www.digitalmars.com Home | Search | C & C++ | D | DMDScript | News Groups | index | prev | next
Archives
D Programming
D
D.gnu
digitalmars.D
digitalmars.D.bugs
digitalmars.D.dtl
digitalmars.D.dwt
digitalmars.D.announce
digitalmars.D.learn
digitalmars.D.debugger
C/C++ Programming
c++
c++.announce
c++.atl
c++.beta
c++.chat
c++.command-line
c++.dos
c++.dos.16-bits
c++.dos.32-bits
c++.idde
c++.mfc
c++.rtl
c++.stl
c++.stl.hp
c++.stl.port
c++.stl.sgi
c++.stlsoft
c++.windows
c++.windows.16-bits
c++.windows.32-bits
c++.wxwindows
digitalmars.empire
digitalmars.DMDScript

c++ - DM 8.23

↑ ↓ ← → Roland <rv ronetech.com> writes:
trying DM 8.23 compiler in DOSX model.

i only replaced bin\SCPPN.EXE and bin\SCPPND.EXE

i had

typedef enum boolean {false,true};

in a header.
it does not compile any more: "identifier expected"

not a big problem, just for information

Roland
Nov 15 2001
↑ ↓ ← → Roland <rv ronetech.com> writes:
Roland a écrit :


 trying DM 8.23 compiler in DOSX model.

 i only replaced bin\SCPPN.EXE and bin\SCPPND.EXE



- i mean bin\SCPPND.DLL

- sorry outpw, inpw are still 32 bit contrary to what compiler.htm says

Regards

Roland
Nov 15 2001
↑ ↓ ← → "Walter" <walter digitalmars.com> writes:
"Roland" <rv ronetech.com> wrote in message
news:3BF39B52.6D08BC5F ronetech.com...

 - sorry outpw, inpw are still 32 bit contrary to what compiler.htm says



I thought I tested & verified it. Could you post an example? -Walter
Nov 15 2001
↑ ↓ ← → Roland <rv ronetech.com> writes:
Walter a écrit :


 "Roland" <rv ronetech.com> wrote in message
 news:3BF39B52.6D08BC5F ronetech.com...

 - sorry outpw, inpw are still 32 bit contrary to what compiler.htm says



 I thought I tested & verified it. Could you post an example? -Walter



little function

int _testbug() {
    int i;
    i1 = inpw(0x03ce);    //16 bit
    outpw(0x03ce,i);        //16 bit
    return i;
}

unassembles

?_testbug  YAHXZ:
  mov EDX,03CEh
  in EAX,DX                        ;32 bit !
  mov ECX,EAX
  mov EAX,ECX
  out DX,EAX                    ;32 bit !
  mov EAX,ECX
  ret

Regard

Roland
Nov 16 2001
↑ ↓ ← → Roland <rv ronetech.com> writes:
Roland a écrit :


 int _testbug() {
     int i;
     i1 = inpw(0x03ce);    //16 bit
     outpw(0x03ce,i);        //16 bit
     return i;
 }

 unassembles

 ?_testbug  YAHXZ:
   mov EDX,03CEh
   in EAX,DX                        ;32 bit !
   mov ECX,EAX
   mov EAX,ECX
   out DX,EAX                    ;32 bit !
   mov EAX,ECX
   ret

 Regard

 Roland



i forgot: model is DOSX (-mx), i use the idde

Roland
Nov 16 2001
↑ ↓ ← → "Walter" <walter digitalmars.com> writes:
ok, I'll check it out. -Walter

"Roland" <rv ronetech.com> wrote in message
news:3BF4DC3B.3661472C ronetech.com...

 Roland a écrit :


 int _testbug() {
     int i;
     i1 = inpw(0x03ce);    file://16 bit
     outpw(0x03ce,i);        file://16 bit
     return i;
 }

 unassembles

 ?_testbug  YAHXZ:
   mov EDX,03CEh
   in EAX,DX                        ;32 bit !
   mov ECX,EAX
   mov EAX,ECX
   out DX,EAX                    ;32 bit !
   mov EAX,ECX
   ret

 Regard

 Roland



 i forgot: model is DOSX (-mx), i use the idde

 Roland
Nov 16 2001
↑ ↓ ← → "John Culver" <jculver btinternet.com> writes:
Hi,
        I can generate this too, but only by using a conio.h to provide the
prototypes
for inp*/outp*:

====== CODE GENERATES inpw/outpw as 16 bit
#include <stdint.h>
// #include <dos.h>
#include <conio.h>

const baseAddress = 0x300;

int main (int argc, char * argv[])
{
  uint8_t  i8;
  uint16_t i16;
  uint32_t i32;

 // Generate 3 inputs
  i8  = inp (baseAddress);
  i16 = inpw(baseAddress);
  i32 = inpl(baseAddress);

 // generate 3 outputs
  outp (baseAddress, i8);
  outpw(baseAddress, i16);
  outpl(baseAddress, i32);

  return 0;
}
====== END of CODE GENERATES inpw/outpw as 16 bit

This variant below (the one I've used) works 100%, fully inline and
_inpw/_outpw are 16
bit with -mx.
The changes are the _ before the inp* and outp* and the use of dos.h rather than
conio.h :

====== CODE that works
#include <stdint.h>
#include <dos.h>
// #include <conio.h>

const baseAddress = 0x300;

int main (int argc, char * argv[])
{
  uint8_t  i8;
  uint16_t i16;
  uint32_t i32;

 // Generate 3 inputs
  i8  = _inp (baseAddress);
  i16 = _inpw(baseAddress);
  i32 = _inpl(baseAddress);

 // generate 3 outputs
  _outp (baseAddress, i8);
  _outpw(baseAddress, i16);
  _outpl(baseAddress, i32);

  return 0;
}
====== END of CODE that works

======== build lines ...
sc -mx -c -C ioTest.Cpp
Obj2Asm ioTest.Obj >Fixed.Dmp
========

The problem appears to be the conio.h on my system, The 'offending' conio.h on
my
system is :
    CONIO    H           8,091  03-17-01 12:15a CONIO.H
I've not checked the exact nature of the inp/outp prototypes in there, but I
assume
that they are substantially different to the updated ones in the dos.h that
appeared
with the 8.23 Beta.


John Culver



"Roland" <rv ronetech.com> wrote in message
news:3BF4DB59.9B791A69 ronetech.com...

 Walter a écrit :


 "Roland" <rv ronetech.com> wrote in message
 news:3BF39B52.6D08BC5F ronetech.com...

 - sorry outpw, inpw are still 32 bit contrary to what compiler.htm says



 I thought I tested & verified it. Could you post an example? -Walter



 little function

 int _testbug() {
     int i;
     i1 = inpw(0x03ce);    //16 bit
     outpw(0x03ce,i);        //16 bit
     return i;
 }

 unassembles

 ?_testbug  YAHXZ:
   mov EDX,03CEh
   in EAX,DX                        ;32 bit !
   mov ECX,EAX
   mov EAX,ECX
   out DX,EAX                    ;32 bit !
   mov EAX,ECX
   ret

 Regard

 Roland
Nov 16 2001
↑ ↓ ← → "John Culver" <jculver btinternet.com> writes:
Hi.
    I've just patched my DM up to the current code set cd81d, then cd823, and
the
incorrect code generation with conio.h remains.


John Culver
Nov 16 2001
↑ ↓ ← → "Laurentiu Pancescu" <plaur crosswinds.net> writes:
"Roland" <rv ronetech.com> wrote in message
news:3BF396E2.CB83AF95 ronetech.com...

 trying DM 8.23 compiler in DOSX model.

 i only replaced bin\SCPPN.EXE and bin\SCPPND.EXE

 i had

 typedef enum boolean {false,true};

 in a header.
 it does not compile any more: "identifier expected"



It should work, and it does for me (I also used 8.23, to be sure).  In C,
not in C++ (true and false are keywords in ISO-C++, they are the only
acceptable values for the bool type).  I wonder why this worked until
now, -Ab is default for some time (8.20, at least, if not sooner).  Probably
your header is included from a C++ file, and the compiler correctly
complains about misuse of a keyword instead of an (expected) identifier.

Regards,
  Laurentiu
Nov 15 2001
↑ ↓ ← → "Walter" <walter digitalmars.com> writes:
This compiles correctly with C, and fails correctly with C++ (since false
and true are now C++ keywords). -Walter

"Roland" <rv ronetech.com> wrote in message
news:3BF396E2.CB83AF95 ronetech.com...

 trying DM 8.23 compiler in DOSX model.

 i only replaced bin\SCPPN.EXE and bin\SCPPND.EXE

 i had

 typedef enum boolean {false,true};

 in a header.
 it does not compile any more: "identifier expected"

 not a big problem, just for information

 Roland
Nov 15 2001
↑ ↓ ← → Roland <rv ronetech.com> writes:
okay, thanks

Roland

Walter a écrit :


 This compiles correctly with C, and fails correctly with C++ (since false
 and true are now C++ keywords). -Walter

 "Roland" <rv ronetech.com> wrote in message
 news:3BF396E2.CB83AF95 ronetech.com...

 trying DM 8.23 compiler in DOSX model.

 i only replaced bin\SCPPN.EXE and bin\SCPPND.EXE

 i had

 typedef enum boolean {false,true};

 in a header.
 it does not compile any more: "identifier expected"

 not a big problem, just for information

 Roland
Nov 16 2001