On 11-4-2010 9:45, Lukas Gebauer wrote: > Just quick note - one assembler function is very funny, when you need to > create code what working on many processor family (Intel, Motorola, > Arm,...) and when same code want to be runned as managed CIL on the .NET > platform. > > You are funny from the totally irelevant performance penalty. But > proposed sendinteger+sendbufer can cause 200ms delay in TCP > communication. It is much more then you lost by little piece of > ineffective code.
To me it looks a bit strange to me too. 1) Both native Delphi and FPC can cast to a record, which is way cheaper than using a refcounted type IFDEF CIL if necessary. 2) FPC has an CPU dependant primitive in the RTL. 3) The http://bugs.freepascal.org/view.php?id=12836 approach might be useful if you don't want to use IFDEFS It is purely arithmetic, and avoids casting, and also inline friendly. ------------------------------------------------------------------------------ Download Intel® Parallel Studio Eval Try the new software tools for yourself. Speed compiling, find bugs proactively, and fine-tune applications for parallel performance. See why Intel Parallel Studio got high marks during beta. http://p.sf.net/sfu/intel-sw-dev _______________________________________________ synalist-public mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/synalist-public
