LLVM Clang 3.8 Compiler Optimization Benchmarks With -Ofast
A few days ago I posted a number of LLVM Clang optimization level benchmarks using the latest code for the upcoming Clang 3.8 release. Those tests went from -O0 to -O3 -march=native, but many Phoronix readers wanted -Ofast so here are those results too.
I didn't include -Ofast in the original tests since I don't know of many using this optimization level within a production capacity considering it has the potential of doing unsafe math as it disregards standards compliance in the name of performance. However, since several readers requested it and I still had this LLVM/Clang 3.8 build around in the same system configuration, I added in extra runs with -Ofast and -Ofast -march=native.
Those are just two of the results... To see all of the latest data about the performance impact of -Ofast relative to other optimization levels for the bleeding-edge LLVM Clang 3.8 compiler, see this OpenBenchmarking.org result file for all of the details and compiler benchmark results.
I didn't include -Ofast in the original tests since I don't know of many using this optimization level within a production capacity considering it has the potential of doing unsafe math as it disregards standards compliance in the name of performance. However, since several readers requested it and I still had this LLVM/Clang 3.8 build around in the same system configuration, I added in extra runs with -Ofast and -Ofast -march=native.
Those are just two of the results... To see all of the latest data about the performance impact of -Ofast relative to other optimization levels for the bleeding-edge LLVM Clang 3.8 compiler, see this OpenBenchmarking.org result file for all of the details and compiler benchmark results.
9 Comments