Calculating relative jnz opcode sizes in a custom code generator

if the destination is more than 0x7F bytes away from the instruction following your Jcc, then use the long form.

But honestly, do you think that an extra 4-bytes really matter, if you consider the greatly added simplicity of *always* using the 6-byte instruction (0F 8x)?

If you start resizing instructions, then you have to make sure that *all* jumps that go beyond the resize/insertion point are also fixed. It's certainly not impossible, but seems like an overkill for what you want.

[redbull]

Hi Doug,

Agreed but the point is to actually make it right now.. My existing code generator works well and now I want to take it to the next level.


Here are some interesting arictles on this topic:

hxxp://compilers.iecc.com/comparch/article/97-05-281

hxxp://compilers.iecc.com/comparch/article/00-10-073

hxxp://compilers.iecc.com/crenshaw/

hxxp://www.pcengines.ch/tp3.htm

hxxp://www.etek.chalmers.se/~e8johan/compiler/

Generating instructions with variable length addresses is a well understood
problem. The PDP-11 Unix assembler did it in about 1973, and there was
a paper by Tom Szymanski in the CACM about that time. I always preferred
the other approach, start with everything long, then shorten until there's
nothing left to shorten, which has the advantage of being fail-safe, all
of the intermediate stages being valid code. -John