|
The _int64 data type is compiler specific and might not port. However, gcc supports a data type such as LONG LONG.
If you are looking at porting code to other platforms, then anything that must be size specific should be done with typedefs or defines.
Example:
#if defined (WIN32)
typedef _int64 INT64
typedef unsigned _int64 UINT64
#elif define (WHATEVER_LINUX)
typedef long long INT64
typedef unsigned long long UINT64
#endif
Then in your code only use INT64 or UINT64.
Tim Smith
I'm going to patent thought. I have yet to see any prior art.
|
|
|
|
|
Thanks, Tim buddy!
George
|
|
|
|
|
First _int64 is Microsofht specific.
Second _int64 represents a 64-bit interger.
If you do a search on MSDN, for _int64, you will find at lease 31 seperate entries. If you search for "microsoft _int64" on google you will find over 200 entries, many of which appear to do with porting. I know of no other sources of information on _int64.
Trust in the code Luke. Yea right!
|
|
|
|
|
Thanks, John buddy!
George
|
|
|
|
|
Hi
I'm inserting all kind of columns in a CListCtrl
with structure LV_COLUMN like this:
ColumnVi = new LV_COLUMN();
CString Title;
ColumnVi->mask = LVCF_TEXT | LVCF_WIDTH;
ColumnVi->fmt = LVCFMT_CENTER;
//Inserting
ColumnVi->cx = 35;
Title = "Nr.";
ColumnVi->pszText = Title.GetBuffer(3); m_PersCtrl.InsertColumn(0,ColumnVi);
//Insert the column into the clistctrl
The width of column i've set to 35, i tested it out and changed the value until it looked good. but the caption can change at some time .. so how can i know how much pixels 1 char. takes or anyone have other ideas to make it more dynamic?
Greetings Jens
|
|
|
|
|
You may combine GetStringWidth and SetColumnWidth of CListCtrl.
rechi
|
|
|
|
|
get the HDC (or CDC if you are using MFC) and then use the GetTextExtent[^] api call to determine the string length
¡El diablo está en mis pantalones! ¡Mire, mire!
Real Mentats use only 100% pure, unfooled around with Sapho Juice(tm)!
|
|
|
|
|
I have a client reporting that some of my code crashes on a few of his machines; I can't reproduce the problem on my end (what else is new) and all I have to work with is the crash address he's reporting.
I've had a look at Wouter Dhondt's excellent article on .MAP files[^]...and if I'm following the steps properly, the crash occurs in a memcpy() call...great, according to a search in my project's .cpp files, I'm calling it explicitely 74 times. Since I can't reproduce the crash there's hardly any point in stepping through all of them. Besides, if I understand correctly, the call to memcpy() might not necessarily be any of my own (ie, another C runtime library function might be doing it; who knows which one it is).
Taking it to the next step, I tried to determine what file/line the crash occurs in according to the .MAP file. Again based on the article, subtracting the load address (0x00400000) and the 0x1000 bytes PE header from the crash address (0x00459083), I end up with 0x00058xxx, which is nowhere near any of the addresses listed in the bottom part of the .MAP file. And since I'm within 0x004xxxxx, I'm pretty confident the crash doesn't occur in a DLL. I *know* I'm using the proper .MAP file as I've rebuilt the project (this time set to generate the .MAP file) and sent the client the recompiled EXE (which turned out to be identical to the previous one anyway).
I've tried to get the client to send me Dr. Watson logs and crash dumps, but none seem to ever get generated despite the options being selected (he's on Win2K SP2).
What next? I don't like the idea of sending a client an EXE with MsgBox() calls...
|
|
|
|
|
Run your release version in debug and and the break the program. Use the debugger to see if you have anything at the address in question. If your release program doesn't have any debug information, it would help to enable the program database debug information.
Tim Smith
I'm going to patent thought. I have yet to see any prior art.
|
|
|
|
|
Well...I have *something*, I'm just not sure what it is. Wouldn't this (the memory at the crash address) be in the code segment?
|
|
|
|
|
The address you listed does seem like an IP, however if I remember the access violation dialog box, it lists a few numbers including the address being referenced and the IP of the instruction.
I just like using the debugger since reading map files can be confusing a hell. I have been doing it for 20 years and I still get confused.
Tim Smith
I'm going to patent thought. I have yet to see any prior art.
|
|
|
|
|
> if I remember the access violation dialog box, it lists a few numbers
> including the address being referenced and the IP of the instruction
That's part of the problem, I don't even get that much info on the crash dialog. Here's the full text (and I just realized I was looking at an old screenshot when I made the original post):
>>>
The instruction at "0x00459083" referenced memory at "0x00c70000". The memory could not be "read".
<<<
According to the .MAP file, the closest function (at 0x004590e0) is memcpy(), and 0x000580e0 isn't part of the .MAP file. So I don't know what to do next...
(I'd kill for a stack trace)
|
|
|
|
|
When you are looking at the map file, ignore the first column. Just take a look at the 3rd column. On an NT system, this will be the load address for the main image.
Tim Smith
I'm going to patent thought. I have yet to see any prior art.
|
|
|
|
|
Yeah, the one entitled "rva+base". Given my crash address (0x00459113), here's what I have:
(...)
0001:00058080 __mbsrchr 00459080 f LIBC:mbsrchr.obj
0001:000580e0 _memcpy 004590e0 f LIBC:memcpy.obj
0001:00058415 __onexit 00459415 f LIBC:onexit.obj
(...)
...so the crash seems to occur in memcpy().
I understand the mapping back to the file/line number in Wouter's article...however in my case I don't have anything close to 0x00058113 (or close to that range) in the section where the .MAP file iterates through each source file (in the bottom half of the .MAP file). I'm wondering if there might be another offset I have to take into account that wasn't discussed in the article...my assumption is that I should be able to find an entry similar to:
nnn 0001:00058xxx
...where nnn is the line number in the source file being mapped.
I hope my description is a little clearer.
|
|
|
|
|
The program is crashing in memcpy. You won't have any source lines associated with that.
Tim Smith
I'm going to patent thought. I have yet to see any prior art.
|
|
|
|
|
Not even a reference to which instance of my memcpy() calls causes the crash? 
Back to my original post then...given this, what are my options? 
|
|
|
|
|
Hi, everyone!
I want to know where can find a nice online
tutorial about raw socket programming on windows
platform?
Thanks in advance,
George
|
|
|
|
|
How raw do you want it? In terms of winsock programming, one excellent reference is CodeProject.
http://www.codeproject.com/internet/
Kuphryn
|
|
|
|
|
Thanks, kuphryn buddy!
George
|
|
|
|
|
Here's a little sump'n sump'n:
http://msdn.microsoft.com/library/default.asp?url=/library/en-us/winsock/winsock/tcp_ip_raw_sockets_2.asp
|
|
|
|
|
Thanks, Philnessosity buddy!
George
|
|
|
|
|
You're welcome, George2 buddy!
|
|
|
|
|
Hi !
I want to create array of objects of a certain class like:
CMyclass object[12];
But i want to intantiate these objects using one argument constructor of CMyClass. Is this possible or i've to create objects on heap ???
|
|
|
|
|
Initialize the array in the following way:
CMyclass object[12] = { CMyclass(argument), CMyclass(argument), CMyclass(argument), ..., CMyclass(argument) };
|
|
|
|
|
or, if you use stl,
std::vector<cmyclass> objects;
objects.resize( 12, CMyClass(argument) );
¡El diablo está en mis pantalones! ¡Mire, mire!
Real Mentats use only 100% pure, unfooled around with Sapho Juice(tm)!
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
