| 1 |
On Mon, 17 Oct 2011 14:39:40 +0200, Florian Philipp wrote: |
| 2 |
|
| 3 |
> > Is it up to an init script to do that either? I'd say no. either way |
| 4 |
> > seems wrong, but having the network config check that the interface is |
| 5 |
> > available before trying to bring it up seems somewhat less wrong. |
| 6 |
|
| 7 |
> Yes, I intended it to return 0 unconditionally. My reasoning was that |
| 8 |
> a) trying anyway doesn't hurt. |
| 9 |
|
| 10 |
Fair enough. |
| 11 |
|
| 12 |
> One thing that I worry more about is that there might be a race |
| 13 |
> condition. Maybe after loading the module, some time is necessary for |
| 14 |
> the interface to appear. I ran into an issue like that while playing |
| 15 |
> around with the zram module. In such a case, the separate init script |
| 16 |
> has a higher chance to succeed than a bash function called some |
| 17 |
> milliseconds before the interface initialization. |
| 18 |
|
| 19 |
You could add a "sleep 1" to preup(). |
| 20 |
|
| 21 |
|
| 22 |
-- |
| 23 |
Neil Bothwick |
| 24 |
|
| 25 |
By the time you can make ends meet, they move the ends. |
Archives