| 1 |
On Wed, 3 Dec 2014 22:53:38 +0400 Alexander V Vershilov wrote: |
| 2 |
> Let me state my idea here. |
| 3 |
> |
| 4 |
> At first I want to mention that author provided 2 different approaches to |
| 5 |
> the solution, simple dependency loop checker and another more complicated |
| 6 |
> algorithm that is a loop breaker. |
| 7 |
|
| 8 |
No, unfortunately you get it wrong. Both earlyloopsolver and |
| 9 |
laterloopsolver break loops and optionally warn users about found |
| 10 |
issues. And they are not alternatives, but complements to each |
| 11 |
other: |
| 12 |
|
| 13 |
Early loop solvers works during dependencies cache generation, but |
| 14 |
it can't solve issues where all dependencies in a loop are hard |
| 15 |
("need" in terms of OpenRC). While later loop solver handles hard |
| 16 |
dependencies at run-time. |
| 17 |
|
| 18 |
More details are available in my another mail in this thread: |
| 19 |
http://article.gmane.org/gmane.linux.gentoo.devel/93968 |
| 20 |
|
| 21 |
> I think that on a boot phase in case of parallel boot rc should try to |
| 22 |
> check if loop exists and it is then print a warning and switch to a |
| 23 |
> sequential boot. |
| 24 |
|
| 25 |
In case of a hard loop this will fix nothing: there is no way to |
| 26 |
determine a correct order of sequential boot in case of a all |
| 27 |
"need" hard loop. Anyway I don't see how switch to sequential boot |
| 28 |
may fix anything, because this doesn't solve order determination |
| 29 |
problem at all. |
| 30 |
|
| 31 |
Best regards, |
| 32 |
Andrew Savchenko |
Archives