cvs commit: src/share/man/man5 fstab.5

Poul-Henning Kamp phk at phk.freebsd.dk
Sat Feb 9 12:02:36 PST 2008


In message <86ir0xc45r.fsf at ds4.des.no>, =?utf-8?Q?Dag-Erling_Sm=C3=B8rgrav?= wr
ites:
>Mike Pritchard <mpp at FreeBSD.org> writes:
>>   Log:
>>   Explain how the passno field in /etc/fstab works with fsck
>>   and quotacheck in some more detail.
>
>One thing that isn't entirely clear: how do fsck and quotacheck behave
>for passno > 1?  I get the impression (both from the man page and from
>previous mailing list discussions) that all file systems with a passno
>above 1 are treated equally, rather than being ranked by their passno.

Last I mucked about with this code, it would handle passes sequentially
but disks[1] in each pass in parallel.

The reason for doing it this way was that there is no way to see from
/dev/ccd0 which drives this would affect, this using passes to control
when ccd devices where handled allowed manual sceduling.

If people have changed it since then, I'm not aware of it.

Poul-Henning

[1] A disk being defined as anything in the device name up to and including
the first numeric sequenc of characters.  Ie: "ad0s4b" is on disk "ad0"

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.


More information about the cvs-src mailing list