1
0
mirror of git://sourceware.org/git/lvm2.git synced 2024-12-25 01:34:38 +03:00
Commit Graph

15679 Commits

Author SHA1 Message Date
Zdenek Kabelac
d5a5792b30 tests: check accepting out-of-range creation_time 2019-05-10 15:07:25 +02:00
Zdenek Kabelac
a397b69ce3 metadata: allow reading metadata with invalid creation_time
lvm2 till version 2.02.169 (commit 78d004efa8)
was printing invalid creation_time argument into metadata on 32bit arch.

However with commit ba9820b142 we started
to properly validate all input numbers and thus we refused to accept
invalid metadata with 'garbage' string - but this results in the
situation where metadata produced on older lvm2 on 32 bit architecture
will become unreadable after upgrade.

To fix this case - extend libdm parser in a way, that whenever we
find error integer value, we also check if the parsed value is not for
creation_time node and in this case we let the metadata pass through
with made-up date 2018-05-24 (release date of 2.02.169).
2019-05-10 15:07:15 +02:00
Zdenek Kabelac
a8921be641 monitoring: fix monitoring change for cluster
Fix bug in table reload for clustered VG.
Function used in lv/vgchange --monitor y|n is using 'somewhat' hackish
shortcut and accesses activation function monitor_dev_for_events()
directly rather through full device refresh to avoid table reload
in case user want to only change monitoring state of device.
However since with old 'mirrors' there is table change dropping
handle_error there was in some cases needed table update.

This was put into monitor_dev_for_events() with assumption
vg_write_lock_held() could be used to decide if the actual monitoring
change comes from read-only lock-holding commands lvchange/vgchange.
However the clustered locking part (clvmd) actually doesn't differentiate
about this concept of having VG openned in read-only or write mode.
So it caused unwanted reloads of mirror tables even in commands like lvconvert.

Thus for clustered locking there is full table reload put into the code and
shortcut applies only for non-clustered locking.

Also the patch tries to avoid calling repeated LV refresh in case the
lv/vgchange uses  --refresh & --monitor.
2019-05-06 15:57:52 +02:00
Zdenek Kabelac
9d7afaaab8 tests: update parm for new kernel 2019-04-30 13:16:31 +02:00
Zdenek Kabelac
71b9cb8e0f tests: extend 2019-04-30 13:16:19 +02:00
Zdenek Kabelac
057d866889 tests: rounding for pools changed to power of 2
Even with 64K chunk support, lvm2 will target power-of-2 chunks.
2019-04-30 13:14:27 +02:00
Zdenek Kabelac
e79e092f8b lv_manip: better work with PERCENT_VG modifier
When using 'lvcreate -l100%VG' and there is big disproportion between
real available space and requested setting - automatically fallback
to 100%FREE.

Difference can be seen when VG is big and already most space was
allocated, so the requestion 100%VG can end (and by spec for % modifier
it's correct) as LV with size of 1%VG.  Usually this is not a big
problem - buit in some cases - like cache-pool allocation, this
can result a big difference for chunksize selection.

With this patch it's more closely match common-sense logic without
the need of reitteration of too big changes in lvm2 core ATM.

TODO: in the future there should be allocator solving all allocations
in a single call.
2019-04-30 12:24:53 +02:00
Zdenek Kabelac
4729b4af0b thin: select chunk size as power of 2
Whenever thin-pool chunk size is unspecified and left for lvm calculation
try to select the size as nearest highest power-of-2 instead of
just being a multiple of 64KiB. When multiple is bigger then 1MiB,
keep using 1MiB multiple.
2019-04-30 12:11:50 +02:00
Zdenek Kabelac
f3be66c002 lv_manip: insert remove layer skips pools
Fixing renaming of subLVs when removing and inserting layers - this
got visible when using stacked VDO pools.
2019-04-30 12:08:36 +02:00
Zdenek Kabelac
2047d405af activation: synchronize before removing devices
Udev is running udev-rule action upon 'resume'.

However lvm2 in special case is doing replacement of
'soon-to-be-removed' device with 'error' target for resuming
and then follows actual removal - the sequence is usually quick,
so when udev start action - it can result in 'strange' error
message in kernel log like:

Process '/usr/sbin/dmsetup info -j 253 -m 17 -c --nameprefixes --noheadings --rows -o name,uuid,suspended' failed with exit code 1.

To avoid this - we need to ensure there is synchronization wait for udev
between 'resume'  and 'remove' part of this process.

However existing code put strict requirement to avoid synchronizing with
udev inside critical section - but this originally came from requirement
to not do anything special while there could be devices in
suspend-state. Now we are able to see differnce between critical section
with or without suspended devices.  For udev synchronization only
suspended devices are prohibited to be there - so slightly relax
condition and allow calling and using 'fs_sync()' even inside critical
section - but there must not be any suspended device.
2019-04-30 12:07:42 +02:00
Zdenek Kabelac
1370277ea7 tests: check auto-growth of thin-pool meta 2019-04-30 12:06:08 +02:00
Zdenek Kabelac
f38cfd09c4 thin: fix maintenance of _pmspare
When metadata grows lvm2 may need to extend also _pmspare volume.
2019-04-30 12:02:49 +02:00
Zdenek Kabelac
515867bbad thin: resize metadata with data
When data are growing, adapt also size of metadata.
As we get way too many reports from users doing huge growths of
data portion while keep metadata small and avoiding using monitoring.

So to enhance the user-experience in case user requests grown of
thin-pool (without passing PV list for growth) - lvm2 will automaticaly
grown also the metadata part of thin-pool (if possible).
2019-04-30 12:02:49 +02:00
Zdenek Kabelac
bcf1aa99e6 thin: introduce estimate_thin_pool_metadata_size
Add function for estimation of thin-pool metadata size for given size of
data. Function is using already existing internal API so it can
be reused for resize of thin-pool data.
2019-04-30 12:02:49 +02:00
Zdenek Kabelac
b41b112a4b libdaemon: ensure threads are reaped before checking shutdown
Since we are checking _shutdown_requested - we expect all threads
are finished - that is currently checked only by checking ->next ptr
being NULL - so this can be NULL only when _reap() function clears
out all already finished threads.

I'm finding this design quite problematic in its core - but as a
'trivial hotfix' - lets _reap() linked list before check for signal.

There is likely a large potentical for few races - but the windows is
very very small - since lvmetad has been already purged from upstream,
lets go with this hotfix.
2019-04-29 13:49:33 +02:00
David Teigland
559cf0cd1e devices: drop open error message
This open error is being printed in more common,
non-error circumstances than expected.  After a
number of complaints make it only a debug message.
2019-04-23 09:42:25 -05:00
Zdenek Kabelac
228e331e27 cleanup: missed string specifier 2019-04-17 11:29:35 +02:00
Zdenek Kabelac
3ead62e24b tests: check vgsplit works with cache 2019-04-16 12:57:18 +02:00
Zdenek Kabelac
fa025cdd9a cache: support vgsplit
Enable vgsplit to work with VG containing cached LVs.
2019-04-16 12:57:18 +02:00
Marian Csontos
f5d1f4f086 Wiping require exclusive actvation
The master branch uses activate_lv only, but on the stable branch
activate_lv_excl_local was used. This patch restores the constraints.

Fix regression from bad cherry pick: 9b04851fc5
2019-04-16 08:26:34 +02:00
Heinz Mauelshagen
699bf86090 lvconvert: can't create cachepool with RAID1 metadata SubLV
Trying to convert an exiting RAID1 LV to cache metadata SubLV
fails because of missing exclusive activation during wiping.

Solve by activating the RAID1 metadata SubLV exclusively.

Resolves: rhbz1698866
2019-04-12 21:38:04 +02:00
Peter Rajnoha
03d6cfdd99 systemd: put back DefaultDependencies=no for lvmpolld and lvmetad socket unit
Previous commit 0a726a7e26 removed this
by mistake - we have to keep the DefaultDependencies=no - the
sockets.target is after sysinit.target.
2019-04-11 12:25:45 +02:00
Zdenek Kabelac
6064b9f1b2 gcc: cleanup const warning 2019-04-10 13:30:34 +02:00
Zdenek Kabelac
6d4f36c2c7 libdaemon: use pselect to avoid condition checking race
To avoid tiny race on checking arrival of signal and entering select
(that can latter remain stuck as signal was already delivered) switch
to use  pselect().

If it would needed, we can eventually add extra code for older systems
without pselect(), but there are probably no such ancient systems in
use.
2019-04-10 13:28:45 +02:00
Zdenek Kabelac
63289b54c7 configure: check for pselect 2019-04-10 13:24:41 +02:00
Peter Rajnoha
0a726a7e26 systemd: add missing Before=shutdown.target to LVM2 services to fix shutdown ordering
We already used Conflicts=shutdown target to stop LVM2 services on shutdown.
But we still missed the ordering - the shutdown.target should be reached
only after all the services are really stopped.

Reported here: https://github.com/lvmteam/lvm2/issues/17
2019-04-09 12:40:55 +02:00
Marian Csontos
b79f1e176f bcache: Fix memory leak 2019-04-04 10:19:15 +02:00
Marian Csontos
81d954df4e post-release 2019-03-22 11:13:08 +01:00
Marian Csontos
7cbee7e9cf pre_release 2019-03-22 11:04:15 +01:00
Marian Csontos
717957ddc5 build: make generate 2019-03-22 11:01:04 +01:00
Heinz Mauelshagen
9b04851fc5 raid: fix (de)activation of RaidLVs with visible SubLVs
There's a small window during creation of a new RaidLV when
rmeta SubLVs are made visible to wipe them in order to prevent
erroneous discovery of stale RAID metadata.  In case a crash
prevents the SubLVs from being committed hidden after such
wiping, the RaidLV can still be activated with the SubLVs visible.
During deactivation though, a deadlock occurs because the visible
SubLVs are deactivated before the RaidLV.

The patch adds _check_raid_sublvs to the raid validation in merge.c,
an activation check to activate.c (paranoid, because the merge.c check
will prevent activation in case of visible SubLVs) and shares the
existing wiping function _clear_lvs in raid_manip.c moved to lv_manip.c
and renamed to activate_and_wipe_lvlist to remove code duplication.
Whilst on it, introduce activate_and_wipe_lv to share with
(lvconvert|lvchange).c.

Resolves: rhbz1633167
(cherry picked from commit dd5716ddf2)

Conflicts:
	WHATS_NEW
	lib/activate/activate.c
	lib/metadata/lv_manip.c
	lib/metadata/raid_manip.c
	tools/lvchange.c
	tools/lvconvert.c
2019-03-21 08:05:23 +01:00
David Teigland
dcf8f3111a pvscan: lvmetad init should set updating before scanning
When pvscan needs to initialize lvmetad (e.g. lvmetad has just
started and is empty), it should set the lvmetad state to "updating"
before it scans any devices.  Otherwise, many parallel pvscans
will try to initialize lvmetad, and in some cases earlier pvscans
with fewer devices information may replace newer pvscans with
more recent information.
2019-03-07 11:07:27 -06:00
David Teigland
ece0b131e5 config: improve scan_lvs description 2019-03-06 13:38:33 -06:00
Alasdair G Kergon
519f4453a5 dmsetup: Fix multi-line concise table parsing
Use the correct loop variable within the loop, instead of reusing the
initial value.  Table lines after the first don't get terminated in
the right place.

Signed-off-by: Kurt Garloff <kurt@garloff.de>
(cherry picked from commit ccfbd505fe)
2019-03-05 13:01:40 +01:00
Zdenek Kabelac
bc6ae7030a dm: migration_threshold for old linked tools
Just like with precending  lvm2 device_mapper patch, ensure
that old users of libdm will also get fixed migration threshold
for caches.

(cherry picked from commit 74ae1c5bc1)

Conflicts:
	WHATS_NEW_DM
2019-03-05 13:01:19 +01:00
Zdenek Kabelac
167aa34926 stats: initilize regions to NULL
Commit 3750b0cff5 used bad: error
path in more occasions thus it now needs regions defined as NULL.

(cherry picked from commit 83c6f7e7e6)
2019-03-05 12:57:50 +01:00
Zdenek Kabelac
8fc64c5ee6 stats: fix error path when region is NULL
We should not call _stats_cleanup_region_ids() when regions
are NULL.
Also add backtracing for  goto.

(cherry picked from commit 3750b0cff5)
2019-03-05 12:57:40 +01:00
Zdenek Kabelac
8d44cd3e47 libdm: add memory barrier
Just for case ensure compiler is not able to optimize
memset() away for resources that are released.

This idea of using memory barrier is taken from openssl.

Other options would be to check for 'explicit_bzero' function.

(cherry picked from commit 55a8d6c86b)

Conflicts:
	device_mapper/ioctl/libdm-iface.c
2019-03-05 12:48:15 +01:00
Zdenek Kabelac
40f57155a3 libdm: print params only for ioctls using them
When preparing ioctl buffer and flatting all parameters,
add table parameters only to ioctl that do process them.

Note: list of ioctl should be kept in sync with kernel code.
(cherry picked from commit 43f8da7699)

Conflicts:
	WHATS_NEW_DM
	device_mapper/ioctl/libdm-iface.c
2019-03-05 12:47:50 +01:00
Zdenek Kabelac
1bef4dfab3 libdm: add DM_DEVICE_ARM_POLL
Expose DM_DEVICE_ARM_POLL via standard API enum.

(cherry picked from commit 1ae5bf2b83)

Conflicts:
	WHATS_NEW_DM
	device_mapper/all.h
	device_mapper/ioctl/libdm-iface.c
2019-03-05 12:46:21 +01:00
Zdenek Kabelac
e974f6866a cleanup: move cast to det_t into MKDEV macro
(cherry picked from commit aa8b2d6a0f)

Conflicts:
	daemons/clvmd/clvmd-common.h
	device_mapper/ioctl/libdm-iface.c
	device_mapper/libdm-common.c
	device_mapper/libdm-deptree.c
2019-03-05 12:39:17 +01:00
Zdenek Kabelac
a93699ece9 cov: remove unused assigns
(cherry picked from commit 70e3d0a613)

Conflicts:
	tools/pvscan.c
	tools/vgchange.c
2019-03-05 12:28:31 +01:00
Zdenek Kabelac
e4bb94a93e cov: hide intentionaly ptr arithmetic report
Only single region count is ever replaced with on-stack uint64_t.

(cherry picked from commit a91ac41b93)
2019-03-05 12:17:18 +01:00
Zdenek Kabelac
93ac80037a cov: mark warning as expected one
(cherry picked from commit 9238b972c5)

Conflicts:
	base/data-struct/radix-tree-adaptive.c
	device_mapper/libdm-file.c
2019-03-05 12:16:33 +01:00
Zdenek Kabelac
c9e5e6800c cov: split check for type assignment
Check that type is always defined, if not make it explicit internal
error (although logged as debug - so catched only with proper lvm.conf
setting).
This ensures later type being NULL can't be dereferenced with coredump.

(cherry picked from commit 79879bd201)
2019-03-05 12:16:07 +01:00
Zdenek Kabelac
f0f68791f3 cov: shutdown warning
Since previous patch reverted coverity patch as this case is intentional,
provide override this coverity warning.

(cherry picked from commit 05b5774827)
2019-03-05 12:15:40 +01:00
Zdenek Kabelac
b39c26ddc3 revert "cov: dm stats missed terminating null"
This reverts commit 20971f7034
as the parsing of 'dmstatus' started to fail on present \0 char.

(cherry picked from commit 6179cab877)
2019-03-05 12:15:16 +01:00
Zdenek Kabelac
d1ae1455b4 cov: ensure vars are set
Make sure, tmp_begin and tmp_end are always set, even for blind
coverity.

(cherry picked from commit 2513661467)

Conflicts:
	device_mapper/libdm-report.c
2019-03-05 12:14:43 +01:00
Marian Csontos
c115d92287 cov: dmstats check for failing malloc
Add missing check for allocation success.

Backported from: 9b71212262
2019-03-05 12:13:08 +01:00
Zdenek Kabelac
ece117ee10 cov: dm stats missed terminating null
Coverity noticed allocating insufficient memory
for the terminating null of the string.

(cherry picked from commit 20971f7034)
2019-03-05 12:10:36 +01:00