diff options
author | Ingo Molnar <mingo@elte.hu> | 2011-02-16 09:44:04 +0100 |
---|---|---|
committer | Ingo Molnar <mingo@elte.hu> | 2011-02-16 09:44:15 +0100 |
commit | 52b8b8d7251f8f7b8ed4a6c623618106d83e18b5 (patch) | |
tree | 13ee81932cba73deb69cd3f386d80e45e6ac23de /arch/x86/mm | |
parent | 02ac81a812fe0575a8475a93bdc22fb291ebad91 (diff) | |
parent | 14392fd329eca9b59d51c0aa5d0acfb4965424d1 (diff) |
Merge branch 'x86/numa' into x86/mm
Merge reason: consolidate it into the more generic x86/mm tree to prevent conflicts.
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Diffstat (limited to 'arch/x86/mm')
-rw-r--r-- | arch/x86/mm/amdtopology_64.c | 4 | ||||
-rw-r--r-- | arch/x86/mm/numa.c | 212 | ||||
-rw-r--r-- | arch/x86/mm/numa_32.c | 7 | ||||
-rw-r--r-- | arch/x86/mm/numa_64.c | 231 | ||||
-rw-r--r-- | arch/x86/mm/srat_32.c | 6 | ||||
-rw-r--r-- | arch/x86/mm/srat_64.c | 12 |
6 files changed, 250 insertions, 222 deletions
diff --git a/arch/x86/mm/amdtopology_64.c b/arch/x86/mm/amdtopology_64.c index 49b334cdd64..2523c3554de 100644 --- a/arch/x86/mm/amdtopology_64.c +++ b/arch/x86/mm/amdtopology_64.c @@ -247,7 +247,7 @@ void __init amd_fake_nodes(const struct bootnode *nodes, int nr_nodes) __acpi_map_pxm_to_node(nid, i); #endif } - memcpy(apicid_to_node, fake_apicid_to_node, sizeof(apicid_to_node)); + memcpy(__apicid_to_node, fake_apicid_to_node, sizeof(__apicid_to_node)); } #endif /* CONFIG_NUMA_EMU */ @@ -287,7 +287,7 @@ int __init amd_scan_nodes(void) int j; for (j = apicid_base; j < cores + apicid_base; j++) - apicid_to_node[(i << bits) + j] = i; + set_apicid_to_node((i << bits) + j, i); setup_node_bootmem(i, nodes[i].start, nodes[i].end); } diff --git a/arch/x86/mm/numa.c b/arch/x86/mm/numa.c index ebf6d7887a3..9559d360fde 100644 --- a/arch/x86/mm/numa.c +++ b/arch/x86/mm/numa.c @@ -26,12 +26,50 @@ static __init int numa_setup(char *opt) early_param("numa", numa_setup); /* - * Which logical CPUs are on which nodes + * apicid, cpu, node mappings */ +s16 __apicid_to_node[MAX_LOCAL_APIC] __cpuinitdata = { + [0 ... MAX_LOCAL_APIC-1] = NUMA_NO_NODE +}; + cpumask_var_t node_to_cpumask_map[MAX_NUMNODES]; EXPORT_SYMBOL(node_to_cpumask_map); /* + * Map cpu index to node index + */ +DEFINE_EARLY_PER_CPU(int, x86_cpu_to_node_map, NUMA_NO_NODE); +EXPORT_EARLY_PER_CPU_SYMBOL(x86_cpu_to_node_map); + +void __cpuinit numa_set_node(int cpu, int node) +{ + int *cpu_to_node_map = early_per_cpu_ptr(x86_cpu_to_node_map); + + /* early setting, no percpu area yet */ + if (cpu_to_node_map) { + cpu_to_node_map[cpu] = node; + return; + } + +#ifdef CONFIG_DEBUG_PER_CPU_MAPS + if (cpu >= nr_cpu_ids || !cpu_possible(cpu)) { + printk(KERN_ERR "numa_set_node: invalid cpu# (%d)\n", cpu); + dump_stack(); + return; + } +#endif + per_cpu(x86_cpu_to_node_map, cpu) = node; + + if (node != NUMA_NO_NODE) + set_cpu_numa_node(cpu, node); +} + +void __cpuinit numa_clear_node(int cpu) +{ + numa_set_node(cpu, NUMA_NO_NODE); +} + +/* * Allocate node_to_cpumask_map based on number of available nodes * Requires node_possible_map to be valid. * @@ -57,7 +95,174 @@ void __init setup_node_to_cpumask_map(void) pr_debug("Node to cpumask map for %d nodes\n", nr_node_ids); } -#ifdef CONFIG_DEBUG_PER_CPU_MAPS +/* + * There are unfortunately some poorly designed mainboards around that + * only connect memory to a single CPU. This breaks the 1:1 cpu->node + * mapping. To avoid this fill in the mapping for all possible CPUs, + * as the number of CPUs is not known yet. We round robin the existing + * nodes. + */ +void __init numa_init_array(void) +{ + int rr, i; + + rr = first_node(node_online_map); + for (i = 0; i < nr_cpu_ids; i++) { + if (early_cpu_to_node(i) != NUMA_NO_NODE) + continue; + numa_set_node(i, rr); + rr = next_node(rr, node_online_map); + if (rr == MAX_NUMNODES) + rr = first_node(node_online_map); + } +} + +static __init int find_near_online_node(int node) +{ + int n, val; + int min_val = INT_MAX; + int best_node = -1; + + for_each_online_node(n) { + val = node_distance(node, n); + + if (val < min_val) { + min_val = val; + best_node = n; + } + } + + return best_node; +} + +/* + * Setup early cpu_to_node. + * + * Populate cpu_to_node[] only if x86_cpu_to_apicid[], + * and apicid_to_node[] tables have valid entries for a CPU. + * This means we skip cpu_to_node[] initialisation for NUMA + * emulation and faking node case (when running a kernel compiled + * for NUMA on a non NUMA box), which is OK as cpu_to_node[] + * is already initialized in a round robin manner at numa_init_array, + * prior to this call, and this initialization is good enough + * for the fake NUMA cases. + * + * Called before the per_cpu areas are setup. + */ +void __init init_cpu_to_node(void) +{ + int cpu; + u16 *cpu_to_apicid = early_per_cpu_ptr(x86_cpu_to_apicid); + + BUG_ON(cpu_to_apicid == NULL); + + for_each_possible_cpu(cpu) { + int node = numa_cpu_node(cpu); + + if (node == NUMA_NO_NODE) + continue; + if (!node_online(node)) + node = find_near_online_node(node); + numa_set_node(cpu, node); + } +} + +#ifndef CONFIG_DEBUG_PER_CPU_MAPS + +# ifndef CONFIG_NUMA_EMU +void __cpuinit numa_add_cpu(int cpu) +{ + cpumask_set_cpu(cpu, node_to_cpumask_map[early_cpu_to_node(cpu)]); +} + +void __cpuinit numa_remove_cpu(int cpu) +{ + cpumask_clear_cpu(cpu, node_to_cpumask_map[early_cpu_to_node(cpu)]); +} +# endif /* !CONFIG_NUMA_EMU */ + +#else /* !CONFIG_DEBUG_PER_CPU_MAPS */ + +int __cpu_to_node(int cpu) +{ + if (early_per_cpu_ptr(x86_cpu_to_node_map)) { + printk(KERN_WARNING + "cpu_to_node(%d): usage too early!\n", cpu); + dump_stack(); + return early_per_cpu_ptr(x86_cpu_to_node_map)[cpu]; + } + return per_cpu(x86_cpu_to_node_map, cpu); +} +EXPORT_SYMBOL(__cpu_to_node); + +/* + * Same function as cpu_to_node() but used if called before the + * per_cpu areas are setup. + */ +int early_cpu_to_node(int cpu) +{ + if (early_per_cpu_ptr(x86_cpu_to_node_map)) + return early_per_cpu_ptr(x86_cpu_to_node_map)[cpu]; + + if (!cpu_possible(cpu)) { + printk(KERN_WARNING + "early_cpu_to_node(%d): no per_cpu area!\n", cpu); + dump_stack(); + return NUMA_NO_NODE; + } + return per_cpu(x86_cpu_to_node_map, cpu); +} + +struct cpumask __cpuinit *debug_cpumask_set_cpu(int cpu, int enable) +{ + int node = early_cpu_to_node(cpu); + struct cpumask *mask; + char buf[64]; + + if (node == NUMA_NO_NODE) { + /* early_cpu_to_node() already emits a warning and trace */ + return NULL; + } + mask = node_to_cpumask_map[node]; + if (!mask) { + pr_err("node_to_cpumask_map[%i] NULL\n", node); + dump_stack(); + return NULL; + } + + cpulist_scnprintf(buf, sizeof(buf), mask); + printk(KERN_DEBUG "%s cpu %d node %d: mask now %s\n", + enable ? "numa_add_cpu" : "numa_remove_cpu", + cpu, node, buf); + return mask; +} + +# ifndef CONFIG_NUMA_EMU +static void __cpuinit numa_set_cpumask(int cpu, int enable) +{ + struct cpumask *mask; + + mask = debug_cpumask_set_cpu(cpu, enable); + if (!mask) + return; + + if (enable) + cpumask_set_cpu(cpu, mask); + else + cpumask_clear_cpu(cpu, mask); +} + +void __cpuinit numa_add_cpu(int cpu) +{ + numa_set_cpumask(cpu, 1); +} + +void __cpuinit numa_remove_cpu(int cpu) +{ + numa_set_cpumask(cpu, 0); +} +# endif /* !CONFIG_NUMA_EMU */ + /* * Returns a pointer to the bitmask of CPUs on Node 'node'. */ @@ -80,4 +285,5 @@ const struct cpumask *cpumask_of_node(int node) return node_to_cpumask_map[node]; } EXPORT_SYMBOL(cpumask_of_node); -#endif + +#endif /* !CONFIG_DEBUG_PER_CPU_MAPS */ diff --git a/arch/x86/mm/numa_32.c b/arch/x86/mm/numa_32.c index 84a3e4c9f27..505bb04654b 100644 --- a/arch/x86/mm/numa_32.c +++ b/arch/x86/mm/numa_32.c @@ -110,6 +110,12 @@ void set_pmd_pfn(unsigned long vaddr, unsigned long pfn, pgprot_t flags); static unsigned long kva_start_pfn; static unsigned long kva_pages; + +int __cpuinit numa_cpu_node(int cpu) +{ + return apic->x86_32_numa_cpu_node(cpu); +} + /* * FLAT - support for basic PC memory model with discontig enabled, essentially * a single node with all available processors in it with a flat @@ -361,6 +367,7 @@ void __init initmem_init(unsigned long start_pfn, unsigned long end_pfn, */ get_memcfg_numa(); + numa_init_array(); kva_pages = roundup(calculate_numa_remap_pages(), PTRS_PER_PTE); diff --git a/arch/x86/mm/numa_64.c b/arch/x86/mm/numa_64.c index 62cb634b5cf..43ad3273561 100644 --- a/arch/x86/mm/numa_64.c +++ b/arch/x86/mm/numa_64.c @@ -26,20 +26,10 @@ EXPORT_SYMBOL(node_data); struct memnode memnode; -s16 apicid_to_node[MAX_LOCAL_APIC] __cpuinitdata = { - [0 ... MAX_LOCAL_APIC-1] = NUMA_NO_NODE -}; - static unsigned long __initdata nodemap_addr; static unsigned long __initdata nodemap_size; /* - * Map cpu index to node index - */ -DEFINE_EARLY_PER_CPU(int, x86_cpu_to_node_map, NUMA_NO_NODE); -EXPORT_EARLY_PER_CPU_SYMBOL(x86_cpu_to_node_map); - -/* * Given a shift value, try to populate memnodemap[] * Returns : * 1 if OK @@ -234,28 +224,6 @@ setup_node_bootmem(int nodeid, unsigned long start, unsigned long end) node_set_online(nodeid); } -/* - * There are unfortunately some poorly designed mainboards around that - * only connect memory to a single CPU. This breaks the 1:1 cpu->node - * mapping. To avoid this fill in the mapping for all possible CPUs, - * as the number of CPUs is not known yet. We round robin the existing - * nodes. - */ -void __init numa_init_array(void) -{ - int rr, i; - - rr = first_node(node_online_map); - for (i = 0; i < nr_cpu_ids; i++) { - if (early_cpu_to_node(i) != NUMA_NO_NODE) - continue; - numa_set_node(i, rr); - rr = next_node(rr, node_online_map); - if (rr == MAX_NUMNODES) - rr = first_node(node_online_map); - } -} - #ifdef CONFIG_NUMA_EMU /* Numa emulation */ static struct bootnode nodes[MAX_NUMNODES] __initdata; @@ -676,115 +644,33 @@ unsigned long __init numa_free_all_bootmem(void) return pages; } -#ifdef CONFIG_NUMA - -static __init int find_near_online_node(int node) +int __cpuinit numa_cpu_node(int cpu) { - int n, val; - int min_val = INT_MAX; - int best_node = -1; - - for_each_online_node(n) { - val = node_distance(node, n); - - if (val < min_val) { - min_val = val; - best_node = n; - } - } + int apicid = early_per_cpu(x86_cpu_to_apicid, cpu); - return best_node; + if (apicid != BAD_APICID) + return __apicid_to_node[apicid]; + return NUMA_NO_NODE; } /* - * Setup early cpu_to_node. + * UGLINESS AHEAD: Currently, CONFIG_NUMA_EMU is 64bit only and makes use + * of 64bit specific data structures. The distinction is artificial and + * should be removed. numa_{add|remove}_cpu() are implemented in numa.c + * for both 32 and 64bit when CONFIG_NUMA_EMU is disabled but here when + * enabled. * - * Populate cpu_to_node[] only if x86_cpu_to_apicid[], - * and apicid_to_node[] tables have valid entries for a CPU. - * This means we skip cpu_to_node[] initialisation for NUMA - * emulation and faking node case (when running a kernel compiled - * for NUMA on a non NUMA box), which is OK as cpu_to_node[] - * is already initialized in a round robin manner at numa_init_array, - * prior to this call, and this initialization is good enough - * for the fake NUMA cases. - * - * Called before the per_cpu areas are setup. + * NUMA emulation is planned to be made generic and the following and other + * related code should be moved to numa.c. */ -void __init init_cpu_to_node(void) -{ - int cpu; - u16 *cpu_to_apicid = early_per_cpu_ptr(x86_cpu_to_apicid); - - BUG_ON(cpu_to_apicid == NULL); - - for_each_possible_cpu(cpu) { - int node; - u16 apicid = cpu_to_apicid[cpu]; - - if (apicid == BAD_APICID) - continue; - node = apicid_to_node[apicid]; - if (node == NUMA_NO_NODE) - continue; - if (!node_online(node)) - node = find_near_online_node(node); - numa_set_node(cpu, node); - } -} -#endif - - -void __cpuinit numa_set_node(int cpu, int node) -{ - int *cpu_to_node_map = early_per_cpu_ptr(x86_cpu_to_node_map); - - /* early setting, no percpu area yet */ - if (cpu_to_node_map) { - cpu_to_node_map[cpu] = node; - return; - } - -#ifdef CONFIG_DEBUG_PER_CPU_MAPS - if (cpu >= nr_cpu_ids || !cpu_possible(cpu)) { - printk(KERN_ERR "numa_set_node: invalid cpu# (%d)\n", cpu); - dump_stack(); - return; - } -#endif - per_cpu(x86_cpu_to_node_map, cpu) = node; - - if (node != NUMA_NO_NODE) - set_cpu_numa_node(cpu, node); -} - -void __cpuinit numa_clear_node(int cpu) -{ - numa_set_node(cpu, NUMA_NO_NODE); -} - -#ifndef CONFIG_DEBUG_PER_CPU_MAPS - -#ifndef CONFIG_NUMA_EMU -void __cpuinit numa_add_cpu(int cpu) -{ - cpumask_set_cpu(cpu, node_to_cpumask_map[early_cpu_to_node(cpu)]); -} - -void __cpuinit numa_remove_cpu(int cpu) -{ - cpumask_clear_cpu(cpu, node_to_cpumask_map[early_cpu_to_node(cpu)]); -} -#else +#ifdef CONFIG_NUMA_EMU +# ifndef CONFIG_DEBUG_PER_CPU_MAPS void __cpuinit numa_add_cpu(int cpu) { unsigned long addr; - u16 apicid; - int physnid; - int nid = NUMA_NO_NODE; + int physnid, nid; - apicid = early_per_cpu(x86_cpu_to_apicid, cpu); - if (apicid != BAD_APICID) - nid = apicid_to_node[apicid]; + nid = numa_cpu_node(cpu); if (nid == NUMA_NO_NODE) nid = early_cpu_to_node(cpu); BUG_ON(nid == NUMA_NO_NODE || !node_online(nid)); @@ -818,53 +704,17 @@ void __cpuinit numa_remove_cpu(int cpu) for_each_online_node(i) cpumask_clear_cpu(cpu, node_to_cpumask_map[i]); } -#endif /* !CONFIG_NUMA_EMU */ - -#else /* CONFIG_DEBUG_PER_CPU_MAPS */ -static struct cpumask __cpuinit *debug_cpumask_set_cpu(int cpu, int enable) -{ - int node = early_cpu_to_node(cpu); - struct cpumask *mask; - char buf[64]; - - mask = node_to_cpumask_map[node]; - if (!mask) { - pr_err("node_to_cpumask_map[%i] NULL\n", node); - dump_stack(); - return NULL; - } - - cpulist_scnprintf(buf, sizeof(buf), mask); - printk(KERN_DEBUG "%s cpu %d node %d: mask now %s\n", - enable ? "numa_add_cpu" : "numa_remove_cpu", - cpu, node, buf); - return mask; -} - -/* - * --------- debug versions of the numa functions --------- - */ -#ifndef CONFIG_NUMA_EMU -static void __cpuinit numa_set_cpumask(int cpu, int enable) -{ - struct cpumask *mask; - - mask = debug_cpumask_set_cpu(cpu, enable); - if (!mask) - return; - - if (enable) - cpumask_set_cpu(cpu, mask); - else - cpumask_clear_cpu(cpu, mask); -} -#else +# else /* !CONFIG_DEBUG_PER_CPU_MAPS */ static void __cpuinit numa_set_cpumask(int cpu, int enable) { int node = early_cpu_to_node(cpu); struct cpumask *mask; int i; + if (node == NUMA_NO_NODE) { + /* early_cpu_to_node() already emits a warning and trace */ + return; + } for_each_online_node(i) { unsigned long addr; @@ -882,7 +732,6 @@ static void __cpuinit numa_set_cpumask(int cpu, int enable) cpumask_clear_cpu(cpu, mask); } } -#endif /* CONFIG_NUMA_EMU */ void __cpuinit numa_add_cpu(int cpu) { @@ -893,39 +742,5 @@ void __cpuinit numa_remove_cpu(int cpu) { numa_set_cpumask(cpu, 0); } - -int __cpu_to_node(int cpu) -{ - if (early_per_cpu_ptr(x86_cpu_to_node_map)) { - printk(KERN_WARNING - "cpu_to_node(%d): usage too early!\n", cpu); - dump_stack(); - return early_per_cpu_ptr(x86_cpu_to_node_map)[cpu]; - } - return per_cpu(x86_cpu_to_node_map, cpu); -} -EXPORT_SYMBOL(__cpu_to_node); - -/* - * Same function as cpu_to_node() but used if called before the - * per_cpu areas are setup. - */ -int early_cpu_to_node(int cpu) -{ - if (early_per_cpu_ptr(x86_cpu_to_node_map)) - return early_per_cpu_ptr(x86_cpu_to_node_map)[cpu]; - - if (!cpu_possible(cpu)) { - printk(KERN_WARNING - "early_cpu_to_node(%d): no per_cpu area!\n", cpu); - dump_stack(); - return NUMA_NO_NODE; - } - return per_cpu(x86_cpu_to_node_map, cpu); -} - -/* - * --------- end of debug versions of the numa functions --------- - */ - -#endif /* CONFIG_DEBUG_PER_CPU_MAPS */ +# endif /* !CONFIG_DEBUG_PER_CPU_MAPS */ +#endif /* CONFIG_NUMA_EMU */ diff --git a/arch/x86/mm/srat_32.c b/arch/x86/mm/srat_32.c index ae96e7b8051..48651c6f657 100644 --- a/arch/x86/mm/srat_32.c +++ b/arch/x86/mm/srat_32.c @@ -57,7 +57,7 @@ struct node_memory_chunk_s { static struct node_memory_chunk_s __initdata node_memory_chunk[MAXCHUNKS]; static int __initdata num_memory_chunks; /* total number of memory chunks */ -static u8 __initdata apicid_to_pxm[MAX_APICID]; +static u8 __initdata apicid_to_pxm[MAX_LOCAL_APIC]; int acpi_numa __initdata; @@ -254,8 +254,8 @@ int __init get_memcfg_from_srat(void) printk(KERN_DEBUG "Number of memory chunks in system = %d\n", num_memory_chunks); - for (i = 0; i < MAX_APICID; i++) - apicid_2_node[i] = pxm_to_node(apicid_to_pxm[i]); + for (i = 0; i < MAX_LOCAL_APIC; i++) + set_apicid_to_node(i, pxm_to_node(apicid_to_pxm[i])); for (j = 0; j < num_memory_chunks; j++){ struct node_memory_chunk_s * chunk = &node_memory_chunk[j]; diff --git a/arch/x86/mm/srat_64.c b/arch/x86/mm/srat_64.c index 4c03e13da13..23498f8b09a 100644 --- a/arch/x86/mm/srat_64.c +++ b/arch/x86/mm/srat_64.c @@ -79,7 +79,7 @@ static __init void bad_srat(void) printk(KERN_ERR "SRAT: SRAT not used.\n"); acpi_numa = -1; for (i = 0; i < MAX_LOCAL_APIC; i++) - apicid_to_node[i] = NUMA_NO_NODE; + set_apicid_to_node(i, NUMA_NO_NODE); for (i = 0; i < MAX_NUMNODES; i++) { nodes[i].start = nodes[i].end = 0; nodes_add[i].start = nodes_add[i].end = 0; @@ -138,7 +138,7 @@ acpi_numa_x2apic_affinity_init(struct acpi_srat_x2apic_cpu_affinity *pa) printk(KERN_INFO "SRAT: PXM %u -> APIC 0x%04x -> Node %u skipped apicid that is too big\n", pxm, apic_id, node); return; } - apicid_to_node[apic_id] = node; + set_apicid_to_node(apic_id, node); node_set(node, cpu_nodes_parsed); acpi_numa = 1; printk(KERN_INFO "SRAT: PXM %u -> APIC 0x%04x -> Node %u\n", @@ -178,7 +178,7 @@ acpi_numa_processor_affinity_init(struct acpi_srat_cpu_affinity *pa) return; } - apicid_to_node[apic_id] = node; + set_apicid_to_node(apic_id, node); node_set(node, cpu_nodes_parsed); acpi_numa = 1; printk(KERN_INFO "SRAT: PXM %u -> APIC 0x%02x -> Node %u\n", @@ -523,7 +523,7 @@ void __init acpi_fake_nodes(const struct bootnode *fake_nodes, int num_nodes) * node, it must now point to the fake node ID. */ for (j = 0; j < MAX_LOCAL_APIC; j++) - if (apicid_to_node[j] == nid && + if (__apicid_to_node[j] == nid && fake_apicid_to_node[j] == NUMA_NO_NODE) fake_apicid_to_node[j] = i; } @@ -534,13 +534,13 @@ void __init acpi_fake_nodes(const struct bootnode *fake_nodes, int num_nodes) * value. */ for (i = 0; i < MAX_LOCAL_APIC; i++) - if (apicid_to_node[i] != NUMA_NO_NODE && + if (__apicid_to_node[i] != NUMA_NO_NODE && fake_apicid_to_node[i] == NUMA_NO_NODE) fake_apicid_to_node[i] = 0; for (i = 0; i < num_nodes; i++) __acpi_map_pxm_to_node(fake_node_to_pxm_map[i], i); - memcpy(apicid_to_node, fake_apicid_to_node, sizeof(apicid_to_node)); + memcpy(__apicid_to_node, fake_apicid_to_node, sizeof(__apicid_to_node)); nodes_clear(nodes_parsed); for (i = 0; i < num_nodes; i++) |