[systemd-devel] Second (erroneous) check of rootfs?

[Chris Murphy]

On Sun, Jan 11, 2015 at 8:29 PM, Andrei Borzenkov <arvidjaar at gmail.com> wrote:
> В Sun, 11 Jan 2015 12:45:17 -0700
> Chris Murphy <lists at colorremedies.com> пишет:
>
>> On Sun, Jan 11, 2015 at 4:57 AM, Andrei Borzenkov <arvidjaar at gmail.com> wrote:
>> > В Sun, 11 Jan 2015 04:15:41 -0700
>>
>> > Does it using systemd *inside* of initrd? From upstream dracut:
>> >
>> >     if ! dracut_module_included "systemd"; then
>> >         inst_hook cmdline 95 "$moddir/parse-block.sh"
>> >         inst_hook pre-udev 30 "$moddir/block-genrules.sh"
>> >         inst_hook mount 99 "$moddir/mount-root.sh"
>> >                                     ^^^^^^^^^^^^^ here it does it
>> >     fi
>> >
>> > Do you have Fedora specific patches to do it differently?
>>
>> So the reason it's mounting all file systems ro by default appears to
>> be strictly because of the ro kernel parameter in the grub.cfg as
>> created by grub-mkconfig.
>
> Again - when using mount-root.sh it will
>
> - *always* mount real root ro
> - fetch other mount options from real root /etc/fstab
> - remount real root rw with correct mount options applied
>
> This all got lost if you use standard systemd-fstab-generator +
> systemd-fsck services.

I don't agree with this, but here is a journalctl -b -l -o
short-monotonic output from Fedora 21 [1], with
systemd.log_level=debug rd.debug set. I do not see any reference to
mount-root.sh from dracut. What I see is systemd honoring kernel
parameter ro.



[1]
https://docs.google.com/a/colorremedies.com/file/d/0B_2Asp8DGjJ9TnlmNVRlLWtETTA/edit
-- 
Chris Murphy