PureBasic Forums - English

I've run into some stuff in the new API's for Vista I don't know how to represent in PureBasic. One of them is UNIT64.

What is UNIT64? How is it represented in PureBasic? Is UNIT64 a quad in PureBasic?

How would a constant something like this be represented:
Simply like this?

UNIT64? or UINT64?
UINT64 is a unsigned Quad, not available in PB, but you can use Quad, is
the same memory but not the same valuerange

UNIT64 must be replaced by UINT64;

Yes, it's a quad;

Reference : http://www.purebasic.fr/english/viewtop ... ght=uint64

UNIT64, so that is a quad. Oki dokee then.

But what about the strange constant like: someconstant (0x00001ui64)

what is the ui64?

Thanks for your help

SFSxOI wrote:what is the ui64?

The "i64" and "ui64" extensions are non-standard, Microsoft compiler specific extensions used for specifing 64bit integer constants.

Kale wrote:

SFSxOI wrote:what is the ui64?

The "i64" and "ui64" extensions are non-standard, Microsoft compiler specific extensions used for specifing 64bit integer constants.

well, crap....I was in the progress of converting the managment portion of the Windows Filtering Platform API for Vista and Server 2008, for use with PureBasic. Was going pretty smooth for a bit, had some basic functions working too, then I encountered the 'ui64' thing. So...if its microsoft compiler specific...well...just Great!

gotta find some way to work around it.


Thanks Folks

SFSxOI I don't see the problem, 0x00001ui64 simply means that the hex value is a Quad.

jack wrote:SFSxOI I don't see the problem, 0x00001ui64 simply means that the hex value is a Quad.

The problem is it's unsigned. There is no problem storing a 64bit number but retrieving a meaningful value from the quad is a little tricky.

I know this thread is a bit old, but rather then starting a new one for the same subject...

I'm just getting back to working on the Windows Filtering Platform API for Vista, had put it aside for a while. I can't figure out how to get a useful value from the ui64 thing. Any ideas anyone?

a problem only occurs when the highest bit is set.

then, PB will interpret it as negative value.

so, when you need to check a uint64 value, you could e.g. check for the highest bit with
If Val < 0

depending on the case you need it, you should get from the documentation if the highest bit could be set by any chance.

declaring a function to use a uint64 does not mean that every value passed will have the highest bit set...

shifting will work differently too, the sign bit doesn't shift with the rest of the value (I don't think, something like that)

with shifting via PB the signbit remains set, iirc.
but you can call the asm command for unsigned shift.

anyhow, checking for single bits doesn't know any difference in signed or unsigned,
so checking a 64bit Flag register will work without problems.

I think the approach to deal with this will strongly depend on the purpose of this uint64 value.

what is the ASM command for unsigned shifting? that would be handy

isnt the result of lpperformancecounter also a UINT64 ?

i found this on a quick google - maybe useful.

http://books.google.ch/books?id=avDkMnu ... 1-PA114,M1

> what is the ASM command for unsigned shifting?

don't know by heart... read it in a topic here not long ago.


> isnt the result of lpperformancecounter also a UINT64 ?

for handling a counter, it is meaningless if you notate it signed or unsigned.
handling the overflow is almost the same.
tho I'm unsure right now, if an overflow of some 64bit counter even can occur within the next thousand years...