Single address space in Long mode

[Combuster]

No, it works with multiple CPUs, multiple address spaces and multithreading.

[jal]

but that means the scheduler must update the MSR when switching tasks.

Not necessarily. You can just have it contain a pointer to a valid temporary stack, then update ESP manually. (instead of finding out the value for ESP, then bothering the slow WRMSR with it)

Indeed. Thanks for pointing it out.


JAL

[Owen]

The convention with syscall/sysret is to use the SWAPGS instruction and hang a "processor control block" off of the kernel's GS.

For this reason, I'm looking forward to clang & llvm maturing, because they have support for declaring a pointer off the GS segment, and therefore better optimized code.