From: Wei Gao via ltp <ltp@lists.linux.it>
To: Li Wang <liwang@redhat.com>
Cc: ltp@lists.linux.it
Subject: Re: [LTP] [PATCH v2 2/2] mount: check path exists before tst_is_mounted in cleanup
Date: Fri, 22 Aug 2025 08:32:08 +0000 [thread overview]
Message-ID: <aKgrCMBfWsOzcwAb@localhost> (raw)
In-Reply-To: <CAEemH2dzFP75VXDLu=ZynXML9EfK-zY10Sd_+0g34zT=ywCkVw@mail.gmail.com>
On Fri, Aug 22, 2025 at 04:12:20PM +0800, Li Wang wrote:
> On Fri, Aug 22, 2025 at 4:03 PM Wei Gao <wegao@suse.com> wrote:
>
> > On Fri, Aug 22, 2025 at 03:48:11PM +0800, Li Wang via ltp wrote:
> > > tst_is_mounted() now internally resolves the mount path via realpath(),
> > > which will fail if the path does not exist. In sometest (e.g. mount05),
> > > MNTPOINT2 may not exist yet during cleanup, calling tst_is_mounted()
> > > directly could produce misleading warnings or errors.
> > >
> > > Add an explicit access(PATH, F_OK) check before calling tst_is_mounted()
> > > in the cleanup, ensuring we only query mounts for existing paths.
> > >
> > > Signed-off-by: Li Wang <liwang@redhat.com>
> > > ---
> > >
> > > Notes:
> > > v1 --> v2:
> > > patch 1/2, keep no change, do not send v2.
> > > patch 2/2, adding two more files for path exist check
> > >
> > > testcases/kernel/containers/mqns/mqns_03.c | 2 +-
> > > testcases/kernel/containers/mqns/mqns_04.c | 2 +-
> > > testcases/kernel/syscalls/mount/mount05.c | 2 +-
> > > 3 files changed, 3 insertions(+), 3 deletions(-)
> > Still more test case need update such as following?
> > 13 111 testcases/kernel/syscalls/mount/mount02.c <<cleanup>>
> > if (tst_is_mounted(MNTPOINT))
> > 14 202 testcases/kernel/syscalls/mount/mount03.c <<cleanup>>
> > if (tst_is_mounted(MNTPOINT))
> > 15 21 testcases/kernel/syscalls/mount/mount04.c <<cleanup>>
> > if (tst_is_mounted(MNTPOINT))
> >
>
> I wasn't encountering these test failures on that, the problem
> only appears when the mounted path is deleted by test.
>
> Did you test and get anything abnormal on your side?
Quick use your patch check mount02/3/4 no issue happen.
>
>
> >
> > Maybe we can just report warning instead of error?
> >
>
> It is Warning but not an error, the 'TERRNO' is only use to
> show the errno for better debugging.
>
> tst_device.c:440: TWARN: realpath(mntpoint2) failed: ENOENT (2)
Sorry i thought this will give an error :)
>
>
>
> >
> > --- a/lib/tst_device.c
> > +++ b/lib/tst_device.c
> > @@ -481,7 +481,7 @@ static int tst_mount_has_opt(const char *path, const
> > char *opt)
> > char abspath[PATH_MAX];
> >
> > if (!realpath(path, abspath)) {
> > - tst_resm(TWARN | TERRNO, "realpath(%s) failed", path);
> > + tst_resm(TWARN, "realpath(%s) failed", path);
> > return 0;
> > }
> >
> > >
> > > diff --git a/testcases/kernel/containers/mqns/mqns_03.c
> > b/testcases/kernel/containers/mqns/mqns_03.c
> > > index 4d3bfc52f..ac8311559 100644
> > > --- a/testcases/kernel/containers/mqns/mqns_03.c
> > > +++ b/testcases/kernel/containers/mqns/mqns_03.c
> > > @@ -130,7 +130,7 @@ static void cleanup(void)
> > > if (!access(MQUEUE2, F_OK))
> > > SAFE_MQ_UNLINK(MQNAME2);
> > >
> > > - if (tst_is_mounted(DEVDIR))
> > > + if (!access(DEVDIR, F_OK) && tst_is_mounted(DEVDIR))
> > > SAFE_UMOUNT(DEVDIR);
> > > }
> > >
> > > diff --git a/testcases/kernel/containers/mqns/mqns_04.c
> > b/testcases/kernel/containers/mqns/mqns_04.c
> > > index d28c330c4..790607ecd 100644
> > > --- a/testcases/kernel/containers/mqns/mqns_04.c
> > > +++ b/testcases/kernel/containers/mqns/mqns_04.c
> > > @@ -123,7 +123,7 @@ static void cleanup(void)
> > > if (!access(MQUEUE2, F_OK))
> > > SAFE_MQ_UNLINK(MQNAME2);
> > >
> > > - if (tst_is_mounted(DEVDIR))
> > > + if (!access(DEVDIR, F_OK) && tst_is_mounted(DEVDIR))
> > > SAFE_UMOUNT(DEVDIR);
> > > }
> > >
> > > diff --git a/testcases/kernel/syscalls/mount/mount05.c
> > b/testcases/kernel/syscalls/mount/mount05.c
> > > index 66e102a32..5585e230e 100644
> > > --- a/testcases/kernel/syscalls/mount/mount05.c
> > > +++ b/testcases/kernel/syscalls/mount/mount05.c
> > > @@ -35,7 +35,7 @@ static void cleanup(void)
> > > if (tst_is_mounted(MNTPOINT1))
> > > SAFE_UMOUNT(MNTPOINT1);
> > >
> > > - if (tst_is_mounted(MNTPOINT2))
> > > + if (!access(MNTPOINT2, F_OK) && tst_is_mounted(MNTPOINT2))
> > > SAFE_UMOUNT(MNTPOINT2);
> > > }
> > >
> > > --
> > > 2.49.0
> > >
> > >
> > > --
> > > Mailing list info: https://lists.linux.it/listinfo/ltp
> >
> >
>
> --
> Regards,
> Li Wang
--
Mailing list info: https://lists.linux.it/listinfo/ltp
next prev parent reply other threads:[~2025-08-22 8:32 UTC|newest]
Thread overview: 9+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-08-22 7:22 [LTP] [PATCH 1/2] tst_device: refine tst_is_mounted() Li Wang via ltp
2025-08-22 7:22 ` [LTP] [PATCH 2/2] mount05: check if mount path exist before tst_is_mounted Li Wang via ltp
2025-08-22 7:28 ` Li Wang via ltp
2025-08-22 7:48 ` [LTP] [PATCH v2 2/2] mount: check path exists before tst_is_mounted in cleanup Li Wang via ltp
2025-08-22 8:02 ` Wei Gao via ltp
2025-08-22 8:12 ` Li Wang via ltp
2025-08-22 8:32 ` Wei Gao via ltp [this message]
2025-08-22 7:45 ` [LTP] [PATCH 1/2] tst_device: refine tst_is_mounted() Wei Gao via ltp
2025-08-26 1:16 ` Li Wang via ltp
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=aKgrCMBfWsOzcwAb@localhost \
--to=ltp@lists.linux.it \
--cc=liwang@redhat.com \
--cc=wegao@suse.com \
/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