Quote of the week
While there are many proponents of "eBPF is good for everything and your grandma" approach, this opinion is not universally shared. One big risk is that this will eventually lead to possibility of having whole drivers / core code written in eBPF, which could potentially lead to decreased maintainability and supportability, also due to big fragmentation of the code (eBPF programs might not necessarily be shipped together with the kernel codebase).— Jiri KosinaThis could potentially be a big risk for distros as well, as we (as a distro vendor) might be very quickly losing control over what is actually running in the context of the kernel they are bound to be supporting.
Posted Jun 17, 2022 14:57 UTC (Fri)
by paulj (subscriber, #341)
[Link] (2 responses)
Posted Jun 25, 2022 22:43 UTC (Sat)
by landley (guest, #6789)
[Link] (1 responses)
Ignoring the previous attempts at microkernels and javaos and such, this is pretty much the same problem I have with rust in the kernel. You want to write a whole new system in rust/go/swift, fine. You want to spread like a cancer through the existing immunoprivileged environment introducing domain crossings and requiring maintainers to understand both sides intimately to get the big picture, I have concerns.
I file "the new stuff added will eventually completely replace the existing stuff" under https://xkcd.com/927/
Posted Jun 26, 2022 1:07 UTC (Sun)
by rahulsundaram (subscriber, #21946)
[Link]
I do have some good news for you, it doesn't spread like cancer. Maintainer have to explicitly take actions to accept the code.
Quote of the week
Quote of the week
Quote of the week
