On 2018-Sep-25 08:01:52 -0700, Larry McVoy wrote: >On Tue, Sep 25, 2018 at 11:00:37AM +0100, Tony Finch wrote: >> Peter Jeremy wrote: >> >> > In the specific case of x86, I would dispute that. The various warts in >> > the x86 instruction set and "architecture" mean that x86 code density is >> > relatively low and on a par with SPARC code. >> >> This paper has a nice survey of instruction set densities, which very much >> disagrees with your statement: >> >> http://web.eece.maine.edu/~vweaver/papers/iccd09/iccd09_density.pdf > >That's a neat paper, I like it, thanks for the pointer. I'm curious >why Peter thought what he thought, my guess would have been more like >what the paper showed, but that was a "hand optimized assembly", maybe >the compilers aren't that good? I dunno, Peter, care to comment? I agree that looks like an interesting paper - I've skimmed it and will have to read it in details. I was thinking back to when I was using a mixture of SPARC and x86 at a previous job. I didn't do any careful analysis, more eyeballing various executables and gut feeling. I no longer have access to that environment. In view of that paper, I'll withdraw my claim since it's not backed up by evidence. -- Peter Jeremy