NOTE:ALIGN 4 on dq params for QueryPerformance functions !

Index > Windows > NOTE:ALIGN 4 on dq params for QueryPerformance functions !

Author

Thread

hopcode

Joined: 04 Mar 2008
Posts: 563
Location: Germany

hopcode 02 Nov 2008, 03:18

Important:
invoke QueryPerformanceCounter,n1
invoke QueryPerformanceFrequency,n2
require ALIGN 4 for the LARGE_INTEGER (dq)
n1 and n2. If not aligned the functions fail!!

Code:

 align 4
 n1 dq 0
 n2 dq 0

02 Nov 2008, 03:18

LocoDelAssembly
Your code has a bug

Joined: 06 May 2005
Posts: 4624
Location: Argentina

LocoDelAssembly 02 Nov 2008, 04:11

True, have you found any documentation backing that? I know it is true, but I never found where the documentation imposes such alignment constraint (maybe I was too blind).

[edit] BTW, align 4 or align 8? I though it was the latter

02 Nov 2008, 04:11

revolution
When all else fails, read the source

Joined: 24 Aug 2004
Posts: 20299
Location: In your JS exploiting you and your system

revolution 02 Nov 2008, 05:10

I always align dwords to 4 and qwords to 8 as a matter of standard. There is really no reason why predefined data declarations would not be aligned.

As to why the functions fails, well that is hard to know. It could be a hang over from the original NT days with the Alpha and Motorola CPUs not being so friendly about alignment. And putting in the necessary alignment fixup code into a high performance counter function is likely to create an overhead that is out of proportion with the intended use.

02 Nov 2008, 05:10

hopcode

Joined: 04 Mar 2008
Posts: 563
Location: Germany

hopcode 07 Nov 2008, 15:52

LocoDelAssembly wrote:

True, have you found any documentation backing that?...BTW, align 4 or align 8? I though it was the latter

No,at the moment. I am investigating...I can only imagine why, but i am not at all sure...
Pheraps this ALIGN 4/8 (4 should be already sufficient) is in a way
related with the structure of KUSER_SHARED_DATA
(look at this thread )
...

07 Nov 2008, 15:52

asmcoder

Joined: 02 Jun 2008
Posts: 784

asmcoder 10 Nov 2008, 09:55

[content deleted]

Last edited by asmcoder on 14 Aug 2009, 14:55; edited 1 time in total

10 Nov 2008, 09:55

bitshifter

Joined: 04 Dec 2007
Posts: 796
Location: Massachusetts, USA

bitshifter 10 Nov 2008, 10:03

That is very strange.
Is the LARGE_INTEGER defined as global or local?
Maybe try pushing 4 word values and using call instead.

10 Nov 2008, 10:03

< Last Thread | Next Thread >

Forum Rules:

You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You cannot attach files in this forum
You can download files in this forum