Re: [PATCH 1/6] describe fixing filesystem corruption in the hurd manual.

[Samuel Thibault]

Applied, thanks!

Joshua Branson, le jeu. 21 sept. 2023 12:48:17 -0400, a ecrit:
> From: "jbranso@dismail.de" <jbranso@dismail.de>
> 
> * doc/hurd.texi (Repairing Filesystems): described fixing filesystem
> corruption.
> * doc/hurd.texi (Shutdown): added the hurd specific halt-hurd command.
> ---
>  doc/hurd.texi | 53 +++++++++++++++++++++++++++++++++++++++++++++++++--
>  1 file changed, 51 insertions(+), 2 deletions(-)
> 
> diff --git a/doc/hurd.texi b/doc/hurd.texi
> index b713850c..ccad3228 100644
> --- a/doc/hurd.texi
> +++ b/doc/hurd.texi
> @@ -777,13 +777,23 @@ If neither @option{--interleave} or @option{--layer} is 
> specified, multiple
>  
>  @node Shutdown
>  @section Shutdown
> -@scindex halt
> +@scindex halt, poweroff, shutdown
>  @scindex reboot
>  
>  You can shut down your GNU/Hurd machine by typing
>  
> +@example
>  $ @kbd{su}
>  # @kbd{halt}
> +@end example
> +
> +If the above commands do not work to poweroff the machine, then you
> +should instead use the hurd specific commands @kbd{halt-hurd},
> +@kbd{poweroff-hurd}, or @kbd{reboot-hurd}, which do not actually shut
> +down daemons, but properly sync data to the disk.  After the data is
> +synced to the disk, you can safely hard shut off the machine (usually
> +via holding the power off button).  To deal with filesystem
> +corruption, see @ref{Repairing Filesystems}.
>  
>  @node Foundations
>  @chapter Foundations
> @@ -3625,8 +3635,47 @@ workstations, as well.
>  @section Repairing Filesystems
>  @pindex fsck
>  
> -FIXME: finish
> +Please note that as of 2023, the Hurd still uses the @code{ext2}
> +filesystem, which is not a journaling filesystem.  This means that
> +filesystem corruption is still possible, normally when the Hurd
> +becomes unresponsive, and you are forced to hard shutoff the
> +machine. Luckily, the @code{/} filesystem is automatically checked and
> +fixed at boot time (though you may have to manually fsck your
> +partitions every now and then).  You will know that your filesystem
> +partitions need to be repaired if the command @command{fsysopts <mount
> +point>} outputs @code{--readonly}.  For example:
> +
> +@example
> +$ fsysopts /home
> +  /hurd/ext2fs --readonly --relatime /dev/hd0s6
> +@end example
> +
> +To fix your filesystems, you should take a look at @code{/etc/fstab}
> +to find your partitions:
> +
> +@example
> +$ cat /etc/fstab
> +@end example
> +
> +Your partition/s will be something like @code{/dev/sd0sN} (or
> +@code{/dev/hd0sN}), where 'N' is a number.  To fix the corrupted
> +filesystem, login as the 'root' user, and run the following command
> +(replacing N with your number).
> +
> +@example
> +# umount /dev/sd0sN
> +# fsck.ext2 /dev/sd0sN
> +@end example
> +
> +If you have many partitions, then you may need to repeat this
> +procedure for all partitions.  Then you can reboot.
> +
> +@example
> +# reboot
> +@end example
>  
> +If the @command{reboot} command fails, which happens occassionally,
> +then use the hurd specific command @command{halt-hurd}.
>  
>  @node Linux Extended 2 FS
>  @section Linux Extended 2 FS
> -- 
> 2.40.1
> 
> 

-- 
Samuel
---
Pour une évaluation indépendante, transparente et rigoureuse !
Je soutiens la Commission d'Évaluation de l'Inria.