]> sourceware.org Git - lvm2.git/log
lvm2.git
6 years agoallocation: add check for passing log allocation
Zdenek Kabelac [Sun, 8 Jul 2018 22:54:16 +0000 (00:54 +0200)]
allocation: add check for passing log allocation

Updates previous commit.

6 years agotests: check how thin-pool allocation works
Zdenek Kabelac [Sun, 8 Jul 2018 21:24:58 +0000 (23:24 +0200)]
tests: check how thin-pool allocation works

Check allocation of thin-pool works on 2PVs, when one is so full,
that even metadata do not fit there (as they need at least 2M,
while 99% of 63MB fills >62MB)

6 years agocleanup: use last_seg
Zdenek Kabelac [Sat, 7 Jul 2018 19:35:07 +0000 (21:35 +0200)]
cleanup: use last_seg

More readable code.

6 years agobuild: libdm preload dir is no longer needed
Zdenek Kabelac [Tue, 3 Jul 2018 09:12:04 +0000 (11:12 +0200)]
build: libdm preload dir is no longer needed

Since we do not build lvm code with libdm, drop preload.

6 years agodev_io: no discard in testmode
Zdenek Kabelac [Sun, 8 Jul 2018 19:11:01 +0000 (21:11 +0200)]
dev_io: no discard in testmode

When lvm2 command is executed in test mode, discard ioctl is skipped.
This may cause even data-loose in case, issuing discard for released
areas was enabled and user 'tested'  lvreduce.

6 years agoallocator: fix thin-pool allocation
Zdenek Kabelac [Sun, 8 Jul 2018 16:34:38 +0000 (18:34 +0200)]
allocator: fix thin-pool allocation

When allocating thin-pool with more then 1 device - try to
allocate 'metadataLV' with reuse of log-type allocation for mirror LV.
It should be naturally place on other device then 'dataLV'.

However due to somewhat hard to follow allocation logic code,
it's been rejected allocation in cases where there was not
enough space for data or metadata on single PV, thus to successed,
usage of segments was mandatory.

While user may use:

allocation/thin_pool_metadata_require_separate_pvs=1

to enforce separe meta and data LV - on default settings, this is not
enable thus segment allocation is meant to work.

NOTE:

As already said - the original intention of this whole  'if()' is unclear,
so try to split this test into multiple more simple tests that are more readable.

TODO: more validation.

6 years agovdo: enhance status parser
Zdenek Kabelac [Wed, 27 Jun 2018 14:18:53 +0000 (16:18 +0200)]
vdo: enhance status parser

Add support for using mempool for allocations inside status parser.
Convert some string pointers to arrays.
Reindent tabs.

6 years agodevice_mapper: relocate code for sending messages
Zdenek Kabelac [Sat, 23 Jun 2018 21:03:25 +0000 (23:03 +0200)]
device_mapper: relocate code for sending messages

To be able to send messages for recently resumed devices,
move code into inner loop.

6 years agodevice_mapper: deactive new nodes when load fails
Zdenek Kabelac [Fri, 29 Jun 2018 15:25:38 +0000 (17:25 +0200)]
device_mapper: deactive new nodes when load fails

When node loading fails, there is not much the caller can do,
since there is 'unknown' set of devices preloaded.

Only suspend during preload knows future precommitted 'metadata',
so it's non-trivial to drop 'preloaded' entries with any later call.

However dm tree tracks newly loaded entries - so in this case it
may simplify the recovery path by dropping preloaded entries so
they are not leaked in the DM table.

6 years agolv_manip: do not check extents for any virtual target
Zdenek Kabelac [Fri, 29 Jun 2018 09:25:08 +0000 (11:25 +0200)]
lv_manip: do not check extents for any virtual target

Allow creation of any virtual segment type with just --virtualsize
specified without any real extent size give.

TODO: likely --type error,zero might be later enhanced to use -V
(along with -L) - but since those targets do not allocate real
space, supporting -V makes sense with them.

6 years agolv_manip: add name of failing LV into error message
Zdenek Kabelac [Fri, 29 Jun 2018 15:26:01 +0000 (17:26 +0200)]
lv_manip: add name of failing LV into error message

6 years agomemlock: extend exception list
Zdenek Kabelac [Sat, 30 Jun 2018 09:05:14 +0000 (11:05 +0200)]
memlock: extend exception list

Amound of linked libraries grows.
Most of them we don't need to lock in, since we are not using
them in locked section, so skip locking them in memory.

6 years agolocking: memory locking ONLY with suspending reason
Zdenek Kabelac [Sat, 30 Jun 2018 08:35:08 +0000 (10:35 +0200)]
locking: memory locking ONLY with suspending reason

It's important to lock memory beforo running SUSPEND ioctl - but whole
lvm preload runs in memory unlocked environment - as in this phase
memory allocation is allowed and is meant to happen.

Once all targets are preload and ready (confirmed from all targets)
we start suspending tree - and here the memory allocation (or i.e.
opening files) is no longer allowed - as it may cause kernel deadlock.

6 years agobuild: drop some more old files
Zdenek Kabelac [Sun, 1 Jul 2018 13:01:41 +0000 (15:01 +0200)]
build: drop some more old files

6 years agobuild: avoid rebuild deps for top-level makefiles
Zdenek Kabelac [Sun, 1 Jul 2018 13:16:18 +0000 (15:16 +0200)]
build: avoid rebuild deps for top-level makefiles

6 years agodmsetup: fix error propagation in _display_info_cols()
Bryn M. Reeves [Thu, 28 Jun 2018 13:25:30 +0000 (14:25 +0100)]
dmsetup: fix error propagation in _display_info_cols()

Commit 3f35146 added a check on the value returned by the
_display_info_cols() function:

  1024         if (!_switches[COLS_ARG])
  1025                 _display_info_long(dmt, &info);
  1026         else
  1027                 r = _display_info_cols(dmt, &info);
  1028
  1029         return r;

This exposes a bug in the dmstats code in _display_info_cols:
the fact that a device has no regions is explicitly not an error
(and is documented as such in the code), but since the return
code is not changed before leaving the function it is now treated
as an error leading to:

  # dmstats list
  Command failed.

When no regions exist.

Set the return code to the correct value before returning.

6 years agoRevert "man: fix lvreduce example"
David Teigland [Wed, 27 Jun 2018 14:19:01 +0000 (09:19 -0500)]
Revert "man: fix lvreduce example"

 -l -3 is correct, meaning reduce by 3.

This reverts commit d5bcc56eefde8349ffee93d0411c6f9fd71ceb84.

6 years agoman: fix lvreduce example
David Teigland [Wed, 27 Jun 2018 13:58:22 +0000 (08:58 -0500)]
man: fix lvreduce example

6 years agobcache: Fix null pointer dereferencing
Marian Csontos [Tue, 26 Jun 2018 15:04:18 +0000 (17:04 +0200)]
bcache: Fix null pointer dereferencing

6 years agodevice_mapper: add new _dm_task_create_device_status
Zdenek Kabelac [Sat, 23 Jun 2018 21:02:24 +0000 (23:02 +0200)]
device_mapper: add new _dm_task_create_device_status

Introduce new function _dm_task_create_device_status for grabbing
status of device for better code sharing.

6 years agodevice_mapper: split code for sending message
Zdenek Kabelac [Sat, 23 Jun 2018 19:00:40 +0000 (21:00 +0200)]
device_mapper: split code for sending message

Move message sending from _thin_pool_node_message to
new _node_message for possible better code sharing.

6 years agodevice_mapper: split _node_send_message
Zdenek Kabelac [Sat, 23 Jun 2018 18:50:36 +0000 (20:50 +0200)]
device_mapper: split _node_send_message

For better code reuse split _node_send_messages into commont
messaging part and separate _thin_pool_node_send_messages.

Patch makes it possible to better reuse common code for messaging
other targets.

6 years agotests: update with --yes
Zdenek Kabelac [Sun, 24 Jun 2018 18:49:59 +0000 (20:49 +0200)]
tests: update with --yes

vgcfgrestore needs to confirm restore while LVs from VG are present.

6 years agolv_manip: use vgmem pool
Zdenek Kabelac [Sat, 23 Jun 2018 09:59:35 +0000 (11:59 +0200)]
lv_manip: use vgmem pool

Switch to vgmem pool for allocation associated with modification
of particular VG.

6 years agocache: use new api function
Zdenek Kabelac [Sat, 23 Jun 2018 09:47:33 +0000 (11:47 +0200)]
cache: use new api function

6 years agolv_manip: add new internal api function
Zdenek Kabelac [Sat, 23 Jun 2018 09:35:22 +0000 (11:35 +0200)]
lv_manip: add new internal api function

6 years agocache: set areas count prior using it
Zdenek Kabelac [Sat, 23 Jun 2018 09:35:34 +0000 (11:35 +0200)]
cache: set areas count prior using it

Set correct counter, so it's not failing on internal error check.

6 years agovcfgrestore: add prompt with active volumes
Zdenek Kabelac [Fri, 22 Jun 2018 16:45:48 +0000 (18:45 +0200)]
vcfgrestore: add prompt with active volumes

Add check for active device with names matching restored VG.
When such devices are present in dm table, prompt user, if he
wish to continue.

6 years agolv_manip: add extra internal error
Zdenek Kabelac [Wed, 20 Jun 2018 08:44:13 +0000 (10:44 +0200)]
lv_manip: add extra internal error

Catch error early, when trying to store data into non-allocated area.

6 years agoutils: add clzll
Zdenek Kabelac [Tue, 19 Jun 2018 17:35:48 +0000 (19:35 +0200)]
utils: add clzll

Check for __builtin_clzll and add wrapper when missing.

6 years agotests: fix rules for mke2fs.conf install
Zdenek Kabelac [Fri, 22 Jun 2018 20:43:58 +0000 (22:43 +0200)]
tests: fix rules for mke2fs.conf install

6 years agobuild: support --disable-silent-rules
Zdenek Kabelac [Fri, 22 Jun 2018 21:30:42 +0000 (23:30 +0200)]
build: support --disable-silent-rules

Add support for standardized option for have verbose builds.
Useful for distro builds where more details can be useful.

6 years agobuild: include configure.h
Zdenek Kabelac [Tue, 19 Jun 2018 18:43:31 +0000 (20:43 +0200)]
build: include configure.h

It's important to consistenly include  configure.h as the 1st. header.
It containts #defines influencing behavior of other included header
files.

6 years agobuild: make generate
Zdenek Kabelac [Tue, 19 Jun 2018 13:07:50 +0000 (15:07 +0200)]
build: make generate

6 years agoradix-tree: squash a pointer arithmetic warning
Joe Thornber [Thu, 21 Jun 2018 16:41:56 +0000 (17:41 +0100)]
radix-tree: squash a pointer arithmetic warning

6 years agoMerge branch 'master' of git+ssh://sourceware.org/git/lvm2
Joe Thornber [Thu, 21 Jun 2018 16:12:09 +0000 (17:12 +0100)]
Merge branch 'master' of git+ssh://sourceware.org/git/lvm2

6 years agoradix-tree: fix bug when erasing elts in remove_prefix
Joe Thornber [Thu, 21 Jun 2018 16:10:05 +0000 (17:10 +0100)]
radix-tree: fix bug when erasing elts in remove_prefix

_erase_elt() now zeroes the last element of the array (ie. sets to
UNSET).  Previously remove() was doing this, but not remove_prefix().

6 years agofilter: use pointers to real addresses
David Teigland [Thu, 21 Jun 2018 15:52:35 +0000 (10:52 -0500)]
filter: use pointers to real addresses

instead of casting values 1 and 2 to pointers
which gcc optimization can have problems with.

6 years agoRemove code for using files as devices
David Teigland [Thu, 21 Jun 2018 14:33:21 +0000 (09:33 -0500)]
Remove code for using files as devices

It appears this has not been used in a long time,
and it seems to have no point since loop devices exist.

6 years agolvmlockd: fix another missing lock_type null check
David Teigland [Thu, 21 Jun 2018 14:00:23 +0000 (09:00 -0500)]
lvmlockd: fix another missing lock_type null check

Same as 347c807f8.

6 years agoradix-tree: More debugging of remove
Joe Thornber [Wed, 20 Jun 2018 09:04:59 +0000 (10:04 +0100)]
radix-tree: More debugging of remove

There's now a pretty printer called radix_tree_dump()

n4, n16, and n48 weren't UNSETting the last entry after
sliding values down.

6 years agoradix_tree: add new test case
Joe Thornber [Tue, 19 Jun 2018 12:38:13 +0000 (13:38 +0100)]
radix_tree: add new test case

Check that value destructors are called by radix_tree_destroy()

6 years agoradix-tree: FIx various bugs to do with removal
Joe Thornber [Tue, 19 Jun 2018 09:19:06 +0000 (10:19 +0100)]
radix-tree: FIx various bugs to do with removal

Add radix_tree_is_well_formed() which does some sanity checking
of the tree.

Call the above a lot in the unit tests.

Fix revealed bugs.

6 years agoscan: work around udev problems by avoiding open RDWR
David Teigland [Wed, 20 Jun 2018 16:32:45 +0000 (11:32 -0500)]
scan: work around udev problems by avoiding open RDWR

udev creates a train wreck of events if we open devices
with RDWR.  Until we can fix/disable/scrap udev, work around
this by opening RDONLY and then closing/reopening RDWR when
a write is needed.  This invalidates the bcache blocks for
the device before writing so it can trigger unnecessary
rereading.

6 years agobcache: remove extraneous error message
David Teigland [Mon, 18 Jun 2018 16:59:57 +0000 (11:59 -0500)]
bcache: remove extraneous error message

an error from io_submit is already recognized by
the caller like errors during completion.

6 years agoPrint advice about changing clustered VGs to shared
David Teigland [Mon, 18 Jun 2018 15:59:11 +0000 (10:59 -0500)]
Print advice about changing clustered VGs to shared

6 years agoDrop --ignoreskippedcluster option
David Teigland [Fri, 15 Jun 2018 20:43:59 +0000 (15:43 -0500)]
Drop --ignoreskippedcluster option

It's no longer needed.  Clustered VGs are now handled in
the same way as foreign VGs, and as shared VGs that
can't be accessed:

- A command processing all VGs sees a clustered VG,
  prints a message ("Skipping clustered VG foo."),
  skips it, and does not fail.

- A command where the clustered VG is explicitly
  named on the command line, prints a message and fails.
  "Cannot access clustered VG foo, see lvmlockd(8)."

The option is listed in the set of ignored options for
the commands that previously accepted it.  (Removing it
entirely would cause commands/scripts to fail if they
set it.)

6 years agoreport: show empty lock_type for none
David Teigland [Fri, 15 Jun 2018 19:14:39 +0000 (14:14 -0500)]
report: show empty lock_type for none

Sometimes lock_type would be displayed as "none"
(after changing it) and sometimes as empty.
Make it consistently empty.

6 years agoRemove unused device error counting
David Teigland [Fri, 15 Jun 2018 19:04:39 +0000 (14:04 -0500)]
Remove unused device error counting

6 years agoconfig: add deprecated version for recently removed settings
David Teigland [Fri, 15 Jun 2018 18:56:26 +0000 (13:56 -0500)]
config: add deprecated version for recently removed settings

assumes that the next version from this branch is 3.0.0

6 years agoscan: use full md filter when md 1.0 devices are present
David Teigland [Fri, 15 Jun 2018 16:42:10 +0000 (11:42 -0500)]
scan: use full md filter when md 1.0 devices are present

The md filter can operate in two native modes:
- normal: reads only the start of each device
- full: reads both the start and end of each device

md 1.0 devices place the superblock at the end of the device,
so components of this version will only be identified and
excluded when lvm uses the full md filter.

Previously, the full md filter was only used in commands
that could write to the device.  Now, the full md filter
is also applied when there is an md 1.0 device present
on the system.  This means the 'pvs' command can avoid
displaying md 1.0 components (at the cost of doubling
the i/o to every device on the system.)

(The md filter can operate in a third mode, using udev,
but this is disabled by default because there have been
problems with reliability of the info returned from udev.)

6 years agoAdd cmd arg to more functions
David Teigland [Fri, 15 Jun 2018 16:03:55 +0000 (11:03 -0500)]
Add cmd arg to more functions

so that it can be used in the filter code

6 years agorpm: drop no longer present clvmd, lvm2app
Zdenek Kabelac [Thu, 14 Jun 2018 22:47:35 +0000 (00:47 +0200)]
rpm: drop no longer present clvmd, lvm2app

6 years agotests: more tolerable makefile
Zdenek Kabelac [Thu, 14 Jun 2018 22:46:54 +0000 (00:46 +0200)]
tests: more tolerable makefile

6 years agoscripts: clvmd gone
Zdenek Kabelac [Thu, 14 Jun 2018 22:46:24 +0000 (00:46 +0200)]
scripts: clvmd gone

6 years agoman: stop installing clvmd man page
Zdenek Kabelac [Thu, 14 Jun 2018 22:46:08 +0000 (00:46 +0200)]
man: stop installing clvmd man page

6 years agoman-generator: drop macro redefines
Zdenek Kabelac [Thu, 14 Jun 2018 21:15:28 +0000 (23:15 +0200)]
man-generator: drop macro redefines

6 years agotests: drop some clvmd refs
Zdenek Kabelac [Thu, 14 Jun 2018 21:14:32 +0000 (23:14 +0200)]
tests: drop some clvmd refs

Do not try to link clvmd binary.
Ensure lib is created new and does not refer old binaries.

6 years agobuild: cmirrord with internal dm lib
Zdenek Kabelac [Thu, 14 Jun 2018 21:14:04 +0000 (23:14 +0200)]
build: cmirrord with internal dm lib

6 years agotests: bigger lv
Zdenek Kabelac [Thu, 14 Jun 2018 19:52:22 +0000 (21:52 +0200)]
tests: bigger lv

Although throttling slows down things considerable, it still could
reach the end before next test so use bigger LV.

6 years agodebug: missing trace
Zdenek Kabelac [Wed, 13 Jun 2018 13:56:58 +0000 (15:56 +0200)]
debug: missing trace

6 years agosystemd: add conficting sockets
Zdenek Kabelac [Tue, 12 Jun 2018 14:27:42 +0000 (16:27 +0200)]
systemd: add conficting sockets

Since we are using "DefaultDependencies=no" we do not get automatic STOP
job on socket connection - so automatically refuse connection on
shutdown by adding this Conflict definition to socket Unit.

6 years agovgchange: start polling with activation
Zdenek Kabelac [Thu, 14 Jun 2018 19:05:41 +0000 (21:05 +0200)]
vgchange: start polling with activation

Shuffle code for better readability as set of conditions was
hard to follow.

Make it obvious the refresh & activate path is handling
monitoring and polling on its own.

So the only --monitor and --poll option needs explicit care.
Option --monitor without option --poll will now as a result
of this patch NOT start polling.

So command: vgchange --monitor n    is no longer a polling starter.

6 years agopvscan: move start of polling into vgchange
Zdenek Kabelac [Thu, 14 Jun 2018 19:07:59 +0000 (21:07 +0200)]
pvscan: move start of polling into vgchange

Restoring polling for activated volumes lost with my recent commit:
75fed05d3ef648583764ff56cc577e0f3eba1bba and move start of polling
directly into _activate_lvs_in_vg() - as there we know exactly
if there was some volume even activated.

Also make it sharing same code for pvscan -aay.

6 years agopvscan: code reshape
Zdenek Kabelac [Thu, 14 Jun 2018 19:07:15 +0000 (21:07 +0200)]
pvscan: code reshape

6 years agovgchange: trace faling activation
Zdenek Kabelac [Thu, 14 Jun 2018 19:04:53 +0000 (21:04 +0200)]
vgchange: trace faling activation

Trace failed activation and directly assign 0 returning failure.

6 years agovgchange: move active assing
Zdenek Kabelac [Thu, 14 Jun 2018 19:00:16 +0000 (21:00 +0200)]
vgchange: move active assing

Make eval of activate_ARG reusable.

6 years agovgchange: fix error code in error path
Zdenek Kabelac [Wed, 13 Jun 2018 13:57:51 +0000 (15:57 +0200)]
vgchange: fix error code in error path

This rather hard to hit error path used wrong return value to signal
real error.

6 years agodevice_mapper: drop unneeded function
Zdenek Kabelac [Thu, 14 Jun 2018 18:47:23 +0000 (20:47 +0200)]
device_mapper: drop unneeded function

Subdir without stats.

6 years agobuild: better srcdir builddir support
Zdenek Kabelac [Thu, 14 Jun 2018 18:57:02 +0000 (20:57 +0200)]
build: better srcdir builddir support

With the move to top-level makefile - there are some issues
with subdir recursive makefile.
Make the building more tolerant for now until fully resolved.

6 years agobuild: drop libdm referring from lvm code
Zdenek Kabelac [Thu, 14 Jun 2018 18:48:08 +0000 (20:48 +0200)]
build: drop libdm referring from lvm code

Avoid adding /libdm  paths into lvm building.

6 years agobuild: drop some lvm references from libdm making
Zdenek Kabelac [Thu, 14 Jun 2018 18:38:08 +0000 (20:38 +0200)]
build: drop some lvm references from libdm making

Some simplification, more may follow...

6 years agobuild: ensure libdm is built before dm-tools
Zdenek Kabelac [Thu, 14 Jun 2018 18:37:33 +0000 (20:37 +0200)]
build: ensure libdm is built before dm-tools

Making libs before entering dm-tools subdir,
so the tool will not link i.e. system library if present.

6 years agoman: update lvmsystemid wording
David Teigland [Thu, 14 Jun 2018 17:34:00 +0000 (12:34 -0500)]
man: update lvmsystemid wording

to refer to "shared VG" instead of "lockd VG".

6 years agoman: updates to lvmlockd
David Teigland [Thu, 14 Jun 2018 17:30:45 +0000 (12:30 -0500)]
man: updates to lvmlockd

The terminology has migrated toward using "shared VG"
rather than "lockd VG".

Also improve the wording in a number of places.

6 years agodevice_mapper: remove libdm-stats.c
Joe Thornber [Thu, 14 Jun 2018 13:32:17 +0000 (14:32 +0100)]
device_mapper: remove libdm-stats.c

We don't use it in lvm.

6 years agodmfilemapd: Move to libdm/dm-tools
Joe Thornber [Thu, 14 Jun 2018 13:27:19 +0000 (14:27 +0100)]
dmfilemapd: Move to libdm/dm-tools

No longer uses any lvm code.

6 years agodmsetup: move to libdm/dm-tools/dmsetup
Joe Thornber [Thu, 14 Jun 2018 12:08:33 +0000 (13:08 +0100)]
dmsetup: move to libdm/dm-tools/dmsetup

links against libdevmapper again.
no longer includes code from lvm.

6 years agolvmlockd: update method for changing clustered VG
David Teigland [Wed, 13 Jun 2018 20:30:28 +0000 (15:30 -0500)]
lvmlockd: update method for changing clustered VG

The previous method for forcibly changing a clustered VG
to a local VG involved using -cn and locking_type 0.
Since those options are deprecated, replace it with
the same command used for other forced lock type changes:
vgchange --locktype none --lockopt force.

6 years agoRemove makefile entries for removed script
David Teigland [Wed, 13 Jun 2018 20:02:12 +0000 (15:02 -0500)]
Remove makefile entries for removed script

6 years agoRemove systemd script for starting shared VG
David Teigland [Tue, 12 Jun 2018 19:35:27 +0000 (14:35 -0500)]
Remove systemd script for starting shared VG

Shared VGs will generally be started and activated by
the resource agent.  Without the agent, this script doesn't
have a good way to know which LVs to activate.

6 years agotests: remove vgconvert usage
David Teigland [Wed, 13 Jun 2018 19:16:28 +0000 (14:16 -0500)]
tests: remove vgconvert usage

6 years agotests: remove metadata-dirs
David Teigland [Wed, 13 Jun 2018 19:14:23 +0000 (14:14 -0500)]
tests: remove metadata-dirs

metadata dirs are removed

6 years agoRemove vgconvert
David Teigland [Wed, 13 Jun 2018 19:12:46 +0000 (14:12 -0500)]
Remove vgconvert

it has no use without lvm1

6 years agofilters: remove cache file in persistent filter
David Teigland [Wed, 13 Jun 2018 19:00:47 +0000 (14:00 -0500)]
filters: remove cache file in persistent filter

It creates problems because it's not always correct,
and it doesn't actually help much.

6 years agoRemove independent metadata areas
David Teigland [Wed, 13 Jun 2018 17:25:19 +0000 (12:25 -0500)]
Remove independent metadata areas

in which metadata is stored in files on the local fs
instead of on PVs.

6 years agoRemove code for loading other metadata formats
David Teigland [Wed, 13 Jun 2018 16:32:29 +0000 (11:32 -0500)]
Remove code for loading other metadata formats

other formats are not used.

6 years agotests: lvmetad-pvscan-cache expect command to fail
David Teigland [Tue, 12 Jun 2018 17:44:23 +0000 (12:44 -0500)]
tests: lvmetad-pvscan-cache expect command to fail

6 years agotests: lvconvert-repair remove cluster test
David Teigland [Tue, 12 Jun 2018 16:35:45 +0000 (11:35 -0500)]
tests: lvconvert-repair remove cluster test

6 years agoRemove the unused lock_hash in lvmcache
David Teigland [Tue, 12 Jun 2018 16:29:56 +0000 (11:29 -0500)]
Remove the unused lock_hash in lvmcache

It kept track of which VGs were locked, but is
no longer used, so remove it.

6 years agoClean up repair and result values in vg_read
David Teigland [Tue, 12 Jun 2018 14:44:37 +0000 (09:44 -0500)]
Clean up repair and result values in vg_read

Fix the confusing mix of input and output values
in the single variable.

6 years agoFix use of orphan lock in commands
David Teigland [Mon, 11 Jun 2018 20:08:23 +0000 (15:08 -0500)]
Fix use of orphan lock in commands

vgreduce, vgremove and vgcfgrestore were acquiring
the orphan lock in the midst of command processing
instead of at the start of the command.  (The orphan
lock moved to being acquired at the start of the
command back when pvcreate/vgcreate/vgextend were
reworked based on pvcreate_each_device.)

vgsplit also needed a small update to avoid reacquiring
a VG lock that it already held (for the new VG name).

6 years agoRemove checking for locked VGs
David Teigland [Mon, 11 Jun 2018 17:25:52 +0000 (12:25 -0500)]
Remove checking for locked VGs

A few places were calling a function to check if a
VG lock was held.  The only place it was actually
needed is for pvcreate which wants to do its own
locking (and scanning) around process_each_pv.

The locking/scanning exceptions for pvcreate in
process_each_pv/vg_read can be enabled by just passing
a couple of flags instead of checking if the VG is
already locked.  This also means that these special
cases won't be enabled unknowingly in other places
where they shouldn't be used.

6 years agolvmlockd: skip repair lock upgrade for non shared vgs
David Teigland [Tue, 12 Jun 2018 14:25:51 +0000 (09:25 -0500)]
lvmlockd: skip repair lock upgrade for non shared vgs

Only attempt lvmlockd lock upgrade for shared VGs.

6 years agobuild: ensure configure.h comes first
Zdenek Kabelac [Mon, 11 Jun 2018 20:38:51 +0000 (22:38 +0200)]
build: ensure configure.h comes first

Fix header order so configure.h is 1st. included header.

6 years agosnapshot: improve checking of merging snapshot
Zdenek Kabelac [Mon, 11 Jun 2018 20:19:20 +0000 (22:19 +0200)]
snapshot: improve checking of merging snapshot

Add runtime detection for 'lvs -o+seg_monitor' and 'vgchange --monitor'.
This fix should avoid unnecessary timeout on systemd shutdown.

6 years agovgchange: start polling with option
Zdenek Kabelac [Mon, 11 Jun 2018 20:18:00 +0000 (22:18 +0200)]
vgchange: start polling with option

Polling start either with '--refresh'
or with '--poll' option specified.

6 years agobuild: use internal libs for lvm2cmd
Zdenek Kabelac [Mon, 11 Jun 2018 20:15:49 +0000 (22:15 +0200)]
build: use internal libs for lvm2cmd

6 years agobuild: link dmeventd plugins with internal libs
Zdenek Kabelac [Mon, 11 Jun 2018 19:09:45 +0000 (21:09 +0200)]
build: link dmeventd plugins with internal libs

6 years agobuild: make generate
Zdenek Kabelac [Fri, 8 Jun 2018 20:16:00 +0000 (22:16 +0200)]
build: make generate

This page took 0.071825 seconds and 5 git commands to generate.