, , , : http://www.atrevido.net/blog/2005/01/28/Inline+X86+ASM+In+C.aspx p >
See comments for additional reservations and other approaches. The advantage of playing with delegates targeting is (AFAIK), you wonโt get a controlled unmanaged transition, as if you were using GetDelegateForFunctionPointer or other interaction approaches.
In the opposite direction, this means that you need to code it manually and make sure that the calling conventions match and so on. It is probably fragile, and the delegate's overhead can make it inappropriate.
source
share