From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752141Ab1GLJQS (ORCPT ); Tue, 12 Jul 2011 05:16:18 -0400 Received: from mail-bw0-f46.google.com ([209.85.214.46]:47219 "EHLO mail-bw0-f46.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751835Ab1GLJQQ (ORCPT ); Tue, 12 Jul 2011 05:16:16 -0400 From: Tejun Heo To: mingo@redhat.com, hpa@zytor.com, tglx@linutronix.de, benh@kernel.crashing.org, yinghai@kernel.org, davem@davemloft.net Cc: linux-kernel@vger.kernel.org, linux-arch@vger.kernel.org, x86@kernel.org, Tejun Heo Subject: [PATCH 03/13] memblock: Add optional region->nid Date: Tue, 12 Jul 2011 11:15:56 +0200 Message-Id: <1310462166-31469-4-git-send-email-tj@kernel.org> X-Mailer: git-send-email 1.7.6 In-Reply-To: <1310462166-31469-1-git-send-email-tj@kernel.org> References: <1310462166-31469-1-git-send-email-tj@kernel.org> Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Add optional region->nid which can be enabled by arch using CONFIG_HAVE_MEMBLOCK_NODE_MAP. When enabled, memblock also carries NUMA node information and replaces early_node_map[]. Newly added memblocks have MAX_NUMNODES as nid. Arch can then call memblock_set_node() to set node information. memblock takes care of merging and node affine allocations w.r.t. node information. When MEMBLOCK_NODE_MAP is enabled, early_node_map[], related data structures and functions to manipulate and iterate it are disabled. memblock version of __next_mem_pfn_range() is provided such that for_each_mem_pfn_range() behaves the same and its users don't have to be updated. Signed-off-by: Tejun Heo Cc: Yinghai Lu Cc: Benjamin Herrenschmidt --- include/linux/memblock.h | 26 +++++++++ include/linux/mm.h | 2 + mm/Kconfig | 3 + mm/memblock.c | 141 ++++++++++++++++++++++++++++++++++++++++------ mm/page_alloc.c | 47 +++++++++------ 5 files changed, 182 insertions(+), 37 deletions(-) diff --git a/include/linux/memblock.h b/include/linux/memblock.h index aa5df9e..e78a9ad 100644 --- a/include/linux/memblock.h +++ b/include/linux/memblock.h @@ -24,6 +24,9 @@ struct memblock_region { phys_addr_t base; phys_addr_t size; +#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP + int nid; +#endif }; struct memblock_type { @@ -58,6 +61,29 @@ extern long memblock_remove(phys_addr_t base, phys_addr_t size); extern long memblock_free(phys_addr_t base, phys_addr_t size); extern long memblock_reserve(phys_addr_t base, phys_addr_t size); +#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP +extern int memblock_set_node(phys_addr_t base, phys_addr_t size, int nid); + +static inline void memblock_set_region_node(struct memblock_region *r, int nid) +{ + r->nid = nid; +} + +static inline int memblock_get_region_node(const struct memblock_region *r) +{ + return r->nid; +} +#else +static inline void memblock_set_region_node(struct memblock_region *r, int nid) +{ +} + +static inline int memblock_get_region_node(const struct memblock_region *r) +{ + return 0; +} +#endif /* CONFIG_HAVE_MEMBLOCK_NODE_MAP */ + /* The numa aware allocator is only available if * CONFIG_ARCH_POPULATES_NODE_MAP is set */ diff --git a/include/linux/mm.h b/include/linux/mm.h index 9ebc65a..ceb1e4a 100644 --- a/include/linux/mm.h +++ b/include/linux/mm.h @@ -1307,12 +1307,14 @@ extern void free_area_init_node(int nid, unsigned long * zones_size, * CONFIG_ARCH_POPULATES_NODE_MAP */ extern void free_area_init_nodes(unsigned long *max_zone_pfn); +#ifndef CONFIG_HAVE_MEMBLOCK_NODE_MAP extern void add_active_range(unsigned int nid, unsigned long start_pfn, unsigned long end_pfn); extern void remove_active_range(unsigned int nid, unsigned long start_pfn, unsigned long end_pfn); extern void remove_all_active_ranges(void); void sort_node_map(void); +#endif unsigned long node_map_pfn_alignment(void); unsigned long __absent_pages_in_range(int nid, unsigned long start_pfn, unsigned long end_pfn); diff --git a/mm/Kconfig b/mm/Kconfig index 8ca47a5..30a5d47 100644 --- a/mm/Kconfig +++ b/mm/Kconfig @@ -131,6 +131,9 @@ config SPARSEMEM_VMEMMAP config HAVE_MEMBLOCK boolean +config HAVE_MEMBLOCK_NODE_MAP + boolean + # eventually, we can have this option just 'select SPARSEMEM' config MEMORY_HOTPLUG bool "Allow for memory hot-add" diff --git a/mm/memblock.c b/mm/memblock.c index 992aa18..766adec 100644 --- a/mm/memblock.c +++ b/mm/memblock.c @@ -161,12 +161,8 @@ int __init_memblock memblock_reserve_reserved_regions(void) static void __init_memblock memblock_remove_region(struct memblock_type *type, unsigned long r) { - unsigned long i; - - for (i = r; i < type->cnt - 1; i++) { - type->regions[i].base = type->regions[i + 1].base; - type->regions[i].size = type->regions[i + 1].size; - } + memmove(&type->regions[r], &type->regions[r + 1], + (type->cnt - (r + 1)) * sizeof(type->regions[r])); type->cnt--; /* Special case for empty arrays */ @@ -174,6 +170,7 @@ static void __init_memblock memblock_remove_region(struct memblock_type *type, u type->cnt = 1; type->regions[0].base = 0; type->regions[0].size = 0; + memblock_set_region_node(&type->regions[0], MAX_NUMNODES); } } @@ -266,7 +263,9 @@ static void __init_memblock memblock_merge_regions(struct memblock_type *type) struct memblock_region *this = &type->regions[i]; struct memblock_region *next = &type->regions[i + 1]; - if (this->base + this->size != next->base) { + if (this->base + this->size != next->base || + memblock_get_region_node(this) != + memblock_get_region_node(next)) { BUG_ON(this->base + this->size > next->base); i++; continue; @@ -290,7 +289,7 @@ static void __init_memblock memblock_merge_regions(struct memblock_type *type) */ static void __init_memblock memblock_insert_region(struct memblock_type *type, int idx, phys_addr_t base, - phys_addr_t size) + phys_addr_t size, int nid) { struct memblock_region *rgn = &type->regions[idx]; @@ -298,6 +297,7 @@ static void __init_memblock memblock_insert_region(struct memblock_type *type, memmove(rgn + 1, rgn, (type->cnt - idx) * sizeof(*rgn)); rgn->base = base; rgn->size = size; + memblock_set_region_node(rgn, nid); type->cnt++; } @@ -327,6 +327,7 @@ static long __init_memblock memblock_add_region(struct memblock_type *type, WARN_ON(type->cnt != 1); type->regions[0].base = base; type->regions[0].size = size; + memblock_set_region_node(&type->regions[0], MAX_NUMNODES); return 0; } repeat: @@ -355,7 +356,7 @@ repeat: nr_new++; if (insert) memblock_insert_region(type, i++, base, - rbase - base); + rbase - base, MAX_NUMNODES); } /* area below @rend is dealt with, forget about it */ base = min(rend, end); @@ -365,7 +366,8 @@ repeat: if (base < end) { nr_new++; if (insert) - memblock_insert_region(type, i, base, end - base); + memblock_insert_region(type, i, base, end - base, + MAX_NUMNODES); } /* @@ -459,6 +461,100 @@ long __init_memblock memblock_reserve(phys_addr_t base, phys_addr_t size) return memblock_add_region(_rgn, base, size); } +#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP +/* + * Common iterator interface used to define for_each_mem_range(). + */ +void __init_memblock __next_mem_pfn_range(int *idx, int nid, + unsigned long *out_start_pfn, + unsigned long *out_end_pfn, int *out_nid) +{ + struct memblock_type *type = &memblock.memory; + struct memblock_region *r; + + while (++*idx < type->cnt) { + r = &type->regions[*idx]; + + if (PFN_UP(r->base) >= PFN_DOWN(r->base + r->size)) + continue; + if (nid == MAX_NUMNODES || nid == r->nid) + break; + } + if (*idx >= type->cnt) { + *idx = -1; + return; + } + + if (out_start_pfn) + *out_start_pfn = PFN_UP(r->base); + if (out_end_pfn) + *out_end_pfn = PFN_DOWN(r->base + r->size); + if (out_nid) + *out_nid = r->nid; +} + +/** + * memblock_set_node - set node ID on memblock regions + * @base: base of area to set node ID for + * @size: size of area to set node ID for + * @nid: node ID to set + * + * Set the nid of memblock memory regions in [@base,@base+@size) to @nid. + * Regions which cross the area boundaries are split as necessary. + * + * RETURNS: + * 0 on success, -errno on failure. + */ +int memblock_set_node(phys_addr_t base, phys_addr_t size, int nid) +{ + struct memblock_type *type = &memblock.memory; + phys_addr_t end = base + size; + int i; + + /* we'll create at most two more regions */ + while (type->cnt + 2 > type->max) + if (memblock_double_array(type) < 0) + return -ENOMEM; + + for (i = 0; i < type->cnt; i++) { + struct memblock_region *rgn = &type->regions[i]; + phys_addr_t rbase = rgn->base; + phys_addr_t rend = rbase + rgn->size; + + if (rbase >= end) + break; + if (rend <= base) + continue; + + if (rbase < base) { + /* + * @rgn intersects from below. Split and continue + * to process the next region - the new top half. + */ + rgn->base = base; + rgn->size = rend - rgn->base; + memblock_insert_region(type, i, rbase, base - rbase, + rgn->nid); + } else if (rend > end) { + /* + * @rgn intersects from above. Split and redo the + * current region - the new bottom half. + */ + rgn->base = end; + rgn->size = rend - rgn->base; + memblock_insert_region(type, i--, rbase, end - rbase, + rgn->nid); + } else { + /* @rgn is fully contained, set ->nid */ + rgn->nid = nid; + } + } + + memblock_merge_regions(type); + return 0; +} +#endif /* CONFIG_HAVE_MEMBLOCK_NODE_MAP */ + phys_addr_t __init __memblock_alloc_base(phys_addr_t size, phys_addr_t align, phys_addr_t max_addr) { phys_addr_t found; @@ -689,19 +785,26 @@ void __init_memblock memblock_set_current_limit(phys_addr_t limit) memblock.current_limit = limit; } -static void __init_memblock memblock_dump(struct memblock_type *region, char *name) +static void __init_memblock memblock_dump(struct memblock_type *type, char *name) { unsigned long long base, size; int i; - pr_info(" %s.cnt = 0x%lx\n", name, region->cnt); - - for (i = 0; i < region->cnt; i++) { - base = region->regions[i].base; - size = region->regions[i].size; + pr_info(" %s.cnt = 0x%lx\n", name, type->cnt); - pr_info(" %s[%#x]\t[%#016llx-%#016llx], %#llx bytes\n", - name, i, base, base + size - 1, size); + for (i = 0; i < type->cnt; i++) { + struct memblock_region *rgn = &type->regions[i]; + char nid_buf[32] = ""; + + base = rgn->base; + size = rgn->size; +#ifdef CONFIG_HAVE_MEMBLOCK_NODE_MAP + if (memblock_get_region_node(rgn) != MAX_NUMNODES) + snprintf(nid_buf, sizeof(nid_buf), " on node %d", + memblock_get_region_node(rgn)); +#endif + pr_info(" %s[%#x]\t[%#016llx-%#016llx], %#llx bytes%s\n", + name, i, base, base + size - 1, size, nid_buf); } } @@ -759,11 +862,13 @@ void __init memblock_init(void) */ memblock.memory.regions[0].base = 0; memblock.memory.regions[0].size = 0; + memblock_set_region_node(&memblock.memory.regions[0], MAX_NUMNODES); memblock.memory.cnt = 1; /* Ditto. */ memblock.reserved.regions[0].base = 0; memblock.reserved.regions[0].size = 0; + memblock_set_region_node(&memblock.reserved.regions[0], MAX_NUMNODES); memblock.reserved.cnt = 1; memblock.current_limit = MEMBLOCK_ALLOC_ANYWHERE; diff --git a/mm/page_alloc.c b/mm/page_alloc.c index 8ab5e5e..3c7ea45 100644 --- a/mm/page_alloc.c +++ b/mm/page_alloc.c @@ -182,28 +182,31 @@ static unsigned long __meminitdata nr_all_pages; static unsigned long __meminitdata dma_reserve; #ifdef CONFIG_ARCH_POPULATES_NODE_MAP - /* - * MAX_ACTIVE_REGIONS determines the maximum number of distinct - * ranges of memory (RAM) that may be registered with add_active_range(). - * Ranges passed to add_active_range() will be merged if possible - * so the number of times add_active_range() can be called is - * related to the number of nodes and the number of holes - */ - #ifdef CONFIG_MAX_ACTIVE_REGIONS - /* Allow an architecture to set MAX_ACTIVE_REGIONS to save memory */ - #define MAX_ACTIVE_REGIONS CONFIG_MAX_ACTIVE_REGIONS - #else - #if MAX_NUMNODES >= 32 - /* If there can be many nodes, allow up to 50 holes per node */ - #define MAX_ACTIVE_REGIONS (MAX_NUMNODES*50) + #ifndef CONFIG_HAVE_MEMBLOCK_NODE_MAP + /* + * MAX_ACTIVE_REGIONS determines the maximum number of distinct ranges + * of memory (RAM) that may be registered with add_active_range(). + * Ranges passed to add_active_range() will be merged if possible so + * the number of times add_active_range() can be called is related to + * the number of nodes and the number of holes + */ + #ifdef CONFIG_MAX_ACTIVE_REGIONS + /* Allow an architecture to set MAX_ACTIVE_REGIONS to save memory */ + #define MAX_ACTIVE_REGIONS CONFIG_MAX_ACTIVE_REGIONS #else - /* By default, allow up to 256 distinct regions */ - #define MAX_ACTIVE_REGIONS 256 + #if MAX_NUMNODES >= 32 + /* If there can be many nodes, allow up to 50 holes per node */ + #define MAX_ACTIVE_REGIONS (MAX_NUMNODES*50) + #else + /* By default, allow up to 256 distinct regions */ + #define MAX_ACTIVE_REGIONS 256 + #endif #endif - #endif - static struct node_active_region __meminitdata early_node_map[MAX_ACTIVE_REGIONS]; - static int __meminitdata nr_nodemap_entries; + static struct node_active_region __meminitdata early_node_map[MAX_ACTIVE_REGIONS]; + static int __meminitdata nr_nodemap_entries; +#endif /* !CONFIG_HAVE_MEMBLOCK_NODE_MAP */ + static unsigned long __meminitdata arch_zone_lowest_possible_pfn[MAX_NR_ZONES]; static unsigned long __meminitdata arch_zone_highest_possible_pfn[MAX_NR_ZONES]; static unsigned long __initdata required_kernelcore; @@ -4268,6 +4271,7 @@ static inline void setup_nr_node_ids(void) } #endif +#ifndef CONFIG_HAVE_MEMBLOCK_NODE_MAP /* * Common iterator interface used to define for_each_mem_pfn_range(). */ @@ -4456,6 +4460,11 @@ void __init sort_node_map(void) sizeof(struct node_active_region), cmp_node_active_region, NULL); } +#else /* !CONFIG_HAVE_MEMBLOCK_NODE_MAP */ +static inline void sort_node_map(void) +{ +} +#endif /** * node_map_pfn_alignment - determine the maximum internode alignment -- 1.7.6