From: Petr Vorel <pvorel@suse.cz>
To: ltp@lists.linux.it
Subject: [LTP] [PATCH v3, 1/2] lib/tst_test.sh: Make tst_umount work with argument that has trailing slash
Date: Mon, 19 Jul 2021 08:45:58 +0200 [thread overview]
Message-ID: <YPUfpkJlrWTCH4p4@pevik> (raw)
In-Reply-To: <60F519D1.9040508@fujitsu.com>
Hi Xu, all,
> Hi Petr
> > Hi all,
> >>> Hi Leo
> >>>> /proc/mounts shows the mount point without trailing slashes, e.g.
> >>>> ~ $ cat /proc/mounts
> >>>> xxx /root/cgroup cgroup rw,relatime,cpu 0 0
> >>>> So current tst_umount would not work with argument that has trailing slash, e.g.
> >>>> tst_umount cgroup/ would give "The device is not mounted".
> >>>> Fix this by filtering out the trailing slash before grepping /proc/mounts.
> >>>> Signed-off-by: Leo Yu-Chi Liang<ycliang@andestech.com>
> >>>> ---
> >>>> testcases/lib/tst_test.sh | 3 ++-
> >>>> 1 file changed, 2 insertions(+), 1 deletion(-)
> >>>> diff --git a/testcases/lib/tst_test.sh b/testcases/lib/tst_test.sh
> >>>> index c6aa2c487..f132512e7 100644
> >>>> --- a/testcases/lib/tst_test.sh
> >>>> +++ b/testcases/lib/tst_test.sh
> >>>> @@ -282,13 +282,14 @@ tst_umount()
> >>>> [ -z "$device" ]&& return
> >>>> + device=${device%/}
> >>>> if ! grep -q "$device" /proc/mounts; then
> >>>> tst_res TINFO "The $device is not mounted, skipping umount"
> >>>> return
> >>>> fi
> >>>> while [ "$i" -lt 50 ]; do
> >>>> - if umount "$device"> /dev/null; then
> >>>> + if umount "$device"/> /dev/null; then
> >>> With removing this(we don't need add "/" here), this patch looks good to me
> >> +1
> > Actually we need to keep / for next patch, right? (cgroup/)
> I guess our ltp tst_umount api should support to umount cgroup or
> cgroup/ like umount command does.
Also the patch always added trailing /:
+ if umount "$device"/ > /dev/null; then
if we keep that, the second patch wouldn't have to add it to "tst_umount
cgroup/" call ("tst_umount cgroup" would be enough).
I'm not sure if we want it. @Cyril?
> > Thus why not just changing argument for grep?
> > - if ! grep -q "$device" /proc/mounts; then
> > + if ! grep -q "${device%/}" /proc/mounts; then
> Yes, it is more easier. But I think it still existed the problme when
> we only use "/" parameters. I guess we should reject this situation.
I would not care about it. 1) no test does 'tst_umount /' 2) error reporting
would be correct: umount: /: target is busy.
Kind regards,
Petr
> code maybe as below:
> --- a/testcases/lib/tst_test.sh
> +++ b/testcases/lib/tst_test.sh
> @@ -282,7 +282,12 @@ tst_umount()
> [ -z "$device" ] && return
> - if ! grep -q "$device" /proc/mounts; then
> + if [ "$device" = "/" ]; then
> + tst_res TINFO "We can not umount / directory"
> + return
> + fi
> +
> + if ! grep -q "${device%/}" /proc/mounts; then
> tst_res TINFO "The $device is not mounted, skipping umount"
> return
> fi
next prev parent reply other threads:[~2021-07-19 6:45 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-07-16 14:01 [LTP] [PATCH v3, 1/2] lib/tst_test.sh: Make tst_umount work with argument that has trailing slash Leo Liang
2021-07-19 5:19 ` xuyang2018.jy
2021-07-19 5:51 ` Petr Vorel
2021-07-19 5:58 ` Petr Vorel
2021-07-19 6:20 ` xuyang2018.jy
2021-07-19 6:45 ` Petr Vorel [this message]
2021-07-19 8:18 ` Leo Liang
2021-07-19 6:49 ` Petr Vorel
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=YPUfpkJlrWTCH4p4@pevik \
--to=pvorel@suse.cz \
--cc=ltp@lists.linux.it \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox