On 10/06/2017 03:46 PM, Wilco Dijkstra wrote: > Martin Liska wrote: > >> There are some numbers for cc1plus: >> >> $ bloaty ./objdir2/gcc/cc1plus -- ./objdir/gcc/cc1plus >> VM SIZE FILE SIZE >> +3.8% +1.11Mi TOTAL +1.03Mi +0.5% > >> insn-attrtab.o: >> VM SIZE FILE SIZE >> +214% +682Ki .rodata +682Ki +214% >> -50.1% -63.3Ki .text -63.3Ki -50.1% > > So is that a 3.8% codesize increase or decrease? If an increase,
Hi. Yes, increase. > I can't see how replacing 63KB of instructions with 682KB of data > is a good tradeoff... There should be an accurate calculation > of the density, taking the switch table width into account (really small > tables can use 1-byte offsets, large tables are typically forced to > use 4-byte offsets). This may need new target callbacks - I changed > PARAM_CASE_VALUES_THRESHOLD on AArch64 to get smaller > code and better performance since the current density calculations > are hardcoded and quite wrong for big tables... Let me return to that in next stage1. Can you please provide more details about how it affects AArach64 target? > > Also what is the codesize difference on SPEC2006/2017? I don't see > any mention of performance impact either... Will prepare that. Thank you, Martin > > Wilco >
