Index
> Projects and Ideas > Tray DefragGoto page Previous 1, 2 |
| Author |
|
|
bitRAKE 20 Dec 2007, 17:58
But you are assuming EBX will be zero at the start of a callback routine.
Edit: this is not true - my mistake.  Last edited by bitRAKE on 20 Dec 2007, 18:09; edited 1 time in total |
|||
20 Dec 2007, 17:58 |
|
|
revolution 20 Dec 2007, 18:02
bitRAKE wrote: But you are assuming EBX will be zero at the start of a callback routine. |
|||
|
20 Dec 2007, 18:02 |
|
|
codegrinder 20 Dec 2007, 21:30
bitRAKE wrote: I noticed you use EBX for 0 throughout the program. I don't know if this is a good assumption. Will windows always restore the process EBX before calling other threads within the process, and WndProcs? It is not my understanding that this is true, nor does the calling convention imply it would be so. Windows preserves ebx, esi and edi registers (edit: and ebp of course). Don't know why is this, but this is true even for Win95. Don't have 64 bit Windows (and 64 bit machine) to test this, almost sure it also does this thing (after seeing Win64 examples from fasm packages). I think this is some C/C++ compiler internal agreement. |
|||
|
20 Dec 2007, 21:30 |
|
|
codegrinder 21 Dec 2007, 00:34
I asked my friend who have got 64 bit PC to help me and we made something like ICQ debugging
 . Looks like we found where the CPU time is consumed (this was in a loop of testing if file is fragmented), but the problem with exiting still exists. Here is version.
|
|||||||||||
|
21 Dec 2007, 00:34 |
|
|
Dopefish 11 Feb 2008, 12:58
This is actually pretty cool and simple. Has there been any progress to this project since your last post?
|
|||
|
11 Feb 2008, 12:58 |
|
| Goto page Previous 1, 2 < Last Thread | Next Thread > |
Forum Rules:
|