random 4.20 update breaking drivers again, trying 5.0 and finding them still broken. Or PowerTOP being broken with the latest 4.20 update (which is crucial for my laptop because of the poor default kernel power management; with mostly only Fedora and Intel trying to make a difference there). There's a copr with the latest LTS kernel which works nicely for me, for F28 and F29.

    Sorry for the negativity, thanks for the links ❤️

    Leave a comment:


  • xxmitsu
    replied
    Originally posted by Compartmentalisation View Post
    I'm considering stepping out of Fedora to get out of mainline hell. Using the latest versions has certainly become better over the last years, but it's still the most buggy software on my laptop.
    If you want to live bleeding edge on fedora, you might be interested in:
    I've been using those for more than a year.

    Code:
    [[email protected] ~]$ glxinfo | grep 'OpenGL renderer' 
    OpenGL renderer string: AMD RAVEN (DRM 3.27.0, 5.0.0-0.rc7.git0.1.fc30.x86_64, LLVM 9.0.0)
    [[email protected] ~]$ uname -a
    Linux localhost.localdomain 5.0.0-0.rc7.git0.1.fc30.x86_64 #1 SMP Mon Feb 18 04:32:16 UTC 2019 x86_64 x86_64 x86_64 GNU/Linux

    Leave a comment:


  • RussianNeuroMancer
    replied
    Originally posted by Compartmentalisation View Post
    Are any NVMe drives old enough to be dying yet?
    This one is not old for sure, but SSD could die regardless of it's age. For example SanDisk SD6SP1M128G1012 (OEM SSD in Dell Venue 11 Pro 7140) died two times (but it was replaced under Dell's warranty) and third death is going on right now (it's not always recognized at boot, and sometimes throw ATA errors to dmesg). I realize that if underlying device is not reliable, FS can't do anything about it, however, this time with btrfs partition that was corrupted twice there seems like no indication of SSD or other hardware-related issues in dmesg. But, maybe I missing something, so I switched to 4.20 for now - let's see if btrfs partition survive another month.

    Leave a comment:


  • rbmorse
    replied
    I had a 500GB Western Digital Black series that was DOA. Replacement unit has been fine over 6 months.

    Leave a comment:


  • Michael
    replied
    Originally posted by Compartmentalisation View Post
    RussianNeuroMancer "I hope that because my SSD is dying or issues between NVME and eGPU connected via TB3, or something like that, otherwise btrfs in Linux 5.0 is unstable beyond anything I seen before."

    Are any NVMe drives old enough to be dying yet? I hope that's more to do with you using RC versions.
    Aside from a Samung 950 PRO that died prematurely on me, I've only had 2 or 3 off-brand NVMe drives die on me so far.

    Leave a comment:


  • starshipeleven
    replied
    Originally posted by Compartmentalisation View Post
    Are any NVMe drives old enough to be dying yet? I hope that's more to do with you using RC versions.
    If they are Samsung, they can die from day one, they don't need to get old.

    More seriously, firmware issues could do that.

    But it's an rc kernel so all bets are off on what is broken.

    Leave a comment:


  • Compartmentalisation
    replied
    RussianNeuroMancer "I hope that because my SSD is dying or issues between NVME and eGPU connected via TB3, or something like that, otherwise btrfs in Linux 5.0 is unstable beyond anything I seen before."

    Are any NVMe drives old enough to be dying yet? I hope that's more to do with you using RC versions.

    geearf "it's been quite a while since I got really excited about a new kernel :/"

    I just wonder how many things will break now. I'm considering stepping out of Fedora to get out of mainline hell. Using the latest versions has certainly become better over the last years, but it's still the most buggy software on my laptop.

    Leave a comment:


  • RussianNeuroMancer
    replied
    Btrfs was mostly stable for me, for years. Yet btrfs partition died twice this month, first time with Linux 5.0rc4 and second time with Linux 5.0rc6, moreover second time btrfs refuse to mount partition even in read-only mode. I hope that because my SSD is dying or issues between NVME and eGPU connected via TB3, or something like that, otherwise btrfs in Linux 5.0 is unstable beyond anything I seen before.

    Leave a comment:


  • geearf
    replied
    I'm sad the only interesting point on this for me is the mice, it's been quite a while since I got really excited about a new kernel :/

    Leave a comment:


  • Brisse
    replied
    Originally posted by brrrrttttt View Post
    I'd be surprised if many others have tried. I'll give it a go once the necessary packages hit Arch repos.
    I tried a couple of times in the past but could not make it work at all.

    Leave a comment:

Working...
X