ib_unregister_device(9)

πŸ“¦ version 7.3.1, in nixpkgs unstable

πŸ’… ib_unregister_device(9) πŸ’…

okay so bestie, we're in the Kernel API Manual today and this page is literally about saying bye to an IB device πŸ‘‹βœ¨


πŸŽ€ NAME πŸŽ€

ib_unregister_device β€” unregister an IB device (yes that's the whole vibe, she's a one-liner and she ate πŸ’…)


✨ SYNOPSIS ✨

void ib_unregister_device (struct ib_device *ib_dev );

like, that's it. one function. no return value. she just does the thing 🫢


🌸 ARGUMENTS 🌸

  • ib_dev β€” the device you're unregistering, aka the one getting the goodbye kiss πŸ˜­πŸ’–

πŸ¦‹ DESCRIPTION πŸ¦‹

So you want to unregister an IB device. Iconic. Here's the tea:

Unregister an IB device. All clients will receive a remove callback. every single one of them, no cap, they all get notified πŸ“£

⚠️ bestie listen to me: callers should call this routine only once, and protect against races with registration. one time. periodt. and do the race protection, please, i'm begging πŸ™

Typically it should only be called as part of a remove callback in an implementation of driver core's struct device_driver and related. that's the main character usage right there πŸ‘‘

And the spicy part 🌢️ β€” if ops.dealloc_driver is used then ib_dev will be freed upon return from this function. like, she's gone gone. don't touch her after this. not even a little. πŸ’€


πŸ”₯ SEE ALSO πŸ”₯

the whole friend group showed up fr fr, look at this roster:

and the source file lives at ./drivers/infiniband/core/device.c β€” go read her, she's giving main character energy πŸ“–πŸ’œ


🌈 mwah πŸ’‹ this page came from ./drivers/infiniband/core, dated October 2026, and she's filed under ib_unregister_device(9). that's the whole glow-up, love u, bye πŸŽ€πŸ«Άβœ¨

πŸ“¦ where this page comes from