Re: About Windows address space

From: Alex McDonald (alex_mcd_at_btopenworld.com)
Date: 04/21/04 

Date: Wed, 21 Apr 2004 19:26:56 +0000 (UTC)

"fa2k" <marius_b@chello.no> wrote in message
news:yhzhc.9653$EV2.89392@amstwist00...
> Hi,
>
> Is there any minimum value for pointers, that is consistent across all
> windows versions? This must include all pointers a user-space program
could
> specify, also those returned by e.g. WinSock. The thing I want to do is
> allow the user (i'm making a library, this is not the end user, but the
one
> who writes "modules" for my IRC proxy) to use either names or handles in
the
> same parameter to a function:
>
> if parameter is less than ADDRESS_MINIMUM
> then is_handle
> else is_string
>
> I hope this qualifies as on topic, even though it it strictly not
assembly.
>
> Thanks,
> Marius
>
>

The effective minimum varies; here, for instance, on my W2K machine the
stack is at 0x00130000 and down for 1 Mb, the loadpoint of an executable is
0x0040000 or greater, and the address range is up to 0x7FFFFFFF, and there
are heaps starting at 0x00130000 and up... and so on. So your handle would
be difficult to differentiate from a string pointer, unless you're far more
restrictive.

Windows does it by limiting handles that can also be strings to a maximum
value of a WORD; see the MAKEINTRESOURCE macro (search msdn.microsoft.com).
That limits your handle to a maximum of 0x0000FFFF. There's nothing below
that address that you can get at in user-land (or would sensibly want to),
in any Windows version. Not all Windows handles are limited to 0x0000FFFF,
btw; just those that also accept string pointers, and most of those seem to
be hangovers from Win3.1. I wouldn't recommend this technique, due to the
limitations.

Why not pass a pair of parameters, one for handles, the other for strings,
and use the first one that isn't NULL? 

-- 
Regards
Alex McDonald

Relevant Pages

  • Re: Windows Update after SP2 installation
    ... that is one of the dll registration entries that was in my ... Microsoft MVP - Windows Shell/User ... > pointers, hr=80040154 ...
    (microsoft.public.windowsupdate)
  • (null)
    ... to printf or similar API. ... maybe when someone decided that some abuses of null pointers were less ... Does a Windows XP API do the same thing? ...
    (microsoft.public.development.device.drivers)
  • Re: Ranting about JVMs default memory limits...
    ... Originally, 32-bit Windows split it half and half, with the user code getting 2GB and the OS getting 2GB. ... But, the OS didn't really need 2GB of address space, and so enabling "LARGEADDRESSAWARE" tells the OS that the application doesn't make any assumptions about the range of pointers and that it can pass user data addresses to the process in the first 1GB of the second 2GB of the full 4GB address space. ... Memory blocks aren't moveable, so allocations can sit in between two free blocks, limiting the size of the largest block that can actually be allocated. ... Depending on the allocation pattern, it's not hard at all to have over 1GB of virtual address space left, but not be able to allocate any single block nearly that large. ...
    (comp.lang.java.programmer)
  • Re: Windows Update after SP2 installation
    ... occur if you are running Windows in safe morde, or if the Windows Installer ... "Patti MacLeod" wrote: ... >> pointers, hr=80040154 ...
    (microsoft.public.windowsupdate)
  • Re: COLOR CHANGE OF CURSOR
    ... As I was saying, besides the mouse pointers that come with Windows, you can ... > mean the mouse pointer rather than the insertion point. ... >>> of my cursor from white to i.e. ...
    (microsoft.public.word.customization.menustoolbars)
Loading