Trying Out The New OpenGL Threaded Dispatch In Mesa 17.1

[anth]

This change is aimed at easing bottlenecking on a single CPU thread, so testing with an older or underclocked CPU might provide some interesting results.

[ConvexEd]

The Borderlands sounds fantastic. I don't know why, but I can barely run it at high frames with a 3570k @ 4.3GHz and a 970 using proprietary drivers.
Maybe I'll just go with AMD and get one of their rebrands this generation if they're priced well.

[grigi]

Michael, I posted my experience with glthread and hitman here: https://www.phoronix.com/forums/foru...112#post939112

A significant improvement in playability, even though there is no real FPS boost.
I may only be seeing it like this because of how underpowered my GPU is in relation to my CPU. (alas 4 1/2 year old notebook)

[Helios747]

The person in this thread who had performance gains had an Intel CPU. And it does appear (At least with our experience with Dolphin) that AMD is having some thread scheduling issues with Ryzen on current Linux kernels and Windows.

See if there's any perf improvements on Intel.

[LeJimster]

The person in this thread who had performance gains had an Intel CPU. And it does appear (At least with our experience with Dolphin) that AMD is having some thread scheduling issues with Ryzen on current Linux kernels and Windows.

See if there's any perf improvements on Intel.

The scheduler issue, at least on the windows side has been largely debunked. However there is an obvious problem on Linux right now. The benchmarks from phoronix and another Linux user that I've seen so far show a bigger drop off in gaming performance compared to windows.

Really strange because the multi threaded performance of Ryzen is really good, even better than intel in nearly all other scenarios.

[L_A_G]

That's rather interesting. maybe something is not working properly?

It's probably conflicting with games trying to do the same thing themselves. I guess it's a good thing the way it's supposed to work is with a whitelist of games that benefit from it and not even try to use it on games that aren't on that whitelist.

[LinuxID10T]

The scheduler issue, at least on the windows side has been largely debunked. However there is an obvious problem on Linux right now. The benchmarks from phoronix and another Linux user that I've seen so far show a bigger drop off in gaming performance compared to windows.

Really strange because the multi threaded performance of Ryzen is really good, even better than intel in nearly all other scenarios.

This is because all these Linux games are ported to OpenGL from DX for the most part. OpenGL isn't as well threaded therefore it doesn't show very good multicore scaling. With Intel having fewer, higher clocked cores even if the IPC is rather similar they get better frame rates (especially for OpenGL.)

[sp82]

I don't believe anymore to this tests. I want to see the frametime graph and measuring the minimum fps not the average fps. The best experience is when the minimum fps do not go below the monitor refresh rate and the input latency is low.
In this new multi-threaded implementation should be interesting to see the minimum fps if changed.

[Michael]

I don't see a Borderlands 2 benchmark here.

As said in the article, not aware of any working benchmark mode for it.

[marek]

I'd say that having the same frame rate with and without glthread on low settings is fairly unlikely. I don't know what went wrong there, whether steam didn't pick up the environment variable or whether PTS wiped out the environment.

Instead of messing with environment variables, you can create ~/.drirc with the following:

<driconf>
    <device>
        <application name="Default">
            <option name="mesa_glthread" value="true" />
        </application>
    </device>
</driconf>

If you already have ~/.drirc with something in it and you don't know why, remove that, because generally ~/.drirc shouldn't mess with default Mesa settings in any way.