[Libre-soc-isa] [Bug 532] discuss iterative approach for statistical analysis of effectiveness of Compressed PowerISA encoding

bugzilla-daemon at libre-soc.org bugzilla-daemon at libre-soc.org
Sun Nov 29 00:04:08 GMT 2020


--- Comment #21 from Luke Kenneth Casson Leighton <lkcl at lkcl.net> ---
(In reply to Alexandre Oliva from comment #20)
> the compression ratio result in the skeleton is meaningless, I just added
> some random insns to the dictionary that encodes the compression decisions. 
> The idea is for it to be filled in with actual insn encoding data, which we
> don't have at all for attempt 2.

ohh ok.

> At ease, I won't mix up the different scripts with the different encoding
> logic.  They will share some structure, but it's going to be significantly
> different scripts, more so because I'm going to codify the proposed attempt
> 1 encodings in it.

(*sotto voice* although it would be longer could you possibly avoid all but
  the most basic of regexs, i just... i can't read them!  or, i can, but it's
  5-10 minutes of looking at them)

btw watch out for "nop" which is "ori r0, r0, 0" of which there are 32,000
in the disasm of bash, and i believe they're data not instructions.

> Yeah, sorry, I messed up in the bug report number.  I still haven't quite
> got into my mind the notion that there are separate ones for these related
> but very different purposes ;-)

we learned over time that the discussions get really long, significantly
slowing down bugzilla and the VM running it.

You are receiving this mail because:
You are on the CC list for the bug.

More information about the Libre-SOC-ISA mailing list