cherrymui

Maybe what happens is that 1. C created thread calls into Go, which calls back to C 2. it uses some C stack, then calls back into Go, which updates...

Thanks for the update! I don't think we want to introduce any autoconf-like things into `go build`, to keep it simple and understandable. So we'd have to use some other...

I don't think we can _just_ skip the bounds check. If C1 calls Go1, which calls C2, which calls Go2, if at Go2 (where `mp.ncgo > 0`) we update the...

[CL 600296](https://go.dev/cl/600296) implements the update+restore logic. Could you try if this works with your case? Thanks.

@chenxiaolong thanks. We could add `__BIONIC__` to the condition. We don't seem to have a policy for which versions of Android we support. Do we require API level 34 for...

As you have read, https://go.dev/doc/go1.22#enhanced_routing_patterns mentions > This change breaks backwards compatibility in small ways... The change is controlled by a [GODEBUG](https://go.dev/doc/godebug) field named httpmuxgo121. Set httpmuxgo121=1 to restore the...

> We can't use AVX in generated code without a feature check. A feature check isn't very expensive, especially compared to writing more than a kilobyte, but it does mean...

Thanks for the report. From the Rust post it sounds like it only does the read if it is called for the main thread. Could you try calling Go from...

Yes, trying calling from a different thread is for validating the assumption. Calling on the main thread should not be slow, either. So if that's the case we'll look into...

Sorry, I haven't got a chance to implement this. My plan is that if we can get an accurate stack bounds, which is the common case with glibc, we just...