On Thu, Aug 20, 2015 at 03:09:05PM +0800, Anand Jain wrote:
>
> (thanks for the off-ML emails from the people who helped me
> to understand).
>
> Dave,
>
> looks like you are suggesting something like..
>
> -------
> +_dmerror_mount_options()
> +{
> + _scratch_options mount
> + echo $SCRATCH_OPTIONS $MOUNT_OPTIONS $SELINUX_MOUNT_OPTIONS \
> + $* $DMERROR_DEV $SCRATCH_MNT
> +}
> +
> +_dmerror_mount()
> +{
> + _mount -t $FSTYP `_dmerror_mount_options $*`
> +}
> --------
Very close. :)
You've got the structure right, but missed one more thing: the
options are still cut'n'paste from _scratch_mount_options, and so
should be /factored/ into a common function...
_scratch_options is only useful to XFS to set external log/realtime
devices, and we don't want to do that for the dmerror device.
Everthing else is almost the same as _scratch_mount_options(),
however. So:
# Used for mounting non-scratch devices (e.g. loop, dm constructs)
# with the safe set of scratch mount options (e.g. loop image may be
# hosted on $SCRATCH_DEV, so can't use external scratch devices).
_common_dev_mount_options()
{
echo $MOUNT_OPTIONS $SELINUX_MOUNT_OPTIONS $*
}
_scratch_mount_options()
{
_scratch_options mount
echo `_common_dev_mount_options $*` $SCRATCH_OPTIONS \
$SCRATCH_DEV $SCRATCH_MNT
}
_dmerror_mount_options()
{
echo `_common_dev_mount_options $*` $DMERROR_DEV $SCRATCH_MNT
}
And to demonstrate why factoring like this is useful, this can also
be applied to _mount_flakey(), and we could also replace most of the
open-coded loop device mounts with a generic "mount device" function
like:
_mount_dev()
{
_mount `_common_dev_mount_options $*`
}
and call it like:
_mount_dev $LOOP_DEV $MNT_POINT
or even
_mount_dev -o loop $MNT_POINT
We can then take it further: if we have a _mount_dev() wrapper we
can factor _common_dev_mount_options() from all the
_*_mount_options() functions. i.e:
_scratch_mount_options()
{
_scratch_options mount
echo $SCRATCH_OPTIONS $SCRATCH_DEV $SCRATCH_MNT
}
_scratch_mount()
{
_mount_dev $* `_scratch_mount_options`
}
The result is we end up with a much cleaner set of mount functions
that are more generic, more extensible and easier to use, as well as
being more maintainable....
> Sorry that the word 'should mirror _scratch_mount()' confused me.
Ok, I'll try to be more clear in future.
Cheers,
Dave.
--
Dave Chinner
david@xxxxxxxxxxxxx
--
To unsubscribe from this list: send the line "unsubscribe fstests" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
