question about upvalue bytecode encoding
- From: Francesco Abbate <francesco.bbt@xxxxxxxxx>
- To: luajit <luajit@xxxxxxxxxxxxx>
- Date: Tue, 25 Mar 2014 23:44:51 +0100
Hi,
I'm working on luajit-lang-toolkit to complete the bytecode generator.
In case someone is interested: I've completed the implementation of
goto statements and included a test suite. I'm now ironing out the
last few remaining issues.
Working on the bytecode encoding of the prototypes I've seen that
upvalues are sometimes encoded with a bitwise or with 0x8000 or
0xc000. I was wondering the difference between these two kinds of
upvalues.
In Nyanga's bytecode generator Richard's was using only the 0x8000 form.
I know that there is a difference between direct upvalues (local
variables of the enclosing functions) and indirect upvalues but the
0x8000 vs 0xc000 is unclear to me.
I would greatly appreciate to have some clarifications or otherwise to
have an indication of where to look in the source code.
Francesco
Other related posts:
