| |
|
|
| |
|
| |
|
|
|
| Michael was so nice and has whom Xpse -Test too once implemented.
here its Results:
328ms with xpse 594ms without xpse
1891ms with xpse interpreter 2297ms without xpse interpreter
I have soebend too time Okrea on the new XPSE adjusted - and the code something rewritten. the whole Program runs IMHO rather ex - works stabiler and power a saubereren local.
I mean there I had to change too time ne good idea. |
|
|
| |
|
|
|
Michael Wodrich |
I mean there I had to change too time ne good idea
without these ständige change would this Forum neither what it now once is.
simply TOP
Best wishes Michael Wodrich |
|
|
| Programmieren, das spannendste Detektivspiel der Welt. | 10/29/06 ▲ |
|
|
|
|
Jörg Sellmeyer | I habs too time tested: Interpreter: Apifunktion: 1422 Call: 1222
Compiliert Api: 290 - 300 Call: 200 - 210
Very interesting discovery.
iF
I mean there I had to change too time ne good idea.
watts indeed time Time! |
|
|
| Windows XP SP2 XProfan X4... und hier mal was ganz anderes als Profan ... | 10/29/06 ▲ |
|
|
|
|
Jac de Lad | means the now, the Def ... slower is as the XPSE-Gedöns (means without jegliche Deklaration and Verheaderung)? |
|
|
| Profan² 2.6 bis XProfan 11.1+XPSE+XPIA+XPRR (und irgendwann XIDE) Core2Duo E8500/T2250, 8192/1024 MB, Radeon HD4850/Radeon XPress 1250, Vista64/XP | 10/29/06 ▲ |
|
|
|
|
Jörg Sellmeyer | Nö, geDeft becomes indeed with Davids Variante. it points only whom deklarierten functions same a solid Einsprungadresse (? or How is it? Funktionhandle?) To and the becomes then through Call apparently faster addressed. Greeting Jörg |
|
|
| Windows XP SP2 XProfan X4... und hier mal was ganz anderes als Profan ... | 10/29/06 ▲ |
|
|
|
|
Jac de Lad | @iF: Defst You the Call-Dingens each time new, if XPSE what recognize? and What is, if the Programmer the Call-Dingens already self in the View source definiert has? |
|
|
| Profan² 2.6 bis XProfan 11.1+XPSE+XPIA+XPRR (und irgendwann XIDE) Core2Duo E8500/T2250, 8192/1024 MB, Radeon HD4850/Radeon XPress 1250, Vista64/XP | 10/29/06 ▲ |
|
|
|
|
| No XPSE Deft nothing.
Statement I in the Nachfolgenden with ApiCall my: appeal of/ one Api which in a Headerfile declared is.
If Roland a ApiCall umwandelt - so does it this over External.
If XPSE a ApiCall umwandelt - so does it this over Call. XPSE power itself there very plenty trouble - declared integrally to that beginning the Program suitable variables to take the Funktionsadressen from whom DLLs, fetch itself to that Programmanfang The Funktionsadressen and GetControlParas (B The ApiCalls in Calls circa which The suitable Funktionsadressen carry.
this Posting shows it still integrally explicit: [...]
and here time the block the new Okrea: CompileMarkSeparationDef __cf1(2) !KERNEL32,GetProcAddress
Def __cf2(1) !KERNEL32,GetModuleHandleA
__cf1&=__cf1(__cf2(user32.dll),GetAsyncKeyState)
__cf2&=__cf1(__cf2(user32.dll),KillTimer)
__cf3&=__cf1(__cf2(user32.dll),GetWindowLongA)
__cf4&=__cf1(__cf2(user32.dll),SetWindowLongA)
__cf5&=__cf1(__cf2(user32.dll),SetClassLongA)
__cf6&=__cf1(__cf2(user32.dll),LoadCursorA)
__cf7&=__cf1(__cf2(user32.dll),SendMessageA)
__cf8&=__cf1(__cf2(user32.dll),SetWindowPos)
__cf9&=__cf1(__cf2(user32.dll),InvalidateRect)
__cf10&=__cf1(__cf2(user32.dll),UpdateWindow)
__cf11&=__cf1(__cf2(user32.dll),CallWindowProcA)
__cf12&=__cf1(__cf2(user32.dll),SetTimer)
__cf13&=__cf1(__cf2(user32.dll),GetCursorPos)
__cf14&=__cf1(__cf2(user32.dll),WindowFromPoint)
__cf15&=__cf1(__cf2(user32.dll),ClientToScreen)
__cf16&=__cf1(__cf2(user32.dll),GetSysColor)
__cf17&=__cf1(__cf2(user32.dll),FindWindowExA)
__cf18&=__cf1(__cf2(kernel32.dll),TerminateProcess)
__cf19&=__cf1(__cf2(kernel32.dll),OpenProcess)
__cf20&=__cf1(__cf2(kernel32.dll),GetCurrentProcessId)
__cf21&=__cf1(__cf2(user32.dll),SetWindowsHookExA)
__cf22&=__cf1(__cf2(user32.dll),GetWindowThreadProcessId)
__cf23&=__cf1(__cf2(user32.dll),UnhookWindowsHookEx class=s2>)
And so a transformed row sees then z.B. so from: CompileMarkSeparation and naturally declared XPSE nothing twice - and only The Apis which objectively used go.
there one The indicated of Headerfiles now even gänzlich omit can is the process the Kompilierens too still significantly faster - means not only the program runs rather separate...
understood? |
|
|
| |
|
|
|
| iF
If XPSE a ApiCall umwandelt - so does it this over Call. XPSE power itself there very plenty trouble - declared integrally to that beginning the Program suitable variables to take the Funktionsadressen from whom DLLs, fetch itself to that Programmanfang The Funktionsadressen and GetControlParas (B The ApiCalls in Calls circa which The suitable Funktionsadressen carry.
an trouble, which auszahlt. who plenty with Headern and API, bypassing sustain to that Geschwindigkeitsvorteil another improved safety hinzu - because it is not More possible, APIs over the Change the Exporttables During low the Laufens the Prozesses To hooken.
the means very: APIs The through DEF and EXTERNAL addressed go, detect The anzuspringende address in the DLL first, if The function in the View source called becomes. through Changing the Headers the DLL in the the API addressed becomes (Change in memory my I), is it possible whom CALL on a Own DLL umzuleiten and the Parameter abzugreifen. with of/ one Codierung one Freischaltungscodes for a shareware could the already something problematically his. there XPSE The Adressen at started the Program already festlegt, nützt a such Attacke nothing more - it go nevertheless The right Adressen angesprungen. |
|
|
| |
|
|