john.brawn added a comment. In http://reviews.llvm.org/D20732#456239, @jyknight wrote:
> This looks okay. The only downside: it will increase sizeof(Decl) on 32bit > platforms, since the layout is: > vtable ptr (4) > NextInContextAndBits ptr (4) > DeclCtx ptr (4) > Loc int (4) > bitfield int (4) > > Totaling 20 bytes, which will now be rounded up to 24 bytes. I don't know how > others feel, but it seems probably fine to me? On 64-bit platforms, it's > already 32 bytes, and that won't be affected. I did an experiment compiling the top 5 largest llvm/clang source files with and without this patch and got the following numbers for "Maximum resident set size (kbytes)" reported by time -v | **Source File** | **Before** | **After** | **Change** | | X86ISelLowering.cpp | 679824 +/- 28 | 679897 +/- 271 | 73 (0.01%) | | SemaDecl.cpp | 705379 +/- 440 | 705590 +/- 77 | 211 (0.03%) | | SemaExpr.cpp | 950729 +/- 6910 | 951654 +/- 5834 | 925 (0.10%) | | DAGCombiner.cpp | 443673 +/- 38 | 443721 +/- 16 | 48 (0.01%) | | SemaDeclCXX.cpp | 850441 +/- 4531 | 851891 +/- 5079 | 1450 (0.17%) | I think a <0.2% increase in memory usage is fine. Repository: rL LLVM http://reviews.llvm.org/D20732 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
