home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!infoserv!steiny!
- From: steiny@steiny.com (Don Steiny)
- Newsgroups: gnu.gcc.help
- Subject: Version compatibility?
- Message-ID: <196@steiny.com>
- Date: 20 Dec 92 02:21:59 GMT
- Distribution: gnu
- Organization: Don Steiny Software
- Lines: 54
-
-
-
- I am attempting to get gcc 2.3.2, libg++2.3, gdb4-7 working on my
- system and I have some questions that someone out there may know the
- answer to.
-
- It seems to me that I have to get gcc working first, then libg++, then
- gdb. (Though the last two may not matter). However, I notice that one
- of the include directories in the makefile for gcc is
- /usr/local/lib/lib-gcc/g++-includes. The directory exists, but it contains
- the includes from gcc 2.2.2, the previous version I got mostly working.
- Is that going to create problems?
-
- This is about the third try on my part (it takes many hours to build
- gcc) to get gcc 2.3.2 working, and I am not sure what I am doing wrong.
-
- First, when I built it and tried to compile using g++ with the -g
- flag I would get an error that L_T51 could not be found. Sure enough the
- label was in the assembler code with a assembler directive. I will submit
- a formal bug report if this is not just due to some obvious screw-up on
- my part.
-
- Second, when I compiled a regular C program, the step option did not
- work in my gdb. I was using 4-6 and had not gotten 4-7 up. I assume that
- problem will go away if I upgrade to 4-7.
-
- gcc does not seem to recognize file suffixes and always gives me
- a link error if I do not explicitly call g++ for C++ programs.
-
- When I try to compile with the new gcc, I get many errors from
- the include files.
-
- I should note that the fixincludes script does not seem to run to
- completion, but I don't think that should make any difference for g++. I
- can just manually create the proper directory and copy the files in and
- it seems to allow gcc to run.
-
- I am on SVR4 on a 486, using a rather obscure UNIX (Microport), but
- I believe it is just the same as USL with some extra drivers and a few
- performance enhancements. In other words, the compiler and libraries are
- just the same.
-
- I think that all of this may just be one problem that is cascading
- and I wonder if my description looks familiar to anyone. If not, I will
- post detailed bug reports.
-
- Thanks for your help!!
-
- -don
- --
- Don Steiny
- Don Steiny Software
- Santa Cruz, CA 95060
- (408) 425-0382
-