Matthew Parkinson

My concern with two paths is the code uses a lot of tail recursive calls that get optimised into jumps. This might be tricky to maintain if there are two...

@nwf the pagemap is updated with the [`BackendAllocator`](https://github.com/microsoft/snmalloc/blob/6ad7f65d19f36eea905b14ae0753114287ff2c09/src/backend/backend.h#L291), which does not hold the global lock. My plan was the use the code in [`ChunkAllocator`](https://github.com/microsoft/snmalloc/blob/main/src/backend/chunkallocator.h). This currently performs a time based...

Is it possible to get a stack trace? My guess is that something about how thread local state is being initialised is interacting badly with our thread local state.

I'm going to leave this issue open until we have CI support to test Android support.

Preliminary support provided by #171.

@zhmt from your comment it looks like you are on Windows. I think you want the static library for your use case. If you set CMAKE with: ``` -DSNMALLOC_STATIC_LIBRARY_PREFIX="" ```...

So Emery Berger's heap layers has a library for redirecting allocators on Windows: https://github.com/emeryberger/Heap-Layers/blob/f71407951a9b68882ad528d6200c2528e39bb31d/wrappers/winwrapper.cpp This looks like we might be able to use this with snmalloc. I don't have capacity...

1) Performance is pretty similar. There are examples where `snmalloc` out performs `mimalloc` and there are examples where `mimalloc` out performs `snmalloc`. `snmalloc` tends to do better in very asymmetric...

So running some of my usual benchmarks does not show any statistically significant difference. I am still happy to take this. I wonder if it makes sense to have a...

This is a really interesting idea. I am not sure what the granularity of NODUMP should be. ``` The Linux implementation requires that the address addr be page- aligned, and...