Junyang Zhang

For your reference, we have attempts months ago to support LLVM instrumentation (aimed for coverage at that time). But that was closed since: 1. We haven't got scripts to dump...

> > Perhaps it's time to consider removing the `vmo` mechanism. > > Agree, but we need a proposal for an alternative. Behold, detailed designs documentation is on it's way!

While the problem you found is valuable, the solution is more like a hack. It seems that the compiler may use SIMD to optimize `copy_from`. I think the core problem...

> I was saying that in the absence of this PR, that is, only changing `fill(0)` to `fill(0u8)` is enough to enable SIMD optimization. It can be observed in `objdump...

> Optimizations in this way are only 10% slower on average than optimizations based on a zero-page This is not intuitive for me. I guess this is because more checks...

> For such large-scaled read-only operations, mapping zeroed page can greatly reduce memory usage, yet KSM is more likely to optimize the shared memory usage. Oh, yes. You could map...

I have revisited the frame part of memory management and leaking is a concern when implementing #713. But I think that alone don't fix all the problems here. Also, [Rust...

Well if it is merged the publishing cannot success because I published 0.4.2 myself and yanked that. Re publishing with the same version is forbidden and the next try when...

> By the way, the README.md contains some outdated information or duplicated one with the Book. I am counting on you to do some house keeping for the current README...

> @junyang-zh It looks like this PR needs to be rebased to the latest main branch in order to pass the CI. Done.