aboutsummaryrefslogtreecommitdiff
path: root/arch/powerpc/mm/numa.c
diff options
context:
space:
mode:
Diffstat (limited to 'arch/powerpc/mm/numa.c')
-rw-r--r--arch/powerpc/mm/numa.c133
1 files changed, 118 insertions, 15 deletions
diff --git a/arch/powerpc/mm/numa.c b/arch/powerpc/mm/numa.c
index c916127f10c..3b181b22cd4 100644
--- a/arch/powerpc/mm/numa.c
+++ b/arch/powerpc/mm/numa.c
@@ -31,6 +31,8 @@
#include <asm/sparsemem.h>
#include <asm/prom.h>
#include <asm/smp.h>
+#include <asm/cputhreads.h>
+#include <asm/topology.h>
#include <asm/firmware.h>
#include <asm/paca.h>
#include <asm/hvcall.h>
@@ -152,9 +154,22 @@ static void __init get_node_active_region(unsigned long pfn,
}
}
-static void map_cpu_to_node(int cpu, int node)
+static void reset_numa_cpu_lookup_table(void)
+{
+ unsigned int cpu;
+
+ for_each_possible_cpu(cpu)
+ numa_cpu_lookup_table[cpu] = -1;
+}
+
+static void update_numa_cpu_lookup_table(unsigned int cpu, int node)
{
numa_cpu_lookup_table[cpu] = node;
+}
+
+static void map_cpu_to_node(int cpu, int node)
+{
+ update_numa_cpu_lookup_table(cpu, node);
dbg("adding cpu %d to node %d\n", cpu, node);
@@ -195,7 +210,7 @@ static const __be32 *of_get_usable_memory(struct device_node *memory)
u32 len;
prop = of_get_property(memory, "linux,drconf-usable-memory", &len);
if (!prop || len < sizeof(unsigned int))
- return 0;
+ return NULL;
return prop;
}
@@ -217,6 +232,7 @@ int __node_distance(int a, int b)
return distance;
}
+EXPORT_SYMBOL(__node_distance);
static void initialize_distance_lookup_table(int nid,
const __be32 *associativity)
@@ -522,11 +538,24 @@ static int of_drconf_to_nid_single(struct of_drconf_cell *drmem,
*/
static int numa_setup_cpu(unsigned long lcpu)
{
- int nid = 0;
- struct device_node *cpu = of_get_cpu_node(lcpu, NULL);
+ int nid;
+ struct device_node *cpu;
+
+ /*
+ * If a valid cpu-to-node mapping is already available, use it
+ * directly instead of querying the firmware, since it represents
+ * the most recent mapping notified to us by the platform (eg: VPHN).
+ */
+ if ((nid = numa_cpu_lookup_table[lcpu]) >= 0) {
+ map_cpu_to_node(lcpu, nid);
+ return nid;
+ }
+
+ cpu = of_get_cpu_node(lcpu, NULL);
if (!cpu) {
WARN_ON(1);
+ nid = 0;
goto out;
}
@@ -542,16 +571,38 @@ out:
return nid;
}
+static void verify_cpu_node_mapping(int cpu, int node)
+{
+ int base, sibling, i;
+
+ /* Verify that all the threads in the core belong to the same node */
+ base = cpu_first_thread_sibling(cpu);
+
+ for (i = 0; i < threads_per_core; i++) {
+ sibling = base + i;
+
+ if (sibling == cpu || cpu_is_offline(sibling))
+ continue;
+
+ if (cpu_to_node(sibling) != node) {
+ WARN(1, "CPU thread siblings %d and %d don't belong"
+ " to the same node!\n", cpu, sibling);
+ break;
+ }
+ }
+}
+
static int cpu_numa_callback(struct notifier_block *nfb, unsigned long action,
void *hcpu)
{
unsigned long lcpu = (unsigned long)hcpu;
- int ret = NOTIFY_DONE;
+ int ret = NOTIFY_DONE, nid;
switch (action) {
case CPU_UP_PREPARE:
case CPU_UP_PREPARE_FROZEN:
- numa_setup_cpu(lcpu);
+ nid = numa_setup_cpu(lcpu);
+ verify_cpu_node_mapping((int)lcpu, nid);
ret = NOTIFY_OK;
break;
#ifdef CONFIG_HOTPLUG_CPU
@@ -670,7 +721,8 @@ static void __init parse_drconf_memory(struct device_node *memory)
node_set_online(nid);
sz = numa_enforce_memory_limit(base, size);
if (sz)
- memblock_set_node(base, sz, nid);
+ memblock_set_node(base, sz,
+ &memblock.memory, nid);
} while (--ranges);
}
}
@@ -760,7 +812,7 @@ new_range:
continue;
}
- memblock_set_node(start, size, nid);
+ memblock_set_node(start, size, &memblock.memory, nid);
if (--ranges)
goto new_range;
@@ -797,7 +849,8 @@ static void __init setup_nonnuma(void)
fake_numa_create_new_node(end_pfn, &nid);
memblock_set_node(PFN_PHYS(start_pfn),
- PFN_PHYS(end_pfn - start_pfn), nid);
+ PFN_PHYS(end_pfn - start_pfn),
+ &memblock.memory, nid);
node_set_online(nid);
}
}
@@ -938,8 +991,7 @@ static void __init mark_reserved_regions_for_nid(int nid)
unsigned long start_pfn = physbase >> PAGE_SHIFT;
unsigned long end_pfn = PFN_UP(physbase + size);
struct node_active_region node_ar;
- unsigned long node_end_pfn = node->node_start_pfn +
- node->node_spanned_pages;
+ unsigned long node_end_pfn = pgdat_end_pfn(node);
/*
* Check to make sure that this memblock.reserved area is
@@ -1068,6 +1120,7 @@ void __init do_init_bootmem(void)
*/
setup_node_to_cpumask_map();
+ reset_numa_cpu_lookup_table();
register_cpu_notifier(&ppc64_numa_nb);
cpu_numa_callback(&ppc64_numa_nb, CPU_UP_PREPARE,
(void *)(unsigned long)boot_cpuid);
@@ -1154,7 +1207,7 @@ static int hot_add_drconf_scn_to_nid(struct device_node *memory,
* represented in the device tree as a node (i.e. memory@XXXX) for
* each memblock.
*/
-int hot_add_node_scn_to_nid(unsigned long scn_addr)
+static int hot_add_node_scn_to_nid(unsigned long scn_addr)
{
struct device_node *memory;
int nid = -1;
@@ -1235,7 +1288,7 @@ static u64 hot_add_drconf_memory_max(void)
struct device_node *memory = NULL;
unsigned int drconf_cell_cnt = 0;
u64 lmb_size = 0;
- const __be32 *dm = 0;
+ const __be32 *dm = NULL;
memory = of_find_node_by_path("/ibm,dynamic-reconfiguration-memory");
if (memory) {
@@ -1446,6 +1499,33 @@ static int update_cpu_topology(void *data)
return 0;
}
+static int update_lookup_table(void *data)
+{
+ struct topology_update_data *update;
+
+ if (!data)
+ return -EINVAL;
+
+ /*
+ * Upon topology update, the numa-cpu lookup table needs to be updated
+ * for all threads in the core, including offline CPUs, to ensure that
+ * future hotplug operations respect the cpu-to-node associativity
+ * properly.
+ */
+ for (update = data; update; update = update->next) {
+ int nid, base, j;
+
+ nid = update->new_nid;
+ base = cpu_first_thread_sibling(update->cpu);
+
+ for (j = 0; j < threads_per_core; j++) {
+ update_numa_cpu_lookup_table(base + j, nid);
+ }
+ }
+
+ return 0;
+}
+
/*
* Update the node maps and sysfs entries for each cpu whose home node
* has changed. Returns 1 when the topology has changed, and 0 otherwise.
@@ -1512,8 +1592,30 @@ int arch_update_cpu_topology(void)
cpu = cpu_last_thread_sibling(cpu);
}
+ /*
+ * In cases where we have nothing to update (because the updates list
+ * is too short or because the new topology is same as the old one),
+ * skip invoking update_cpu_topology() via stop-machine(). This is
+ * necessary (and not just a fast-path optimization) since stop-machine
+ * can end up electing a random CPU to run update_cpu_topology(), and
+ * thus trick us into setting up incorrect cpu-node mappings (since
+ * 'updates' is kzalloc()'ed).
+ *
+ * And for the similar reason, we will skip all the following updating.
+ */
+ if (!cpumask_weight(&updated_cpus))
+ goto out;
+
stop_machine(update_cpu_topology, &updates[0], &updated_cpus);
+ /*
+ * Update the numa-cpu lookup table with the new mappings, even for
+ * offline CPUs. It is best to perform this update from the stop-
+ * machine context.
+ */
+ stop_machine(update_lookup_table, &updates[0],
+ cpumask_of(raw_smp_processor_id()));
+
for (ud = &updates[0]; ud; ud = ud->next) {
unregister_cpu_under_node(ud->cpu, ud->old_nid);
register_cpu_under_node(ud->cpu, ud->new_nid);
@@ -1525,6 +1627,7 @@ int arch_update_cpu_topology(void)
changed = 1;
}
+out:
kfree(updates);
return changed;
}
@@ -1535,7 +1638,7 @@ static void topology_work_fn(struct work_struct *work)
}
static DECLARE_WORK(topology_work, topology_work_fn);
-void topology_schedule_update(void)
+static void topology_schedule_update(void)
{
schedule_work(&topology_work);
}
@@ -1698,7 +1801,7 @@ static const struct file_operations topology_ops = {
static int topology_update_init(void)
{
start_topology_update();
- proc_create("powerpc/topology_updates", 644, NULL, &topology_ops);
+ proc_create("powerpc/topology_updates", 0644, NULL, &topology_ops);
return 0;
}
generated by cgit v1.2.3-70-g09d2 (git 2.46.0) at 2025-11-04 19:06:36 +0000