Fix: resources: typo in HealthSMART meta-data

Build: resources: add a make target for validating agent meta-data

Low: resources: remove no-longer-relevant DTD reference from meta-data

Low: resources: add SystemHealth parameter to its meta-data

Merge branch '2.1' of https://github.com/ClusterLabs/pacemaker into fixes

Fix: fencing: delegate shouldn't matter when checking equivalent fencing

Low: Fix print spec for unsigned int count

Fix: fencer: avoid memory leaks when processing topology requests

Fix: fencer: get target-by-attribute working again

Low: CTS: Fix a bug in an awk command used in ctslab.

Low: libcrmcommon: fix a wrong separator in the .warning CSS declaration

Low: NLS: update po files for 2.1.4 release

Fix: fencer: avoid use-after-free with self-fencing and topology

Build: update version information for 2.1.4

Build: libraries: update shared object versions for 2.1.4

Doc: ChangeLog: update for 2.1.4-rc1 release

Doc: ChangeLog: update for 2.1.3 release

Refactor: libcrmcommon: add crm_str() equivalent with a default string

Log: all: use pcmk__s() instead of crm_str()

Test: cli: update expected output

Doc: Pacemaker Development: replace crm_str() reference

API: libcrmcommon: deprecate crm_str()

Refactor: libstonithd: use pcmk__s() to simplify variable initializations

Refactor: controller: drop redundant condition

Fix: fencer: avoid memory leaks when processing topology requests

Refactor: fencer: split handle_topology_change() into two functions

Low: fencer: always add device ID to device addition notification

Low: fencer: always add device ID to level addition notification

Low: fencer: always add device ID to level removal notification

Low: fencing: always add notification type to client notification data

API: libstonithd: deprecate stonith_event_t:type

API: libstonithd: deprecate stonith_event_t:message

Low: fencing: add device/level description to configuration change notifications

Low: fencing: make stonith__event_description() handle all notification types

Log: cts-fence-helper: use stonith__event_description() to log notifications

Doc: controller: drop incorrect comment

Refactor: fencing: avoid duplicate variable for notification type

Low: fencer: avoid redundant result in XML

Low: libstonithd: use case-sensitive comparisons where appropriate

Doc: libstonithd: clarify a comment

Test: cts: install new cts-cli data files

Low: libcrmcommon: avoid memory leak if IPC server fails to run

Low: cts-lab: fix variable syntax in messages

Doc: ChangeLog: update for 2.1.3-rc2 release

Fix: schemas: Use fence-event-2.15.rng

Test: cts-fencing: update expected total timeouts

Low: schemas: Copy API schemas in preparation for fix

Low: scheduler: don't support versioned instance attributes in operations

Low: schemas: allow Promoted/Unpromoted in crm_resource output

Low: schemas: copy crm_resource API schema in preparation for changes

API: libstonithd: deprecate stonith_t:call_timeout

Refactor: fencing: make "stonith-event" message take int instead of gboolean

Refactor: fencing: pass show options to "stonith-event" messages

Refactor: fencing: functionize creating a history entry description

Feature: tools: respect crm_mon --show-detail for fencing actions

Refactor: fencing: pass show options to stonith__history_description()

Log: fencer: log device only if not null

Log: fencer: log raw XML at trace level

Log: fencer: lower eligibility logs to info level

Log: fencer: downgrade topology failure message to info level

Log: controller: downgrade successful fencing logs to info

Fix: fencing: delegate shouldn't matter when checking equivalent fencing

Feature: fencer: deprecate stand-alone mode

Log: executor: improve messages after fencing connection loss

Low: tools: stonith_admin should display device/level errors

Test: cts-lab: drop redundant pattern

Low: fencing: prefer same delegate when checking equivalent fencing

Log: fencer: make completion logs more consistent

Low: fencing: show delegate for later successful fencing actions

Low: tools: add more detail to stonith_admin error messages

Low: libcib: Check for NULL before calling g_ascii_strdown.

Log: scheduler: make some trace logs a little more useful

Fix: scheduler: make multiple-active="stop_unexpected" actually work

Test: scheduler: add a second regression test for multiple…

Revert "Refactor: scheduler: add expected node to primitive variant data"

Doc: ChangeLog: update for 2.1.3-rc1 release

Build: libraries: update shared object versions for 2.1.3

Build: update version information for 2.1.3

Doc: xml: update README for 2.1.3

Test: scheduler: update whitespace in output

Doc: scheduler: update function name in comments

Test: scheduler: avoid deprecated node type in regression test

Doc: configure: document when DBusBasicValue became available

API: libpe_status: deprecate node_ping

API: libpe_status: deprecate pe_rsc_starting and pe_rsc_stopping

Refactor: libcrmcommon: ensure proper includes are available

Fix: fencer: avoid core dump at exit

Low: fencer: don't use constant within translatable string literal

Fix: libcrmcommon: fix NLS issues with option meta-data

Low: NLS: alphabetize POTFILES

Low: NLS: update translation files for current code base

Refactor: libraries: move health strategy validation function

API: libpe_status: add pe_node_shared_s:data_set

Test: libcrmcommon: add unit test for pcmk__validate_health_strategy()