To: Ingo Molnar From 354a59039cc393a6d7266490c8fc05f7f4c9ef25 Mon Sep 17 00:00:00 2001 From: Daniel J Blueman Date: Mon, 5 Oct 2015 11:53:27 +0800 Subject: [PATCH v2] x86/apic: Use smaller array for __apicid_to_node[] mapping The Intel x2APIC spec states the upper 16-bits of APIC ID is the cluster ID [1, p2-12], intended for future distributed systems. Beyond the legacy 8-bit APIC ID, Numascale NumaConnect uses 4-bits for the position of a server on each axis of a multi-dimension torus; SGI NUMAlink also structures the APIC ID space. Instead, define an array based on NR_CPUs to achieve a 1:1 mapping and perform linear search; we see "ACPI: NR_CPUS/possible_cpus limit of X reached. Processor 8/0x16 ignored." when config-limited. This addresses the binary bloat and the present artificial APIC ID limits. With CONFIG_NR_CPUS=256, we save ~64KB of vmlinux data: $ size vmlinux vmlinux-patched text data bss dec hex filename 18232877 1849656 2281472 22364005 1553f65 vmlinux 18233034 1786168 2281472 22300674 1544802 vmlinux-patched Tested on a 256-core system with a 20-bit APIC ID space, and on a 48-core legacy 8-bit APIC ID system with and without CONFIG_NUMA, CONFIG_NUMA_EMU and CONFIG_AMD_NUMA. v2: Improved readability by moving static variable out; integrated Denys's numa emulation fix Signed-off-by: Daniel J Blueman CC: Denys Vlasenko CC: Ingo Molnar CC: Thomas Gleixner CC: Jiang Liu CC: Len Brown CC: Steffen Persvold CC: linux-kernel@vger.kernel.org CC: x86@kernel.org [1] http://www.intel.com/content/dam/doc/specification-update/64-architecture-x2apic-specification.pdf -- 2.5.0 diff --git a/arch/x86/include/asm/numa.h b/arch/x86/include/asm/numa.h index 01b493e..33becb8 100644 --- a/arch/x86/include/asm/numa.h +++ b/arch/x86/include/asm/numa.h @@ -17,6 +17,11 @@ */ #define NODE_MIN_SIZE (4*1024*1024) +struct apicid_to_node { + int apicid; + s16 node; +}; + extern int numa_off; /* @@ -27,17 +32,13 @@ extern int numa_off; * should be accessed by the accessors - set_apicid_to_node() and * numa_cpu_node(). */ -extern s16 __apicid_to_node[MAX_LOCAL_APIC]; +extern struct apicid_to_node __apicid_to_node[NR_CPUS]; extern nodemask_t numa_nodes_parsed __initdata; extern int __init numa_add_memblk(int nodeid, u64 start, u64 end); extern void __init numa_set_distance(int from, int to, int distance); -static inline void set_apicid_to_node(int apicid, s16 node) -{ - __apicid_to_node[apicid] = node; -} - +extern void set_apicid_to_node(int apicid, s16 node); extern int numa_cpu_node(int cpu); #else /* CONFIG_NUMA */ diff --git a/arch/x86/kernel/cpu/amd.c b/arch/x86/kernel/cpu/amd.c index a07956a..45a2dff 100644 --- a/arch/x86/kernel/cpu/amd.c +++ b/arch/x86/kernel/cpu/amd.c @@ -277,12 +277,13 @@ static int nearby_node(int apicid) int i, node; for (i = apicid - 1; i >= 0; i--) { - node = __apicid_to_node[i]; + node = __apicid_to_node[i].node; if (node != NUMA_NO_NODE && node_online(node)) return node; } - for (i = apicid + 1; i < MAX_LOCAL_APIC; i++) { - node = __apicid_to_node[i]; + for (i = apicid + 1; i < NR_CPUS; i++) { + node = __apicid_to_node[i].node; + if (node != NUMA_NO_NODE && node_online(node)) return node; } @@ -434,8 +435,8 @@ static void srat_detect_node(struct cpuinfo_x86 *c) */ int ht_nodeid = c->initial_apicid; - if (__apicid_to_node[ht_nodeid] != NUMA_NO_NODE) - node = __apicid_to_node[ht_nodeid]; + if (__apicid_to_node[ht_nodeid].node != NUMA_NO_NODE) + node = __apicid_to_node[ht_nodeid].node; /* Pick a nearby node */ if (!node_online(node)) node = nearby_node(apicid); diff --git a/arch/x86/mm/numa.c b/arch/x86/mm/numa.c index d04f809..7f511b7 100644 --- a/arch/x86/mm/numa.c +++ b/arch/x86/mm/numa.c @@ -26,6 +26,7 @@ nodemask_t numa_nodes_parsed __initdata; struct pglist_data *node_data[MAX_NUMNODES] __read_mostly; EXPORT_SYMBOL(node_data); +static unsigned apicids; static struct numa_meminfo numa_meminfo #ifndef CONFIG_MEMORY_HOTPLUG __initdata @@ -56,16 +57,31 @@ early_param("numa", numa_setup); /* * apicid, cpu, node mappings */ -s16 __apicid_to_node[MAX_LOCAL_APIC] = { - [0 ... MAX_LOCAL_APIC-1] = NUMA_NO_NODE +struct apicid_to_node __apicid_to_node[NR_CPUS] = { + [0 ... NR_CPUS-1] = {-1, NUMA_NO_NODE} }; +void set_apicid_to_node(int apicid, s16 node) +{ + /* Protect against small kernel on large system */ + if (apicids >= NR_CPUS) + return; + + __apicid_to_node[apicids].apicid = apicid; + __apicid_to_node[apicids].node = node; + apicids++; +} + int numa_cpu_node(int cpu) { - int apicid = early_per_cpu(x86_cpu_to_apicid, cpu); + int ent, apicid = early_per_cpu(x86_cpu_to_apicid, cpu); + if (apicid == BAD_APICID) + return NUMA_NO_NODE; + + for (ent = 0; ent < NR_CPUS; ent++) + if (__apicid_to_node[ent].apicid == apicid) + return __apicid_to_node[ent].node; - if (apicid != BAD_APICID) - return __apicid_to_node[apicid]; return NUMA_NO_NODE; } @@ -607,9 +623,6 @@ static int __init numa_init(int (*init_func)(void)) int i; int ret; - for (i = 0; i < MAX_LOCAL_APIC; i++) - set_apicid_to_node(i, NUMA_NO_NODE); - nodes_clear(numa_nodes_parsed); nodes_clear(node_possible_map); nodes_clear(node_online_map); diff --git a/arch/x86/mm/numa_emulation.c b/arch/x86/mm/numa_emulation.c index a8f90ce..1a0e112 100644 --- a/arch/x86/mm/numa_emulation.c +++ b/arch/x86/mm/numa_emulation.c @@ -399,12 +399,12 @@ void __init numa_emulation(struct numa_meminfo *numa_meminfo, int numa_dist_cnt) * back to zero just in case. */ for (i = 0; i < ARRAY_SIZE(__apicid_to_node); i++) { - if (__apicid_to_node[i] == NUMA_NO_NODE) + if (__apicid_to_node[i].node == 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]) + if (__apicid_to_node[i].node == emu_nid_to_phys[j]) break; - __apicid_to_node[i] = j < ARRAY_SIZE(emu_nid_to_phys) ? j : 0; + __apicid_to_node[i].node = j < ARRAY_SIZE(emu_nid_to_phys) ? j : 0; } /* make sure all emulated nodes are mapped to a physical node */