www.digitalmars.com         C & C++   DMDScript  

digitalmars.D.bugs - [Issue 814] New: lazy argument + variadic arguments

reply d-bugmail puremagic.com writes:
http://d.puremagic.com/issues/show_bug.cgi?id=814

           Summary: lazy argument + variadic arguments
           Product: D
           Version: 1.00
          Platform: PC
        OS/Version: Linux
            Status: NEW
          Severity: normal
          Priority: P2
         Component: DMD
        AssignedTo: bugzilla digitalmars.com
        ReportedBy: ville.mattila iki.fi


Hello,

 I'm very new to d. I'm testing lazy evaluation and it seems like
 I've found bug w/ dmd  compiler on linux. Here are the details.


-- foo.d ---

import std.stdio;
import std.stdarg;
import std.string, std.conv, std.stream;
void foo(lazy void expression, ...)
{
  //expression();
  if (_arguments.length > 0) {
    _arguments[0].print();
    //writefln("got vararg:" ~ _arguments[0].name);
    if ((_arguments[0]) == typeid(char[])) {
      writefln("got vararg:" ~ std.conv.toString(_arguments.length));
      char[] tmp_msg = va_arg!(char[])(_argptr);
      writefln("GOOTT vararg: " ~ tmp_msg);

    }
  }
}

int main(char[][] args)
{
  foo(writefln("barz"),"foo");
  return(0);
}

--- foo.d ----

./foo
char[]
got vararg:1
Segmentation fault (core dumped)
mulperi mulperi-desktop:/usr/local/src/apps/d/dmd/samples/d$ dmd --help
Digital Mars D Compiler v1.0
Copyright (c) 1999-2007 by Digital Mars written by Walter Bright
Documentation: www.digitalmars.com/d/index.html
Usage:
  dmd files.d ... { -switch }

I'm usring xubuntu 6.10 w/ latest patches.

 - Ville


-- 
Jan 07 2007
next sibling parent d-bugmail puremagic.com writes:
http://d.puremagic.com/issues/show_bug.cgi?id=814






Added to DStress as
http://dstress.kuehne.cn/run/l/lazy_04_A.d
http://dstress.kuehne.cn/run/l/lazy_04_B.d
http://dstress.kuehne.cn/run/l/lazy_04_C.d


-- 
Apr 05 2007
prev sibling next sibling parent d-bugmail puremagic.com writes:
http://d.puremagic.com/issues/show_bug.cgi?id=814


Don <clugdbug yahoo.com.au> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |clugdbug yahoo.com.au
            Summary|lazy argument + variadic    |lazy argument + variadic
                   |arguments                   |arguments = segfault





Reduced test case shows it's an out-by-1 error in grabbing the arguments. The
length ends up in the ptr field.
Probably, it's neglecting the fact that lazy arguments have a pointer as well
as the value, or something similar.

import std.stdarg;

void foo(lazy int expr, ...){
    char[] tmp_msg = va_arg!(char[])(_argptr);
    if (cast(int)(tmp_msg.ptr)=="food_for_thought".length)
         assert(0, "length is in the pointer!");
}

void main(){
  foo(515,"food_for_thought");
}

Applies to D2 as well as D1.

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
Jul 02 2009
prev sibling next sibling parent d-bugmail puremagic.com writes:
http://d.puremagic.com/issues/show_bug.cgi?id=814






In fact, by replacing the T arg line below with the commented out-line, it gets
the correct arguments. So in fact, it's just getting the size of the lazy
parameter wrong when it's calculating _argptr.
-------------
alias void* va_list;

template va_arg(T)
{
    T va_arg(inout va_list _argptr)
    {
    T arg = *cast(T*)_argptr; // original from Phobos
//    T arg = *(cast(T*)(cast(int *)_argptr+1));
    _argptr = _argptr + ((T.sizeof + int.sizeof - 1) & ~(int.sizeof - 1));
    return arg;
    }
}


void foo(lazy int expr, ...){
    char[] tmp_msg = va_arg!(char[])(_argptr);
    if (cast(int)(tmp_msg.ptr)=="food_for_thought".length)
         assert(0, "length is in the pointer!");
    assert(tmp_msg=="food_for_thought");

}

int bar() { return 3; }
void main(){
  foo(bar(),"food_for_thought");
}

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
Jul 02 2009
prev sibling next sibling parent d-bugmail puremagic.com writes:
http://d.puremagic.com/issues/show_bug.cgi?id=814


Don <clugdbug yahoo.com.au> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
           Keywords|                            |patch





PATCH: func.c, line 1211. Need to account for a lazy last parameter.


-    offset = p->type->size();
+    if (p->storage_class & STClazy) {
+    // If the last parameter is lazy, it's the size of a delegate
+    offset = sizeof(int(*)(int)) + sizeof(void *);
+    } else {
+    offset = p->type->size();
+    }

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
Jul 20 2009
prev sibling parent d-bugmail puremagic.com writes:
http://d.puremagic.com/issues/show_bug.cgi?id=814


Walter Bright <bugzilla digitalmars.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
                 CC|                            |bugzilla digitalmars.com
         Resolution|                            |FIXED



02:12:02 PDT ---
Fixed dmd 1.048 and 2.033

-- 
Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
Oct 06 2009