Announcement

Collapse
No announcement yet.

Linux 5.1-rc1 Kernel Released After A "Fairly Normal" Merge Window

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

  • #11
    Originally posted by clapbr View Post
    Still no fix for the AMD memory clock flickering Come on devs, it cripples faster displays and is 3 years old almost.
    My guess is that there's some project manager at AMD that isn't giving developers time to go back and fix/debug issues for older GPUs, as they want to focus their limited resources on new GPU enablement.

    Comment


    • #12
      Originally posted by Space Heater View Post

      My guess is that there's some project manager at AMD that isn't giving developers time to go back and fix/debug issues for older GPUs, as they want to focus their limited resources on new GPU enablement.
      I follow the ML and I don't think that's the case, they are constantly fixing other Polaris issues, some of them less important than this, but not abandoned by any means. It seems some specific bugs just end in a blackhole that no one cares digging, maybe for it's technical difficulty, maybe for "luck".

      Comment


      • #13
        Originally posted by shmerl View Post

        I mean compression itself. btrfs has a forced and autodetect mode. But the later is not very smart, and doesn't compress files if first few blocks aren't compressing well.
        Zstd's internal skipper is much faster than btrfs's internal algorithm. compress-force is therefore faster even for incompressible data. Except maybe mounting without compression.

        Comment


        • #14
          Originally posted by Spam View Post

          Zstd's internal skipper is much faster than btrfs's internal algorithm. compress-force is therefore faster even for incompressible data. Except maybe mounting without compression.
          Good to know, thanks!

          Comment


          • #15
            Originally posted by clapbr View Post

            I follow the ML and I don't think that's the case, they are constantly fixing other Polaris issues, some of them less important than this, but not abandoned by any means. It seems some specific bugs just end in a blackhole that no one cares digging, maybe for it's technical difficulty, maybe for "luck".
            VI and later cards are pretty well supported. Sea Islands are somewhat supported - generally things that are quick fixes are done, while things requiring lots of investigation are not. SI cards are largely not supported at this point, beyond getting things to compile.

            Comment

            Working...
            X