Got this issue on a Radeon 290 with AMDGPU drivers. Works fine with kernel 4.14 but not 4.15. Here are some pictures of it.
https://imgur.com/a/4wMQ5

@stigarn Intel CPU as well?

Yes using Core i5 4690k

Disabling Intel IOMMU (intel_iommu=off)

If this fixes it, we'll have to look into why. So far this looks like an interaction issue between the IOMMU (which is responsible for things like PCI-E) and the AMDGPU driver.

If anyone sees this task and has a non-Intel CPU and is experiencing the same rendering problems, please let us know

The only relevant thing I could find was this PDF mentioning DMA remapping faults for the Xeon mentioned in the inxi output. But I'm not using integrated graphics so I'm not sure how that'd be related. https://www.thailand.intel.com/content/dam/www/public/us/en/documents/specification-updates/xeon-e3-1200v5-spec-update.pdf

Also, I have a Core i5, not a Xeon. Maybe they have that bit of hardware in common?

Strongly suspect this is an AMDGPU regression and we need to enable CONFIG_DRM_AMD_DC_PRE_VEGA for your POLARIS.

I don't know if it helps, but seeing I have an AMD POLARIS card on Intel, and have had no rendering issues, I thought it worth mentioning. Apparently it's not affecting all.

~  inxi -CGS                                                                              1062ms  Tue 20 Feb 2018 09:33:04 AM MST
System:    Host: mirkwood Kernel: 4.15.3-53.current x86_64 bits: 64 Desktop: KDE Plasma 5.12.1 Distro: Solus 3
CPU:       Quad core Intel Core i5-3470 (-MCP-) cache: 6144 KB
           clock speeds: max: 3600 MHz 1: 1596 MHz 2: 1596 MHz 3: 1596 MHz 4: 1596 MHz
Graphics:  Card: Advanced Micro Devices [AMD/ATI] Lexa PRO [Radeon RX 550]
           Display Server: x11 (X.Org 1.19.6 ) drivers: ati,amdgpu (unloaded: modesetting,fbdev,vesa,radeon)
           Resolution: 1920x1080@60.00hz
           OpenGL: renderer: Radeon RX 550 Series (POLARIS12 / DRM 3.23.0 / 4.15.3-53.current, LLVM 5.0.1)
           version: 4.5 Mesa 17.3.4

It's a System76 Wild-Dog, "built to run Linux", so I don't know if they do something in the way the system is put together that negates the problem.

And I'm using HAWAII.

➜ ~ inxi -CGS
System: Host: solus Kernel: 4.14.18-51.current x86_64 bits: 64 Desktop: Budgie 10.4 Distro: Solus 3
CPU: Quad core Intel Core i5-4690K (-MCP-) cache: 6144 KB clock speeds: max: 3900 MHz 1: 1872 MHz 2: 1676 MHz 3: 2514 MHz 4: 2482 MHz
Graphics: Card-1: Intel Xeon E3-1200 v3/4th Gen Core Processor Integrated Graphics Controller
Card-2: Advanced Micro Devices [AMD/ATI] Hawaii PRO [Radeon R9 290/390]
Display Server: x11 (X.Org 1.19.6 ) drivers: intel,amdgpu (unloaded: modesetting,fbdev,vesa)
Resolution: 1920x1080@60.00hz, 1920x1080@60.00hz
OpenGL: renderer: AMD Radeon R9 200 Series (HAWAII / DRM 3.19.0 / 4.14.18-51.current, LLVM 5.0.1)
version: 4.5 Mesa 17.3.4

I was going to boot from the previous kernel again and get the dmesg from there to compare, but it's not in my boot menu anymore. Would that be helpful? If so, how do I get clr-boot-manager to show the last kernel again, or has it been removed from the system now?

Just updated my system and tried kernel Linux solus 4.15.5-56.current #1 SMP Fri Feb 23 14:11:51 UTC 2018 x86_64 GNU/Linux and the problem still exists.

@stigarn, have you tried disabling intel_iommu? It works for me, as without it the problem still occurs on that latest kernel.

@ikey, missed your comment earlier. I tried CONFIG_DRM_AMD_DC_PRE_VEGA=on just now (without intel_iommu disabled) and it didn't seem to make a difference?

@yaomtc Works fine for me too when I disabled that. So using latest kernel disabling intel_iommu solves the issue.

Still broken in latest kernel: Linux solus 4.15.7-60.current #1 SMP PREEMPT Thu Mar 1 13:02:55 UTC 2018 x86_64 GNU/Linux

Still broken
Linux solus 4.15.11-61.current #1 SMP PREEMPT Mon Mar 19 16:15:42 UTC 2018 x86_64 GNU/Linux

Broken in plasma too T6082

Same for linux-current-4.15.15-63 on an AMD Radeon R9 M370X with an Intel i7-4870HQ processor.

STILL broken in Linux solus 4.15.17-65.current #1 SMP PREEMPT Fri Apr 20 11:36:00 UTC 2018 x86_64 GNU/Linux Can someone please look at this. This was reported two months ago now.

They are a small team and this issue doesn't seem to affect many people, so I'm guessing it's not the highest priority, that's OK. Also, we're not paying. Well, I'm paying them a few bucks on Patreon but that doesn't mean anything. Of course I'm sure they'd like to fix it, but this is one open issue out of many, it has a quick workaround, and they've got to choose how they spend their time.

Regarding the workaround, appending intel_iommu=off to GRUB_CMDLINE_LINUX_DEFAULT in /etc/default/grub and running update-grub does not solve the issue for me. Am I missing something or is that workaround only for certain GPUs?

Well unlike me and Stigarn (we both use the i5-4690k) you have an i7. It might help to name the specific model.

In T5867#113321, @Siilwyn wrote:

Regarding the workaround, appending intel_iommu=off to GRUB_CMDLINE_LINUX_DEFAULT in /etc/default/grub and running update-grub does not solve the issue for me. Am I missing something or is that workaround only for certain GPUs?

Try this
echo 'intel_iommu=off' | sudo tee /etc/kernel/cmdline sudo clr-boot-manager update

Solus adds kernel parameters in a different way.

Thank you @stigarn that worked! Now I can stop unchecking linux-current from the updates.

Tried with latest updates and still broken.

On a fresh install of the latest Plasma ISO, the problem didn't occur, but after I ran through the updates I was reminded what this bug looks like.

Still broken in 4.18

This is even more broken in 5.0.7-114 tested without intel_iommu=off and now it hangs on boot. I just see the mess in the picture but can't do anything but hard reset the box. Not even ctrl+alt+f1 takes me to cli.

I don't have Intel anymore, so I can't test, sorry.