From mboxrd@z Thu Jan  1 00:00:00 1970
From: Jan Stancek <jstancek@redhat.com>
Date: Wed, 26 Jun 2019 05:59:37 -0400 (EDT)
Subject: [LTP] [PATCH v4 2/2] pkey: add test for memory protection keys
In-Reply-To: <CAEemH2d_7fq2RxEwZEMSUgErM7XG26RTsV1ogLTiLvALc810Gw@mail.gmail.com>
References: <20190626071538.29486-1-liwang@redhat.com>
 <20190626071538.29486-2-liwang@redhat.com>
 <285629943.30156546.1561537194409.JavaMail.zimbra@redhat.com>
 <CAEemH2d_7fq2RxEwZEMSUgErM7XG26RTsV1ogLTiLvALc810Gw@mail.gmail.com>
Message-ID: <1695359151.30173908.1561543177520.JavaMail.zimbra@redhat.com>
List-Id: <ltp.lists.linux.it>
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-Transfer-Encoding: 7bit
To: ltp@lists.linux.it


----- Original Message -----
> On Wed, Jun 26, 2019 at 4:20 PM Jan Stancek <jstancek@redhat.com> wrote:
> 
> >
> > ----- Original Message -----
> > > +
> > > +static void setup(void)
> > > +{
> > > +     int i, fd;
> > > +
> > > +     if (access(PATH_VM_NRHPS, F_OK)) {
> > > +             tst_res(TINFO, "Huge page is not supported");
> > > +             size = getpagesize();
> > > +             no_hugepage = 1;
> > > +     } else {
> > > +             int val;
> > > +
> > > +             SAFE_FILE_PRINTF(PATH_VM_NRHPS, "%d", 1);
> >
> > This is still SAFE write, which may trigger TBROK:
> >
> > # ./pkey01
> > tst_test.c:1100: INFO: Timeout per run is 0h 05m 00s
> > safe_file_ops.c:301: BROK: Failed to close FILE
> > '/proc/sys/vm/nr_hugepages' at pkey01.c:67: EOPNOTSUPP
> > safe_macros.c:773: WARN: pkey01.c:91: umount(tmp_pkey) failed: ENOENT
> > safe_macros.c:184: WARN: pkey01.c:92: rmdir(tmp_pkey) failed: ENOENT
> >
> 
> Er, sorry about still not working here.
> 
> If a system(e.g PowerKVM guest) configured with NO huge page support, then
> the file '/proc/sys/vm/nr_hugepages' exist but EOPNOTSUPP to read/write. In
> this key01, perhaps I shouldn't use the "/proc/.../nr_hugepages" to detect
> that at the beginning.
> 
> Seems the correct way is to use "/sys/kernel/mm/hugepages/" for huge page
> detecting and leave the SAFE_FILE_* still in setup(). Because we need to
> catch the "/proc/.../nr_hugepages" open/close or read/write issue in
> testing.

That should work, I see we used that in some tests already.

> 
> What do you think? Should we mask the EOPNOTSUPP as a TCONF skipping or
> others?
> 
> --
> Regards,
> Li Wang
>