Gunnar Dalsnes wrote:
Walking the list by handcoding is not an optimization.
Those macros are
just as fast as doing it manually.
Yes it is, not always but especially the kernel
sometimes have time
critical operations where any unnecessary overhead could have an impact
on performance.
Cleaner code (less code) is the point here, not
knowing how to deal with
lists. Less code = better code.
That's not always true. I remember the old
Nt/KeWaitForSingleObject
implementation.... total mess and slow as hell...
At least I'm not going to use these non-standard macros.
- Thomas