import pacemaker-2.1.5-8.el8

.gitignore

@@ -0,0 +1,2 @@
+SOURCES/nagios-agents-metadata-105ab8a.tar.gz
+SOURCES/pacemaker-a3f4479.tar.gz

.pacemaker.metadata

@@ -0,0 +1,2 @@
+ea6c0a27fd0ae8ce02f84a11f08a0d79377041c3 SOURCES/nagios-agents-metadata-105ab8a.tar.gz
+883efa27f94c6a07942f51cf7c8959c5fbb624fe SOURCES/pacemaker-a3f4479.tar.gz

SOURCES/002-remote-regression.patch

@@ -0,0 +1,98 @@
+From d8e08729ad5e3dc62f774172f992210902fc0ed4 Mon Sep 17 00:00:00 2001
+From: Ken Gaillot <kgaillot@redhat.com>
+Date: Mon, 23 Jan 2023 14:25:56 -0600
+Subject: [PATCH] High: executor: fix regression in remote node shutdown
+
+This reverts the essential part of d61494347, which was based on misdiagnosing
+a remote node shutdown issue. Initially, it was thought that a "TLS server
+session ended" log just after a remote node requested shutdown indicated that
+the proxy connection coincidentally dropped at that moment. It actually is the
+routine stopping of accepting new proxy connections, and existing when that
+happens makes the remote node exit immediately without waiting for the
+all-clear from the cluster.
+
+Fixes T361
+---
+ daemons/execd/pacemaker-execd.c | 19 +------------------
+ daemons/execd/pacemaker-execd.h |  3 +--
+ daemons/execd/remoted_tls.c     |  6 +-----
+ 3 files changed, 3 insertions(+), 25 deletions(-)
+
+diff --git a/daemons/execd/pacemaker-execd.c b/daemons/execd/pacemaker-execd.c
+index db12674f13..491808974a 100644
+--- a/daemons/execd/pacemaker-execd.c
++++ b/daemons/execd/pacemaker-execd.c
+@@ -1,5 +1,5 @@
+ /*
+- * Copyright 2012-2022 the Pacemaker project contributors
++ * Copyright 2012-2023 the Pacemaker project contributors
+  *
+  * The version control history for this file may have further details.
+  *
+@@ -305,23 +305,6 @@ lrmd_exit(gpointer data)
+     return FALSE;
+ }
+ 
+-/*!
+- * \internal
+- * \brief Clean up and exit if shutdown has started
+- *
+- * \return Doesn't return
+- */
+-void
+-execd_exit_if_shutting_down(void)
+-{
+-#ifdef PCMK__COMPILE_REMOTE
+-    if (shutting_down) {
+-        crm_warn("exit because TLS connection was closed and 'shutting_down' set");
+-        lrmd_exit(NULL);
+-    }
+-#endif
+-}
+-
+ /*!
+  * \internal
+  * \brief Request cluster shutdown if appropriate, otherwise exit immediately
+diff --git a/daemons/execd/pacemaker-execd.h b/daemons/execd/pacemaker-execd.h
+index 6646ae29e3..f78e8dcdde 100644
+--- a/daemons/execd/pacemaker-execd.h
++++ b/daemons/execd/pacemaker-execd.h
+@@ -1,5 +1,5 @@
+ /*
+- * Copyright 2012-2022 the Pacemaker project contributors
++ * Copyright 2012-2023 the Pacemaker project contributors
+  *
+  * The version control history for this file may have further details.
+  *
+@@ -105,6 +105,5 @@ void remoted_spawn_pidone(int argc, char **argv, char **envp);
+ int process_lrmd_alert_exec(pcmk__client_t *client, uint32_t id,
+                             xmlNode *request);
+ void lrmd_drain_alerts(GMainLoop *mloop);
+-void execd_exit_if_shutting_down(void);
+ 
+ #endif // PACEMAKER_EXECD__H
+diff --git a/daemons/execd/remoted_tls.c b/daemons/execd/remoted_tls.c
+index 6f4b2d0062..c65e3f394d 100644
+--- a/daemons/execd/remoted_tls.c
++++ b/daemons/execd/remoted_tls.c
+@@ -1,5 +1,5 @@
+ /*
+- * Copyright 2012-2022 the Pacemaker project contributors
++ * Copyright 2012-2023 the Pacemaker project contributors
+  *
+  * The version control history for this file may have further details.
+  *
+@@ -250,10 +250,6 @@ static void
+ tls_server_dropped(gpointer user_data)
+ {
+     crm_notice("TLS server session ended");
+-    /* If we are in the process of shutting down, then we should actually exit.
+-     * bz#1804259
+-     */
+-    execd_exit_if_shutting_down();
+     return;
+ }
+ 
+-- 
+2.31.1
+

SOURCES/004-g_source_remove.patch

@@ -0,0 +1,107 @@
+From 45617b727e280cac384a28ae3d96145e066e6197 Mon Sep 17 00:00:00 2001
+From: Reid Wahl <nrwahl@protonmail.com>
+Date: Fri, 3 Feb 2023 12:08:57 -0800
+Subject: [PATCH 01/02] Fix: fencer: Prevent double g_source_remove of op_timer_one
+
+QE observed a rarely reproducible core dump in the fencer during
+Pacemaker shutdown, in which we try to g_source_remove() an op timer
+that's already been removed.
+
+free_stonith_remote_op_list()
+-> g_hash_table_destroy()
+-> g_hash_table_remove_all_nodes()
+-> clear_remote_op_timers()
+-> g_source_remove()
+-> crm_glib_handler()
+-> "Source ID 190 was not found when attempting to remove it"
+
+The likely cause is that request_peer_fencing() doesn't set
+op->op_timer_one to 0 after calling g_source_remove() on it, so if that
+op is still in the stonith_remote_op_list at shutdown with the same
+timer, clear_remote_op_timers() tries to remove the source for
+op_timer_one again.
+
+There are only five locations that call g_source_remove() on a
+remote_fencing_op_t timer.
+* Three of them are in clear_remote_op_timers(), which first 0-checks
+  the timer and then sets it to 0 after g_source_remove().
+* One is in remote_op_query_timeout(), which does the same.
+* The last is the one we fix here in request_peer_fencing().
+
+I don't know all the conditions of QE's test scenario at this point.
+What I do know:
+* have-watchdog=true
+* stonith-watchdog-timeout=10
+* no explicit topology
+* fence agent script is missing for the configured fence device
+* requested fencing of one node
+* cluster shutdown
+
+Fixes RHBZ2166967
+
+Signed-off-by: Reid Wahl <nrwahl@protonmail.com>
+---
+ daemons/fenced/fenced_remote.c | 1 +
+ 1 file changed, 1 insertion(+)
+
+diff --git a/daemons/fenced/fenced_remote.c b/daemons/fenced/fenced_remote.c
+index d61b5bd..b7426ff 100644
+--- a/daemons/fenced/fenced_remote.c
++++ b/daemons/fenced/fenced_remote.c
+@@ -1825,6 +1825,7 @@ request_peer_fencing(remote_fencing_op_t *op, peer_device_info_t *peer)
+         op->state = st_exec;
+         if (op->op_timer_one) {
+             g_source_remove(op->op_timer_one);
++            op->op_timer_one = 0;
+         }
+ 
+         if (!((stonith_watchdog_timeout_ms > 0)
+-- 
+2.31.1
+
+From 0291db4750322ec7f01ae6a4a2a30abca9d8e19e Mon Sep 17 00:00:00 2001
+From: Reid Wahl <nrwahl@protonmail.com>
+Date: Wed, 15 Feb 2023 22:30:27 -0800
+Subject: [PATCH 02/02] Fix: fencer: Avoid double source remove of op_timer_total
+
+remote_op_timeout() returns G_SOURCE_REMOVE, which tells GLib to remove
+the source from the main loop after returning. Currently this function
+is used as the callback only when creating op->op_timer_total.
+
+If we don't set op->op_timer_total to 0 before returning from
+remote_op_timeout(), then we can get an assertion and core dump from
+GLib when the op's timers are being cleared (either during op
+finalization or during fencer shutdown). This is because
+clear_remote_op_timers() sees that op->op_timer_total != 0 and tries to
+remove the source, but the source has already been removed.
+
+Note that we're already (correctly) zeroing op->op_timer_one and
+op->query_timeout as appropriate in their respective callback functions.
+
+Fortunately, GLib doesn't care whether the source has already been
+removed before we return G_SOURCE_REMOVE from a callback. So it's safe
+to call finalize_op() (which removes all the op's timer sources) from
+within a callback.
+
+Fixes RHBZ#2166967
+
+Signed-off-by: Reid Wahl <nrwahl@protonmail.com>
+---
+ daemons/fenced/fenced_remote.c | 2 ++
+ 1 file changed, 2 insertions(+)
+
+diff --git a/daemons/fenced/fenced_remote.c b/daemons/fenced/fenced_remote.c
+index b7426ff88..adea3d7d8 100644
+--- a/daemons/fenced/fenced_remote.c
++++ b/daemons/fenced/fenced_remote.c
+@@ -718,6 +718,8 @@ remote_op_timeout(gpointer userdata)
+ {
+     remote_fencing_op_t *op = userdata;
+ 
++    op->op_timer_total = 0;
++
+     if (op->state == st_done) {
+         crm_debug("Action '%s' targeting %s for client %s already completed "
+                   CRM_XS " id=%.8s",
+-- 
+2.39.0

SOURCES/005-query-null.patch

@@ -0,0 +1,151 @@
+From 0d15568a538349ac41028db6b506d13dd23e8732 Mon Sep 17 00:00:00 2001
+From: Chris Lumens <clumens@redhat.com>
+Date: Tue, 14 Feb 2023 14:00:37 -0500
+Subject: [PATCH] High: libcrmcommon: Fix handling node=NULL in
+ pcmk__attrd_api_query.
+
+According to the header file, if node is NULL, pcmk__attrd_api_query
+should query the value of the given attribute on all cluster nodes.
+This is also what the server expects and how attrd_updater is supposed
+to work.
+
+However, pcmk__attrd_api_query has no way of letting callers decide
+whether they want to query all nodes or whether they want to use the
+local node.  We were passing NULL for the node name, which it took to
+mean it should look up the local node name.  This calls
+pcmk__node_attr_target, which probes the local cluster name and returns
+that to pcmk__attrd_api_query.  If it returns non-NULL, that value will
+then be put into the XML IPC call which means the server will only
+return the value for that node.
+
+In testing this was usually fine.  However, in pratice, the methods
+pcmk__node_attr_target uses to figure out the local cluster node name
+involves checking the OCF_RESKEY_CRM_meta_on_node environment variable
+among others.
+
+This variable was never set in testing, but can be set in the real
+world.  This leads to circumstances where the user did "attrd_updater -QA"
+expecting to get the values on all nodes, but instead only got the value
+on the local cluster node.
+
+In pacemaker-2.1.4 and prior, pcmk__node_attr_target was simply never
+called if the node was NULL but was called otherwise.
+
+The fix is to modify pcmk__attrd_api_query to take an option for
+querying all nodes.  If that's present, we'll query all nodes.  If it's
+not present, we'll look at the given node name - NULL means look it up,
+anything else means just that node.
+
+Regression in 2.1.5 introduced by eb20a65577
+---
+ include/crm/common/attrd_internal.h     |  6 +++++-
+ include/crm/common/ipc_attrd_internal.h |  7 +++++--
+ lib/common/ipc_attrd.c                  | 12 ++++++++----
+ tools/attrd_updater.c                   |  5 +++--
+ 4 files changed, 21 insertions(+), 9 deletions(-)
+
+diff --git a/include/crm/common/attrd_internal.h b/include/crm/common/attrd_internal.h
+index 389be48..7337c38 100644
+--- a/include/crm/common/attrd_internal.h
++++ b/include/crm/common/attrd_internal.h
+@@ -1,5 +1,5 @@
+ /*
+- * Copyright 2004-2022 the Pacemaker project contributors
++ * Copyright 2004-2023 the Pacemaker project contributors
+  *
+  * The version control history for this file may have further details.
+  *
+@@ -25,6 +25,10 @@ enum pcmk__node_attr_opts {
+     pcmk__node_attr_perm           = (1 << 5),
+     pcmk__node_attr_sync_local     = (1 << 6),
+     pcmk__node_attr_sync_cluster   = (1 << 7),
++    // pcmk__node_attr_utilization is 8, but that has not been backported.
++    // I'm leaving the gap here in case we backport that in the future and
++    // also to avoid problems on mixed-version clusters.
++    pcmk__node_attr_query_all      = (1 << 9),
+ };
+ 
+ #define pcmk__set_node_attr_flags(node_attr_flags, flags_to_set) do {   \
+diff --git a/include/crm/common/ipc_attrd_internal.h b/include/crm/common/ipc_attrd_internal.h
+index 2c6713f..b1b7584 100644
+--- a/include/crm/common/ipc_attrd_internal.h
++++ b/include/crm/common/ipc_attrd_internal.h
+@@ -1,5 +1,5 @@
+ /*
+- * Copyright 2022 the Pacemaker project contributors
++ * Copyright 2022-2023 the Pacemaker project contributors
+  *
+  * The version control history for this file may have further details.
+  *
+@@ -110,10 +110,13 @@ int pcmk__attrd_api_purge(pcmk_ipc_api_t *api, const char *node);
+  *
+  * \param[in,out] api           Connection to pacemaker-attrd
+  * \param[in]     node          Look up the attribute for this node
+- *                              (or NULL for all nodes)
++ *                              (or NULL for the local node)
+  * \param[in]     name          Attribute name
+  * \param[in]     options       Bitmask of pcmk__node_attr_opts
+  *
++ * \note Passing pcmk__node_attr_query_all will cause the function to query
++ *       the value of \p name on all nodes, regardless of the value of \p node.
++ *
+  * \return Standard Pacemaker return code
+  */
+ int pcmk__attrd_api_query(pcmk_ipc_api_t *api, const char *node, const char *name,
+diff --git a/lib/common/ipc_attrd.c b/lib/common/ipc_attrd.c
+index 4606509..dece49b 100644
+--- a/lib/common/ipc_attrd.c
++++ b/lib/common/ipc_attrd.c
+@@ -1,5 +1,5 @@
+ /*
+- * Copyright 2011-2022 the Pacemaker project contributors
++ * Copyright 2011-2023 the Pacemaker project contributors
+  *
+  * The version control history for this file may have further details.
+  *
+@@ -332,10 +332,14 @@ pcmk__attrd_api_query(pcmk_ipc_api_t *api, const char *node, const char *name,
+         return EINVAL;
+     }
+ 
+-    target = pcmk__node_attr_target(node);
++    if (pcmk_is_set(options, pcmk__node_attr_query_all)) {
++        node = NULL;
++    } else {
++        target = pcmk__node_attr_target(node);
+ 
+-    if (target != NULL) {
+-        node = target;
++        if (target != NULL) {
++            node = target;
++        }
+     }
+ 
+     request = create_attrd_op(NULL);
+diff --git a/tools/attrd_updater.c b/tools/attrd_updater.c
+index 3cd766d..cbd341d 100644
+--- a/tools/attrd_updater.c
++++ b/tools/attrd_updater.c
+@@ -376,6 +376,7 @@ attrd_event_cb(pcmk_ipc_api_t *attrd_api, enum pcmk_ipc_event event_type,
+ static int
+ send_attrd_query(pcmk__output_t *out, const char *attr_name, const char *attr_node, gboolean query_all)
+ {
++    uint32_t options = pcmk__node_attr_none;
+     pcmk_ipc_api_t *attrd_api = NULL;
+     int rc = pcmk_rc_ok;
+ 
+@@ -400,10 +401,10 @@ send_attrd_query(pcmk__output_t *out, const char *attr_name, const char *attr_no
+ 
+     /* Decide which node(s) to query */
+     if (query_all == TRUE) {
+-        attr_node = NULL;
++        options |= pcmk__node_attr_query_all;
+     }
+ 
+-    rc = pcmk__attrd_api_query(attrd_api, attr_node, attr_name, 0);
++    rc = pcmk__attrd_api_query(attrd_api, attr_node, attr_name, options);
+ 
+     if (rc != pcmk_rc_ok) {
+         g_set_error(&error, PCMK__RC_ERROR, rc, "Could not query value of %s: %s (%d)",
+-- 
+2.31.1
+