1 files changed, 258 insertions, 442 deletions

diff --git a/arch/powerpc/platforms/pseries/hotplug-cpu.c b/arch/powerpc/platforms/pseries/hotplug-cpu.c
index 3e8cbfe7a80f..e0a7ac5db15d 100644
--- a/arch/powerpc/platforms/pseries/hotplug-cpu.c
+++ b/arch/powerpc/platforms/pseries/hotplug-cpu.c
@@ -35,53 +35,15 @@
 #include <asm/topology.h>
 
 #include "pseries.h"
-#include "offline_states.h"
 
 /* This version can't take the spinlock, because it never returns */
 static int rtas_stop_self_token = RTAS_UNKNOWN_SERVICE;
 
-static DEFINE_PER_CPU(enum cpu_state_vals, preferred_offline_state) =
-							CPU_STATE_OFFLINE;
-static DEFINE_PER_CPU(enum cpu_state_vals, current_state) = CPU_STATE_OFFLINE;
-
-static enum cpu_state_vals default_offline_state = CPU_STATE_OFFLINE;
-
-static bool cede_offline_enabled __read_mostly = true;
-
 /*
- * Enable/disable cede_offline when available.
+ * Record the CPU ids used on each nodes.
+ * Protected by cpu_add_remove_lock.
  */
-static int __init setup_cede_offline(char *str)
-{
-	return (kstrtobool(str, &cede_offline_enabled) == 0);
-}
-
-__setup("cede_offline=", setup_cede_offline);
-
-enum cpu_state_vals get_cpu_current_state(int cpu)
-{
-	return per_cpu(current_state, cpu);
-}
-
-void set_cpu_current_state(int cpu, enum cpu_state_vals state)
-{
-	per_cpu(current_state, cpu) = state;
-}
-
-enum cpu_state_vals get_preferred_offline_state(int cpu)
-{
-	return per_cpu(preferred_offline_state, cpu);
-}
-
-void set_preferred_offline_state(int cpu, enum cpu_state_vals state)
-{
-	per_cpu(preferred_offline_state, cpu) = state;
-}
-
-void set_default_offline_state(int cpu)
-{
-	per_cpu(preferred_offline_state, cpu) = default_offline_state;
-}
+static cpumask_var_t node_recorded_ids_map[MAX_NUMNODES];
 
 static void rtas_stop_self(void)
 {
@@ -91,19 +53,14 @@ static void rtas_stop_self(void)
 
 	BUG_ON(rtas_stop_self_token == RTAS_UNKNOWN_SERVICE);
 
-	printk("cpu %u (hwid %u) Ready to die...\n",
-	       smp_processor_id(), hard_smp_processor_id());
-
 	rtas_call_unlocked(&args, rtas_stop_self_token, 0, 1, NULL);
 
 	panic("Alas, I survived.\n");
 }
 
-static void pseries_mach_cpu_die(void)
+static void pseries_cpu_offline_self(void)
 {
-	unsigned int cpu = smp_processor_id();
 	unsigned int hwcpu = hard_smp_processor_id();
-	u8 cede_latency_hint = 0;
 
 	local_irq_disable();
 	idle_task_exit();
@@ -112,49 +69,6 @@ static void pseries_mach_cpu_die(void)
 	else
 		xics_teardown_cpu();
 
-	if (get_preferred_offline_state(cpu) == CPU_STATE_INACTIVE) {
-		set_cpu_current_state(cpu, CPU_STATE_INACTIVE);
-		if (ppc_md.suspend_disable_cpu)
-			ppc_md.suspend_disable_cpu();
-
-		cede_latency_hint = 2;
-
-		get_lppaca()->idle = 1;
-		if (!lppaca_shared_proc(get_lppaca()))
-			get_lppaca()->donate_dedicated_cpu = 1;
-
-		while (get_preferred_offline_state(cpu) == CPU_STATE_INACTIVE) {
-			while (!prep_irq_for_idle()) {
-				local_irq_enable();
-				local_irq_disable();
-			}
-
-			extended_cede_processor(cede_latency_hint);
-		}
-
-		local_irq_disable();
-
-		if (!lppaca_shared_proc(get_lppaca()))
-			get_lppaca()->donate_dedicated_cpu = 0;
-		get_lppaca()->idle = 0;
-
-		if (get_preferred_offline_state(cpu) == CPU_STATE_ONLINE) {
-			unregister_slb_shadow(hwcpu);
-
-			hard_irq_disable();
-			/*
-			 * Call to start_secondary_resume() will not return.
-			 * Kernel stack will be reset and start_secondary()
-			 * will be called to continue the online operation.
-			 */
-			start_secondary_resume();
-		}
-	}
-
-	/* Requested state is CPU_STATE_OFFLINE at this point */
-	WARN_ON(get_preferred_offline_state(cpu) != CPU_STATE_OFFLINE);
-
-	set_cpu_current_state(cpu, CPU_STATE_OFFLINE);
 	unregister_slb_shadow(hwcpu);
 	rtas_stop_self();
 
@@ -179,6 +93,9 @@ static int pseries_cpu_disable(void)
 		xive_smp_disable_cpu();
 	else
 		xics_migrate_irqs_away();
+
+	cleanup_cpu_mmu_context();
+
 	return 0;
 }
 
@@ -191,114 +108,180 @@ static int pseries_cpu_disable(void)
  * to self-destroy so that the cpu-offline thread can send the CPU_DEAD
  * notifications.
  *
- * OTOH, pseries_mach_cpu_die() is called by the @cpu when it wants to
+ * OTOH, pseries_cpu_offline_self() is called by the @cpu when it wants to
  * self-destruct.
  */
 static void pseries_cpu_die(unsigned int cpu)
 {
-	int tries;
 	int cpu_status = 1;
 	unsigned int pcpu = get_hard_smp_processor_id(cpu);
+	unsigned long timeout = jiffies + msecs_to_jiffies(120000);
 
-	if (get_preferred_offline_state(cpu) == CPU_STATE_INACTIVE) {
-		cpu_status = 1;
-		for (tries = 0; tries < 5000; tries++) {
-			if (get_cpu_current_state(cpu) == CPU_STATE_INACTIVE) {
-				cpu_status = 0;
-				break;
-			}
-			msleep(1);
-		}
-	} else if (get_preferred_offline_state(cpu) == CPU_STATE_OFFLINE) {
+	while (true) {
+		cpu_status = smp_query_cpu_stopped(pcpu);
+		if (cpu_status == QCSS_STOPPED ||
+		    cpu_status == QCSS_HARDWARE_ERROR)
+			break;
 
-		for (tries = 0; tries < 25; tries++) {
-			cpu_status = smp_query_cpu_stopped(pcpu);
-			if (cpu_status == QCSS_STOPPED ||
-			    cpu_status == QCSS_HARDWARE_ERROR)
-				break;
-			cpu_relax();
+		if (time_after(jiffies, timeout)) {
+			pr_warn("CPU %i (hwid %i) didn't die after 120 seconds\n",
+				cpu, pcpu);
+			timeout = jiffies + msecs_to_jiffies(120000);
 		}
+
+		cond_resched();
 	}
 
-	if (cpu_status != 0) {
-		printk("Querying DEAD? cpu %i (%i) shows %i\n",
-		       cpu, pcpu, cpu_status);
+	if (cpu_status == QCSS_HARDWARE_ERROR) {
+		pr_warn("CPU %i (hwid %i) reported error while dying\n",
+			cpu, pcpu);
 	}
 
-	/* Isolation and deallocation are definitely done by
-	 * drslot_chrp_cpu.  If they were not they would be
-	 * done here.  Change isolate state to Isolate and
-	 * change allocation-state to Unusable.
-	 */
 	paca_ptrs[cpu]->cpu_start = 0;
 }
 
+/**
+ * find_cpu_id_range - found a linear ranger of @nthreads free CPU ids.
+ * @nthreads : the number of threads (cpu ids)
+ * @assigned_node : the node it belongs to or NUMA_NO_NODE if free ids from any
+ *                  node can be peek.
+ * @cpu_mask: the returned CPU mask.
+ *
+ * Returns 0 on success.
+ */
+static int find_cpu_id_range(unsigned int nthreads, int assigned_node,
+			     cpumask_var_t *cpu_mask)
+{
+	cpumask_var_t candidate_mask;
+	unsigned int cpu, node;
+	int rc = -ENOSPC;
+
+	if (!zalloc_cpumask_var(&candidate_mask, GFP_KERNEL))
+		return -ENOMEM;
+
+	cpumask_clear(*cpu_mask);
+	for (cpu = 0; cpu < nthreads; cpu++)
+		cpumask_set_cpu(cpu, *cpu_mask);
+
+	BUG_ON(!cpumask_subset(cpu_present_mask, cpu_possible_mask));
+
+	/* Get a bitmap of unoccupied slots. */
+	cpumask_xor(candidate_mask, cpu_possible_mask, cpu_present_mask);
+
+	if (assigned_node != NUMA_NO_NODE) {
+		/*
+		 * Remove free ids previously assigned on the other nodes. We
+		 * can walk only online nodes because once a node became online
+		 * it is not turned offlined back.
+		 */
+		for_each_online_node(node) {
+			if (node == assigned_node)
+				continue;
+			cpumask_andnot(candidate_mask, candidate_mask,
+				       node_recorded_ids_map[node]);
+		}
+	}
+
+	if (cpumask_empty(candidate_mask))
+		goto out;
+
+	while (!cpumask_empty(*cpu_mask)) {
+		if (cpumask_subset(*cpu_mask, candidate_mask))
+			/* Found a range where we can insert the new cpu(s) */
+			break;
+		cpumask_shift_left(*cpu_mask, *cpu_mask, nthreads);
+	}
+
+	if (!cpumask_empty(*cpu_mask))
+		rc = 0;
+
+out:
+	free_cpumask_var(candidate_mask);
+	return rc;
+}
+
 /*
  * Update cpu_present_mask and paca(s) for a new cpu node.  The wrinkle
- * here is that a cpu device node may represent up to two logical cpus
+ * here is that a cpu device node may represent multiple logical cpus
  * in the SMT case.  We must honor the assumption in other code that
  * the logical ids for sibling SMT threads x and y are adjacent, such
  * that x^1 == y and y^1 == x.
  */
 static int pseries_add_processor(struct device_node *np)
 {
-	unsigned int cpu;
-	cpumask_var_t candidate_mask, tmp;
-	int err = -ENOSPC, len, nthreads, i;
+	int len, nthreads, node, cpu, assigned_node;
+	int rc = 0;
+	cpumask_var_t cpu_mask;
 	const __be32 *intserv;
 
 	intserv = of_get_property(np, "ibm,ppc-interrupt-server#s", &len);
 	if (!intserv)
 		return 0;
 
-	zalloc_cpumask_var(&candidate_mask, GFP_KERNEL);
-	zalloc_cpumask_var(&tmp, GFP_KERNEL);
-
 	nthreads = len / sizeof(u32);
-	for (i = 0; i < nthreads; i++)
-		cpumask_set_cpu(i, tmp);
 
-	cpu_maps_update_begin();
+	if (!alloc_cpumask_var(&cpu_mask, GFP_KERNEL))
+		return -ENOMEM;
 
-	BUG_ON(!cpumask_subset(cpu_present_mask, cpu_possible_mask));
+	/*
+	 * Fetch from the DT nodes read by dlpar_configure_connector() the NUMA
+	 * node id the added CPU belongs to.
+	 */
+	node = of_node_to_nid(np);
+	if (node < 0 || !node_possible(node))
+		node = first_online_node;
 
-	/* Get a bitmap of unoccupied slots. */
-	cpumask_xor(candidate_mask, cpu_possible_mask, cpu_present_mask);
-	if (cpumask_empty(candidate_mask)) {
-		/* If we get here, it most likely means that NR_CPUS is
-		 * less than the partition's max processors setting.
+	BUG_ON(node == NUMA_NO_NODE);
+	assigned_node = node;
+
+	cpu_maps_update_begin();
+
+	rc = find_cpu_id_range(nthreads, node, &cpu_mask);
+	if (rc && nr_node_ids > 1) {
+		/*
+		 * Try again, considering the free CPU ids from the other node.
 		 */
-		printk(KERN_ERR "Cannot add cpu %pOF; this system configuration"
-		       " supports %d logical cpus.\n", np,
-		       num_possible_cpus());
-		goto out_unlock;
+		node = NUMA_NO_NODE;
+		rc = find_cpu_id_range(nthreads, NUMA_NO_NODE, &cpu_mask);
 	}
 
-	while (!cpumask_empty(tmp))
-		if (cpumask_subset(tmp, candidate_mask))
-			/* Found a range where we can insert the new cpu(s) */
-			break;
-		else
-			cpumask_shift_left(tmp, tmp, nthreads);
-
-	if (cpumask_empty(tmp)) {
-		printk(KERN_ERR "Unable to find space in cpu_present_mask for"
-		       " processor %pOFn with %d thread(s)\n", np,
-		       nthreads);
-		goto out_unlock;
+	if (rc) {
+		pr_err("Cannot add cpu %pOF; this system configuration"
+		       " supports %d logical cpus.\n", np, num_possible_cpus());
+		goto out;
 	}
 
-	for_each_cpu(cpu, tmp) {
+	for_each_cpu(cpu, cpu_mask) {
 		BUG_ON(cpu_present(cpu));
 		set_cpu_present(cpu, true);
 		set_hard_smp_processor_id(cpu, be32_to_cpu(*intserv++));
 	}
-	err = 0;
-out_unlock:
+
+	/* Record the newly used CPU ids for the associate node. */
+	cpumask_or(node_recorded_ids_map[assigned_node],
+		   node_recorded_ids_map[assigned_node], cpu_mask);
+
+	/*
+	 * If node is set to NUMA_NO_NODE, CPU ids have be reused from
+	 * another node, remove them from its mask.
+	 */
+	if (node == NUMA_NO_NODE) {
+		cpu = cpumask_first(cpu_mask);
+		pr_warn("Reusing free CPU ids %d-%d from another node\n",
+			cpu, cpu + nthreads - 1);
+		for_each_online_node(node) {
+			if (node == assigned_node)
+				continue;
+			cpumask_andnot(node_recorded_ids_map[node],
+				       node_recorded_ids_map[node],
+				       cpu_mask);
+		}
+	}
+
+out:
 	cpu_maps_update_done();
-	free_cpumask_var(candidate_mask);
-	free_cpumask_var(tmp);
-	return err;
+	free_cpumask_var(cpu_mask);
+	return rc;
 }
 
 /*
@@ -359,28 +342,27 @@ static int dlpar_offline_cpu(struct device_node *dn)
 			if (get_hard_smp_processor_id(cpu) != thread)
 				continue;
 
-			if (get_cpu_current_state(cpu) == CPU_STATE_OFFLINE)
+			if (!cpu_online(cpu))
 				break;
 
-			if (get_cpu_current_state(cpu) == CPU_STATE_ONLINE) {
-				set_preferred_offline_state(cpu,
-							    CPU_STATE_OFFLINE);
-				cpu_maps_update_done();
-				timed_topology_update(1);
-				rc = device_offline(get_cpu_device(cpu));
-				if (rc)
-					goto out;
-				cpu_maps_update_begin();
-				break;
-			}
-
 			/*
-			 * The cpu is in CPU_STATE_INACTIVE.
-			 * Upgrade it's state to CPU_STATE_OFFLINE.
+			 * device_offline() will return -EBUSY (via cpu_down()) if there
+			 * is only one CPU left. Check it here to fail earlier and with a
+			 * more informative error message, while also retaining the
+			 * cpu_add_remove_lock to be sure that no CPUs are being
+			 * online/offlined during this check.
 			 */
-			set_preferred_offline_state(cpu, CPU_STATE_OFFLINE);
-			WARN_ON(plpar_hcall_norets(H_PROD, thread) != H_SUCCESS);
-			__cpu_die(cpu);
+			if (num_online_cpus() == 1) {
+				pr_warn("Unable to remove last online CPU %pOFn\n", dn);
+				rc = -EBUSY;
+				goto out_unlock;
+			}
+
+			cpu_maps_update_done();
+			rc = device_offline(get_cpu_device(cpu));
+			if (rc)
+				goto out;
+			cpu_maps_update_begin();
 			break;
 		}
 		if (cpu == num_possible_cpus()) {
@@ -388,6 +370,7 @@ static int dlpar_offline_cpu(struct device_node *dn)
 				thread);
 		}
 	}
+out_unlock:
 	cpu_maps_update_done();
 
 out:
@@ -414,11 +397,8 @@ static int dlpar_online_cpu(struct device_node *dn)
 		for_each_present_cpu(cpu) {
 			if (get_hard_smp_processor_id(cpu) != thread)
 				continue;
-			BUG_ON(get_cpu_current_state(cpu)
-					!= CPU_STATE_OFFLINE);
 			cpu_maps_update_done();
-			timed_topology_update(1);
-			find_and_online_cpu_nid(cpu);
+			find_and_update_cpu_nid(cpu);
 			rc = device_online(get_cpu_device(cpu));
 			if (rc) {
 				dlpar_offline_cpu(dn);
@@ -536,6 +516,27 @@ static bool valid_cpu_drc_index(struct device_node *parent, u32 drc_index)
 	return found;
 }
 
+static int pseries_cpuhp_attach_nodes(struct device_node *dn)
+{
+	struct of_changeset cs;
+	int ret;
+
+	/*
+	 * This device node is unattached but may have siblings; open-code the
+	 * traversal.
+	 */
+	for (of_changeset_init(&cs); dn != NULL; dn = dn->sibling) {
+		ret = of_changeset_attach_node(&cs, dn);
+		if (ret)
+			goto out;
+	}
+
+	ret = of_changeset_apply(&cs);
+out:
+	of_changeset_destroy(&cs);
+	return ret;
+}
+
 static ssize_t dlpar_cpu_add(u32 drc_index)
 {
 	struct device_node *dn, *parent;
@@ -578,7 +579,7 @@ static ssize_t dlpar_cpu_add(u32 drc_index)
 		return -EINVAL;
 	}
 
-	rc = dlpar_attach_node(dn, parent);
+	rc = pseries_cpuhp_attach_nodes(dn);
 
 	/* Regardless we are done with parent now */
 	of_node_put(parent);
@@ -595,6 +596,8 @@ static ssize_t dlpar_cpu_add(u32 drc_index)
 		return saved_rc;
 	}
 
+	update_numa_distance(dn);
+
 	rc = dlpar_online_cpu(dn);
 	if (rc) {
 		saved_rc = rc;
@@ -613,6 +616,60 @@ static ssize_t dlpar_cpu_add(u32 drc_index)
 	return rc;
 }
 
+static unsigned int pseries_cpuhp_cache_use_count(const struct device_node *cachedn)
+{
+	unsigned int use_count = 0;
+	struct device_node *dn, *tn;
+
+	WARN_ON(!of_node_is_type(cachedn, "cache"));
+
+	for_each_of_cpu_node(dn) {
+		tn = of_find_next_cache_node(dn);
+		of_node_put(tn);
+		if (tn == cachedn)
+			use_count++;
+	}
+
+	for_each_node_by_type(dn, "cache") {
+		tn = of_find_next_cache_node(dn);
+		of_node_put(tn);
+		if (tn == cachedn)
+			use_count++;
+	}
+
+	return use_count;
+}
+
+static int pseries_cpuhp_detach_nodes(struct device_node *cpudn)
+{
+	struct device_node *dn;
+	struct of_changeset cs;
+	int ret = 0;
+
+	of_changeset_init(&cs);
+	ret = of_changeset_detach_node(&cs, cpudn);
+	if (ret)
+		goto out;
+
+	dn = cpudn;
+	while ((dn = of_find_next_cache_node(dn))) {
+		if (pseries_cpuhp_cache_use_count(dn) > 1) {
+			of_node_put(dn);
+			break;
+		}
+
+		ret = of_changeset_detach_node(&cs, dn);
+		of_node_put(dn);
+		if (ret)
+			goto out;
+	}
+
+	ret = of_changeset_apply(&cs);
+out:
+	of_changeset_destroy(&cs);
+	return ret;
+}
+
 static ssize_t dlpar_cpu_remove(struct device_node *dn, u32 drc_index)
 {
 	int rc;
@@ -634,7 +691,7 @@ static ssize_t dlpar_cpu_remove(struct device_node *dn, u32 drc_index)
 		return rc;
 	}
 
-	rc = dlpar_detach_node(dn);
+	rc = pseries_cpuhp_detach_nodes(dn);
 	if (rc) {
 		int saved_rc = rc;
 
@@ -686,258 +743,32 @@ static int dlpar_cpu_remove_by_index(u32 drc_index)
 	return rc;
 }
 
-static int find_dlpar_cpus_to_remove(u32 *cpu_drcs, int cpus_to_remove)
-{
-	struct device_node *dn;
-	int cpus_found = 0;
-	int rc;
-
-	/* We want to find cpus_to_remove + 1 CPUs to ensure we do not
-	 * remove the last CPU.
-	 */
-	for_each_node_by_type(dn, "cpu") {
-		cpus_found++;
-
-		if (cpus_found > cpus_to_remove) {
-			of_node_put(dn);
-			break;
-		}
-
-		/* Note that cpus_found is always 1 ahead of the index
-		 * into the cpu_drcs array, so we use cpus_found - 1
-		 */
-		rc = of_property_read_u32(dn, "ibm,my-drc-index",
-					  &cpu_drcs[cpus_found - 1]);
-		if (rc) {
-			pr_warn("Error occurred getting drc-index for %pOFn\n",
-				dn);
-			of_node_put(dn);
-			return -1;
-		}
-	}
-
-	if (cpus_found < cpus_to_remove) {
-		pr_warn("Failed to find enough CPUs (%d of %d) to remove\n",
-			cpus_found, cpus_to_remove);
-	} else if (cpus_found == cpus_to_remove) {
-		pr_warn("Cannot remove all CPUs\n");
-	}
-
-	return cpus_found;
-}
-
-static int dlpar_cpu_remove_by_count(u32 cpus_to_remove)
-{
-	u32 *cpu_drcs;
-	int cpus_found;
-	int cpus_removed = 0;
-	int i, rc;
-
-	pr_debug("Attempting to hot-remove %d CPUs\n", cpus_to_remove);
-
-	cpu_drcs = kcalloc(cpus_to_remove, sizeof(*cpu_drcs), GFP_KERNEL);
-	if (!cpu_drcs)
-		return -EINVAL;
-
-	cpus_found = find_dlpar_cpus_to_remove(cpu_drcs, cpus_to_remove);
-	if (cpus_found <= cpus_to_remove) {
-		kfree(cpu_drcs);
-		return -EINVAL;
-	}
-
-	for (i = 0; i < cpus_to_remove; i++) {
-		rc = dlpar_cpu_remove_by_index(cpu_drcs[i]);
-		if (rc)
-			break;
-
-		cpus_removed++;
-	}
-
-	if (cpus_removed != cpus_to_remove) {
-		pr_warn("CPU hot-remove failed, adding back removed CPUs\n");
-
-		for (i = 0; i < cpus_removed; i++)
-			dlpar_cpu_add(cpu_drcs[i]);
-
-		rc = -EINVAL;
-	} else {
-		rc = 0;
-	}
-
-	kfree(cpu_drcs);
-	return rc;
-}
-
-static int find_drc_info_cpus_to_add(struct device_node *cpus,
-				     struct property *info,
-				     u32 *cpu_drcs, u32 cpus_to_add)
-{
-	struct of_drc_info drc;
-	const __be32 *value;
-	u32 count, drc_index;
-	int cpus_found = 0;
-	int i, j;
-
-	if (!info)
-		return -1;
-
-	value = of_prop_next_u32(info, NULL, &count);
-	if (value)
-		value++;
-
-	for (i = 0; i < count; i++) {
-		of_read_drc_info_cell(&info, &value, &drc);
-		if (strncmp(drc.drc_type, "CPU", 3))
-			break;
-
-		drc_index = drc.drc_index_start;
-		for (j = 0; j < drc.num_sequential_elems; j++) {
-			if (dlpar_cpu_exists(cpus, drc_index))
-				continue;
-
-			cpu_drcs[cpus_found++] = drc_index;
-
-			if (cpus_found == cpus_to_add)
-				return cpus_found;
-
-			drc_index += drc.sequential_inc;
-		}
-	}
-
-	return cpus_found;
-}
-
-static int find_drc_index_cpus_to_add(struct device_node *cpus,
-				      u32 *cpu_drcs, u32 cpus_to_add)
-{
-	int cpus_found = 0;
-	int index, rc;
-	u32 drc_index;
-
-	/* Search the ibm,drc-indexes array for possible CPU drcs to
-	 * add. Note that the format of the ibm,drc-indexes array is
-	 * the number of entries in the array followed by the array
-	 * of drc values so we start looking at index = 1.
-	 */
-	index = 1;
-	while (cpus_found < cpus_to_add) {
-		rc = of_property_read_u32_index(cpus, "ibm,drc-indexes",
-						index++, &drc_index);
-
-		if (rc)
-			break;
-
-		if (dlpar_cpu_exists(cpus, drc_index))
-			continue;
-
-		cpu_drcs[cpus_found++] = drc_index;
-	}
-
-	return cpus_found;
-}
-
-static int dlpar_cpu_add_by_count(u32 cpus_to_add)
-{
-	struct device_node *parent;
-	struct property *info;
-	u32 *cpu_drcs;
-	int cpus_added = 0;
-	int cpus_found;
-	int i, rc;
-
-	pr_debug("Attempting to hot-add %d CPUs\n", cpus_to_add);
-
-	cpu_drcs = kcalloc(cpus_to_add, sizeof(*cpu_drcs), GFP_KERNEL);
-	if (!cpu_drcs)
-		return -EINVAL;
-
-	parent = of_find_node_by_path("/cpus");
-	if (!parent) {
-		pr_warn("Could not find CPU root node in device tree\n");
-		kfree(cpu_drcs);
-		return -1;
-	}
-
-	info = of_find_property(parent, "ibm,drc-info", NULL);
-	if (info)
-		cpus_found = find_drc_info_cpus_to_add(parent, info, cpu_drcs, cpus_to_add);
-	else
-		cpus_found = find_drc_index_cpus_to_add(parent, cpu_drcs, cpus_to_add);
-
-	of_node_put(parent);
-
-	if (cpus_found < cpus_to_add) {
-		pr_warn("Failed to find enough CPUs (%d of %d) to add\n",
-			cpus_found, cpus_to_add);
-		kfree(cpu_drcs);
-		return -EINVAL;
-	}
-
-	for (i = 0; i < cpus_to_add; i++) {
-		rc = dlpar_cpu_add(cpu_drcs[i]);
-		if (rc)
-			break;
-
-		cpus_added++;
-	}
-
-	if (cpus_added < cpus_to_add) {
-		pr_warn("CPU hot-add failed, removing any added CPUs\n");
-
-		for (i = 0; i < cpus_added; i++)
-			dlpar_cpu_remove_by_index(cpu_drcs[i]);
-
-		rc = -EINVAL;
-	} else {
-		rc = 0;
-	}
-
-	kfree(cpu_drcs);
-	return rc;
-}
-
-int dlpar_cpu_readd(int cpu)
-{
-	struct device_node *dn;
-	struct device *dev;
-	u32 drc_index;
-	int rc;
-
-	dev = get_cpu_device(cpu);
-	dn = dev->of_node;
-
-	rc = of_property_read_u32(dn, "ibm,my-drc-index", &drc_index);
-
-	rc = dlpar_cpu_remove_by_index(drc_index);
-	if (!rc)
-		rc = dlpar_cpu_add(drc_index);
-
-	return rc;
-}
-
 int dlpar_cpu(struct pseries_hp_errorlog *hp_elog)
 {
-	u32 count, drc_index;
+	u32 drc_index;
 	int rc;
 
-	count = hp_elog->_drc_u.drc_count;
 	drc_index = hp_elog->_drc_u.drc_index;
 
 	lock_device_hotplug();
 
 	switch (hp_elog->action) {
 	case PSERIES_HP_ELOG_ACTION_REMOVE:
-		if (hp_elog->id_type == PSERIES_HP_ELOG_ID_DRC_COUNT)
-			rc = dlpar_cpu_remove_by_count(count);
-		else if (hp_elog->id_type == PSERIES_HP_ELOG_ID_DRC_INDEX)
+		if (hp_elog->id_type == PSERIES_HP_ELOG_ID_DRC_INDEX) {
 			rc = dlpar_cpu_remove_by_index(drc_index);
+			/*
+			 * Setting the isolation state of an UNISOLATED/CONFIGURED
+			 * device to UNISOLATE is a no-op, but the hypervisor can
+			 * use it as a hint that the CPU removal failed.
+			 */
+			if (rc)
+				dlpar_unisolate_drc(drc_index);
+		}
 		else
 			rc = -EINVAL;
 		break;
 	case PSERIES_HP_ELOG_ACTION_ADD:
-		if (hp_elog->id_type == PSERIES_HP_ELOG_ID_DRC_COUNT)
-			rc = dlpar_cpu_add_by_count(count);
-		else if (hp_elog->id_type == PSERIES_HP_ELOG_ID_DRC_INDEX)
+		if (hp_elog->id_type == PSERIES_HP_ELOG_ID_DRC_INDEX)
 			rc = dlpar_cpu_add(drc_index);
 		else
 			rc = -EINVAL;
@@ -1013,28 +844,10 @@ static struct notifier_block pseries_smp_nb = {
 	.notifier_call = pseries_smp_notifier,
 };
 
-#define MAX_CEDE_LATENCY_LEVELS		4
-#define	CEDE_LATENCY_PARAM_LENGTH	10
-#define CEDE_LATENCY_PARAM_MAX_LENGTH	\
-	(MAX_CEDE_LATENCY_LEVELS * CEDE_LATENCY_PARAM_LENGTH * sizeof(char))
-#define CEDE_LATENCY_TOKEN		45
-
-static char cede_parameters[CEDE_LATENCY_PARAM_MAX_LENGTH];
-
-static int parse_cede_parameters(void)
-{
-	memset(cede_parameters, 0, CEDE_LATENCY_PARAM_MAX_LENGTH);
-	return rtas_call(rtas_token("ibm,get-system-parameter"), 3, 1,
-			 NULL,
-			 CEDE_LATENCY_TOKEN,
-			 __pa(cede_parameters),
-			 CEDE_LATENCY_PARAM_MAX_LENGTH);
-}
-
 static int __init pseries_cpu_hotplug_init(void)
 {
-	int cpu;
 	int qcss_tok;
+	unsigned int node;
 
 #ifdef CONFIG_ARCH_CPU_PROBE_RELEASE
 	ppc_md.cpu_probe = dlpar_cpu_probe;
@@ -1051,20 +864,23 @@ static int __init pseries_cpu_hotplug_init(void)
 		return 0;
 	}
 
-	ppc_md.cpu_die = pseries_mach_cpu_die;
+	smp_ops->cpu_offline_self = pseries_cpu_offline_self;
 	smp_ops->cpu_disable = pseries_cpu_disable;
 	smp_ops->cpu_die = pseries_cpu_die;
 
 	/* Processors can be added/removed only on LPAR */
 	if (firmware_has_feature(FW_FEATURE_LPAR)) {
-		of_reconfig_notifier_register(&pseries_smp_nb);
-		cpu_maps_update_begin();
-		if (cede_offline_enabled && parse_cede_parameters() == 0) {
-			default_offline_state = CPU_STATE_INACTIVE;
-			for_each_online_cpu(cpu)
-				set_default_offline_state(cpu);
+		for_each_node(node) {
+			if (!alloc_cpumask_var_node(&node_recorded_ids_map[node],
+						    GFP_KERNEL, node))
+				return -ENOMEM;
+
+			/* Record ids of CPU added at boot time */
+			cpumask_copy(node_recorded_ids_map[node],
+				     cpumask_of_node(node));
 		}
-		cpu_maps_update_done();
+
+		of_reconfig_notifier_register(&pseries_smp_nb);
 	}
 
 	return 0;