Linux Kernel: OpenZFS and Fake Frames

posted by Roy Schestowitz on Jul 13, 2025

• Kernel Space

  • Michael Kjörling ☛ The case of OpenZFS and the variable name mixup

    Look at the code in question for a moment. Forget everything about who wrote it and what software it is part of. The function, named vdev_raidz_asize_to_psize, as one might expect uses an input variable named asize (along with some other data) to calculate an output termed psize. This in itself is not a strange pattern: a function to encapsulate some fairly complex math which is needed in different places, giving the calculation a name and a single source of truth.

    But after going to all that trouble, the return statement at the end returns asize to the caller. But asize was the input.

    Oops.

  • University of Toronto ☛ The development version of OpenZFS is sometimes dangerous, illustrated

    I've used OpenZFS on my office and home desktops (on Linux) for what is a long time now, and over that time I've consistently used the development version of OpenZFS, updating to the latest git tip on a regular basis (cf). There have been occasional issues but I've said, and continue to say, that the code that goes into the development version is generally well tested and I usually don't worry too much about it. But I do worry somewhat, and I do things like read every commit message for the development version and I sometimes hold off on updating my version if a particular significant change has recently landed.

  • Rob Norris ☛ An (almost) catastrophic OpenZFS bug and the humans that made it (and Rust is here too)

    Now, I made it a bit easy for you by showing you the troublesome function. I actually found it while testing a new feature I’m working on for a client, with some pretty aggressive allocator fragmentation settings in place, and I started getting weird errors, and wondered what I’d broken. It took almost two days to track it down.

    Now in a lot of ways, it’s no big deal. Bugs happen, as do near-misses. It had been on the main development branch for a couple of months, but never on a real release.

• Graphics Stack

  • Matt Birchler ☛ Fake frames?

    As some in the video game community would say, this is a “fake frame” from the game. Why? Well, the screenshot is 4k, but I was playing with DLSS enabled, so this is actually more like a 1440p (or maybe even 1080p) frame that was upscaled in real time to 4k. Also, I was using frame generation which doubled my frame rate, so there’s a 50% chance this frame technically wasn’t even rasterized at all by the GPU. Here’s a few more screenshots, all of which could be “fake frames” so the odds are quite high that at least one of these is a dirty fake frame.