On Thu, Jan 27, 2011 at 08:18:34PM +0100, J. Roeleveld wrote: > On Thursday 27 January 2011 19:56:23 Allan Gottlieb wrote: > > On Thu, Jan 27 2011, Neil Bothwick wrote: > > > On Thu, 27 Jan 2011 17:09:27 +0200, Nikos Chantziaras wrote: > > >> > So on a 20 package world update, only 19 are faster while the 20th > > >> > runs at the same speed? Where's the loss there? Even if the last were > > >> > slower, it would be worth it. > > >> > > >> Given the amount of time unpack/configure/install of most packages > > >> needs (very short), my observation is that it would not be worth it. > > > > > > Even if that were true, how much time would you have to save to justify > > > adding -j 2 to EMERGE_DEFAULTS in make.conf? > > > > > > But it's not true, large packages spend a lot of time on these phases of > > > the install. > > > > OK I'm convinced since I know that those phases do take noticeable time. > > > > I have a "4" processor i7 model 620 (2 cores, doubled for > > hyperthreading) and have set MAKEOPTS="-j5". > > If I add -jobs=2 to EMERGE_DEFAULT_OPTS, should I lower > > MAKEOPTS to 3 (to 4)? > > > > thanks, > > allan > > You could, as if you leave it at -j5, you can end up with 2 * 5 = 10 > processed, eg: similar as if running with MAKEOPTS="-j10" > > I think the option that YoYo came with is a good compromise: > # MAKEOPTS="-j -l5" emerge -j --load-average=5 > > Next time I am doing a big upgrade, I'm going to test that to see how it > behaves.
I was just building the whole system for my notebook in a chroot on my desktop machine ( I use FEATURES=buildpkg to build binary packages in the chroot on a fast desktop machine and then upgrade the notebook with the binary packages) and I used exactly that (-j -l5 for bot make and emerge). Can't say if it really is better or not ;) but most of the time all four cores were busy, though sometimes I saw even 6 or 7 gcc-s simultaneously in top ;) emerge was running 3 to 4 jobs most of the time, sometimes dropping to 1 and once I saw it emerging about 10 parallel packages ;) (mostly small things, which I guess were doing a lot of unpacking/configuring/installing but almost nothing of compiling ;) btw, just now I got this error from dev-lang/v8: SCons error: option -j: invalid integer value: '-l4' seems scons honors MAKEOPTS, but doesn't understand the "loadaverage" version (-j -l4) yoyo
