Hi, On Fri, 2009-07-03 at 10:31 -0500, Carsten Neumann wrote: > Hello Gerrit, > > Gerrit Voß wrote: > > On Fri, 2009-07-03 at 10:03 -0500, Carsten Neumann wrote: > >> quick question since there was some discussion about the build speed on > >> win: do you pass /MP > >> (<http://msdn.microsoft.com/en-us/library/bb385193.aspx>) as a compiler > >> option? > > > > no, but I'm maxing out the one CPU I have already, its permanent at > > 100%, and I'm not sure how it impacts the IO performance, which I > > suspect a little of being a bottleneck. > > hm, probably IO is a limiting factor, but it is sad to see some cores > idling and only one making painstakingly slow progress ;)
;) My windows machine is only a single core. Usually that is enough except for the last few days. My multicore/multicpu machines are all Linux + one MacPro ;). But there it's not really a problem/issue, my linux laptop, where I spend 99% of my dev time, builds OpenSG a lot faster that the windows desktop ;) > >> Not sure if there are any hidden/undocumented pitfalls, a quick try with > >> setting it manually seemed to go through fine. > > > > I can add it. How responsive is the system while doing this ?. With one > > processor and compiler running it is quite painful. > > I've not noticed a problem with responsiveness, e.g. aborting a build > worked fine for me. This is on a 4 core machine. tja, the usual rule I had once was on windows 1 cpu/core for compile and at least 1 (better more) spare if you still want to do something else ;) kind regards, gerrit ------------------------------------------------------------------------------ _______________________________________________ Opensg-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/opensg-users
