Twice now when I come back to my new Framework 13 Ryzen running Fedora 39 beta, the screen will blank/flicker after resuming from sleep. It alternates between the whole screen whiting out or just half, in a strange blocky pattern. Happens both with the built-in screen as well as with an external monitor.
See videos: https://photos.app.goo.gl/FsZX6KpW2aWXtL9j8
Note: changing the refresh rate from 60hz to 30hz doesn’t help.
How much RAM do you have and which kernel version? It sounds a bit like the 64Gb+ issue described here: What DRAM/memory is supported by Framework Laptop 13 (AMD Ryzen™ 7040 Series)?
32 GB of RAM (purchased from Framework, 1 SODIMM)
Kernel version 6.5.7-300.fc39.x86_64
Hello, I can confirm this is happening in Fedora 39 (stable) on a framework 13 AMD.
I have 16gb of ram and I am on kernel 6.5.12-300.fc39.x86_64
Had the same problem and could solve it by changing a VRAM setting in the BIOS: [RESPONDED] AMD Ryzen 7040 (7840U) - Arch Linux amdgpu errors, blank screen on opening Steam - #12 by Remy_Garrigue
I’ve had the same thing happened to me twice. I already have UMA_GAME_OPTIMIZED set. This is with Fedora 39 and 32 gigs of RAM from Framework.
I have the same issue, i had this running arch and now with fedora.
It makes the sleep function unusable as it happens everytime on wakeup.
When this issue happens the following errors are constantly repeated in the dmesg log
[37057.877500] amd_iommu_report_page_fault: 31522 callbacks suppressed
[37057.877515] amdgpu 0000:c1:00.0: AMD-Vi: Event logged [IO_PAGE_FAULT domain=0x0005 address=0xfffcc9f0000 flags=0x0000]
[37057.877529] amdgpu 0000:c1:00.0: AMD-Vi: Event logged [IO_PAGE_FAULT domain=0x0005 address=0xfffcc9f1000 flags=0x0000]
[37057.877538] amdgpu 0000:c1:00.0: AMD-Vi: Event logged [IO_PAGE_FAULT domain=0x0005 address=0xfffcc9f2000 flags=0x0000]
[37057.877546] amdgpu 0000:c1:00.0: AMD-Vi: Event logged [IO_PAGE_FAULT domain=0x0005 address=0xfffcc9f3000 flags=0x0000]
[37057.877555] amdgpu 0000:c1:00.0: AMD-Vi: Event logged [IO_PAGE_FAULT domain=0x0005 address=0xfffcc9f4000 flags=0x0000]
[37057.877566] amdgpu 0000:c1:00.0: AMD-Vi: Event logged [IO_PAGE_FAULT domain=0x0005 address=0xfffcc9f5000 flags=0x0000]
[37057.877576] amdgpu 0000:c1:00.0: AMD-Vi: Event logged [IO_PAGE_FAULT domain=0x0005 address=0xfffcc9f6000 flags=0x0000]
[37057.877585] amdgpu 0000:c1:00.0: AMD-Vi: Event logged [IO_PAGE_FAULT domain=0x0005 address=0xfffcc9f7000 flags=0x0000]
[37057.877593] amdgpu 0000:c1:00.0: AMD-Vi: Event logged [IO_PAGE_FAULT domain=0x0005 address=0xfffcc9f8000 flags=0x0000]
[37057.877603] amdgpu 0000:c1:00.0: AMD-Vi: Event logged [IO_PAGE_FAULT domain=0x0005 address=0xfffb7045000 flags=0x0000]
When this happens i can access a tty console with having this issue.
The complete dmesg log can be found here
I recently tried setting amdgpu.sg_display=0 as it showed up as a solution to similar problems. I haven’t seen the white flicker screen since and it has been several days. I’m new to Linux though so not sure if this is on the right track.
I can confirm the same problem on Arch Linux.
Today, I’ve installed the new Linux kernel 6.7.0 (Manjaro) with various improvements for AMD.
As of now, the flickering seems to be gone.