Actually no, I found reports of VirtualBox users hitting this issue but
they couldn't figure out why. bounds.h was being deleted which was where
the build was dying. So this can happen in the field though I don't know
what the triggers are.
This is why distros distribute kernel headers only to begin with, so the user
is burdened with the extra disk space and is still able to build external
modules. Unless dkms could verify beyond a shadow of a doubt that what it's
operating on is the full source then it should never do this. Even then it
could be a bad idea because dkms will never know why that source is there
and what it's current state is.
On 06/30/2010 04:35 AM, Jerone Young wrote:
> Does this only happen with chroot enviroment?
>
Actually no, I found reports of VirtualBox users hitting this issue but
they couldn't figure out why. bounds.h was being deleted which was where
the build was dying. So this can happen in the field though I don't know
what the triggers are.
http:// www.virtualbox. org/ticket/ 1935
Then I found this: /lists. linux-foundatio n.org/pipermail /bugme- janitors/ 2008-September/ 011296. html
https:/
and realized what had happened.
This is why distros distribute kernel headers only to begin with, so the user
is burdened with the extra disk space and is still able to build external
modules. Unless dkms could verify beyond a shadow of a doubt that what it's
operating on is the full source then it should never do this. Even then it
could be a bad idea because dkms will never know why that source is there
and what it's current state is.
On 06/30/2010 04:35 AM, Jerone Young wrote:
> Does this only happen with chroot enviroment?
>