search for: johnskal

...ict-aliasing -std=c++11 -fPIC .... Elapsed: 106.058, Result code 0 /usr/local/bin/clang++ -fno-common -fno-strict-aliasing -std=c++11 -dynamiclib -O3 .... Elapsed: 0.049, Result code 0 env DYLD_LIBRARY_PATH=build/release/lib/rtl:$DYLD_LIBRARY_PATH build/release/bin/flx_run '/Users/johnskaller/felix/test/regress/rt/tuple-02.dylib' Elapsed: 0.007, Result code 0 With -O2: Compile: Elapsed: 106.918, Result code 0 Link: Elapsed: 0.048, Result code 0 Run: Elapsed: 0.010, Result code 0 With -O1: Compile: Elapsed: 13.664, Result code 0 Link: Elapsed: 0.042, Result code...

....... > Elapsed: 106.058, Result code 0 > > /usr/local/bin/clang++ -fno-common -fno-strict-aliasing -std=c++11 -dynamiclib -O3 .... > Elapsed: 0.049, Result code 0 > > env DYLD_LIBRARY_PATH=build/release/lib/rtl:$DYLD_LIBRARY_PATH build/release/bin/flx_run '/Users/johnskaller/felix/test/regress/rt/tuple-02.dylib' > Elapsed: 0.007, Result code 0 > > > With -O2: > Compile: Elapsed: 106.918, Result code 0 > Link: Elapsed: 0.048, Result code 0 > Run: Elapsed: 0.010, Result code 0 > > With -O1: > Compile: Elapsed: 13.664,...

Adding LLVMdev, since this is intimately related to the optimization passes. > I think this is roughly because some function level optimisations are > worse than O(N) in the number of instructions. Please profile this and mail llvmdev regarding passes with significantly superlinear behavior (e.g. O(n^2)). My understanding is that these kinds of algorithmic problems are generally considered