Chengjun Chen

> It is ok to leave this feature as a TODO. It is your call. I have done most parts of this encoded clock id and leave the implementations for...

> @cchanging Do you have any clue about this bug? It's not clear yet through this information. I'll look into it.

> The "user" and "sys" columns seem to display random values. I find out the reason. `time` command uses `wait4` syscall to acquire the user time and sys time through...

@skpupil By using #814 and adding dummy `vfork`, I can get the correct user time and sys time from `time` command.

This is an issue in current system. Theoretically, there shouldn't be such a long period of interrupt disabling during startup. With KVM enabled, this interrupt disable time is approximately 5...

Currently during the current boot process, some instructions like copying a large file may still result in an excessively long period of disabled interrupts, leading to this issue. If the...

> The following program takes nearly 20 seconds to run on Asterinas Thanks for your issues. The reason for the slow execution of this program is due to the handling...

I agree，We do indeed require the kernel space to have this capability. In theory, the guard page doesn't need additional physical pages. The current implementation is primarily constrained by the...

Currently we can not support running sysbench in command-line interactive mode, but we can run sysbench with specified tests directly. You can check the README.md in /regression/benchmark to see detial...

@skpupil , this appears to be a memory-related bug. Currently two obvious bugs are due to the `resize` of the VMO and bugs with copy-on-write during fork. You can try...