* Re: [PATCH 1/2] xfstests: Split MOUNT_OPTIONS to TEST_FS_MOUNT_OPTS and MOUNT_OPTIONS
[not found] <1505722270-24648-1-git-send-email-gujx@cn.fujitsu.com>
@ 2017-09-22 8:18 ` Eryu Guan
0 siblings, 0 replies; only message in thread
From: Eryu Guan @ 2017-09-22 8:18 UTC (permalink / raw)
To: Gu Jinxiang; +Cc: fstests, linux-btrfs
On Mon, Sep 18, 2017 at 04:11:09PM +0800, Gu Jinxiang wrote:
> Resovle the inconsistent of mount option.
> Btrfs use MOUNT_OPTIONS for both scrath_dev and test_dev. Change to
> MOUNT_OPTIONS for scratch mount, and TEST_FS_MOUNT_OPTS for test dev
> mount.
>
> Signed-off-by: Gu Jinxiang <gujx@cn.fujitsu.com>
> ---
> As mentioned by https://patchwork.kernel.org/patch/9742039/, the usage
> of MOUNT_OPTIONS is inconsistent.
Sorry for the late reply, because I don't think this is the right fix,
and I was trying to sort out & refactor the mount option settings
through fstests, but found that it required more time than I thought..
The new _check_test_btrfs_filesystem() you introduced here is a (almost)
complete copy of _check_btrfs_filesystem, the only difference is that
it's mounting $device with $TEST_FS_MOUNT_OPTS instead of
$MOUNT_OPTIONS. At least you could factor out a common helper so you
don't have to repeat 99% of the code.
But that only fixes _check_test_btrfs_filesystems, there're similar
problems in _check_generic_filesystems and other places that do mount
operation. I'd like to see a well-defined interface to return the correct
mount options to callers that want to do mount operations.
I'll look into this and see if I can work something out. I really
appreciate if anyone has other thoughts/suggestions!
Thanks,
Eryu
^ permalink raw reply [flat|nested] only message in thread