2023-03-22  Jan Friesse  <jfriesse@redhat.com>

	qdevice: Destroy non blocking client on failure
	This fixes regression introduced in patch
	8217e33e86bca79dc307e112f0c0c4a278f9a568 which made qdevice assert when
	PR_GetError in nss_sock_non_blocking_client_try_next called by
	qdevice_model_net_pre_poll_loop returns error other than
	PR_IN_PROGRESS_ERROR. This is usually not happening - only way to get
	such error seems to be to ifdown ALL network interfaces other than
	loopback and returned error is then PR_NETWORK_UNREACHABLE_ERROR
	(-5980).

	Fix is rather simple (and present in pre
	8217e33e86bca79dc307e112f0c0c4a278f9a568) - just call
	nss_sock_non_blocking_client_destroy.

2022-07-21  Jan Friesse  <jfriesse@redhat.com>

	configure: Modernize configure.ac a bit
	... to make 2.71 happy. Also increase minimum version to 2.69 (10 years
	old version so should be compatible enough).

2022-03-02  Jan Friesse  <jfriesse@redhat.com>

	unix-socket: Check minimal length of socket path

2021-04-09  liangxin1300  <XLiang@suse.com>

	man: replace votequorum_poll for actually used fn
	Actually used function is votequorum_qdevice_poll

	Reviewed-by: Jan Friesse <jfriesse@redhat.com>

2021-04-07  liangxin1300  <XLiang@suse.com>

	qnetd: sort by node_id when add new client
	Reviewed-by: Jan Friesse <jfriesse@redhat.com>

2021-03-17  Jan Friesse  <jfriesse@redhat.com>

	timer-list: Use correct english term children

	tests: Fix test-timer-list NULL check

2021-03-08  Jan Friesse  <jfriesse@redhat.com>

	tests: Enhance test-timer-list

2020-11-23  Jan Friesse  <jfriesse@redhat.com>

	qnetd: Don't alloc host_addr
	getopt will return pointer to argv so there is no need to dup optarg.

2020-11-18  Jan Friesse  <jfriesse@redhat.com>

	qnetd: Move client schedule disconnect handling
	Client disconnect used to be per client fd in
	the qnetd_client_net_socket_poll_loop_set_events_cb. Problem is, that
	disconnect calls algorithm which may send message to other client
	with fd which was already processed in the pr-poll-loop so POLLOUT is
	not set till new loop exec is called (and that usually happens
	because old one timeouts). To reproduce this problem use
	ffsplit and make qnetd disconnect one of the clients - ffsplit
	needs to send ack/nack votes, but it doesn't send them during
	first iteration and waits for dpd timeout.

	qnetd-algo-ffsplit: Simplify KAP Tie-breaker logic
	Also make it more reliable.

	qnetd: Improve dead peer detection
	Previously dead peer detection timer was scheduled every dpd_interval,
	added dpd_interval to all of the clients timestamp and if timestamp was
	larger than client hearbeat interval * 1.2 then check if client sent
	some message. If so, flag was reset.

	This method was source of number of problems so instead different method
	is now used.

	Every single client has its own timer with timeout based on
	(configurable) dpd_interval_coefficient and multiplied with
	client heartbeat timeout. When message is received from client timer is
	rescheduled. When timer callback is called (= client doesn't sent
	message during timeout) then client is disconnected.

2020-11-13  Jan Friesse  <jfriesse@redhat.com>

	qnet-config: Add space to string concat

	timer-list: Add functions for get and set interval

	timer-list: Rename delete and reschedule ops
	Add entry to the name so it is more evident change is happening to the
	entry and not to the list.

	utils: Add utils_strtod

	qdevice: Handle configurations without ring0_addr
	Configuration without ring0_addr is valid for new Corosync.

	Big thanks to Fabian-Gruenbichler who reported the problem and
	Oyvind Albrigtsen for englishify the error message.

	qdevice: Configuration without nodelist is invalid

	qdevice-cmap: Load clear node high bit only once

	qdevice-cmap: Fix clear high node bit typo
	It's really bit, not byte.

2020-11-11  Jan Friesse  <jfriesse@redhat.com>

	qdevice-net-ipc-cmd: Fix compiler warning

2020-11-11  Fabio M. Di Nitto  <fdinitto@redhat.com>

	devel: add corosync-qdevice.pc file for pcs to use
	add corosync-qdevice-devel package for future

	Reviewed-by: Jan Friesse <jfriesse@redhat.com>

2020-11-05  Jan Friesse  <jfriesse@redhat.com>

	tests: Fix assert problems

	timer-list: Improve efficiency of delete operation
	Position in entries array, heap_pos, is added to the entry.
	This has to be kept in sync for every move so new internal set/get
	functions are added too.

	This removes need for searching for entry.

2020-11-04  Jan Friesse  <jfriesse@redhat.com>

	test-timer-list: Ignore poll errors

	tlv: Check dynar_cat result

	test-process-list: Fix few bugs

	msg: Check cat result on adding msg type and size

	timer-list: Implement heap based timer-list
	Previous timer-list was naive implementation of priority queue and very
	slow when number of timers increased. This was not a problem because
	only few timers were used. But with removal of dpd timer and replacement
	with per-connection timer this may become problematic.

	Solution is to use binary heap based priority queue which is much
	faster.

	pr-poll-loop: Add queue header include
	Also add same includes to qnetd-alog-utils header file.

2020-10-27  Jan Friesse  <jfriesse@redhat.com>

	qdevice-ipc: Fix dereference bug

	qnet: Add support for keep active partition vote
	This patch adds qdevice-net part of keep active partition tie breaker
	functionality. It's enabled by default.

	When tie happens prefer partition with members of
	previously active (quorate) partition. This is hard-coded
	behavior of LMS algorithm so this setting affects only
	FFSplit algorithm. By default it is disabled for backwards
	compatibility.

	This solves problem with FFSplit when node A (with lowest id) is killed,
	node B gets vote and then node A starts up and creates single node
	membership and gets vote.

	qnetd: Fix NULL dereference of client
	Shouldn't happen but be rather safe.

	Also add more comments.

	qdevice-net-heuristics: Fix log message

	qdevice: Fix set option and set option reply
	To match the specification add heartbeat timeout only when requested.

	Also add qdevice client method to send option message.

2020-09-22  Jan Friesse  <jfriesse@redhat.com>

	LICENSE: Update copyright date

	qnetd: Add support for keep active partition vote
	When tie happens prefer partition with members of
	previously active (quorate) partition. This is hard-coded
	behavior of LMS algorithm so this setting affects only
	FFSplit algorithm. By default it is disabled for backwards
	compatibility.

	This solves problem with FFSplit when node A (with lowest id) is killed,
	node B gets vote and then node A starts up and creates single node
	membership and gets vote.

2020-09-21  Jan Friesse  <jfriesse@redhat.com>

	README: Fix typos

	timer-list: Add test

	timer-list: Return error on adding NULL callback

	qnetd: Fix dpd timer
	With default config of running dpd timer every 10 second and waiting for
	2 * client_timeout to clear message received flag and then waiting
	another 2 * client_timeout without message received it was possible that
	client was marked as a dead after more than 40 seconds making qdevice to
	stop sending votequorum hearbeat for too long so corosync lost votes
	from qdevice.

	This patch is simpler solution which just changes default dpd timer to
	1 second and timeout to 1.2 * client_timeout.

	qdevice-votequorum: Fix typo in log message

2020-08-31  Jan Friesse  <jfriesse@redhat.com>

	qdevice: Port qdevice to use pr-poll-loop

	qdevice-net: Log adds newline automatically

	qnetd: Return error code based on ipc closed
	So restore pre pr-poll-loop behavior.

	pr-poll-loop: Fix set_events_cb return code
	When events is set to 0 and set_events return -2 it was changed to -1.
	Solution is to check, if return code was 0 and only if so, change return
	code to -1 if events is 0.

2020-08-28  Jan Friesse  <jfriesse@redhat.com>

	qdevice: Fix connect heuristics result callback
	Previous patch 8dbf1bc8b0f348e468e7c70448705ac92dab68ea was wrong
	because it fixed the crash but made qdevice not work at all.

	Correct solution is to test, if state is
	QDEVICE_NET_INSTANCE_STATE_WAITING_INIT_REPLY.

	pr-poll-loop: Add pre poll callbacks

	pr-poll-loop: Pass PRPollDesc for prfd events

	pr-poll-loop: Add support for PR_POLL_EXCEPT
	Map PR_POLL_EXCEPT to POLLPRI (as NSPR does).

	qnetd: Move pr_poll_loop_exec call to function

2020-08-25  Jan Friesse  <jfriesse@redhat.com>

	qnetd: Log pr_poll_loop_add,del errors properly

	qnetd: Remove unneeded pprio include

	qnetd: Remove write callback on listening sockets
	IPC and TLS sockets are read only, so write callbacks should never
	happen (specifically tested in pr-poll-loop test) so remove them.

	qdevice: Initial port to use pr-poll-loop
	Only qdevice_instance_wait_for_initial_heuristics_exec_result is ported
	for now.

	pr-poll-loop: Return error code if PR_Poll fails

	heuristics: Remove qdevice instance pointer
	Heuristics is designed to be component of its own, which doesn't depend
	on qdevice_instance. Removing qdevice_instance pointer was easy as soon
	as exec notifier got two user data pointers.

2020-08-19  Jan Friesse  <jfriesse@redhat.com>

	tests: Enhance pr-poll-loop test

	tests: Add pr-poll-loop test

	pr-poll-loop: Do not add FD when events is empty

2020-08-12  Jan Friesse  <jfriesse@redhat.com>

	qdevice: Fix connect heuristics result callback
	Qdevice may crash if server disconnect before connect
	heuristics callback is processed.

	This is quite hard to reproduce without adding sleep before
	qdevice_net_heuristics_exec_after_connect call.

	Solution is to check if qdevice is still connected and if it doesn't,
	just throw away connect heuristics results.

	qdevice: Add space before bracket

	qdevice: Use EXIT_SUCCESS and EXIT_FAILURE codes

	qnetd: Do not call ffsplit_do on shutdown

	qnetd: Migrate main loop to pr-poll-loop

	pr-poll-loop: Add main poll loop based on PR_Poll
	Add main loop abstraction based on PR_Poll with set_events, read, write
	and error callbacks.

2020-07-15  Jan Friesse  <jfriesse@redhat.com>

	configure: Use default systemd path with prefix
	Reviewed-by: Fabio M. Di Nitto <fdinitto@redhat.com>

	build: Use git-version-gen during specfile build
	Instead of copying parts of git-version-gen for spec target use
	git-version-gen directly and parse final version into components
	(rpmver, alphatag, numcomm) and use them.

	Main reason is to simplify code a bit (sed scripts are a bit repetitive
	tho), reuse the code and also allow building of RPM from dist tarball
	generated from non-tagged commit or dirty git (not very useful).

	The code relies on fact, that hyphen is never used in tagged release
	name.

	Reviewed-by: Fabio M. Di Nitto <fdinitto@redhat.com>

	build: Update git-version-gen
	Reviewed-by: Fabio M. Di Nitto <fdinitto@redhat.com>

2020-03-13  liangxin1300  <XLiang@suse.com>

	qdevice: Change log level to NOTICE on PASS
	Previously when heuristics result became PASS it was logged as an error.
	It shouldn't be and error message so change it to LOG_NOTICE level.

	Reviewed-by: Jan Friesse <jfriesse@redhat.com>

2019-11-27  Jan Friesse  <jfriesse@redhat.com>

	qdevice: Add prototype of __vsyslog_chk

2019-09-12  Kumabuchi Kenji  <k.kumabuchi+curvygrin@gmail.com>

	init: change restart to on-failure for qdevice
	Reviewed-by: Jan Friesse <jfriesse@redhat.com>

2019-08-14  Jan Friesse  <jfriesse@redhat.com>

	qdevice: Add chk variant of vsyslog to test-log

2019-08-13  Jan Friesse  <jfriesse@redhat.com>

	qdevice: Add header files to list of test sources

	qdevice: Add log test

	qdevice: Free memory used by log

	tests: Add utils_parse_bool_str test

2019-08-12  Jan Friesse  <jfriesse@redhat.com>

	qdevice: Configurable log priority bump

	qnetd: Use log-common for nodelist debug dump

	qdevice: Merge msg_decode_error functions

	qdevice: Import log instead of qdevice-log
	and move qdevice-log-debug to log-common to share it with qnetd.

	qdevice: Use log instead of libqb log

	qnetd: Add function to set log target

	qnetd: Check log initialization error

	qnetd: Fix double -d description

	qnetd: Rename qnetd-log.c to log.c
	And remove qnetd_ prefix from function names.

2019-06-10  Jan Friesse  <jfriesse@redhat.com>

	qdevice: Fix qdevice_heuristics_destroy error
	When daemonization is used, heuristics worker is not a child of
	the corosync-qdevice process any longer so it's not possible
	to wait for its exit.

	configure: Add user-flags configure option
	configure.ac adds by optimization and debug flags by default. This is
	not always intended so conifgure option is added.

	Support for using this option in specfile is added too.

	Also fix GDB_CFLAGS typo. GDB_FLAGS (without C)
	is the correct name of variable to print in the summary.

2019-04-25  Jan Friesse  <jfriesse@redhat.com>

	man: Fix corosync-qdevice default connect_timeout

2019-03-20  Jan Friesse  <jfriesse@redhat.com>

	spec: Use install -p and add license

2019-03-19  Jan Friesse  <jfriesse@redhat.com>

	qnetd: Check existence of NSS DB dir before fork
	Previously, when user tried start corosync-qnetd without
	initialized NSS database then generic (not very helpful
	and misleading) NSS error was logged
	"NSS error (-8015): The certificate/key database is in an old,
	unsupported format.".

	Solution is to check if it's possible to open NSS DB directory and
	display (usually much more informative) result of strerror function.

	Such check is called before fork, so init system can return error code
	during start.

	Reviewed-by: Christine Caulfield <ccaulfie@redhat.com>

2018-11-26  Jan Friesse  <jfriesse@redhat.com>

	certutils: Fix warnings found by coverity

2018-11-15  Jan Friesse  <jfriesse@redhat.com>

	init: Fix init scripts to work with containers
	Previously init scripts were not using pid file so pidof was used. This
	is usually not a problem, but when containers are used it may result to
	killing improper instance when issued on host.

	Solution is to always use pidfile.

	Also try to use LSB complaint status codes.

	Reviewed-by: Christine Caulfield <ccaulfie@redhat.com>

2018-11-07  Jan Friesse  <jfriesse@redhat.com>

	configure: move to AC_COMPILE_IFELSE
	from AC_PREPROC_IFELSE which is strongly discouraged.

	Our detection system was very weak and recent versions of clang did
	show that PREPROC_IFELFE (cpp) would enable warning options that
	the compiler does not support (clang).

	Use a full compilation test to detect what works and what doesn't.

	Based on knet patch 88491f27375a9e8aceb946853a1abf4d23ebb8f3.

	Reviewed-by: Fabio M. Di Nitto <fdinitto@redhat.com>

2018-09-03  Jan Friesse  <jfriesse@redhat.com>

	git-version-gen: Fail on UNKNOWN version
	Reviewed-by: Christine Caulfield <ccaulfie@redhat.com>

	build: Support for git archive stored tags
	Attempt to solve problem with git archive generated tarballs
	(used for example by github when release is downloaded) which are no
	longer git tree and (in contrast to officially released tarballs) also
	doesn't contain .tarball-version file so git-version-gen script simply
	cannot obtain valid version info.

	Solution is based on using gitattributes which is instructs git to
	replace string in the .gitarchivever file by known ref names.
	git-version-gen is enhanced to support this file and tries to parse
	any string which looks like "tag: v[0-9]+.[0-9]+.[0-9]". If such string
	is found it's used as a version. This file is used as a last attempt and
	other methods (.tarball-version, git abbrev) have precedence.

	Based on idea stated by Jan Pokorný <jpokorny@redhat.com>.

	Reviewed-by: Christine Caulfield <ccaulfie@redhat.com>

2018-08-27  Jan Friesse  <jfriesse@redhat.com>

	qdevice: Propagate error to exit code
	Net model never returned error when qdevice_model_run was called. This
	was incorrect because with exception of local ipc close all other
	disconnect reasons are errors.

	Solution is to return proper error code.

	Also instead of exit right after qdevice_model_run it's better to store
	result value, try clean resources and use stored value to return correct
	exit code.

2018-08-22  Jan Friesse  <jfriesse@redhat.com>

	build: Remove WITH_LIST
	WITH_LIST is replaced in favor of bcond generator.

	spec: Autogenerate bcond options based on config
	Inspired by kronosnet spec file generator.

2018-08-21  Jan Friesse  <jfriesse@redhat.com>

	spec: add exlicit docdir
	OpenSUSE does not ship docs in the normal dir and their rpm macro
	does not appear to set it for us.

	Based on corosync 30c7f3319f003efa1abeb2173eef9e5c7943ae5b.

	spec: Use different BuildRequires for SUSE

	init: disable stderr output in systemd unit file
	Usually both syslog and stderr are enabled so log entries are duplicated
	in journal. Solution is to use similar patch as corosync
	c34208ad402b45f52b5d3ee8d2a08df0779ec9aa and disable stderr.

	tests: Enlarge timeout for process-list test

2018-08-09  Jan Friesse  <jfriesse@redhat.com>

	spec: Add explicit gcc build dependency

	config: Remove unneeded qb includes

	qnetd-certutil: Add -G option

	man: Document NSS database conversion
	This is not needed at least for cert8 -> cert9, but it's still nice to
	have it documented. Also document NSS_IGNORE_SYSTEM_POLICY=1 workaround.

	qdevice-net-certutil: Implement scp wrapper
	Standard scp doesn't handle copy of file from remote machine to remote
	machine very well when agent forwarding is used and no key exists
	between the machines.

	certutil: Add support for new NSS database format
	New NSS database uses files cert9.db, key4.db and pkcs11.txt. Add
	support for these files into cert utilities.

	gitignore: Add testsuite results

2018-08-08  Jan Friesse  <jfriesse@redhat.com>

	qdevice-cmap: Fix compiler warning on Sparc/ARM

	tests: Notify process-list after trap was created
	Previously there was no waiting for establish of bash trap.
	This may result in situation when sigint/term was send before trap
	making test fail.

	Solution is to catch the sigusr1 and sigusr2 signals in tests and wait
	till these signals are sent by bash after trap is established.

2018-05-02  Jan Friesse  <jfriesse@redhat.com>

	unix-socket: Fix snprintf warning
	Base N on sizeof sun_path instead of strlen of input path.

2018-04-27  Jan Friesse  <jfriesse@redhat.com>

	tests: enlarge process-list timeouts

	tests: process-list add stdlib include

	tests: Fix process-list test to work on FreeBSD
	FreeBSD true and false commands are installed in the /usr/bin prefix
	instead of Linux /bin.

2018-04-20  Ferenc Wágner  <wferi@debian.org>

	Fix typo: sucesfully -> successfully
	Reviewed-by: Jan Friesse <jfriesse@redhat.com>

2018-04-19  Bin Liu  <bliu@suse.com>

	qdevice: optarg should be str in init_from_cmap
	Reviewed-by: Jan Friesse <jfriesse@redhat.com>

2018-04-12  Jan Friesse  <jfriesse@redhat.com>

	spec: Remove unused clean section

2018-04-03  Jan Friesse  <jfriesse@redhat.com>

	msgio: Remove unused code
	Blocking version of send was used in the very early phases of qdevice
	development and it's not used any longer, so it make sense to delete
	it.

2018-03-23  Jan Friesse  <jfriesse@redhat.com>

	msgio: Fix reading of msg longer than PR_INT32_MAX

2018-03-02  Jan Friesse  <jfriesse@redhat.com>

	qnetd: Replace strtol by strtonum

	qdevice: Replace strtol by strtonum

	Add safer wrapper of strtoll

	Check if user_data can be dereferenced

	Nodelist is set into string not array

	Fix NULL pointer dereference

	Quote certutils scripts properly
	Also remove unused variables.

	init: Quote subshell result properly

2018-01-25  Jan Friesse  <jfriesse@redhat.com>

	spec: Modernize spec file a bit
	Thanks to Robert-André Mauchin <zebob.m@gmail.com> for spec file review
	and comments.

2018-01-23  Jan Friesse  <jfriesse@redhat.com>

	Initial import from corosync codebase
	Used the code from corosync master
	(31ddba64a2726bcedf81eb84df2e2da4846832f7)