New GCC Back-End Proposed For WebAssembly
When it comes to compiling C/C++ code to WebAssembly (WASM), LLVM/Clang and other LLVM-based tooling has dominated the space. Nearly a decade ago was a proposal for a GCC WebAssembly back-end that ultimately never ended up being merged while now there is a new proposal for a WebAssembly back-end for the GNU toolchain.
An open-source developer this week posted a patch series introducing a new GCC back-end for WebAssembly. This has yet to receive approval from the GCC Steering Committee but is the first time seeing WASM activity for GCC in years.
The back-end is still in early stages of development and lacks some features like WASM reference types, tables, exceptions, debug info, data sections, and other features, but at least far enough along for announcing on the GCC patches mailing list.
This GCC Wiki page lays out the early details on this back-end along with instructions on building this GCC WebAssembly back-end plus related WASM components.
The patch series for GCC in its current form is around three thousand lines of code and is being treated as a "request for comments" (RFC) at this time. It will be interesting to see where this leads and after all these years later if GCC will end up finally having a WASM back-end and viable alternative to the LLVM-based WASM world.
An open-source developer this week posted a patch series introducing a new GCC back-end for WebAssembly. This has yet to receive approval from the GCC Steering Committee but is the first time seeing WASM activity for GCC in years.
The back-end is still in early stages of development and lacks some features like WASM reference types, tables, exceptions, debug info, data sections, and other features, but at least far enough along for announcing on the GCC patches mailing list.
This GCC Wiki page lays out the early details on this back-end along with instructions on building this GCC WebAssembly back-end plus related WASM components.
The patch series for GCC in its current form is around three thousand lines of code and is being treated as a "request for comments" (RFC) at this time. It will be interesting to see where this leads and after all these years later if GCC will end up finally having a WASM back-end and viable alternative to the LLVM-based WASM world.