^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1) // SPDX-License-Identifier: GPL-2.0-or-later
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 2) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 3) * pSeries NUMA support
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 4) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 5) * Copyright (C) 2002 Anton Blanchard <anton@au.ibm.com>, IBM
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 6) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 7) #define pr_fmt(fmt) "numa: " fmt
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 8)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 9) #include <linux/threads.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 10) #include <linux/memblock.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 11) #include <linux/init.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 12) #include <linux/mm.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 13) #include <linux/mmzone.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 14) #include <linux/export.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 15) #include <linux/nodemask.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 16) #include <linux/cpu.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 17) #include <linux/notifier.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 18) #include <linux/of.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 19) #include <linux/pfn.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 20) #include <linux/cpuset.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 21) #include <linux/node.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 22) #include <linux/stop_machine.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 23) #include <linux/proc_fs.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 24) #include <linux/seq_file.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 25) #include <linux/uaccess.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 26) #include <linux/slab.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 27) #include <asm/cputhreads.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 28) #include <asm/sparsemem.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 29) #include <asm/prom.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 30) #include <asm/smp.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 31) #include <asm/topology.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 32) #include <asm/firmware.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 33) #include <asm/paca.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 34) #include <asm/hvcall.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 35) #include <asm/setup.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 36) #include <asm/vdso.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 37) #include <asm/drmem.h>
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 38)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 39) static int numa_enabled = 1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 40)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 41) static char *cmdline __initdata;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 42)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 43) static int numa_debug;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 44) #define dbg(args...) if (numa_debug) { printk(KERN_INFO args); }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 45)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 46) int numa_cpu_lookup_table[NR_CPUS];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 47) cpumask_var_t node_to_cpumask_map[MAX_NUMNODES];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 48) struct pglist_data *node_data[MAX_NUMNODES];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 49)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 50) EXPORT_SYMBOL(numa_cpu_lookup_table);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 51) EXPORT_SYMBOL(node_to_cpumask_map);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 52) EXPORT_SYMBOL(node_data);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 53)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 54) static int min_common_depth;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 55) static int n_mem_addr_cells, n_mem_size_cells;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 56) static int form1_affinity;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 57)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 58) #define MAX_DISTANCE_REF_POINTS 4
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 59) static int distance_ref_points_depth;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 60) static const __be32 *distance_ref_points;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 61) static int distance_lookup_table[MAX_NUMNODES][MAX_DISTANCE_REF_POINTS];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 62)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 63) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 64) * Allocate node_to_cpumask_map based on number of available nodes
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 65) * Requires node_possible_map to be valid.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 66) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 67) * Note: cpumask_of_node() is not valid until after this is done.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 68) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 69) static void __init setup_node_to_cpumask_map(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 70) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 71) unsigned int node;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 72)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 73) /* setup nr_node_ids if not done yet */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 74) if (nr_node_ids == MAX_NUMNODES)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 75) setup_nr_node_ids();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 76)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 77) /* allocate the map */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 78) for_each_node(node)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 79) alloc_bootmem_cpumask_var(&node_to_cpumask_map[node]);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 80)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 81) /* cpumask_of_node() will now work */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 82) dbg("Node to cpumask map for %u nodes\n", nr_node_ids);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 83) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 84)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 85) static int __init fake_numa_create_new_node(unsigned long end_pfn,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 86) unsigned int *nid)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 87) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 88) unsigned long long mem;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 89) char *p = cmdline;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 90) static unsigned int fake_nid;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 91) static unsigned long long curr_boundary;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 92)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 93) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 94) * Modify node id, iff we started creating NUMA nodes
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 95) * We want to continue from where we left of the last time
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 96) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 97) if (fake_nid)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 98) *nid = fake_nid;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 99) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 100) * In case there are no more arguments to parse, the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 101) * node_id should be the same as the last fake node id
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 102) * (we've handled this above).
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 103) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 104) if (!p)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 105) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 106)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 107) mem = memparse(p, &p);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 108) if (!mem)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 109) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 110)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 111) if (mem < curr_boundary)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 112) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 113)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 114) curr_boundary = mem;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 115)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 116) if ((end_pfn << PAGE_SHIFT) > mem) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 117) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 118) * Skip commas and spaces
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 119) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 120) while (*p == ',' || *p == ' ' || *p == '\t')
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 121) p++;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 122)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 123) cmdline = p;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 124) fake_nid++;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 125) *nid = fake_nid;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 126) dbg("created new fake_node with id %d\n", fake_nid);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 127) return 1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 128) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 129) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 130) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 131)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 132) static void reset_numa_cpu_lookup_table(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 133) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 134) unsigned int cpu;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 135)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 136) for_each_possible_cpu(cpu)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 137) numa_cpu_lookup_table[cpu] = -1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 138) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 139)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 140) static void map_cpu_to_node(int cpu, int node)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 141) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 142) update_numa_cpu_lookup_table(cpu, node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 143)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 144) dbg("adding cpu %d to node %d\n", cpu, node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 145)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 146) if (!(cpumask_test_cpu(cpu, node_to_cpumask_map[node])))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 147) cpumask_set_cpu(cpu, node_to_cpumask_map[node]);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 148) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 149)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 150) #if defined(CONFIG_HOTPLUG_CPU) || defined(CONFIG_PPC_SPLPAR)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 151) static void unmap_cpu_from_node(unsigned long cpu)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 152) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 153) int node = numa_cpu_lookup_table[cpu];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 154)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 155) dbg("removing cpu %lu from node %d\n", cpu, node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 156)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 157) if (cpumask_test_cpu(cpu, node_to_cpumask_map[node])) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 158) cpumask_clear_cpu(cpu, node_to_cpumask_map[node]);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 159) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 160) printk(KERN_ERR "WARNING: cpu %lu not found in node %d\n",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 161) cpu, node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 162) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 163) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 164) #endif /* CONFIG_HOTPLUG_CPU || CONFIG_PPC_SPLPAR */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 165)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 166) int cpu_distance(__be32 *cpu1_assoc, __be32 *cpu2_assoc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 167) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 168) int dist = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 169)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 170) int i, index;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 171)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 172) for (i = 0; i < distance_ref_points_depth; i++) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 173) index = be32_to_cpu(distance_ref_points[i]);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 174) if (cpu1_assoc[index] == cpu2_assoc[index])
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 175) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 176) dist++;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 177) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 178)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 179) return dist;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 180) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 181)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 182) /* must hold reference to node during call */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 183) static const __be32 *of_get_associativity(struct device_node *dev)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 184) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 185) return of_get_property(dev, "ibm,associativity", NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 186) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 187)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 188) int __node_distance(int a, int b)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 189) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 190) int i;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 191) int distance = LOCAL_DISTANCE;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 192)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 193) if (!form1_affinity)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 194) return ((a == b) ? LOCAL_DISTANCE : REMOTE_DISTANCE);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 195)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 196) for (i = 0; i < distance_ref_points_depth; i++) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 197) if (distance_lookup_table[a][i] == distance_lookup_table[b][i])
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 198) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 199)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 200) /* Double the distance for each NUMA level */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 201) distance *= 2;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 202) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 203)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 204) return distance;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 205) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 206) EXPORT_SYMBOL(__node_distance);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 207)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 208) static void initialize_distance_lookup_table(int nid,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 209) const __be32 *associativity)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 210) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 211) int i;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 212)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 213) if (!form1_affinity)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 214) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 215)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 216) for (i = 0; i < distance_ref_points_depth; i++) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 217) const __be32 *entry;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 218)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 219) entry = &associativity[be32_to_cpu(distance_ref_points[i]) - 1];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 220) distance_lookup_table[nid][i] = of_read_number(entry, 1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 221) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 222) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 223)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 224) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 225) * Returns nid in the range [0..nr_node_ids], or -1 if no useful NUMA
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 226) * info is found.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 227) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 228) static int associativity_to_nid(const __be32 *associativity)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 229) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 230) int nid = NUMA_NO_NODE;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 231)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 232) if (!numa_enabled)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 233) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 234)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 235) if (of_read_number(associativity, 1) >= min_common_depth)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 236) nid = of_read_number(&associativity[min_common_depth], 1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 237)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 238) /* POWER4 LPAR uses 0xffff as invalid node */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 239) if (nid == 0xffff || nid >= nr_node_ids)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 240) nid = NUMA_NO_NODE;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 241)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 242) if (nid > 0 &&
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 243) of_read_number(associativity, 1) >= distance_ref_points_depth) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 244) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 245) * Skip the length field and send start of associativity array
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 246) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 247) initialize_distance_lookup_table(nid, associativity + 1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 248) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 249)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 250) out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 251) return nid;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 252) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 253)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 254) /* Returns the nid associated with the given device tree node,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 255) * or -1 if not found.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 256) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 257) static int of_node_to_nid_single(struct device_node *device)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 258) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 259) int nid = NUMA_NO_NODE;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 260) const __be32 *tmp;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 261)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 262) tmp = of_get_associativity(device);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 263) if (tmp)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 264) nid = associativity_to_nid(tmp);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 265) return nid;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 266) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 267)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 268) /* Walk the device tree upwards, looking for an associativity id */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 269) int of_node_to_nid(struct device_node *device)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 270) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 271) int nid = NUMA_NO_NODE;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 272)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 273) of_node_get(device);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 274) while (device) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 275) nid = of_node_to_nid_single(device);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 276) if (nid != -1)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 277) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 278)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 279) device = of_get_next_parent(device);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 280) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 281) of_node_put(device);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 282)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 283) return nid;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 284) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 285) EXPORT_SYMBOL(of_node_to_nid);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 286)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 287) static int __init find_min_common_depth(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 288) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 289) int depth;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 290) struct device_node *root;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 291)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 292) if (firmware_has_feature(FW_FEATURE_OPAL))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 293) root = of_find_node_by_path("/ibm,opal");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 294) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 295) root = of_find_node_by_path("/rtas");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 296) if (!root)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 297) root = of_find_node_by_path("/");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 298)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 299) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 300) * This property is a set of 32-bit integers, each representing
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 301) * an index into the ibm,associativity nodes.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 302) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 303) * With form 0 affinity the first integer is for an SMP configuration
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 304) * (should be all 0's) and the second is for a normal NUMA
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 305) * configuration. We have only one level of NUMA.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 306) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 307) * With form 1 affinity the first integer is the most significant
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 308) * NUMA boundary and the following are progressively less significant
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 309) * boundaries. There can be more than one level of NUMA.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 310) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 311) distance_ref_points = of_get_property(root,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 312) "ibm,associativity-reference-points",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 313) &distance_ref_points_depth);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 314)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 315) if (!distance_ref_points) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 316) dbg("NUMA: ibm,associativity-reference-points not found.\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 317) goto err;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 318) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 319)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 320) distance_ref_points_depth /= sizeof(int);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 321)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 322) if (firmware_has_feature(FW_FEATURE_OPAL) ||
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 323) firmware_has_feature(FW_FEATURE_TYPE1_AFFINITY)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 324) dbg("Using form 1 affinity\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 325) form1_affinity = 1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 326) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 327)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 328) if (form1_affinity) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 329) depth = of_read_number(distance_ref_points, 1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 330) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 331) if (distance_ref_points_depth < 2) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 332) printk(KERN_WARNING "NUMA: "
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 333) "short ibm,associativity-reference-points\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 334) goto err;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 335) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 336)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 337) depth = of_read_number(&distance_ref_points[1], 1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 338) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 339)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 340) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 341) * Warn and cap if the hardware supports more than
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 342) * MAX_DISTANCE_REF_POINTS domains.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 343) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 344) if (distance_ref_points_depth > MAX_DISTANCE_REF_POINTS) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 345) printk(KERN_WARNING "NUMA: distance array capped at "
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 346) "%d entries\n", MAX_DISTANCE_REF_POINTS);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 347) distance_ref_points_depth = MAX_DISTANCE_REF_POINTS;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 348) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 349)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 350) of_node_put(root);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 351) return depth;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 352)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 353) err:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 354) of_node_put(root);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 355) return -1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 356) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 357)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 358) static void __init get_n_mem_cells(int *n_addr_cells, int *n_size_cells)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 359) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 360) struct device_node *memory = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 361)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 362) memory = of_find_node_by_type(memory, "memory");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 363) if (!memory)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 364) panic("numa.c: No memory nodes found!");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 365)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 366) *n_addr_cells = of_n_addr_cells(memory);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 367) *n_size_cells = of_n_size_cells(memory);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 368) of_node_put(memory);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 369) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 370)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 371) static unsigned long read_n_cells(int n, const __be32 **buf)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 372) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 373) unsigned long result = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 374)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 375) while (n--) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 376) result = (result << 32) | of_read_number(*buf, 1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 377) (*buf)++;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 378) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 379) return result;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 380) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 381)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 382) struct assoc_arrays {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 383) u32 n_arrays;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 384) u32 array_sz;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 385) const __be32 *arrays;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 386) };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 387)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 388) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 389) * Retrieve and validate the list of associativity arrays for drconf
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 390) * memory from the ibm,associativity-lookup-arrays property of the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 391) * device tree..
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 392) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 393) * The layout of the ibm,associativity-lookup-arrays property is a number N
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 394) * indicating the number of associativity arrays, followed by a number M
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 395) * indicating the size of each associativity array, followed by a list
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 396) * of N associativity arrays.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 397) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 398) static int of_get_assoc_arrays(struct assoc_arrays *aa)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 399) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 400) struct device_node *memory;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 401) const __be32 *prop;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 402) u32 len;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 403)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 404) memory = of_find_node_by_path("/ibm,dynamic-reconfiguration-memory");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 405) if (!memory)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 406) return -1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 407)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 408) prop = of_get_property(memory, "ibm,associativity-lookup-arrays", &len);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 409) if (!prop || len < 2 * sizeof(unsigned int)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 410) of_node_put(memory);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 411) return -1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 412) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 413)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 414) aa->n_arrays = of_read_number(prop++, 1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 415) aa->array_sz = of_read_number(prop++, 1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 416)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 417) of_node_put(memory);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 418)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 419) /* Now that we know the number of arrays and size of each array,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 420) * revalidate the size of the property read in.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 421) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 422) if (len < (aa->n_arrays * aa->array_sz + 2) * sizeof(unsigned int))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 423) return -1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 424)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 425) aa->arrays = prop;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 426) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 427) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 428)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 429) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 430) * This is like of_node_to_nid_single() for memory represented in the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 431) * ibm,dynamic-reconfiguration-memory node.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 432) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 433) int of_drconf_to_nid_single(struct drmem_lmb *lmb)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 434) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 435) struct assoc_arrays aa = { .arrays = NULL };
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 436) int default_nid = NUMA_NO_NODE;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 437) int nid = default_nid;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 438) int rc, index;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 439)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 440) if ((min_common_depth < 0) || !numa_enabled)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 441) return default_nid;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 442)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 443) rc = of_get_assoc_arrays(&aa);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 444) if (rc)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 445) return default_nid;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 446)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 447) if (min_common_depth <= aa.array_sz &&
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 448) !(lmb->flags & DRCONF_MEM_AI_INVALID) && lmb->aa_index < aa.n_arrays) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 449) index = lmb->aa_index * aa.array_sz + min_common_depth - 1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 450) nid = of_read_number(&aa.arrays[index], 1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 451)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 452) if (nid == 0xffff || nid >= nr_node_ids)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 453) nid = default_nid;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 454)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 455) if (nid > 0) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 456) index = lmb->aa_index * aa.array_sz;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 457) initialize_distance_lookup_table(nid,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 458) &aa.arrays[index]);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 459) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 460) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 461)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 462) return nid;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 463) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 464)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 465) #ifdef CONFIG_PPC_SPLPAR
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 466) static int vphn_get_nid(long lcpu)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 467) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 468) __be32 associativity[VPHN_ASSOC_BUFSIZE] = {0};
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 469) long rc, hwid;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 470)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 471) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 472) * On a shared lpar, device tree will not have node associativity.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 473) * At this time lppaca, or its __old_status field may not be
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 474) * updated. Hence kernel cannot detect if its on a shared lpar. So
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 475) * request an explicit associativity irrespective of whether the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 476) * lpar is shared or dedicated. Use the device tree property as a
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 477) * fallback. cpu_to_phys_id is only valid between
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 478) * smp_setup_cpu_maps() and smp_setup_pacas().
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 479) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 480) if (firmware_has_feature(FW_FEATURE_VPHN)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 481) if (cpu_to_phys_id)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 482) hwid = cpu_to_phys_id[lcpu];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 483) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 484) hwid = get_hard_smp_processor_id(lcpu);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 485)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 486) rc = hcall_vphn(hwid, VPHN_FLAG_VCPU, associativity);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 487) if (rc == H_SUCCESS)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 488) return associativity_to_nid(associativity);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 489) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 490)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 491) return NUMA_NO_NODE;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 492) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 493) #else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 494) static int vphn_get_nid(long unused)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 495) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 496) return NUMA_NO_NODE;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 497) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 498) #endif /* CONFIG_PPC_SPLPAR */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 499)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 500) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 501) * Figure out to which domain a cpu belongs and stick it there.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 502) * Return the id of the domain used.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 503) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 504) static int numa_setup_cpu(unsigned long lcpu)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 505) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 506) struct device_node *cpu;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 507) int fcpu = cpu_first_thread_sibling(lcpu);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 508) int nid = NUMA_NO_NODE;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 509)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 510) if (!cpu_present(lcpu)) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 511) set_cpu_numa_node(lcpu, first_online_node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 512) return first_online_node;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 513) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 514)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 515) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 516) * If a valid cpu-to-node mapping is already available, use it
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 517) * directly instead of querying the firmware, since it represents
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 518) * the most recent mapping notified to us by the platform (eg: VPHN).
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 519) * Since cpu_to_node binding remains the same for all threads in the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 520) * core. If a valid cpu-to-node mapping is already available, for
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 521) * the first thread in the core, use it.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 522) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 523) nid = numa_cpu_lookup_table[fcpu];
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 524) if (nid >= 0) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 525) map_cpu_to_node(lcpu, nid);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 526) return nid;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 527) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 528)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 529) nid = vphn_get_nid(lcpu);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 530) if (nid != NUMA_NO_NODE)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 531) goto out_present;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 532)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 533) cpu = of_get_cpu_node(lcpu, NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 534)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 535) if (!cpu) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 536) WARN_ON(1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 537) if (cpu_present(lcpu))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 538) goto out_present;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 539) else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 540) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 541) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 542)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 543) nid = of_node_to_nid_single(cpu);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 544) of_node_put(cpu);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 545)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 546) out_present:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 547) if (nid < 0 || !node_possible(nid))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 548) nid = first_online_node;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 549)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 550) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 551) * Update for the first thread of the core. All threads of a core
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 552) * have to be part of the same node. This not only avoids querying
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 553) * for every other thread in the core, but always avoids a case
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 554) * where virtual node associativity change causes subsequent threads
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 555) * of a core to be associated with different nid. However if first
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 556) * thread is already online, expect it to have a valid mapping.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 557) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 558) if (fcpu != lcpu) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 559) WARN_ON(cpu_online(fcpu));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 560) map_cpu_to_node(fcpu, nid);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 561) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 562)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 563) map_cpu_to_node(lcpu, nid);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 564) out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 565) return nid;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 566) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 567)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 568) static void verify_cpu_node_mapping(int cpu, int node)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 569) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 570) int base, sibling, i;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 571)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 572) /* Verify that all the threads in the core belong to the same node */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 573) base = cpu_first_thread_sibling(cpu);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 574)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 575) for (i = 0; i < threads_per_core; i++) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 576) sibling = base + i;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 577)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 578) if (sibling == cpu || cpu_is_offline(sibling))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 579) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 580)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 581) if (cpu_to_node(sibling) != node) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 582) WARN(1, "CPU thread siblings %d and %d don't belong"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 583) " to the same node!\n", cpu, sibling);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 584) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 585) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 586) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 587) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 588)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 589) /* Must run before sched domains notifier. */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 590) static int ppc_numa_cpu_prepare(unsigned int cpu)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 591) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 592) int nid;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 593)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 594) nid = numa_setup_cpu(cpu);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 595) verify_cpu_node_mapping(cpu, nid);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 596) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 597) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 598)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 599) static int ppc_numa_cpu_dead(unsigned int cpu)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 600) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 601) #ifdef CONFIG_HOTPLUG_CPU
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 602) unmap_cpu_from_node(cpu);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 603) #endif
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 604) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 605) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 606)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 607) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 608) * Check and possibly modify a memory region to enforce the memory limit.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 609) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 610) * Returns the size the region should have to enforce the memory limit.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 611) * This will either be the original value of size, a truncated value,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 612) * or zero. If the returned value of size is 0 the region should be
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 613) * discarded as it lies wholly above the memory limit.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 614) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 615) static unsigned long __init numa_enforce_memory_limit(unsigned long start,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 616) unsigned long size)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 617) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 618) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 619) * We use memblock_end_of_DRAM() in here instead of memory_limit because
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 620) * we've already adjusted it for the limit and it takes care of
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 621) * having memory holes below the limit. Also, in the case of
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 622) * iommu_is_off, memory_limit is not set but is implicitly enforced.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 623) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 624)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 625) if (start + size <= memblock_end_of_DRAM())
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 626) return size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 627)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 628) if (start >= memblock_end_of_DRAM())
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 629) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 630)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 631) return memblock_end_of_DRAM() - start;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 632) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 633)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 634) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 635) * Reads the counter for a given entry in
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 636) * linux,drconf-usable-memory property
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 637) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 638) static inline int __init read_usm_ranges(const __be32 **usm)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 639) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 640) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 641) * For each lmb in ibm,dynamic-memory a corresponding
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 642) * entry in linux,drconf-usable-memory property contains
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 643) * a counter followed by that many (base, size) duple.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 644) * read the counter from linux,drconf-usable-memory
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 645) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 646) return read_n_cells(n_mem_size_cells, usm);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 647) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 648)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 649) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 650) * Extract NUMA information from the ibm,dynamic-reconfiguration-memory
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 651) * node. This assumes n_mem_{addr,size}_cells have been set.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 652) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 653) static int __init numa_setup_drmem_lmb(struct drmem_lmb *lmb,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 654) const __be32 **usm,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 655) void *data)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 656) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 657) unsigned int ranges, is_kexec_kdump = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 658) unsigned long base, size, sz;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 659) int nid;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 660)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 661) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 662) * Skip this block if the reserved bit is set in flags (0x80)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 663) * or if the block is not assigned to this partition (0x8)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 664) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 665) if ((lmb->flags & DRCONF_MEM_RESERVED)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 666) || !(lmb->flags & DRCONF_MEM_ASSIGNED))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 667) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 668)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 669) if (*usm)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 670) is_kexec_kdump = 1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 671)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 672) base = lmb->base_addr;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 673) size = drmem_lmb_size();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 674) ranges = 1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 675)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 676) if (is_kexec_kdump) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 677) ranges = read_usm_ranges(usm);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 678) if (!ranges) /* there are no (base, size) duple */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 679) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 680) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 681)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 682) do {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 683) if (is_kexec_kdump) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 684) base = read_n_cells(n_mem_addr_cells, usm);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 685) size = read_n_cells(n_mem_size_cells, usm);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 686) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 687)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 688) nid = of_drconf_to_nid_single(lmb);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 689) fake_numa_create_new_node(((base + size) >> PAGE_SHIFT),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 690) &nid);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 691) node_set_online(nid);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 692) sz = numa_enforce_memory_limit(base, size);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 693) if (sz)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 694) memblock_set_node(base, sz, &memblock.memory, nid);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 695) } while (--ranges);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 696)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 697) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 698) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 699)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 700) static int __init parse_numa_properties(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 701) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 702) struct device_node *memory;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 703) int default_nid = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 704) unsigned long i;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 705)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 706) if (numa_enabled == 0) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 707) printk(KERN_WARNING "NUMA disabled by user\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 708) return -1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 709) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 710)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 711) min_common_depth = find_min_common_depth();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 712)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 713) if (min_common_depth < 0) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 714) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 715) * if we fail to parse min_common_depth from device tree
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 716) * mark the numa disabled, boot with numa disabled.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 717) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 718) numa_enabled = false;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 719) return min_common_depth;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 720) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 721)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 722) dbg("NUMA associativity depth for CPU/Memory: %d\n", min_common_depth);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 723)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 724) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 725) * Even though we connect cpus to numa domains later in SMP
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 726) * init, we need to know the node ids now. This is because
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 727) * each node to be onlined must have NODE_DATA etc backing it.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 728) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 729) for_each_present_cpu(i) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 730) struct device_node *cpu;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 731) int nid = vphn_get_nid(i);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 732)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 733) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 734) * Don't fall back to default_nid yet -- we will plug
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 735) * cpus into nodes once the memory scan has discovered
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 736) * the topology.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 737) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 738) if (nid == NUMA_NO_NODE) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 739) cpu = of_get_cpu_node(i, NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 740) BUG_ON(!cpu);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 741) nid = of_node_to_nid_single(cpu);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 742) of_node_put(cpu);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 743) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 744)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 745) /* node_set_online() is an UB if 'nid' is negative */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 746) if (likely(nid >= 0))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 747) node_set_online(nid);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 748) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 749)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 750) get_n_mem_cells(&n_mem_addr_cells, &n_mem_size_cells);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 751)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 752) for_each_node_by_type(memory, "memory") {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 753) unsigned long start;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 754) unsigned long size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 755) int nid;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 756) int ranges;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 757) const __be32 *memcell_buf;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 758) unsigned int len;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 759)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 760) memcell_buf = of_get_property(memory,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 761) "linux,usable-memory", &len);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 762) if (!memcell_buf || len <= 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 763) memcell_buf = of_get_property(memory, "reg", &len);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 764) if (!memcell_buf || len <= 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 765) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 766)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 767) /* ranges in cell */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 768) ranges = (len >> 2) / (n_mem_addr_cells + n_mem_size_cells);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 769) new_range:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 770) /* these are order-sensitive, and modify the buffer pointer */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 771) start = read_n_cells(n_mem_addr_cells, &memcell_buf);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 772) size = read_n_cells(n_mem_size_cells, &memcell_buf);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 773)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 774) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 775) * Assumption: either all memory nodes or none will
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 776) * have associativity properties. If none, then
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 777) * everything goes to default_nid.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 778) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 779) nid = of_node_to_nid_single(memory);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 780) if (nid < 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 781) nid = default_nid;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 782)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 783) fake_numa_create_new_node(((start + size) >> PAGE_SHIFT), &nid);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 784) node_set_online(nid);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 785)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 786) size = numa_enforce_memory_limit(start, size);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 787) if (size)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 788) memblock_set_node(start, size, &memblock.memory, nid);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 789)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 790) if (--ranges)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 791) goto new_range;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 792) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 793)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 794) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 795) * Now do the same thing for each MEMBLOCK listed in the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 796) * ibm,dynamic-memory property in the
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 797) * ibm,dynamic-reconfiguration-memory node.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 798) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 799) memory = of_find_node_by_path("/ibm,dynamic-reconfiguration-memory");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 800) if (memory) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 801) walk_drmem_lmbs(memory, NULL, numa_setup_drmem_lmb);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 802) of_node_put(memory);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 803) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 804)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 805) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 806) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 807)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 808) static void __init setup_nonnuma(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 809) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 810) unsigned long top_of_ram = memblock_end_of_DRAM();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 811) unsigned long total_ram = memblock_phys_mem_size();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 812) unsigned long start_pfn, end_pfn;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 813) unsigned int nid = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 814) int i;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 815)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 816) printk(KERN_DEBUG "Top of RAM: 0x%lx, Total RAM: 0x%lx\n",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 817) top_of_ram, total_ram);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 818) printk(KERN_DEBUG "Memory hole size: %ldMB\n",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 819) (top_of_ram - total_ram) >> 20);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 820)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 821) for_each_mem_pfn_range(i, MAX_NUMNODES, &start_pfn, &end_pfn, NULL) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 822) fake_numa_create_new_node(end_pfn, &nid);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 823) memblock_set_node(PFN_PHYS(start_pfn),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 824) PFN_PHYS(end_pfn - start_pfn),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 825) &memblock.memory, nid);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 826) node_set_online(nid);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 827) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 828) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 829)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 830) void __init dump_numa_cpu_topology(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 831) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 832) unsigned int node;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 833) unsigned int cpu, count;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 834)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 835) if (!numa_enabled)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 836) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 837)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 838) for_each_online_node(node) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 839) pr_info("Node %d CPUs:", node);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 840)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 841) count = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 842) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 843) * If we used a CPU iterator here we would miss printing
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 844) * the holes in the cpumap.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 845) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 846) for (cpu = 0; cpu < nr_cpu_ids; cpu++) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 847) if (cpumask_test_cpu(cpu,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 848) node_to_cpumask_map[node])) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 849) if (count == 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 850) pr_cont(" %u", cpu);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 851) ++count;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 852) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 853) if (count > 1)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 854) pr_cont("-%u", cpu - 1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 855) count = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 856) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 857) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 858)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 859) if (count > 1)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 860) pr_cont("-%u", nr_cpu_ids - 1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 861) pr_cont("\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 862) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 863) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 864)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 865) /* Initialize NODE_DATA for a node on the local memory */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 866) static void __init setup_node_data(int nid, u64 start_pfn, u64 end_pfn)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 867) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 868) u64 spanned_pages = end_pfn - start_pfn;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 869) const size_t nd_size = roundup(sizeof(pg_data_t), SMP_CACHE_BYTES);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 870) u64 nd_pa;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 871) void *nd;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 872) int tnid;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 873)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 874) nd_pa = memblock_phys_alloc_try_nid(nd_size, SMP_CACHE_BYTES, nid);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 875) if (!nd_pa)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 876) panic("Cannot allocate %zu bytes for node %d data\n",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 877) nd_size, nid);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 878)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 879) nd = __va(nd_pa);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 880)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 881) /* report and initialize */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 882) pr_info(" NODE_DATA [mem %#010Lx-%#010Lx]\n",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 883) nd_pa, nd_pa + nd_size - 1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 884) tnid = early_pfn_to_nid(nd_pa >> PAGE_SHIFT);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 885) if (tnid != nid)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 886) pr_info(" NODE_DATA(%d) on node %d\n", nid, tnid);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 887)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 888) node_data[nid] = nd;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 889) memset(NODE_DATA(nid), 0, sizeof(pg_data_t));
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 890) NODE_DATA(nid)->node_id = nid;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 891) NODE_DATA(nid)->node_start_pfn = start_pfn;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 892) NODE_DATA(nid)->node_spanned_pages = spanned_pages;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 893) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 894)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 895) static void __init find_possible_nodes(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 896) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 897) struct device_node *rtas;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 898) const __be32 *domains = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 899) int prop_length, max_nodes;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 900) u32 i;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 901)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 902) if (!numa_enabled)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 903) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 904)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 905) rtas = of_find_node_by_path("/rtas");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 906) if (!rtas)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 907) return;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 908)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 909) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 910) * ibm,current-associativity-domains is a fairly recent property. If
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 911) * it doesn't exist, then fallback on ibm,max-associativity-domains.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 912) * Current denotes what the platform can support compared to max
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 913) * which denotes what the Hypervisor can support.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 914) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 915) * If the LPAR is migratable, new nodes might be activated after a LPM,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 916) * so we should consider the max number in that case.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 917) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 918) if (!of_get_property(of_root, "ibm,migratable-partition", NULL))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 919) domains = of_get_property(rtas,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 920) "ibm,current-associativity-domains",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 921) &prop_length);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 922) if (!domains) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 923) domains = of_get_property(rtas, "ibm,max-associativity-domains",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 924) &prop_length);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 925) if (!domains)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 926) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 927) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 928)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 929) max_nodes = of_read_number(&domains[min_common_depth], 1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 930) pr_info("Partition configured for %d NUMA nodes.\n", max_nodes);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 931)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 932) for (i = 0; i < max_nodes; i++) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 933) if (!node_possible(i))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 934) node_set(i, node_possible_map);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 935) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 936)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 937) prop_length /= sizeof(int);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 938) if (prop_length > min_common_depth + 2)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 939) coregroup_enabled = 1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 940)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 941) out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 942) of_node_put(rtas);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 943) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 944)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 945) void __init mem_topology_setup(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 946) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 947) int cpu;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 948)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 949) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 950) * Linux/mm assumes node 0 to be online at boot. However this is not
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 951) * true on PowerPC, where node 0 is similar to any other node, it
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 952) * could be cpuless, memoryless node. So force node 0 to be offline
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 953) * for now. This will prevent cpuless, memoryless node 0 showing up
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 954) * unnecessarily as online. If a node has cpus or memory that need
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 955) * to be online, then node will anyway be marked online.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 956) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 957) node_set_offline(0);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 958)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 959) if (parse_numa_properties())
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 960) setup_nonnuma();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 961)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 962) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 963) * Modify the set of possible NUMA nodes to reflect information
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 964) * available about the set of online nodes, and the set of nodes
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 965) * that we expect to make use of for this platform's affinity
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 966) * calculations.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 967) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 968) nodes_and(node_possible_map, node_possible_map, node_online_map);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 969)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 970) find_possible_nodes();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 971)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 972) setup_node_to_cpumask_map();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 973)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 974) reset_numa_cpu_lookup_table();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 975)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 976) for_each_possible_cpu(cpu) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 977) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 978) * Powerpc with CONFIG_NUMA always used to have a node 0,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 979) * even if it was memoryless or cpuless. For all cpus that
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 980) * are possible but not present, cpu_to_node() would point
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 981) * to node 0. To remove a cpuless, memoryless dummy node,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 982) * powerpc need to make sure all possible but not present
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 983) * cpu_to_node are set to a proper node.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 984) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 985) numa_setup_cpu(cpu);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 986) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 987) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 988)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 989) void __init initmem_init(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 990) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 991) int nid;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 992)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 993) max_low_pfn = memblock_end_of_DRAM() >> PAGE_SHIFT;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 994) max_pfn = max_low_pfn;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 995)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 996) memblock_dump_all();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 997)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 998) for_each_online_node(nid) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 999) unsigned long start_pfn, end_pfn;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1000)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1001) get_pfn_range_for_nid(nid, &start_pfn, &end_pfn);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1002) setup_node_data(nid, start_pfn, end_pfn);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1003) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1004)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1005) sparse_init();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1006)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1007) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1008) * We need the numa_cpu_lookup_table to be accurate for all CPUs,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1009) * even before we online them, so that we can use cpu_to_{node,mem}
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1010) * early in boot, cf. smp_prepare_cpus().
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1011) * _nocalls() + manual invocation is used because cpuhp is not yet
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1012) * initialized for the boot CPU.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1013) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1014) cpuhp_setup_state_nocalls(CPUHP_POWER_NUMA_PREPARE, "powerpc/numa:prepare",
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1015) ppc_numa_cpu_prepare, ppc_numa_cpu_dead);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1016) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1017)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1018) static int __init early_numa(char *p)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1019) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1020) if (!p)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1021) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1022)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1023) if (strstr(p, "off"))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1024) numa_enabled = 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1025)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1026) if (strstr(p, "debug"))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1027) numa_debug = 1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1028)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1029) p = strstr(p, "fake=");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1030) if (p)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1031) cmdline = p + strlen("fake=");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1032)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1033) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1034) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1035) early_param("numa", early_numa);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1036)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1037) #ifdef CONFIG_MEMORY_HOTPLUG
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1038) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1039) * Find the node associated with a hot added memory section for
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1040) * memory represented in the device tree by the property
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1041) * ibm,dynamic-reconfiguration-memory/ibm,dynamic-memory.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1042) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1043) static int hot_add_drconf_scn_to_nid(unsigned long scn_addr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1044) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1045) struct drmem_lmb *lmb;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1046) unsigned long lmb_size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1047) int nid = NUMA_NO_NODE;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1048)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1049) lmb_size = drmem_lmb_size();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1050)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1051) for_each_drmem_lmb(lmb) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1052) /* skip this block if it is reserved or not assigned to
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1053) * this partition */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1054) if ((lmb->flags & DRCONF_MEM_RESERVED)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1055) || !(lmb->flags & DRCONF_MEM_ASSIGNED))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1056) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1057)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1058) if ((scn_addr < lmb->base_addr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1059) || (scn_addr >= (lmb->base_addr + lmb_size)))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1060) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1061)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1062) nid = of_drconf_to_nid_single(lmb);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1063) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1064) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1065)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1066) return nid;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1067) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1068)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1069) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1070) * Find the node associated with a hot added memory section for memory
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1071) * represented in the device tree as a node (i.e. memory@XXXX) for
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1072) * each memblock.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1073) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1074) static int hot_add_node_scn_to_nid(unsigned long scn_addr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1075) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1076) struct device_node *memory;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1077) int nid = NUMA_NO_NODE;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1078)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1079) for_each_node_by_type(memory, "memory") {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1080) unsigned long start, size;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1081) int ranges;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1082) const __be32 *memcell_buf;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1083) unsigned int len;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1084)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1085) memcell_buf = of_get_property(memory, "reg", &len);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1086) if (!memcell_buf || len <= 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1087) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1088)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1089) /* ranges in cell */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1090) ranges = (len >> 2) / (n_mem_addr_cells + n_mem_size_cells);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1091)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1092) while (ranges--) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1093) start = read_n_cells(n_mem_addr_cells, &memcell_buf);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1094) size = read_n_cells(n_mem_size_cells, &memcell_buf);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1095)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1096) if ((scn_addr < start) || (scn_addr >= (start + size)))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1097) continue;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1098)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1099) nid = of_node_to_nid_single(memory);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1100) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1101) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1102)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1103) if (nid >= 0)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1104) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1105) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1106)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1107) of_node_put(memory);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1108)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1109) return nid;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1110) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1111)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1112) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1113) * Find the node associated with a hot added memory section. Section
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1114) * corresponds to a SPARSEMEM section, not an MEMBLOCK. It is assumed that
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1115) * sections are fully contained within a single MEMBLOCK.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1116) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1117) int hot_add_scn_to_nid(unsigned long scn_addr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1118) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1119) struct device_node *memory = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1120) int nid;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1121)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1122) if (!numa_enabled)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1123) return first_online_node;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1124)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1125) memory = of_find_node_by_path("/ibm,dynamic-reconfiguration-memory");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1126) if (memory) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1127) nid = hot_add_drconf_scn_to_nid(scn_addr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1128) of_node_put(memory);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1129) } else {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1130) nid = hot_add_node_scn_to_nid(scn_addr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1131) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1132)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1133) if (nid < 0 || !node_possible(nid))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1134) nid = first_online_node;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1135)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1136) return nid;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1137) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1138)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1139) static u64 hot_add_drconf_memory_max(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1140) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1141) struct device_node *memory = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1142) struct device_node *dn = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1143) const __be64 *lrdr = NULL;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1144)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1145) dn = of_find_node_by_path("/rtas");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1146) if (dn) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1147) lrdr = of_get_property(dn, "ibm,lrdr-capacity", NULL);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1148) of_node_put(dn);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1149) if (lrdr)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1150) return be64_to_cpup(lrdr);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1151) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1152)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1153) memory = of_find_node_by_path("/ibm,dynamic-reconfiguration-memory");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1154) if (memory) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1155) of_node_put(memory);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1156) return drmem_lmb_memory_max();
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1157) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1158) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1159) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1160)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1161) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1162) * memory_hotplug_max - return max address of memory that may be added
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1163) *
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1164) * This is currently only used on systems that support drconfig memory
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1165) * hotplug.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1166) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1167) u64 memory_hotplug_max(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1168) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1169) return max(hot_add_drconf_memory_max(), memblock_end_of_DRAM());
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1170) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1171) #endif /* CONFIG_MEMORY_HOTPLUG */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1172)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1173) /* Virtual Processor Home Node (VPHN) support */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1174) #ifdef CONFIG_PPC_SPLPAR
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1175) static int topology_inited;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1176)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1177) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1178) * Retrieve the new associativity information for a virtual processor's
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1179) * home node.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1180) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1181) static long vphn_get_associativity(unsigned long cpu,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1182) __be32 *associativity)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1183) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1184) long rc;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1185)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1186) rc = hcall_vphn(get_hard_smp_processor_id(cpu),
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1187) VPHN_FLAG_VCPU, associativity);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1188)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1189) switch (rc) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1190) case H_SUCCESS:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1191) dbg("VPHN hcall succeeded. Reset polling...\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1192) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1193)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1194) case H_FUNCTION:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1195) pr_err_ratelimited("VPHN unsupported. Disabling polling...\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1196) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1197) case H_HARDWARE:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1198) pr_err_ratelimited("hcall_vphn() experienced a hardware fault "
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1199) "preventing VPHN. Disabling polling...\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1200) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1201) case H_PARAMETER:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1202) pr_err_ratelimited("hcall_vphn() was passed an invalid parameter. "
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1203) "Disabling polling...\n");
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1204) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1205) default:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1206) pr_err_ratelimited("hcall_vphn() returned %ld. Disabling polling...\n"
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1207) , rc);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1208) break;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1209) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1210) out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1211) return rc;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1212) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1213)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1214) int find_and_online_cpu_nid(int cpu)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1215) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1216) __be32 associativity[VPHN_ASSOC_BUFSIZE] = {0};
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1217) int new_nid;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1218)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1219) /* Use associativity from first thread for all siblings */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1220) if (vphn_get_associativity(cpu, associativity))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1221) return cpu_to_node(cpu);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1222)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1223) new_nid = associativity_to_nid(associativity);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1224) if (new_nid < 0 || !node_possible(new_nid))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1225) new_nid = first_online_node;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1226)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1227) if (NODE_DATA(new_nid) == NULL) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1228) #ifdef CONFIG_MEMORY_HOTPLUG
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1229) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1230) * Need to ensure that NODE_DATA is initialized for a node from
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1231) * available memory (see memblock_alloc_try_nid). If unable to
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1232) * init the node, then default to nearest node that has memory
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1233) * installed. Skip onlining a node if the subsystems are not
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1234) * yet initialized.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1235) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1236) if (!topology_inited || try_online_node(new_nid))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1237) new_nid = first_online_node;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1238) #else
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1239) /*
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1240) * Default to using the nearest node that has memory installed.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1241) * Otherwise, it would be necessary to patch the kernel MM code
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1242) * to deal with more memoryless-node error conditions.
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1243) */
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1244) new_nid = first_online_node;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1245) #endif
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1246) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1247)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1248) pr_debug("%s:%d cpu %d nid %d\n", __FUNCTION__, __LINE__,
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1249) cpu, new_nid);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1250) return new_nid;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1251) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1252)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1253) int cpu_to_coregroup_id(int cpu)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1254) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1255) __be32 associativity[VPHN_ASSOC_BUFSIZE] = {0};
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1256) int index;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1257)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1258) if (cpu < 0 || cpu > nr_cpu_ids)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1259) return -1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1260)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1261) if (!coregroup_enabled)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1262) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1263)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1264) if (!firmware_has_feature(FW_FEATURE_VPHN))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1265) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1266)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1267) if (vphn_get_associativity(cpu, associativity))
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1268) goto out;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1269)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1270) index = of_read_number(associativity, 1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1271) if (index > min_common_depth + 1)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1272) return of_read_number(&associativity[index - 1], 1);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1273)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1274) out:
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1275) return cpu_to_core_id(cpu);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1276) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1277)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1278) static int topology_update_init(void)
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1279) {
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1280) topology_inited = 1;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1281) return 0;
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1282) }
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1283) device_initcall(topology_update_init);
^8f3ce5b39 (kx 2023-10-28 12:00:06 +0300 1284) #endif /* CONFIG_PPC_SPLPAR */
Orange Pi5 kernel
Deprecated Linux kernel 5.10.110 for OrangePi 5/5B/5+ boards
3 Commits
0 Branches
0 Tags