| 1 |
On Wednesday, November 22, 2017 2:12:31 PM CET Raffaele Belardi wrote: |
| 2 |
> Jeremi Piotrowski wrote: |
| 3 |
> > That being said: if you do a world rebuild you will have lots of packages |
| 4 |
> > that spend ~40 seconds doing their autoconf run, only to build 2-3 sources |
| 5 |
> > files. On an 8-core machine at work, I get good results using parallel |
| 6 |
> > emerge jobs (emerge -jX). For your 6-core AMD CPU (assuming it actually |
| 7 |
> > has 12 threads) I'd start with 'emerge -j3' and MAKEOPTS='-j12 -l16'. |
| 8 |
> > That should get you a nice speedup, but may require a bit more ram. |
| 9 |
> |
| 10 |
> emerge -j3 is something I did not think of, I'll try it. But won't that |
| 11 |
> break portage's carefully crafted package dependencies? I suppose you could |
| 12 |
> get occasional build failures? |
| 13 |
|
| 14 |
No, it won't. |
| 15 |
If it has multiple packages that can be built (taking dependencies into |
| 16 |
account) it will. If all packages in the queue are waiting for a single |
| 17 |
package to be installed, all those will wait till that package is finished. |
| 18 |
|
| 19 |
> I'm using MAKEOPTS=-j7, I thought 2 threads per CPU (hyperthreading?) was an |
| 20 |
> Intel thing only. |
| 21 |
|
| 22 |
I thought so too. |
| 23 |
|
| 24 |
Don't forget to add the "load-average" option to both, otherwise there is a |
| 25 |
very likely chance you will have 3 * 7 = 21 build processes running at the |
| 26 |
same time. |
| 27 |
|
| 28 |
Also, I have had situations where using "-j" actually caused some issues where |
| 29 |
it was picked up by the wrong process or it has a different meaning for some |
| 30 |
other process. Specifying the long version (--jobs) has been reliable for me. |
| 31 |
|
| 32 |
-- |
| 33 |
Joost |
Archives