As was pointed out in the first post, MASM is fast when generating OMF.

I do know that OMF has an LIDATA record which specifies repetitive data with a (repeat-count, data0-dataN) format. So an OMF assembler doesn't need to generate the entire data block (and hands the performance hit over to the linker).
Posted on 2002-02-13 21:55:42 by tank
Assembling uninitialized data should be a very quick operation no matter the size. The label name, type, size and offset relitive to the current segment are stored in a list of some type, and the assembler moves on!
Posted on 2002-02-13 23:08:09 by bitRAKE
I've posted my bin2o tool version 0.4 on my site ( ) .
Get it from the bottom of the tools section. No sourcecode, as it's
a bit messier than what I would like ;). 4096byte .exe written in
C++ without any tricks. Yeah, using oop and all.
Posted on 2002-02-16 11:45:32 by f0dder
Looks like a nice toy f0dder, I don't suppose you could translate this crap into assembler could you ? It may make the toy a lot more useful to more people if they understood what it was meant to do and how to do it.

Converts a binary file to (ms/win32) COFF object file, accessible as a
static array (C-style extern "C" unsigned char data[]; ).

Posted on 2002-02-17 04:51:22 by hutch--
"extern mydata:byte" :)
Posted on 2002-02-17 04:59:02 by f0dder