linux/arch/x86/mm/srat_64.c

353 lines
8.6 KiB
C
Raw Normal View History

/*
* ACPI 3.0 based NUMA setup
* Copyright 2004 Andi Kleen, SuSE Labs.
*
* Reads the ACPI SRAT table to figure out what memory belongs to which CPUs.
*
* Called from acpi_numa_init while reading the SRAT and SLIT tables.
* Assumes all memory regions belonging to a single proximity domain
* are in one chunk. Holes between them will be included in the node.
*/
#include <linux/kernel.h>
#include <linux/acpi.h>
#include <linux/mmzone.h>
#include <linux/bitmap.h>
#include <linux/module.h>
#include <linux/topology.h>
[PATCH] x86_64: Reserve SRAT hotadd memory on x86-64 From: Keith Mannthey, Andi Kleen Implement memory hotadd without sparsemem. The memory in the SRAT hotadd area is just preserved instead and can be activated later. There are a few restrictions: - Only one continuous hotadd area allowed per node The main problem is dealing with the many buggy SRAT tables that are out there. The strategy here is to reject anything suspicious. Originally from Keith Mannthey, with several hacks and changes by AK and also contributions from Andrew Morton [ TBD: Problems pointed out by KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>: 1) Goto's rebuild_zonelist patch will not work if CONFIG_MEMORY_HOTPLUG=n. Rebuilding zonelist is necessary when the system has just memory < 4G at boot, and hot add memory > 4G. because x86_64 has DMA32, ZONE_NORAML is not included into zonelist at boot time if system doesn't have memory >4G at boot. [AK: should just force the higher zones at boot time when SRAT tells us] 2) zone and node's spanned_pages and present_pages are not incremented. They should be. For example, our server (ia64/Fujitsu PrimeQuest) can equip memory from 4G to 1T(maybe 2T in future), and SRAT will *always* say we have possible 1T +memory. (Microsoft requires "write all possible memory in SRAT") When we reserve memmap for possible 1T memory, Linux will not work well in +minimum 4G configuraion ;) [AK: needs limiting to 5-10% of max memory] ] Signed-off-by: Andi Kleen <ak@suse.de> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-04-07 17:49:18 +00:00
#include <linux/bootmem.h>
#include <linux/memblock.h>
[PATCH] x86_64: Reserve SRAT hotadd memory on x86-64 From: Keith Mannthey, Andi Kleen Implement memory hotadd without sparsemem. The memory in the SRAT hotadd area is just preserved instead and can be activated later. There are a few restrictions: - Only one continuous hotadd area allowed per node The main problem is dealing with the many buggy SRAT tables that are out there. The strategy here is to reject anything suspicious. Originally from Keith Mannthey, with several hacks and changes by AK and also contributions from Andrew Morton [ TBD: Problems pointed out by KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>: 1) Goto's rebuild_zonelist patch will not work if CONFIG_MEMORY_HOTPLUG=n. Rebuilding zonelist is necessary when the system has just memory < 4G at boot, and hot add memory > 4G. because x86_64 has DMA32, ZONE_NORAML is not included into zonelist at boot time if system doesn't have memory >4G at boot. [AK: should just force the higher zones at boot time when SRAT tells us] 2) zone and node's spanned_pages and present_pages are not incremented. They should be. For example, our server (ia64/Fujitsu PrimeQuest) can equip memory from 4G to 1T(maybe 2T in future), and SRAT will *always* say we have possible 1T +memory. (Microsoft requires "write all possible memory in SRAT") When we reserve memmap for possible 1T memory, Linux will not work well in +minimum 4G configuraion ;) [AK: needs limiting to 5-10% of max memory] ] Signed-off-by: Andi Kleen <ak@suse.de> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-04-07 17:49:18 +00:00
#include <linux/mm.h>
#include <asm/proto.h>
#include <asm/numa.h>
#include <asm/e820.h>
#include <asm/apic.h>
#include <asm/uv/uv.h>
int acpi_numa __initdata;
static struct acpi_table_slit *acpi_slit;
static struct bootnode nodes_add[MAX_NUMNODES];
static __init int setup_node(int pxm)
{
return acpi_map_pxm_to_node(pxm);
}
static __init void bad_srat(void)
{
printk(KERN_ERR "SRAT: SRAT not used.\n");
acpi_numa = -1;
memset(nodes_add, 0, sizeof(nodes_add));
}
static __init inline int srat_disabled(void)
{
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");
acpi_slit = __va(phys);
memcpy(acpi_slit, slit, length);
memblock_x86_reserve_range(phys, phys + length, "ACPI SLIT");
}
/* Callback for Proximity Domain -> x2APIC mapping */
void __init
acpi_numa_x2apic_affinity_init(struct acpi_srat_x2apic_cpu_affinity *pa)
{
int pxm, node;
int apic_id;
if (srat_disabled())
return;
if (pa->header.length < sizeof(struct acpi_srat_x2apic_cpu_affinity)) {
bad_srat();
return;
}
if ((pa->flags & ACPI_SRAT_CPU_ENABLED) == 0)
return;
pxm = pa->proximity_domain;
node = setup_node(pxm);
if (node < 0) {
printk(KERN_ERR "SRAT: Too many proximity domains %x\n", pxm);
bad_srat();
return;
}
apic_id = pa->apic_id;
x86, acpi: Parse all SRAT cpu entries even above the cpu number limitation Recent Intel new system have different order in MADT, aka will list all thread0 at first, then all thread1. But SRAT table still old order, it will list cpus in one socket all together. If the user have compiled limited NR_CPUS or boot with nr_cpus=, could have missed to put some cpus apic id to node mapping into apicid_to_node[]. for example for 4 sockets system with 64 cpus with nr_cpus=32 will get crash... [ 9.106288] Total of 32 processors activated (136190.88 BogoMIPS). [ 9.235021] divide error: 0000 [#1] SMP [ 9.235315] last sysfs file: [ 9.235481] CPU 1 [ 9.235592] Modules linked in: [ 9.245398] [ 9.245478] Pid: 2, comm: kthreadd Not tainted 2.6.37-rc1-tip-yh-01782-ge92ef79-dirty #274 /Sun Fire x4800 [ 9.265415] RIP: 0010:[<ffffffff81075a8f>] [<ffffffff81075a8f>] select_task_rq_fair+0x4f0/0x623 ... [ 9.645938] RIP [<ffffffff81075a8f>] select_task_rq_fair+0x4f0/0x623 [ 9.665356] RSP <ffff88103f8d1c40> [ 9.665568] ---[ end trace 2296156d35fdfc87 ]--- So let just parse all cpu entries in SRAT. Also add apicid checking with MAX_LOCAL_APIC, in case We could out of boundaries of apicid_to_node[]. it fixes following bug too. https://bugzilla.kernel.org/show_bug.cgi?id=22662 -v2: expand to 32bit according to hpa need to add MAX_LOCAL_APIC for 32bit Reported-and-Tested-by: Wu Fengguang <fengguang.wu@intel.com> Reported-by: Bjorn Helgaas <bjorn.helgaas@hp.com> Tested-by: Myron Stowe <myron.stowe@hp.com> Signed-off-by: Yinghai Lu <yinghai@kernel.org> LKML-Reference: <4D0AD486.9020704@kernel.org> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
2010-12-17 03:09:58 +00:00
if (apic_id >= MAX_LOCAL_APIC) {
printk(KERN_INFO "SRAT: PXM %u -> APIC 0x%04x -> Node %u skipped apicid that is too big\n", pxm, apic_id, node);
return;
}
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);
}
/* Callback for Proximity Domain -> LAPIC mapping */
void __init
acpi_numa_processor_affinity_init(struct acpi_srat_cpu_affinity *pa)
{
int pxm, node;
int apic_id;
if (srat_disabled())
return;
if (pa->header.length != sizeof(struct acpi_srat_cpu_affinity)) {
bad_srat();
return;
}
if ((pa->flags & ACPI_SRAT_CPU_ENABLED) == 0)
return;
pxm = pa->proximity_domain_lo;
node = setup_node(pxm);
if (node < 0) {
printk(KERN_ERR "SRAT: Too many proximity domains %x\n", pxm);
bad_srat();
return;
}
if (get_uv_system_type() >= UV_X2APIC)
apic_id = (pa->apic_id << 8) | pa->local_sapic_eid;
else
apic_id = pa->apic_id;
x86, acpi: Parse all SRAT cpu entries even above the cpu number limitation Recent Intel new system have different order in MADT, aka will list all thread0 at first, then all thread1. But SRAT table still old order, it will list cpus in one socket all together. If the user have compiled limited NR_CPUS or boot with nr_cpus=, could have missed to put some cpus apic id to node mapping into apicid_to_node[]. for example for 4 sockets system with 64 cpus with nr_cpus=32 will get crash... [ 9.106288] Total of 32 processors activated (136190.88 BogoMIPS). [ 9.235021] divide error: 0000 [#1] SMP [ 9.235315] last sysfs file: [ 9.235481] CPU 1 [ 9.235592] Modules linked in: [ 9.245398] [ 9.245478] Pid: 2, comm: kthreadd Not tainted 2.6.37-rc1-tip-yh-01782-ge92ef79-dirty #274 /Sun Fire x4800 [ 9.265415] RIP: 0010:[<ffffffff81075a8f>] [<ffffffff81075a8f>] select_task_rq_fair+0x4f0/0x623 ... [ 9.645938] RIP [<ffffffff81075a8f>] select_task_rq_fair+0x4f0/0x623 [ 9.665356] RSP <ffff88103f8d1c40> [ 9.665568] ---[ end trace 2296156d35fdfc87 ]--- So let just parse all cpu entries in SRAT. Also add apicid checking with MAX_LOCAL_APIC, in case We could out of boundaries of apicid_to_node[]. it fixes following bug too. https://bugzilla.kernel.org/show_bug.cgi?id=22662 -v2: expand to 32bit according to hpa need to add MAX_LOCAL_APIC for 32bit Reported-and-Tested-by: Wu Fengguang <fengguang.wu@intel.com> Reported-by: Bjorn Helgaas <bjorn.helgaas@hp.com> Tested-by: Myron Stowe <myron.stowe@hp.com> Signed-off-by: Yinghai Lu <yinghai@kernel.org> LKML-Reference: <4D0AD486.9020704@kernel.org> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
2010-12-17 03:09:58 +00:00
if (apic_id >= MAX_LOCAL_APIC) {
printk(KERN_INFO "SRAT: PXM %u -> APIC 0x%02x -> Node %u skipped apicid that is too big\n", pxm, apic_id, node);
return;
}
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);
}
#ifdef CONFIG_MEMORY_HOTPLUG_SPARSE
static inline int save_add_info(void) {return 1;}
#else
static inline int save_add_info(void) {return 0;}
#endif
[PATCH] x86_64: Reserve SRAT hotadd memory on x86-64 From: Keith Mannthey, Andi Kleen Implement memory hotadd without sparsemem. The memory in the SRAT hotadd area is just preserved instead and can be activated later. There are a few restrictions: - Only one continuous hotadd area allowed per node The main problem is dealing with the many buggy SRAT tables that are out there. The strategy here is to reject anything suspicious. Originally from Keith Mannthey, with several hacks and changes by AK and also contributions from Andrew Morton [ TBD: Problems pointed out by KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>: 1) Goto's rebuild_zonelist patch will not work if CONFIG_MEMORY_HOTPLUG=n. Rebuilding zonelist is necessary when the system has just memory < 4G at boot, and hot add memory > 4G. because x86_64 has DMA32, ZONE_NORAML is not included into zonelist at boot time if system doesn't have memory >4G at boot. [AK: should just force the higher zones at boot time when SRAT tells us] 2) zone and node's spanned_pages and present_pages are not incremented. They should be. For example, our server (ia64/Fujitsu PrimeQuest) can equip memory from 4G to 1T(maybe 2T in future), and SRAT will *always* say we have possible 1T +memory. (Microsoft requires "write all possible memory in SRAT") When we reserve memmap for possible 1T memory, Linux will not work well in +minimum 4G configuraion ;) [AK: needs limiting to 5-10% of max memory] ] Signed-off-by: Andi Kleen <ak@suse.de> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-04-07 17:49:18 +00:00
/*
* Update nodes_add[]
* This code supports one contiguous hot add area per node
[PATCH] x86_64: Reserve SRAT hotadd memory on x86-64 From: Keith Mannthey, Andi Kleen Implement memory hotadd without sparsemem. The memory in the SRAT hotadd area is just preserved instead and can be activated later. There are a few restrictions: - Only one continuous hotadd area allowed per node The main problem is dealing with the many buggy SRAT tables that are out there. The strategy here is to reject anything suspicious. Originally from Keith Mannthey, with several hacks and changes by AK and also contributions from Andrew Morton [ TBD: Problems pointed out by KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>: 1) Goto's rebuild_zonelist patch will not work if CONFIG_MEMORY_HOTPLUG=n. Rebuilding zonelist is necessary when the system has just memory < 4G at boot, and hot add memory > 4G. because x86_64 has DMA32, ZONE_NORAML is not included into zonelist at boot time if system doesn't have memory >4G at boot. [AK: should just force the higher zones at boot time when SRAT tells us] 2) zone and node's spanned_pages and present_pages are not incremented. They should be. For example, our server (ia64/Fujitsu PrimeQuest) can equip memory from 4G to 1T(maybe 2T in future), and SRAT will *always* say we have possible 1T +memory. (Microsoft requires "write all possible memory in SRAT") When we reserve memmap for possible 1T memory, Linux will not work well in +minimum 4G configuraion ;) [AK: needs limiting to 5-10% of max memory] ] Signed-off-by: Andi Kleen <ak@suse.de> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-04-07 17:49:18 +00:00
*/
static void __init
update_nodes_add(int node, unsigned long start, unsigned long end)
[PATCH] x86_64: Reserve SRAT hotadd memory on x86-64 From: Keith Mannthey, Andi Kleen Implement memory hotadd without sparsemem. The memory in the SRAT hotadd area is just preserved instead and can be activated later. There are a few restrictions: - Only one continuous hotadd area allowed per node The main problem is dealing with the many buggy SRAT tables that are out there. The strategy here is to reject anything suspicious. Originally from Keith Mannthey, with several hacks and changes by AK and also contributions from Andrew Morton [ TBD: Problems pointed out by KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>: 1) Goto's rebuild_zonelist patch will not work if CONFIG_MEMORY_HOTPLUG=n. Rebuilding zonelist is necessary when the system has just memory < 4G at boot, and hot add memory > 4G. because x86_64 has DMA32, ZONE_NORAML is not included into zonelist at boot time if system doesn't have memory >4G at boot. [AK: should just force the higher zones at boot time when SRAT tells us] 2) zone and node's spanned_pages and present_pages are not incremented. They should be. For example, our server (ia64/Fujitsu PrimeQuest) can equip memory from 4G to 1T(maybe 2T in future), and SRAT will *always* say we have possible 1T +memory. (Microsoft requires "write all possible memory in SRAT") When we reserve memmap for possible 1T memory, Linux will not work well in +minimum 4G configuraion ;) [AK: needs limiting to 5-10% of max memory] ] Signed-off-by: Andi Kleen <ak@suse.de> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-04-07 17:49:18 +00:00
{
unsigned long s_pfn = start >> PAGE_SHIFT;
unsigned long e_pfn = end >> PAGE_SHIFT;
int changed = 0;
[PATCH] x86_64: Reserve SRAT hotadd memory on x86-64 From: Keith Mannthey, Andi Kleen Implement memory hotadd without sparsemem. The memory in the SRAT hotadd area is just preserved instead and can be activated later. There are a few restrictions: - Only one continuous hotadd area allowed per node The main problem is dealing with the many buggy SRAT tables that are out there. The strategy here is to reject anything suspicious. Originally from Keith Mannthey, with several hacks and changes by AK and also contributions from Andrew Morton [ TBD: Problems pointed out by KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>: 1) Goto's rebuild_zonelist patch will not work if CONFIG_MEMORY_HOTPLUG=n. Rebuilding zonelist is necessary when the system has just memory < 4G at boot, and hot add memory > 4G. because x86_64 has DMA32, ZONE_NORAML is not included into zonelist at boot time if system doesn't have memory >4G at boot. [AK: should just force the higher zones at boot time when SRAT tells us] 2) zone and node's spanned_pages and present_pages are not incremented. They should be. For example, our server (ia64/Fujitsu PrimeQuest) can equip memory from 4G to 1T(maybe 2T in future), and SRAT will *always* say we have possible 1T +memory. (Microsoft requires "write all possible memory in SRAT") When we reserve memmap for possible 1T memory, Linux will not work well in +minimum 4G configuraion ;) [AK: needs limiting to 5-10% of max memory] ] Signed-off-by: Andi Kleen <ak@suse.de> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-04-07 17:49:18 +00:00
struct bootnode *nd = &nodes_add[node];
/* I had some trouble with strange memory hotadd regions breaking
the boot. Be very strict here and reject anything unexpected.
If you want working memory hotadd write correct SRATs.
The node size check is a basic sanity check to guard against
mistakes */
if ((signed long)(end - start) < NODE_MIN_SIZE) {
printk(KERN_ERR "SRAT: Hotplug area too small\n");
return;
[PATCH] x86_64: Reserve SRAT hotadd memory on x86-64 From: Keith Mannthey, Andi Kleen Implement memory hotadd without sparsemem. The memory in the SRAT hotadd area is just preserved instead and can be activated later. There are a few restrictions: - Only one continuous hotadd area allowed per node The main problem is dealing with the many buggy SRAT tables that are out there. The strategy here is to reject anything suspicious. Originally from Keith Mannthey, with several hacks and changes by AK and also contributions from Andrew Morton [ TBD: Problems pointed out by KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>: 1) Goto's rebuild_zonelist patch will not work if CONFIG_MEMORY_HOTPLUG=n. Rebuilding zonelist is necessary when the system has just memory < 4G at boot, and hot add memory > 4G. because x86_64 has DMA32, ZONE_NORAML is not included into zonelist at boot time if system doesn't have memory >4G at boot. [AK: should just force the higher zones at boot time when SRAT tells us] 2) zone and node's spanned_pages and present_pages are not incremented. They should be. For example, our server (ia64/Fujitsu PrimeQuest) can equip memory from 4G to 1T(maybe 2T in future), and SRAT will *always* say we have possible 1T +memory. (Microsoft requires "write all possible memory in SRAT") When we reserve memmap for possible 1T memory, Linux will not work well in +minimum 4G configuraion ;) [AK: needs limiting to 5-10% of max memory] ] Signed-off-by: Andi Kleen <ak@suse.de> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-04-07 17:49:18 +00:00
}
/* This check might be a bit too strict, but I'm keeping it for now. */
if (absent_pages_in_range(s_pfn, e_pfn) != e_pfn - s_pfn) {
printk(KERN_ERR
"SRAT: Hotplug area %lu -> %lu has existing memory\n",
s_pfn, e_pfn);
return;
[PATCH] x86_64: Reserve SRAT hotadd memory on x86-64 From: Keith Mannthey, Andi Kleen Implement memory hotadd without sparsemem. The memory in the SRAT hotadd area is just preserved instead and can be activated later. There are a few restrictions: - Only one continuous hotadd area allowed per node The main problem is dealing with the many buggy SRAT tables that are out there. The strategy here is to reject anything suspicious. Originally from Keith Mannthey, with several hacks and changes by AK and also contributions from Andrew Morton [ TBD: Problems pointed out by KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>: 1) Goto's rebuild_zonelist patch will not work if CONFIG_MEMORY_HOTPLUG=n. Rebuilding zonelist is necessary when the system has just memory < 4G at boot, and hot add memory > 4G. because x86_64 has DMA32, ZONE_NORAML is not included into zonelist at boot time if system doesn't have memory >4G at boot. [AK: should just force the higher zones at boot time when SRAT tells us] 2) zone and node's spanned_pages and present_pages are not incremented. They should be. For example, our server (ia64/Fujitsu PrimeQuest) can equip memory from 4G to 1T(maybe 2T in future), and SRAT will *always* say we have possible 1T +memory. (Microsoft requires "write all possible memory in SRAT") When we reserve memmap for possible 1T memory, Linux will not work well in +minimum 4G configuraion ;) [AK: needs limiting to 5-10% of max memory] ] Signed-off-by: Andi Kleen <ak@suse.de> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-04-07 17:49:18 +00:00
}
/* Looks good */
if (nd->start == nd->end) {
nd->start = start;
nd->end = end;
[PATCH] x86_64: Reserve SRAT hotadd memory on x86-64 From: Keith Mannthey, Andi Kleen Implement memory hotadd without sparsemem. The memory in the SRAT hotadd area is just preserved instead and can be activated later. There are a few restrictions: - Only one continuous hotadd area allowed per node The main problem is dealing with the many buggy SRAT tables that are out there. The strategy here is to reject anything suspicious. Originally from Keith Mannthey, with several hacks and changes by AK and also contributions from Andrew Morton [ TBD: Problems pointed out by KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>: 1) Goto's rebuild_zonelist patch will not work if CONFIG_MEMORY_HOTPLUG=n. Rebuilding zonelist is necessary when the system has just memory < 4G at boot, and hot add memory > 4G. because x86_64 has DMA32, ZONE_NORAML is not included into zonelist at boot time if system doesn't have memory >4G at boot. [AK: should just force the higher zones at boot time when SRAT tells us] 2) zone and node's spanned_pages and present_pages are not incremented. They should be. For example, our server (ia64/Fujitsu PrimeQuest) can equip memory from 4G to 1T(maybe 2T in future), and SRAT will *always* say we have possible 1T +memory. (Microsoft requires "write all possible memory in SRAT") When we reserve memmap for possible 1T memory, Linux will not work well in +minimum 4G configuraion ;) [AK: needs limiting to 5-10% of max memory] ] Signed-off-by: Andi Kleen <ak@suse.de> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-04-07 17:49:18 +00:00
changed = 1;
} else {
if (nd->start == end) {
nd->start = start;
[PATCH] x86_64: Reserve SRAT hotadd memory on x86-64 From: Keith Mannthey, Andi Kleen Implement memory hotadd without sparsemem. The memory in the SRAT hotadd area is just preserved instead and can be activated later. There are a few restrictions: - Only one continuous hotadd area allowed per node The main problem is dealing with the many buggy SRAT tables that are out there. The strategy here is to reject anything suspicious. Originally from Keith Mannthey, with several hacks and changes by AK and also contributions from Andrew Morton [ TBD: Problems pointed out by KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>: 1) Goto's rebuild_zonelist patch will not work if CONFIG_MEMORY_HOTPLUG=n. Rebuilding zonelist is necessary when the system has just memory < 4G at boot, and hot add memory > 4G. because x86_64 has DMA32, ZONE_NORAML is not included into zonelist at boot time if system doesn't have memory >4G at boot. [AK: should just force the higher zones at boot time when SRAT tells us] 2) zone and node's spanned_pages and present_pages are not incremented. They should be. For example, our server (ia64/Fujitsu PrimeQuest) can equip memory from 4G to 1T(maybe 2T in future), and SRAT will *always* say we have possible 1T +memory. (Microsoft requires "write all possible memory in SRAT") When we reserve memmap for possible 1T memory, Linux will not work well in +minimum 4G configuraion ;) [AK: needs limiting to 5-10% of max memory] ] Signed-off-by: Andi Kleen <ak@suse.de> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-04-07 17:49:18 +00:00
changed = 1;
}
if (nd->end == start) {
nd->end = end;
[PATCH] x86_64: Reserve SRAT hotadd memory on x86-64 From: Keith Mannthey, Andi Kleen Implement memory hotadd without sparsemem. The memory in the SRAT hotadd area is just preserved instead and can be activated later. There are a few restrictions: - Only one continuous hotadd area allowed per node The main problem is dealing with the many buggy SRAT tables that are out there. The strategy here is to reject anything suspicious. Originally from Keith Mannthey, with several hacks and changes by AK and also contributions from Andrew Morton [ TBD: Problems pointed out by KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>: 1) Goto's rebuild_zonelist patch will not work if CONFIG_MEMORY_HOTPLUG=n. Rebuilding zonelist is necessary when the system has just memory < 4G at boot, and hot add memory > 4G. because x86_64 has DMA32, ZONE_NORAML is not included into zonelist at boot time if system doesn't have memory >4G at boot. [AK: should just force the higher zones at boot time when SRAT tells us] 2) zone and node's spanned_pages and present_pages are not incremented. They should be. For example, our server (ia64/Fujitsu PrimeQuest) can equip memory from 4G to 1T(maybe 2T in future), and SRAT will *always* say we have possible 1T +memory. (Microsoft requires "write all possible memory in SRAT") When we reserve memmap for possible 1T memory, Linux will not work well in +minimum 4G configuraion ;) [AK: needs limiting to 5-10% of max memory] ] Signed-off-by: Andi Kleen <ak@suse.de> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-04-07 17:49:18 +00:00
changed = 1;
}
if (!changed)
printk(KERN_ERR "SRAT: Hotplug zone not continuous. Partly ignored\n");
}
[PATCH] x86_64: Reserve SRAT hotadd memory on x86-64 From: Keith Mannthey, Andi Kleen Implement memory hotadd without sparsemem. The memory in the SRAT hotadd area is just preserved instead and can be activated later. There are a few restrictions: - Only one continuous hotadd area allowed per node The main problem is dealing with the many buggy SRAT tables that are out there. The strategy here is to reject anything suspicious. Originally from Keith Mannthey, with several hacks and changes by AK and also contributions from Andrew Morton [ TBD: Problems pointed out by KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>: 1) Goto's rebuild_zonelist patch will not work if CONFIG_MEMORY_HOTPLUG=n. Rebuilding zonelist is necessary when the system has just memory < 4G at boot, and hot add memory > 4G. because x86_64 has DMA32, ZONE_NORAML is not included into zonelist at boot time if system doesn't have memory >4G at boot. [AK: should just force the higher zones at boot time when SRAT tells us] 2) zone and node's spanned_pages and present_pages are not incremented. They should be. For example, our server (ia64/Fujitsu PrimeQuest) can equip memory from 4G to 1T(maybe 2T in future), and SRAT will *always* say we have possible 1T +memory. (Microsoft requires "write all possible memory in SRAT") When we reserve memmap for possible 1T memory, Linux will not work well in +minimum 4G configuraion ;) [AK: needs limiting to 5-10% of max memory] ] Signed-off-by: Andi Kleen <ak@suse.de> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-04-07 17:49:18 +00:00
if (changed) {
node_set(node, numa_nodes_parsed);
printk(KERN_INFO "SRAT: hot plug zone found %Lx - %Lx\n",
nd->start, nd->end);
}
[PATCH] x86_64: Reserve SRAT hotadd memory on x86-64 From: Keith Mannthey, Andi Kleen Implement memory hotadd without sparsemem. The memory in the SRAT hotadd area is just preserved instead and can be activated later. There are a few restrictions: - Only one continuous hotadd area allowed per node The main problem is dealing with the many buggy SRAT tables that are out there. The strategy here is to reject anything suspicious. Originally from Keith Mannthey, with several hacks and changes by AK and also contributions from Andrew Morton [ TBD: Problems pointed out by KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>: 1) Goto's rebuild_zonelist patch will not work if CONFIG_MEMORY_HOTPLUG=n. Rebuilding zonelist is necessary when the system has just memory < 4G at boot, and hot add memory > 4G. because x86_64 has DMA32, ZONE_NORAML is not included into zonelist at boot time if system doesn't have memory >4G at boot. [AK: should just force the higher zones at boot time when SRAT tells us] 2) zone and node's spanned_pages and present_pages are not incremented. They should be. For example, our server (ia64/Fujitsu PrimeQuest) can equip memory from 4G to 1T(maybe 2T in future), and SRAT will *always* say we have possible 1T +memory. (Microsoft requires "write all possible memory in SRAT") When we reserve memmap for possible 1T memory, Linux will not work well in +minimum 4G configuraion ;) [AK: needs limiting to 5-10% of max memory] ] Signed-off-by: Andi Kleen <ak@suse.de> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-04-07 17:49:18 +00:00
}
/* Callback for parsing of the Proximity Domain <-> Memory Area mappings */
void __init
acpi_numa_memory_affinity_init(struct acpi_srat_mem_affinity *ma)
{
unsigned long start, end;
int node, pxm;
if (srat_disabled())
return;
if (ma->header.length != sizeof(struct acpi_srat_mem_affinity)) {
bad_srat();
return;
}
if ((ma->flags & ACPI_SRAT_MEM_ENABLED) == 0)
return;
if ((ma->flags & ACPI_SRAT_MEM_HOT_PLUGGABLE) && !save_add_info())
[PATCH] x86_64: Reserve SRAT hotadd memory on x86-64 From: Keith Mannthey, Andi Kleen Implement memory hotadd without sparsemem. The memory in the SRAT hotadd area is just preserved instead and can be activated later. There are a few restrictions: - Only one continuous hotadd area allowed per node The main problem is dealing with the many buggy SRAT tables that are out there. The strategy here is to reject anything suspicious. Originally from Keith Mannthey, with several hacks and changes by AK and also contributions from Andrew Morton [ TBD: Problems pointed out by KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>: 1) Goto's rebuild_zonelist patch will not work if CONFIG_MEMORY_HOTPLUG=n. Rebuilding zonelist is necessary when the system has just memory < 4G at boot, and hot add memory > 4G. because x86_64 has DMA32, ZONE_NORAML is not included into zonelist at boot time if system doesn't have memory >4G at boot. [AK: should just force the higher zones at boot time when SRAT tells us] 2) zone and node's spanned_pages and present_pages are not incremented. They should be. For example, our server (ia64/Fujitsu PrimeQuest) can equip memory from 4G to 1T(maybe 2T in future), and SRAT will *always* say we have possible 1T +memory. (Microsoft requires "write all possible memory in SRAT") When we reserve memmap for possible 1T memory, Linux will not work well in +minimum 4G configuraion ;) [AK: needs limiting to 5-10% of max memory] ] Signed-off-by: Andi Kleen <ak@suse.de> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-04-07 17:49:18 +00:00
return;
start = ma->base_address;
end = start + ma->length;
pxm = ma->proximity_domain;
node = setup_node(pxm);
if (node < 0) {
printk(KERN_ERR "SRAT: Too many proximity domains.\n");
bad_srat();
return;
}
if (numa_add_memblk(node, start, end) < 0) {
bad_srat();
return;
}
[PATCH] x86_64: Reserve SRAT hotadd memory on x86-64 From: Keith Mannthey, Andi Kleen Implement memory hotadd without sparsemem. The memory in the SRAT hotadd area is just preserved instead and can be activated later. There are a few restrictions: - Only one continuous hotadd area allowed per node The main problem is dealing with the many buggy SRAT tables that are out there. The strategy here is to reject anything suspicious. Originally from Keith Mannthey, with several hacks and changes by AK and also contributions from Andrew Morton [ TBD: Problems pointed out by KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>: 1) Goto's rebuild_zonelist patch will not work if CONFIG_MEMORY_HOTPLUG=n. Rebuilding zonelist is necessary when the system has just memory < 4G at boot, and hot add memory > 4G. because x86_64 has DMA32, ZONE_NORAML is not included into zonelist at boot time if system doesn't have memory >4G at boot. [AK: should just force the higher zones at boot time when SRAT tells us] 2) zone and node's spanned_pages and present_pages are not incremented. They should be. For example, our server (ia64/Fujitsu PrimeQuest) can equip memory from 4G to 1T(maybe 2T in future), and SRAT will *always* say we have possible 1T +memory. (Microsoft requires "write all possible memory in SRAT") When we reserve memmap for possible 1T memory, Linux will not work well in +minimum 4G configuraion ;) [AK: needs limiting to 5-10% of max memory] ] Signed-off-by: Andi Kleen <ak@suse.de> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-04-07 17:49:18 +00:00
printk(KERN_INFO "SRAT: Node %u PXM %u %lx-%lx\n", node, pxm,
start, end);
[PATCH] x86_64: Reserve SRAT hotadd memory on x86-64 From: Keith Mannthey, Andi Kleen Implement memory hotadd without sparsemem. The memory in the SRAT hotadd area is just preserved instead and can be activated later. There are a few restrictions: - Only one continuous hotadd area allowed per node The main problem is dealing with the many buggy SRAT tables that are out there. The strategy here is to reject anything suspicious. Originally from Keith Mannthey, with several hacks and changes by AK and also contributions from Andrew Morton [ TBD: Problems pointed out by KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>: 1) Goto's rebuild_zonelist patch will not work if CONFIG_MEMORY_HOTPLUG=n. Rebuilding zonelist is necessary when the system has just memory < 4G at boot, and hot add memory > 4G. because x86_64 has DMA32, ZONE_NORAML is not included into zonelist at boot time if system doesn't have memory >4G at boot. [AK: should just force the higher zones at boot time when SRAT tells us] 2) zone and node's spanned_pages and present_pages are not incremented. They should be. For example, our server (ia64/Fujitsu PrimeQuest) can equip memory from 4G to 1T(maybe 2T in future), and SRAT will *always* say we have possible 1T +memory. (Microsoft requires "write all possible memory in SRAT") When we reserve memmap for possible 1T memory, Linux will not work well in +minimum 4G configuraion ;) [AK: needs limiting to 5-10% of max memory] ] Signed-off-by: Andi Kleen <ak@suse.de> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-04-07 17:49:18 +00:00
if (ma->flags & ACPI_SRAT_MEM_HOT_PLUGGABLE)
update_nodes_add(node, start, end);
}
void __init acpi_numa_arch_fixup(void) {}
int __init x86_acpi_numa_init(void)
{
int ret;
ret = acpi_numa_init();
if (ret < 0)
return ret;
return srat_disabled() ? -EINVAL : 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
};
/*
* 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;
}
x86, numa: Fix cpu to node mapping for sparse node ids NUMA boot code assumes that physical node ids start at 0, but the DIMMs that the apic id represents may not be reachable. If this is the case, node 0 is never online and cpus never end up getting appropriately assigned to a node. This causes the cpumask of all online nodes to be empty and machines crash with kernel code assuming online nodes have valid cpus. The fix is to appropriately map all the address ranges for physical nodes and ensure the cpu to node mapping function checks all possible nodes (up to MAX_NUMNODES) instead of simply checking nodes 0-N, where N is the number of physical nodes, for valid address ranges. This requires no longer "compressing" the address ranges of nodes in the physical node map from 0-N, but rather leave indices in physnodes[] to represent the actual node id of the physical node. Accordingly, the topology exported by both amd_get_nodes() and acpi_get_nodes() no longer must return the number of nodes to iterate through; all such iterations will now be to MAX_NUMNODES. This change also passes the end address of system RAM (which may be different from normal operation if mem= is specified on the command line) before the physnodes[] array is populated. ACPI parsed nodes are truncated to fit within the address range that respect the mem= boundaries and even some physical nodes may become unreachable in such cases. When NUMA emulation does succeed, any apicid to node mapping that exists for unreachable nodes are given default values so that proximity domains can still be assigned. This is important for node_distance() to function as desired. Signed-off-by: David Rientjes <rientjes@google.com> LKML-Reference: <alpine.DEB.2.00.1012221702090.3701@chino.kir.corp.google.com> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
2010-12-23 01:23:56 +00:00
/*
* 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 &&
x86, numa: Fix cpu to node mapping for sparse node ids NUMA boot code assumes that physical node ids start at 0, but the DIMMs that the apic id represents may not be reachable. If this is the case, node 0 is never online and cpus never end up getting appropriately assigned to a node. This causes the cpumask of all online nodes to be empty and machines crash with kernel code assuming online nodes have valid cpus. The fix is to appropriately map all the address ranges for physical nodes and ensure the cpu to node mapping function checks all possible nodes (up to MAX_NUMNODES) instead of simply checking nodes 0-N, where N is the number of physical nodes, for valid address ranges. This requires no longer "compressing" the address ranges of nodes in the physical node map from 0-N, but rather leave indices in physnodes[] to represent the actual node id of the physical node. Accordingly, the topology exported by both amd_get_nodes() and acpi_get_nodes() no longer must return the number of nodes to iterate through; all such iterations will now be to MAX_NUMNODES. This change also passes the end address of system RAM (which may be different from normal operation if mem= is specified on the command line) before the physnodes[] array is populated. ACPI parsed nodes are truncated to fit within the address range that respect the mem= boundaries and even some physical nodes may become unreachable in such cases. When NUMA emulation does succeed, any apicid to node mapping that exists for unreachable nodes are given default values so that proximity domains can still be assigned. This is important for node_distance() to function as desired. Signed-off-by: David Rientjes <rientjes@google.com> LKML-Reference: <alpine.DEB.2.00.1012221702090.3701@chino.kir.corp.google.com> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
2010-12-23 01:23:56 +00:00
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));
for (i = 0; i < num_nodes; i++)
if (fake_nodes[i].start != fake_nodes[i].end)
node_set(i, numa_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)
{
int i, ret = 0;
for_each_node(i)
if (nodes_add[i].start <= start && nodes_add[i].end > start)
ret = i;
return ret;
}
EXPORT_SYMBOL_GPL(memory_add_physaddr_to_nid);
#endif