Announcement

Collapse
No announcement yet.

More AMDGPU DC Patches Posted As It Looks Unlikely It Will Land For Linux 4.13

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

  • M-Bab
    replied
    DarkFoss You're welcome. I personally learned way more after publishing the kernel than before when I was only building the stuff for myself. I now added 4.12 but didn't dare to publish drm-next yet: It is based on a 4.13 release candidate and I assume things won't get more stable if I combine the bleeding edge AMD code with an unreleased kernel ;-)

    Leave a comment:


  • DarkFoss
    replied
    M-Bab Just wanted to say Thank You again. The 4-11 staging was a fun ride. The only issues I had from 4.11-10 to 4.11-12 were coming out of S3. I didn't bother reporting since looking through the commits at agd5f's site I could see possible fixes were already coming down the line, 1st one landed 7/20 in Ubuntu's drm-next. With yesterdays drm-next and todays drm-tip 4.13.0-994-generic my Crosshair V Formula-Z /BIOS 2201 is probably as close to golden status as it will get.

    Leave a comment:


  • beniwtv
    replied
    Hopefully there will be a 4.13-amd-staging soon
    Right now I can either have VR working (Mesa needs the amdgpu module in 4.13rc1+ for that), or have working sound (via 4.11 staging), and no VR.

    Leave a comment:


  • Laughing1
    replied
    So kernel 4.14 or 4.15 or beyond?

    Leave a comment:


  • M-Bab
    replied
    DarkFoss Thanks for being a brave tester and user of this bleeding edge kernel. So far 4.11.5+ is working amazingly for me too - it fixed an issue with Tyranny I had since kernel 4.9.13+(!).

    Whenever you notice problems that are undoubtedly linked to the kernel feel free to open an issue on github. Then I try to track it down a bit and report the bug upstream to the AMD devs.

    Leave a comment:


  • DarkFoss
    replied
    M-Bab Thank you for rolling the 11 staging branch.
    The difference from 4.11.4+ to 4.11.5+ is nice the vulcan error I had with the Talos Principle is gone. I see a Code Guard error (happened with 4.11.4 too)with Serious Sam Fusion but the benchmark runs anyway. It doesn't happen with the generic Ubuntu kernels.

    The best surprise is the constant spam during start-up is fixed, no more kernel: IO_PAGE_FAULT device=01:00.0 (My 1st Fury X ) it was present in 4.11.4+.
    That error is still in today's daily roll of drm/next 4.12.0-996.21706162316-generic_amd64 and all Ubuntu kernels I've tested.

    It's just amazing how everything is coming together. A big Thank You to all the hard working devs.

    Leave a comment:


  • Mystro256
    replied
    Originally posted by bridgman View Post
    Isn't Mystro256 building packages off the staging trees ?
    Indeed but only for Fedora though. There's also someone who's doing the same for Debian/Ubuntu:

    https://github.com/M-Bab

    Leave a comment:


  • lordheavy
    replied
    Originally posted by zanny View Post
    Is there an amdgpu-dc repo for Arch anyone knows of? I'd love to just try it, but the Arch wiki doesn't have one listed.

    See http://pkgbuild.com/~lcarlier/mesa-git/

    Leave a comment:


  • bridgman
    replied
    Isn't Mystro256 building packages off the staging trees ?

    Leave a comment:


  • chithanh
    replied
    Originally posted by marek View Post
    Or wait for a PPA, though there already might be one.
    There is http://packages.amd.com/ but it contains only ROCm code, not DC/DAL.
    Plus from the package filenames it appears that things there are based off Linux 4.9.0 which would mean that the kernels there may be behind on security patches.

    Leave a comment:

Working...
X