linux-kernel - Desktop responsiveness degradation over uptime despite idle system

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [day] [month] [year] [list]

Message-ID: <4lGQjffnwMTshiIyWReUpga_OuxTtTf2itjY9C-SqqWpnFKo5itBUraov58lTgRULiU2r8jQ0twcLjgxoHnNYc4DsaxozTh8-GNMBQscltA=@protonmail.com>
Date: Thu, 04 Dec 2025 13:34:38 +0000
From: Zone <getzone@...tonmail.com>
To: "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Desktop responsiveness degradation over uptime despite idle system

Hi all,

I've been noticing a pattern on my desktop that I'm curious whether others have observed or have insights into.

After a fresh reboot, desktop responsiveness is excellent, window operations, application switching, and general UI interactions all feel snappy. But as uptime increases into weeks, this responsiveness gradually degrades. What starts as a crisp experience slowly turns visibly sluggish.

The puzzling part: when I notice this degradation, the system is essentially idle. CPU utilization is minimal, I'm not actively doing anything demanding and also the various DE components are idle.

I've asked a few friends who also run Linux desktops if they've noticed the same thing, and they have. We're running different DEs, I'm on GNOME, others are on KDE or lighter-weight setups, and we're all on modern, reasonably high-spec hardware with ample RAM. So it doesn't seem DE-specific.

My suspicion is that this is related to browser usage. I'm a heavy desktop user, and I tend to run multiple browser instances (Brave and Firefox) with a lot of tabs open. But here's the thing: I actively use tab suspension to keep idle tabs from consuming resources, and the slowdown is noticeable even when the browsers are just sitting there idle. More importantly, I've tried fully exiting all browser instances when responsiveness gets bad, and it doesn't help. The sluggishness persists. The only thing that clears it is a full reboot.

That makes me think some sort of debt is accumulating over time and does not change when the browsers exit. Memory fragmentation? some kernel resource that's not being reclaimed? something in the mm path? I really don't know.

I'm not reporting this as a definitive bug, more just trying to understand if this is a known pattern or something worth investigating further. If there's debugging info that would be useful to collect, I'm happy to do that.

Kernel: 6.17 Dist: Ubuntu 25.10 (up to date) DE: GNOME Browsers: Brave, Firefox

I regularly update kernels when major releases happen, and I do notice that kernels as of late have improved the responsiveness and snappiness of the DE of a freshly booted machine substantially, however, over time the situation I’m seeing has been happening for many years across different hardware.

Is anyone else noticing this? It would really be great to squash this one.

Cheers