Re:The Helios microkernel
by Santosh83 on Thursday June 16, 2022 @03:41AM
(#31763454)
Linus is not a great believer in the microkernel architecture. Quite apart from that while certain things like FS can be shifted to userspace through FUSE, a ground-up rearchitecting into a microkernel may be extremely difficult for Linux at this stage. It would be more or less a rewrite of most of the core kernel I guess... and then the question of changes to the syscalls, which also Linus doesn't encourage.
[ Parent
| on HN ]
Generally his skepticism was driven by 2 things around the time the microkernel debate happened:
- Microkernels need more development effort for equal functionality, something that was a constraint when he was doing Linux on his own.
- The context switching penalties on 386s were ginormous - it's much more reasonable now, I'd even say it's not much more penalizing than thread switching.
- We live in a multi-master world with IOMMU - tons of devices have virtual address spaces on their own, with most of them doing DMA transfers.
What Linus believes in or does not believe is hardly relevant though, what matters is what is the best solution going forward. Sunk costs are always smaller than the future costs.
Re:The Helios microkernel
by CameronNemo on Thursday June 16, 2022 @01:18PM
(#31768568)
Linus was not a believer something like 20 years ago. I wouldn't be surprised if his stance has significantly softened.