diff options
author | Linus Torvalds <torvalds@linux-foundation.org> | 2011-03-16 02:49:10 (GMT) |
---|---|---|
committer | Linus Torvalds <torvalds@linux-foundation.org> | 2011-03-16 02:49:10 (GMT) |
commit | 181f977d134a9f8e3f8839f42af655b045fc059e (patch) | |
tree | 5d9bb67c62ef1476c18ed350106a84c02f0dd8e4 | |
parent | d5d42399bd7b66bd6b55363b311810504110c967 (diff) | |
parent | 25542c646afbf14c43fa7d2b443055cadb73b07a (diff) | |
download | linux-181f977d134a9f8e3f8839f42af655b045fc059e.tar.xz |
Merge branch 'x86-mm-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip
* 'x86-mm-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip: (93 commits)
x86, tlb, UV: Do small micro-optimization for native_flush_tlb_others()
x86-64, NUMA: Don't call numa_set_distanc() for all possible node combinations during emulation
x86-64, NUMA: Don't assume phys node 0 is always online in numa_emulation()
x86-64, NUMA: Clean up initmem_init()
x86-64, NUMA: Fix numa_emulation code with node0 without RAM
x86-64, NUMA: Revert NUMA affine page table allocation
x86: Work around old gas bug
x86-64, NUMA: Better explain numa_distance handling
x86-64, NUMA: Fix distance table handling
mm: Move early_node_map[] reverse scan helpers under HAVE_MEMBLOCK
x86-64, NUMA: Fix size of numa_distance array
x86: Rename e820_table_* to pgt_buf_*
bootmem: Move __alloc_memory_core_early() to nobootmem.c
bootmem: Move contig_page_data definition to bootmem.c/nobootmem.c
bootmem: Separate out CONFIG_NO_BOOTMEM code into nobootmem.c
x86-64, NUMA: Seperate out numa_alloc_distance() from numa_set_distance()
x86-64, NUMA: Add proper function comments to global functions
x86-64, NUMA: Move NUMA emulation into numa_emulation.c
x86-64, NUMA: Prepare numa_emulation() for moving NUMA emulation into a separate file
x86-64, NUMA: Do not scan two times for setup_node_bootmem()
...
Fix up conflicts in arch/x86/kernel/smpboot.c
63 files changed, 2306 insertions, 1796 deletions
diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index a42660c..159c2ff 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -1709,7 +1709,7 @@ config HAVE_ARCH_EARLY_PFN_TO_NID depends on NUMA config USE_PERCPU_NUMA_NODE_ID - def_bool X86_64 + def_bool y depends on NUMA menu "Power management and ACPI options" diff --git a/arch/x86/include/asm/acpi.h b/arch/x86/include/asm/acpi.h index 4ea15ca..b964ec4 100644 --- a/arch/x86/include/asm/acpi.h +++ b/arch/x86/include/asm/acpi.h @@ -186,15 +186,7 @@ struct bootnode; #ifdef CONFIG_ACPI_NUMA extern int acpi_numa; -extern void acpi_get_nodes(struct bootnode *physnodes, unsigned long start, - unsigned long end); -extern int acpi_scan_nodes(unsigned long start, unsigned long end); -#define NR_NODE_MEMBLKS (MAX_NUMNODES*2) - -#ifdef CONFIG_NUMA_EMU -extern void acpi_fake_nodes(const struct bootnode *fake_nodes, - int num_nodes); -#endif +extern int x86_acpi_numa_init(void); #endif /* CONFIG_ACPI_NUMA */ #define acpi_unlazy_tlb(x) leave_mm(x) diff --git a/arch/x86/include/asm/amd_nb.h b/arch/x86/include/asm/amd_nb.h index 64dc82e..e264ae5 100644 --- a/arch/x86/include/asm/amd_nb.h +++ b/arch/x86/include/asm/amd_nb.h @@ -9,23 +9,20 @@ struct amd_nb_bus_dev_range { u8 dev_limit; }; -extern struct pci_device_id amd_nb_misc_ids[]; +extern const struct pci_device_id amd_nb_misc_ids[]; extern const struct amd_nb_bus_dev_range amd_nb_bus_dev_ranges[]; struct bootnode; extern int early_is_amd_nb(u32 value); extern int amd_cache_northbridges(void); extern void amd_flush_garts(void); -extern int amd_numa_init(unsigned long start_pfn, unsigned long end_pfn); -extern int amd_scan_nodes(void); - -#ifdef CONFIG_NUMA_EMU -extern void amd_fake_nodes(const struct bootnode *nodes, int nr_nodes); -extern void amd_get_nodes(struct bootnode *nodes); -#endif +extern int amd_numa_init(void); +extern int amd_get_subcaches(int); +extern int amd_set_subcaches(int, int); struct amd_northbridge { struct pci_dev *misc; + struct pci_dev *link; }; struct amd_northbridge_info { @@ -37,6 +34,7 @@ extern struct amd_northbridge_info amd_northbridges; #define AMD_NB_GART 0x1 #define AMD_NB_L3_INDEX_DISABLE 0x2 +#define AMD_NB_L3_PARTITIONING 0x4 #ifdef CONFIG_AMD_NB diff --git a/arch/x86/include/asm/apic.h b/arch/x86/include/asm/apic.h index 3c89694..b8a3484 100644 --- a/arch/x86/include/asm/apic.h +++ b/arch/x86/include/asm/apic.h @@ -307,8 +307,6 @@ struct apic { void (*setup_apic_routing)(void); int (*multi_timer_check)(int apic, int irq); - int (*apicid_to_node)(int logical_apicid); - int (*cpu_to_logical_apicid)(int cpu); int (*cpu_present_to_apicid)(int mps_cpu); void (*apicid_to_cpu_present)(int phys_apicid, physid_mask_t *retmap); void (*setup_portio_remap)(void); @@ -356,6 +354,23 @@ struct apic { void (*icr_write)(u32 low, u32 high); void (*wait_icr_idle)(void); u32 (*safe_wait_icr_idle)(void); + +#ifdef CONFIG_X86_32 + /* + * Called very early during boot from get_smp_config(). It should + * return the logical apicid. x86_[bios]_cpu_to_apicid is + * initialized before this function is called. + * + * If logical apicid can't be determined that early, the function + * may return BAD_APICID. Logical apicid will be configured after + * init_apic_ldr() while bringing up CPUs. Note that NUMA affinity + * won't be applied properly during early boot in this case. + */ + int (*x86_32_early_logical_apicid)(int cpu); + + /* determine CPU -> NUMA node mapping */ + int (*x86_32_numa_cpu_node)(int cpu); +#endif }; /* @@ -503,6 +518,11 @@ extern struct apic apic_noop; extern struct apic apic_default; +static inline int noop_x86_32_early_logical_apicid(int cpu) +{ + return BAD_APICID; +} + /* * Set up the logical destination ID. * @@ -522,7 +542,7 @@ static inline int default_phys_pkg_id(int cpuid_apic, int index_msb) return cpuid_apic >> index_msb; } -extern int default_apicid_to_node(int logical_apicid); +extern int default_x86_32_numa_cpu_node(int cpu); #endif @@ -558,12 +578,6 @@ static inline void default_ioapic_phys_id_map(physid_mask_t *phys_map, physid_ma *retmap = *phys_map; } -/* Mapping from cpu number to logical apicid */ -static inline int default_cpu_to_logical_apicid(int cpu) -{ - return 1 << cpu; -} - static inline int __default_cpu_present_to_apicid(int mps_cpu) { if (mps_cpu < nr_cpu_ids && cpu_present(mps_cpu)) @@ -596,8 +610,4 @@ extern int default_check_phys_apicid_present(int phys_apicid); #endif /* CONFIG_X86_LOCAL_APIC */ -#ifdef CONFIG_X86_32 -extern u8 cpu_2_logical_apicid[NR_CPUS]; -#endif - #endif /* _ASM_X86_APIC_H */ diff --git a/arch/x86/include/asm/entry_arch.h b/arch/x86/include/asm/entry_arch.h index 57650ab..1cd6d26 100644 --- a/arch/x86/include/asm/entry_arch.h +++ b/arch/x86/include/asm/entry_arch.h @@ -16,10 +16,13 @@ BUILD_INTERRUPT(call_function_single_interrupt,CALL_FUNCTION_SINGLE_VECTOR) BUILD_INTERRUPT(irq_move_cleanup_interrupt,IRQ_MOVE_CLEANUP_VECTOR) BUILD_INTERRUPT(reboot_interrupt,REBOOT_VECTOR) -.irpc idx, "01234567" +.irp idx,0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15, \ + 16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31 +.if NUM_INVALIDATE_TLB_VECTORS > \idx BUILD_INTERRUPT3(invalidate_interrupt\idx, (INVALIDATE_TLB_VECTOR_START)+\idx, smp_invalidate_interrupt) +.endif .endr #endif diff --git a/arch/x86/include/asm/hw_irq.h b/arch/x86/include/asm/hw_irq.h index 0274ec5..bb9efe8 100644 --- a/arch/x86/include/asm/hw_irq.h +++ b/arch/x86/include/asm/hw_irq.h @@ -45,6 +45,30 @@ extern void invalidate_interrupt4(void); extern void invalidate_interrupt5(void); extern void invalidate_interrupt6(void); extern void invalidate_interrupt7(void); +extern void invalidate_interrupt8(void); +extern void invalidate_interrupt9(void); +extern void invalidate_interrupt10(void); +extern void invalidate_interrupt11(void); +extern void invalidate_interrupt12(void); +extern void invalidate_interrupt13(void); +extern void invalidate_interrupt14(void); +extern void invalidate_interrupt15(void); +extern void invalidate_interrupt16(void); +extern void invalidate_interrupt17(void); +extern void invalidate_interrupt18(void); +extern void invalidate_interrupt19(void); +extern void invalidate_interrupt20(void); +extern void invalidate_interrupt21(void); +extern void invalidate_interrupt22(void); +extern void invalidate_interrupt23(void); +extern void invalidate_interrupt24(void); +extern void invalidate_interrupt25(void); +extern void invalidate_interrupt26(void); +extern void invalidate_interrupt27(void); +extern void invalidate_interrupt28(void); +extern void invalidate_interrupt29(void); +extern void invalidate_interrupt30(void); +extern void invalidate_interrupt31(void); extern void irq_move_cleanup_interrupt(void); extern void reboot_interrupt(void); diff --git a/arch/x86/include/asm/init.h b/arch/x86/include/asm/init.h index 36fb1a6a..8dbe353 100644 --- a/arch/x86/include/asm/init.h +++ b/arch/x86/include/asm/init.h @@ -11,8 +11,8 @@ kernel_physical_mapping_init(unsigned long start, unsigned long page_size_mask); -extern unsigned long __initdata e820_table_start; -extern unsigned long __meminitdata e820_table_end; -extern unsigned long __meminitdata e820_table_top; +extern unsigned long __initdata pgt_buf_start; +extern unsigned long __meminitdata pgt_buf_end; +extern unsigned long __meminitdata pgt_buf_top; #endif /* _ASM_X86_INIT_32_H */ diff --git a/arch/x86/include/asm/ipi.h b/arch/x86/include/asm/ipi.h index 0b72282..615fa90 100644 --- a/arch/x86/include/asm/ipi.h +++ b/arch/x86/include/asm/ipi.h @@ -123,10 +123,6 @@ extern void default_send_IPI_mask_sequence_phys(const struct cpumask *mask, int vector); extern void default_send_IPI_mask_allbutself_phys(const struct cpumask *mask, int vector); -extern void default_send_IPI_mask_sequence_logical(const struct cpumask *mask, - int vector); -extern void default_send_IPI_mask_allbutself_logical(const struct cpumask *mask, - int vector); /* Avoid include hell */ #define NMI_VECTOR 0x02 @@ -150,6 +146,10 @@ static inline void __default_local_send_IPI_all(int vector) } #ifdef CONFIG_X86_32 +extern void default_send_IPI_mask_sequence_logical(const struct cpumask *mask, + int vector); +extern void default_send_IPI_mask_allbutself_logical(const struct cpumask *mask, + int vector); extern void default_send_IPI_mask_logical(const struct cpumask *mask, int vector); extern void default_send_IPI_allbutself(int vector); diff --git a/arch/x86/include/asm/irq_vectors.h b/arch/x86/include/asm/irq_vectors.h index 6af0894..6e976ee 100644 --- a/arch/x86/include/asm/irq_vectors.h +++ b/arch/x86/include/asm/irq_vectors.h @@ -1,6 +1,7 @@ #ifndef _ASM_X86_IRQ_VECTORS_H #define _ASM_X86_IRQ_VECTORS_H +#include <linux/threads.h> /* * Linux IRQ vector layout. * @@ -16,8 +17,8 @@ * Vectors 0 ... 31 : system traps and exceptions - hardcoded events * Vectors 32 ... 127 : device interrupts * Vector 128 : legacy int80 syscall interface - * Vectors 129 ... 237 : device interrupts - * Vectors 238 ... 255 : special interrupts + * Vectors 129 ... INVALIDATE_TLB_VECTOR_START-1 : device interrupts + * Vectors INVALIDATE_TLB_VECTOR_START ... 255 : special interrupts * * 64-bit x86 has per CPU IDT tables, 32-bit has one shared IDT table. * @@ -96,37 +97,43 @@ #define THRESHOLD_APIC_VECTOR 0xf9 #define REBOOT_VECTOR 0xf8 -/* f0-f7 used for spreading out TLB flushes: */ -#define INVALIDATE_TLB_VECTOR_END 0xf7 -#define INVALIDATE_TLB_VECTOR_START 0xf0 -#define NUM_INVALIDATE_TLB_VECTORS 8 - -/* - * Local APIC timer IRQ vector is on a different priority level, - * to work around the 'lost local interrupt if more than 2 IRQ - * sources per level' errata. - */ -#define LOCAL_TIMER_VECTOR 0xef - /* * Generic system vector for platform specific use */ -#define X86_PLATFORM_IPI_VECTOR 0xed +#define X86_PLATFORM_IPI_VECTOR 0xf7 /* * IRQ work vector: */ -#define IRQ_WORK_VECTOR 0xec +#define IRQ_WORK_VECTOR 0xf6 -#define UV_BAU_MESSAGE 0xea +#define UV_BAU_MESSAGE 0xf5 /* * Self IPI vector for machine checks */ -#define MCE_SELF_VECTOR 0xeb +#define MCE_SELF_VECTOR 0xf4 /* Xen vector callback to receive events in a HVM domain */ -#define XEN_HVM_EVTCHN_CALLBACK 0xe9 +#define XEN_HVM_EVTCHN_CALLBACK 0xf3 + +/* + * Local APIC timer IRQ vector is on a different priority level, + * to work around the 'lost local interrupt if more than 2 IRQ + * sources per level' errata. + */ +#define LOCAL_TIMER_VECTOR 0xef + +/* up to 32 vectors used for spreading out TLB flushes: */ +#if NR_CPUS <= 32 +# define NUM_INVALIDATE_TLB_VECTORS (NR_CPUS) +#else +# define NUM_INVALIDATE_TLB_VECTORS (32) +#endif + +#define INVALIDATE_TLB_VECTOR_END (0xee) +#define INVALIDATE_TLB_VECTOR_START \ + (INVALIDATE_TLB_VECTOR_END-NUM_INVALIDATE_TLB_VECTORS+1) #define NR_VECTORS 256 diff --git a/arch/x86/include/asm/mpspec.h b/arch/x86/include/asm/mpspec.h index 0c90dd9..9c7d95f 100644 --- a/arch/x86/include/asm/mpspec.h +++ b/arch/x86/include/asm/mpspec.h @@ -25,7 +25,6 @@ extern int pic_mode; #define MAX_IRQ_SOURCES 256 extern unsigned int def_to_bigsmp; -extern u8 apicid_2_node[]; #ifdef CONFIG_X86_NUMAQ extern int mp_bus_id_to_node[MAX_MP_BUSSES]; @@ -33,8 +32,6 @@ extern int mp_bus_id_to_local[MAX_MP_BUSSES]; extern int quad_local_to_mp_bus_id [NR_CPUS/4][4]; #endif -#define MAX_APICID 256 - #else /* CONFIG_X86_64: */ #define MAX_MP_BUSSES 256 diff --git a/arch/x86/include/asm/numa.h b/arch/x86/include/asm/numa.h index 27da400..3d4dab4 100644 --- a/arch/x86/include/asm/numa.h +++ b/arch/x86/include/asm/numa.h @@ -1,5 +1,57 @@ +#ifndef _ASM_X86_NUMA_H +#define _ASM_X86_NUMA_H + +#include <asm/topology.h> +#include <asm/apicdef.h> + +#ifdef CONFIG_NUMA + +#define NR_NODE_MEMBLKS (MAX_NUMNODES*2) + +/* + * __apicid_to_node[] stores the raw mapping between physical apicid and + * node and is used to initialize cpu_to_node mapping. + * + * The mapping may be overridden by apic->numa_cpu_node() on 32bit and thus + * should be accessed by the accessors - set_apicid_to_node() and + * numa_cpu_node(). + */ +extern s16 __apicid_to_node[MAX_LOCAL_APIC]; + +static inline void set_apicid_to_node(int apicid, s16 node) +{ + __apicid_to_node[apicid] = node; +} +#else /* CONFIG_NUMA */ +static inline void set_apicid_to_node(int apicid, s16 node) +{ +} +#endif /* CONFIG_NUMA */ + #ifdef CONFIG_X86_32 # include "numa_32.h" #else # include "numa_64.h" #endif + +#ifdef CONFIG_NUMA +extern void __cpuinit numa_set_node(int cpu, int node); +extern void __cpuinit numa_clear_node(int cpu); +extern void __init numa_init_array(void); +extern void __init init_cpu_to_node(void); +extern void __cpuinit numa_add_cpu(int cpu); +extern void __cpuinit numa_remove_cpu(int cpu); +#else /* CONFIG_NUMA */ +static inline void numa_set_node(int cpu, int node) { } +static inline void numa_clear_node(int cpu) { } +static inline void numa_init_array(void) { } +static inline void init_cpu_to_node(void) { } +static inline void numa_add_cpu(int cpu) { } +static inline void numa_remove_cpu(int cpu) { } +#endif /* CONFIG_NUMA */ + +#ifdef CONFIG_DEBUG_PER_CPU_MAPS +struct cpumask __cpuinit *debug_cpumask_set_cpu(int cpu, int enable); +#endif + +#endif /* _ASM_X86_NUMA_H */ diff --git a/arch/x86/include/asm/numa_32.h b/arch/x86/include/asm/numa_32.h index b0ef2b4..c6beed1 100644 --- a/arch/x86/include/asm/numa_32.h +++ b/arch/x86/include/asm/numa_32.h @@ -4,7 +4,12 @@ extern int numa_off; extern int pxm_to_nid(int pxm); -extern void numa_remove_cpu(int cpu); + +#ifdef CONFIG_NUMA +extern int __cpuinit numa_cpu_node(int cpu); +#else /* CONFIG_NUMA */ +static inline int numa_cpu_node(int cpu) { return NUMA_NO_NODE; } +#endif /* CONFIG_NUMA */ #ifdef CONFIG_HIGHMEM extern void set_highmem_pages_init(void); diff --git a/arch/x86/include/asm/numa_64.h b/arch/x86/include/asm/numa_64.h index 0493be3..344eb17 100644 --- a/arch/x86/include/asm/numa_64.h +++ b/arch/x86/include/asm/numa_64.h @@ -2,23 +2,16 @@ #define _ASM_X86_NUMA_64_H #include <linux/nodemask.h> -#include <asm/apicdef.h> struct bootnode { u64 start; u64 end; }; -extern int compute_hash_shift(struct bootnode *nodes, int numblks, - int *nodeids); - #define ZONE_ALIGN (1UL << (MAX_ORDER+PAGE_SHIFT)) -extern void numa_init_array(void); extern int numa_off; -extern s16 apicid_to_node[MAX_LOCAL_APIC]; - extern unsigned long numa_free_all_bootmem(void); extern void setup_node_bootmem(int nodeid, unsigned long start, unsigned long end); @@ -31,11 +24,11 @@ extern void setup_node_bootmem(int nodeid, unsigned long start, */ #define NODE_MIN_SIZE (4*1024*1024) -extern void __init init_cpu_to_node(void); -extern void __cpuinit numa_set_node(int cpu, int node); -extern void __cpuinit numa_clear_node(int cpu); -extern void __cpuinit numa_add_cpu(int cpu); -extern void __cpuinit numa_remove_cpu(int cpu); +extern nodemask_t numa_nodes_parsed __initdata; + +extern int __cpuinit numa_cpu_node(int cpu); +extern int __init numa_add_memblk(int nodeid, u64 start, u64 end); +extern void __init numa_set_distance(int from, int to, int distance); #ifdef CONFIG_NUMA_EMU #define FAKE_NODE_MIN_SIZE ((u64)32 << 20) @@ -43,11 +36,7 @@ extern void __cpuinit numa_remove_cpu(int cpu); void numa_emu_cmdline(char *); #endif /* CONFIG_NUMA_EMU */ #else -static inline void init_cpu_to_node(void) { } -static inline void numa_set_node(int cpu, int node) { } -static inline void numa_clear_node(int cpu) { } -static inline void numa_add_cpu(int cpu, int node) { } -static inline void numa_remove_cpu(int cpu) { } +static inline int numa_cpu_node(int cpu) { return NUMA_NO_NODE; } #endif #endif /* _ASM_X86_NUMA_64_H */ diff --git a/arch/x86/include/asm/page_types.h b/arch/x86/include/asm/page_types.h index 1df6621..bce688d 100644 --- a/arch/x86/include/asm/page_types.h +++ b/arch/x86/include/asm/page_types.h @@ -2,6 +2,7 @@ #define _ASM_X86_PAGE_DEFS_H #include <linux/const.h> +#include <linux/types.h> /* PAGE_SHIFT determines the page size */ #define PAGE_SHIFT 12 @@ -45,11 +46,15 @@ extern int devmem_is_allowed(unsigned long pagenr); extern unsigned long max_low_pfn_mapped; extern unsigned long max_pfn_mapped; +static inline phys_addr_t get_max_mapped(void) +{ + return (phys_addr_t)max_pfn_mapped << PAGE_SHIFT; +} + extern unsigned long init_memory_mapping(unsigned long start, unsigned long end); -extern void initmem_init(unsigned long start_pfn, unsigned long end_pfn, - int acpi, int k8); +extern void initmem_init(void); extern void free_initmem(void); #endif /* !__ASSEMBLY__ */ diff --git a/arch/x86/include/asm/smp.h b/arch/x86/include/asm/smp.h index 99fa8b4..73b11bc 100644 --- a/arch/x86/include/asm/smp.h +++ b/arch/x86/include/asm/smp.h @@ -55,6 +55,9 @@ static inline struct cpumask *cpu_llc_shared_mask(int cpu) DECLARE_EARLY_PER_CPU(u16, x86_cpu_to_apicid); DECLARE_EARLY_PER_CPU(u16, x86_bios_cpu_apicid); +#if defined(CONFIG_X86_LOCAL_APIC) && defined(CONFIG_X86_32) +DECLARE_EARLY_PER_CPU(int, x86_cpu_to_logical_apicid); +#endif /* Static state in head.S used to set up a CPU */ extern unsigned long stack_start; /* Initial stack pointer address */ diff --git a/arch/x86/include/asm/topology.h b/arch/x86/include/asm/topology.h index 21899cc3..910a708 100644 --- a/arch/x86/include/asm/topology.h +++ b/arch/x86/include/asm/topology.h @@ -47,21 +47,6 @@ #include <asm/mpspec.h> -#ifdef CONFIG_X86_32 - -/* Mappings between logical cpu number and node number */ -extern int cpu_to_node_map[]; - -/* Returns the number of the node containing CPU 'cpu' */ -static inline int __cpu_to_node(int cpu) -{ - return cpu_to_node_map[cpu]; -} -#define early_cpu_to_node __cpu_to_node -#define cpu_to_node __cpu_to_node - -#else /* CONFIG_X86_64 */ - /* Mappings between logical cpu number and node number */ DECLARE_EARLY_PER_CPU(int, x86_cpu_to_node_map); @@ -84,8 +69,6 @@ static inline int early_cpu_to_node(int cpu) #endif /* !CONFIG_DEBUG_PER_CPU_MAPS */ -#endif /* CONFIG_X86_64 */ - /* Mappings between node number and cpus on that node. */ extern cpumask_var_t node_to_cpumask_map[MAX_NUMNODES]; @@ -155,7 +138,7 @@ extern unsigned long node_remap_size[]; .balance_interval = 1, \ } -#ifdef CONFIG_X86_64_ACPI_NUMA +#ifdef CONFIG_X86_64 extern int __node_distance(int, int); #define node_distance(a, b) __node_distance(a, b) #endif diff --git a/arch/x86/kernel/acpi/boot.c b/arch/x86/kernel/acpi/boot.c index 3e6e2d6..9a966c5 100644 --- a/arch/x86/kernel/acpi/boot.c +++ b/arch/x86/kernel/acpi/boot.c @@ -595,14 +595,8 @@ static void acpi_map_cpu2node(acpi_handle handle, int cpu, int physid) nid = acpi_get_node(handle); if (nid == -1 || !node_online(nid)) return; -#ifdef CONFIG_X86_64 - apicid_to_node[physid] = nid; + set_apicid_to_node(physid, nid); numa_set_node(cpu, nid); -#else /* CONFIG_X86_32 */ - apicid_2_node[physid] = nid; - cpu_to_node_map[cpu] = nid; -#endif - #endif } diff --git a/arch/x86/kernel/amd_nb.c b/arch/x86/kernel/amd_nb.c index 0a99f71..ed3c2e5 100644 --- a/arch/x86/kernel/amd_nb.c +++ b/arch/x86/kernel/amd_nb.c @@ -12,7 +12,7 @@ static u32 *flush_words; -struct pci_device_id amd_nb_misc_ids[] = { +const struct pci_device_id amd_nb_misc_ids[] = { { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_K8_NB_MISC) }, { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_10H_NB_MISC) }, { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_15H_NB_MISC) }, @@ -20,6 +20,11 @@ struct pci_device_id amd_nb_misc_ids[] = { }; EXPORT_SYMBOL(amd_nb_misc_ids); +static struct pci_device_id amd_nb_link_ids[] = { + { PCI_DEVICE(PCI_VENDOR_ID_AMD, PCI_DEVICE_ID_AMD_15H_NB_LINK) }, + {} +}; + const struct amd_nb_bus_dev_range amd_nb_bus_dev_ranges[] __initconst = { { 0x00, 0x18, 0x20 }, { 0xff, 0x00, 0x20 }, @@ -31,7 +36,7 @@ struct amd_northbridge_info amd_northbridges; EXPORT_SYMBOL(amd_northbridges); static struct pci_dev *next_northbridge(struct pci_dev *dev, - struct pci_device_id *ids) + const struct pci_device_id *ids) { do { dev = pci_get_device(PCI_ANY_ID, PCI_ANY_ID, dev); @@ -45,7 +50,7 @@ int amd_cache_northbridges(void) { int i = 0; struct amd_northbridge *nb; - struct pci_dev *misc; + struct pci_dev *misc, *link; if (amd_nb_num()) return 0; @@ -64,10 +69,12 @@ int amd_cache_northbridges(void) amd_northbridges.nb = nb; amd_northbridges.num = i; - misc = NULL; + link = misc = NULL; for (i = 0; i != amd_nb_num(); i++) { node_to_amd_nb(i)->misc = misc = next_northbridge(misc, amd_nb_misc_ids); + node_to_amd_nb(i)->link = link = + next_northbridge(link, amd_nb_link_ids); } /* some CPU families (e.g. family 0x11) do not support GART */ @@ -85,6 +92,13 @@ int amd_cache_northbridges(void) boot_cpu_data.x86_mask >= 0x1)) amd_northbridges.flags |= AMD_NB_L3_INDEX_DISABLE; + if (boot_cpu_data.x86 == 0x15) + amd_northbridges.flags |= AMD_NB_L3_INDEX_DISABLE; + + /* L3 cache partitioning is supported on family 0x15 */ + if (boot_cpu_data.x86 == 0x15) + amd_northbridges.flags |= AMD_NB_L3_PARTITIONING; + return 0; } EXPORT_SYMBOL_GPL(amd_cache_northbridges); @@ -93,8 +107,9 @@ EXPORT_SYMBOL_GPL(amd_cache_northbridges); they're useless anyways */ int __init early_is_amd_nb(u32 device) { - struct pci_device_id *id; + const struct pci_device_id *id; u32 vendor = device & 0xffff; + device >>= 16; for (id = amd_nb_misc_ids; id->vendor; id++) if (vendor == id->vendor && device == id->device) @@ -102,6 +117,65 @@ int __init early_is_amd_nb(u32 device) return 0; } +int amd_get_subcaches(int cpu) +{ + struct pci_dev *link = node_to_amd_nb(amd_get_nb_id(cpu))->link; + unsigned int mask; + int cuid = 0; + + if (!amd_nb_has_feature(AMD_NB_L3_PARTITIONING)) + return 0; + + pci_read_config_dword(link, 0x1d4, &mask); + +#ifdef CONFIG_SMP + cuid = cpu_data(cpu).compute_unit_id; +#endif + return (mask >> (4 * cuid)) & 0xf; +} + +int amd_set_subcaches(int cpu, int mask) +{ + static unsigned int reset, ban; + struct amd_northbridge *nb = node_to_amd_nb(amd_get_nb_id(cpu)); + unsigned int reg; + int cuid = 0; + + if (!amd_nb_has_feature(AMD_NB_L3_PARTITIONING) || mask > 0xf) + return -EINVAL; + + /* if necessary, collect reset state of L3 partitioning and BAN mode */ + if (reset == 0) { + pci_read_config_dword(nb->link, 0x1d4, &reset); + pci_read_config_dword(nb->misc, 0x1b8, &ban); + ban &= 0x180000; + } + + /* deactivate BAN mode if any subcaches are to be disabled */ + if (mask != 0xf) { + pci_read_config_dword(nb->misc, 0x1b8, ®); + pci_write_config_dword(nb->misc, 0x1b8, reg & ~0x180000); + } + +#ifdef CONFIG_SMP + cuid = cpu_data(cpu).compute_unit_id; +#endif + mask <<= 4 * cuid; + mask |= (0xf ^ (1 << cuid)) << 26; + + pci_write_config_dword(nb->link, 0x1d4, mask); + + /* reset BAN mode if L3 partitioning returned to reset state */ + pci_read_config_dword(nb->link, 0x1d4, ®); + if (reg == reset) { + pci_read_config_dword(nb->misc, 0x1b8, ®); + reg &= ~0x180000; + pci_write_config_dword(nb->misc, 0x1b8, reg | ban); + } + + return 0; +} + int amd_cache_gart(void) { int i; diff --git a/arch/x86/kernel/aperture_64.c b/arch/x86/kernel/aperture_64.c index 5955a78..7b1e8e1 100644 --- a/arch/x86/kernel/aperture_64.c +++ b/arch/x86/kernel/aperture_64.c @@ -13,7 +13,7 @@ #include <linux/kernel.h> #include <linux/types.h> #include <linux/init.h> -#include <linux/bootmem.h> +#include <linux/memblock.h> #include <linux/mmzone.h> #include <linux/pci_ids.h> #include <linux/pci.h> @@ -57,7 +57,7 @@ static void __init insert_aperture_resource(u32 aper_base, u32 aper_size) static u32 __init allocate_aperture(void) { u32 aper_size; - void *p; + unsigned long addr; /* aper_size should <= 1G */ if (fallback_aper_order > 5) @@ -83,27 +83,26 @@ static u32 __init allocate_aperture(void) * so don't use 512M below as gart iommu, leave the space for kernel * code for safe */ - p = __alloc_bootmem_nopanic(aper_size, aper_size, 512ULL<<20); + addr = memblock_find_in_range(0, 1ULL<<32, aper_size, 512ULL<<20); + if (addr == MEMBLOCK_ERROR || addr + aper_size > 0xffffffff) { + printk(KERN_ERR + "Cannot allocate aperture memory hole (%lx,%uK)\n", + addr, aper_size>>10); + return 0; + } + memblock_x86_reserve_range(addr, addr + aper_size, "aperture64"); /* * Kmemleak should not scan this block as it may not be mapped via the * kernel direct mapping. */ - kmemleak_ignore(p); - if (!p || __pa(p)+aper_size > 0xffffffff) { - printk(KERN_ERR - "Cannot allocate aperture memory hole (%p,%uK)\n", - p, aper_size>>10); - if (p) - free_bootmem(__pa(p), aper_size); - return 0; - } + kmemleak_ignore(phys_to_virt(addr)); printk(KERN_INFO "Mapping aperture over %d KB of RAM @ %lx\n", - aper_size >> 10, __pa(p)); - insert_aperture_resource((u32)__pa(p), aper_size); - register_nosave_region((u32)__pa(p) >> PAGE_SHIFT, - (u32)__pa(p+aper_size) >> PAGE_SHIFT); + aper_size >> 10, addr); + insert_aperture_resource((u32)addr, aper_size); + register_nosave_region(addr >> PAGE_SHIFT, + (addr+aper_size) >> PAGE_SHIFT); - return (u32)__pa(p); + return (u32)addr; } diff --git a/arch/x86/kernel/apic/apic.c b/arch/x86/kernel/apic/apic.c index 48dcd2e..562a832 100644 --- a/arch/x86/kernel/apic/apic.c +++ b/arch/x86/kernel/apic/apic.c @@ -79,6 +79,15 @@ EXPORT_EARLY_PER_CPU_SYMBOL(x86_cpu_to_apicid); EXPORT_EARLY_PER_CPU_SYMBOL(x86_bios_cpu_apicid); #ifdef CONFIG_X86_32 + +/* + * On x86_32, the mapping between cpu and logical apicid may vary + * depending on apic in use. The following early percpu variable is + * used for the mapping. This is where the behaviors of x86_64 and 32 + * actually diverge. Let's keep it ugly for now. + */ +DEFINE_EARLY_PER_CPU(int, x86_cpu_to_logical_apicid, BAD_APICID); + /* * Knob to control our willingness to enable the local APIC. * @@ -1238,6 +1247,19 @@ void __cpuinit setup_local_APIC(void) */ apic->init_apic_ldr(); +#ifdef CONFIG_X86_32 + /* + * APIC LDR is initialized. If logical_apicid mapping was + * initialized during get_smp_config(), make sure it matches the + * actual value. + */ + i = early_per_cpu(x86_cpu_to_logical_apicid, cpu); + WARN_ON(i != BAD_APICID && i != logical_smp_processor_id()); + /* always use the value from LDR */ + early_per_cpu(x86_cpu_to_logical_apicid, cpu) = + logical_smp_processor_id(); +#endif + /* * Set Task Priority to 'accept all'. We never change this * later on. @@ -1979,7 +2001,10 @@ void __cpuinit generic_processor_info(int apicid, int version) early_per_cpu(x86_cpu_to_apicid, cpu) = apicid; early_per_cpu(x86_bios_cpu_apicid, cpu) = apicid; #endif - +#ifdef CONFIG_X86_32 + early_per_cpu(x86_cpu_to_logical_apicid, cpu) = + apic->x86_32_early_logical_apicid(cpu); +#endif set_cpu_possible(cpu, true); set_cpu_present(cpu, true); } @@ -2000,10 +2025,14 @@ void default_init_apic_ldr(void) } #ifdef CONFIG_X86_32 -int default_apicid_to_node(int logical_apicid) +int default_x86_32_numa_cpu_node(int cpu) { -#ifdef CONFIG_SMP - return apicid_2_node[hard_smp_processor_id()]; +#ifdef CONFIG_NUMA + int apicid = early_per_cpu(x86_cpu_to_apicid, cpu); + + if (apicid != BAD_APICID) + return __apicid_to_node[apicid]; + return NUMA_NO_NODE; #else return 0; #endif diff --git a/arch/x86/kernel/apic/apic_flat_64.c b/arch/x86/kernel/apic/apic_flat_64.c index 09d3b17..5652d31 100644 --- a/arch/x86/kernel/apic/apic_flat_64.c +++ b/arch/x86/kernel/apic/apic_flat_64.c @@ -185,8 +185,6 @@ struct apic apic_flat = { .ioapic_phys_id_map = NULL, .setup_apic_routing = NULL, .multi_timer_check = NULL, - .apicid_to_node = NULL, - .cpu_to_logical_apicid = NULL, .cpu_present_to_apicid = default_cpu_present_to_apicid, .apicid_to_cpu_present = NULL, .setup_portio_remap = NULL, @@ -337,8 +335,6 @@ struct apic apic_physflat = { .ioapic_phys_id_map = NULL, .setup_apic_routing = NULL, .multi_timer_check = NULL, - .apicid_to_node = NULL, - .cpu_to_logical_apicid = NULL, .cpu_present_to_apicid = default_cpu_present_to_apicid, .apicid_to_cpu_present = NULL, .setup_portio_remap = NULL, diff --git a/arch/x86/kernel/apic/apic_noop.c b/arch/x86/kernel/apic/apic_noop.c index e31b9ff..f1baa2d 100644 --- a/arch/x86/kernel/apic/apic_noop.c +++ b/arch/x86/kernel/apic/apic_noop.c @@ -54,11 +54,6 @@ static u64 noop_apic_icr_read(void) return 0; } -static int noop_cpu_to_logical_apicid(int cpu) -{ - return 0; -} - static int noop_phys_pkg_id(int cpuid_apic, int index_msb) { return 0; @@ -113,12 +108,6 @@ static void noop_vector_allocation_domain(int cpu, struct cpumask *retmask) cpumask_set_cpu(cpu, retmask); } -int noop_apicid_to_node(int logical_apicid) -{ - /* we're always on node 0 */ - return 0; -} - static u32 noop_apic_read(u32 reg) { WARN_ON_ONCE((cpu_has_apic && !disable_apic)); @@ -130,6 +119,14 @@ static void noop_apic_write(u32 reg, u32 v) WARN_ON_ONCE(cpu_has_apic && !disable_apic); } +#ifdef CONFIG_X86_32 +static int noop_x86_32_numa_cpu_node(int cpu) +{ + /* we're always on node 0 */ + return 0; +} +#endif + struct apic apic_noop = { .name = "noop", .probe = noop_probe, @@ -153,9 +150,7 @@ struct apic apic_noop = { .ioapic_phys_id_map = default_ioapic_phys_id_map, .setup_apic_routing = NULL, .multi_timer_check = NULL, - .apicid_to_node = noop_apicid_to_node, - .cpu_to_logical_apicid = noop_cpu_to_logical_apicid, .cpu_present_to_apicid = default_cpu_present_to_apicid, .apicid_to_cpu_present = physid_set_mask_of_physid, @@ -197,4 +192,9 @@ struct apic apic_noop = { .icr_write = noop_apic_icr_write, .wait_icr_idle = noop_apic_wait_icr_idle, .safe_wait_icr_idle = noop_safe_apic_wait_icr_idle, + +#ifdef CONFIG_X86_32 + .x86_32_early_logical_apicid = noop_x86_32_early_logical_apicid, + .x86_32_numa_cpu_node = noop_x86_32_numa_cpu_node, +#endif }; diff --git a/arch/x86/kernel/apic/bigsmp_32.c b/arch/x86/kernel/apic/bigsmp_32.c index cb804c5..541a2e4 100644 --- a/arch/x86/kernel/apic/bigsmp_32.c +++ b/arch/x86/kernel/apic/bigsmp_32.c @@ -45,6 +45,12 @@ static unsigned long bigsmp_check_apicid_present(int bit) return 1; } +static int bigsmp_early_logical_apicid(int cpu) +{ + /* on bigsmp, logical apicid is the same as physical */ + return early_per_cpu(x86_cpu_to_apicid, cpu); +} + static inline unsigned long calculate_ldr(int cpu) { unsigned long val, id; @@ -80,11 +86,6 @@ static void bigsmp_setup_apic_routing(void) nr_ioapics); } -static int bigsmp_apicid_to_node(int logical_apicid) -{ - return apicid_2_node[hard_smp_processor_id()]; -} - static int bigsmp_cpu_present_to_apicid(int mps_cpu) { if (mps_cpu < nr_cpu_ids) @@ -93,14 +94,6 @@ static int bigsmp_cpu_present_to_apicid(int mps_cpu) return BAD_APICID; } -/* Mapping from cpu number to logical apicid */ -static inline int bigsmp_cpu_to_logical_apicid(int cpu) -{ - if (cpu >= nr_cpu_ids) - return BAD_APICID; - return cpu_physical_id(cpu); -} - static void bigsmp_ioapic_phys_id_map(physid_mask_t *phys_map, physid_mask_t *retmap) { /* For clustered we don't have a good way to do this yet - hack */ @@ -115,7 +108,11 @@ static int bigsmp_check_phys_apicid_present(int phys_apicid) /* As we are using single CPU as destination, pick only one CPU here */ static unsigned int bigsmp_cpu_mask_to_apicid(const struct cpumask *cpumask) { - return bigsmp_cpu_to_logical_apicid(cpumask_first(cpumask)); + int cpu = cpumask_first(cpumask); + + if (cpu < nr_cpu_ids) + return cpu_physical_id(cpu); + return BAD_APICID; } static unsigned int bigsmp_cpu_mask_to_apicid_and(const struct cpumask *cpumask, @@ -129,9 +126,9 @@ static unsigned int bigsmp_cpu_mask_to_apicid_and(const struct cpumask *cpumask, */ for_each_cpu_and(cpu, cpumask, andmask) { if (cpumask_test_cpu(cpu, cpu_online_mask)) - break; + return cpu_physical_id(cpu); } - return bigsmp_cpu_to_logical_apicid(cpu); + return BAD_APICID; } static int bigsmp_phys_pkg_id(int cpuid_apic, int index_msb) @@ -219,8 +216,6 @@ struct apic apic_bigsmp = { .ioapic_phys_id_map = bigsmp_ioapic_phys_id_map, .setup_apic_routing = bigsmp_setup_apic_routing, .multi_timer_check = NULL, - .apicid_to_node = bigsmp_apicid_to_node, - .cpu_to_logical_apicid = bigsmp_cpu_to_logical_apicid, .cpu_present_to_apicid = bigsmp_cpu_present_to_apicid, .apicid_to_cpu_present = physid_set_mask_of_physid, .setup_portio_remap = NULL, @@ -256,4 +251,7 @@ struct apic apic_bigsmp = { .icr_write = native_apic_icr_write, .wait_icr_idle = native_apic_wait_icr_idle, .safe_wait_icr_idle = native_safe_apic_wait_icr_idle, + + .x86_32_early_logical_apicid = bigsmp_early_logical_apicid, + .x86_32_numa_cpu_node = default_x86_32_numa_cpu_node, }; diff --git a/arch/x86/kernel/apic/es7000_32.c b/arch/x86/kernel/apic/es7000_32.c index 8593582..3e9de48 100644 --- a/arch/x86/kernel/apic/es7000_32.c +++ b/arch/x86/kernel/apic/es7000_32.c @@ -460,6 +460,12 @@ static unsigned long es7000_check_apicid_present(int bit) return physid_isset(bit, phys_cpu_present_map); } +static int es7000_early_logical_apicid(int cpu) +{ + /* on es7000, logical apicid is the same as physical */ + return early_per_cpu(x86_bios_cpu_apicid, cpu); +} + static unsigned long calculate_ldr(int cpu) { unsigned long id = per_cpu(x86_bios_cpu_apicid, cpu); @@ -504,12 +510,11 @@ static void es7000_setup_apic_routing(void) nr_ioapics, cpumask_bits(es7000_target_cpus())[0]); } -static int es7000_apicid_to_node(int logical_apicid) +static int es7000_numa_cpu_node(int cpu) { return 0; } - static int es7000_cpu_present_to_apicid(int mps_cpu) { if (!mps_cpu) @@ -528,18 +533,6 @@ static void es7000_apicid_to_cpu_present(int phys_apicid, physid_mask_t *retmap) ++cpu_id; } -/* Mapping from cpu number to logical apicid */ -static int es7000_cpu_to_logical_apicid(int cpu) -{ -#ifdef CONFIG_SMP - if (cpu >= nr_cpu_ids) - return BAD_APICID; - return cpu_2_logical_apicid[cpu]; -#else - return logical_smp_processor_id(); -#endif -} - static void es7000_ioapic_phys_id_map(physid_mask_t *phys_map, physid_mask_t *retmap) { /* For clustered we don't have a good way to do this yet - hack */ @@ -561,7 +554,7 @@ static unsigned int es7000_cpu_mask_to_apicid(const struct cpumask *cpumask) * The cpus in the mask must all be on the apic cluster. */ for_each_cpu(cpu, cpumask) { - int new_apicid = es7000_cpu_to_logical_apicid(cpu); + int new_apicid = early_per_cpu(x86_cpu_to_logical_apicid, cpu); if (round && APIC_CLUSTER(apicid) != APIC_CLUSTER(new_apicid)) { WARN(1, "Not a valid mask!"); @@ -578,7 +571,7 @@ static unsigned int es7000_cpu_mask_to_apicid_and(const struct cpumask *inmask, const struct cpumask *andmask) { - int apicid = es7000_cpu_to_logical_apicid(0); + int apicid = early_per_cpu(x86_cpu_to_logical_apicid, 0); cpumask_var_t cpumask; if (!alloc_cpumask_var(&cpumask, GFP_ATOMIC)) @@ -655,8 +648,6 @@ struct apic __refdata apic_es7000_cluster = { .ioapic_phys_id_map = es7000_ioapic_phys_id_map, .setup_apic_routing = es7000_setup_apic_routing, .multi_timer_check = NULL, - .apicid_to_node = es7000_apicid_to_node, - .cpu_to_logical_apicid = es7000_cpu_to_logical_apicid, .cpu_present_to_apicid = es7000_cpu_present_to_apicid, .apicid_to_cpu_present = es7000_apicid_to_cpu_present, .setup_portio_remap = NULL, @@ -695,6 +686,9 @@ struct apic __refdata apic_es7000_cluster = { .icr_write = native_apic_icr_write, .wait_icr_idle = native_apic_wait_icr_idle, .safe_wait_icr_idle = native_safe_apic_wait_icr_idle, + + .x86_32_early_logical_apicid = es7000_early_logical_apicid, + .x86_32_numa_cpu_node = es7000_numa_cpu_node, }; struct apic __refdata apic_es7000 = { @@ -720,8 +714,6 @@ struct apic __refdata apic_es7000 = { .ioapic_phys_id_map = es7000_ioapic_phys_id_map, .setup_apic_routing = es7000_setup_apic_routing, .multi_timer_check = NULL, - .apicid_to_node = es7000_apicid_to_node, - .cpu_to_logical_apicid = es7000_cpu_to_logical_apicid, .cpu_present_to_apicid = es7000_cpu_present_to_apicid, .apicid_to_cpu_present = es7000_apicid_to_cpu_present, .setup_portio_remap = NULL, @@ -758,4 +750,7 @@ struct apic __refdata apic_es7000 = { .icr_write = native_apic_icr_write, .wait_icr_idle = native_apic_wait_icr_idle, .safe_wait_icr_idle = native_safe_apic_wait_icr_idle, + + .x86_32_early_logical_apicid = es7000_early_logical_apicid, + .x86_32_numa_cpu_node = es7000_numa_cpu_node, }; diff --git a/arch/x86/kernel/apic/ipi.c b/arch/x86/kernel/apic/ipi.c index 08385e0..cce91bf 100644 --- a/arch/x86/kernel/apic/ipi.c +++ b/arch/x86/kernel/apic/ipi.c @@ -56,6 +56,8 @@ void default_send_IPI_mask_allbutself_phys(const struct cpumask *mask, local_irq_restore(flags); } +#ifdef CONFIG_X86_32 + void default_send_IPI_mask_sequence_logical(const struct cpumask *mask, int vector) { @@ -71,8 +73,8 @@ void default_send_IPI_mask_sequence_logical(const struct cpumask *mask, local_irq_save(flags); for_each_cpu(query_cpu, mask) __default_send_IPI_dest_field( - apic->cpu_to_logical_apicid(query_cpu), vector, - apic->dest_logical); + early_per_cpu(x86_cpu_to_logical_apicid, query_cpu), + vector, apic->dest_logical); local_irq_restore(flags); } @@ -90,14 +92,12 @@ void default_send_IPI_mask_allbutself_logical(const struct cpumask *mask, if (query_cpu == this_cpu) continue; __default_send_IPI_dest_field( - apic->cpu_to_logical_apicid(query_cpu), vector, - apic->dest_logical); + early_per_cpu(x86_cpu_to_logical_apicid, query_cpu), + vector, apic->dest_logical); } local_irq_restore(flags); } -#ifdef CONFIG_X86_32 - /* * This is only used on smaller machines. */ diff --git a/arch/x86/kernel/apic/numaq_32.c b/arch/x86/kernel/apic/numaq_32.c index 960f26a..6273eee 100644 --- a/arch/x86/kernel/apic/numaq_32.c +++ b/arch/x86/kernel/apic/numaq_32.c @@ -373,13 +373,6 @@ static inline void numaq_ioapic_phys_id_map(physid_mask_t *phys_map, physid_mask return physids_promote(0xFUL, retmap); } -static inline int numaq_cpu_to_logical_apicid(int cpu) -{ - if (cpu >= nr_cpu_ids) - return BAD_APICID; - return cpu_2_logical_apicid[cpu]; -} - /* * Supporting over 60 cpus on NUMA-Q requires a locality-dependent * cpu to APIC ID relation to properly interact with the intelligent @@ -398,6 +391,15 @@ static inline int numaq_apicid_to_node(int logical_apicid) return logical_apicid >> 4; } +static int numaq_numa_cpu_node(int cpu) +{ + int logical_apicid = early_per_cpu(x86_cpu_to_logical_apicid, cpu); + + if (logical_apicid != BAD_APICID) + return numaq_apicid_to_node(logical_apicid); + return NUMA_NO_NODE; +} + static void numaq_apicid_to_cpu_present(int logical_apicid, physid_mask_t *retmap) { int node = numaq_apicid_to_node(logical_apicid); @@ -508,8 +510,6 @@ struct apic __refdata apic_numaq = { .ioapic_phys_id_map = numaq_ioapic_phys_id_map, .setup_apic_routing = numaq_setup_apic_routing, .multi_timer_check = numaq_multi_timer_check, - .apicid_to_node = numaq_apicid_to_node, - .cpu_to_logical_apicid = numaq_cpu_to_logical_apicid, .cpu_present_to_apicid = numaq_cpu_present_to_apicid, .apicid_to_cpu_present = numaq_apicid_to_cpu_present, .setup_portio_remap = numaq_setup_portio_remap, @@ -547,4 +547,7 @@ struct apic __refdata apic_numaq = { .icr_write = native_apic_icr_write, .wait_icr_idle = native_apic_wait_icr_idle, .safe_wait_icr_idle = native_safe_apic_wait_icr_idle, + + .x86_32_early_logical_apicid = noop_x86_32_early_logical_apicid, + .x86_32_numa_cpu_node = numaq_numa_cpu_node, }; diff --git a/arch/x86/kernel/apic/probe_32.c b/arch/x86/kernel/apic/probe_32.c index 99d2fe0..fc84c7b 100644 --- a/arch/x86/kernel/apic/probe_32.c +++ b/arch/x86/kernel/apic/probe_32.c @@ -77,6 +77,11 @@ void __init default_setup_apic_routing(void) apic->setup_apic_routing(); } +static int default_x86_32_early_logical_apicid(int cpu) +{ + return 1 << cpu; +} + static void setup_apic_flat_routing(void) { #ifdef CONFIG_X86_IO_APIC @@ -130,8 +135,6 @@ struct apic apic_default = { .ioapic_phys_id_map = default_ioapic_phys_id_map, .setup_apic_routing = setup_apic_flat_routing, .multi_timer_check = NULL, - .apicid_to_node = default_apicid_to_node, - .cpu_to_logical_apicid = default_cpu_to_logical_apicid, .cpu_present_to_apicid = default_cpu_present_to_apicid, .apicid_to_cpu_present = physid_set_mask_of_physid, .setup_portio_remap = NULL, @@ -167,6 +170,9 @@ struct apic apic_default = { .icr_write = native_apic_icr_write, .wait_icr_idle = native_apic_wait_icr_idle, .safe_wait_icr_idle = native_safe_apic_wait_icr_idle, + + .x86_32_early_logical_apicid = default_x86_32_early_logical_apicid, + .x86_32_numa_cpu_node = default_x86_32_numa_cpu_node, }; extern struct apic apic_numaq; diff --git a/arch/x86/kernel/apic/summit_32.c b/arch/x86/kernel/apic/summit_32.c index 9b41926..e4b8059 100644 --- a/arch/x86/kernel/apic/summit_32.c +++ b/arch/x86/kernel/apic/summit_32.c @@ -194,11 +194,10 @@ static unsigned long summit_check_apicid_present(int bit) return 1; } -static void summit_init_apic_ldr(void) +static int summit_early_logical_apicid(int cpu) { - unsigned long val, id; int count = 0; - u8 my_id = (u8)hard_smp_processor_id(); + u8 my_id = early_per_cpu(x86_cpu_to_apicid, cpu); u8 my_cluster = APIC_CLUSTER(my_id); #ifdef CONFIG_SMP u8 lid; @@ -206,7 +205,7 @@ static void summit_init_apic_ldr(void) /* Create logical APIC IDs by counting CPUs already in cluster. */ for (count = 0, i = nr_cpu_ids; --i >= 0; ) { - lid = cpu_2_logical_apicid[i]; + lid = early_per_cpu(x86_cpu_to_logical_apicid, i); if (lid != BAD_APICID && APIC_CLUSTER(lid) == my_cluster) ++count; } @@ -214,7 +213,15 @@ static void summit_init_apic_ldr(void) /* We only have a 4 wide bitmap in cluster mode. If a deranged * BIOS puts 5 CPUs in one APIC cluster, we're hosed. */ BUG_ON(count >= XAPIC_DEST_CPUS_SHIFT); - id = my_cluster | (1UL << count); + return my_cluster | (1UL << count); +} + +static void summit_init_apic_ldr(void) +{ + int cpu = smp_processor_id(); + unsigned long id = early_per_cpu(x86_cpu_to_logical_apicid, cpu); + unsigned long val; + apic_write(APIC_DFR, SUMMIT_APIC_DFR_VALUE); val = apic_read(APIC_LDR) & ~APIC_LDR_MASK; val |= SET_APIC_LOGICAL_ID(id); @@ -232,27 +239,6 @@ static void summit_setup_apic_routing(void) nr_ioapics); } -static int summit_apicid_to_node(int logical_apicid) -{ -#ifdef CONFIG_SMP - return apicid_2_node[hard_smp_processor_id()]; -#else - return 0; -#endif -} - -/* Mapping from cpu number to logical apicid */ -static inline int summit_cpu_to_logical_apicid(int cpu) -{ -#ifdef CONFIG_SMP - if (cpu >= nr_cpu_ids) - return BAD_APICID; - return cpu_2_logical_apicid[cpu]; -#else - return logical_smp_processor_id(); -#endif -} - static int summit_cpu_present_to_apicid(int mps_cpu) { if (mps_cpu < nr_cpu_ids) @@ -286,7 +272,7 @@ static unsigned int summit_cpu_mask_to_apicid(const struct cpumask *cpumask) * The cpus in the mask must all be on the apic cluster. */ for_each_cpu(cpu, cpumask) { - int new_apicid = summit_cpu_to_logical_apicid(cpu); + int new_apicid = early_per_cpu(x86_cpu_to_logical_apicid, cpu); if (round && APIC_CLUSTER(apicid) != APIC_CLUSTER(new_apicid)) { printk("%s: Not a valid mask!\n", __func__); @@ -301,7 +287,7 @@ static unsigned int summit_cpu_mask_to_apicid(const struct cpumask *cpumask) static unsigned int summit_cpu_mask_to_apicid_and(const struct cpumask *inmask, const struct cpumask *andmask) { - int apicid = summit_cpu_to_logical_apicid(0); + int apicid = early_per_cpu(x86_cpu_to_logical_apicid, 0); cpumask_var_t cpumask; if (!alloc_cpumask_var(&cpumask, GFP_ATOMIC)) @@ -528,8 +514,6 @@ struct apic apic_summit = { .ioapic_phys_id_map = summit_ioapic_phys_id_map, .setup_apic_routing = summit_setup_apic_routing, .multi_timer_check = NULL, - .apicid_to_node = summit_apicid_to_node, - .cpu_to_logical_apicid = summit_cpu_to_logical_apicid, .cpu_present_to_apicid = summit_cpu_present_to_apicid, .apicid_to_cpu_present = summit_apicid_to_cpu_present, .setup_portio_remap = NULL, @@ -565,4 +549,7 @@ struct apic apic_summit = { .icr_write = native_apic_icr_write, .wait_icr_idle = native_apic_wait_icr_idle, .safe_wait_icr_idle = native_safe_apic_wait_icr_idle, + + .x86_32_early_logical_apicid = summit_early_logical_apicid, + .x86_32_numa_cpu_node = default_x86_32_numa_cpu_node, }; diff --git a/arch/x86/kernel/apic/x2apic_cluster.c b/arch/x86/kernel/apic/x2apic_cluster.c index cf69c59..90949bb 100644 --- a/arch/x86/kernel/apic/x2apic_cluster.c +++ b/arch/x86/kernel/apic/x2apic_cluster.c @@ -206,8 +206,6 @@ struct apic apic_x2apic_cluster = { .ioapic_phys_id_map = NULL, .setup_apic_routing = NULL, .multi_timer_check = NULL, - .apicid_to_node = NULL, - .cpu_to_logical_apicid = NULL, .cpu_present_to_apicid = default_cpu_present_to_apicid, .apicid_to_cpu_present = NULL, .setup_portio_remap = NULL, diff --git a/arch/x86/kernel/apic/x2apic_phys.c b/arch/x86/kernel/apic/x2apic_phys.c index 8972f38..c7e6d66 100644 --- a/arch/x86/kernel/apic/x2apic_phys.c +++ b/arch/x86/kernel/apic/x2apic_phys.c @@ -195,8 +195,6 @@ struct apic apic_x2apic_phys = { .ioapic_phys_id_map = NULL, .setup_apic_routing = NULL, .multi_timer_check = NULL, - .apicid_to_node = NULL, - .cpu_to_logical_apicid = NULL, .cpu_present_to_apicid = default_cpu_present_to_apicid, .apicid_to_cpu_present = NULL, .setup_portio_remap = NULL, diff --git a/arch/x86/kernel/apic/x2apic_uv_x.c b/arch/x86/kernel/apic/x2apic_uv_x.c index bd16b58..3c28928 100644 --- a/arch/x86/kernel/apic/x2apic_uv_x.c +++ b/arch/x86/kernel/apic/x2apic_uv_x.c @@ -338,8 +338,6 @@ struct apic __refdata apic_x2apic_uv_x = { .ioapic_phys_id_map = NULL, .setup_apic_routing = NULL, .multi_timer_check = NULL, - .apicid_to_node = NULL, - .cpu_to_logical_apicid = NULL, .cpu_present_to_apicid = default_cpu_present_to_apicid, .apicid_to_cpu_present = NULL, .setup_portio_remap = NULL, diff --git a/arch/x86/kernel/cpu/amd.c b/arch/x86/kernel/cpu/amd.c index 7c7bedb..f771ab6 100644 --- a/arch/x86/kernel/cpu/amd.c +++ b/arch/x86/kernel/cpu/amd.c @@ -233,18 +233,22 @@ static void __cpuinit init_amd_k7(struct cpuinfo_x86 *c) } #endif -#if defined(CONFIG_NUMA) && defined(CONFIG_X86_64) +#ifdef CONFIG_NUMA +/* + * To workaround broken NUMA config. Read the comment in + * srat_detect_node(). + */ static int __cpuinit nearby_node(int apicid) { int i, node; for (i = apicid - 1; i >= 0; i--) { - node = apicid_to_node[i]; + node = __apicid_to_node[i]; if (node != NUMA_NO_NODE && node_online(node)) return node; } for (i = apicid + 1; i < MAX_LOCAL_APIC; i++) { - node = apicid_to_node[i]; + node = __apicid_to_node[i]; if (node != NUMA_NO_NODE && node_online(node)) return node; } @@ -261,7 +265,7 @@ static int __cpuinit nearby_node(int apicid) #ifdef CONFIG_X86_HT static void __cpuinit amd_get_topology(struct cpuinfo_x86 *c) { - u32 nodes; + u32 nodes, cores_per_cu = 1; u8 node_id; int cpu = smp_processor_id(); @@ -276,6 +280,7 @@ static void __cpuinit amd_get_topology(struct cpuinfo_x86 *c) /* get compute unit information */ smp_num_siblings = ((ebx >> 8) & 3) + 1; c->compute_unit_id = ebx & 0xff; + cores_per_cu += ((ebx >> 8) & 3); } else if (cpu_has(c, X86_FEATURE_NODEID_MSR)) { u64 value; @@ -288,15 +293,18 @@ static void __cpuinit amd_get_topology(struct cpuinfo_x86 *c) /* fixup multi-node processor information */ if (nodes > 1) { u32 cores_per_node; + u32 cus_per_node; set_cpu_cap(c, X86_FEATURE_AMD_DCM); cores_per_node = c->x86_max_cores / nodes; + cus_per_node = cores_per_node / cores_per_cu; /* store NodeID, use llc_shared_map to store sibling info */ per_cpu(cpu_llc_id, cpu) = node_id; - /* core id to be in range from 0 to (cores_per_node - 1) */ - c->cpu_core_id = c->cpu_core_id % cores_per_node; + /* core id has to be in the [0 .. cores_per_node - 1] range */ + c->cpu_core_id %= cores_per_node; + c->compute_unit_id %= cus_per_node; } } #endif @@ -334,31 +342,40 @@ EXPORT_SYMBOL_GPL(amd_get_nb_id); static void __cpuinit srat_detect_node(struct cpuinfo_x86 *c) { -#if defined(CONFIG_NUMA) && defined(CONFIG_X86_64) +#ifdef CONFIG_NUMA int cpu = smp_processor_id(); int node; unsigned apicid = c->apicid; - node = per_cpu(cpu_llc_id, cpu); + node = numa_cpu_node(cpu); + if (node == NUMA_NO_NODE) + node = per_cpu(cpu_llc_id, cpu); - if (apicid_to_node[apicid] != NUMA_NO_NODE) - node = apicid_to_node[apicid]; if (!node_online(node)) { - /* Two possibilities here: - - The CPU is missing memory and no node was created. - In that case try picking one from a nearby CPU - - The APIC IDs differ from the HyperTransport node IDs - which the K8 northbridge parsing fills in. - Assume they are all increased by a constant offset, - but in the same order as the HT nodeids. - If that doesn't result in a usable node fall back to the - path for the previous case. */ - + /* + * Two possibilities here: + * + * - The CPU is missing memory and no node was created. In + * that case try picking one from a nearby CPU. + * + * - The APIC IDs differ from the HyperTransport node IDs + * which the K8 northbridge parsing fills in. Assume + * they are all increased by a constant offset, but in + * the same order as the HT nodeids. If that doesn't + * result in a usable node fall back to the path for the + * previous case. + * + * This workaround operates directly on the mapping between + * APIC ID and NUMA node, assuming certain relationship + * between APIC ID, HT node ID and NUMA topology. As going + * through CPU mapping may alter the outcome, directly + * access __apicid_to_node[]. + */ int ht_nodeid = c->initial_apicid; if (ht_nodeid >= 0 && - apicid_to_node[ht_nodeid] != NUMA_NO_NODE) - node = apicid_to_node[ht_nodeid]; + __apicid_to_node[ht_nodeid] != NUMA_NO_NODE) + node = __apicid_to_node[ht_nodeid]; /* Pick a nearby node */ if (!node_online(node)) node = nearby_node(apicid); diff --git a/arch/x86/kernel/cpu/common.c b/arch/x86/kernel/cpu/common.c index 5d98c46..e2ced00 100644 --- a/arch/x86/kernel/cpu/common.c +++ b/arch/x86/kernel/cpu/common.c @@ -869,7 +869,7 @@ static void __cpuinit identify_cpu(struct cpuinfo_x86 *c) select_idle_routine(c); -#if defined(CONFIG_NUMA) && defined(CONFIG_X86_64) +#ifdef CONFIG_NUMA numa_add_cpu(smp_processor_id()); #endif } diff --git a/arch/x86/kernel/cpu/intel.c b/arch/x86/kernel/cpu/intel.c index d16c2c5..df86bc8 100644 --- a/arch/x86/kernel/cpu/intel.c +++ b/arch/x86/kernel/cpu/intel.c @@ -276,14 +276,13 @@ static void __cpuinit intel_workarounds(struct cpuinfo_x86 *c) static void __cpuinit srat_detect_node(struct cpuinfo_x86 *c) { -#if defined(CONFIG_NUMA) && defined(CONFIG_X86_64) +#ifdef CONFIG_NUMA unsigned node; int cpu = smp_processor_id(); - int apicid = cpu_has_apic ? hard_smp_processor_id() : c->apicid; /* Don't do the funky fallback heuristics the AMD version employs for now. */ - node = apicid_to_node[apicid]; + node = numa_cpu_node(cpu); if (node == NUMA_NO_NODE || !node_online(node)) { /* reuse the value from init_cpu_to_node() */ node = cpu_to_node(cpu); diff --git a/arch/x86/kernel/cpu/intel_cacheinfo.c b/arch/x86/kernel/cpu/intel_cacheinfo.c index 5419a26..1ce1af28 100644 --- a/arch/x86/kernel/cpu/intel_cacheinfo.c +++ b/arch/x86/kernel/cpu/intel_cacheinfo.c @@ -304,8 +304,9 @@ amd_cpuid4(int leaf, union _cpuid4_leaf_eax *eax, struct _cache_attr { struct attribute attr; - ssize_t (*show)(struct _cpuid4_info *, char *); - ssize_t (*store)(struct _cpuid4_info *, const char *, size_t count); + ssize_t (*show)(struct _cpuid4_info *, char *, unsigned int); + ssize_t (*store)(struct _cpuid4_info *, const char *, size_t count, + unsigned int); }; #ifdef CONFIG_AMD_NB @@ -400,7 +401,8 @@ static ssize_t show_cache_disable(struct _cpuid4_info *this_leaf, char *buf, #define SHOW_CACHE_DISABLE(slot) \ static ssize_t \ -show_cache_disable_##slot(struct _cpuid4_info *this_leaf, char *buf) \ +show_cache_disable_##slot(struct _cpuid4_info *this_leaf, char *buf, \ + unsigned int cpu) \ { \ return show_cache_disable(this_leaf, buf, slot); \ } @@ -512,7 +514,8 @@ static ssize_t store_cache_disable(struct _cpuid4_info *this_leaf, #define STORE_CACHE_DISABLE(slot) \ static ssize_t \ store_cache_disable_##slot(struct _cpuid4_info *this_leaf, \ - const char *buf, size_t count) \ + const char *buf, size_t count, \ + unsigned int cpu) \ { \ return store_cache_disable(this_leaf, buf, count, slot); \ } @@ -524,6 +527,39 @@ static struct _cache_attr cache_disable_0 = __ATTR(cache_disable_0, 0644, static struct _cache_attr cache_disable_1 = __ATTR(cache_disable_1, 0644, show_cache_disable_1, store_cache_disable_1); +static ssize_t +show_subcaches(struct _cpuid4_info *this_leaf, char *buf, unsigned int cpu) +{ + if (!this_leaf->l3 || !amd_nb_has_feature(AMD_NB_L3_PARTITIONING)) + return -EINVAL; + + return sprintf(buf, "%x\n", amd_get_subcaches(cpu)); +} + +static ssize_t +store_subcaches(struct _cpuid4_info *this_leaf, const char *buf, size_t count, + unsigned int cpu) +{ + unsigned long val; + + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + if (!this_leaf->l3 || !amd_nb_has_feature(AMD_NB_L3_PARTITIONING)) + return -EINVAL; + + if (strict_strtoul(buf, 16, &val) < 0) + return -EINVAL; + + if (amd_set_subcaches(cpu, val)) + return -EINVAL; + + return count; +} + +static struct _cache_attr subcaches = + __ATTR(subcaches, 0644, show_subcaches, store_subcaches); + #else /* CONFIG_AMD_NB */ #define amd_init_l3_cache(x, y) #endif /* CONFIG_AMD_NB */ @@ -532,9 +568,9 @@ static int __cpuinit cpuid4_cache_lookup_regs(int index, struct _cpuid4_info_regs *this_leaf) { - union _cpuid4_leaf_eax eax; - union _cpuid4_leaf_ebx ebx; - union _cpuid4_leaf_ecx ecx; + union _cpuid4_leaf_eax eax; + union _cpuid4_leaf_ebx ebx; + union _cpuid4_leaf_ecx ecx; unsigned edx; if (boot_cpu_data.x86_vendor == X86_VENDOR_AMD) { @@ -870,8 +906,8 @@ static DEFINE_PER_CPU(struct _index_kobject *, ici_index_kobject); #define INDEX_KOBJECT_PTR(x, y) (&((per_cpu(ici_index_kobject, x))[y])) #define show_one_plus(file_name, object, val) \ -static ssize_t show_##file_name \ - (struct _cpuid4_info *this_leaf, char *buf) \ +static ssize_t show_##file_name(struct _cpuid4_info *this_leaf, char *buf, \ + unsigned int cpu) \ { \ return sprintf(buf, "%lu\n", (unsigned long)this_leaf->object + val); \ } @@ -882,7 +918,8 @@ show_one_plus(physical_line_partition, ebx.split.physical_line_partition, 1); show_one_plus(ways_of_associativity, ebx.split.ways_of_associativity, 1); show_one_plus(number_of_sets, ecx.split.number_of_sets, 1); -static ssize_t show_size(struct _cpuid4_info *this_leaf, char *buf) +static ssize_t show_size(struct _cpuid4_info *this_leaf, char *buf, + unsigned int cpu) { return sprintf(buf, "%luK\n", this_leaf->size / 1024); } @@ -906,17 +943,20 @@ static ssize_t show_shared_cpu_map_func(struct _cpuid4_info *this_leaf, return n; } -static inline ssize_t show_shared_cpu_map(struct _cpuid4_info *leaf, char *buf) +static inline ssize_t show_shared_cpu_map(struct _cpuid4_info *leaf, char *buf, + unsigned int cpu) { return show_shared_cpu_map_func(leaf, 0, buf); } -static inline ssize_t show_shared_cpu_list(struct _cpuid4_info *leaf, char *buf) +static inline ssize_t show_shared_cpu_list(struct _cpuid4_info *leaf, char *buf, + unsigned int cpu) { return show_shared_cpu_map_func(leaf, 1, buf); } -static ssize_t show_type(struct _cpuid4_info *this_leaf, char *buf) +static ssize_t show_type(struct _cpuid4_info *this_leaf, char *buf, + unsigned int cpu) { switch (this_leaf->eax.split.type) { case CACHE_TYPE_DATA: @@ -974,6 +1014,9 @@ static struct attribute ** __cpuinit amd_l3_attrs(void) if (amd_nb_has_feature(AMD_NB_L3_INDEX_DISABLE)) n += 2; + if (amd_nb_has_feature(AMD_NB_L3_PARTITIONING)) + n += 1; + attrs = kzalloc(n * sizeof (struct attribute *), GFP_KERNEL); if (attrs == NULL) return attrs = default_attrs; @@ -986,6 +1029,9 @@ static struct attribute ** __cpuinit amd_l3_attrs(void) attrs[n++] = &cache_disable_1.attr; } + if (amd_nb_has_feature(AMD_NB_L3_PARTITIONING)) + attrs[n++] = &subcaches.attr; + return attrs; } #endif @@ -998,7 +1044,7 @@ static ssize_t show(struct kobject *kobj, struct attribute *attr, char *buf) ret = fattr->show ? fattr->show(CPUID4_INFO_IDX(this_leaf->cpu, this_leaf->index), - buf) : + buf, this_leaf->cpu) : 0; return ret; } @@ -1012,7 +1058,7 @@ static ssize_t store(struct kobject *kobj, struct attribute *attr, ret = fattr->store ? fattr->store(CPUID4_INFO_IDX(this_leaf->cpu, this_leaf->index), - buf, count) : + buf, count, this_leaf->cpu) : 0; return ret; } diff --git a/arch/x86/kernel/e820.c b/arch/x86/kernel/e820.c index 294f26d..0b5e2b5 100644 --- a/arch/x86/kernel/e820.c +++ b/arch/x86/kernel/e820.c @@ -847,15 +847,21 @@ static int __init parse_memopt(char *p) if (!p) return -EINVAL; -#ifdef CONFIG_X86_32 if (!strcmp(p, "nopentium")) { +#ifdef CONFIG_X86_32 setup_clear_cpu_cap(X86_FEATURE_PSE); return 0; - } +#else + printk(KERN_WARNING "mem=nopentium ignored! (only supported on x86_32)\n"); + return -EINVAL; #endif + } userdef = 1; mem_size = memparse(p, &p); + /* don't remove all of memory when handling "mem={invalid}" param */ + if (mem_size == 0) + return -EINVAL; e820_remove_range(mem_size, ULLONG_MAX - mem_size, E820_RAM, 1); return 0; diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S index c32cbbc..b72b4a6 100644 --- a/arch/x86/kernel/entry_64.S +++ b/arch/x86/kernel/entry_64.S @@ -977,9 +977,12 @@ apicinterrupt X86_PLATFORM_IPI_VECTOR \ x86_platform_ipi smp_x86_platform_ipi #ifdef CONFIG_SMP -.irpc idx, "01234567" +.irp idx,0,1,2,3,4,5,6,7,8,9,10,11,12,13,14,15, \ + 16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31 +.if NUM_INVALIDATE_TLB_VECTORS > \idx apicinterrupt (INVALIDATE_TLB_VECTOR_START)+\idx \ invalidate_interrupt\idx smp_invalidate_interrupt +.endif .endr #endif diff --git a/arch/x86/kernel/irqinit.c b/arch/x86/kernel/irqinit.c index 1cc302d..d30854b 100644 --- a/arch/x86/kernel/irqinit.c +++ b/arch/x86/kernel/irqinit.c @@ -166,14 +166,77 @@ static void __init smp_intr_init(void) alloc_intr_gate(RESCHEDULE_VECTOR, reschedule_interrupt); /* IPIs for invalidation */ - alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+0, invalidate_interrupt0); - alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+1, invalidate_interrupt1); - alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+2, invalidate_interrupt2); - alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+3, invalidate_interrupt3); - alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+4, invalidate_interrupt4); - alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+5, invalidate_interrupt5); - alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+6, invalidate_interrupt6); - alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+7, invalidate_interrupt7); +#define ALLOC_INVTLB_VEC(NR) \ + alloc_intr_gate(INVALIDATE_TLB_VECTOR_START+NR, \ + invalidate_interrupt##NR) + + switch (NUM_INVALIDATE_TLB_VECTORS) { + default: + ALLOC_INVTLB_VEC(31); + case 31: + ALLOC_INVTLB_VEC(30); + case 30: + ALLOC_INVTLB_VEC(29); + case 29: + ALLOC_INVTLB_VEC(28); + case 28: + ALLOC_INVTLB_VEC(27); + case 27: + ALLOC_INVTLB_VEC(26); + case 26: + ALLOC_INVTLB_VEC(25); + case 25: + ALLOC_INVTLB_VEC(24); + case 24: + ALLOC_INVTLB_VEC(23); + case 23: + ALLOC_INVTLB_VEC(22); + case 22: + ALLOC_INVTLB_VEC(21); + case 21: + ALLOC_INVTLB_VEC(20); + case 20: + ALLOC_INVTLB_VEC(19); + case 19: + ALLOC_INVTLB_VEC(18); + case 18: + ALLOC_INVTLB_VEC(17); + case 17: + ALLOC_INVTLB_VEC(16); + case 16: + ALLOC_INVTLB_VEC(15); + case 15: + ALLOC_INVTLB_VEC(14); + case 14: + ALLOC_INVTLB_VEC(13); + case 13: + ALLOC_INVTLB_VEC(12); + case 12: + ALLOC_INVTLB_VEC(11); + case 11: + ALLOC_INVTLB_VEC(10); + case 10: + ALLOC_INVTLB_VEC(9); + case 9: + ALLOC_INVTLB_VEC(8); + case 8: + ALLOC_INVTLB_VEC(7); + case 7: + ALLOC_INVTLB_VEC(6); + case 6: + ALLOC_INVTLB_VEC(5); + case 5: + ALLOC_INVTLB_VEC(4); + case 4: + ALLOC_INVTLB_VEC(3); + case 3: + ALLOC_INVTLB_VEC(2); + case 2: + ALLOC_INVTLB_VEC(1); + case 1: + ALLOC_INVTLB_VEC(0); + break; + } /* IPI for generic function call */ alloc_intr_gate(CALL_FUNCTION_VECTOR, call_function_interrupt); diff --git a/arch/x86/kernel/setup.c b/arch/x86/kernel/setup.c index d3cfe26c..c3a606c 100644 --- a/arch/x86/kernel/setup.c +++ b/arch/x86/kernel/setup.c @@ -293,10 +293,32 @@ static void __init init_gbpages(void) else direct_gbpages = 0; } + +static void __init cleanup_highmap_brk_end(void) +{ + pud_t *pud; + pmd_t *pmd; + + mmu_cr4_features = read_cr4(); + + /* + * _brk_end cannot change anymore, but it and _end may be + * located on different 2M pages. cleanup_highmap(), however, + * can only consider _end when it runs, so destroy any + * mappings beyond _brk_end here. + */ + pud = pud_offset(pgd_offset_k(_brk_end), _brk_end); + pmd = pmd_offset(pud, _brk_end - 1); + while (++pmd <= pmd_offset(pud, (unsigned long)_end - 1)) + pmd_clear(pmd); +} #else static inline void init_gbpages(void) { } +static inline void cleanup_highmap_brk_end(void) +{ +} #endif static void __init reserve_brk(void) @@ -307,6 +329,8 @@ static void __init reserve_brk(void) /* Mark brk area as locked down and no longer taking any new allocations */ _brk_start = 0; + + cleanup_highmap_brk_end(); } #ifdef CONFIG_BLK_DEV_INITRD @@ -680,15 +704,6 @@ static int __init parse_reservelow(char *p) early_param("reservelow", parse_reservelow); -static u64 __init get_max_mapped(void) -{ - u64 end = max_pfn_mapped; - - end <<= PAGE_SHIFT; - - return end; -} - /* * Determine if we were loaded by an EFI loader. If so, then we have also been * passed the efi memmap, systab, etc., so we should use these data structures @@ -704,8 +719,6 @@ static u64 __init get_max_mapped(void) void __init setup_arch(char **cmdline_p) { - int acpi = 0; - int amd = 0; unsigned long flags; #ifdef CONFIG_X86_32 @@ -984,19 +997,7 @@ void __init setup_arch(char **cmdline_p) early_acpi_boot_init(); -#ifdef CONFIG_ACPI_NUMA - /* - * Parse SRAT to discover nodes. - */ - acpi = acpi_numa_init(); -#endif - -#ifdef CONFIG_AMD_NUMA - if (!acpi) - amd = !amd_numa_init(0, max_pfn); -#endif - - initmem_init(0, max_pfn, acpi, amd); + initmem_init(); memblock_find_dma_reserve(); dma32_reserve_bootmem(); @@ -1040,9 +1041,7 @@ void __init setup_arch(char **cmdline_p) prefill_possible_map(); -#ifdef CONFIG_X86_64 init_cpu_to_node(); -#endif init_apic_mappings(); ioapic_and_gsi_init(); diff --git a/arch/x86/kernel/setup_percpu.c b/arch/x86/kernel/setup_percpu.c index 002b796..71f4727 100644 --- a/arch/x86/kernel/setup_percpu.c +++ b/arch/x86/kernel/setup_percpu.c @@ -225,10 +225,15 @@ void __init setup_per_cpu_areas(void) per_cpu(x86_bios_cpu_apicid, cpu) = early_per_cpu_map(x86_bios_cpu_apicid, cpu); #endif +#ifdef CONFIG_X86_32 + per_cpu(x86_cpu_to_logical_apicid, cpu) = + early_per_cpu_map(x86_cpu_to_logical_apicid, cpu); +#endif #ifdef CONFIG_X86_64 per_cpu(irq_stack_ptr, cpu) = per_cpu(irq_stack_union.irq_stack, cpu) + IRQ_STACK_SIZE - 64; +#endif #ifdef CONFIG_NUMA per_cpu(x86_cpu_to_node_map, cpu) = early_per_cpu_map(x86_cpu_to_node_map, cpu); @@ -242,7 +247,6 @@ void __init setup_per_cpu_areas(void) */ set_cpu_numa_node(cpu, early_cpu_to_node(cpu)); #endif -#endif /* * Up to this point, the boot CPU has been using .init.data * area. Reload any changed state for the boot CPU. @@ -256,7 +260,10 @@ void __init setup_per_cpu_areas(void) early_per_cpu_ptr(x86_cpu_to_apicid) = NULL; early_per_cpu_ptr(x86_bios_cpu_apicid) = NULL; #endif -#if defined(CONFIG_X86_64) && defined(CONFIG_NUMA) +#ifdef CONFIG_X86_32 + early_per_cpu_ptr(x86_cpu_to_logical_apicid) = NULL; +#endif +#ifdef CONFIG_NUMA early_per_cpu_ptr(x86_cpu_to_node_map) = NULL; #endif diff --git a/arch/x86/kernel/smpboot.c b/arch/x86/kernel/smpboot.c index 16ce426..e9efdfd 100644 --- a/arch/x86/kernel/smpboot.c +++ b/arch/x86/kernel/smpboot.c @@ -72,10 +72,6 @@ #include <asm/smpboot_hooks.h> #include <asm/i8259.h> -#ifdef CONFIG_X86_32 -u8 apicid_2_node[MAX_APICID]; -#endif - /* State of each CPU */ DEFINE_PER_CPU(int, cpu_state) = { 0 }; @@ -139,62 +135,6 @@ EXPORT_PER_CPU_SYMBOL(cpu_info); atomic_t init_deasserted; -#if defined(CONFIG_NUMA) && defined(CONFIG_X86_32) -/* 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); - -/* set up a mapping between cpu and node. */ -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]); - cpu_to_node_map[cpu] = node; -} - -/* undo a mapping between cpu and node. */ -static void unmap_cpu_to_node(int cpu) -{ - int node; - - 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]); - cpu_to_node_map[cpu] = 0; -} -#else /* !(CONFIG_NUMA && CONFIG_X86_32) */ -#define map_cpu_to_node(cpu, node) ({}) -#define unmap_cpu_to_node(cpu) ({}) -#endif - -#ifdef CONFIG_X86_32 -static int boot_cpu_logical_apicid; - -u8 cpu_2_logical_apicid[NR_CPUS] __read_mostly = - { [0 ... NR_CPUS-1] = BAD_APICID }; - -static void map_cpu_to_logical_apicid(void) -{ - int cpu = smp_processor_id(); - int apicid = logical_smp_processor_id(); - int node = apic->apicid_to_node(apicid); - - if (!node_online(node)) - node = first_online_node; - - cpu_2_logical_apicid[cpu] = apicid; - map_cpu_to_node(cpu, node); -} - -void numa_remove_cpu(int cpu) -{ - cpu_2_logical_apicid[cpu] = BAD_APICID; - unmap_cpu_to_node(cpu); -} -#else -#define map_cpu_to_logical_apicid() do {} while (0) -#endif - /* * Report back to the Boot Processor. * Running on AP. @@ -262,7 +202,6 @@ static void __cpuinit smp_callin(void) apic->smp_callin_clear_local_apic(); setup_local_APIC(); end_local_APIC_setup(); - map_cpu_to_logical_apicid(); /* * Need to setup vector mappings before we enable interrupts. @@ -397,6 +336,7 @@ void __cpuinit set_cpu_sibling_map(int cpu) if (cpu_has(c, X86_FEATURE_TOPOEXT)) { if (c->phys_proc_id == o->phys_proc_id && + per_cpu(cpu_llc_id, cpu) == per_cpu(cpu_llc_id, i) && c->compute_unit_id == o->compute_unit_id) link_thread_siblings(cpu, i); } else if (c->phys_proc_id == o->phys_proc_id && @@ -951,7 +891,6 @@ static __init void disable_smp(void) physid_set_mask_of_physid(boot_cpu_physical_apicid, &phys_cpu_present_map); else physid_set_mask_of_physid(0, &phys_cpu_present_map); - map_cpu_to_logical_apicid(); cpumask_set_cpu(0, cpu_sibling_mask(0)); cpumask_set_cpu(0, cpu_core_mask(0)); } @@ -1087,9 +1026,7 @@ void __init native_smp_prepare_cpus(unsigned int max_cpus) smp_store_cpu_info(0); /* Final full version of the data */ cpumask_copy(cpu_callin_mask, cpumask_of(0)); mb(); -#ifdef CONFIG_X86_32 - boot_cpu_logical_apicid = logical_smp_processor_id(); -#endif + current_thread_info()->cpu = 0; /* needed? */ for_each_possible_cpu(i) { zalloc_cpumask_var(&per_cpu(cpu_sibling_map, i), GFP_KERNEL); @@ -1130,8 +1067,6 @@ void __init native_smp_prepare_cpus(unsigned int max_cpus) bsp_end_local_APIC_setup(); - map_cpu_to_logical_apicid(); - if (apic->setup_portio_remap) apic->setup_portio_remap(); diff --git a/arch/x86/mm/Makefile b/arch/x86/mm/Makefile index 09df2f9..3e608ed 100644 --- a/arch/x86/mm/Makefile +++ b/arch/x86/mm/Makefile @@ -25,6 +25,7 @@ obj-$(CONFIG_MMIOTRACE_TEST) += testmmiotrace.o obj-$(CONFIG_NUMA) += numa.o numa_$(BITS).o obj-$(CONFIG_AMD_NUMA) += amdtopology_64.o obj-$(CONFIG_ACPI_NUMA) += srat_$(BITS).o +obj-$(CONFIG_NUMA_EMU) += numa_emulation.o obj-$(CONFIG_HAVE_MEMBLOCK) += memblock.o diff --git a/arch/x86/mm/amdtopology_64.c b/arch/x86/mm/amdtopology_64.c index f21962c..0919c26 100644 --- a/arch/x86/mm/amdtopology_64.c +++ b/arch/x86/mm/amdtopology_64.c @@ -26,9 +26,7 @@ #include <asm/apic.h> #include <asm/amd_nb.h> -static struct bootnode __initdata nodes[8]; static unsigned char __initdata nodeids[8]; -static nodemask_t __initdata nodes_parsed = NODE_MASK_NONE; static __init int find_northbridge(void) { @@ -51,7 +49,7 @@ static __init int find_northbridge(void) return num; } - return -1; + return -ENOENT; } static __init void early_get_boot_cpu_id(void) @@ -69,17 +67,18 @@ static __init void early_get_boot_cpu_id(void) #endif } -int __init amd_numa_init(unsigned long start_pfn, unsigned long end_pfn) +int __init amd_numa_init(void) { - unsigned long start = PFN_PHYS(start_pfn); - unsigned long end = PFN_PHYS(end_pfn); + unsigned long start = PFN_PHYS(0); + unsigned long end = PFN_PHYS(max_pfn); unsigned numnodes; unsigned long prevbase; - int i, nb, found = 0; + int i, j, nb; u32 nodeid, reg; + unsigned int bits, cores, apicid_base; if (!early_pci_allowed()) - return -1; + return -EINVAL; nb = find_northbridge(); if (nb < 0) @@ -90,7 +89,7 @@ int __init amd_numa_init(unsigned long start_pfn, unsigned long end_pfn) reg = read_pci_config(0, nb, 0, 0x60); numnodes = ((reg >> 4) & 0xF) + 1; if (numnodes <= 1) - return -1; + return -ENOENT; pr_info("Number of physical nodes %d\n", numnodes); @@ -121,9 +120,9 @@ int __init amd_numa_init(unsigned long start_pfn, unsigned long end_pfn) if ((base >> 8) & 3 || (limit >> 8) & 3) { pr_err("Node %d using interleaving mode %lx/%lx\n", nodeid, (base >> 8) & 3, (limit >> 8) & 3); - return -1; + return -EINVAL; } - if (node_isset(nodeid, nodes_parsed)) { + if (node_isset(nodeid, numa_nodes_parsed)) { pr_info("Node %d already present, skipping\n", nodeid); continue; @@ -160,117 +159,28 @@ int __init amd_numa_init(unsigned long start_pfn, unsigned long end_pfn) if (prevbase > base) { pr_err("Node map not sorted %lx,%lx\n", prevbase, base); - return -1; + return -EINVAL; } pr_info("Node %d MemBase %016lx Limit %016lx\n", nodeid, base, limit); - found++; - - nodes[nodeid].start = base; - nodes[nodeid].end = limit; - prevbase = base; - - node_set(nodeid, nodes_parsed); - } - - if (!found) - return -1; - return 0; -} - -#ifdef CONFIG_NUMA_EMU -static s16 fake_apicid_to_node[MAX_LOCAL_APIC] __initdata = { - [0 ... MAX_LOCAL_APIC-1] = NUMA_NO_NODE -}; - -void __init amd_get_nodes(struct bootnode *physnodes) -{ - int i; - - for_each_node_mask(i, nodes_parsed) { - physnodes[i].start = nodes[i].start; - physnodes[i].end = nodes[i].end; + numa_add_memblk(nodeid, base, limit); + node_set(nodeid, numa_nodes_parsed); } -} - -static int __init find_node_by_addr(unsigned long addr) -{ - int ret = NUMA_NO_NODE; - int i; - - for (i = 0; i < 8; i++) - if (addr >= nodes[i].start && addr < nodes[i].end) { - ret = i; - break; - } - return ret; -} -/* - * For NUMA emulation, fake proximity domain (_PXM) to node id mappings must be - * setup to represent the physical topology but reflect the emulated - * environment. For each emulated node, the real node which it appears on is - * found and a fake pxm to nid mapping is created which mirrors the actual - * locality. node_distance() then represents the correct distances between - * emulated nodes by using the fake acpi mappings to pxms. - */ -void __init amd_fake_nodes(const struct bootnode *nodes, int nr_nodes) -{ - unsigned int bits; - unsigned int cores; - unsigned int apicid_base = 0; - int i; + if (!nodes_weight(numa_nodes_parsed)) + return -ENOENT; + /* + * We seem to have valid NUMA configuration. Map apicids to nodes + * using the coreid bits from early_identify_cpu. + */ bits = boot_cpu_data.x86_coreid_bits; cores = 1 << bits; - early_get_boot_cpu_id(); - if (boot_cpu_physical_apicid > 0) - apicid_base = boot_cpu_physical_apicid; - - for (i = 0; i < nr_nodes; i++) { - int index; - int nid; - int j; - - nid = find_node_by_addr(nodes[i].start); - if (nid == NUMA_NO_NODE) - continue; - - index = nodeids[nid] << bits; - if (fake_apicid_to_node[index + apicid_base] == NUMA_NO_NODE) - for (j = apicid_base; j < cores + apicid_base; j++) - fake_apicid_to_node[index + j] = i; -#ifdef CONFIG_ACPI_NUMA - __acpi_map_pxm_to_node(nid, i); -#endif - } - memcpy(apicid_to_node, fake_apicid_to_node, sizeof(apicid_to_node)); -} -#endif /* CONFIG_NUMA_EMU */ - -int __init amd_scan_nodes(void) -{ - unsigned int bits; - unsigned int cores; - unsigned int apicid_base; - int i; - - BUG_ON(nodes_empty(nodes_parsed)); - node_possible_map = nodes_parsed; - memnode_shift = compute_hash_shift(nodes, 8, NULL); - if (memnode_shift < 0) { - pr_err("No NUMA node hash function found. Contact maintainer\n"); - return -1; - } - pr_info("Using node hash shift of %d\n", memnode_shift); - - /* use the coreid bits from early_identify_cpu */ - bits = boot_cpu_data.x86_coreid_bits; - cores = (1<<bits); apicid_base = 0; + /* get the APIC ID of the BSP early for systems with apicid lifting */ early_get_boot_cpu_id(); if (boot_cpu_physical_apicid > 0) { @@ -278,17 +188,9 @@ int __init amd_scan_nodes(void) apicid_base = boot_cpu_physical_apicid; } - for_each_node_mask(i, node_possible_map) { - int j; - - memblock_x86_register_active_regions(i, - nodes[i].start >> PAGE_SHIFT, - nodes[i].end >> PAGE_SHIFT); + for_each_node_mask(i, numa_nodes_parsed) for (j = apicid_base; j < cores + apicid_base; j++) - apicid_to_node[(i << bits) + j] = i; - setup_node_bootmem(i, nodes[i].start, nodes[i].end); - } + set_apicid_to_node((i << bits) + j, i); - numa_init_array(); return 0; } diff --git a/arch/x86/mm/init.c b/arch/x86/mm/init.c index 947f42ab..286d289 100644 --- a/arch/x86/mm/init.c +++ b/arch/x86/mm/init.c @@ -18,9 +18,9 @@ DEFINE_PER_CPU(struct mmu_gather, mmu_gathers); -unsigned long __initdata e820_table_start; -unsigned long __meminitdata e820_table_end; -unsigned long __meminitdata e820_table_top; +unsigned long __initdata pgt_buf_start; +unsigned long __meminitdata pgt_buf_end; +unsigned long __meminitdata pgt_buf_top; int after_bootmem; @@ -33,7 +33,7 @@ int direct_gbpages static void __init find_early_table_space(unsigned long end, int use_pse, int use_gbpages) { - unsigned long puds, pmds, ptes, tables, start; + unsigned long puds, pmds, ptes, tables, start = 0, good_end = end; phys_addr_t base; puds = (end + PUD_SIZE - 1) >> PUD_SHIFT; @@ -65,29 +65,20 @@ static void __init find_early_table_space(unsigned long end, int use_pse, #ifdef CONFIG_X86_32 /* for fixmap */ tables += roundup(__end_of_fixed_addresses * sizeof(pte_t), PAGE_SIZE); -#endif - /* - * RED-PEN putting page tables only on node 0 could - * cause a hotspot and fill up ZONE_DMA. The page tables - * need roughly 0.5KB per GB. - */ -#ifdef CONFIG_X86_32 - start = 0x7000; -#else - start = 0x8000; + good_end = max_pfn_mapped << PAGE_SHIFT; #endif - base = memblock_find_in_range(start, max_pfn_mapped<<PAGE_SHIFT, - tables, PAGE_SIZE); + + base = memblock_find_in_range(start, good_end, tables, PAGE_SIZE); if (base == MEMBLOCK_ERROR) panic("Cannot find space for the kernel page tables"); - e820_table_start = base >> PAGE_SHIFT; - e820_table_end = e820_table_start; - e820_table_top = e820_table_start + (tables >> PAGE_SHIFT); + pgt_buf_start = base >> PAGE_SHIFT; + pgt_buf_end = pgt_buf_start; + pgt_buf_top = pgt_buf_start + (tables >> PAGE_SHIFT); printk(KERN_DEBUG "kernel direct mapping tables up to %lx @ %lx-%lx\n", - end, e820_table_start << PAGE_SHIFT, e820_table_top << PAGE_SHIFT); + end, pgt_buf_start << PAGE_SHIFT, pgt_buf_top << PAGE_SHIFT); } struct map_range { @@ -279,30 +270,11 @@ unsigned long __init_refok init_memory_mapping(unsigned long start, load_cr3(swapper_pg_dir); #endif -#ifdef CONFIG_X86_64 - if (!after_bootmem && !start) { - pud_t *pud; - pmd_t *pmd; - - mmu_cr4_features = read_cr4(); - - /* - * _brk_end cannot change anymore, but it and _end may be - * located on different 2M pages. cleanup_highmap(), however, - * can only consider _end when it runs, so destroy any - * mappings beyond _brk_end here. - */ - pud = pud_offset(pgd_offset_k(_brk_end), _brk_end); - pmd = pmd_offset(pud, _brk_end - 1); - while (++pmd <= pmd_offset(pud, (unsigned long)_end - 1)) - pmd_clear(pmd); - } -#endif __flush_tlb_all(); - if (!after_bootmem && e820_table_end > e820_table_start) - memblock_x86_reserve_range(e820_table_start << PAGE_SHIFT, - e820_table_end << PAGE_SHIFT, "PGTABLE"); + if (!after_bootmem && pgt_buf_end > pgt_buf_start) + memblock_x86_reserve_range(pgt_buf_start << PAGE_SHIFT, + pgt_buf_end << PAGE_SHIFT, "PGTABLE"); if (!after_bootmem) early_memtest(start, end); diff --git a/arch/x86/mm/init_32.c b/arch/x86/mm/init_32.c index c821074..73ad7eb 100644 --- a/arch/x86/mm/init_32.c +++ b/arch/x86/mm/init_32.c @@ -62,10 +62,10 @@ bool __read_mostly __vmalloc_start_set = false; static __init void *alloc_low_page(void) { - unsigned long pfn = e820_table_end++; + unsigned long pfn = pgt_buf_end++; void *adr; - if (pfn >= e820_table_top) + if (pfn >= pgt_buf_top) panic("alloc_low_page: ran out of memory"); adr = __va(pfn * PAGE_SIZE); @@ -163,8 +163,8 @@ static pte_t *__init page_table_kmap_check(pte_t *pte, pmd_t *pmd, if (pmd_idx_kmap_begin != pmd_idx_kmap_end && (vaddr >> PMD_SHIFT) >= pmd_idx_kmap_begin && (vaddr >> PMD_SHIFT) <= pmd_idx_kmap_end - && ((__pa(pte) >> PAGE_SHIFT) < e820_table_start - || (__pa(pte) >> PAGE_SHIFT) >= e820_table_end)) { + && ((__pa(pte) >> PAGE_SHIFT) < pgt_buf_start + || (__pa(pte) >> PAGE_SHIFT) >= pgt_buf_end)) { pte_t *newpte; int i; @@ -644,8 +644,7 @@ void __init find_low_pfn_range(void) } #ifndef CONFIG_NEED_MULTIPLE_NODES -void __init initmem_init(unsigned long start_pfn, unsigned long end_pfn, - int acpi, int k8) +void __init initmem_init(void) { #ifdef CONFIG_HIGHMEM highstart_pfn = highend_pfn = max_pfn; diff --git a/arch/x86/mm/init_64.c b/arch/x86/mm/init_64.c index c14a542..a08a62c 100644 --- a/arch/x86/mm/init_64.c +++ b/arch/x86/mm/init_64.c @@ -314,7 +314,7 @@ void __init cleanup_highmap(void) static __ref void *alloc_low_page(unsigned long *phys) { - unsigned long pfn = e820_table_end++; + unsigned long pfn = pgt_buf_end++; void *adr; if (after_bootmem) { @@ -324,7 +324,7 @@ static __ref void *alloc_low_page(unsigned long *phys) return adr; } - if (pfn >= e820_table_top) + if (pfn >= pgt_buf_top) panic("alloc_low_page: ran out of memory"); adr = early_memremap(pfn * PAGE_SIZE, PAGE_SIZE); @@ -333,12 +333,28 @@ static __ref void *alloc_low_page(unsigned long *phys) return adr; } +static __ref void *map_low_page(void *virt) +{ + void *adr; + unsigned long phys, left; + + if (after_bootmem) + return virt; + + phys = __pa(virt); + left = phys & (PAGE_SIZE - 1); + adr = early_memremap(phys & PAGE_MASK, PAGE_SIZE); + adr = (void *)(((unsigned long)adr) | left); + + return adr; +} + static __ref void unmap_low_page(void *adr) { if (after_bootmem) return; - early_iounmap(adr, PAGE_SIZE); + early_iounmap((void *)((unsigned long)adr & PAGE_MASK), PAGE_SIZE); } static unsigned long __meminit @@ -386,15 +402,6 @@ phys_pte_init(pte_t *pte_page, unsigned long addr, unsigned long end, } static unsigned long __meminit -phys_pte_update(pmd_t *pmd, unsigned long address, unsigned long end, - pgprot_t prot) -{ - pte_t *pte = (pte_t *)pmd_page_vaddr(*pmd); - - return phys_pte_init(pte, address, end, prot); -} - -static unsigned long __meminit phys_pmd_init(pmd_t *pmd_page, unsigned long address, unsigned long end, unsigned long page_size_mask, pgprot_t prot) { @@ -420,8 +427,10 @@ phys_pmd_init(pmd_t *pmd_page, unsigned long address, unsigned long end, if (pmd_val(*pmd)) { if (!pmd_large(*pmd)) { spin_lock(&init_mm.page_table_lock); - last_map_addr = phys_pte_update(pmd, address, + pte = map_low_page((pte_t *)pmd_page_vaddr(*pmd)); + last_map_addr = phys_pte_init(pte, address, end, prot); + unmap_low_page(pte); spin_unlock(&init_mm.page_table_lock); continue; } @@ -468,18 +477,6 @@ phys_pmd_init(pmd_t *pmd_page, unsigned long address, unsigned long end, } static unsigned long __meminit -phys_pmd_update(pud_t *pud, unsigned long address, unsigned long end, - unsigned long page_size_mask, pgprot_t prot) -{ - pmd_t *pmd = pmd_offset(pud, 0); - unsigned long last_map_addr; - - last_map_addr = phys_pmd_init(pmd, address, end, page_size_mask, prot); - __flush_tlb_all(); - return last_map_addr; -} - -static unsigned long __meminit phys_pud_init(pud_t *pud_page, unsigned long addr, unsigned long end, unsigned long page_size_mask) { @@ -504,8 +501,11 @@ phys_pud_init(pud_t *pud_page, unsigned long addr, unsigned long end, if (pud_val(*pud)) { if (!pud_large(*pud)) { - last_map_addr = phys_pmd_update(pud, addr, end, + pmd = map_low_page(pmd_offset(pud, 0)); + last_map_addr = phys_pmd_init(pmd, addr, end, page_size_mask, prot); + unmap_low_page(pmd); + __flush_tlb_all(); continue; } /* @@ -553,17 +553,6 @@ phys_pud_init(pud_t *pud_page, unsigned long addr, unsigned long end, return last_map_addr; } -static unsigned long __meminit -phys_pud_update(pgd_t *pgd, unsigned long addr, unsigned long end, - unsigned long page_size_mask) -{ - pud_t *pud; - - pud = (pud_t *)pgd_page_vaddr(*pgd); - - return phys_pud_init(pud, addr, end, page_size_mask); -} - unsigned long __meminit kernel_physical_mapping_init(unsigned long start, unsigned long end, @@ -587,8 +576,10 @@ kernel_physical_mapping_init(unsigned long start, next = end; if (pgd_val(*pgd)) { - last_map_addr = phys_pud_update(pgd, __pa(start), + pud = map_low_page((pud_t *)pgd_page_vaddr(*pgd)); + last_map_addr = phys_pud_init(pud, __pa(start), __pa(end), page_size_mask); + unmap_low_page(pud); continue; } @@ -612,10 +603,9 @@ kernel_physical_mapping_init(unsigned long start, } #ifndef CONFIG_NUMA -void __init initmem_init(unsigned long start_pfn, unsigned long end_pfn, - int acpi, int k8) +void __init initmem_init(void) { - memblock_x86_register_active_regions(0, start_pfn, end_pfn); + memblock_x86_register_active_regions(0, 0, max_pfn); } #endif diff --git a/arch/x86/mm/numa.c b/arch/x86/mm/numa.c index ebf6d78..9559d36 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 84a3e4c..bde3906 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 @@ -346,8 +352,7 @@ static void init_remap_allocator(int nid) (ulong) node_remap_end_vaddr[nid]); } -void __init initmem_init(unsigned long start_pfn, unsigned long end_pfn, - int acpi, int k8) +void __init initmem_init(void) { int nid; long kva_target_pfn; @@ -361,6 +366,7 @@ 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 1337c51..9ec0f20 100644 --- a/arch/x86/mm/numa_64.c +++ b/arch/x86/mm/numa_64.c @@ -13,31 +13,30 @@ #include <linux/module.h> #include <linux/nodemask.h> #include <linux/sched.h> +#include <linux/acpi.h> #include <asm/e820.h> #include <asm/proto.h> #include <asm/dma.h> -#include <asm/numa.h> #include <asm/acpi.h> #include <asm/amd_nb.h> +#include "numa_internal.h" + struct pglist_data *node_data[MAX_NUMNODES] __read_mostly; EXPORT_SYMBOL(node_data); -struct memnode memnode; +nodemask_t numa_nodes_parsed __initdata; -s16 apicid_to_node[MAX_LOCAL_APIC] __cpuinitdata = { - [0 ... MAX_LOCAL_APIC-1] = NUMA_NO_NODE -}; +struct memnode memnode; 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); +static struct numa_meminfo numa_meminfo __initdata; + +static int numa_distance_cnt; +static u8 *numa_distance; /* * Given a shift value, try to populate memnodemap[] @@ -46,16 +45,15 @@ EXPORT_EARLY_PER_CPU_SYMBOL(x86_cpu_to_node_map); * 0 if memnodmap[] too small (of shift too small) * -1 if node overlap or lost ram (shift too big) */ -static int __init populate_memnodemap(const struct bootnode *nodes, - int numnodes, int shift, int *nodeids) +static int __init populate_memnodemap(const struct numa_meminfo *mi, int shift) { unsigned long addr, end; int i, res = -1; memset(memnodemap, 0xff, sizeof(s16)*memnodemapsize); - for (i = 0; i < numnodes; i++) { - addr = nodes[i].start; - end = nodes[i].end; + for (i = 0; i < mi->nr_blks; i++) { + addr = mi->blk[i].start; + end = mi->blk[i].end; if (addr >= end) continue; if ((end >> shift) >= memnodemapsize) @@ -63,12 +61,7 @@ static int __init populate_memnodemap(const struct bootnode *nodes, do { if (memnodemap[addr >> shift] != NUMA_NO_NODE) return -1; - - if (!nodeids) - memnodemap[addr >> shift] = i; - else - memnodemap[addr >> shift] = nodeids[i]; - + memnodemap[addr >> shift] = mi->blk[i].nid; addr += (1UL << shift); } while (addr < end); res = 1; @@ -86,7 +79,7 @@ static int __init allocate_cachealigned_memnodemap(void) addr = 0x8000; nodemap_size = roundup(sizeof(s16) * memnodemapsize, L1_CACHE_BYTES); - nodemap_addr = memblock_find_in_range(addr, max_pfn<<PAGE_SHIFT, + nodemap_addr = memblock_find_in_range(addr, get_max_mapped(), nodemap_size, L1_CACHE_BYTES); if (nodemap_addr == MEMBLOCK_ERROR) { printk(KERN_ERR @@ -106,16 +99,15 @@ static int __init allocate_cachealigned_memnodemap(void) * The LSB of all start and end addresses in the node map is the value of the * maximum possible shift. */ -static int __init extract_lsb_from_nodes(const struct bootnode *nodes, - int numnodes) +static int __init extract_lsb_from_nodes(const struct numa_meminfo *mi) { int i, nodes_used = 0; unsigned long start, end; unsigned long bitfield = 0, memtop = 0; - for (i = 0; i < numnodes; i++) { - start = nodes[i].start; - end = nodes[i].end; + for (i = 0; i < mi->nr_blks; i++) { + start = mi->blk[i].start; + end = mi->blk[i].end; if (start >= end) continue; bitfield |= start; @@ -131,18 +123,17 @@ static int __init extract_lsb_from_nodes(const struct bootnode *nodes, return i; } -int __init compute_hash_shift(struct bootnode *nodes, int numnodes, - int *nodeids) +static int __init compute_hash_shift(const struct numa_meminfo *mi) { int shift; - shift = extract_lsb_from_nodes(nodes, numnodes); + shift = extract_lsb_from_nodes(mi); if (allocate_cachealigned_memnodemap()) return -1; printk(KERN_DEBUG "NUMA: Using %d for the hash shift.\n", shift); - if (populate_memnodemap(nodes, numnodes, shift, nodeids) != 1) { + if (populate_memnodemap(mi, shift) != 1) { printk(KERN_INFO "Your memory is not aligned you need to " "rebuild your kernel with a bigger NODEMAPSIZE " "shift=%d\n", shift); @@ -188,6 +179,63 @@ static void * __init early_node_mem(int nodeid, unsigned long start, return NULL; } +static int __init numa_add_memblk_to(int nid, u64 start, u64 end, + struct numa_meminfo *mi) +{ + /* ignore zero length blks */ + if (start == end) + return 0; + + /* whine about and ignore invalid blks */ + if (start > end || nid < 0 || nid >= MAX_NUMNODES) { + pr_warning("NUMA: Warning: invalid memblk node %d (%Lx-%Lx)\n", + nid, start, end); + return 0; + } + + if (mi->nr_blks >= NR_NODE_MEMBLKS) { + pr_err("NUMA: too many memblk ranges\n"); + return -EINVAL; + } + + mi->blk[mi->nr_blks].start = start; + mi->blk[mi->nr_blks].end = end; + mi->blk[mi->nr_blks].nid = nid; + mi->nr_blks++; + return 0; +} + +/** + * numa_remove_memblk_from - Remove one numa_memblk from a numa_meminfo + * @idx: Index of memblk to remove + * @mi: numa_meminfo to remove memblk from + * + * Remove @idx'th numa_memblk from @mi by shifting @mi->blk[] and + * decrementing @mi->nr_blks. + */ +void __init numa_remove_memblk_from(int idx, struct numa_meminfo *mi) +{ + mi->nr_blks--; + memmove(&mi->blk[idx], &mi->blk[idx + 1], + (mi->nr_blks - idx) * sizeof(mi->blk[0])); +} + +/** + * numa_add_memblk - Add one numa_memblk to numa_meminfo + * @nid: NUMA node ID of the new memblk + * @start: Start address of the new memblk + * @end: End address of the new memblk + * + * Add a new memblk to the default numa_meminfo. + * + * RETURNS: + * 0 on success, -errno on failure. + */ +int __init numa_add_memblk(int nid, u64 start, u64 end) +{ + return numa_add_memblk_to(nid, start, end, &numa_meminfo); +} + /* Initialize bootmem allocator for a node */ void __init setup_node_bootmem(int nodeid, unsigned long start, unsigned long end) @@ -234,692 +282,386 @@ 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. +/** + * numa_cleanup_meminfo - Cleanup a numa_meminfo + * @mi: numa_meminfo to clean up + * + * Sanitize @mi by merging and removing unncessary memblks. Also check for + * conflicts and clear unused memblks. + * + * RETURNS: + * 0 on success, -errno on failure. */ -void __init numa_init_array(void) +int __init numa_cleanup_meminfo(struct numa_meminfo *mi) { - int rr, i; + const u64 low = 0; + const u64 high = (u64)max_pfn << PAGE_SHIFT; + int i, j, k; - 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; -static struct bootnode physnodes[MAX_NUMNODES] __cpuinitdata; -static char *cmdline __initdata; + for (i = 0; i < mi->nr_blks; i++) { + struct numa_memblk *bi = &mi->blk[i]; -void __init numa_emu_cmdline(char *str) -{ - cmdline = str; -} + /* make sure all blocks are inside the limits */ + bi->start = max(bi->start, low); + bi->end = min(bi->end, high); -static int __init setup_physnodes(unsigned long start, unsigned long end, - int acpi, int amd) -{ - int ret = 0; - int i; - - memset(physnodes, 0, sizeof(physnodes)); -#ifdef CONFIG_ACPI_NUMA - if (acpi) - acpi_get_nodes(physnodes, start, end); -#endif -#ifdef CONFIG_AMD_NUMA - if (amd) - amd_get_nodes(physnodes); -#endif - /* - * Basic sanity checking on the physical node map: there may be errors - * if the SRAT or AMD code incorrectly reported the topology or the mem= - * kernel parameter is used. - */ - for (i = 0; i < MAX_NUMNODES; i++) { - if (physnodes[i].start == physnodes[i].end) - continue; - if (physnodes[i].start > end) { - physnodes[i].end = physnodes[i].start; - continue; - } - if (physnodes[i].end < start) { - physnodes[i].start = physnodes[i].end; + /* and there's no empty block */ + if (bi->start == bi->end) { + numa_remove_memblk_from(i--, mi); continue; } - if (physnodes[i].start < start) - physnodes[i].start = start; - if (physnodes[i].end > end) - physnodes[i].end = end; - ret++; - } - /* - * If no physical topology was detected, a single node is faked to cover - * the entire address space. - */ - if (!ret) { - physnodes[ret].start = start; - physnodes[ret].end = end; - ret = 1; - } - return ret; -} - -static void __init fake_physnodes(int acpi, int amd, int nr_nodes) -{ - int i; - - BUG_ON(acpi && amd); -#ifdef CONFIG_ACPI_NUMA - if (acpi) - acpi_fake_nodes(nodes, nr_nodes); -#endif -#ifdef CONFIG_AMD_NUMA - if (amd) - amd_fake_nodes(nodes, nr_nodes); -#endif - if (!acpi && !amd) - for (i = 0; i < nr_cpu_ids; i++) - numa_set_node(i, 0); -} - -/* - * Setups up nid to range from addr to addr + size. If the end - * boundary is greater than max_addr, then max_addr is used instead. - * The return value is 0 if there is additional memory left for - * allocation past addr and -1 otherwise. addr is adjusted to be at - * the end of the node. - */ -static int __init setup_node_range(int nid, u64 *addr, u64 size, u64 max_addr) -{ - int ret = 0; - nodes[nid].start = *addr; - *addr += size; - if (*addr >= max_addr) { - *addr = max_addr; - ret = -1; - } - nodes[nid].end = *addr; - node_set(nid, node_possible_map); - printk(KERN_INFO "Faking node %d at %016Lx-%016Lx (%LuMB)\n", nid, - nodes[nid].start, nodes[nid].end, - (nodes[nid].end - nodes[nid].start) >> 20); - return ret; -} - -/* - * Sets up nr_nodes fake nodes interleaved over physical nodes ranging from addr - * to max_addr. The return value is the number of nodes allocated. - */ -static int __init split_nodes_interleave(u64 addr, u64 max_addr, int nr_nodes) -{ - nodemask_t physnode_mask = NODE_MASK_NONE; - u64 size; - int big; - int ret = 0; - int i; - - if (nr_nodes <= 0) - return -1; - if (nr_nodes > MAX_NUMNODES) { - pr_info("numa=fake=%d too large, reducing to %d\n", - nr_nodes, MAX_NUMNODES); - nr_nodes = MAX_NUMNODES; - } - - size = (max_addr - addr - memblock_x86_hole_size(addr, max_addr)) / nr_nodes; - /* - * Calculate the number of big nodes that can be allocated as a result - * of consolidating the remainder. - */ - big = ((size & ~FAKE_NODE_MIN_HASH_MASK) * nr_nodes) / - FAKE_NODE_MIN_SIZE; - - size &= FAKE_NODE_MIN_HASH_MASK; - if (!size) { - pr_err("Not enough memory for each node. " - "NUMA emulation disabled.\n"); - return -1; - } - - for (i = 0; i < MAX_NUMNODES; i++) - if (physnodes[i].start != physnodes[i].end) - node_set(i, physnode_mask); - - /* - * Continue to fill physical nodes with fake nodes until there is no - * memory left on any of them. - */ - while (nodes_weight(physnode_mask)) { - for_each_node_mask(i, physnode_mask) { - u64 end = physnodes[i].start + size; - u64 dma32_end = PFN_PHYS(MAX_DMA32_PFN); - - if (ret < big) - end += FAKE_NODE_MIN_SIZE; + for (j = i + 1; j < mi->nr_blks; j++) { + struct numa_memblk *bj = &mi->blk[j]; + unsigned long start, end; /* - * Continue to add memory to this fake node if its - * non-reserved memory is less than the per-node size. + * See whether there are overlapping blocks. Whine + * about but allow overlaps of the same nid. They + * will be merged below. */ - while (end - physnodes[i].start - - memblock_x86_hole_size(physnodes[i].start, end) < size) { - end += FAKE_NODE_MIN_SIZE; - if (end > physnodes[i].end) { - end = physnodes[i].end; - break; + if (bi->end > bj->start && bi->start < bj->end) { + if (bi->nid != bj->nid) { + pr_err("NUMA: node %d (%Lx-%Lx) overlaps with node %d (%Lx-%Lx)\n", + bi->nid, bi->start, bi->end, + bj->nid, bj->start, bj->end); + return -EINVAL; } + pr_warning("NUMA: Warning: node %d (%Lx-%Lx) overlaps with itself (%Lx-%Lx)\n", + bi->nid, bi->start, bi->end, + bj->start, bj->end); } /* - * If there won't be at least FAKE_NODE_MIN_SIZE of - * non-reserved memory in ZONE_DMA32 for the next node, - * this one must extend to the boundary. - */ - if (end < dma32_end && dma32_end - end - - memblock_x86_hole_size(end, dma32_end) < FAKE_NODE_MIN_SIZE) - end = dma32_end; - - /* - * If there won't be enough non-reserved memory for the - * next node, this one must extend to the end of the - * physical node. + * Join together blocks on the same node, holes + * between which don't overlap with memory on other + * nodes. */ - if (physnodes[i].end - end - - memblock_x86_hole_size(end, physnodes[i].end) < size) - end = physnodes[i].end; - - /* - * Avoid allocating more nodes than requested, which can - * happen as a result of rounding down each node's size - * to FAKE_NODE_MIN_SIZE. - */ - if (nodes_weight(physnode_mask) + ret >= nr_nodes) - end = physnodes[i].end; - - if (setup_node_range(ret++, &physnodes[i].start, - end - physnodes[i].start, - physnodes[i].end) < 0) - node_clear(i, physnode_mask); + if (bi->nid != bj->nid) + continue; + start = max(min(bi->start, bj->start), low); + end = min(max(bi->end, bj->end), high); + for (k = 0; k < mi->nr_blks; k++) { + struct numa_memblk *bk = &mi->blk[k]; + + if (bi->nid == bk->nid) + continue; + if (start < bk->end && end > bk->start) + break; + } + if (k < mi->nr_blks) + continue; + printk(KERN_INFO "NUMA: Node %d [%Lx,%Lx) + [%Lx,%Lx) -> [%lx,%lx)\n", + bi->nid, bi->start, bi->end, bj->start, bj->end, + start, end); + bi->start = start; + bi->end = end; + numa_remove_memblk_from(j--, mi); } } - return ret; -} - -/* - * Returns the end address of a node so that there is at least `size' amount of - * non-reserved memory or `max_addr' is reached. - */ -static u64 __init find_end_of_node(u64 start, u64 max_addr, u64 size) -{ - u64 end = start + size; - while (end - start - memblock_x86_hole_size(start, end) < size) { - end += FAKE_NODE_MIN_SIZE; - if (end > max_addr) { - end = max_addr; - break; - } + for (i = mi->nr_blks; i < ARRAY_SIZE(mi->blk); i++) { + mi->blk[i].start = mi->blk[i].end = 0; + mi->blk[i].nid = NUMA_NO_NODE; } - return end; + + return 0; } /* - * Sets up fake nodes of `size' interleaved over physical nodes ranging from - * `addr' to `max_addr'. The return value is the number of nodes allocated. + * Set nodes, which have memory in @mi, in *@nodemask. */ -static int __init split_nodes_size_interleave(u64 addr, u64 max_addr, u64 size) +static void __init numa_nodemask_from_meminfo(nodemask_t *nodemask, + const struct numa_meminfo *mi) { - nodemask_t physnode_mask = NODE_MASK_NONE; - u64 min_size; - int ret = 0; int i; - if (!size) - return -1; - /* - * The limit on emulated nodes is MAX_NUMNODES, so the size per node is - * increased accordingly if the requested size is too small. This - * creates a uniform distribution of node sizes across the entire - * machine (but not necessarily over physical nodes). - */ - min_size = (max_addr - addr - memblock_x86_hole_size(addr, max_addr)) / - MAX_NUMNODES; - min_size = max(min_size, FAKE_NODE_MIN_SIZE); - if ((min_size & FAKE_NODE_MIN_HASH_MASK) < min_size) - min_size = (min_size + FAKE_NODE_MIN_SIZE) & - FAKE_NODE_MIN_HASH_MASK; - if (size < min_size) { - pr_err("Fake node size %LuMB too small, increasing to %LuMB\n", - size >> 20, min_size >> 20); - size = min_size; - } - size &= FAKE_NODE_MIN_HASH_MASK; - - for (i = 0; i < MAX_NUMNODES; i++) - if (physnodes[i].start != physnodes[i].end) - node_set(i, physnode_mask); - /* - * Fill physical nodes with fake nodes of size until there is no memory - * left on any of them. - */ - while (nodes_weight(physnode_mask)) { - for_each_node_mask(i, physnode_mask) { - u64 dma32_end = MAX_DMA32_PFN << PAGE_SHIFT; - u64 end; - - end = find_end_of_node(physnodes[i].start, - physnodes[i].end, size); - /* - * If there won't be at least FAKE_NODE_MIN_SIZE of - * non-reserved memory in ZONE_DMA32 for the next node, - * this one must extend to the boundary. - */ - if (end < dma32_end && dma32_end - end - - memblock_x86_hole_size(end, dma32_end) < FAKE_NODE_MIN_SIZE) - end = dma32_end; + for (i = 0; i < ARRAY_SIZE(mi->blk); i++) + if (mi->blk[i].start != mi->blk[i].end && + mi->blk[i].nid != NUMA_NO_NODE) + node_set(mi->blk[i].nid, *nodemask); +} - /* - * If there won't be enough non-reserved memory for the - * next node, this one must extend to the end of the - * physical node. - */ - if (physnodes[i].end - end - - memblock_x86_hole_size(end, physnodes[i].end) < size) - end = physnodes[i].end; +/** + * numa_reset_distance - Reset NUMA distance table + * + * The current table is freed. The next numa_set_distance() call will + * create a new one. + */ +void __init numa_reset_distance(void) +{ + size_t size = numa_distance_cnt * numa_distance_cnt * sizeof(numa_distance[0]); - /* - * Setup the fake node that will be allocated as bootmem - * later. If setup_node_range() returns non-zero, there - * is no more memory available on this physical node. - */ - if (setup_node_range(ret++, &physnodes[i].start, - end - physnodes[i].start, - physnodes[i].end) < 0) - node_clear(i, physnode_mask); - } - } - return ret; + /* numa_distance could be 1LU marking allocation failure, test cnt */ + if (numa_distance_cnt) + memblock_x86_free_range(__pa(numa_distance), + __pa(numa_distance) + size); + numa_distance_cnt = 0; + numa_distance = NULL; /* enable table creation */ } -/* - * Sets up the system RAM area from start_pfn to last_pfn according to the - * numa=fake command-line option. - */ -static int __init numa_emulation(unsigned long start_pfn, - unsigned long last_pfn, int acpi, int amd) +static int __init numa_alloc_distance(void) { - u64 addr = start_pfn << PAGE_SHIFT; - u64 max_addr = last_pfn << PAGE_SHIFT; - int num_nodes; - int i; + nodemask_t nodes_parsed; + size_t size; + int i, j, cnt = 0; + u64 phys; - /* - * If the numa=fake command-line contains a 'M' or 'G', it represents - * the fixed node size. Otherwise, if it is just a single number N, - * split the system RAM into N fake nodes. - */ - if (strchr(cmdline, 'M') || strchr(cmdline, 'G')) { - u64 size; + /* size the new table and allocate it */ + nodes_parsed = numa_nodes_parsed; + numa_nodemask_from_meminfo(&nodes_parsed, &numa_meminfo); - size = memparse(cmdline, &cmdline); - num_nodes = split_nodes_size_interleave(addr, max_addr, size); - } else { - unsigned long n; + for_each_node_mask(i, nodes_parsed) + cnt = i; + cnt++; + size = cnt * cnt * sizeof(numa_distance[0]); - n = simple_strtoul(cmdline, NULL, 0); - num_nodes = split_nodes_interleave(addr, max_addr, n); + phys = memblock_find_in_range(0, (u64)max_pfn_mapped << PAGE_SHIFT, + size, PAGE_SIZE); + if (phys == MEMBLOCK_ERROR) { + pr_warning("NUMA: Warning: can't allocate distance table!\n"); + /* don't retry until explicitly reset */ + numa_distance = (void *)1LU; + return -ENOMEM; } + memblock_x86_reserve_range(phys, phys + size, "NUMA DIST"); - if (num_nodes < 0) - return num_nodes; - memnode_shift = compute_hash_shift(nodes, num_nodes, NULL); - if (memnode_shift < 0) { - memnode_shift = 0; - printk(KERN_ERR "No NUMA hash function found. NUMA emulation " - "disabled.\n"); - return -1; - } + numa_distance = __va(phys); + numa_distance_cnt = cnt; + + /* fill with the default distances */ + for (i = 0; i < cnt; i++) + for (j = 0; j < cnt; j++) + numa_distance[i * cnt + j] = i == j ? + LOCAL_DISTANCE : REMOTE_DISTANCE; + printk(KERN_DEBUG "NUMA: Initialized distance table, cnt=%d\n", cnt); - /* - * We need to vacate all active ranges that may have been registered for - * the e820 memory map. - */ - remove_all_active_ranges(); - for_each_node_mask(i, node_possible_map) { - memblock_x86_register_active_regions(i, nodes[i].start >> PAGE_SHIFT, - nodes[i].end >> PAGE_SHIFT); - setup_node_bootmem(i, nodes[i].start, nodes[i].end); - } - setup_physnodes(addr, max_addr, acpi, amd); - fake_physnodes(acpi, amd, num_nodes); - numa_init_array(); return 0; } -#endif /* CONFIG_NUMA_EMU */ -void __init initmem_init(unsigned long start_pfn, unsigned long last_pfn, - int acpi, int amd) +/** + * numa_set_distance - Set NUMA distance from one NUMA to another + * @from: the 'from' node to set distance + * @to: the 'to' node to set distance + * @distance: NUMA distance + * + * Set the distance from node @from to @to to @distance. If distance table + * doesn't exist, one which is large enough to accomodate all the currently + * known nodes will be created. + * + * If such table cannot be allocated, a warning is printed and further + * calls are ignored until the distance table is reset with + * numa_reset_distance(). + * + * If @from or @to is higher than the highest known node at the time of + * table creation or @distance doesn't make sense, the call is ignored. + * This is to allow simplification of specific NUMA config implementations. + */ +void __init numa_set_distance(int from, int to, int distance) { - int i; - - nodes_clear(node_possible_map); - nodes_clear(node_online_map); - -#ifdef CONFIG_NUMA_EMU - setup_physnodes(start_pfn << PAGE_SHIFT, last_pfn << PAGE_SHIFT, - acpi, amd); - if (cmdline && !numa_emulation(start_pfn, last_pfn, acpi, amd)) + if (!numa_distance && numa_alloc_distance() < 0) return; - setup_physnodes(start_pfn << PAGE_SHIFT, last_pfn << PAGE_SHIFT, - acpi, amd); - nodes_clear(node_possible_map); - nodes_clear(node_online_map); -#endif -#ifdef CONFIG_ACPI_NUMA - if (!numa_off && acpi && !acpi_scan_nodes(start_pfn << PAGE_SHIFT, - last_pfn << PAGE_SHIFT)) + if (from >= numa_distance_cnt || to >= numa_distance_cnt) { + printk_once(KERN_DEBUG "NUMA: Debug: distance out of bound, from=%d to=%d distance=%d\n", + from, to, distance); return; - nodes_clear(node_possible_map); - nodes_clear(node_online_map); -#endif + } -#ifdef CONFIG_AMD_NUMA - if (!numa_off && amd && !amd_scan_nodes()) + if ((u8)distance != distance || + (from == to && distance != LOCAL_DISTANCE)) { + pr_warn_once("NUMA: Warning: invalid distance parameter, from=%d to=%d distance=%d\n", + from, to, distance); return; - nodes_clear(node_possible_map); - nodes_clear(node_online_map); -#endif - printk(KERN_INFO "%s\n", - numa_off ? "NUMA turned off" : "No NUMA configuration found"); + } - printk(KERN_INFO "Faking a node at %016lx-%016lx\n", - start_pfn << PAGE_SHIFT, - last_pfn << PAGE_SHIFT); - /* setup dummy node covering all memory */ - memnode_shift = 63; - memnodemap = memnode.embedded_map; - memnodemap[0] = 0; - node_set_online(0); - node_set(0, node_possible_map); - for (i = 0; i < nr_cpu_ids; i++) - numa_set_node(i, 0); - memblock_x86_register_active_regions(0, start_pfn, last_pfn); - setup_node_bootmem(0, start_pfn << PAGE_SHIFT, last_pfn << PAGE_SHIFT); + numa_distance[from * numa_distance_cnt + to] = distance; } -unsigned long __init numa_free_all_bootmem(void) +int __node_distance(int from, int to) { - unsigned long pages = 0; - int i; + if (from >= numa_distance_cnt || to >= numa_distance_cnt) + return from == to ? LOCAL_DISTANCE : REMOTE_DISTANCE; + return numa_distance[from * numa_distance_cnt + to]; +} +EXPORT_SYMBOL(__node_distance); - for_each_online_node(i) - pages += free_all_bootmem_node(NODE_DATA(i)); +/* + * Sanity check to catch more bad NUMA configurations (they are amazingly + * common). Make sure the nodes cover all memory. + */ +static bool __init numa_meminfo_cover_memory(const struct numa_meminfo *mi) +{ + unsigned long numaram, e820ram; + int i; - pages += free_all_memory_core_early(MAX_NUMNODES); + numaram = 0; + for (i = 0; i < mi->nr_blks; i++) { + unsigned long s = mi->blk[i].start >> PAGE_SHIFT; + unsigned long e = mi->blk[i].end >> PAGE_SHIFT; + numaram += e - s; + numaram -= __absent_pages_in_range(mi->blk[i].nid, s, e); + if ((long)numaram < 0) + numaram = 0; + } - return pages; + e820ram = max_pfn - (memblock_x86_hole_size(0, + max_pfn << PAGE_SHIFT) >> PAGE_SHIFT); + /* We seem to lose 3 pages somewhere. Allow 1M of slack. */ + if ((long)(e820ram - numaram) >= (1 << (20 - PAGE_SHIFT))) { + printk(KERN_ERR "NUMA: nodes only cover %luMB of your %luMB e820 RAM. Not used.\n", + (numaram << PAGE_SHIFT) >> 20, + (e820ram << PAGE_SHIFT) >> 20); + return false; + } + return true; } -#ifdef CONFIG_NUMA - -static __init int find_near_online_node(int node) +static int __init numa_register_memblks(struct numa_meminfo *mi) { - int n, val; - int min_val = INT_MAX; - int best_node = -1; + int i, nid; - for_each_online_node(n) { - val = node_distance(node, n); + /* Account for nodes with cpus and no memory */ + node_possible_map = numa_nodes_parsed; + numa_nodemask_from_meminfo(&node_possible_map, mi); + if (WARN_ON(nodes_empty(node_possible_map))) + return -EINVAL; - if (val < min_val) { - min_val = val; - best_node = n; + memnode_shift = compute_hash_shift(mi); + if (memnode_shift < 0) { + printk(KERN_ERR "NUMA: No NUMA node hash function found. Contact maintainer\n"); + return -EINVAL; + } + + for (i = 0; i < mi->nr_blks; i++) + memblock_x86_register_active_regions(mi->blk[i].nid, + mi->blk[i].start >> PAGE_SHIFT, + mi->blk[i].end >> PAGE_SHIFT); + + /* for out of order entries */ + sort_node_map(); + if (!numa_meminfo_cover_memory(mi)) + return -EINVAL; + + /* Finally register nodes. */ + for_each_node_mask(nid, node_possible_map) { + u64 start = (u64)max_pfn << PAGE_SHIFT; + u64 end = 0; + + for (i = 0; i < mi->nr_blks; i++) { + if (nid != mi->blk[i].nid) + continue; + start = min(mi->blk[i].start, start); + end = max(mi->blk[i].end, end); } + + if (start < end) + setup_node_bootmem(nid, start, end); } - return best_node; + return 0; } -/* - * Setup early cpu_to_node. +/** + * dummy_numma_init - Fallback dummy NUMA init * - * 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. + * Used if there's no underlying NUMA architecture, NUMA initialization + * fails, or NUMA is disabled on the command line. * - * Called before the per_cpu areas are setup. + * Must online at least one node and add memory blocks that cover all + * allowed memory. This function must not fail. */ -void __init init_cpu_to_node(void) +static int __init dummy_numa_init(void) { - int cpu; - u16 *cpu_to_apicid = early_per_cpu_ptr(x86_cpu_to_apicid); - - BUG_ON(cpu_to_apicid == NULL); + printk(KERN_INFO "%s\n", + numa_off ? "NUMA turned off" : "No NUMA configuration found"); + printk(KERN_INFO "Faking a node at %016lx-%016lx\n", + 0LU, max_pfn << PAGE_SHIFT); - for_each_possible_cpu(cpu) { - int node; - u16 apicid = cpu_to_apicid[cpu]; + node_set(0, numa_nodes_parsed); + numa_add_memblk(0, 0, (u64)max_pfn << PAGE_SHIFT); - 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); - } + return 0; } -#endif - -void __cpuinit numa_set_node(int cpu, int node) +static int __init numa_init(int (*init_func)(void)) { - 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; + int i; + int ret; - if (node != NUMA_NO_NODE) - set_cpu_numa_node(cpu, node); -} + for (i = 0; i < MAX_LOCAL_APIC; i++) + set_apicid_to_node(i, NUMA_NO_NODE); -void __cpuinit numa_clear_node(int cpu) -{ - numa_set_node(cpu, NUMA_NO_NODE); -} - -#ifndef CONFIG_DEBUG_PER_CPU_MAPS + nodes_clear(numa_nodes_parsed); + nodes_clear(node_possible_map); + nodes_clear(node_online_map); + memset(&numa_meminfo, 0, sizeof(numa_meminfo)); + remove_all_active_ranges(); + numa_reset_distance(); -#ifndef CONFIG_NUMA_EMU -void __cpuinit numa_add_cpu(int cpu) -{ - cpumask_set_cpu(cpu, node_to_cpumask_map[early_cpu_to_node(cpu)]); -} + ret = init_func(); + if (ret < 0) + return ret; + ret = numa_cleanup_meminfo(&numa_meminfo); + if (ret < 0) + return ret; -void __cpuinit numa_remove_cpu(int cpu) -{ - cpumask_clear_cpu(cpu, node_to_cpumask_map[early_cpu_to_node(cpu)]); -} -#else -void __cpuinit numa_add_cpu(int cpu) -{ - unsigned long addr; - u16 apicid; - int physnid; - int nid = NUMA_NO_NODE; + numa_emulation(&numa_meminfo, numa_distance_cnt); - nid = early_cpu_to_node(cpu); - BUG_ON(nid == NUMA_NO_NODE || !node_online(nid)); + ret = numa_register_memblks(&numa_meminfo); + if (ret < 0) + return ret; - /* - * Use the starting address of the emulated node to find which physical - * node it is allocated on. - */ - addr = node_start_pfn(nid) << PAGE_SHIFT; - for (physnid = 0; physnid < MAX_NUMNODES; physnid++) - if (addr >= physnodes[physnid].start && - addr < physnodes[physnid].end) - break; + for (i = 0; i < nr_cpu_ids; i++) { + int nid = early_cpu_to_node(i); - /* - * Map the cpu to each emulated node that is allocated on the physical - * node of the cpu's apic id. - */ - for_each_online_node(nid) { - addr = node_start_pfn(nid) << PAGE_SHIFT; - if (addr >= physnodes[physnid].start && - addr < physnodes[physnid].end) - cpumask_set_cpu(cpu, node_to_cpumask_map[nid]); + if (nid == NUMA_NO_NODE) + continue; + if (!node_online(nid)) + numa_clear_node(i); } + numa_init_array(); + return 0; } -void __cpuinit numa_remove_cpu(int cpu) +void __init initmem_init(void) { - int i; + int ret; - 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; + if (!numa_off) { +#ifdef CONFIG_ACPI_NUMA + ret = numa_init(x86_acpi_numa_init); + if (!ret) + return; +#endif +#ifdef CONFIG_AMD_NUMA + ret = numa_init(amd_numa_init); + if (!ret) + return; +#endif } - 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; + numa_init(dummy_numa_init); } -/* - * --------- 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 -static void __cpuinit numa_set_cpumask(int cpu, int enable) +unsigned long __init numa_free_all_bootmem(void) { - int node = early_cpu_to_node(cpu); - struct cpumask *mask; + unsigned long pages = 0; int i; - for_each_online_node(i) { - unsigned long addr; - - addr = node_start_pfn(i) << PAGE_SHIFT; - if (addr < physnodes[node].start || - addr >= physnodes[node].end) - continue; - mask = debug_cpumask_set_cpu(cpu, enable); - if (!mask) - return; - - if (enable) - cpumask_set_cpu(cpu, mask); - else - cpumask_clear_cpu(cpu, mask); - } -} -#endif /* CONFIG_NUMA_EMU */ + for_each_online_node(i) + pages += free_all_bootmem_node(NODE_DATA(i)); -void __cpuinit numa_add_cpu(int cpu) -{ - numa_set_cpumask(cpu, 1); -} + pages += free_all_memory_core_early(MAX_NUMNODES); -void __cpuinit numa_remove_cpu(int cpu) -{ - numa_set_cpumask(cpu, 0); + return pages; } -int __cpu_to_node(int cpu) +int __cpuinit numa_cpu_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); + int apicid = early_per_cpu(x86_cpu_to_apicid, cpu); -/* - * 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); + if (apicid != BAD_APICID) + return __apicid_to_node[apicid]; + return NUMA_NO_NODE; } - -/* - * --------- end of debug versions of the numa functions --------- - */ - -#endif /* CONFIG_DEBUG_PER_CPU_MAPS */ diff --git a/arch/x86/mm/numa_emulation.c b/arch/x86/mm/numa_emulation.c new file mode 100644 index 0000000..ad091e4 --- /dev/null +++ b/arch/x86/mm/numa_emulation.c @@ -0,0 +1,494 @@ +/* + * NUMA emulation + */ +#include <linux/kernel.h> +#include <linux/errno.h> +#include <linux/topology.h> +#include <linux/memblock.h> +#include <asm/dma.h> + +#include "numa_internal.h" + +static int emu_nid_to_phys[MAX_NUMNODES] __cpuinitdata; +static char *emu_cmdline __initdata; + +void __init numa_emu_cmdline(char *str) +{ + emu_cmdline = str; +} + +static int __init emu_find_memblk_by_nid(int nid, const struct numa_meminfo *mi) +{ + int i; + + for (i = 0; i < mi->nr_blks; i++) + if (mi->blk[i].nid == nid) + return i; + return -ENOENT; +} + +/* + * Sets up nid to range from @start to @end. The return value is -errno if + * something went wrong, 0 otherwise. + */ +static int __init emu_setup_memblk(struct numa_meminfo *ei, + struct numa_meminfo *pi, + int nid, int phys_blk, u64 size) +{ + struct numa_memblk *eb = &ei->blk[ei->nr_blks]; + struct numa_memblk *pb = &pi->blk[phys_blk]; + + if (ei->nr_blks >= NR_NODE_MEMBLKS) { + pr_err("NUMA: Too many emulated memblks, failing emulation\n"); + return -EINVAL; + } + + ei->nr_blks++; + eb->start = pb->start; + eb->end = pb->start + size; + eb->nid = nid; + + if (emu_nid_to_phys[nid] == NUMA_NO_NODE) + emu_nid_to_phys[nid] = pb->nid; + + pb->start += size; + if (pb->start >= pb->end) { + WARN_ON_ONCE(pb->start > pb->end); + numa_remove_memblk_from(phys_blk, pi); + } + + printk(KERN_INFO "Faking node %d at %016Lx-%016Lx (%LuMB)\n", nid, + eb->start, eb->end, (eb->end - eb->start) >> 20); + return 0; +} + +/* + * Sets up nr_nodes fake nodes interleaved over physical nodes ranging from addr + * to max_addr. The return value is the number of nodes allocated. + */ +static int __init split_nodes_interleave(struct numa_meminfo *ei, + struct numa_meminfo *pi, + u64 addr, u64 max_addr, int nr_nodes) +{ + nodemask_t physnode_mask = NODE_MASK_NONE; + u64 size; + int big; + int nid = 0; + int i, ret; + + if (nr_nodes <= 0) + return -1; + if (nr_nodes > MAX_NUMNODES) { + pr_info("numa=fake=%d too large, reducing to %d\n", + nr_nodes, MAX_NUMNODES); + nr_nodes = MAX_NUMNODES; + } + + size = (max_addr - addr - memblock_x86_hole_size(addr, max_addr)) / nr_nodes; + /* + * Calculate the number of big nodes that can be allocated as a result + * of consolidating the remainder. + */ + big = ((size & ~FAKE_NODE_MIN_HASH_MASK) * nr_nodes) / + FAKE_NODE_MIN_SIZE; + + size &= FAKE_NODE_MIN_HASH_MASK; + if (!size) { + pr_err("Not enough memory for each node. " + "NUMA emulation disabled.\n"); + return -1; + } + + for (i = 0; i < pi->nr_blks; i++) + node_set(pi->blk[i].nid, physnode_mask); + + /* + * Continue to fill physical nodes with fake nodes until there is no + * memory left on any of them. + */ + while (nodes_weight(physnode_mask)) { + for_each_node_mask(i, physnode_mask) { + u64 dma32_end = PFN_PHYS(MAX_DMA32_PFN); + u64 start, limit, end; + int phys_blk; + + phys_blk = emu_find_memblk_by_nid(i, pi); + if (phys_blk < 0) { + node_clear(i, physnode_mask); + continue; + } + start = pi->blk[phys_blk].start; + limit = pi->blk[phys_blk].end; + end = start + size; + + if (nid < big) + end += FAKE_NODE_MIN_SIZE; + + /* + * Continue to add memory to this fake node if its + * non-reserved memory is less than the per-node size. + */ + while (end - start - + memblock_x86_hole_size(start, end) < size) { + end += FAKE_NODE_MIN_SIZE; + if (end > limit) { + end = limit; + break; + } + } + + /* + * If there won't be at least FAKE_NODE_MIN_SIZE of + * non-reserved memory in ZONE_DMA32 for the next node, + * this one must extend to the boundary. + */ + if (end < dma32_end && dma32_end - end - + memblock_x86_hole_size(end, dma32_end) < FAKE_NODE_MIN_SIZE) + end = dma32_end; + + /* + * If there won't be enough non-reserved memory for the + * next node, this one must extend to the end of the + * physical node. + */ + if (limit - end - + memblock_x86_hole_size(end, limit) < size) + end = limit; + + ret = emu_setup_memblk(ei, pi, nid++ % nr_nodes, + phys_blk, + min(end, limit) - start); + if (ret < 0) + return ret; + } + } + return 0; +} + +/* + * Returns the end address of a node so that there is at least `size' amount of + * non-reserved memory or `max_addr' is reached. + */ +static u64 __init find_end_of_node(u64 start, u64 max_addr, u64 size) +{ + u64 end = start + size; + + while (end - start - memblock_x86_hole_size(start, end) < size) { + end += FAKE_NODE_MIN_SIZE; + if (end > max_addr) { + end = max_addr; + break; + } + } + return end; +} + +/* + * Sets up fake nodes of `size' interleaved over physical nodes ranging from + * `addr' to `max_addr'. The return value is the number of nodes allocated. + */ +static int __init split_nodes_size_interleave(struct numa_meminfo *ei, + struct numa_meminfo *pi, + u64 addr, u64 max_addr, u64 size) +{ + nodemask_t physnode_mask = NODE_MASK_NONE; + u64 min_size; + int nid = 0; + int i, ret; + + if (!size) + return -1; + /* + * The limit on emulated nodes is MAX_NUMNODES, so the size per node is + * increased accordingly if the requested size is too small. This + * creates a uniform distribution of node sizes across the entire + * machine (but not necessarily over physical nodes). + */ + min_size = (max_addr - addr - memblock_x86_hole_size(addr, max_addr)) / + MAX_NUMNODES; + min_size = max(min_size, FAKE_NODE_MIN_SIZE); + if ((min_size & FAKE_NODE_MIN_HASH_MASK) < min_size) + min_size = (min_size + FAKE_NODE_MIN_SIZE) & + FAKE_NODE_MIN_HASH_MASK; + if (size < min_size) { + pr_err("Fake node size %LuMB too small, increasing to %LuMB\n", + size >> 20, min_size >> 20); + size = min_size; + } + size &= FAKE_NODE_MIN_HASH_MASK; + + for (i = 0; i < pi->nr_blks; i++) + node_set(pi->blk[i].nid, physnode_mask); + + /* + * Fill physical nodes with fake nodes of size until there is no memory + * left on any of them. + */ + while (nodes_weight(physnode_mask)) { + for_each_node_mask(i, physnode_mask) { + u64 dma32_end = MAX_DMA32_PFN << PAGE_SHIFT; + u64 start, limit, end; + int phys_blk; + + phys_blk = emu_find_memblk_by_nid(i, pi); + if (phys_blk < 0) { + node_clear(i, physnode_mask); + continue; + } + start = pi->blk[phys_blk].start; + limit = pi->blk[phys_blk].end; + + end = find_end_of_node(start, limit, size); + /* + * If there won't be at least FAKE_NODE_MIN_SIZE of + * non-reserved memory in ZONE_DMA32 for the next node, + * this one must extend to the boundary. + */ + if (end < dma32_end && dma32_end - end - + memblock_x86_hole_size(end, dma32_end) < FAKE_NODE_MIN_SIZE) + end = dma32_end; + + /* + * If there won't be enough non-reserved memory for the + * next node, this one must extend to the end of the + * physical node. + */ + if (limit - end - + memblock_x86_hole_size(end, limit) < size) + end = limit; + + ret = emu_setup_memblk(ei, pi, nid++ % MAX_NUMNODES, + phys_blk, + min(end, limit) - start); + if (ret < 0) + return ret; + } + } + return 0; +} + +/** + * numa_emulation - Emulate NUMA nodes + * @numa_meminfo: NUMA configuration to massage + * @numa_dist_cnt: The size of the physical NUMA distance table + * + * Emulate NUMA nodes according to the numa=fake kernel parameter. + * @numa_meminfo contains the physical memory configuration and is modified + * to reflect the emulated configuration on success. @numa_dist_cnt is + * used to determine the size of the physical distance table. + * + * On success, the following modifications are made. + * + * - @numa_meminfo is updated to reflect the emulated nodes. + * + * - __apicid_to_node[] is updated such that APIC IDs are mapped to the + * emulated nodes. + * + * - NUMA distance table is rebuilt to represent distances between emulated + * nodes. The distances are determined considering how emulated nodes + * are mapped to physical nodes and match the actual distances. + * + * - emu_nid_to_phys[] reflects how emulated nodes are mapped to physical + * nodes. This is used by numa_add_cpu() and numa_remove_cpu(). + * + * If emulation is not enabled or fails, emu_nid_to_phys[] is filled with + * identity mapping and no other modification is made. + */ +void __init numa_emulation(struct numa_meminfo *numa_meminfo, int numa_dist_cnt) +{ + static struct numa_meminfo ei __initdata; + static struct numa_meminfo pi __initdata; + const u64 max_addr = max_pfn << PAGE_SHIFT; + u8 *phys_dist = NULL; + size_t phys_size = numa_dist_cnt * numa_dist_cnt * sizeof(phys_dist[0]); + int max_emu_nid, dfl_phys_nid; + int i, j, ret; + + if (!emu_cmdline) + goto no_emu; + + memset(&ei, 0, sizeof(ei)); + pi = *numa_meminfo; + + for (i = 0; i < MAX_NUMNODES; i++) + emu_nid_to_phys[i] = NUMA_NO_NODE; + + /* + * If the numa=fake command-line contains a 'M' or 'G', it represents + * the fixed node size. Otherwise, if it is just a single number N, + * split the system RAM into N fake nodes. + */ + if (strchr(emu_cmdline, 'M') || strchr(emu_cmdline, 'G')) { + u64 size; + + size = memparse(emu_cmdline, &emu_cmdline); + ret = split_nodes_size_interleave(&ei, &pi, 0, max_addr, size); + } else { + unsigned long n; + + n = simple_strtoul(emu_cmdline, NULL, 0); + ret = split_nodes_interleave(&ei, &pi, 0, max_addr, n); + } + + if (ret < 0) + goto no_emu; + + if (numa_cleanup_meminfo(&ei) < 0) { + pr_warning("NUMA: Warning: constructed meminfo invalid, disabling emulation\n"); + goto no_emu; + } + + /* copy the physical distance table */ + if (numa_dist_cnt) { + u64 phys; + + phys = memblock_find_in_range(0, + (u64)max_pfn_mapped << PAGE_SHIFT, + phys_size, PAGE_SIZE); + if (phys == MEMBLOCK_ERROR) { + pr_warning("NUMA: Warning: can't allocate copy of distance table, disabling emulation\n"); + goto no_emu; + } + memblock_x86_reserve_range(phys, phys + phys_size, "TMP NUMA DIST"); + phys_dist = __va(phys); + + for (i = 0; i < numa_dist_cnt; i++) + for (j = 0; j < numa_dist_cnt; j++) + phys_dist[i * numa_dist_cnt + j] = + node_distance(i, j); + } + + /* + * Determine the max emulated nid and the default phys nid to use + * for unmapped nodes. + */ + max_emu_nid = 0; + dfl_phys_nid = NUMA_NO_NODE; + for (i = 0; i < ARRAY_SIZE(emu_nid_to_phys); i++) { + if (emu_nid_to_phys[i] != NUMA_NO_NODE) { + max_emu_nid = i; + if (dfl_phys_nid == NUMA_NO_NODE) + dfl_phys_nid = emu_nid_to_phys[i]; + } + } + if (dfl_phys_nid == NUMA_NO_NODE) { + pr_warning("NUMA: Warning: can't determine default physical node, disabling emulation\n"); + goto no_emu; + } + + /* commit */ + *numa_meminfo = ei; + + /* + * Transform __apicid_to_node table to use emulated nids by + * reverse-mapping phys_nid. The maps should always exist but fall + * back to zero just in case. + */ + for (i = 0; i < ARRAY_SIZE(__apicid_to_node); i++) { + if (__apicid_to_node[i] == NUMA_NO_NODE) + continue; + for (j = 0; j < ARRAY_SIZE(emu_nid_to_phys); j++) + if (__apicid_to_node[i] == emu_nid_to_phys[j]) + break; + __apicid_to_node[i] = j < ARRAY_SIZE(emu_nid_to_phys) ? j : 0; + } + + /* make sure all emulated nodes are mapped to a physical node */ + for (i = 0; i < ARRAY_SIZE(emu_nid_to_phys); i++) + if (emu_nid_to_phys[i] == NUMA_NO_NODE) + emu_nid_to_phys[i] = dfl_phys_nid; + + /* transform distance table */ + numa_reset_distance(); + for (i = 0; i < max_emu_nid + 1; i++) { + for (j = 0; j < max_emu_nid + 1; j++) { + int physi = emu_nid_to_phys[i]; + int physj = emu_nid_to_phys[j]; + int dist; + + if (physi >= numa_dist_cnt || physj >= numa_dist_cnt) + dist = physi == physj ? + LOCAL_DISTANCE : REMOTE_DISTANCE; + else + dist = phys_dist[physi * numa_dist_cnt + physj]; + + numa_set_distance(i, j, dist); + } + } + + /* free the copied physical distance table */ + if (phys_dist) + memblock_x86_free_range(__pa(phys_dist), __pa(phys_dist) + phys_size); + return; + +no_emu: + /* No emulation. Build identity emu_nid_to_phys[] for numa_add_cpu() */ + for (i = 0; i < ARRAY_SIZE(emu_nid_to_phys); i++) + emu_nid_to_phys[i] = i; +} + +#ifndef CONFIG_DEBUG_PER_CPU_MAPS +void __cpuinit numa_add_cpu(int cpu) +{ + int physnid, nid; + + nid = early_cpu_to_node(cpu); + BUG_ON(nid == NUMA_NO_NODE || !node_online(nid)); + + physnid = emu_nid_to_phys[nid]; + + /* + * Map the cpu to each emulated node that is allocated on the physical + * node of the cpu's apic id. + */ + for_each_online_node(nid) + if (emu_nid_to_phys[nid] == physnid) + cpumask_set_cpu(cpu, node_to_cpumask_map[nid]); +} + +void __cpuinit numa_remove_cpu(int cpu) +{ + int i; + + for_each_online_node(i) + cpumask_clear_cpu(cpu, node_to_cpumask_map[i]); +} +#else /* !CONFIG_DEBUG_PER_CPU_MAPS */ +static void __cpuinit numa_set_cpumask(int cpu, int enable) +{ + struct cpumask *mask; + int nid, physnid, i; + + nid = early_cpu_to_node(cpu); + if (nid == NUMA_NO_NODE) { + /* early_cpu_to_node() already emits a warning and trace */ + return; + } + + physnid = emu_nid_to_phys[nid]; + + for_each_online_node(i) { + if (emu_nid_to_phys[nid] != physnid) + continue; + + 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_DEBUG_PER_CPU_MAPS */ diff --git a/arch/x86/mm/numa_internal.h b/arch/x86/mm/numa_internal.h new file mode 100644 index 0000000..ef2d973 --- /dev/null +++ b/arch/x86/mm/numa_internal.h @@ -0,0 +1,31 @@ +#ifndef __X86_MM_NUMA_INTERNAL_H +#define __X86_MM_NUMA_INTERNAL_H + +#include <linux/types.h> +#include <asm/numa.h> + +struct numa_memblk { + u64 start; + u64 end; + int nid; +}; + +struct numa_meminfo { + int nr_blks; + struct numa_memblk blk[NR_NODE_MEMBLKS]; +}; + +void __init numa_remove_memblk_from(int idx, struct numa_meminfo *mi); +int __init numa_cleanup_meminfo(struct numa_meminfo *mi); +void __init numa_reset_distance(void); + +#ifdef CONFIG_NUMA_EMU +void __init numa_emulation(struct numa_meminfo *numa_meminfo, + int numa_dist_cnt); +#else +static inline void numa_emulation(struct numa_meminfo *numa_meminfo, + int numa_dist_cnt) +{ } +#endif + +#endif /* __X86_MM_NUMA_INTERNAL_H */ diff --git a/arch/x86/mm/srat_32.c b/arch/x86/mm/srat_32.c index ae96e7b8..48651c6 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 603d285..8e9d339 100644 --- a/arch/x86/mm/srat_64.c +++ b/arch/x86/mm/srat_64.c @@ -26,88 +26,34 @@ int acpi_numa __initdata; -static struct acpi_table_slit *acpi_slit; - -static nodemask_t nodes_parsed __initdata; -static nodemask_t cpu_nodes_parsed __initdata; -static struct bootnode nodes[MAX_NUMNODES] __initdata; static struct bootnode nodes_add[MAX_NUMNODES]; -static int num_node_memblks __initdata; -static struct bootnode node_memblk_range[NR_NODE_MEMBLKS] __initdata; -static int memblk_nodeid[NR_NODE_MEMBLKS] __initdata; - static __init int setup_node(int pxm) { return acpi_map_pxm_to_node(pxm); } -static __init int conflicting_memblks(unsigned long start, unsigned long end) -{ - int i; - for (i = 0; i < num_node_memblks; i++) { - struct bootnode *nd = &node_memblk_range[i]; - if (nd->start == nd->end) - continue; - if (nd->end > start && nd->start < end) - return memblk_nodeid[i]; - if (nd->end == end && nd->start == start) - return memblk_nodeid[i]; - } - return -1; -} - -static __init void cutoff_node(int i, unsigned long start, unsigned long end) -{ - struct bootnode *nd = &nodes[i]; - - if (nd->start < start) { - nd->start = start; - if (nd->end < nd->start) - nd->start = nd->end; - } - if (nd->end > end) { - nd->end = end; - if (nd->start > nd->end) - nd->start = nd->end; - } -} - static __init void bad_srat(void) { - int i; 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; - for (i = 0; i < MAX_NUMNODES; i++) { - nodes[i].start = nodes[i].end = 0; - nodes_add[i].start = nodes_add[i].end = 0; - } - remove_all_active_ranges(); + memset(nodes_add, 0, sizeof(nodes_add)); } static __init inline int srat_disabled(void) { - return numa_off || acpi_numa < 0; + return acpi_numa < 0; } /* Callback for SLIT parsing */ void __init acpi_numa_slit_init(struct acpi_table_slit *slit) { - unsigned length; - unsigned long phys; - - length = slit->header.length; - phys = memblock_find_in_range(0, max_pfn_mapped<<PAGE_SHIFT, length, - PAGE_SIZE); - - if (phys == MEMBLOCK_ERROR) - panic(" Can not save slit!\n"); + int i, j; - acpi_slit = __va(phys); - memcpy(acpi_slit, slit, length); - memblock_x86_reserve_range(phys, phys + length, "ACPI SLIT"); + for (i = 0; i < slit->locality_count; i++) + for (j = 0; j < slit->locality_count; j++) + numa_set_distance(pxm_to_node(i), pxm_to_node(j), + slit->entry[slit->locality_count * i + j]); } /* Callback for Proximity Domain -> x2APIC mapping */ @@ -138,8 +84,8 @@ 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; - node_set(node, cpu_nodes_parsed); + set_apicid_to_node(apic_id, node); + node_set(node, numa_nodes_parsed); acpi_numa = 1; printk(KERN_INFO "SRAT: PXM %u -> APIC 0x%04x -> Node %u\n", pxm, apic_id, node); @@ -178,8 +124,8 @@ acpi_numa_processor_affinity_init(struct acpi_srat_cpu_affinity *pa) return; } - apicid_to_node[apic_id] = node; - node_set(node, cpu_nodes_parsed); + set_apicid_to_node(apic_id, node); + node_set(node, numa_nodes_parsed); acpi_numa = 1; printk(KERN_INFO "SRAT: PXM %u -> APIC 0x%02x -> Node %u\n", pxm, apic_id, node); @@ -241,7 +187,7 @@ update_nodes_add(int node, unsigned long start, unsigned long end) } if (changed) { - node_set(node, cpu_nodes_parsed); + node_set(node, numa_nodes_parsed); printk(KERN_INFO "SRAT: hot plug zone found %Lx - %Lx\n", nd->start, nd->end); } @@ -251,10 +197,8 @@ update_nodes_add(int node, unsigned long start, unsigned long end) void __init acpi_numa_memory_affinity_init(struct acpi_srat_mem_affinity *ma) { - struct bootnode *nd, oldnode; unsigned long start, end; int node, pxm; - int i; if (srat_disabled()) return; @@ -276,300 +220,31 @@ acpi_numa_memory_affinity_init(struct acpi_srat_mem_affinity *ma) bad_srat(); return; } - i = conflicting_memblks(start, end); - if (i == node) { - printk(KERN_WARNING - "SRAT: Warning: PXM %d (%lx-%lx) overlaps with itself (%Lx-%Lx)\n", - pxm, start, end, nodes[i].start, nodes[i].end); - } else if (i >= 0) { - printk(KERN_ERR - "SRAT: PXM %d (%lx-%lx) overlaps with PXM %d (%Lx-%Lx)\n", - pxm, start, end, node_to_pxm(i), - nodes[i].start, nodes[i].end); + + if (numa_add_memblk(node, start, end) < 0) { bad_srat(); return; } - nd = &nodes[node]; - oldnode = *nd; - if (!node_test_and_set(node, nodes_parsed)) { - nd->start = start; - nd->end = end; - } else { - if (start < nd->start) - nd->start = start; - if (nd->end < end) - nd->end = end; - } printk(KERN_INFO "SRAT: Node %u PXM %u %lx-%lx\n", node, pxm, start, end); - if (ma->flags & ACPI_SRAT_MEM_HOT_PLUGGABLE) { + if (ma->flags & ACPI_SRAT_MEM_HOT_PLUGGABLE) update_nodes_add(node, start, end); - /* restore nodes[node] */ - *nd = oldnode; - if ((nd->start | nd->end) == 0) - node_clear(node, nodes_parsed); - } - - node_memblk_range[num_node_memblks].start = start; - node_memblk_range[num_node_memblks].end = end; - memblk_nodeid[num_node_memblks] = node; - num_node_memblks++; -} - -/* Sanity check to catch more bad SRATs (they are amazingly common). - Make sure the PXMs cover all memory. */ -static int __init nodes_cover_memory(const struct bootnode *nodes) -{ - int i; - unsigned long pxmram, e820ram; - - pxmram = 0; - for_each_node_mask(i, nodes_parsed) { - unsigned long s = nodes[i].start >> PAGE_SHIFT; - unsigned long e = nodes[i].end >> PAGE_SHIFT; - pxmram += e - s; - pxmram -= __absent_pages_in_range(i, s, e); - if ((long)pxmram < 0) - pxmram = 0; - } - - e820ram = max_pfn - (memblock_x86_hole_size(0, max_pfn<<PAGE_SHIFT)>>PAGE_SHIFT); - /* We seem to lose 3 pages somewhere. Allow 1M of slack. */ - if ((long)(e820ram - pxmram) >= (1<<(20 - PAGE_SHIFT))) { - printk(KERN_ERR - "SRAT: PXMs only cover %luMB of your %luMB e820 RAM. Not used.\n", - (pxmram << PAGE_SHIFT) >> 20, - (e820ram << PAGE_SHIFT) >> 20); - return 0; - } - return 1; } void __init acpi_numa_arch_fixup(void) {} -#ifdef CONFIG_NUMA_EMU -void __init acpi_get_nodes(struct bootnode *physnodes, unsigned long start, - unsigned long end) -{ - int i; - - for_each_node_mask(i, nodes_parsed) { - cutoff_node(i, start, end); - physnodes[i].start = nodes[i].start; - physnodes[i].end = nodes[i].end; - } -} -#endif /* CONFIG_NUMA_EMU */ - -/* Use the information discovered above to actually set up the nodes. */ -int __init acpi_scan_nodes(unsigned long start, unsigned long end) +int __init x86_acpi_numa_init(void) { - int i; - - if (acpi_numa <= 0) - return -1; - - /* First clean up the node list */ - for (i = 0; i < MAX_NUMNODES; i++) - cutoff_node(i, start, end); - - /* - * Join together blocks on the same node, holes between - * which don't overlap with memory on other nodes. - */ - for (i = 0; i < num_node_memblks; ++i) { - int j, k; - - for (j = i + 1; j < num_node_memblks; ++j) { - unsigned long start, end; - - if (memblk_nodeid[i] != memblk_nodeid[j]) - continue; - start = min(node_memblk_range[i].end, - node_memblk_range[j].end); - end = max(node_memblk_range[i].start, - node_memblk_range[j].start); - for (k = 0; k < num_node_memblks; ++k) { - if (memblk_nodeid[i] == memblk_nodeid[k]) - continue; - if (start < node_memblk_range[k].end && - end > node_memblk_range[k].start) - break; - } - if (k < num_node_memblks) - continue; - start = min(node_memblk_range[i].start, - node_memblk_range[j].start); - end = max(node_memblk_range[i].end, - node_memblk_range[j].end); - printk(KERN_INFO "SRAT: Node %d " - "[%Lx,%Lx) + [%Lx,%Lx) -> [%lx,%lx)\n", - memblk_nodeid[i], - node_memblk_range[i].start, - node_memblk_range[i].end, - node_memblk_range[j].start, - node_memblk_range[j].end, - start, end); - node_memblk_range[i].start = start; - node_memblk_range[i].end = end; - k = --num_node_memblks - j; - memmove(memblk_nodeid + j, memblk_nodeid + j+1, - k * sizeof(*memblk_nodeid)); - memmove(node_memblk_range + j, node_memblk_range + j+1, - k * sizeof(*node_memblk_range)); - --j; - } - } - - memnode_shift = compute_hash_shift(node_memblk_range, num_node_memblks, - memblk_nodeid); - if (memnode_shift < 0) { - printk(KERN_ERR - "SRAT: No NUMA node hash function found. Contact maintainer\n"); - bad_srat(); - return -1; - } - - for (i = 0; i < num_node_memblks; i++) - memblock_x86_register_active_regions(memblk_nodeid[i], - node_memblk_range[i].start >> PAGE_SHIFT, - node_memblk_range[i].end >> PAGE_SHIFT); - - /* for out of order entries in SRAT */ - sort_node_map(); - if (!nodes_cover_memory(nodes)) { - bad_srat(); - return -1; - } + int ret; - /* Account for nodes with cpus and no memory */ - nodes_or(node_possible_map, nodes_parsed, cpu_nodes_parsed); - - /* Finally register nodes */ - for_each_node_mask(i, node_possible_map) - setup_node_bootmem(i, nodes[i].start, nodes[i].end); - /* Try again in case setup_node_bootmem missed one due - to missing bootmem */ - for_each_node_mask(i, node_possible_map) - if (!node_online(i)) - setup_node_bootmem(i, nodes[i].start, nodes[i].end); - - for (i = 0; i < nr_cpu_ids; i++) { - int node = early_cpu_to_node(i); - - if (node == NUMA_NO_NODE) - continue; - if (!node_online(node)) - numa_clear_node(i); - } - numa_init_array(); - return 0; -} - -#ifdef CONFIG_NUMA_EMU -static int fake_node_to_pxm_map[MAX_NUMNODES] __initdata = { - [0 ... MAX_NUMNODES-1] = PXM_INVAL -}; -static s16 fake_apicid_to_node[MAX_LOCAL_APIC] __initdata = { - [0 ... MAX_LOCAL_APIC-1] = NUMA_NO_NODE -}; -static int __init find_node_by_addr(unsigned long addr) -{ - int ret = NUMA_NO_NODE; - int i; - - for_each_node_mask(i, nodes_parsed) { - /* - * Find the real node that this emulated node appears on. For - * the sake of simplicity, we only use a real node's starting - * address to determine which emulated node it appears on. - */ - if (addr >= nodes[i].start && addr < nodes[i].end) { - ret = i; - break; - } - } - return ret; + ret = acpi_numa_init(); + if (ret < 0) + return ret; + return srat_disabled() ? -EINVAL : 0; } -/* - * In NUMA emulation, we need to setup proximity domain (_PXM) to node ID - * mappings that respect the real ACPI topology but reflect our emulated - * environment. For each emulated node, we find which real node it appears on - * and create PXM to NID mappings for those fake nodes which mirror that - * locality. SLIT will now represent the correct distances between emulated - * nodes as a result of the real topology. - */ -void __init acpi_fake_nodes(const struct bootnode *fake_nodes, int num_nodes) -{ - int i, j; - - for (i = 0; i < num_nodes; i++) { - int nid, pxm; - - nid = find_node_by_addr(fake_nodes[i].start); - if (nid == NUMA_NO_NODE) - continue; - pxm = node_to_pxm(nid); - if (pxm == PXM_INVAL) - continue; - fake_node_to_pxm_map[i] = pxm; - /* - * For each apicid_to_node mapping that exists for this real - * node, it must now point to the fake node ID. - */ - for (j = 0; j < MAX_LOCAL_APIC; j++) - if (apicid_to_node[j] == nid && - fake_apicid_to_node[j] == NUMA_NO_NODE) - fake_apicid_to_node[j] = i; - } - - /* - * If there are apicid-to-node mappings for physical nodes that do not - * have a corresponding emulated node, it should default to a guaranteed - * value. - */ - for (i = 0; i < MAX_LOCAL_APIC; i++) - 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)); - - nodes_clear(nodes_parsed); - for (i = 0; i < num_nodes; i++) - if (fake_nodes[i].start != fake_nodes[i].end) - node_set(i, nodes_parsed); -} - -static int null_slit_node_compare(int a, int b) -{ - return node_to_pxm(a) == node_to_pxm(b); -} -#else -static int null_slit_node_compare(int a, int b) -{ - return a == b; -} -#endif /* CONFIG_NUMA_EMU */ - -int __node_distance(int a, int b) -{ - int index; - - if (!acpi_slit) - return null_slit_node_compare(a, b) ? LOCAL_DISTANCE : - REMOTE_DISTANCE; - index = acpi_slit->locality_count * node_to_pxm(a); - return acpi_slit->entry[index + node_to_pxm(b)]; -} - -EXPORT_SYMBOL(__node_distance); - #if defined(CONFIG_MEMORY_HOTPLUG_SPARSE) || defined(CONFIG_ACPI_HOTPLUG_MEMORY) int memory_add_physaddr_to_nid(u64 start) { diff --git a/arch/x86/mm/tlb.c b/arch/x86/mm/tlb.c index 6acc724..d6c0418 100644 --- a/arch/x86/mm/tlb.c +++ b/arch/x86/mm/tlb.c @@ -179,12 +179,8 @@ static void flush_tlb_others_ipi(const struct cpumask *cpumask, sender = this_cpu_read(tlb_vector_offset); f = &flush_state[sender]; - /* - * Could avoid this lock when - * num_online_cpus() <= NUM_INVALIDATE_TLB_VECTORS, but it is - * probably not worth checking this for a cache-hot lock. - */ - raw_spin_lock(&f->tlbstate_lock); + if (nr_cpu_ids > NUM_INVALIDATE_TLB_VECTORS) + raw_spin_lock(&f->tlbstate_lock); f->flush_mm = mm; f->flush_va = va; @@ -202,7 +198,8 @@ static void flush_tlb_others_ipi(const struct cpumask *cpumask, f->flush_mm = NULL; f->flush_va = 0; - raw_spin_unlock(&f->tlbstate_lock); + if (nr_cpu_ids > NUM_INVALIDATE_TLB_VECTORS) + raw_spin_unlock(&f->tlbstate_lock); } void native_flush_tlb_others(const struct cpumask *cpumask, @@ -211,11 +208,10 @@ void native_flush_tlb_others(const struct cpumask *cpumask, if (is_uv_system()) { unsigned int cpu; - cpu = get_cpu(); + cpu = smp_processor_id(); cpumask = uv_flush_tlb_others(cpumask, mm, va, cpu); if (cpumask) flush_tlb_others_ipi(cpumask, mm, va); - put_cpu(); return; } flush_tlb_others_ipi(cpumask, mm, va); diff --git a/arch/x86/pci/amd_bus.c b/arch/x86/pci/amd_bus.c index e27dffb..026e493 100644 --- a/arch/x86/pci/amd_bus.c +++ b/arch/x86/pci/amd_bus.c @@ -350,7 +350,7 @@ static int __init early_fill_mp_bus_info(void) #define ENABLE_CF8_EXT_CFG (1ULL << 46) -static void enable_pci_io_ecs(void *unused) +static void __cpuinit enable_pci_io_ecs(void *unused) { u64 reg; rdmsrl(MSR_AMD64_NB_CFG, reg); diff --git a/arch/x86/xen/mmu.c b/arch/x86/xen/mmu.c index 832765c..3f6f334 100644 --- a/arch/x86/xen/mmu.c +++ b/arch/x86/xen/mmu.c @@ -1491,7 +1491,7 @@ static __init pte_t mask_rw_pte(pte_t *ptep, pte_t pte) * early_ioremap fixmap slot, make sure it is RO. */ if (!is_early_ioremap_ptep(ptep) && - pfn >= e820_table_start && pfn < e820_table_end) + pfn >= pgt_buf_start && pfn < pgt_buf_end) pte = pte_wrprotect(pte); return pte; diff --git a/drivers/acpi/numa.c b/drivers/acpi/numa.c index 5eb25eb..3b5c318 100644 --- a/drivers/acpi/numa.c +++ b/drivers/acpi/numa.c @@ -274,7 +274,7 @@ acpi_table_parse_srat(enum acpi_srat_type id, int __init acpi_numa_init(void) { - int ret = 0; + int cnt = 0; /* * Should not limit number with cpu num that is from NR_CPUS or nr_cpus= @@ -288,7 +288,7 @@ int __init acpi_numa_init(void) acpi_parse_x2apic_affinity, 0); acpi_table_parse_srat(ACPI_SRAT_TYPE_CPU_AFFINITY, acpi_parse_processor_affinity, 0); - ret = acpi_table_parse_srat(ACPI_SRAT_TYPE_MEMORY_AFFINITY, + cnt = acpi_table_parse_srat(ACPI_SRAT_TYPE_MEMORY_AFFINITY, acpi_parse_memory_affinity, NR_NODE_MEMBLKS); } @@ -297,7 +297,10 @@ int __init acpi_numa_init(void) acpi_table_parse(ACPI_SIG_SLIT, acpi_parse_slit); acpi_numa_arch_fixup(); - return ret; + + if (cnt <= 0) + return cnt ?: -ENOENT; + return 0; } int acpi_get_pxm(acpi_handle h) diff --git a/include/linux/mm.h b/include/linux/mm.h index f6385fc..679300c 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -1309,8 +1309,6 @@ int add_from_early_node_map(struct range *range, int az, int nr_range, int nid); u64 __init find_memory_core_early(int nid, u64 size, u64 align, u64 goal, u64 limit); -void *__alloc_memory_core_early(int nodeid, u64 size, u64 align, - u64 goal, u64 limit); typedef int (*work_fn_t)(unsigned long, unsigned long, void *); extern void work_with_active_regions(int nid, work_fn_t work_fn, void *data); extern void sparse_memory_present_with_active_regions(int nid); diff --git a/include/linux/pci_ids.h b/include/linux/pci_ids.h index 3adb06e..580de67 100644 --- a/include/linux/pci_ids.h +++ b/include/linux/pci_ids.h @@ -518,6 +518,7 @@ #define PCI_DEVICE_ID_AMD_11H_NB_MISC 0x1303 #define PCI_DEVICE_ID_AMD_11H_NB_LINK 0x1304 #define PCI_DEVICE_ID_AMD_15H_NB_MISC 0x1603 +#define PCI_DEVICE_ID_AMD_15H_NB_LINK 0x1604 #define PCI_DEVICE_ID_AMD_CNB17H_F3 0x1703 #define PCI_DEVICE_ID_AMD_LANCE 0x2000 #define PCI_DEVICE_ID_AMD_LANCE_HOME 0x2001 diff --git a/mm/Makefile b/mm/Makefile index 2b1b575..42a8326 100644 --- a/mm/Makefile +++ b/mm/Makefile @@ -7,7 +7,7 @@ mmu-$(CONFIG_MMU) := fremap.o highmem.o madvise.o memory.o mincore.o \ mlock.o mmap.o mprotect.o mremap.o msync.o rmap.o \ vmalloc.o pagewalk.o pgtable-generic.o -obj-y := bootmem.o filemap.o mempool.o oom_kill.o fadvise.o \ +obj-y := filemap.o mempool.o oom_kill.o fadvise.o \ maccess.o page_alloc.o page-writeback.o \ readahead.o swap.o truncate.o vmscan.o shmem.o \ prio_tree.o util.o mmzone.o vmstat.o backing-dev.o \ @@ -15,6 +15,12 @@ obj-y := bootmem.o filemap.o mempool.o oom_kill.o fadvise.o \ $(mmu-y) obj-y += init-mm.o +ifdef CONFIG_NO_BOOTMEM + obj-y += nobootmem.o +else + obj-y += bootmem.o +endif + obj-$(CONFIG_HAVE_MEMBLOCK) += memblock.o obj-$(CONFIG_BOUNCE) += bounce.o diff --git a/mm/bootmem.c b/mm/bootmem.c index 13b0caa..07aeb89 100644 --- a/mm/bootmem.c +++ b/mm/bootmem.c @@ -23,6 +23,13 @@ #include "internal.h" +#ifndef CONFIG_NEED_MULTIPLE_NODES +struct pglist_data __refdata contig_page_data = { + .bdata = &bootmem_node_data[0] +}; +EXPORT_SYMBOL(contig_page_data); +#endif + unsigned long max_low_pfn; unsigned long min_low_pfn; unsigned long max_pfn; @@ -35,7 +42,6 @@ unsigned long max_pfn; unsigned long saved_max_pfn; #endif -#ifndef CONFIG_NO_BOOTMEM bootmem_data_t bootmem_node_data[MAX_NUMNODES] __initdata; static struct list_head bdata_list __initdata = LIST_HEAD_INIT(bdata_list); @@ -146,7 +152,7 @@ unsigned long __init init_bootmem(unsigned long start, unsigned long pages) min_low_pfn = start; return init_bootmem_core(NODE_DATA(0)->bdata, start, 0, pages); } -#endif + /* * free_bootmem_late - free bootmem pages directly to page allocator * @addr: starting address of the range @@ -171,53 +177,6 @@ void __init free_bootmem_late(unsigned long addr, unsigned long size) } } -#ifdef CONFIG_NO_BOOTMEM -static void __init __free_pages_memory(unsigned long start, unsigned long end) -{ - int i; - unsigned long start_aligned, end_aligned; - int order = ilog2(BITS_PER_LONG); - - start_aligned = (start + (BITS_PER_LONG - 1)) & ~(BITS_PER_LONG - 1); - end_aligned = end & ~(BITS_PER_LONG - 1); - - if (end_aligned <= start_aligned) { - for (i = start; i < end; i++) - __free_pages_bootmem(pfn_to_page(i), 0); - - return; - } - - for (i = start; i < start_aligned; i++) - __free_pages_bootmem(pfn_to_page(i), 0); - - for (i = start_aligned; i < end_aligned; i += BITS_PER_LONG) - __free_pages_bootmem(pfn_to_page(i), order); - - for (i = end_aligned; i < end; i++) - __free_pages_bootmem(pfn_to_page(i), 0); -} - -unsigned long __init free_all_memory_core_early(int nodeid) -{ - int i; - u64 start, end; - unsigned long count = 0; - struct range *range = NULL; - int nr_range; - - nr_range = get_free_all_memory_range(&range, nodeid); - - for (i = 0; i < nr_range; i++) { - start = range[i].start; - end = range[i].end; - count += end - start; - __free_pages_memory(start, end); - } - - return count; -} -#else static unsigned long __init free_all_bootmem_core(bootmem_data_t *bdata) { int aligned; @@ -278,7 +237,6 @@ static unsigned long __init free_all_bootmem_core(bootmem_data_t *bdata) return count; } -#endif /** * free_all_bootmem_node - release a node's free pages to the buddy allocator @@ -289,12 +247,7 @@ static unsigned long __init free_all_bootmem_core(bootmem_data_t *bdata) unsigned long __init free_all_bootmem_node(pg_data_t *pgdat) { register_page_bootmem_info_node(pgdat); -#ifdef CONFIG_NO_BOOTMEM - /* free_all_memory_core_early(MAX_NUMNODES) will be called later */ - return 0; -#else return free_all_bootmem_core(pgdat->bdata); -#endif } /** @@ -304,16 +257,6 @@ unsigned long __init free_all_bootmem_node(pg_data_t *pgdat) */ unsigned long __init free_all_bootmem(void) { -#ifdef CONFIG_NO_BOOTMEM - /* - * We need to use MAX_NUMNODES instead of NODE_DATA(0)->node_id - * because in some case like Node0 doesnt have RAM installed - * low ram will be on Node1 - * Use MAX_NUMNODES will make sure all ranges in early_node_map[] - * will be used instead of only Node0 related - */ - return free_all_memory_core_early(MAX_NUMNODES); -#else unsigned long total_pages = 0; bootmem_data_t *bdata; @@ -321,10 +264,8 @@ unsigned long __init free_all_bootmem(void) total_pages += free_all_bootmem_core(bdata); return total_pages; -#endif } -#ifndef CONFIG_NO_BOOTMEM static void __init __free(bootmem_data_t *bdata, unsigned long sidx, unsigned long eidx) { @@ -419,7 +360,6 @@ static int __init mark_bootmem(unsigned long start, unsigned long end, } BUG(); } -#endif /** * free_bootmem_node - mark a page range as usable @@ -434,10 +374,6 @@ static int __init mark_bootmem(unsigned long start, unsigned long end, void __init free_bootmem_node(pg_data_t *pgdat, unsigned long physaddr, unsigned long size) { -#ifdef CONFIG_NO_BOOTMEM - kmemleak_free_part(__va(physaddr), size); - memblock_x86_free_range(physaddr, physaddr + size); -#else unsigned long start, end; kmemleak_free_part(__va(physaddr), size); @@ -446,7 +382,6 @@ void __init free_bootmem_node(pg_data_t *pgdat, unsigned long physaddr, end = PFN_DOWN(physaddr + size); mark_bootmem_node(pgdat->bdata, start, end, 0, 0); -#endif } /** @@ -460,10 +395,6 @@ void __init free_bootmem_node(pg_data_t *pgdat, unsigned long physaddr, */ void __init free_bootmem(unsigned long addr, unsigned long size) { -#ifdef CONFIG_NO_BOOTMEM - kmemleak_free_part(__va(addr), size); - memblock_x86_free_range(addr, addr + size); -#else unsigned long start, end; kmemleak_free_part(__va(addr), size); @@ -472,7 +403,6 @@ void __init free_bootmem(unsigned long addr, unsigned long size) end = PFN_DOWN(addr + size); mark_bootmem(start, end, 0, 0); -#endif } /** @@ -489,17 +419,12 @@ void __init free_bootmem(unsigned long addr, unsigned long size) int __init reserve_bootmem_node(pg_data_t *pgdat, unsigned long physaddr, unsigned long size, int flags) { -#ifdef CONFIG_NO_BOOTMEM - panic("no bootmem"); - return 0; -#else unsigned long start, end; start = PFN_DOWN(physaddr); end = PFN_UP(physaddr + size); return mark_bootmem_node(pgdat->bdata, start, end, 1, flags); -#endif } /** @@ -515,20 +440,14 @@ int __init reserve_bootmem_node(pg_data_t *pgdat, unsigned long physaddr, int __init reserve_bootmem(unsigned long addr, unsigned long size, int flags) { -#ifdef CONFIG_NO_BOOTMEM - panic("no bootmem"); - return 0; -#else unsigned long start, end; start = PFN_DOWN(addr); end = PFN_UP(addr + size); return mark_bootmem(start, end, 1, flags); -#endif } -#ifndef CONFIG_NO_BOOTMEM int __weak __init reserve_bootmem_generic(unsigned long phys, unsigned long len, int flags) { @@ -685,33 +604,12 @@ static void * __init alloc_arch_preferred_bootmem(bootmem_data_t *bdata, #endif return NULL; } -#endif static void * __init ___alloc_bootmem_nopanic(unsigned long size, unsigned long align, unsigned long goal, unsigned long limit) { -#ifdef CONFIG_NO_BOOTMEM - void *ptr; - - if (WARN_ON_ONCE(slab_is_available())) - return kzalloc(size, GFP_NOWAIT); - -restart: - - ptr = __alloc_memory_core_early(MAX_NUMNODES, size, align, goal, limit); - - if (ptr) - return ptr; - - if (goal != 0) { - goal = 0; - goto restart; - } - - return NULL; -#else bootmem_data_t *bdata; void *region; @@ -737,7 +635,6 @@ restart: } return NULL; -#endif } /** @@ -758,10 +655,6 @@ void * __init __alloc_bootmem_nopanic(unsigned long size, unsigned long align, { unsigned long limit = 0; -#ifdef CONFIG_NO_BOOTMEM - limit = -1UL; -#endif - return ___alloc_bootmem_nopanic(size, align, goal, limit); } @@ -798,14 +691,9 @@ void * __init __alloc_bootmem(unsigned long size, unsigned long align, { unsigned long limit = 0; -#ifdef CONFIG_NO_BOOTMEM - limit = -1UL; -#endif - return ___alloc_bootmem(size, align, goal, limit); } -#ifndef CONFIG_NO_BOOTMEM static void * __init ___alloc_bootmem_node(bootmem_data_t *bdata, unsigned long size, unsigned long align, unsigned long goal, unsigned long limit) @@ -822,7 +710,6 @@ static void * __init ___alloc_bootmem_node(bootmem_data_t *bdata, return ___alloc_bootmem(size, align, goal, limit); } -#endif /** * __alloc_bootmem_node - allocate boot memory from a specific node @@ -842,24 +729,10 @@ static void * __init ___alloc_bootmem_node(bootmem_data_t *bdata, void * __init __alloc_bootmem_node(pg_data_t *pgdat, unsigned long size, unsigned long align, unsigned long goal) { - void *ptr; - if (WARN_ON_ONCE(slab_is_available())) return kzalloc_node(size, GFP_NOWAIT, pgdat->node_id); -#ifdef CONFIG_NO_BOOTMEM - ptr = __alloc_memory_core_early(pgdat->node_id, size, align, - goal, -1ULL); - if (ptr) - return ptr; - - ptr = __alloc_memory_core_early(MAX_NUMNODES, size, align, - goal, -1ULL); -#else - ptr = ___alloc_bootmem_node(pgdat->bdata, size, align, goal, 0); -#endif - - return ptr; + return ___alloc_bootmem_node(pgdat->bdata, size, align, goal, 0); } void * __init __alloc_bootmem_node_high(pg_data_t *pgdat, unsigned long size, @@ -880,13 +753,8 @@ void * __init __alloc_bootmem_node_high(pg_data_t *pgdat, unsigned long size, unsigned long new_goal; new_goal = MAX_DMA32_PFN << PAGE_SHIFT; -#ifdef CONFIG_NO_BOOTMEM - ptr = __alloc_memory_core_early(pgdat->node_id, size, align, - new_goal, -1ULL); -#else ptr = alloc_bootmem_core(pgdat->bdata, size, align, new_goal, 0); -#endif if (ptr) return ptr; } @@ -907,16 +775,6 @@ void * __init __alloc_bootmem_node_high(pg_data_t *pgdat, unsigned long size, void * __init alloc_bootmem_section(unsigned long size, unsigned long section_nr) { -#ifdef CONFIG_NO_BOOTMEM - unsigned long pfn, goal, limit; - - pfn = section_nr_to_pfn(section_nr); - goal = pfn << PAGE_SHIFT; - limit = section_nr_to_pfn(section_nr + 1) << PAGE_SHIFT; - - return __alloc_memory_core_early(early_pfn_to_nid(pfn), size, - SMP_CACHE_BYTES, goal, limit); -#else bootmem_data_t *bdata; unsigned long pfn, goal, limit; @@ -926,7 +784,6 @@ void * __init alloc_bootmem_section(unsigned long size, bdata = &bootmem_node_data[early_pfn_to_nid(pfn)]; return alloc_bootmem_core(bdata, size, SMP_CACHE_BYTES, goal, limit); -#endif } #endif @@ -938,16 +795,11 @@ void * __init __alloc_bootmem_node_nopanic(pg_data_t *pgdat, unsigned long size, if (WARN_ON_ONCE(slab_is_available())) return kzalloc_node(size, GFP_NOWAIT, pgdat->node_id); -#ifdef CONFIG_NO_BOOTMEM - ptr = __alloc_memory_core_early(pgdat->node_id, size, align, - goal, -1ULL); -#else ptr = alloc_arch_preferred_bootmem(pgdat->bdata, size, align, goal, 0); if (ptr) return ptr; ptr = alloc_bootmem_core(pgdat->bdata, size, align, goal, 0); -#endif if (ptr) return ptr; @@ -995,21 +847,9 @@ void * __init __alloc_bootmem_low(unsigned long size, unsigned long align, void * __init __alloc_bootmem_low_node(pg_data_t *pgdat, unsigned long size, unsigned long align, unsigned long goal) { - void *ptr; - if (WARN_ON_ONCE(slab_is_available())) return kzalloc_node(size, GFP_NOWAIT, pgdat->node_id); -#ifdef CONFIG_NO_BOOTMEM - ptr = __alloc_memory_core_early(pgdat->node_id, size, align, + return ___alloc_bootmem_node(pgdat->bdata, size, align, goal, ARCH_LOW_ADDRESS_LIMIT); - if (ptr) - return ptr; - ptr = __alloc_memory_core_early(MAX_NUMNODES, size, align, - goal, ARCH_LOW_ADDRESS_LIMIT); -#else - ptr = ___alloc_bootmem_node(pgdat->bdata, size, align, - goal, ARCH_LOW_ADDRESS_LIMIT); -#endif - return ptr; } diff --git a/mm/nobootmem.c b/mm/nobootmem.c new file mode 100644 index 0000000..e2bdb070 --- /dev/null +++ b/mm/nobootmem.c @@ -0,0 +1,435 @@ +/* + * bootmem - A boot-time physical memory allocator and configurator + * + * Copyright (C) 1999 Ingo Molnar + * 1999 Kanoj Sarcar, SGI + * 2008 Johannes Weiner + * + * Access to this subsystem has to be serialized externally (which is true + * for the boot process anyway). + */ +#include <linux/init.h> +#include <linux/pfn.h> +#include <linux/slab.h> +#include <linux/bootmem.h> +#include <linux/module.h> +#include <linux/kmemleak.h> +#include <linux/range.h> +#include <linux/memblock.h> + +#include <asm/bug.h> +#include <asm/io.h> +#include <asm/processor.h> + +#include "internal.h" + +#ifndef CONFIG_NEED_MULTIPLE_NODES +struct pglist_data __refdata contig_page_data; +EXPORT_SYMBOL(contig_page_data); +#endif + +unsigned long max_low_pfn; +unsigned long min_low_pfn; +unsigned long max_pfn; + +#ifdef CONFIG_CRASH_DUMP +/* + * If we have booted due to a crash, max_pfn will be a very low value. We need + * to know the amount of memory that the previous kernel used. + */ +unsigned long saved_max_pfn; +#endif + +static void * __init __alloc_memory_core_early(int nid, u64 size, u64 align, + u64 goal, u64 limit) +{ + void *ptr; + u64 addr; + + if (limit > memblock.current_limit) + limit = memblock.current_limit; + + addr = find_memory_core_early(nid, size, align, goal, limit); + + if (addr == MEMBLOCK_ERROR) + return NULL; + + ptr = phys_to_virt(addr); + memset(ptr, 0, size); + memblock_x86_reserve_range(addr, addr + size, "BOOTMEM"); + /* + * The min_count is set to 0 so that bootmem allocated blocks + * are never reported as leaks. + */ + kmemleak_alloc(ptr, size, 0, 0); + return ptr; +} + +/* + * free_bootmem_late - free bootmem pages directly to page allocator + * @addr: starting address of the range + * @size: size of the range in bytes + * + * This is only useful when the bootmem allocator has already been torn + * down, but we are still initializing the system. Pages are given directly + * to the page allocator, no bootmem metadata is updated because it is gone. + */ +void __init free_bootmem_late(unsigned long addr, unsigned long size) +{ + unsigned long cursor, end; + + kmemleak_free_part(__va(addr), size); + + cursor = PFN_UP(addr); + end = PFN_DOWN(addr + size); + + for (; cursor < end; cursor++) { + __free_pages_bootmem(pfn_to_page(cursor), 0); + totalram_pages++; + } +} + +static void __init __free_pages_memory(unsigned long start, unsigned long end) +{ + int i; + unsigned long start_aligned, end_aligned; + int order = ilog2(BITS_PER_LONG); + + start_aligned = (start + (BITS_PER_LONG - 1)) & ~(BITS_PER_LONG - 1); + end_aligned = end & ~(BITS_PER_LONG - 1); + + if (end_aligned <= start_aligned) { + for (i = start; i < end; i++) + __free_pages_bootmem(pfn_to_page(i), 0); + + return; + } + + for (i = start; i < start_aligned; i++) + __free_pages_bootmem(pfn_to_page(i), 0); + + for (i = start_aligned; i < end_aligned; i += BITS_PER_LONG) + __free_pages_bootmem(pfn_to_page(i), order); + + for (i = end_aligned; i < end; i++) + __free_pages_bootmem(pfn_to_page(i), 0); +} + +unsigned long __init free_all_memory_core_early(int nodeid) +{ + int i; + u64 start, end; + unsigned long count = 0; + struct range *range = NULL; + int nr_range; + + nr_range = get_free_all_memory_range(&range, nodeid); + + for (i = 0; i < nr_range; i++) { + start = range[i].start; + end = range[i].end; + count += end - start; + __free_pages_memory(start, end); + } + + return count; +} + +/** + * free_all_bootmem_node - release a node's free pages to the buddy allocator + * @pgdat: node to be released + * + * Returns the number of pages actually released. + */ +unsigned long __init free_all_bootmem_node(pg_data_t *pgdat) +{ + register_page_bootmem_info_node(pgdat); + + /* free_all_memory_core_early(MAX_NUMNODES) will be called later */ + return 0; +} + +/** + * free_all_bootmem - release free pages to the buddy allocator + * + * Returns the number of pages actually released. + */ +unsigned long __init free_all_bootmem(void) +{ + /* + * We need to use MAX_NUMNODES instead of NODE_DATA(0)->node_id + * because in some case like Node0 doesnt have RAM installed + * low ram will be on Node1 + * Use MAX_NUMNODES will make sure all ranges in early_node_map[] + * will be used instead of only Node0 related + */ + return free_all_memory_core_early(MAX_NUMNODES); +} + +/** + * free_bootmem_node - mark a page range as usable + * @pgdat: node the range resides on + * @physaddr: starting address of the range + * @size: size of the range in bytes + * + * Partial pages will be considered reserved and left as they are. + * + * The range must reside completely on the specified node. + */ +void __init free_bootmem_node(pg_data_t *pgdat, unsigned long physaddr, + unsigned long size) +{ + kmemleak_free_part(__va(physaddr), size); + memblock_x86_free_range(physaddr, physaddr + size); +} + +/** + * free_bootmem - mark a page range as usable + * @addr: starting address of the range + * @size: size of the range in bytes + * + * Partial pages will be considered reserved and left as they are. + * + * The range must be contiguous but may span node boundaries. + */ +void __init free_bootmem(unsigned long addr, unsigned long size) +{ + kmemleak_free_part(__va(addr), size); + memblock_x86_free_range(addr, addr + size); +} + +static void * __init ___alloc_bootmem_nopanic(unsigned long size, + unsigned long align, + unsigned long goal, + unsigned long limit) +{ + void *ptr; + + if (WARN_ON_ONCE(slab_is_available())) + return kzalloc(size, GFP_NOWAIT); + +restart: + + ptr = __alloc_memory_core_early(MAX_NUMNODES, size, align, goal, limit); + + if (ptr) + return ptr; + + if (goal != 0) { + goal = 0; + goto restart; + } + + return NULL; +} + +/** + * __alloc_bootmem_nopanic - allocate boot memory without panicking + * @size: size of the request in bytes + * @align: alignment of the region + * @goal: preferred starting address of the region + * + * The goal is dropped if it can not be satisfied and the allocation will + * fall back to memory below @goal. + * + * Allocation may happen on any node in the system. + * + * Returns NULL on failure. + */ +void * __init __alloc_bootmem_nopanic(unsigned long size, unsigned long align, + unsigned long goal) +{ + unsigned long limit = -1UL; + + return ___alloc_bootmem_nopanic(size, align, goal, limit); +} + +static void * __init ___alloc_bootmem(unsigned long size, unsigned long align, + unsigned long goal, unsigned long limit) +{ + void *mem = ___alloc_bootmem_nopanic(size, align, goal, limit); + + if (mem) + return mem; + /* + * Whoops, we cannot satisfy the allocation request. + */ + printk(KERN_ALERT "bootmem alloc of %lu bytes failed!\n", size); + panic("Out of memory"); + return NULL; +} + +/** + * __alloc_bootmem - allocate boot memory + * @size: size of the request in bytes + * @align: alignment of the region + * @goal: preferred starting address of the region + * + * The goal is dropped if it can not be satisfied and the allocation will + * fall back to memory below @goal. + * + * Allocation may happen on any node in the system. + * + * The function panics if the request can not be satisfied. + */ +void * __init __alloc_bootmem(unsigned long size, unsigned long align, + unsigned long goal) +{ + unsigned long limit = -1UL; + + return ___alloc_bootmem(size, align, goal, limit); +} + +/** + * __alloc_bootmem_node - allocate boot memory from a specific node + * @pgdat: node to allocate from + * @size: size of the request in bytes + * @align: alignment of the region + * @goal: preferred starting address of the region + * + * The goal is dropped if it can not be satisfied and the allocation will + * fall back to memory below @goal. + * + * Allocation may fall back to any node in the system if the specified node + * can not hold the requested memory. + * + * The function panics if the request can not be satisfied. + */ +void * __init __alloc_bootmem_node(pg_data_t *pgdat, unsigned long size, + unsigned long align, unsigned long goal) +{ + void *ptr; + + if (WARN_ON_ONCE(slab_is_available())) + return kzalloc_node(size, GFP_NOWAIT, pgdat->node_id); + + ptr = __alloc_memory_core_early(pgdat->node_id, size, align, + goal, -1ULL); + if (ptr) + return ptr; + + return __alloc_memory_core_early(MAX_NUMNODES, size, align, + goal, -1ULL); +} + +void * __init __alloc_bootmem_node_high(pg_data_t *pgdat, unsigned long size, + unsigned long align, unsigned long goal) +{ +#ifdef MAX_DMA32_PFN + unsigned long end_pfn; + + if (WARN_ON_ONCE(slab_is_available())) + return kzalloc_node(size, GFP_NOWAIT, pgdat->node_id); + + /* update goal according ...MAX_DMA32_PFN */ + end_pfn = pgdat->node_start_pfn + pgdat->node_spanned_pages; + + if (end_pfn > MAX_DMA32_PFN + (128 >> (20 - PAGE_SHIFT)) && + (goal >> PAGE_SHIFT) < MAX_DMA32_PFN) { + void *ptr; + unsigned long new_goal; + + new_goal = MAX_DMA32_PFN << PAGE_SHIFT; + ptr = __alloc_memory_core_early(pgdat->node_id, size, align, + new_goal, -1ULL); + if (ptr) + return ptr; + } +#endif + + return __alloc_bootmem_node(pgdat, size, align, goal); + +} + +#ifdef CONFIG_SPARSEMEM +/** + * alloc_bootmem_section - allocate boot memory from a specific section + * @size: size of the request in bytes + * @section_nr: sparse map section to allocate from + * + * Return NULL on failure. + */ +void * __init alloc_bootmem_section(unsigned long size, + unsigned long section_nr) +{ + unsigned long pfn, goal, limit; + + pfn = section_nr_to_pfn(section_nr); + goal = pfn << PAGE_SHIFT; + limit = section_nr_to_pfn(section_nr + 1) << PAGE_SHIFT; + + return __alloc_memory_core_early(early_pfn_to_nid(pfn), size, + SMP_CACHE_BYTES, goal, limit); +} +#endif + +void * __init __alloc_bootmem_node_nopanic(pg_data_t *pgdat, unsigned long size, + unsigned long align, unsigned long goal) +{ + void *ptr; + + if (WARN_ON_ONCE(slab_is_available())) + return kzalloc_node(size, GFP_NOWAIT, pgdat->node_id); + + ptr = __alloc_memory_core_early(pgdat->node_id, size, align, + goal, -1ULL); + if (ptr) + return ptr; + + return __alloc_bootmem_nopanic(size, align, goal); +} + +#ifndef ARCH_LOW_ADDRESS_LIMIT +#define ARCH_LOW_ADDRESS_LIMIT 0xffffffffUL +#endif + +/** + * __alloc_bootmem_low - allocate low boot memory + * @size: size of the request in bytes + * @align: alignment of the region + * @goal: preferred starting address of the region + * + * The goal is dropped if it can not be satisfied and the allocation will + * fall back to memory below @goal. + * + * Allocation may happen on any node in the system. + * + * The function panics if the request can not be satisfied. + */ +void * __init __alloc_bootmem_low(unsigned long size, unsigned long align, + unsigned long goal) +{ + return ___alloc_bootmem(size, align, goal, ARCH_LOW_ADDRESS_LIMIT); +} + +/** + * __alloc_bootmem_low_node - allocate low boot memory from a specific node + * @pgdat: node to allocate from + * @size: size of the request in bytes + * @align: alignment of the region + * @goal: preferred starting address of the region + * + * The goal is dropped if it can not be satisfied and the allocation will + * fall back to memory below @goal. + * + * Allocation may fall back to any node in the system if the specified node + * can not hold the requested memory. + * + * The function panics if the request can not be satisfied. + */ +void * __init __alloc_bootmem_low_node(pg_data_t *pgdat, unsigned long size, + unsigned long align, unsigned long goal) +{ + void *ptr; + + if (WARN_ON_ONCE(slab_is_available())) + return kzalloc_node(size, GFP_NOWAIT, pgdat->node_id); + + ptr = __alloc_memory_core_early(pgdat->node_id, size, align, + goal, ARCH_LOW_ADDRESS_LIMIT); + if (ptr) + return ptr; + + return __alloc_memory_core_early(MAX_NUMNODES, size, align, + goal, ARCH_LOW_ADDRESS_LIMIT); +} diff --git a/mm/page_alloc.c b/mm/page_alloc.c index cdef1d4..bd76256 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -3699,13 +3699,45 @@ void __init free_bootmem_with_active_regions(int nid, } #ifdef CONFIG_HAVE_MEMBLOCK +/* + * Basic iterator support. Return the last range of PFNs for a node + * Note: nid == MAX_NUMNODES returns last region regardless of node + */ +static int __meminit last_active_region_index_in_nid(int nid) +{ + int i; + + for (i = nr_nodemap_entries - 1; i >= 0; i--) + if (nid == MAX_NUMNODES || early_node_map[i].nid == nid) + return i; + + return -1; +} + +/* + * Basic iterator support. Return the previous active range of PFNs for a node + * Note: nid == MAX_NUMNODES returns next region regardless of node + */ +static int __meminit previous_active_region_index_in_nid(int index, int nid) +{ + for (index = index - 1; index >= 0; index--) + if (nid == MAX_NUMNODES || early_node_map[index].nid == nid) + return index; + + return -1; +} + +#define for_each_active_range_index_in_nid_reverse(i, nid) \ + for (i = last_active_region_index_in_nid(nid); i != -1; \ + i = previous_active_region_index_in_nid(i, nid)) + u64 __init find_memory_core_early(int nid, u64 size, u64 align, u64 goal, u64 limit) { int i; /* Need to go over early_node_map to find out good range for node */ - for_each_active_range_index_in_nid(i, nid) { + for_each_active_range_index_in_nid_reverse(i, nid) { u64 addr; u64 ei_start, ei_last; u64 final_start, final_end; @@ -3748,34 +3780,6 @@ int __init add_from_early_node_map(struct range *range, int az, return nr_range; } -#ifdef CONFIG_NO_BOOTMEM -void * __init __alloc_memory_core_early(int nid, u64 size, u64 align, - u64 goal, u64 limit) -{ - void *ptr; - u64 addr; - - if (limit > memblock.current_limit) - limit = memblock.current_limit; - - addr = find_memory_core_early(nid, size, align, goal, limit); - - if (addr == MEMBLOCK_ERROR) - return NULL; - - ptr = phys_to_virt(addr); - memset(ptr, 0, size); - memblock_x86_reserve_range(addr, addr + size, "BOOTMEM"); - /* - * The min_count is set to 0 so that bootmem allocated blocks - * are never reported as leaks. - */ - kmemleak_alloc(ptr, size, 0, 0); - return ptr; -} -#endif - - void __init work_with_active_regions(int nid, work_fn_t work_fn, void *data) { int i; @@ -4809,15 +4813,6 @@ void __init set_dma_reserve(unsigned long new_dma_reserve) dma_reserve = new_dma_reserve; } -#ifndef CONFIG_NEED_MULTIPLE_NODES -struct pglist_data __refdata contig_page_data = { -#ifndef CONFIG_NO_BOOTMEM - .bdata = &bootmem_node_data[0] -#endif - }; -EXPORT_SYMBOL(contig_page_data); -#endif - void __init free_area_init(unsigned long *zones_size) { free_area_init_node(0, zones_size, |