Dan -- > >I've continued to play with the DO_OP simplification patch I posted > >before. > > Looks good, at least as far as I can tell from the diff. Go ahead and apply > it, and I'll probably tweak from there.
I just committed this. It still wins vs. stock on my machine, I hope YMDV too much... (NOTE: The C-based disassembler was split into a separate patch. It isn't ready for prime-time). Here's what's included: * test_prog and pdump are linked with only the files the need, not al the files. * build_interp_starter.pl writes out op_info.[hc], too. These files contain op name, arg count and arg type information. This will be useful for multiple things in the future, including better trace output and the crystalizing loader. * The DO_OP macro is now simpler and doesn't require explicit temporaries. * interpreter.c is modified to use the new op info in tracing output. * The interpreter just points its opcode_funcs and opcode_info members at the default tables rather than allocating memory on the fly. Regards, -- Gregor _____________________________________________________________________ / perl -e 'srand(-2091643526); print chr rand 90 for (0..4)' \ Gregor N. Purdy [EMAIL PROTECTED] Focus Research, Inc. http://www.focusresearch.com/ 8080 Beckett Center Drive #203 513-860-3570 vox West Chester, OH 45069 513-860-3579 fax \_____________________________________________________________________/