| Message ID | 20141110160309.GI2863@dhcp-13-216.nay.redhat.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
I get newly enabled test 294 passing on cifs mounts to Samba. What is best way to get the mount options passed to the mount of scratch dev so it doesn't prompt for it? (seems to ignore the mount options in TEST_FS_MOUNT_OPTS so prompts multiple times for password of scratch mount) On Mon, Nov 10, 2014 at 10:03 AM, Eryu Guan <eguan@redhat.com> wrote: > On Mon, Nov 10, 2014 at 09:32:34AM -0600, Steve French wrote: >> Failures: generic/035 generic/074 generic/120 generic/184 generic/192 >> generic/193 generic/215 generic/237 generic/258 generic/294 >> generic/306 generic/307 generic/313 generic/317 generic/319 >> >> Test 258 is a Samba server bug (and does work to Windows server). 074 >> and 215 are passing for me for both CIFS and SMB3, 035 was only >> failing for SMB2/SMB3 (and NFS) for me, not CIFS > > generic/074 is passing for me too now. > > generic/215 fails like: > --- tests/generic/215.out 2013-10-19 01:23:10.352848003 +0800 > +++ /root/xfstests/results//generic/215.out.bad 2014-11-10 23:59:41.619045914 +0800 > @@ -5,4 +5,6 @@ > 4096+0 records out > > writing via mmap > +FAIL: mtime not update after mapped write > +FAIL: ctime not update after mapped write > *** done > >> Test 184 passes to Samba, but requires "sfu" (mount.cifs mount option) >> if mounting with SMB2/SMB3 or mounting to Windows (in order to emulate >> FIFO and device files) >> >> 192 and 193 and 313 are time stamp related problems. >> >> Are you using at least 3.17 kernel? We put various fixes in 3.17 to >> address problems fsx pointed out. > > Yes, I'm using 3.18-rc2+ kernel. > > Thanks, > Eryu
[ please don't top post. ] On Wed, Nov 12, 2014 at 05:26:13PM -0600, Steve French wrote: > I get newly enabled test 294 passing on cifs mounts to Samba. > > What is best way to get the mount options passed to the mount of > scratch dev so it doesn't prompt for it? # MOUNT_OPTIONS="-o ...." ./check -g auto Cheers, Dave.
On Wed, Nov 12, 2014 at 9:28 PM, Dave Chinner <david@fromorbit.com> wrote: > On Wed, Nov 12, 2014 at 05:26:13PM -0600, Steve French wrote: >> I get newly enabled test 294 passing on cifs mounts to Samba. >> >> What is best way to get the mount options passed to the mount of >> scratch dev so it doesn't prompt for it? > > # MOUNT_OPTIONS="-o ...." ./check -g auto much better - that worked - thanks I had been using TEST_FS_MOUNT_OPTS (so I could have multiple sections in local.config) rather than MOUNT_OPTIONS
On Wed, Nov 12, 2014 at 10:08:34PM -0600, Steve French wrote: > On Wed, Nov 12, 2014 at 9:28 PM, Dave Chinner <david@fromorbit.com> wrote: > > On Wed, Nov 12, 2014 at 05:26:13PM -0600, Steve French wrote: > >> I get newly enabled test 294 passing on cifs mounts to Samba. > >> > >> What is best way to get the mount options passed to the mount of > >> scratch dev so it doesn't prompt for it? > > > > # MOUNT_OPTIONS="-o ...." ./check -g auto > > much better - that worked - thanks > > I had been using TEST_FS_MOUNT_OPTS (so I could have multiple sections > in local.config) rather > than MOUNT_OPTIONS BTW, why aren't you using CIFS_MOUNT_OPTIONS to define the login creds for both the test and scratch devices? That's what it was added for.... Cheers, Dave.
On Wed, Nov 12, 2014 at 10:27 PM, Dave Chinner <david@fromorbit.com> wrote: > On Wed, Nov 12, 2014 at 10:08:34PM -0600, Steve French wrote: >> On Wed, Nov 12, 2014 at 9:28 PM, Dave Chinner <david@fromorbit.com> wrote: >> > On Wed, Nov 12, 2014 at 05:26:13PM -0600, Steve French wrote: >> >> I get newly enabled test 294 passing on cifs mounts to Samba. >> >> >> >> What is best way to get the mount options passed to the mount of >> >> scratch dev so it doesn't prompt for it? >> > >> > # MOUNT_OPTIONS="-o ...." ./check -g auto >> >> much better - that worked - thanks >> >> I had been using TEST_FS_MOUNT_OPTS (so I could have multiple sections >> in local.config) rather >> than MOUNT_OPTIONS > > BTW, why aren't you using CIFS_MOUNT_OPTIONS to define the login > creds for both the test and scratch devices? That's what it was > added for.... When specifying multiple sections in local.config (a common case for us is to have one section to run tests with cifs dialect, then run again with smb2, then again with smb3 dialect etc.). Pavel had indicated that to do this you must use TEST_FS_MOUNT_OPTS instead of CIFS_MOUNT_OPTIONS if you have multiple sections. I haven't found a way around this. Specifying the username/password in TEST_FS_MOUNT_OPTS does work for the TEST mount but not for the SCRATCH mount (but as you mentioned does work if you specify MOUNT_OPTIONS)
On Wed, Nov 12, 2014 at 10:40:35PM -0600, Steve French wrote: > On Wed, Nov 12, 2014 at 10:27 PM, Dave Chinner <david@fromorbit.com> wrote: > > On Wed, Nov 12, 2014 at 10:08:34PM -0600, Steve French wrote: > >> On Wed, Nov 12, 2014 at 9:28 PM, Dave Chinner <david@fromorbit.com> wrote: > >> > On Wed, Nov 12, 2014 at 05:26:13PM -0600, Steve French wrote: > >> >> I get newly enabled test 294 passing on cifs mounts to Samba. > >> >> > >> >> What is best way to get the mount options passed to the mount of > >> >> scratch dev so it doesn't prompt for it? > >> > > >> > # MOUNT_OPTIONS="-o ...." ./check -g auto > >> > >> much better - that worked - thanks > >> > >> I had been using TEST_FS_MOUNT_OPTS (so I could have multiple sections > >> in local.config) rather > >> than MOUNT_OPTIONS > > > > BTW, why aren't you using CIFS_MOUNT_OPTIONS to define the login > > creds for both the test and scratch devices? That's what it was > > added for.... > > When specifying multiple sections in local.config (a common case for us > is to have one section to run tests with cifs dialect, then run again with > smb2, then again with smb3 dialect etc.). Pavel had indicated that to do this > you must use TEST_FS_MOUNT_OPTS instead of CIFS_MOUNT_OPTIONS > if you have multiple sections. I haven't found a way around this. Adding CIFS_MOUNT_OPTIONS in each section that requires a different definition should just work. If it doesn't, then please fix get_next_config() so it does and post the patch.... -Dave.
--- tests/generic/215.out 2013-10-19 01:23:10.352848003 +0800 +++ /root/xfstests/results//generic/215.out.bad 2014-11-10 23:59:41.619045914 +0800 @@ -5,4 +5,6 @@ 4096+0 records out writing via mmap +FAIL: mtime not update after mapped write +FAIL: ctime not update after mapped write *** done