[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: ddd 3.x* Compilation problems with gcc 3*
From: |
R. Bernstein |
Subject: |
Re: ddd 3.x* Compilation problems with gcc 3* |
Date: |
Thu, 21 Nov 2002 01:20:59 -0500 |
Just compared the changes you report with ones that I just made to get
DDD to work bash debugging. From the information you give, I now
understand better and precisly the problems are going on. (and I
confess that some of your workarounds are better than mine).
A complete patched source can be found here:
http://sourceforge.net/project/showfiles.php?group_id=61395
under "ddd-for-bash".
So based on your work, further investigation reveals this.
1. ios::iostate. There is a configure a test for the presence of
this. However the test is returning negative because g++ now
warns that this include is deprecated (and gcc is passed
-Werror).One possibility I experimented with is to change that to
-Werror -Wno-deprecated. If this is done (or -Werror removed), then
HAVE_IOSTATE will be set correctly automatically. It however may
cause other problems.
2. Perhaps what's going on here is that Motif defines names in
incompatible way with Athena various names. DDD says it supports
just Athena widgets, but I'd imagine that you can't support both
simultaneously, nor do I think it makes sense to do so. So probably
when Motif is enabled the configuration should disable Athena.
3. I believe someone else posted a query to this problem back in March:
http://mail.gnu.org/pipermail/ddd/2002-March/001657.html
The suggestion there was to move the #include <fstream.h>
line in logplayer.C up to before the various #include "localheader" lines.
The local include "streampos.h" I think checks to see if something perhaps
from <fstream.h> has been included, so moving the order around is probably
preferable to removing a line.