• manfred toppoint.de (56/56) Mar 24 2004 Hello,
• s31552 mail.ecc.u-tokyo.ac.jp (42/46) Mar 24 2004 I found the minimal code of this problem.
• s31552 mail.ecc.u-tokyo.ac.jp (19/19) Mar 24 2004 I'm sorry I post continuously.

manfred toppoint.de writes:

Hello, 
 
in file socket.d i get an error message: 
gdc -c socket.d 
socket.d:
In function `add': 
socket.d:0: interner Compiler-Fehler: Speicherzugriffsfehler
Please submit a full bug report, 
with preprocessed source if appropriate. 
See
<URL:http://gcc.gnu.org/bugs.html> for instructions. 
 
the function add start
in line  780 . 
 
Here is a code snip 
 
typedef int socket_t = -1; 
const
socket_t INVALID_SOCKET = socket_t.init; 
void add(socket_t s) 
//      /*
in 
        { 
                version(Win32) 
                {
assert(count < max); //added too many sockets; specify a higher max in the
constructor 
                } 
        } 
        body 
//        */ 
        {
version(Win32) 
                { 
                        uint c = count;
first[c] = s; 
                        count = c + 1; 
                }
else version(linux) 
                {
//bts(cast(uint*)&first[fdelt(s)], cast(uint)s % nfdbits); 
                }
else 
                { 
                        static assert(0);
}
} 
 
 
You can put this in a file and try to compile there appear the same error
message. 
I believe the problem is the "in" and "body" statements. 
 
Manfred 
 

s31552 mail.ecc.u-tokyo.ac.jp writes:

 I believe the problem is the "in" and "body" statements. 

I found the minimal code of this problem.

void bug()
in {
    version (Win32) { // it's safe when it is 'linux' (on Linux)
    }
}
body {
}

And I found another fatal bug:

void bug() {
    char[][int] a;
    int[] i = a.keys;
}

 gdc -c bug_assoc_key.d 

bug_assoc_key.d: In function `bug':
bug_assoc_key.d:0: internal compiler error: in emit_move_insn, at expr.c:3165
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://gcc.gnu.org/bugs.html> for instructions.

And not fatal bug:

void bug() {
    typedef double D;
    D d = cast(D)0.0;
}

 gdc -c bug_typedef_double_cast.d

bug_typedef_double_cast.d: In function `bug':
bug_typedef_double_cast.d:0: internal compiler error: Segmentation fault
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://gcc.gnu.org/bugs.html> for instructions.

And I cannot judge whether it's a bug or not:

void bug() {
    int d = 1;
    d /= 1.0;
}

 gdc -c bug_double_div.d

bug_double_div.d: In function `bug':
bug_double_div.d:0: internal compiler error: in expand_fix, at optabs.c:4905
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://gcc.gnu.org/bugs.html> for instructions.

------------------
 shinichiro.h
  s31552 mail.ecc.u-tokyo.ac.jp
  http://user.ecc.u-tokyo.ac.jp/~s31552/wp/

s31552 mail.ecc.u-tokyo.ac.jp writes:

I'm sorry I post continuously.

I found another fatal bug:

class Base {
        void bug() {}
}
class Derived : Base {
        void bug() { super.bug(); }    // call Derived::bug!
}
int main() {
        Base b = new Derived();
        b.bug();
        return 0;
}

It can be compiled and linked, but it fail with segmentation fault.
Because super.bug() call Derived::bug() recursively.

------------------
 shinichiro.h
  s31552 mail.ecc.u-tokyo.ac.jp
  http://user.ecc.u-tokyo.ac.jp/~s31552/wp/