1 |
On Monday 14 May 2012 18:42:07 Zac Medico wrote: |
2 |
> On 05/14/2012 01:10 PM, Mike Frysinger wrote: |
3 |
> > On Monday 14 May 2012 15:08:32 Zac Medico wrote: |
4 |
> >> Actually, the inode_var_name thing will not work unless it's all in |
5 |
> >> one process. |
6 |
> > |
7 |
> > hmm, true, but that's the level we currently parallelize at, so it's |
8 |
> > fine. we do one subprocess per ELF and that includes the |
9 |
> > strip/splitdebug/splitsources. |
10 |
> > |
11 |
> > parallelizing more than on a per-ELF basis will require much finer |
12 |
> > grained queues which, while possible, would make the file much harder to |
13 |
> > hack on and extend. and i'm not sure we'd see that much of a gain. |
14 |
> |
15 |
> The thing is, in the case of hardlinks, we're parallelizing multiple |
16 |
> times on the *same* elf. Anyway, I've fixed it by using a directory full |
17 |
> of hardlinks, in these commits: |
18 |
|
19 |
well, realistically speaking, hardlinking has been broken before the |
20 |
parallelization work. |
21 |
https://bugs.gentoo.org/400767 |
22 |
|
23 |
> http://git.overlays.gentoo.org/gitweb/?p=proj/portage.git;a=commit;h=ad9442 |
24 |
> 75b88a50d2a1f694826b127cceb9221e78 |
25 |
> http://git.overlays.gentoo.org/gitweb/?p=proj/portage.git;a=commit;h=9ed00 |
26 |
> a9e70a3705164a5349145ff467e5c40ddfd |
27 |
|
28 |
i'll go through it and see what's what |
29 |
-mike |