mingw-users Mailing List for MinGW - Minimalist GNU for Windows

huangliang wrote:
> 
> Hi all. I write a program which opens many text files, do some
> operations and then close them. Each time the program opens one
> file, reads its contents and then closes it. I use a loop to deal
> with all files. But every time I run it, the program will terminate
> after it has processed a number of (for example 641) files throwing
> the error message "abnormal program termination". I am sure I have
> closed the opened file after having read it. Who can tell what's
> wrong with it?

To find out what's wrong, use standard debugging techniques such as:

1. Add statements to your program to print the values of variables
as your program runs. Make sure you flush each print statement,
because abnormal termination may prevent the system from doing that
for you (use something like fflush() or std::endl). The output will
show the state of your program right before it terminates.

2. Remove statements from your program until it does not terminate
abnormally. If it looks like
  for(int j = 0; j < 1000; ++j) {
    // (a) open a file
    // (b) read file contents into a variable
    // (c) perform some operation that changes that variable
    // (d) store that changed variable somewhere
    }
then remove first (d), then (c), then (b), then (a). The last
statement you had to remove may be the problem.

3. Follow the advice Alberto Luaces <al...@na...> gave in
his 27 Dec 2003 13:52:17 +0100 reply to your other message:

| Hello, what information does the debugger provides? (Compile with
| '-g'  option, run 'gdb your_exe.exe', type 'run' command and wait
| until it crashes; then type 'bt' to inspect the stack frame and
| know what function is the culprit).

4. I suspect it doesn't actually crash: the message
  abnormal program termination
could come from a function such as abort(), as in this program:
  #include <cstdlib>
  void bar() {std::abort();}
  void foo() {bar();}
  int main() {foo();}
Still, Alberto's advice helps you:
  C:/tmp[0]$/MinGW/bin/g++ -ggdb -O0 abnormal.cpp
  C:/tmp[0]$/gdb/bin/gdb.exe --quiet a.exe
  (gdb) break 'abort'
  Breakpoint 1 at 0x401800
  (gdb) r
  Starting program: C:\tmp/a.exe

  Breakpoint 1, 0x00401800 in abort ()
  (gdb) bt
  #0  0x00401800 in abort ()
  #1  0x004012db in bar() () at abnormal.cpp:2
  #2  0x004012e9 in foo() () at abnormal.cpp:3
  #3  0x0040130f in main () at abnormal.cpp:4

Of course, you probably don't call abort() explicitly, but it
may be called by some library routine that you use, such as
assert(). You may need to set breakpoints on similar functions
like std::terminate().

Hi,

I write a simple Dll COM server example with MingW, http://www.anzwers.org/free/devlily/, they include most basic problems i have encoutered and you might encouter too.

Happy new years,
Bobber Cheng

__________________________________________________________________
New! Unlimited Access from the Netscape Internet Service.
Beta test the new Netscape Internet Service for only $1.00 per month until 3/1/04.
Sign up today at http://isp.netscape.com/register
Act now to get a personalized email address!

Netscape. Just the Net You Need.
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Manu wrote:
> Stephan T. Lavavej wrote:
> 
> 
> 
>>[Manu]
>>
>>>worse, the program's output doesn't show up at the right
>>>time, but just before the program exits.
>>
>>>printf("Hello world !");
>>
>>You are not printing a newline.  Thus, printf does not need to flush the
>>buffer.  The buffer will be flushed upon successful program termination.
>>
>>If you ask printf to print a newline, it will flush its buffer (this is
>>different behavior from cout).
> 
> 
> Same behavior with or without a newline, within MSYS or with redirected
> I/O. OTOH, it works as expected in a DOS box.
> 

Search for ``flushall'' in the search box at www.mingw.org.  It should 
answer this question.

> What is so different between MinGW's GDB 5.2.1 and 6.0?
> Was it configured differently or is it GDB itself?
> 

See Chris Johns' /mingw/doc/gdb/PORTNOTES and /mingw/doc/gdb/PORTPATCH.

Earnie.
-- 
http://www.mingw.org
http://sourceforge.net/projects/mingw
https://sourceforge.net/donate/index.php?user_id=15438

GC>   http://gcc.gnu.org/ml/java-patches/2003-q4/msg00430.html
Greate thanks for that link. I found what I need at
http://libunicows.sourceforge.net/ mentioned there. 

Roman Miklashesvky wrote:
> 
> As you know, this dll allows unicode programs to work under Windows
> 9x/ME. But mingw distribution does not include an .a file for this
> dll. Also I failed to create it myself.

Try showing what you tried that did not work. That may make it
easier for someone else to help.

> And the second question is: why libunicows.a file was not included in
> mingw?

Perhaps because of the library's license:
  http://gcc.gnu.org/ml/java-patches/2003-q4/msg00430.html

As you know, this dll allows unicode programs to work under Windows
9x/ME. But mingw distribution does not include an .a file for this
dll. Also I failed to create it myself. This dll exports a lot of
fuctions and there for it is not possible to create .def file
manually. What can I do?
And the second question is: why libunicows.a file was not included in
mingw? 

Hello. I have got anothet related question. Shouldn't that variable be declared as volatile?
I mean to avoid all possible optimization issues... you know.

---
Jan Ringos, Tr...@mx...
http://tringi.mx-3.cz

senior programmer-analyst
n.v.t. MX-3, www.mx-3.cz



----- Original Message ----- 
From: "Samuel Bandara" <sam...@ur...>
To: <min...@li...>
Sent: Monday, December 29, 2003 5:23 PM
Subject: [Mingw-users] Re: __declspec(allocate("myshare"))


> Success! The GCC manual describes the following windows-specific feature on:
> http://gcc.gnu.org/onlinedocs/gcc-3.3.2/gcc/Variable-Attributes.html.
> Defining
> 
> int foobar __attribute__((section ("fooshare"), shared)) = 42;
> 
> like that, the same value is shared by all instances of my DLL, no matter in
> which process context the module is run. Luke, thank you very much.
> 
> Samuel
> 
> > Try looking in the GCC manual for __attribute__((shared)) and also search
> > this mailing list for examples.
> >
> > Luke
> 
> > > in a hook procedure called back by the system - in the context of the
> > > process that raised the respective event, however - I need to access
> > > values stored in the context of my own process. Someone advised me
> > > to use the following construction:
> > >
> > > #pragma data_seg("fooshare")
> > > __declspec(allocate("fooshare")) int foobar = 42;
> > > #pragma data_seg()
> > > #pragma comment(linker, "/SECTION:fooshare,RWS")
> > >
> > > With Visual C++ this works fine, with MinGW it does not. In fact, it's
> > > Microsoft specific. MinGW just warns me that it ignores 'allocate', and
> my
> > > process and the hooked process see different values. Can anybody give a
> > > hint, please?
> 
> 
> 
> 
> -------------------------------------------------------
> This SF.net email is sponsored by: IBM Linux Tutorials.
> Become an expert in LINUX or just sharpen your skills.  Sign up for IBM's
> Free Linux Tutorials.  Learn everything from the bash shell to sys admin.
> Click now! http://ads.osdn.com/?ad_id=1278&alloc_id=3371&op=click
> _______________________________________________
> MinGW-users mailing list
> Min...@li...
> 
> You may change your MinGW Account Options or unsubscribe at:
> https://lists.sourceforge.net/lists/listinfo/mingw-users
> 

Success! The GCC manual describes the following windows-specific feature on:
http://gcc.gnu.org/onlinedocs/gcc-3.3.2/gcc/Variable-Attributes.html.
Defining

int foobar __attribute__((section ("fooshare"), shared)) = 42;

like that, the same value is shared by all instances of my DLL, no matter in
which process context the module is run. Luke, thank you very much.

Samuel

> Try looking in the GCC manual for __attribute__((shared)) and also search
> this mailing list for examples.
>
> Luke

> > in a hook procedure called back by the system - in the context of the
> > process that raised the respective event, however - I need to access
> > values stored in the context of my own process. Someone advised me
> > to use the following construction:
> >
> > #pragma data_seg("fooshare")
> > __declspec(allocate("fooshare")) int foobar = 42;
> > #pragma data_seg()
> > #pragma comment(linker, "/SECTION:fooshare,RWS")
> >
> > With Visual C++ this works fine, with MinGW it does not. In fact, it's
> > Microsoft specific. MinGW just warns me that it ignores 'allocate', and
my
> > process and the hooked process see different values. Can anybody give a
> > hint, please?

Thanx Manu, that got both problems.

For the combobox, I misinterpreted the "height" parameter to be the
displayed height of the combobox control on the dialog box when the
selection list is not active.  This is not spelled out very clearly in
Microsoft's nor anybody else's documentation on the combobox.

As for the tabbing problem, I was confused as the the use of the
accellerator in the code, I originally referenced "hAccelTable" to the
dialog box and then simplified it to:

while (GetMessage(&Msg,NULL,0,0)){
    if ((!IsWindow(hwnd)) || (!IsDialogMessage(hwnd,&Msg))){
        TranslateMessage(&Msg);
        DispatchMessage(&Msg);
    }
}

Without the accellerator, the function works great - as you pointed out the
"!IsDialogMessage(...)" function seems to be important factor.

Thanx for your help on this,

myke

A few notes:

Myke Predko wrote:

[...]
>   COMBOBOX ID_PORT, 78, 2, 70, 14, CBS_DROPDOWN | WS_TABSTOP
                                                                ^^ you should set the height of
                                                                     the dropdown list.
                                                                     eg: 78, 2, 70, 60

[...]
> Along with this, when I press the "Tab"
> key, the focus does not change from button, or control to the next one.
[...]

Your message loop should be like:

while (GetMessage(&Msg,NULL,0,0)){
    if (!TranslateAccelerator(hwnd,hAccelTable,&Msg)){
        if ((!IsWindow(hwnd)) || (!IsDialogMessage(hwnd,&Msg))){
                                                ^^^^^^^^^^^^^^
                                                process dialog messages
            TranslateMessage(&Msg);
            DispatchMessage(&Msg);
        }
    }
}


IsDialogMessage:
The IsDialogMessage function determines whether a message is intended
for the specified dialog box and, if it is, processes the message.

See a working example from FoosYerDoos:
http://www.foosyerdoos.fsnet.co.uk/MainFiles/DynaContainFiles/Code/Resources.html

Manu.

Hi Folks,

I originally sent this email to the cygwin user's group that suggested that
I send it to the mingw-users list - I am going to leave it in its entirety
even though it sounds like the problem is with how I am invoking gcc (from
cygwin and X-Windows bash).  This is my first question to the group, so
hopefully I will provide all the required information.

I'm attempting to build a small programmer interface for Win32 using gcc
under cygwin.  Along with gcc, I have been using gdb running under ddd for
symbolic debugging.

The application is to be based on a dialog box (form) type and so far, I
have put in some pushbuttons and a couple of combobox dropdown lists using
the resource file:

#include "windows.h"
#include "resource.h"

IDD_MYDIALOG DIALOG 6, 21, 153, 88
STYLE DS_MODALFRAME | WS_POPUP | WS_VISIBLE | WS_CAPTION | WS_SYSMENU
CAPTION "Programmer"
FONT 8, "Arial"
BEGIN
  DEFPUSHBUTTON "FILE LOAD", ID_FILE_OPEN, 3, 2, 70, 14, WS_TABSTOP
  DEFPUSHBUTTON "PROGRAM", ID_PROGRAM, 3, 19, 70, 14, WS_TABSTOP
  DEFPUSHBUTTON "VERIFY", ID_VERIFY, 3, 36, 70, 14, WS_TABSTOP
  DEFPUSHBUTTON "BLANK CHECK", ID_BLANK, 3, 53, 70, 14, WS_TABSTOP
  DEFPUSHBUTTON "QUIT", ID_QUIT, 3, 70, 70, 14, WS_TABSTOP
  COMBOBOX ID_PORT, 78, 2, 70, 14, CBS_DROPDOWN | WS_TABSTOP
  COMBOBOX ID_DEVICE, 78, 15, 70, 14, CBS_DROPDOWN | WS_TABSTOP
  LTEXT "Select File to Download into MCU", ID_STATUS, 78, 28, 70, 56,
WS_BORDER
END


To Update the "ID_PORT" comboboxes, I use the following WM_INITDIALOG code:

SendMessage(hdlgPortListCtrl, CB_ADDSTRING, 0, (LPARAM)(LPCTSTR)
TEXT("COM1:"));
SendMessage(hdlgPortListCtrl, CB_ADDSTRING, 0, (LPARAM)(LPCTSTR)
TEXT("COM2:"));
SendMessage(hdlgPortListCtrl, CB_ADDSTRING, 0, (LPARAM)(LPCTSTR)
TEXT("COM3:"));
SendMessage(hdlgPortListCtrl, CB_SETCURSEL, 0, 0);


The resource file and program file compile without any problems using the
following two statements:

windres Programmer.rc -O coff -o Programmer.res
gcc -mwindows -mno-cygwin -g -o Programmer.exe Programmer.c
Programmer.res -lgdi32 -luser32


When I click on the dropdown button on either one of the comboboxes, nothing
happens unless I double click in which case the initial text is highlighted
with a blinking cursor at the end.  Along with this, when I press the "Tab"
key, the focus does not change from button, or control to the next one.  As
I mention below, I have seen one instance of a problem with the combobox on
the list but none on the list indicate the issues with the problem with
tabbing.

To try and find the problem, I ran the resource file through Microsoft's
rc.exe and attempted to bind it to a compiled version of "Programmer.c"
(using the same gcc statement above with "Programmer.res" removed) using
rsrc but I end up with a "Checksum Field != 0" error (which I can't find any
explanation of).

I did a bit level compare on the windres produced .res file and the rc
produced .res file and found that in one of the combobox controls
(ID_DEVICE), the bit pattern was different:

rc:
da 07 ff ff 85 00 00 00 00 00 00 00 00 00 82 50 00 00 00 00 4e 00 1c 00 46
00 38 00
windres:
da 07 ff ff 85 00 00 00 00 00 00 00 00 00 80 50 00 00 00 00 4e 00 1c 00 46
00 38 00

The ID_PORT combobox control's bit pattern was identical between the two
versions.

I changed the corresponding windres .res file to the Microsoft rc .res file
and re-ran the gcc statement above with no difference.

I've just updated my tools to the latest levels from cygwin.com and when I
check the versions of windres and gcc, I get:

$ windres --version
GNU windres 2.14.90 20030901
Copyright 2002 Free Software Foundation, Inc.
This program is free software; you may redistribute it under the terms of
the GNU General Public License.  This program has absolutely no warranty.

$ gcc --version
gcc (GCC) 3.3.1 (cygming special)
Copyright (C) 2003 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.


Looking through the (cygwin) archives, I've noticed that Danny Smith,
earlier this
year, has encountered a problem and a potential patch for the windres
processing the combobox under the thread:

"Fix a windres (rcparse.y) parsing bug for controls without text fields"


Hopefully somebody can:
1.  Tell me this is a problem with windres
2.  Point out what I am doing wrong with the windres or gcc specification
3.  Explain what I'm doing wrong with rc.exe so I can dispense with windres
altogether

Thanx for your consideration and help!

myke

Read and respond to this message at: 
https://sourceforge.net/forum/message.php?msg_id=2350706
By: j0h0

thanks once again
greets,     witti

______________________________________________________________________
You are receiving this email because you elected to monitor this forum.
To stop monitoring this forum, login to SourceForge.net and visit: 
https://sourceforge.net/forum/unmonitor.php?forum_id=286529

Read and respond to this message at: 
https://sourceforge.net/forum/message.php?msg_id=2350698
By: derekbaker

Put it in the declaration only, not there and also in the definition.


Derek


______________________________________________________________________
You are receiving this email because you elected to monitor this forum.
To stop monitoring this forum, login to SourceForge.net and visit: 
https://sourceforge.net/forum/unmonitor.php?forum_id=286529

Read and respond to this message at: 
https://sourceforge.net/forum/message.php?msg_id=2350685
By: j0h0

hi there!
i tried to use a default argument in a function and i can´t figure out what´s
wrong. (sorry once again a newbee´s question)

int MidiFile::ReadIntInFileText(int start, int count = 1) { /*....*/ }

the error message is:
MidiFile.cpp:70: default argument given for parameter 2 of `int 
   MidiFile::ReadIntInFileText(int, int = 1)'

MidiFile.h:26: after previous specification in `int 
   MidiFile::ReadIntInFileText(int, int = 1)'

thanks for helping me! *g*

______________________________________________________________________
You are receiving this email because you elected to monitor this forum.
To stop monitoring this forum, login to SourceForge.net and visit: 
https://sourceforge.net/forum/unmonitor.php?forum_id=286529

In that case you installed GCC into the MSYS directory, which is not
allowed. See README.rtf and MSYS_WELCOME.rtf

Luke

----- Original Message ----- 
From: "Michael Ribbons" <et...@ho...>
To: <min...@li...>
Sent: Sunday, December 28, 2003 4:40 AM
Subject: Re: [Mingw-users] Base MingW


>
> > You specified no input files. This is a command-line compiler.
> > Consult the manual at
> >   http://gcc.gnu.org/
> >
>
> In fact I did specify input files, perhaps I should have specified the
> command I used ;) Actually gcc.exe is working ok from within cmd.exe.

> You specified no input files. This is a command-line compiler.
> Consult the manual at
>   http://gcc.gnu.org/
>

In fact I did specify input files, perhaps I should have specified the
command I used ;) Actually gcc.exe is working ok from within cmd.exe.

Michael Ribbons wrote:
> 
> I just need a simple c++ environment that I can use to do some of the very
> simple (base) c++ things, File I/O, cout & cin, possibly sockets.
> 
> Which packages will I need for this,

Go here
  http://sourceforge.net/project/showfiles.php?group_id=2435
and get the latest MinGW*.exe, which is probably the file
with the greatest number of downloads.

> and what is the process for
> installing them?

Just run the program you downloaded. It installs gcc etc.

> Currently I have a gcc.exe but when i run gcc in the MSYS
> environment, I get
> gcc.exe: no input files

You specified no input files. This is a command-line compiler.
Consult the manual at
  http://gcc.gnu.org/

Try looking in the GCC manual for __attribute__((shared)) and also search
this mailing list for examples.

Luke

----- Original Message ----- 
From: "Samuel Bandara" <sam...@ur...>
To: <min...@li...>
Sent: Saturday, December 27, 2003 9:52 PM
Subject: [Mingw-users] __declspec(allocate("myshare"))


> Hello to you all,
>
> in a hook procedure called back by the system - in the context of the
> process that raised the respective event, however - I need to access
values
> stored in the context of my own process. Someone advised me to use the
> following construction:
>
> #pragma data_seg("fooshare")
>
> __declspec(allocate("fooshare")) int foobar = 42;
>
> #pragma data_seg()
> #pragma comment(linker, "/SECTION:fooshare,RWS")
>
> With Visual C++ this works fine, with MinGW it does not. In fact, it's
> Microsoft specific. MinGW just warns me that it ignores 'allocate', and my
> process and the hooked process see different values. Can anybody give a
> hint, please?
>
> Samuel

Hello to you all,

in a hook procedure called back by the system - in the context of the
process that raised the respective event, however - I need to access values
stored in the context of my own process. Someone advised me to use the
following construction:

#pragma data_seg("fooshare")

__declspec(allocate("fooshare")) int foobar = 42;

#pragma data_seg()
#pragma comment(linker, "/SECTION:fooshare,RWS")

With Visual C++ this works fine, with MinGW it does not. In fact, it's
Microsoft specific. MinGW just warns me that it ignores 'allocate', and my
process and the hooked process see different values. Can anybody give a
hint, please?

Samuel

Hello, what information does the debugger provides? (Compile with '-g' 
option, run 'gdb your_exe.exe', type 'run' command and wait until it 
crashes; then type 'bt' to inspect the stack frame and know what function 
is the culprit).

Alberto

El Sáb 27 Dic 2003 06:35, huangliang escribió:
> I wrote a program to parse MIME emails and I use this program to
> automatically parse and print emails in a directory in which I place
> about 1000 emails. Every time it crashed after it has parsed about 500
> emails. The windows show me the error like "you can't access memory at
> 0xXXXXXXXX". My program parse every single email well, and I really
> don't know what's the problem is? Anyone can give me some advice?

If you downloaded Mingw, as it seems, you have all what you need. 'gcc' is 
a command-line compiler, not an IDE. There are several IDEs for Mingw on 
the Net: Visual-Mingw, Dev-C++, ...

Alberto

El Sáb 27 Dic 2003 04:29, Michael Ribbons escribió:
> Hi,
> I just need a simple c++ environment that I can use to do some of the
> very simple (base) c++ things, File I/O, cout & cin, possibly sockets.
>
> Which packages will I need for this, and what is the process for
> installing them? Currently I have a gcc.exe but when i run gcc in the
> MSYS environment, I get
> gcc.exe: no input files
>
> Also, the only make utility I seem to have is the one that comes with
> MSYS.
>
> Is MingW what I need? Cygwin seems far too big for my purposes.
>
> Thanks
> Michael Ribbons
>

Thank you very much, Manu. Your code clarified things a lot. 
PeekNamedPipe() was what I was looking for, because at first I incorrectly 
thought that ReadFile would wait until all the data were sent to the pipe.

Alberto

El Vie 26 Dic 2003 16:46, Manu escribió:
> Well, I can share a bit of my experience, this is how I proceed in
> Visual-MinGW:
> CPipe::_peek() calls PeekNamedPipe to know if we have
> something to read, then CPipe::_read() fills the output buffer
> (COutBuffer).
>
> You can freely use the code in /libtemp, it is currently GPL'ed,
> but it will be moved into WinUI library soon, then it will be switched
> to public domain. As the copyright holder, I give you acknowledgement
> to use it in your app, in case it is helpfull.
> http://cvs.sourceforge.net/viewcvs.py/visual-mingw/visual-mingw/src/libt
>emp/
>
> A Google search with "PeekNamedPipe", "ReadFile", etc... will give
> you much more inputs.
>
> Enjoy!
>
> Manu.

SSB3cm90ZSBhIHByb2dyYW0gdG8gcGFyc2UgTUlNRSBlbWFpbHMgYW5kIEkgdXNlIHRoaXMgcHJv
Z3JhbSB0byBhdXRvbWF0aWNhbGx5IHBhcnNlIGFuZCBwcmludCBlbWFpbHMgaW4gYSBkaXJlY3Rv
cnkgaW4gd2hpY2ggSSBwbGFjZSBhYm91dCAxMDAwIGVtYWlscy4gRXZlcnkgdGltZSBpdCBjcmFz
aGVkIGFmdGVyIGl0IGhhcyBwYXJzZWQgYWJvdXQgNTAwIGVtYWlscy4gVGhlIHdpbmRvd3Mgc2hv
dyBtZSB0aGUgZXJyb3IgbGlrZSAieW91IGNhbid0IGFjY2VzcyBtZW1vcnkgYXQgMHhYWFhYWFhY
WCIuIE15IHByb2dyYW0gcGFyc2UgZXZlcnkgc2luZ2xlIGVtYWlsIHdlbGwsIGFuZCBJIHJlYWxs
eSBkb24ndCBrbm93IHdoYXQncyB0aGUgcHJvYmxlbSBpcz8gQW55b25lIGNhbiBnaXZlIG1lIHNv
bWUgYWR2aWNlPw==

Hi,
I just need a simple c++ environment that I can use to do some of the very
simple (base) c++ things, File I/O, cout & cin, possibly sockets.

Which packages will I need for this, and what is the process for
installing them? Currently I have a gcc.exe but when i run gcc in the MSYS
environment, I get
gcc.exe: no input files

Also, the only make utility I seem to have is the one that comes with MSYS.

Is MingW what I need? Cygwin seems far too big for my purposes.

Thanks
Michael Ribbons