labath added a comment. Hmm... how much is a "significant"? I tried `time optimized/lldb --file debug/clang -o "br set -n dump"` as a benchmark and I saw a 2-3% slowdown (the variance between individual samples is ~1%, so this is statistically significant).
I tried playing around with this, and I was able to get about 1% speedup (borderline significant) by changing the llvm's implementation to a lookup table (presumably, the speedup comes from the fact that the llvm version handles more forms, most likely the ones whose size is dwarf-version-dependent). However, this required making some non-llvm-y changes like ditching llvm::Optional as a return type, as that saved a couple of instructions (I should probably double check that with a top-of-tree compiler, to see if it can optimize it better). https://reviews.llvm.org/D44739 _______________________________________________ lldb-commits mailing list lldb-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
