| 1 |
On Wed, Sep 14, 2016 at 5:03 PM, Holger Wünsche |
| 2 |
<diegoldeneenteml@×××××.com> wrote: |
| 3 |
> |
| 4 |
> I think the problem might be fstab or the point, where the initramfs gives controll to the kernel. |
| 5 |
|
| 6 |
The initramfs doesn't ever really give control to the kernel (well, at |
| 7 |
least not any more than any process does anytime it invokes a system |
| 8 |
call). The last thing the initramfs does is exec init, which then |
| 9 |
assumes control. |
| 10 |
|
| 11 |
> Since if I am not mistaken (I just say what I recall reading while searching for a solution) the initramfs just gets the system running and then the kernel should remount the root-partition. |
| 12 |
|
| 13 |
If you're using an initramfs the kernel will not mount anything at any |
| 14 |
time unless some process with sufficient capabilities asks it to. The |
| 15 |
initramfs typically mounts the root partition, and then execs init. |
| 16 |
Anything beyond that is done by init or whatever processes it spawns, |
| 17 |
such as openrc. Offhand I don't remember if modern initramfs |
| 18 |
solutions mount root as read-write; openrc will probably check for |
| 19 |
this in any case and remount if it is read-only (I could be wrong on |
| 20 |
that). |
| 21 |
|
| 22 |
If something other than root isn't mounted correctly, the fault |
| 23 |
probably lies in your fstab or openrc, or you're missing a necessary |
| 24 |
driver/etc. |
| 25 |
|
| 26 |
-- |
| 27 |
Rich |
Archives