dwf%hope@LANL.GOV (David W. Forslund) (09/24/88)
I've tried building g++ 1.27.0 on a Sun3 running SunOS4.0. I compiled it with gcc 1.26 and applied the combine.c patch and the tm-sun3.h patch. I also fixed the commented out line in the Makefile for building ld++ so that it had the "right" options for the Sun3 instead of the Sun4. G++ builds and compiles everything fine including libg++.a. However, when I try to load anything including the tests, I get Undefined symbols: start Script started on Fri Sep 23 15:39:00 1988 make tests hope% (cd tests; make GXX="../../g++/g++ -B../../g++/ -L..") ../../g++/g++ -B../../g++/ -L.. -g -fchar-charconst -B../../g++/ -c test1.cc ../../g++/g++ -B../../g++/ -L.. test1.o -o test1 Undefined symbols: start ../../g++/g++ -B../../g++/ -L.. -g -fchar-charconst -B../../g++/ -c test2.cc ../../g++/g++ -B../../g++/ -L.. test2.o -o test2 -lg++ -lm Undefined symbols: start ../../g++/g++ -B../../g++/ -L.. -g -fchar-charconst -B../../g++/ -c test3.cc ^C *** tests not removed. hope% g++ -fchar-charconst -v -c test1.cc g++ version 1.27.0 /usr/local/lib/gcc-cpp+ -v -undef -D__GNU__ -D__GNUG__ -Dmc68000 -Dsun -Dunix test1.cc /tmp/cca00425.cpp GNU CPP version 1.27.0 /usr/local/lib/gcc-c++ /tmp/cca00425.cpp -quiet -dumpbase test1.cc -fchar-charconst -noreg -version -o /tmp/cca00425.s gGNU C++ version 1.27.0 (68k, MIT syntax) compiled by GNU C version 1.26. as -mc68020 /tmp/cca00425.s -o test1.o hope% g++ -v -g -o test1 test1.o -lg++ -lm g++ version 1.27.0 /usr/local/lib/gcc-ld++ -C -o test1 -e start -dc -dp /usr/local/lib/crt0+.o /lib/Mcrt1.o test1.o -lg++ -lm -lg++ /usr/local/lib/gcc-gnulib+ -lg -lc Undefined symbols: start hope% ^D script done on Fri Sep 23 15:41:23 1988 Any ideas? Dave Forslund (dwf@lanl.gov)
metz@iam.unibe.ch (Igor Metz) (09/26/88)
Hi,
I have some problems with the latest g++ version. I compile it on my
SUN 3/260 (SunOS 3.5) and I get the following warnings from gcc 1.27
===== the beginning ======
gcc -g -c fold-const.c
In function fold:
fold-const.c:1435: warning: comparison is always 0 due to limited range of data type
fold-const.c:1452: warning: comparison is always 0 due to limited range of data type
gcc -g -c expr.c
In function expand_increment:
expr.c:3258: warning: comparison is always 0 due to limited range of data type
expr.c:3258: warning: comparison is always 0 due to limited range of data type
gcc -g -c class.c
In function instantiate_type:
class.c:3454: warning: case value out of range
class.c:3455: warning: case value out of range
class.c:3456: warning: case value out of range
class.c:3457: warning: case value out of range
gcc -g -c method.c
In function dump_init:
method.c:690: warning: case value out of range
method.c:691: warning: case value out of range
method.c:696: warning: case value out of range
method.c:697: warning: case value out of range
gcc -g -c search.c
In function my_tree_cons:
search.c:172: warning: initialization between incompatible pointer types
In function my_build_string:
search.c:186: warning: initialization between incompatible pointer types
==== the end ====
When I then try to compile the following little program, gcc-c++ abort()'s
itself (see below for typescript of gdb session), when it tries to expand
an expression:
==== the beginning ====
#include <stream.h>
const int dim = 10;
class Vektor {
short vec[dim];
public:
Vektor();
short& operator[](int i) { return(vec[i]); }
};
Vektor::Vektor()
{
for (int i= 0; i<dim; i++)
vec[i] = 0; // or any other value needed
}
class Matrix {
Vektor mat[dim];
public :
Vektor& operator[](int i) { return(mat[i]); }
};
main()
{
Matrix m;
int i, j;
for (i=0; i<dim; i++)
for (j=0; j<dim; j++)
m[i][j] = j + i*dim;
for (i=0; i<dim; i++) {
for (j=0; j<dim; j++) cout << m[i][j] << "\t";
cout << "\n";
}
}
==== the end ====
==== the beginning ====
Script started on Mon Sep 26 11:39:01 1988
gina:/u-gina/metz/g++/g++% gdb ~/lib/gcc-c++ core
GDB 2.3, Copyright (C) 1987 Free Software Foundation, Inc.
There is ABSOLUTELY NO WARRANTY for GDB; type "info warranty" for details.
GDB is free software and you are welcome to distribute copies of it
under certain conditions; type "info copying" to see the conditions.
Reading symbol data from /u-gina/metz/lib/gcc-c++...done.
Core file is from "gcc-c++".
Type "help" for a list of commands.
(gdb) whe
#0 0xa3f84 in kill ()
#1 0x337a0 in expand_expr (exp=(tree) 0xd4848, target=(rtx) 0x0,
tmode=VOIDmode, modifier=EXPAND_NORMAL) (expr.c line 3018)
#2 0x372ba in expand_expr_stmt (exp=(tree) 0xd4848) (stmt.c line 752)
#3 0xef90 in yyparse () (parse.y line 1802)
#4 0x3162 in compile_file (name=(char *) 0xefffc30 "/tmp/cca00218.cpp")
(toplev.c line 905)
#5 0x4a28 in main (argc=9, argv=(char **) 0xefffbc8,
envp=(char **) 0xefffbf0) (toplev.c line 1716)
(gdb) up 1
#1 0x337a0 in expand_expr (exp=(tree) 0xd4848, target=(rtx) 0x0,
tmode=VOIDmode, modifier=EXPAND_NORMAL) (expr.c line 3018)
3018 abort ();
(gdb) li
3013
3014 case ERROR_MARK:
3015 return const0_rtx;
3016
3017 default:
3018 abort ();
3019 }
3020
3021 /* Here to do an ordinary binary operator, generating an instruction
3022 from the optab already placed in `this_optab'. */
(gdb) i lo
code = -124
exp = (tree) 0xd4848
ignore = 1
mode = SImode
negate_1 = 251656602
op0 = (rtx) 0xefff698
op1 = (rtx) 0xefffbf0
original_target = (rtx) 0x0
subtarget = (rtx) 0x0
temp = (rtx) 0x9
this_optab = (optab) 0xefff698
type = (tree) 0xd2d90
(gdb) p *exp
$1 = {common = {uid = 13992, chain = 0x0, type = 0xd2d90, code = -2079326208,
external_attr = 0, public_attr = 0, static_attr = 0, volatile_attr = 1,
packed_attr = 0, readonly_attr = 0, literal_attr = 0, nonlocal_attr = 0,
permanent_attr = 0, addressable_attr = 0, regdecl_attr = 0, this_vol_attr = 0,
unsigned_attr = 0, asm_written_attr = 0, overloaded_attr = 0, private_attr =
0, protected_attr = 0, virtual_attr = 0, inline_attr = 0, everused_attr = 0,
has_ctor_attr = 0, needs_ctor_attr = 0, has_dtor_attr = 0, needs_dtor_attr =
0}, int_cst = {common =
{"\000\0006\250\000\000\000\000\000\r-\220\204\020\000\000"}, int_cst_low =
1478668, int_cst_high = 883658}, real_cst = {common =
{"\000\0006\250\000\000\000\000\000\r-\220\204\020\000\000"}, rtl = 0x16900c,
real_cst = 1.87512e-308}, string = {common =
{"\000\0006\250\000\000\000\000\000\r-\220\204\020\000\000"}, rtl = 0x16900c,
length = 883658, pointer = 0xd480c ""}, complex = {common =
{"\000\0006\250\000\000\000\000\000\r-\220\204\020\000\000"}, rtl = 0x16900c,
real = 0xd7bca, imag = 0xd480c}, identifier = {common =
{"\000\0006\250\000\000\000\000\000\r-\220\204\020\000\000"}, length =
1478668, pointer = 0xd7bca "", global_value = 0xd480c, class_value = 0xd486c,
local_value = 0xc91a8, label_value = 0x36ab, implicit_decl = 0x0}, decl =
{common = {"\000\0006\250\000\000\000\000\000\r-\220\204\020\000\000"},
filename = 0x16900c "", linenum = 883658, size = 0xd480c, mode = 870508,
size_unit = 0 '\000', align = 12 '\f', voffset_unit = 145 '\221', name =
0x36ab, original_name = 0x0, context = 0x1664c2, offset = 855638016, voffset =
0xd3824, arguments = 0x166556, result = 0x36ad, initial = 0x0, rtl = 0xd328a,
frame_size = 956301312, saved_insns = 0xd486c, block_symtab_address =
1478668}, type = {common =
{"\000\0006\250\000\000\000\000\000\r-\220\204\020\000\000"}, values =
0x16900c, sep = 0xd7bca, size = 0xd480c, size_unit = 0 '\000', align = 13
'\r', sep_unit = 72 'H', mode = 1811942545, pointer_to = 0xc91a8, reference_to
= 0x36ab, parse_info = 0, symtab_address = 1467586, name = 0x33000000, max =
0xd3824, next_variant = 0x166556, main_variant = 0x36ad}, classtype =
{type_info =
{"\000\0006\250\000\000\000\000\000\r-\220\204\020\000\000\000\026\220\f\000\r{\312\000\rH\f\000\rHl\000\f\221\250\000\0006\253\000\000\000\000\000\026d\3023\000\000\000\000\r8$\000\026eV\000\0006\255"},
type_flags = {has_type_conversion_attr = 0, has_int_conversion_attr = 0,
has_float_conversion_attr = 0, gets_init_ref_attr = 0, gets_new_attr = 0,
gets_delete_attr = 0, has_assignment_attr = 0, gets_assignment_attr = 0,
has_assign_ref_attr = 0, gets_assign_ref_attr = 0, has_wrapper = 0,
needs_wrapper = 0, has_wrapper_pred = 0, has_method_call_overloaded = 0,
gets_init_aggr_attr = 0, has_call_overloaded = 0, has_array_ref_overloaded =
0, any_assigns_this = 0, none_assign_this = 0, uses_multiple_inheritance = 0,
marked = 0, marked2 = 0, marked3 = 0, marked4 = 0, declared_class = 0, dummy_6
= 0}, offset = 0xd328a, vtable = 0x39000000, virtuals = 0xd486c, vfield =
0x16900c, vsize = 13998, n_parents = 0, cid = 863632, types = 0x25020000,
n_ancestors = 0, by_bfs = 0x0, for_bfs = 0x36af, by_dfs = 0x0, for_dfs =
0xd328a, tags = 0x75000000, main_class_variant = 0xd489c, next_class_variant =
0x36b0, memoized_table_entry = 0x0, via_pub_or_virt = 0xd328a "", sep =
0x25020000, size = 0x0, size_unit = 0 '\000', align = 0 '\000', sep_unit = 0
'\000', mode = 54, base_init_list = 0x36b1, member_init_list = 0x0}, list =
{common = {"\000\0006\250\000\000\000\000\000\r-\220\204\020\000\000"},
purpose = 0x16900c, value = 0xd7bca}, exp = {common =
{"\000\0006\250\000\000\000\000\000\r-\220\204\020\000\000"}, operands =
{0x16900c}}, stmt = {common =
{"\000\0006\250\000\000\000\000\000\r-\220\204\020\000\000"}, filename =
0x16900c "", linenum = 883658, body = 0xd480c}, if_stmt = {common =
{"\000\0006\250\000\000\000\000\000\r-\220\204\020\000\000"}, filename =
0x16900c "", linenum = 883658, cond = 0xd480c, thenpart = 0xd486c, elsepart =
0xc91a8}, bind_stmt = {common =
{"\000\0006\250\000\000\000\000\000\r-\220\204\020\000\000"}, filename =
0x16900c "", linenum = 883658, body = 0xd480c, vars = 0xd486c, supercontext =
0xc91a8, bind_size = 0x36ab, type_tags = 0x0}, case_stmt = {common =
{"\000\0006\250\000\000\000\000\000\r-\220\204\020\000\000"}, filename =
0x16900c "", linenum = 883658, index = 0xd480c, case_list = 0xd486c}}
(gdb) q
==== the end ====
In my earlier experiments gcc-cpp+ crashed in malloc until I used the GNUEmacs
version of malloc.
Is there anybody out there who can help???
Thank you in advance
Igor Metz X400: metz@iam.unibe.ch
Institut fuer Informatik ARPA: metz%iam.unibe.ch@relay.cs.net
und angewandte Mathematik UUCP: ..!uunet!mcvax!iam.unibe.ch!metz
Universitaet Bern
Switzerland Phone: (+41) 31 65 49 02metz@iam.unibe.ch (Igor Metz) (09/26/88)
Hi, I'm answering my question myself! I asked > I have some problems with the latest g++ version. I compile it on my > SUN 3/260 (SunOS 3.5) and I get the following warnings from gcc 1.27 The problem was, that I had the wrong version of gcc on my search path (1.18 instead of 1.27). Sorry for bothering you all! Regards, Igor P.S. Is there anybody out there who has a *good* C++-mode for Emacs? I have hacked c-mode.el, but I'm not very happy with it. If anybody wants it, let me know. Igor Metz X400: metz@iam.unibe.ch Institut fuer Informatik ARPA: metz%iam.unibe.ch@relay.cs.net und angewandte Mathematik UUCP: ..!uunet!mcvax!iam.unibe.ch!metz Universitaet Bern Switzerland Phone: (+41) 31 65 49 02