On Sat, Jan 09, 2010 at 06:35:45PM -0500, Ted Unangst wrote: > On Sat, Jan 9, 2010 at 6:22 PM, Joerg Sonnenberger > <jo...@britannica.bec.de> wrote: > > You have analysed the situation correctly. The problem is that the > > compiler does not know that the signed numvnodes is never negative, so > > it creates different code. E.g. on AMD64 it is 6 instructions for the > > signed division by 2 compared to one instruction for the explicit shift > > or the unsigned division. > > That's a good argument for making numvnodes unsigned, but not for > obfuscating the code.
With respect to 'obfuscating', I would agree with you, except using shift to divide (or multiply) by two, or multiples of two, is very common. --patrick
