1
0
mirror of git://sourceware.org/git/lvm2.git synced 2024-12-22 17:35:59 +03:00
Commit Graph

9709 Commits

Author SHA1 Message Date
Peter Rajnoha
a976226e81 cleanup: remove compiler warning about possible uninitialized variable use
filters/filter-usable.c:22: warning: "ucp.check_..." may be used uninitialized in this function

This can't actually be hit in real, but let's clean this up for the compiler
to be happy again.
2014-10-02 13:21:24 +02:00
Peter Rajnoha
5011cac9cf filters: add cmd->full_filter - composite of cmd->filter and cmd->lvmetad_filter
There are actually three filter chains if lvmetad is used:
  - cmd->lvmetad_filter used when when scanning devices for lvmetad
  - cmd->filter used when processing lvmetad responses
  - cmd->full_fiilter (which is just cmd->lvmetad_filter + cmd->filter chained together) used
    for remaining situations

This patch adds the third one - "cmd->full_filter" - currently this is
used if device processing does not fall into any of the groups before,
for example, devices which does not have the PV label yet and we're just
creating a new one or we're processing the devices where the list of the
devices (PVs) is not returned by lvmetad initially.

Currently, the cmd->full_filter is used exactly in these functions:
  - lvmcache_label_scan
  - _pvcreate_check
  - pvcreate_vol
  - lvmdiskscan
  - pvscan
  - _process_each_label

If lvmetad is used, then simply cmd->full_filter == cmd->filter because
cmd->lvmetad_filter is NULL in this case.
2014-10-02 13:06:46 +02:00
Peter Rajnoha
d3fb69c3f3 cleanup: remove symlink to profile in test dir for CLEAN target 2014-10-02 09:30:38 +02:00
David Teigland
1cdb8766a7 tools: Add ENABLE_ALL_DEVS flag.
The ENABLE_ALL_DEVS flag is added to the command structure
for commands that should process all devs (pvs and non-pvs)
when they call process_each_pv and the command includes the
--all arg.  This will be used in a later process_each_pv patch.
2014-10-01 22:58:23 +01:00
David Teigland
e6ab275aa0 tools: Add ALL_VGS_IS_DEFAULT flag.
The ALL_VGS_IS_DEFAULT flag is added to the command structure
for commands that should process all vgs when they call
process_each_vg or process_each_lv with no args.
This will be used in later patches to process_each functions.
2014-10-01 22:58:00 +01:00
David Teigland
d66f257452 man: lvmthin mention alternate syntax
Users will probably run across the alternate syntax,
so we should explain how it relates to what is used here.
2014-10-01 10:44:18 -05:00
David Teigland
33b96bef5b Revert "Revert "man: lvmcache should use clearer cache pool options""
This reverts commit f120c954fc.
2014-10-01 10:43:53 -05:00
Peter Rajnoha
a5f01dad22 filters: refresh filters when lvmetad use is toggled
We need to use proper filter chain when we disable lvmetad use
explicitly in the code by calling lvmetad_set_active(0) while
overriding existing configuration. We need to reinitialize filters
in this case so proper filter chain is used. The same applies
for the other way round - when we enable lvmetad use explicitly in
the code (though this is not yet used).
2014-09-30 16:08:05 +02:00
Peter Rajnoha
c2981cf921 filters: use usable device filter and separate lvmetad filter chain so it's not reevaluated for any lvmetad response
With this change, the filter chains used look like this now:

  A) When *lvmetad is not used*:
    - persistent filter -> regex filter -> sysfs filter ->
      global regex filter -> type filter ->
      usable device filter(FILTER_MODE_NO_LVMETAD) ->
      mpath component filter -> partitioned filter ->
      md component filter

  B) When *lvmetad is used* (two separate filter chains):
     - the lvmetad filter chain used when scanning devs for lvmetad update:
       sysfs filter -> global regex filter -> type filter ->
       usable device filter(FILTER_MODE_PRE_LVMETAD) ->
       mpath component filter -> partitioned filter ->
       md component filter

     - the filter chain used for lvmetad responses:
       persistent filter -> usable device filter(FILTER_MODE_POST_LVMETAD) ->
       regex filter
2014-09-30 13:22:11 +02:00
Peter Rajnoha
8a843d0d97 filters: add "usable device" filter
Usable device filter is responsible for filtering out unusable DM devices.
The filter has 3 modes of operation:

  - FILTER_MODE_NO_LVMETAD:
    When this mode is used, we check DM device usability by looking:
      - whether device is empty
      - whether device is blocked
      - whether device is suspended (only on devices/ignore_suspended_devices=1)
      - whether device uses an error target
      - whether device name/uuid is reserved

  - FILTER_MODE_PRE_LVMETAD:
    When this mode is used, we check DM device usability by looking:
      - whether device is empty
      - whether device is suspended (only on devices/ignore_suspended_devices=1)
      - whether device uses an error target
      - whether device name/uuid is reserved

  - FILTER_MODE_POST_LVMETAD:
    When this mode is used, we check DM device usability by looking:
      - whether device is blocked
      - whether device is suspended (only on devices/ignore_suspended_devices=1)

These modes will be used by subsequent patch to create different
instances of this filter, depending on lvmetad use.
2014-09-30 13:11:58 +02:00
Peter Rajnoha
00d8ab8492 refactor: make it possible to select what to check exactly when calling device_is_usable fn
Currently, there are 5 things that device_is_usable function checks
(for DM devices only, of course):
  - is device empty?
  - is device blocked? (mirror)
  - is device suspended?
  - is device composed of an error target?
  - is device name/uuid reserved?

If answer to any of these questions is "yes", then the device is not usable.
This patch just adds possibility to choose what to check for exactly - the
device_is_usable function now accepts struct dev_usable_check_params make
this selection possible. This is going to be used by subsequent patches.
2014-09-30 13:11:58 +02:00
Petr Rockai
fbc28cc7ad conf: Update comments on lvmetad+filters in example.conf.in. 2014-09-30 11:39:07 +02:00
David Teigland
9f3c11b39a man: lvmthin remove unnecessary fixme
The existing method for single step thin-pool + thin creation
makes sense.
2014-09-29 12:28:00 -05:00
David Teigland
9646c3359f Revert "Revert "man: lvmthin should use clearer thin pool options""
This reverts commit 17a1869df5.

We've agreed on the clearer syntax.
2014-09-29 12:26:16 -05:00
Zdenek Kabelac
d1be66ba37 valgrind: don't eat mem with valgrind
When compiled with valgrind pool support - don't waste time
with preallocation of memory - it just waste of CPU cycles to
trace access to this memory.

We also may get slightly better estimation about real memory usage
during command processing.
2014-09-28 13:49:01 +02:00
Zdenek Kabelac
f3e9ff7179 cleanup: drop unused variable 2014-09-28 13:49:01 +02:00
Zdenek Kabelac
ad60805ffd lvconvert: switch to validate_lvname_param
Use new toollib function for validation.
2014-09-28 13:49:01 +02:00
Zdenek Kabelac
0d4baeba18 toollib: introduce validate_lvname_param
Function for parsing and validating of lvname parameter.
2014-09-28 13:49:01 +02:00
Zdenek Kabelac
89e1190ef0 cleanup: rename func
Update name of function read_and_validate_major_minor
and put it into the right header file toollib.h.

(In release update for f09f85d027)
2014-09-28 13:49:01 +02:00
Zdenek Kabelac
911d6efa51 toollib: refactor extract_vgname
Split internals of extract_vgname into _extract_vgname.
This common code will be used for other similar function.

Reuse skip_dev_dir() instead of less mature coded to skip
device dir.

Instead of duplicating full vg/lv name - allocate string
only vg portion of lv name.
2014-09-28 13:49:01 +02:00
Zdenek Kabelac
b0dde9e8f0 toollib: refactor skip_dev_dir
Detect dev dir just once.
2014-09-28 13:49:01 +02:00
Zdenek Kabelac
f120c954fc Revert "man: lvmcache should use clearer cache pool options"
Revert: --cachepool  and --thipnpool is the right way

they both take arg

This is a major difference from  --cache   and --thin
2014-09-27 15:10:44 +02:00
Zdenek Kabelac
17a1869df5 Revert "man: lvmthin should use clearer thin pool options"
Revert: --cachepool  and --thipnpool is the right way

they both take arg

This is a major difference from  --cache   and --thin
2014-09-27 15:08:31 +02:00
David Teigland
95df06d721 man: lvmthin should use clearer thin pool options
Previously, this was the recommended form for creating a thin pool:

lvconvert --thinpool VG/ThinDataLV --poolmetadata VG/ThinMetaLV

but this is confusing, because --thinpool does not actually take
an arg, and is more naturally used to specify an existing thin pool.

The new recommended form is:

lvconvert --type thin-pool --poolmetadata VG/ThinMetaLV VG/ThinDataLV
2014-09-26 16:05:30 -05:00
David Teigland
b53504ed37 man: lvmcache should use clearer cache pool options
Previously, this was the recommended form for creating a cache pool:

lvconvert --cachepool VG/CacheDataLV --poolmetadata VG/CacheMetaLV

but this is confusing, because --cachepool does not actually take
an arg, and is more natually used to specify an existing cache pool.

The new recommended form is:

lvconvert --type cache-pool --poolmetadata VG/CacheMetaLV VG/CacheDataLV
2014-09-26 14:40:40 -05:00
Zdenek Kabelac
26dd17f834 lvchange: missed bits for a8aee7dba2
lv_info is not used here any more.
This code should have been committed with
lv_check_not_in_use API change commit.
2014-09-24 16:30:02 +02:00
Heinz Mauelshagen
45f57477f4 cleanup: Use segtype.h definitions of segment type names wherever possible
We are not using already defined segement type names where we could.

There is a lot of other places in device-mapper and LVM2 we have those
hardcoded so we should better finally have a common interface in
libdevmapper to avoid this.
2014-09-24 15:24:41 +02:00
Zdenek Kabelac
21aa850b43 tests: add helper funcionality for using gdb
When there is need for using gdb within test suite
just prefix command with LVM_DBG and run test in
standard shell (for stdin).
2014-09-24 10:54:48 +02:00
Zdenek Kabelac
9acbb0695f tests: drop * from dmsetup
We cannot use shell expansion for dmsetup here.
2014-09-24 10:54:48 +02:00
Zdenek Kabelac
bc5031c283 debug: add debug message
Since we leave error printing on the called of deactivation,
at least put in debug log for this case.
2014-09-24 10:54:48 +02:00
Zdenek Kabelac
edb3902c25 debug: show stacktrace on error path 2014-09-24 10:54:48 +02:00
Zdenek Kabelac
2bfd986ea3 debug: drop printing debug trace without labeler
Not having labeller on device is not reason eo generate backtrace in log.
2014-09-24 10:54:48 +02:00
Zdenek Kabelac
4319e06a0f debug: add missing stack trace 2014-09-24 10:54:48 +02:00
Zdenek Kabelac
7531a9169e debug: monitor_dev_for_events stack trace 2014-09-24 10:54:48 +02:00
Zdenek Kabelac
03aeb86762 cleanup: reindent
Save some code lines.
2014-09-24 10:54:48 +02:00
Zdenek Kabelac
3cef6dd721 cleanup: drop duplicate const usage
const char == const char const.
2014-09-24 10:54:48 +02:00
Zdenek Kabelac
f809fa5a78 cleanup: drop uneeded backup call
lv_update_and_reload already handle backup.
2014-09-24 10:54:48 +02:00
Zdenek Kabelac
ec4ffeb51c cleanup: use supplied cmd pointer 2014-09-24 10:54:48 +02:00
Zdenek Kabelac
30777fa46f cleanup: fix typo and use vg mempool
Since we are updating VG -  we can use its own mempool.
2014-09-24 10:54:47 +02:00
Zdenek Kabelac
914be0696d cleanup: replace error with print message
These are not error messages.
They are informing a user about missconfigured
options which do not change resulting error status.
2014-09-24 10:54:47 +02:00
Zdenek Kabelac
bc0a3e2355 cleanup: simplier _lv_passes_volumes_filter
Don't recreate string and just check components directly.
2014-09-24 10:54:47 +02:00
Zdenek Kabelac
a121340139 cleanup: use find_lv_in_vg_by_lvid
Replace another loop with existion function.

TODO: consider using hash maps for lvname & lvid lookups.
2014-09-24 10:54:47 +02:00
Zdenek Kabelac
a8aee7dba2 activate: update lv_check_not_in_use: API
Use of lv_info() internally in lv_check_not_in_use(),
so it always could use with_open_count properly.

Skip sysfs() testing in open_count == 0 case.

Accept just 'lv' pointer like other functions.

The function has 'built-in' lv_is_active_locally check,
which however is not what we need to check in many place.
For now at least remotely active snapshot merge is
detected and for this case merge on next activation is scheduled.
2014-09-24 10:54:47 +02:00
Zdenek Kabelac
c96665e6a8 lvconvert: reoder old snapshot merge
Move check for snapshot-merge support before archiving.

Split code on 2 paths - with merge_on_activate
using vg_write & vg_commit
and lv_update_reload call for instant merging.

Move printing after backup.
2014-09-24 10:54:47 +02:00
Zdenek Kabelac
84cdf85bd2 cleanup: constify activation usage of lv pointer
Let's enforce cheking of write access to LV by compiler.
Activation part does never need to write anything to LV
so keep LV pointer const.
2014-09-24 10:54:47 +02:00
Zdenek Kabelac
af18e7cfbf cleanup: move printing after backup
Print successful result after making backup.
In this place we are sure to not keep memlock.
2014-09-24 10:54:47 +02:00
Zdenek Kabelac
9f82a7d114 validation: move check for spare correctness
Check lv feature when we check LVs
2014-09-24 10:54:47 +02:00
Zdenek Kabelac
93fbef1a45 vgchange: use NULL 'feature' of lv_info call
When NULL is passed for info the call itself
already does the same evaluation.
2014-09-24 10:54:47 +02:00
Zdenek Kabelac
5b70bdc104 vgchange: add sync point
Before leaving _activate_lvs_in_vg() wait till devices
are active - so we do not print message about active
devices earlier then it really happens for a user.
2014-09-24 10:54:47 +02:00
Zdenek Kabelac
13c07c685a thin and cache: improve conversion validation
More validations before any thin or cache related conversion begins.

We allow to use and stack:

pool data: cache or raid
pool metadata: raid

pool:  linear, striped
cache: linear, striped, raid
thin(extorig): linear, origin, cow, virtual, thin
2014-09-24 10:53:44 +02:00