Derek Schuff

Is it actually necessary to have the em-foo wrapper? Is there some reason we can't just tell CMake the location of the clang-scan-deps tool directly?

> > Is it actually necessary to have the em-foo wrapper? Is there some reason we can't just tell CMake the location of the clang-scan-deps tool directly? > > The...

> But how would clang know where the emscripten sysroot is without being told like this in the wrapper? I think it might make sense for CMake to pass the...

Actually I meant having CMake have the information about the sysroot in our toolchain file (since that already has alll the special knowledge about emscripten/emsdk) do it for both clang...

@sbc100 I think we still want this. IIRC it appeared to be working but there was a failure in the CMake tests that we never debugged. This has been an...

When loading a split module, the loader code around [here](https://github.com/emscripten-core/emscripten/blob/main/src/preamble.js#L535) patches the indirect function table, overwriting the pointers to the proxy with the real functions. When wasm-split creates a second...

These section names are very generic, but they seem to have very specific meaning wrt OCI. It's easy to imagine some tool that's unaware of OCI using one of them....

> This PR implements the mapping from the `-ffast-math` compiler flag to the `wasm-opt --fast-math` optimization flag, as requested in issue #21497. Did anyone actually request this? Can you say...

> Many applications use floating point math extensively so -ffast-math flag enables aggressive optimizations that can significantly improve performance around 10-30% improvement without code changes and the issue #21497 requesting...

Sorry I'm a little late to this conversation: I'm also very interested and happy that there is a GCC port in the works. I do have a few concerns and...