Runtime initialization
PeterPan
xuesongguo@sohu-inc.com
Thu Jan 10 22:26:00 GMT 2002
hi Rayiner Hashem.
I think your problem is uninitailized object or bad point, or bad linkage.
as i know, there is nothing special at runtime needed, if your object is
initailized correctly , its vtable should be ok.
is your object a static one or a dynamic ? if a static, do you sure you are
calling it's virtaul functions after it constructed ?
did you call the function from another static object's constructor (this may
couse proble) ?
if it's a dynamic one, do you sure not calling it from a NULL point ?
---------------------------------------------------------
1uNCEE
Guo Xuesong
Sohu-inc
phone: +86 13910010559
email: xuesongguo@sohu-inc.com
-----Original Message-----
From: gcc-help-owner@gcc.gnu.org [mailto:gcc-help-owner@gcc.gnu.org]On
Behalf Of Rayiner Hashem
Sent: Friday, January 11, 2002 7:51 AM
To: gcc-help@gcc.gnu.org
Subject: Runtime initialization
I have a few questions about what kind of runtime support g++ needs. I'm
(trying) to use full-blown C++ in an OS kernel, and I'm having some problems
invoking virtual functions. Basically, the problem occurs when I try to call
the function itself, the CPU triple-faults and reboots. I'm thinking it has
something to do with the vtables (I don't touch them anywhere else in the
code). I would appreciate some info on how GCC manages vtables and
specifically, what, if anything, needs to be done to the vtable sections
(the gnu.linkonce.d.__vtable*) before virtual functions can be used.
Thanks in advance,
Rayiner H.
More information about the Gcc-help
mailing list