Open-Source "Terakan" Vulkan Driver For Radeon HD 6000 Series Shown On Windows

[phoronix]

Phoronix: Open-Source "Terakan" Vulkan Driver For Radeon HD 6000 Series Shown On Windows

The past year there's been an independent open-source driver developer working on "Terakan" as a Vulkan driver for old Radeon HD 6000 series GPUs. These pre-GCN GPUs never received any official Vulkan driver support from AMD but thanks to open-source and a strong desire to pull off such a feat, Vitaliy Kuzmin "Triang3l" has been pursuing this challenge and has been pulling off some basic results. The work so far has been predominantly been carried out with the open-source Linux graphics stack while this weekend the Terakan driver was demonstrated under Microsoft Windows...

https://www.phoronix.com/news/Terakan-HD-6000-VLK-Windows

[dragorth]

There are lots of side effects to such a thing, for instance, could Zink run on this and offer up to date OpenGL on it? For low demanding applications, such as Older Windows Games, Emulation software, and older machines, this puts it in the running to allow things like Wayland to do basic Vulkan displays.

However, Mesa is something that is used on more than just Linux. In fact, it is used when porting Linux GPU DRM drivers in other OSes. Things like NetBSD and FreeBSD have compatibility layers that take these drivers and allow these devices to work on their systems.
GenodeOS does the same thing, and other projects, such as AmigaOS 4 has drivers based on this, and MorphOS are in the process of adding drivers for this class of card.

[NateHubbard]

There are lots of side effects to such a thing, for instance, could Zink run on this and offer up to date OpenGL on it? For low demanding applications, such as Older Windows Games, Emulation software, and older machines, this puts it in the running to allow things like Wayland to do basic Vulkan displays.

Maybe there is some particular case where you'd want to run something that is both newer and less demanding at the same time, but it usually doesn't work that way.

[dragorth]

Maybe there is some particular case where you'd want to run something that is both newer and less demanding at the same time, but it usually doesn't work that way.

I gave the simplest example of window managers already, do you want more? Font rendering in text editors, ebook readers, which frankly text reading is the majority occupation of most on a desktop system, old school games and emulators, word processors, spreadsheets, most office work in fact, and all of these are just on Linux.

On MorphOS, they are stuck with much older cards than these currently, currently stuck at the ATI X1900 line of PCI-E cards as the latest stable cards, which is why they are porting some drivers and Mesa code to enable cards of this caliber.

This will simply allow them to do crazy things like use 4K displays on Power Mac G5s and have some hardware decode for their Wayfarer Browser based off the latest Chromium releases.

There are whole businesses that run on hardware that is less powerful than this.

[Developer12]

So many people seem to have caught michael's terminal benchmarker-brain syndrome.

Vulkan is rapidly becoming the *only* graphics API used on linux. Not just for high-spec games but for everything from window managers to gui toolkits and from there to regular old productivity tools, misc config menus, and random stuff like that. SDL is used for a lot of random boring apps, not just games. What happens if/when they deprecate openGL entirely? That's totally on the table for the next 5-10 years.

Most of the stuff the average user runs on a linux desktop is perfectly capable of being run with *software rendering* if push comes to shove, but when more and more stuff is deprecating the openGL backend a vulkan implementation becomes mandatory. At that point anyone with this older hardware has to choose between this GPU-accelerated vulkan solution and the vulkan equivalent of LLVM-pipe. Clearly, having an implementation of vulkan with at least *basic* GPU acceleration is preferable.

[NateHubbard]

I gave the simplest example of window managers already, do you want more?

Right, what I mean though is that those cards already supported OpenGL 4.5 anyway. Maybe you'd find something that requires 4.6, but isn't a modern game or something, but it just doesn't seem likely to me.
I suppose a compositor running vulkan makes the most sense.

[Daktyl198]

I really wish that some effort was made to properly port Mesa drivers to Windows. Specifically AMD drivers. I know Mesa's architecture allows for it to run on Windows, but nobody is bug testing and producing prebuilt installers for them.

[dragorth]

Right, what I mean though is that those cards already supported OpenGL 4.5 anyway. Maybe you'd find something that requires 4.6, but isn't a modern game or something, but it just doesn't seem likely to me.
I suppose a compositor running vulkan makes the most sense.

You are making the assumption that the current drivers will continue to work, which isn't a good assumption to make, really.

I have recently been watching a coder on Twitch who has been working on his T2SDE distro that still supports Itanium and Alpha and lots of other platforms, and the biggest takeaway I have from someone actually maintaining a distro is upstream is constantly breaking older hardware, because it isn't tested, it is just assumed it works.

Getting everything onto one common code path can help with that to some extent, but the reality is, the kernel devs are always chasing the new shiny. I appreciate it, but I also can't use the latest version of software on my Power Mac G4 MDD anymore, because the drivers and software are not up to snuff.

When you have 100 times or more people making changes to upstream, the people maintaining older hardware really can't keep up.

And frankly, I want older hardware to be usable. It is easier to hack on it, it is easier to learn how to design parts for it, it is easier to boot from it.

[rene]

certainly shows how inefficient and lazy big mega cooperations are, ...

[rene]

You are making the assumption that the current drivers will continue to work, which isn't a good assumption to make, really.

I have recently been watching a coder on Twitch who has been working on his T2SDE distro that still supports Itanium and Alpha and lots of other platforms, and the biggest takeaway I have from someone actually maintaining a distro is upstream is constantly breaking older hardware, because it isn't tested, it is just assumed it works.

Getting everything onto one common code path can help with that to some extent, but the reality is, the kernel devs are always chasing the new shiny. I appreciate it, but I also can't use the latest version of software on my Power Mac G4 MDD anymore, because the drivers and software are not up to snuff.

When you have 100 times or more people making changes to upstream, the people maintaining older hardware really can't keep up.

And frankly, I want older hardware to be usable. It is easier to hack on it, it is easier to learn how to design parts for it, it is easier to boot from it.

oh, yeah, this guy, I think I watched him, too ;-) https://www.twitch.tv/rxrbln/clip/Ge...lXWf14iWl6TH8Y - this https://t2sde.org is also pretty cool, run on all my collection of vintage & retro hw, too, ... ¯\_(ツ)_/¯