It is possible that mdadm creates a new subarray containing failed
devices. This may happen if a device has failed, but the meta data
containing that information hasn't been written out yet.
This code tests for this situation, and handles it in the monitor.
Signed-off-by: Martin Wilck <mwilck@arcor.de>
Signed-off-by: NeilBrown <neilb@suse.de>
The recent change to skip over invalid conf entries was bad because
it could leave garbage on the disk.
But we don't to write each entry separately as the writes a O_DIRECT
and so synchronous so it takes way too long.
So allocate a large buffer (probably the one used to read the config records)
and fill that then write it all at once.
Reported-by: Martin Wilck <mwilck@arcor.de>
Signed-off-by: NeilBrown <neilb@suse.de>
We should skip known failed disks when allocating space for
new arrays. This fixes the problem with 10ddf-fail-spare.
Signed-off-by: Martin Wilck <mwilck@arcor.de>
Signed-off-by: NeilBrown <neilb@suse.de>
Commit c7079c84 arrange for DDF to forget about any device
that is failed and not still marked as part of any array.
However such devices could still be part of the container and this
removal and updating of 'pdnum' can result in multiple devices having
the same pdnum. This in turn easily leads to confusion and
corruption.
So only discard pd entries for devices which are failed, not listed in
any virtual device, and for which we don't have a handle on the
device.
pd entries will not get removed until a new device is added after
the device has been removed from the container, either by
"mdadm --remove" or by assembling without the failed devices.
Reported-by: Albert Pauw <albert.pauw@gmail.com>
Analysed-by: Martin Wilck <mwilck@arcor.de>
Signed-off-by: NeilBrown <neilb@suse.de>
Recent commit 273989b93a
skipped writing some large blocks of 0xFF, but didn't seek
over the space, so subsequent data was written wrongly.
When we don't write, we need to seek.
Signed-off-by: NeilBrown <neilb@suse.de>
With the previous patch, mdmon will provide the layout property for us.
Signed-off-by: Martin Wilck <mwilck@arcor.de>
Signed-off-by: NeilBrown <neilb@suse.de>
Adds more verbose debugging in ddf_set_disk, to understand failures
better.
Signed-off-by: Martin Wilck <mwilck@arcor.de>
Signed-off-by: NeilBrown <neilb@suse.de>
Try to determine problem if load_ddf_header fails. May be useful
for determining compatibility problems with Fake RAID BIOSes.
Signed-off-by: Martin Wilck <mwilck@arcor.de>
Signed-off-by: NeilBrown <neilb@suse.de>
I needed this for tracking a bug with wrong offsets after array
creation.
Signed-off-by: Martin Wilck <mwilck@arcor.de>
Signed-off-by: NeilBrown <neilb@suse.de>
In particular, include refnum for better tracking. This makes
it a little easier for humans to track what happened to which disk.
Signed-off-by: Martin Wilck <mwilck@arcor.de>
Signed-off-by: NeilBrown <neilb@suse.de>
Move the check for good drives in the dl loop - otherwise dl
may be NULL and mdmon may crash.
Signed-off-by: Martin Wilck <mwilck@arcor.de>
Signed-off-by: NeilBrown <neilb@suse.de>
This call to validate_geometry is really rather gratuitous.
It is purely about the fact that super0 cannot use more than 4TB.
So just make it an explicit test - less confusing that way.
With this, validate_geometry is only called from Create, which
makes it easier to reason about.
Also validate_geometry is now never passed NULL for the 'chunk'
parameter, so we can remove those annoying tests for NULL.
Signed-off-by: NeilBrown <neilb@suse.de>
Metadata updates for secondary RAID (RAID10) need to cover
all BVDs. Compare with code in write_init_super_ddf().
Signed-off-by: NeilBrown <neilb@suse.de>
Fix a bug that caused the wrong conf record to be used to derive
data offset and size on secondary RAID (RAID10).
Signed-off-by: NeilBrown <neilb@suse.de>
Factor out single disk from __write_init_super_ddf to a new function
_write_super_to_disk. Use this function in store_super_ddf.
Signed-off-by: NeilBrown <neilb@suse.de>
Increase seq number only when there's actually a metadata change.
This is better then increasing it at every write.
This also fixes another endianness bug.
Signed-off-by: NeilBrown <neilb@suse.de>
Currently we compare fields between primary and secondary
superblocks, before we check if the primary is even valid.
This is a bit backwards, so reverse it.
Signed-off-by: NeilBrown <neilb@suse.de>
The "indirect" code path for adding VDs was not working correctly
for secondary RAID level. The "other BVDs" were not transmitted
to mdmon. Thus mdmon wouldn't build up correct information, and
RAID creation would fail when mdmon was already running on the container.
This patch fixes this.
Signed-off-by: NeilBrown <neilb@suse.de>
The return value of disk.raid_disk may be wrong.
The old code was using raiddisk, which is only valid with auto
layout. This leads to errors when arrays are created with
specified disks and mdmon is already running, like this:
mdadm -CR /dev/md/container -n5 $d1 $d2 $d3 $d4 $d5
mdadm -CR /dev/md/r5 -n5 -l5 /dev/md/container -z 5000
mdadm -CR /dev/md/r1 -n2 -l1 $d1 $d2
=> resulting array will use wrong disks
This patch fixes that.
Signed-off-by: NeilBrown <neilb@suse.de>
Use refnum rather than raiddisk for identifying the physical disk.
raiddisk should only be used for auto-layout.
Signed-off-by: NeilBrown <neilb@suse.de>
Implement kill_subarray, for mdmon running and not running.
The way Kill_subarray() is implemented, this requires that the
DDF layer uses "currentconf" to remember the last subarray
queried with container_content(), and use it as the one to kill.
I don't like this much but IMSM does it the same way.
Signed-off-by: NeilBrown <neilb@suse.de>
commit d682f344 inserted this call to "Kill" in write_init_super_ddf:
"Matching the functionality already in super0 and super1, when
we first create a container, remove any other recognisable metadata to
ensure it doesn't cause confusion."
But we should do this only at first container creation, not when
subarrays are created later.
Signed-off-by: NeilBrown <neilb@suse.de>
With secondary RAID level, disks may belong to other BVDs in
a given conf record. This needs to be taken into account
when fixing the vlist.
Signed-off-by: Martin Wilck <mwilck@arcor.de>
Signed-off-by: NeilBrown <neilb@suse.de>
For secondary RAID, we need to check which BVD needs to be updated.
Signed-off-by: Martin Wilck <mwilck@arcor.de>
Signed-off-by: NeilBrown <neilb@suse.de>
For debugging DDF structure changes, it is important to be able
to identify VCs by their GUIDs.
Signed-off-by: Martin Wilck <mwilck@arcor.de>
Signed-off-by: NeilBrown <neilb@suse.de>
If there are already disks in the container, reserve the same amount
of workspace as the first disk. Fill in the primary/secondary/
workspace LBA values.
Signed-off-by: Martin Wilck <mwilck@arcor.de>
Signed-off-by: NeilBrown <neilb@suse.de>
The used slots in the phys disk table aren't necessarily the
first ones. Rather, unused entries are represented by entries
where the GUID is all 0xff.
Signed-off-by: Martin Wilck <mwilck@arcor.de>
Signed-off-by: NeilBrown <neilb@suse.de>
Use get_pd_index_from_refnum() in get_extents to determine
matching VD. This will ensure RAID 10 (secondary RAID level)
support, too.
This also fixes a bug in the previous get_extents() code (missing
__be16_to_cpu for conf.prim_elmnt_count).
DDF test case (10ddf-create) verified.
Signed-off-by: Martin Wilck <mwilck@arcor.de>
Signed-off-by: NeilBrown <neilb@suse.de>
mwilck@arcor.de
NeilBrown