Announcement

Collapse
No announcement yet.

It Looks Like Raven Ridge Desktop APUs Will Work Better With Linux 4.17

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

  • #11
    Mez' New DC also gives me some troubles like artifacts >73Hz and it also seems that DRI3 isn't working correctly anymore. However, I'm lucky and don't have to rely on it for my DL-DVI display, the DPM improvements are also present with old DC.
    With old DC, I'm basically 100% free of trouble with 4.16 drm-next kernel. I guess I won't be using new DC for the next few months at least, but AMD guys seem busy improving it, so who knows.

    Comment


    • #12
      Originally posted by davidbepo View Post
      good to hear my 2400g still needs more polishing, both on linux and on windows
      I'm in the market for a Ryzen APU notebook this year. What distribution and kernel/Mesa are you using, and can you comment on what does/does not work and what was difficult to make work?

      Comment


      • #13
        I finally got my Raven Ridge running (2400G) but I had to wait until an A-Series arrived so I could flash the BIOS for RR support.

        While this one will be a Win10 box, I did boot several Linux versions off USB3 to see if any issues cropped up. None that I could see, but it was limited use.

        My quibble is in two places. While the socket is AM4, the CPU mounting methods between the older A-Series and the newer Ryzen series makes for some minor gymnastics to get the board ready.

        In my particular case, the A-Series and AMD cooler use a leaf spring latch (which they have used for several years). The Ryzen 2400G kit uses the newer 4 sprung screws. Both mount to the cooler plate on the back side of the planar. So CPU swaps can't be done while it is in case. While I don't foresee anymore swaps in the future, it is suggested to flash the board outside the case before you put the RR CPU in. Also note that once the BIOS is updated and the RR is installed the PC boots twice to get the 2400G recognized.

        My other quibble is with driver support. Even though it is new, I keep getting a driver load issue for the Vega. This is causing all the GPU Compute benchmarks to fail. The driver reports that there is "inadequate resources available for the Vega graphics to start properly". Everything else works just fine. I can crank the resolutions up and down, change the memory sizes, but can't run any OpenCL activities.

        I am troubleshooting the resource issue currently and already have trashed the motherboard twice requiring a CMOS reset. It may be a board/BIOS issue and I will eventually track it down.

        But this being a very new CPU, the board support appears to be somewhat immature even though the socket is the same.

        Comment


        • #14
          Originally posted by aufkrawall View Post
          Mez' New DC also gives me some troubles like artifacts >73Hz and it also seems that DRI3 isn't working correctly anymore. However, I'm lucky and don't have to rely on it for my DL-DVI display, the DPM improvements are also present with old DC.
          With old DC, I'm basically 100% free of trouble with 4.16 drm-next kernel. I guess I won't be using new DC for the next few months at least, but AMD guys seem busy improving it, so who knows.
          I had an issue with DRI3 as well a few days ago.
          1st, X wouldn't start, I had to purge Padoka ppa (unstable and stable) on tty. Then X started with amdgpudrmfb or llvmpipe as the driver in use (software rendering) and no DRI3.
          I finally had DRI3 back to normal by using Oibaf ppa and install libegl-mesa0 (which removed some conflicting packages) after searching on the Internet. Now, I can log in normally (sometimes restarting gdm service is needed). Very unstable in any case.

          Comment


          • #15
            Originally posted by edwaleni View Post
            I finally got my Raven Ridge running (2400G) but I had to wait until an A-Series arrived so I could flash the BIOS for RR support.

            While this one will be a Win10 box, I did boot several Linux versions off USB3 to see if any issues cropped up. None that I could see, but it was limited use.

            My quibble is in two places. While the socket is AM4, the CPU mounting methods between the older A-Series and the newer Ryzen series makes for some minor gymnastics to get the board ready.

            In my particular case, the A-Series and AMD cooler use a leaf spring latch (which they have used for several years). The Ryzen 2400G kit uses the newer 4 sprung screws. Both mount to the cooler plate on the back side of the planar. So CPU swaps can't be done while it is in case. While I don't foresee anymore swaps in the future, it is suggested to flash the board outside the case before you put the RR CPU in. Also note that once the BIOS is updated and the RR is installed the PC boots twice to get the 2400G recognized.

            My other quibble is with driver support. Even though it is new, I keep getting a driver load issue for the Vega. This is causing all the GPU Compute benchmarks to fail. The driver reports that there is "inadequate resources available for the Vega graphics to start properly". Everything else works just fine. I can crank the resolutions up and down, change the memory sizes, but can't run any OpenCL activities.

            I am troubleshooting the resource issue currently and already have trashed the motherboard twice requiring a CMOS reset. It may be a board/BIOS issue and I will eventually track it down.

            But this being a very new CPU, the board support appears to be somewhat immature even though the socket is the same.
            I use Antec C400 cooler for 2400G which uses stock AMD mountings(the spring loaded bracket). This cooler supports almost all intel sockets also starting from 775. Performance is as good as CM hyper 212 and have thick copper plate.

            Comment


            • #16
              Originally posted by edwaleni View Post
              I finally got my Raven Ridge running (2400G) but I had to wait until an A-Series arrived so I could flash the BIOS for RR support.

              ....

              But this being a very new CPU, the board support appears to be somewhat immature even though the socket is the same.
              Could you mention the board model? Just to know what board you're talking about.

              Comment


              • #17
                We are getting there, if it wasnt for Meltdown i suspect Racen Ridge would already be stable.

                In any event im running an HP X360 with Raven Ridge and have bee trying various distros via a USB thumb drive. Ubuntu 18.04 is starting to look much better. Ive been running it for about a wekk and last night i did a complete upgrade. After the upgrade ubuntu is running much better and this is from a testing thumb drive.

                Now much better doesnt mean perfect!!!!!!! However id call it usable where as almost nothing else ive tried works worth a damn. This includes multiple attempts with Rawhide and other bleeding edge distros. So if Raven Ridge is in your machine the only thing i can suggest at the moment is the absolute most recent Ubuntue 18.4 series.

                By the way folks, the HP running Windows 10 has been getting constsnt updates. This includes stuff from MS and constant updates from HP for both the AMD chips and other supporting chips. Some of the updates have had a big impact so dont be thinking Linux is that far behind.

                Comment


                • #18
                  Originally posted by starshipeleven View Post
                  Could you mention the board model? Just to know what board you're talking about.
                  My bad. I meant to include it.

                  MSI B350M PRO-VDH

                  I already pinged MSI to request a change in their BIOS so that it can boot into safe mode with an older AGESA so it can be flashed. By the fact the BIOS had to reboot the board twice to recognize the Raven Ridge install makes me think they can't do it. The boot sequence on the A-Series is very quick. The sequence under the Raven Ridge is much longer before you see a screen. Watching the boot monitor, it turns on everything, stops the CPU briefly (along with the CPU fan), then restarts it, then completes.

                  If I didn't know better I would say the BIOS checks to see if it is a A-Series first, stops and restarts it as a Ryzen. MSI does warn you that after the RR upgrade, UEFI will take awhile to load on first boot.

                  In one case after resetting the CMOS, I couldn't get video on the VGA port. After some tweaks UEFI came up finally, but was full of artifacts and wouldn't accept changes. I finally had to dig out of storage an old Radeon HD4550 PCIe and boom, everything worked and UEFI went back to defaults. In case anyone asks, the NVidia card did not work in this circumstance.

                  As far as I can tell, its telling me something is on IRQ7 that can't be shared and the Vega doesn't like it. Also of note, the Vega is not an IGP on Raven Ridge. It is attached via an internal 16x lane in the SOC. Also something I haven't seen before, UEFI has a cryptomining above 4Gb option.

                  Comment


                  • #19
                    Originally posted by Lanz View Post

                    I'm in the market for a Ryzen APU notebook this year. What distribution and kernel/Mesa are you using, and can you comment on what does/does not work and what was difficult to make work?
                    im on manjaro kernel 4.16 git, mesa git
                    what works: everything, but some things like games have HEAVY and frequent framerate drops, that might be because im on llvm 5.0
                    there are also some system hangs on both linux and windows

                    Comment


                    • #20
                      Originally posted by rene View Post
                      so far no hangs nor display corruption with an Lenovo ideapad 720s w/ AMD Ryzen 5 2500U with Radeon Vega Mobile Gfx: https://www.youtube.com/watch?v=utviLZ5c9SE
                      Does this device wake up fine from suspend-to-RAM aka ACPI S3 sleep?
                      (Should you not know how to test: Just type "echo mem >/sys/power/state" in a terminal console, wait until the power LED blinks, then press any or the "function" key to resume.)

                      Comment

                      Working...
                      X