When xp MCE is loaded and the kernel expands downward through memory or
atleast at some point an instance of the media player must be loaded into
memory. Does anyone happen to know the starting and end address of the media
player? Debug should be able to show something.
Bill

Bill Cunningham wrote:
> When xp MCE is loaded and the kernel expands downward through memory or
> atleast at some point an instance of the media player must be loaded into
> memory. Does anyone happen to know the starting and end address of the media
> player? Debug should be able to show something.
> Bill

I'd be surprised if ANY application loads into the same memory (virtual
OR actual) on two different occasions. In fact, with Window's use of the
page file an application's code and data segments might be non
continuous, out of sequence and even shift around as they are written to
the page file and read back into physical memory.

"RobertVA" <(E-Mail Removed)> wrote in message
news:%(E-Mail Removed)...
> I'd be surprised if ANY application loads into the same memory (virtual OR
> actual) on two different occasions. In fact, with Window's use of the page
> file an application's code and data segments might be non continuous, out
> of sequence and even shift around as they are written to the page file and
> read back into physical memory.

Things have become so complicated.

Bill

"RobertVA" <(E-Mail Removed)> wrote in message
news:%(E-Mail Removed)...

> I'd be surprised if ANY application loads into the same memory (virtual OR
> actual) on two different occasions. In fact, with Window's use of the page
> file an application's code and data segments might be non continuous, out
> of sequence and even shift around as they are written to the page file and
> read back into physical memory.

There must be linkage of shared libraries. Surely in machine language
the binary address that would start the player in memory would have a flag
to other parts of the player.

Bill

Bill Cunningham wrote:
> "RobertVA" <(E-Mail Removed)> wrote in message
> news:%(E-Mail Removed)...
>> I'd be surprised if ANY application loads into the same memory (virtual OR
>> actual) on two different occasions. In fact, with Window's use of the page
>> file an application's code and data segments might be non continuous, out
>> of sequence and even shift around as they are written to the page file and
>> read back into physical memory.
>
> Things have become so complicated.
>
> Bill
>
>
That's just PART of it. You've got peripherals like mice and keyboards
generating interrupts, peripheral input getting stored into event ques,
the OS has to figure out which window and control the pointer was over
when the mouse was clicked, keeping track of which window and control
has the focus, the OS calling redraw routines in the applications,
applications calling dialog boxes in the OS and applications
"displaying" output on the printer! Pretty much the same thing goes on
in other GUI operating systems going back into the eighties.

It's also pretty amazing how easy it is to accomplish something in
Visual BASIC when Visual C++ requires several large complex source
files, partially generated by Microsoft's wizards, for an identical purpose.

Bill Cunningham wrote:
> "RobertVA" <(E-Mail Removed)> wrote in message
> news:%(E-Mail Removed)...
>
>> I'd be surprised if ANY application loads into the same memory (virtual OR
>> actual) on two different occasions. In fact, with Window's use of the page
>> file an application's code and data segments might be non continuous, out
>> of sequence and even shift around as they are written to the page file and
>> read back into physical memory.
>
> There must be linkage of shared libraries. Surely in machine language
> the binary address that would start the player in memory would have a flag
> to other parts of the player.
>
> Bill
>
>
I'm not up on all the details, but an application's executables and
libraries contain tables of all the code entry points the developer
wanted to make available from other code modules. When the OS loads the
file it assigns each entry point a reference number, sometimes called a
"handle" that other modules use to access that entry point. When that
code segment is necessary the other modules absolutely HAVE to look the
handle up in the OS's list to find the required entry point. It's like
having a bunch of acquaintances that move around A LOT and needing to do
an on line directory search for their current address EVERY time you go
to visit them, even if you left their residence five minutes ago.