Maintaining a monolithic kernel like Linux is an ongoing balancing act: integrating new features without burdening the codebase with components nobody uses anymore. With version 7.2, developers have turned the scalpel on a driver that dates back to a prehistoric era of computing: support for PROFIBUS, the fieldbus born from 1990s industrial automation. The offending code was ported from SCO Unix in 1998 and, since then, had received no significant updates. No active user has been identified on modern kernel branches, and so the driver has been shredded together with two other obsolete components removed from the char/misc subsystem.

A technology that made industrial history

PROFIBUS (Process Field Bus) is an internationally standardized serial communication protocol, widely used to connect sensors, actuators, and programmable logic controllers (PLCs) in factories. Developed in the late 1980s and ratified as EN 50170, it dominated manufacturing plants for decades. Its presence in the Linux kernel was a relic of when the open-source system began to be adopted as an embedded alternative in heavy industry. But the particular driver in question, for PROFIBUS network cards, had been stuck on code over twenty-five years old, unable to adapt to modern kernel interfaces and with no developers willing to maintain it.

Necessary cleanup for kernel health

The Linux community has well-tested mechanisms for deciding obsolescence: a driver without maintenance, with no known users on current kernel branches, gets flagged for removal after a notice period. The PROFIBUS case is emblematic because it combines two factors: very old code and no signs of life from the real world. This isn't a whim but a hygiene measure that reduces the attack surface (unchecked code can hide latent vulnerabilities) and lightens the load on maintainers, who must ensure compatibility with new kernel versions. Every line of code removed is a line that doesn't need to be tested, reviewed, or adapted to architectural changes.

What it means for those managing on-prem machines

In industrial contexts and on-premise environments, where machines with a decade or more of service often survive, the removal of legacy drivers can create friction. Those still using PROFIBUS hardware on Linux face a fork in the road: stick with outdated kernels (perhaps customized), isolate the machine from the network, or migrate to newer solutions like PROFINET. It's the classic trade-off between operational stability and technological upgrade that every infrastructure manager knows. On one side is the cost and risk of touching a working system, on the other the need to receive security patches and keep the operating system up-to-date. The news of the removal, though unnoticed by most, is a wake-up call for those planning the longevity of their plants: monitoring upstream support of the drivers in use is not optional.

Kernel evolution and lessons for on-prem

Linux 7.2 won't be remembered for this single pruning, but every trim helps keep the kernel healthy and long-lived. For the on-premise world – not just industrial but also private data centers, edge computing, and air-gapped environments – the message is clear: dependencies on orphaned hardware components represent technical debt that must eventually be paid. Those evaluating local deployments should include driver abandonment risk in their Total Cost of Ownership, opting for hardware with active upstream support and known maintenance timelines. Similarly, for those inheriting existing infrastructure, it pays to run periodic audits of the kernel modules in use, leveraging configuration analysis tools to identify at-risk modules. This isn't just about security: it's a strategy of technological sovereignty, where the ability to update and control the software stack also depends on the vitality of community support.

The removal of the PROFIBUS driver is a small note in the kernel's lengthy release notes, but it tells a larger story: open-source discipline knows when it's time to turn the page, even at the cost of leaving behind some pieces of computer archaeology. For those building the future of on-prem IT, it's a reminder to take nothing for granted.