[systemd-devel] [PATCH] fstab-generator: do not check btrfs and xfs

[Kai Krakow]

Stefan G. Weichinger <lists at xunil.at> schrieb:

> Am 29.06.2014 23:52, schrieb Tom Gundersen:
>> On Sun, Jun 29, 2014 at 11:31 PM, Stefan G. Weichinger
>> <lists at xunil.at> wrote:
>> 
>>> This is how gentoo currently implements things. So the devs there
>>> should link it to /bin/true ? We could open a bug for that at
>>> bugs.gentoo.org ...
>> 
>> IMHO both xfs and btrfs should just not ship a fsck helper at all,
>> not even as a symlink. This workaround made sense at some point, but
>> now I believe both systemd _and_ fsck itself can deal gracefully with
>> a missing fsck helper.
> 
> I'll be happy to file that as a gentoo bug in the next few days.
> 
> AFAI understand it is more a "cosmetic" issue now ... fsck "fails" / is
> not executed for these 2 filesystems currently but gentoo linux boots.

I don't think that issueing a bug to BGO is the right thing because after 
looking at the upstream source, it is not "how Gentoo does things", this 
"fsck.btrfs" is baked into upstream.

It does not even fail. It checks if the device node exists, then returns 
exit code 0. It prints a warning if not called with "-a" option.

Have a look yourself:

https://github.com/josefbacik/btrfs-progs/blob/master/fsck.btrfs

I, for myself, followed upstream recommendation and set passno=0 so I even 
didn't notice the OR's problem. However, I scanned my old journals and found 
this:

Jun 11 00:19:27 jupiter systemd-fsck[287]: fsck.btrfs doesn't exist, not 
checking file system.

So, in the past "fsck.btrfs" was not installed, and systemd didn't like it 
that much back then. However, it has been handled gracefully - systemd just 
ignored that.

I conclude that, while executing a no-op binary/shell is pointless, nothing 
needs to be changed the way it is now (at least not in systemd). It is there 
as a safe fall-back / compat option, and it does not hurt. People wanting to 
optimize for that should simply set passno=0.

If distros want to change that, they should probably ship a patch with the 
package that replaces the script by a symlink or removes it completely. So, 
maybe yes, report to BGO. But I don't feel its currently worth the hassle, 
especially since Gentoo supports more init systems than just systemd and 
openrc, and that change may be an incompatible change.

-- 
Replies to list only preferred.