From 6470aff619fbb9dff8dfe8afa5033084cd55ca20 Mon Sep 17 00:00:00 2001 From: Brian Gerst Date: Tue, 27 Jan 2009 12:56:47 +0900 Subject: x86: move 64-bit NUMA code Impact: Code movement, no functional change. Move the 64-bit NUMA code from setup_percpu.c to numa_64.c Signed-off-by: Brian Gerst Signed-off-by: Tejun Heo --- arch/x86/mm/numa_64.c | 217 ++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 217 insertions(+) (limited to 'arch/x86/mm/numa_64.c') diff --git a/arch/x86/mm/numa_64.c b/arch/x86/mm/numa_64.c index 71a14f89f89..08d140fbc31 100644 --- a/arch/x86/mm/numa_64.c +++ b/arch/x86/mm/numa_64.c @@ -20,6 +20,12 @@ #include #include +#ifdef CONFIG_DEBUG_PER_CPU_MAPS +# define DBG(x...) printk(KERN_DEBUG x) +#else +# define DBG(x...) +#endif + struct pglist_data *node_data[MAX_NUMNODES] __read_mostly; EXPORT_SYMBOL(node_data); @@ -33,6 +39,21 @@ int numa_off __initdata; static unsigned long __initdata nodemap_addr; static unsigned long __initdata nodemap_size; +DEFINE_PER_CPU(int, node_number) = 0; +EXPORT_PER_CPU_SYMBOL(node_number); + +/* + * 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); + +/* + * Which logical CPUs are on which nodes + */ +cpumask_t *node_to_cpumask_map; +EXPORT_SYMBOL(node_to_cpumask_map); + /* * Given a shift value, try to populate memnodemap[] * Returns : @@ -640,3 +661,199 @@ void __init init_cpu_to_node(void) #endif +/* + * Allocate node_to_cpumask_map based on number of available nodes + * Requires node_possible_map to be valid. + * + * Note: node_to_cpumask() is not valid until after this is done. + * (Use CONFIG_DEBUG_PER_CPU_MAPS to check this.) + */ +void __init setup_node_to_cpumask_map(void) +{ + unsigned int node, num = 0; + cpumask_t *map; + + /* setup nr_node_ids if not done yet */ + if (nr_node_ids == MAX_NUMNODES) { + for_each_node_mask(node, node_possible_map) + num = node; + nr_node_ids = num + 1; + } + + /* allocate the map */ + map = alloc_bootmem_low(nr_node_ids * sizeof(cpumask_t)); + DBG("node_to_cpumask_map at %p for %d nodes\n", map, nr_node_ids); + + pr_debug("Node to cpumask map at %p for %d nodes\n", + map, nr_node_ids); + + /* node_to_cpumask() will now work */ + node_to_cpumask_map = 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 || !per_cpu_offset(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) + per_cpu(node_number, cpu) = node; +} + +void __cpuinit numa_clear_node(int cpu) +{ + numa_set_node(cpu, NUMA_NO_NODE); +} + +#ifndef CONFIG_DEBUG_PER_CPU_MAPS + +void __cpuinit numa_add_cpu(int cpu) +{ + cpu_set(cpu, node_to_cpumask_map[early_cpu_to_node(cpu)]); +} + +void __cpuinit numa_remove_cpu(int cpu) +{ + cpu_clear(cpu, node_to_cpumask_map[early_cpu_to_node(cpu)]); +} + +#else /* CONFIG_DEBUG_PER_CPU_MAPS */ + +/* + * --------- debug versions of the numa functions --------- + */ +static void __cpuinit numa_set_cpumask(int cpu, int enable) +{ + int node = early_cpu_to_node(cpu); + cpumask_t *mask; + char buf[64]; + + if (node_to_cpumask_map == NULL) { + printk(KERN_ERR "node_to_cpumask_map NULL\n"); + dump_stack(); + return; + } + + mask = &node_to_cpumask_map[node]; + if (enable) + cpu_set(cpu, *mask); + else + cpu_clear(cpu, *mask); + + 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); +} + +void __cpuinit numa_add_cpu(int cpu) +{ + numa_set_cpumask(cpu, 1); +} + +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 (!per_cpu_offset(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); +} + + +/* empty cpumask */ +static const cpumask_t cpu_mask_none; + +/* + * Returns a pointer to the bitmask of CPUs on Node 'node'. + */ +const cpumask_t *cpumask_of_node(int node) +{ + if (node_to_cpumask_map == NULL) { + printk(KERN_WARNING + "cpumask_of_node(%d): no node_to_cpumask_map!\n", + node); + dump_stack(); + return (const cpumask_t *)&cpu_online_map; + } + if (node >= nr_node_ids) { + printk(KERN_WARNING + "cpumask_of_node(%d): node > nr_node_ids(%d)\n", + node, nr_node_ids); + dump_stack(); + return &cpu_mask_none; + } + return &node_to_cpumask_map[node]; +} +EXPORT_SYMBOL(cpumask_of_node); + +/* + * Returns a bitmask of CPUs on Node 'node'. + * + * Side note: this function creates the returned cpumask on the stack + * so with a high NR_CPUS count, excessive stack space is used. The + * node_to_cpumask_ptr function should be used whenever possible. + */ +cpumask_t node_to_cpumask(int node) +{ + if (node_to_cpumask_map == NULL) { + printk(KERN_WARNING + "node_to_cpumask(%d): no node_to_cpumask_map!\n", node); + dump_stack(); + return cpu_online_map; + } + if (node >= nr_node_ids) { + printk(KERN_WARNING + "node_to_cpumask(%d): node > nr_node_ids(%d)\n", + node, nr_node_ids); + dump_stack(); + return cpu_mask_none; + } + return node_to_cpumask_map[node]; +} +EXPORT_SYMBOL(node_to_cpumask); + +/* + * --------- end of debug versions of the numa functions --------- + */ + +#endif /* CONFIG_DEBUG_PER_CPU_MAPS */ -- cgit v1.2.3-70-g09d2 From 44581a28e805a31661469c4b466b9cd14b36e7b6 Mon Sep 17 00:00:00 2001 From: Brian Gerst Date: Sun, 8 Feb 2009 09:58:40 -0500 Subject: x86: fix abuse of per_cpu_offset Impact: bug fix Don't use per_cpu_offset() to determine if it valid to access a per-cpu variable for a given cpu number. It is not a valid assumption on x86-64 anymore. Use cpu_possible() instead. Signed-off-by: Brian Gerst Signed-off-by: Ingo Molnar --- arch/x86/mm/numa_64.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) (limited to 'arch/x86/mm/numa_64.c') diff --git a/arch/x86/mm/numa_64.c b/arch/x86/mm/numa_64.c index 08d140fbc31..deb1c1ab786 100644 --- a/arch/x86/mm/numa_64.c +++ b/arch/x86/mm/numa_64.c @@ -702,7 +702,7 @@ void __cpuinit numa_set_node(int cpu, int node) } #ifdef CONFIG_DEBUG_PER_CPU_MAPS - if (cpu >= nr_cpu_ids || !per_cpu_offset(cpu)) { + if (cpu >= nr_cpu_ids || !cpu_possible(cpu)) { printk(KERN_ERR "numa_set_node: invalid cpu# (%d)\n", cpu); dump_stack(); return; @@ -790,7 +790,7 @@ 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 (!per_cpu_offset(cpu)) { + if (!cpu_possible(cpu)) { printk(KERN_WARNING "early_cpu_to_node(%d): no per_cpu area!\n", cpu); dump_stack(); -- cgit v1.2.3-70-g09d2 From b9c4398ed43a7ed023e091610c23ba7412aec2a8 Mon Sep 17 00:00:00 2001 From: Rusty Russell Date: Fri, 13 Mar 2009 14:49:52 +1030 Subject: cpumask: remove x86's node_to_cpumask now everyone uses cpumask_of_node Impact: cleanup Signed-off-by: Rusty Russell --- arch/x86/include/asm/topology.h | 41 ----------------------------------------- arch/x86/mm/numa_64.c | 26 -------------------------- 2 files changed, 67 deletions(-) (limited to 'arch/x86/mm/numa_64.c') diff --git a/arch/x86/include/asm/topology.h b/arch/x86/include/asm/topology.h index fa4aa42e976..216a960d4f1 100644 --- a/arch/x86/include/asm/topology.h +++ b/arch/x86/include/asm/topology.h @@ -57,17 +57,6 @@ static inline int cpu_to_node(int cpu) } #define early_cpu_to_node(cpu) cpu_to_node(cpu) -/* Returns a bitmask of CPUs on Node 'node'. - * - * Side note: this function creates the returned cpumask on the stack - * so with a high NR_CPUS count, excessive stack space is used. The - * cpumask_of_node function should be used whenever possible. - */ -static inline cpumask_t node_to_cpumask(int node) -{ - return node_to_cpumask_map[node]; -} - /* Returns a bitmask of CPUs on Node 'node'. */ static inline const struct cpumask *cpumask_of_node(int node) { @@ -92,7 +81,6 @@ DECLARE_PER_CPU(int, node_number); extern int cpu_to_node(int cpu); extern int early_cpu_to_node(int cpu); extern const cpumask_t *cpumask_of_node(int node); -extern cpumask_t node_to_cpumask(int node); #else /* !CONFIG_DEBUG_PER_CPU_MAPS */ @@ -114,26 +102,10 @@ static inline const cpumask_t *cpumask_of_node(int node) return &node_to_cpumask_map[node]; } -/* Returns a bitmask of CPUs on Node 'node'. */ -static inline cpumask_t node_to_cpumask(int node) -{ - return node_to_cpumask_map[node]; -} - #endif /* !CONFIG_DEBUG_PER_CPU_MAPS */ extern void setup_node_to_cpumask_map(void); -/* - * Replace default node_to_cpumask_ptr with optimized version - * Deprecated: use "const struct cpumask *mask = cpumask_of_node(node)" - */ -#define node_to_cpumask_ptr(v, node) \ - const cpumask_t *v = cpumask_of_node(node) - -#define node_to_cpumask_ptr_next(v, node) \ - v = cpumask_of_node(node) - #endif /* CONFIG_X86_64 */ /* @@ -212,10 +184,6 @@ static inline const cpumask_t *cpumask_of_node(int node) { return &cpu_online_map; } -static inline cpumask_t node_to_cpumask(int node) -{ - return cpu_online_map; -} static inline int node_to_first_cpu(int node) { return first_cpu(cpu_online_map); @@ -223,15 +191,6 @@ static inline int node_to_first_cpu(int node) static inline void setup_node_to_cpumask_map(void) { } -/* - * Replace default node_to_cpumask_ptr with optimized version - * Deprecated: use "const struct cpumask *mask = cpumask_of_node(node)" - */ -#define node_to_cpumask_ptr(v, node) \ - const cpumask_t *v = cpumask_of_node(node) - -#define node_to_cpumask_ptr_next(v, node) \ - v = cpumask_of_node(node) #endif #include diff --git a/arch/x86/mm/numa_64.c b/arch/x86/mm/numa_64.c index 64c9cf043cd..48bf396b6e7 100644 --- a/arch/x86/mm/numa_64.c +++ b/arch/x86/mm/numa_64.c @@ -826,32 +826,6 @@ const cpumask_t *cpumask_of_node(int node) } EXPORT_SYMBOL(cpumask_of_node); -/* - * Returns a bitmask of CPUs on Node 'node'. - * - * Side note: this function creates the returned cpumask on the stack - * so with a high NR_CPUS count, excessive stack space is used. The - * node_to_cpumask_ptr function should be used whenever possible. - */ -cpumask_t node_to_cpumask(int node) -{ - if (node_to_cpumask_map == NULL) { - printk(KERN_WARNING - "node_to_cpumask(%d): no node_to_cpumask_map!\n", node); - dump_stack(); - return cpu_online_map; - } - if (node >= nr_node_ids) { - printk(KERN_WARNING - "node_to_cpumask(%d): node > nr_node_ids(%d)\n", - node, nr_node_ids); - dump_stack(); - return cpu_mask_none; - } - return node_to_cpumask_map[node]; -} -EXPORT_SYMBOL(node_to_cpumask); - /* * --------- end of debug versions of the numa functions --------- */ -- cgit v1.2.3-70-g09d2 From 71ee73e72228775a076a502b3c92028fa59e2889 Mon Sep 17 00:00:00 2001 From: Rusty Russell Date: Fri, 13 Mar 2009 14:49:52 +1030 Subject: x86: unify 32 and 64-bit node_to_cpumask_map Impact: cleanup We take the 64-bit code and use it on 32-bit as well. The new file is called mm/numa.c. In a minor cleanup, we use cpu_none_mask instead of declaring a local cpu_mask_none. Signed-off-by: Rusty Russell --- arch/x86/include/asm/topology.h | 30 +++++++---------- arch/x86/kernel/smpboot.c | 5 --- arch/x86/mm/Makefile | 2 +- arch/x86/mm/numa.c | 71 +++++++++++++++++++++++++++++++++++++++++ arch/x86/mm/numa_64.c | 69 --------------------------------------- 5 files changed, 83 insertions(+), 94 deletions(-) create mode 100644 arch/x86/mm/numa.c (limited to 'arch/x86/mm/numa_64.c') diff --git a/arch/x86/include/asm/topology.h b/arch/x86/include/asm/topology.h index 216a960d4f1..dc31d929da0 100644 --- a/arch/x86/include/asm/topology.h +++ b/arch/x86/include/asm/topology.h @@ -44,9 +44,6 @@ #ifdef CONFIG_X86_32 -/* Mappings between node number and cpus on that node. */ -extern cpumask_t node_to_cpumask_map[]; - /* Mappings between logical cpu number and node number */ extern int cpu_to_node_map[]; @@ -57,19 +54,8 @@ static inline int cpu_to_node(int cpu) } #define early_cpu_to_node(cpu) cpu_to_node(cpu) -/* Returns a bitmask of CPUs on Node 'node'. */ -static inline const struct cpumask *cpumask_of_node(int node) -{ - return &node_to_cpumask_map[node]; -} - -static inline void setup_node_to_cpumask_map(void) { } - #else /* CONFIG_X86_64 */ -/* Mappings between node number and cpus on that node. */ -extern cpumask_t *node_to_cpumask_map; - /* Mappings between logical cpu number and node number */ DECLARE_EARLY_PER_CPU(int, x86_cpu_to_node_map); @@ -80,7 +66,6 @@ DECLARE_PER_CPU(int, node_number); #ifdef CONFIG_DEBUG_PER_CPU_MAPS extern int cpu_to_node(int cpu); extern int early_cpu_to_node(int cpu); -extern const cpumask_t *cpumask_of_node(int node); #else /* !CONFIG_DEBUG_PER_CPU_MAPS */ @@ -96,18 +81,25 @@ static inline int early_cpu_to_node(int cpu) return early_per_cpu(x86_cpu_to_node_map, cpu); } +#endif /* !CONFIG_DEBUG_PER_CPU_MAPS */ + +#endif /* CONFIG_X86_64 */ + +/* Mappings between node number and cpus on that node. */ +extern cpumask_t *node_to_cpumask_map; + +#ifdef CONFIG_DEBUG_PER_CPU_MAPS +extern const cpumask_t *cpumask_of_node(int node); +#else /* Returns a pointer to the cpumask of CPUs on Node 'node'. */ static inline const cpumask_t *cpumask_of_node(int node) { return &node_to_cpumask_map[node]; } - -#endif /* !CONFIG_DEBUG_PER_CPU_MAPS */ +#endif extern void setup_node_to_cpumask_map(void); -#endif /* CONFIG_X86_64 */ - /* * Returns the number of the node containing Node 'node'. This * architecture is flat, so it is a pretty simple function! diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c index 7f051c170ad..c55639b1ab8 100644 --- a/arch/x86/kernel/smpboot.c +++ b/arch/x86/kernel/smpboot.c @@ -115,11 +115,6 @@ EXPORT_PER_CPU_SYMBOL(cpu_info); atomic_t init_deasserted; #if defined(CONFIG_NUMA) && defined(CONFIG_X86_32) - -/* which logical CPUs are on which nodes */ -cpumask_t node_to_cpumask_map[MAX_NUMNODES] __read_mostly = - { [0 ... MAX_NUMNODES-1] = CPU_MASK_NONE }; -EXPORT_SYMBOL(node_to_cpumask_map); /* which node each logical CPU is on */ int cpu_to_node_map[NR_CPUS] __read_mostly = { [0 ... NR_CPUS-1] = 0 }; EXPORT_SYMBOL(cpu_to_node_map); diff --git a/arch/x86/mm/Makefile b/arch/x86/mm/Makefile index 08537747cb5..fdd30d08ab5 100644 --- a/arch/x86/mm/Makefile +++ b/arch/x86/mm/Makefile @@ -14,7 +14,7 @@ obj-$(CONFIG_MMIOTRACE) += mmiotrace.o mmiotrace-y := kmmio.o pf_in.o mmio-mod.o obj-$(CONFIG_MMIOTRACE_TEST) += testmmiotrace.o -obj-$(CONFIG_NUMA) += numa_$(BITS).o +obj-$(CONFIG_NUMA) += numa.o numa_$(BITS).o obj-$(CONFIG_K8_NUMA) += k8topology_64.o obj-$(CONFIG_ACPI_NUMA) += srat_$(BITS).o diff --git a/arch/x86/mm/numa.c b/arch/x86/mm/numa.c new file mode 100644 index 00000000000..f3a19e939e8 --- /dev/null +++ b/arch/x86/mm/numa.c @@ -0,0 +1,71 @@ +/* Common code for 32 and 64-bit NUMA */ +#include +#include +#include + +#ifdef CONFIG_DEBUG_PER_CPU_MAPS +# define DBG(x...) printk(KERN_DEBUG x) +#else +# define DBG(x...) +#endif + +/* + * Which logical CPUs are on which nodes + */ +cpumask_t *node_to_cpumask_map; +EXPORT_SYMBOL(node_to_cpumask_map); + +/* + * Allocate node_to_cpumask_map based on number of available nodes + * Requires node_possible_map to be valid. + * + * Note: node_to_cpumask() is not valid until after this is done. + * (Use CONFIG_DEBUG_PER_CPU_MAPS to check this.) + */ +void __init setup_node_to_cpumask_map(void) +{ + unsigned int node, num = 0; + cpumask_t *map; + + /* setup nr_node_ids if not done yet */ + if (nr_node_ids == MAX_NUMNODES) { + for_each_node_mask(node, node_possible_map) + num = node; + nr_node_ids = num + 1; + } + + /* allocate the map */ + map = alloc_bootmem_low(nr_node_ids * sizeof(cpumask_t)); + DBG("node_to_cpumask_map at %p for %d nodes\n", map, nr_node_ids); + + pr_debug("Node to cpumask map at %p for %d nodes\n", + map, nr_node_ids); + + /* node_to_cpumask() will now work */ + node_to_cpumask_map = map; +} + +#ifdef CONFIG_DEBUG_PER_CPU_MAPS +/* + * Returns a pointer to the bitmask of CPUs on Node 'node'. + */ +const cpumask_t *cpumask_of_node(int node) +{ + if (node_to_cpumask_map == NULL) { + printk(KERN_WARNING + "cpumask_of_node(%d): no node_to_cpumask_map!\n", + node); + dump_stack(); + return cpu_online_mask; + } + if (node >= nr_node_ids) { + printk(KERN_WARNING + "cpumask_of_node(%d): node > nr_node_ids(%d)\n", + node, nr_node_ids); + dump_stack(); + return cpu_none_mask; + } + return &node_to_cpumask_map[node]; +} +EXPORT_SYMBOL(cpumask_of_node); +#endif diff --git a/arch/x86/mm/numa_64.c b/arch/x86/mm/numa_64.c index 48bf396b6e7..eee14907886 100644 --- a/arch/x86/mm/numa_64.c +++ b/arch/x86/mm/numa_64.c @@ -20,12 +20,6 @@ #include #include -#ifdef CONFIG_DEBUG_PER_CPU_MAPS -# define DBG(x...) printk(KERN_DEBUG x) -#else -# define DBG(x...) -#endif - struct pglist_data *node_data[MAX_NUMNODES] __read_mostly; EXPORT_SYMBOL(node_data); @@ -48,12 +42,6 @@ EXPORT_PER_CPU_SYMBOL(node_number); DEFINE_EARLY_PER_CPU(int, x86_cpu_to_node_map, NUMA_NO_NODE); EXPORT_EARLY_PER_CPU_SYMBOL(x86_cpu_to_node_map); -/* - * Which logical CPUs are on which nodes - */ -cpumask_t *node_to_cpumask_map; -EXPORT_SYMBOL(node_to_cpumask_map); - /* * Given a shift value, try to populate memnodemap[] * Returns : @@ -661,36 +649,6 @@ void __init init_cpu_to_node(void) #endif -/* - * Allocate node_to_cpumask_map based on number of available nodes - * Requires node_possible_map to be valid. - * - * Note: node_to_cpumask() is not valid until after this is done. - * (Use CONFIG_DEBUG_PER_CPU_MAPS to check this.) - */ -void __init setup_node_to_cpumask_map(void) -{ - unsigned int node, num = 0; - cpumask_t *map; - - /* setup nr_node_ids if not done yet */ - if (nr_node_ids == MAX_NUMNODES) { - for_each_node_mask(node, node_possible_map) - num = node; - nr_node_ids = num + 1; - } - - /* allocate the map */ - map = alloc_bootmem_low(nr_node_ids * sizeof(cpumask_t)); - DBG("node_to_cpumask_map at %p for %d nodes\n", map, nr_node_ids); - - pr_debug("Node to cpumask map at %p for %d nodes\n", - map, nr_node_ids); - - /* node_to_cpumask() will now work */ - node_to_cpumask_map = map; -} - void __cpuinit numa_set_node(int cpu, int node) { int *cpu_to_node_map = early_per_cpu_ptr(x86_cpu_to_node_map); @@ -799,33 +757,6 @@ int early_cpu_to_node(int cpu) return per_cpu(x86_cpu_to_node_map, cpu); } - -/* empty cpumask */ -static const cpumask_t cpu_mask_none; - -/* - * Returns a pointer to the bitmask of CPUs on Node 'node'. - */ -const cpumask_t *cpumask_of_node(int node) -{ - if (node_to_cpumask_map == NULL) { - printk(KERN_WARNING - "cpumask_of_node(%d): no node_to_cpumask_map!\n", - node); - dump_stack(); - return (const cpumask_t *)&cpu_online_map; - } - if (node >= nr_node_ids) { - printk(KERN_WARNING - "cpumask_of_node(%d): node > nr_node_ids(%d)\n", - node, nr_node_ids); - dump_stack(); - return &cpu_mask_none; - } - return &node_to_cpumask_map[node]; -} -EXPORT_SYMBOL(cpumask_of_node); - /* * --------- end of debug versions of the numa functions --------- */ -- cgit v1.2.3-70-g09d2 From c032ef60d1aa9af33730b7a35bbea751b131adc1 Mon Sep 17 00:00:00 2001 From: Rusty Russell Date: Fri, 13 Mar 2009 14:49:53 +1030 Subject: cpumask: convert node_to_cpumask_map[] to cpumask_var_t Impact: reduce kernel memory usage when CONFIG_CPUMASK_OFFSTACK=y Straightforward conversion: done for 32 and 64 bit kernels. node_to_cpumask_map is now a cpumask_var_t array. 64-bit used to be a dynamic cpumask_t array, and 32-bit used to be a static cpumask_t array. Signed-off-by: Rusty Russell --- arch/x86/include/asm/topology.h | 6 +++--- arch/x86/kernel/smpboot.c | 4 ++-- arch/x86/mm/numa.c | 28 ++++++++++++---------------- arch/x86/mm/numa_64.c | 14 +++++++------- 4 files changed, 24 insertions(+), 28 deletions(-) (limited to 'arch/x86/mm/numa_64.c') diff --git a/arch/x86/include/asm/topology.h b/arch/x86/include/asm/topology.h index dc31d929da0..f8b833e1257 100644 --- a/arch/x86/include/asm/topology.h +++ b/arch/x86/include/asm/topology.h @@ -86,15 +86,15 @@ static inline int early_cpu_to_node(int cpu) #endif /* CONFIG_X86_64 */ /* Mappings between node number and cpus on that node. */ -extern cpumask_t *node_to_cpumask_map; +extern cpumask_var_t node_to_cpumask_map[MAX_NUMNODES]; #ifdef CONFIG_DEBUG_PER_CPU_MAPS extern const cpumask_t *cpumask_of_node(int node); #else /* Returns a pointer to the cpumask of CPUs on Node 'node'. */ -static inline const cpumask_t *cpumask_of_node(int node) +static inline const struct cpumask *cpumask_of_node(int node) { - return &node_to_cpumask_map[node]; + return node_to_cpumask_map[node]; } #endif diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c index c55639b1ab8..5a58a45ac1e 100644 --- a/arch/x86/kernel/smpboot.c +++ b/arch/x86/kernel/smpboot.c @@ -123,7 +123,7 @@ EXPORT_SYMBOL(cpu_to_node_map); static void map_cpu_to_node(int cpu, int node) { printk(KERN_INFO "Mapping cpu %d to node %d\n", cpu, node); - cpumask_set_cpu(cpu, &node_to_cpumask_map[node]); + cpumask_set_cpu(cpu, node_to_cpumask_map[node]); cpu_to_node_map[cpu] = node; } @@ -134,7 +134,7 @@ static void unmap_cpu_to_node(int cpu) printk(KERN_INFO "Unmapping cpu %d from all nodes\n", cpu); for (node = 0; node < MAX_NUMNODES; node++) - cpumask_clear_cpu(cpu, &node_to_cpumask_map[node]); + cpumask_clear_cpu(cpu, node_to_cpumask_map[node]); cpu_to_node_map[cpu] = 0; } #else /* !(CONFIG_NUMA && CONFIG_X86_32) */ diff --git a/arch/x86/mm/numa.c b/arch/x86/mm/numa.c index f3a19e939e8..429dc2d191f 100644 --- a/arch/x86/mm/numa.c +++ b/arch/x86/mm/numa.c @@ -12,7 +12,7 @@ /* * Which logical CPUs are on which nodes */ -cpumask_t *node_to_cpumask_map; +cpumask_var_t node_to_cpumask_map[MAX_NUMNODES]; EXPORT_SYMBOL(node_to_cpumask_map); /* @@ -25,7 +25,6 @@ EXPORT_SYMBOL(node_to_cpumask_map); void __init setup_node_to_cpumask_map(void) { unsigned int node, num = 0; - cpumask_t *map; /* setup nr_node_ids if not done yet */ if (nr_node_ids == MAX_NUMNODES) { @@ -35,14 +34,11 @@ void __init setup_node_to_cpumask_map(void) } /* allocate the map */ - map = alloc_bootmem_low(nr_node_ids * sizeof(cpumask_t)); - DBG("node_to_cpumask_map at %p for %d nodes\n", map, nr_node_ids); + for (node = 0; node < nr_node_ids; node++) + alloc_bootmem_cpumask_var(&node_to_cpumask_map[node]); - pr_debug("Node to cpumask map at %p for %d nodes\n", - map, nr_node_ids); - - /* node_to_cpumask() will now work */ - node_to_cpumask_map = map; + /* cpumask_of_node() will now work */ + pr_debug("Node to cpumask map for %d nodes\n", nr_node_ids); } #ifdef CONFIG_DEBUG_PER_CPU_MAPS @@ -51,13 +47,6 @@ void __init setup_node_to_cpumask_map(void) */ const cpumask_t *cpumask_of_node(int node) { - if (node_to_cpumask_map == NULL) { - printk(KERN_WARNING - "cpumask_of_node(%d): no node_to_cpumask_map!\n", - node); - dump_stack(); - return cpu_online_mask; - } if (node >= nr_node_ids) { printk(KERN_WARNING "cpumask_of_node(%d): node > nr_node_ids(%d)\n", @@ -65,6 +54,13 @@ const cpumask_t *cpumask_of_node(int node) dump_stack(); return cpu_none_mask; } + if (node_to_cpumask_map[node] == NULL) { + printk(KERN_WARNING + "cpumask_of_node(%d): no node_to_cpumask_map!\n", + node); + dump_stack(); + return cpu_online_mask; + } return &node_to_cpumask_map[node]; } EXPORT_SYMBOL(cpumask_of_node); diff --git a/arch/x86/mm/numa_64.c b/arch/x86/mm/numa_64.c index eee14907886..9d2b3d2625c 100644 --- a/arch/x86/mm/numa_64.c +++ b/arch/x86/mm/numa_64.c @@ -681,12 +681,12 @@ void __cpuinit numa_clear_node(int cpu) void __cpuinit numa_add_cpu(int cpu) { - cpu_set(cpu, node_to_cpumask_map[early_cpu_to_node(cpu)]); + cpumask_set_cpu(cpu, node_to_cpumask_map[early_cpu_to_node(cpu)]); } void __cpuinit numa_remove_cpu(int cpu) { - cpu_clear(cpu, node_to_cpumask_map[early_cpu_to_node(cpu)]); + cpumask_clear_cpu(cpu, node_to_cpumask_map[early_cpu_to_node(cpu)]); } #else /* CONFIG_DEBUG_PER_CPU_MAPS */ @@ -700,17 +700,17 @@ static void __cpuinit numa_set_cpumask(int cpu, int enable) cpumask_t *mask; char buf[64]; - if (node_to_cpumask_map == NULL) { - printk(KERN_ERR "node_to_cpumask_map NULL\n"); + mask = node_to_cpumask_map[node]; + if (mask == NULL) { + printk(KERN_ERR "node_to_cpumask_map[%i] NULL\n", node); dump_stack(); return; } - mask = &node_to_cpumask_map[node]; if (enable) - cpu_set(cpu, *mask); + cpumask_set_cpu(cpu, mask); else - cpu_clear(cpu, *mask); + cpumask_clear_cpu(cpu, mask); cpulist_scnprintf(buf, sizeof(buf), mask); printk(KERN_DEBUG "%s cpu %d node %d: mask now %s\n", -- cgit v1.2.3-70-g09d2 From 73e907de7d5cecef43d9949ab8f4fdca508168c7 Mon Sep 17 00:00:00 2001 From: Rusty Russell Date: Fri, 13 Mar 2009 14:49:57 +1030 Subject: cpumask: remove x86 cpumask_t uses. Impact: cleanup We are removing cpumask_t in favour of struct cpumask: mainly as a marker of what code is now CONFIG_CPUMASK_OFFSTACK-safe. The only non-trivial change here is vector_allocation_domain(): explicitly clear the mask and set the first word, rather than using assignment. Signed-off-by: Rusty Russell --- arch/x86/include/asm/topology.h | 4 ++-- arch/x86/kernel/apic/es7000_32.c | 6 +++--- arch/x86/kernel/apic/numaq_32.c | 6 +++--- arch/x86/kernel/apic/summit_32.c | 6 +++--- arch/x86/mm/numa.c | 2 +- arch/x86/mm/numa_64.c | 2 +- 6 files changed, 13 insertions(+), 13 deletions(-) (limited to 'arch/x86/mm/numa_64.c') diff --git a/arch/x86/include/asm/topology.h b/arch/x86/include/asm/topology.h index 1ce1e1afa80..e3f4198371a 100644 --- a/arch/x86/include/asm/topology.h +++ b/arch/x86/include/asm/topology.h @@ -89,7 +89,7 @@ static inline int early_cpu_to_node(int cpu) extern cpumask_var_t node_to_cpumask_map[MAX_NUMNODES]; #ifdef CONFIG_DEBUG_PER_CPU_MAPS -extern const cpumask_t *cpumask_of_node(int node); +extern const struct cpumask *cpumask_of_node(int node); #else /* Returns a pointer to the cpumask of CPUs on Node 'node'. */ static inline const struct cpumask *cpumask_of_node(int node) @@ -172,7 +172,7 @@ static inline int early_cpu_to_node(int cpu) return 0; } -static inline const cpumask_t *cpumask_of_node(int node) +static inline const struct cpumask *cpumask_of_node(int node) { return cpu_online_mask; } diff --git a/arch/x86/kernel/apic/es7000_32.c b/arch/x86/kernel/apic/es7000_32.c index 12c8e19ef96..1c11b819f24 100644 --- a/arch/x86/kernel/apic/es7000_32.c +++ b/arch/x86/kernel/apic/es7000_32.c @@ -410,7 +410,7 @@ static void es7000_enable_apic_mode(void) WARN(1, "Command failed, status = %x\n", mip_status); } -static void es7000_vector_allocation_domain(int cpu, cpumask_t *retmask) +static void es7000_vector_allocation_domain(int cpu, struct cpumask *retmask) { /* Careful. Some cpus do not strictly honor the set of cpus * specified in the interrupt destination when using lowest @@ -456,7 +456,7 @@ static int es7000_apic_id_registered(void) return 1; } -static const cpumask_t *target_cpus_cluster(void) +static const struct cpumask *target_cpus_cluster(void) { return cpu_all_mask; } @@ -573,7 +573,7 @@ static int es7000_check_phys_apicid_present(int cpu_physical_apicid) return 1; } -static unsigned int es7000_cpu_mask_to_apicid(const cpumask_t *cpumask) +static unsigned int es7000_cpu_mask_to_apicid(const struct cpumask *cpumask) { unsigned int round = 0; int cpu, uninitialized_var(apicid); diff --git a/arch/x86/kernel/apic/numaq_32.c b/arch/x86/kernel/apic/numaq_32.c index 9562de1b888..533e59c6fc8 100644 --- a/arch/x86/kernel/apic/numaq_32.c +++ b/arch/x86/kernel/apic/numaq_32.c @@ -334,7 +334,7 @@ static inline void numaq_smp_callin_clear_local_apic(void) clear_local_APIC(); } -static inline const cpumask_t *numaq_target_cpus(void) +static inline const struct cpumask *numaq_target_cpus(void) { return cpu_all_mask; } @@ -427,7 +427,7 @@ static inline int numaq_check_phys_apicid_present(int boot_cpu_physical_apicid) * We use physical apicids here, not logical, so just return the default * physical broadcast to stop people from breaking us */ -static inline unsigned int numaq_cpu_mask_to_apicid(const cpumask_t *cpumask) +static unsigned int numaq_cpu_mask_to_apicid(const struct cpumask *cpumask) { return 0x0F; } @@ -462,7 +462,7 @@ static int probe_numaq(void) return found_numaq; } -static void numaq_vector_allocation_domain(int cpu, cpumask_t *retmask) +static void numaq_vector_allocation_domain(int cpu, struct cpumask *retmask) { /* Careful. Some cpus do not strictly honor the set of cpus * specified in the interrupt destination when using lowest diff --git a/arch/x86/kernel/apic/summit_32.c b/arch/x86/kernel/apic/summit_32.c index 278e863c71b..9cfe1f415d8 100644 --- a/arch/x86/kernel/apic/summit_32.c +++ b/arch/x86/kernel/apic/summit_32.c @@ -182,7 +182,7 @@ static inline int is_WPEG(struct rio_detail *rio){ #define SUMMIT_APIC_DFR_VALUE (APIC_DFR_CLUSTER) -static const cpumask_t *summit_target_cpus(void) +static const struct cpumask *summit_target_cpus(void) { /* CPU_MASK_ALL (0xff) has undefined behaviour with * dest_LowestPrio mode logical clustered apic interrupt routing @@ -285,7 +285,7 @@ static int summit_check_phys_apicid_present(int boot_cpu_physical_apicid) return 1; } -static unsigned int summit_cpu_mask_to_apicid(const cpumask_t *cpumask) +static unsigned int summit_cpu_mask_to_apicid(const struct cpumask *cpumask) { unsigned int round = 0; int cpu, apicid = 0; @@ -342,7 +342,7 @@ static int probe_summit(void) return 0; } -static void summit_vector_allocation_domain(int cpu, cpumask_t *retmask) +static void summit_vector_allocation_domain(int cpu, struct cpumask *retmask) { /* Careful. Some cpus do not strictly honor the set of cpus * specified in the interrupt destination when using lowest diff --git a/arch/x86/mm/numa.c b/arch/x86/mm/numa.c index 429dc2d191f..ce255e32a59 100644 --- a/arch/x86/mm/numa.c +++ b/arch/x86/mm/numa.c @@ -45,7 +45,7 @@ void __init setup_node_to_cpumask_map(void) /* * Returns a pointer to the bitmask of CPUs on Node 'node'. */ -const cpumask_t *cpumask_of_node(int node) +const struct cpumask *cpumask_of_node(int node) { if (node >= nr_node_ids) { printk(KERN_WARNING diff --git a/arch/x86/mm/numa_64.c b/arch/x86/mm/numa_64.c index 9d2b3d2625c..d73aaa89237 100644 --- a/arch/x86/mm/numa_64.c +++ b/arch/x86/mm/numa_64.c @@ -697,7 +697,7 @@ void __cpuinit numa_remove_cpu(int cpu) static void __cpuinit numa_set_cpumask(int cpu, int enable) { int node = early_cpu_to_node(cpu); - cpumask_t *mask; + struct cpumask *mask; char buf[64]; mask = node_to_cpumask_map[node]; -- cgit v1.2.3-70-g09d2