[M3commit] CVS Update: cm3

[Daniel Alejandro Benavides D.]

Hi all:

There are known techniques for avoid this to happen or tu use brute
force approaches (even when they don't work as resorted in p. 43):

http://fg-metriken.gi.de/fileadmin/user_upload/news/news98_2.pdf



There are approaches for typeless abstract interpretation, but for functional correctness of the abstract machine we have to develop execution models
and improve low-level error-catching like above explained if it can get enough , I guess the
issue is in what in-memory approach are you thinking 

Thanks in advance

--- El vie, 7/12/12, Jay Krell <jkrell at elego.de> escribió:

De: Jay Krell <jkrell at elego.de>
Asunto: [M3commit] CVS Update: cm3
Para: m3commit at elegosoft.com
Fecha: viernes, 7 de diciembre, 2012 06:47

CVSROOT:    /usr/cvs
Changes by:    jkrell at birch.    12/12/07 11:47:21

Modified files:
    cm3/m3-sys/m3back/src/: M3C.m3 

Log message:
    Punt and make check_range a helper function.
    That reduces warnings.
    Doing better here seems elusive.
    There are still range warnings to fix, probably check_lo/hi/index.
    This does mean we could probably dispense with most of my recent
    work on forming a higher level in-memory form.
    Also, ideally check_range helper would call reportfault, but I don't
    think we have that name by this time in the code.
    todo: capture the value into locals for all the check_ operations.
    (requires another pass to allocate the locals)

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://m3lists.elegosoft.com/pipermail/m3commit/attachments/20121209/5ff3db16/attachment-0002.html>