Any recommendations for improving performance on Slackware current?
SlackwareThis Forum is for the discussion of Slackware Linux.
Notices
Welcome to LinuxQuestions.org, a friendly and active Linux Community.
You are currently viewing LQ as a guest. By joining our community you will have the ability to post topics, receive our newsletter, use the advanced search, subscribe to threads and access many other special features. Registration is quick, simple and absolutely free. Join our community today!
Note that registered members see fewer ads, and ContentLink is completely disabled once you log in.
If you have any problems with the registration process or your account login, please contact us. If you need to reset your password, click here.
Having a problem logging in? Please visit this page to clear all LQ-related cookies.
Get a virtual cloud desktop with the Linux distro that you want in less than five minutes with Shells! With over 10 pre-installed distros to choose from, the worry-free installation life is here! Whether you are a digital nomad or just looking for flexibility, Shells can put your Linux machine on the device that you want to use.
Exclusive for LQ members, get up to 45% off per month. Click here for more info.
Any recommendations for improving performance on Slackware current?
Hello, I'm currently running Slackware current on my gaming machine, it has a Ryzen 7 3500U and AMD Picasso GPU. I'm using the mesa and vulkan drivers along with the 5.18 kernel. What more could be done to squeeze a bit more performance out of my machine, a custom kernel, messing with Xorg configs?
At the moment I'm running KDE-plasma on my Slackware64-current laptop. If you wanted to use less system resources you could use a light window manager like Fluxbox. A suggestion.
If you are using a default kernel, afaik, whether generic or huge, you could benefit from a custom kernel. How much you could benefit would depend on how well you know all your hardware. Changing settings to your specific CPU type as opposed to generic, timing, pre-emptible (low latency) all make differences a person involved in heavier workloads, especially mulktimedia work, can feel as well as see in benchmarks.
As then list of supported hardware only increases since extremely old hardware still enjoys support and even though that is less a load than it used to be back before on-demand modules, there is a reason embedded systems can be so fast - tiny kernels only supporting what's actually there. Figuring out every chep your PC might have in 2022 is a massive job, one I stopped considering with 2.4 kernels.
Is it possible to compile already made custom build kernels like Xanmod from source, using the make oldconfig route? If not, I have no issue making one from the latest builds.
But as always there are differing opinions on the above.
Changing to a lighter window manager and building a custom kernel with make localmodconfig and make localyesconfig?
I'd still go through the kernel config with make menuconfig anyway.
Definitely use wired ethernet instead of wireless and modify the home router with better QOS marking for the gaming application.
A long time ago and on the home router, I gave the kids static IPs with 50% bandwidth while reserving the remainder for the better half and I ( ͡~ ͜ʖ ͡°)
Is messing around with the real-time scheduler still worth it with the new kernels?
The RT patchset won't necessarily get you better "performance" but it will help with interactivity. Its real purpose is to provide more predictable (and lower) latencies through preemption and priority. More kernel processes are preemptible too, with some locks converted to mutexes and stuff. Even soft interrupts are preemptible.
A prime example of something a desktop user might like is being able to, say, give something more priority and actually have the scheduler bias adhere to that instead of being all "fuzzy" about it.
I use it on my manjaroo system where I play games. Why? Mostly because giving pulseaudio RT priority through rtkit-daemon actually works with the RT kernel. I was able to remove all workarounds for audio crackling and the like from all my games. I no longer need PULSE_LATENCY_MSEC= (induces latency by buffering behind the number of milliseconds) or wine dll overrides to work around poor audio behaviour.
That's really the only benefit I can qualify other than saying my games are all performing well. It would not directly lead to "higher framerates" but possibly help with situations where there might be brief hitching or something.
Another thing... let's say I've got a big assed LTO build of Firefox or something going, niced +10. I can't even bloody tell I'm compiling. I once forgot I even had a build going on another virtual desktop and quit my session. Oops lol.
For most usage, the normal kernel with preempt=full should be good.
LinuxQuestions.org is looking for people interested in writing
Editorials, Articles, Reviews, and more. If you'd like to contribute
content, let us know.