[lfs-support] LFS Bootup Sequence/etc

Baho Utot baho-utot at columbus.rr.com
Fri Aug 31 04:57:07 PDT 2012


On 08/30/2012 09:19 PM, Bruce Dubbs wrote:
> Baho Utot wrote:
>> On 08/30/2012 08:41 PM, Bruce Dubbs wrote:
>>> Baho Utot wrote:
>>>> On 08/30/2012 05:28 PM, Bruce Dubbs wrote:
>>>> /run was in place so the only thing missing, was it was on the root file
>>>> system on the ssd, it just was not mounted in tmpfs.
>>>>
>>>> Why should that cause an error?
>>> The very first command of the first boot script is:
>>>
>>> # Make sure /run/var is available before logging any messages
>>> if ! mountpoint /run >/dev/null; then
>>>              mount -n /run || failed=1
>>> fi
>> Ok but I am still not clear as to why it needs to be a mount point.
>>
>> It would work just the same as if it was not or is it required by the
>> kernel, that it be tmpfs?
> You are free to change it if you want.  However, virtually all other
> distros use it AFAIK.  Note that some applications may become confused
> if items from a previous boot are left in /run.  Also, if /run is on the
> root partition, things will not go well because there will be attempts
> to write to a (generally) read only partition.  One point of /run is to
> give a writable location early in the boot process.

That then might be why I had the issue upon booting.

>
>>> Do you have
>>>
>>> CONFIG_TMPFS=y
>> I mount /tmp as tmpfs so that points to that it is indeed set.
> OK, then it's set.
>
>> The default config gets me a login prompt and for now that is OK.
> I don't know what you've done (or not done).  Use /etc/sysconfig/rc.site
> and set IPROMPT="yes" and then step through the boot scripts.

Don't need to do that as everthing is working just fine.   I just was 
confused as to why it didn't function properly when /run was not tmpfs


>     -- Bruce
>




More information about the lfs-support mailing list