| 1 |
On Sun, 8 Jun 2014 09:04:20 -0400 |
| 2 |
Rich Freeman <rich0@g.o> wrote: |
| 3 |
|
| 4 |
> > c) EJOBS variable |
| 5 |
> > Bug #273101 [17], gentoo-dev discussion [18] |
| 6 |
> > - Discussion was in 2008. Is there (still) consensus? |
| 7 |
|
| 8 |
> The only thing that might be worth noting is that distcc users may |
| 9 |
> have an interest in distinguishing between gcc jobs and everything |
| 10 |
> else. I once messed with dictcc with very high job numbers and it |
| 11 |
> worked great when make hit a directory full of .c files, and not so |
| 12 |
> great when make/ant/whatever tried to run 50 instances of java in |
| 13 |
> parallel. |
| 14 |
|
| 15 |
Not just distcc! We already regularly see problems with a combination of |
| 16 |
(a) tons of RAM, (b) a dozen CPUs and (b) matching MAKEOPTS and (d) |
| 17 |
-pipe, where concurrent linker jobs surprisingly consume all of the |
| 18 |
RAM and one or more linker jobs segfault or get killed. Anyone's |
| 19 |
MAKEOPTS calculation should already include all of those factors. |
| 20 |
|
| 21 |
|
| 22 |
Regards, |
| 23 |
jer |
Archives