Wait - it’s 8-24x faster than O0 while producing code on par with O3???
Edit: LLM to LLVM
Seems like a pretty neat fast compiler backend for LLVM. Why the extra buzzwords?
In the llvm/llvm-project repository
git switch origin/release/19.x
cmake -GNinja -S. -B/tmp/out/custom -DLLVM_TARGETS_TO_BUILD='X86;AArch64' -DLLVM_ENABLE_PROJECTS=clang -DLLVM_ENABLE_PLUGINS=off -DCMAKE_BUILD_TYPE=Release -DLLVM_LINK_LLVM_DYLIB=on
# consider -DCLANG_ENABLE_OBJC_REWRITER=off -DCLANG_ENABLE_STATIC_ANALYZER=off -DCLANG_ENABLE_ARCMT=off -DCLANG_PLUGIN_SUPPORT=off
ninja -C /tmp/out/custom clang LLVM FileCheck # build clang and libLLVM.so and test utilities
git submodule update --init
cmake -GNinja -S. -Bout/debug -DCMAKE_BUILD_TYPE=Debug -DCMAKE_EXPORT_COMPILE_COMMANDS=on -DCMAKE_PREFIX_PATH=/tmp/out/custom -DCMAKE_CXX_COMPILER=$HOME/Stable/bin/clang++ -DCMAKE_C_COMPILER=$HOME/Stable/bin/clangThere are some failures:
``` % /tmp/out/custom/bin/llvm-lit out/debug/tpde/test/filetest ... Failed Tests (5): TPDE FileTests :: codegen/eh-frame-arm64.tir TPDE FileTests :: codegen/eh-frame-x64.tir TPDE FileTests :: codegen/simple_ret.tir TPDE FileTests :: codegen/tbz.tir TPDE FileTests :: tir/duplicate_funcs.tir ```
BarakWidawsky•1d ago
npalli•1d ago
The TPDE-based back-end compiles 4.27x faster than Cranelift and 2.68x faster than Cranelift with its fast register allocator, but is 1.74x slower than Winch
cfallin•1d ago
This isn't too surprising to me, as the person who wrote Cranelift's current regalloc (hi!) -- regalloc is super important to run-time perf, so for Wasmtime's use-case at least, we've judged that it's worth the compile time.
TPDE is pretty cool and it's great to see more exploration in compiler architectures!