linux/arch/x86/mm/numa_64.c

649 lines
16 KiB
C
Raw Normal View History

/*
* Generic VM initialization for x86-64 NUMA setups.
* Copyright 2002,2003 Andi Kleen, SuSE Labs.
*/
#include <linux/kernel.h>
#include <linux/mm.h>
#include <linux/string.h>
#include <linux/init.h>
#include <linux/bootmem.h>
#include <linux/mmzone.h>
#include <linux/ctype.h>
#include <linux/module.h>
#include <linux/nodemask.h>
#include <asm/e820.h>
#include <asm/proto.h>
#include <asm/dma.h>
#include <asm/numa.h>
#include <asm/acpi.h>
#ifndef Dprintk
#define Dprintk(x...)
#endif
struct pglist_data *node_data[MAX_NUMNODES] __read_mostly;
bootmem_data_t plat_node_bdata[MAX_NUMNODES];
struct memnode memnode;
unsigned char cpu_to_node[NR_CPUS] __read_mostly = {
[0 ... NR_CPUS-1] = NUMA_NO_NODE
};
unsigned char apicid_to_node[MAX_LOCAL_APIC] __cpuinitdata = {
[0 ... MAX_LOCAL_APIC-1] = NUMA_NO_NODE
};
cpumask_t node_to_cpumask[MAX_NUMNODES] __read_mostly;
int numa_off __initdata;
unsigned long __initdata nodemap_addr;
unsigned long __initdata nodemap_size;
/*
* Given a shift value, try to populate memnodemap[]
* Returns :
* 1 if OK
* 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 i;
int res = -1;
unsigned long addr, end;
memset(memnodemap, 0xff, memnodemapsize);
for (i = 0; i < numnodes; i++) {
addr = nodes[i].start;
end = nodes[i].end;
if (addr >= end)
continue;
if ((end >> shift) >= memnodemapsize)
return 0;
do {
if (memnodemap[addr >> shift] != 0xff)
return -1;
memnodemap[addr >> shift] = i;
addr += (1UL << shift);
} while (addr < end);
res = 1;
}
return res;
}
static int __init allocate_cachealigned_memnodemap(void)
{
unsigned long pad, pad_addr;
memnodemap = memnode.embedded_map;
if (memnodemapsize <= 48)
return 0;
pad = L1_CACHE_BYTES - 1;
pad_addr = 0x8000;
nodemap_size = pad + memnodemapsize;
nodemap_addr = find_e820_area(pad_addr, end_pfn<<PAGE_SHIFT,
nodemap_size);
if (nodemap_addr == -1UL) {
printk(KERN_ERR
"NUMA: Unable to allocate Memory to Node hash map\n");
nodemap_addr = nodemap_size = 0;
return -1;
}
pad_addr = (nodemap_addr + pad) & ~pad;
memnodemap = phys_to_virt(pad_addr);
printk(KERN_DEBUG "NUMA: Allocated memnodemap from %lx - %lx\n",
nodemap_addr, nodemap_addr + nodemap_size);
return 0;
}
/*
* 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)
{
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;
if (start >= end)
continue;
bitfield |= start;
nodes_used++;
if (end > memtop)
memtop = end;
}
if (nodes_used <= 1)
i = 63;
else
i = find_first_bit(&bitfield, sizeof(unsigned long)*8);
memnodemapsize = (memtop >> i)+1;
return i;
}
int __init compute_hash_shift(struct bootnode *nodes, int numnodes)
{
int shift;
shift = extract_lsb_from_nodes(nodes, numnodes);
if (allocate_cachealigned_memnodemap())
return -1;
printk(KERN_DEBUG "NUMA: Using %d for the hash shift.\n",
shift);
if (populate_memnodemap(nodes, numnodes, shift) != 1) {
printk(KERN_INFO
"Your memory is not aligned you need to rebuild your kernel "
"with a bigger NODEMAPSIZE shift=%d\n",
shift);
return -1;
}
return shift;
}
#ifdef CONFIG_SPARSEMEM
int early_pfn_to_nid(unsigned long pfn)
{
return phys_to_nid(pfn << PAGE_SHIFT);
}
#endif
static void * __init
early_node_mem(int nodeid, unsigned long start, unsigned long end,
unsigned long size)
{
unsigned long mem = find_e820_area(start, end, size);
void *ptr;
if (mem != -1L)
return __va(mem);
ptr = __alloc_bootmem_nopanic(size,
SMP_CACHE_BYTES, __pa(MAX_DMA_ADDRESS));
if (ptr == 0) {
printk(KERN_ERR "Cannot find %lu bytes in node %d\n",
size, nodeid);
return NULL;
}
return ptr;
}
/* Initialize bootmem allocator for a node */
void __init setup_node_bootmem(int nodeid, unsigned long start, unsigned long end)
{
unsigned long start_pfn, end_pfn, bootmap_pages, bootmap_size, bootmap_start;
unsigned long nodedata_phys;
void *bootmap;
const int pgdat_size = round_up(sizeof(pg_data_t), PAGE_SIZE);
start = round_up(start, ZONE_ALIGN);
printk(KERN_INFO "Bootmem setup node %d %016lx-%016lx\n", nodeid, start, end);
start_pfn = start >> PAGE_SHIFT;
end_pfn = end >> PAGE_SHIFT;
node_data[nodeid] = early_node_mem(nodeid, start, end, pgdat_size);
if (node_data[nodeid] == NULL)
return;
nodedata_phys = __pa(node_data[nodeid]);
memset(NODE_DATA(nodeid), 0, sizeof(pg_data_t));
NODE_DATA(nodeid)->bdata = &plat_node_bdata[nodeid];
NODE_DATA(nodeid)->node_start_pfn = start_pfn;
NODE_DATA(nodeid)->node_spanned_pages = end_pfn - start_pfn;
/* Find a place for the bootmem map */
bootmap_pages = bootmem_bootmap_pages(end_pfn - start_pfn);
bootmap_start = round_up(nodedata_phys + pgdat_size, PAGE_SIZE);
bootmap = early_node_mem(nodeid, bootmap_start, end,
bootmap_pages<<PAGE_SHIFT);
if (bootmap == NULL) {
if (nodedata_phys < start || nodedata_phys >= end)
free_bootmem((unsigned long)node_data[nodeid],pgdat_size);
node_data[nodeid] = NULL;
return;
}
bootmap_start = __pa(bootmap);
Dprintk("bootmap start %lu pages %lu\n", bootmap_start, bootmap_pages);
bootmap_size = init_bootmem_node(NODE_DATA(nodeid),
bootmap_start >> PAGE_SHIFT,
start_pfn, end_pfn);
free_bootmem_with_active_regions(nodeid, end);
reserve_bootmem_node(NODE_DATA(nodeid), nodedata_phys, pgdat_size);
reserve_bootmem_node(NODE_DATA(nodeid), bootmap_start, bootmap_pages<<PAGE_SHIFT);
[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
#ifdef CONFIG_ACPI_NUMA
srat_reserve_add_area(nodeid);
#endif
node_set_online(nodeid);
}
/* Initialize final allocator for a zone */
void __init setup_node_zones(int nodeid)
{
unsigned long start_pfn, end_pfn, memmapsize, limit;
[PATCH] x86_64: Add 4GB DMA32 zone Add a new 4GB GFP_DMA32 zone between the GFP_DMA and GFP_NORMAL zones. As a bit of historical background: when the x86-64 port was originally designed we had some discussion if we should use a 16MB DMA zone like i386 or a 4GB DMA zone like IA64 or both. Both was ruled out at this point because it was in early 2.4 when VM is still quite shakey and had bad troubles even dealing with one DMA zone. We settled on the 16MB DMA zone mainly because we worried about older soundcards and the floppy. But this has always caused problems since then because device drivers had trouble getting enough DMA able memory. These days the VM works much better and the wide use of NUMA has proven it can deal with many zones successfully. So this patch adds both zones. This helps drivers who need a lot of memory below 4GB because their hardware is not accessing more (graphic drivers - proprietary and free ones, video frame buffer drivers, sound drivers etc.). Previously they could only use IOMMU+16MB GFP_DMA, which was not enough memory. Another common problem is that hardware who has full memory addressing for >4GB misses it for some control structures in memory (like transmit rings or other metadata). They tended to allocate memory in the 16MB GFP_DMA or the IOMMU/swiotlb then using pci_alloc_consistent, but that can tie up a lot of precious 16MB GFPDMA/IOMMU/swiotlb memory (even on AMD systems the IOMMU tends to be quite small) especially if you have many devices. With the new zone pci_alloc_consistent can just put this stuff into memory below 4GB which works better. One argument was still if the zone should be 4GB or 2GB. The main motivation for 2GB would be an unnamed not so unpopular hardware raid controller (mostly found in older machines from a particular four letter company) who has a strange 2GB restriction in firmware. But that one works ok with swiotlb/IOMMU anyways, so it doesn't really need GFP_DMA32. I chose 4GB to be compatible with IA64 and because it seems to be the most common restriction. The new zone is so far added only for x86-64. For other architectures who don't set up this new zone nothing changes. Architectures can set a compatibility define in Kconfig CONFIG_DMA_IS_DMA32 that will define GFP_DMA32 as GFP_DMA. Otherwise it's a nop because on 32bit architectures it's normally not needed because GFP_NORMAL (=0) is DMA able enough. One problem is still that GFP_DMA means different things on different architectures. e.g. some drivers used to have #ifdef ia64 use GFP_DMA (trusting it to be 4GB) #elif __x86_64__ (use other hacks like the swiotlb because 16MB is not enough) ... . This was quite ugly and is now obsolete. These should be now converted to use GFP_DMA32 unconditionally. I haven't done this yet. Or best only use pci_alloc_consistent/dma_alloc_coherent which will use GFP_DMA32 transparently. Signed-off-by: Andi Kleen <ak@suse.de> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-11-05 16:25:53 +00:00
start_pfn = node_start_pfn(nodeid);
end_pfn = node_end_pfn(nodeid);
Dprintk(KERN_INFO "Setting up memmap for node %d %lx-%lx\n",
[PATCH] x86_64: Add 4GB DMA32 zone Add a new 4GB GFP_DMA32 zone between the GFP_DMA and GFP_NORMAL zones. As a bit of historical background: when the x86-64 port was originally designed we had some discussion if we should use a 16MB DMA zone like i386 or a 4GB DMA zone like IA64 or both. Both was ruled out at this point because it was in early 2.4 when VM is still quite shakey and had bad troubles even dealing with one DMA zone. We settled on the 16MB DMA zone mainly because we worried about older soundcards and the floppy. But this has always caused problems since then because device drivers had trouble getting enough DMA able memory. These days the VM works much better and the wide use of NUMA has proven it can deal with many zones successfully. So this patch adds both zones. This helps drivers who need a lot of memory below 4GB because their hardware is not accessing more (graphic drivers - proprietary and free ones, video frame buffer drivers, sound drivers etc.). Previously they could only use IOMMU+16MB GFP_DMA, which was not enough memory. Another common problem is that hardware who has full memory addressing for >4GB misses it for some control structures in memory (like transmit rings or other metadata). They tended to allocate memory in the 16MB GFP_DMA or the IOMMU/swiotlb then using pci_alloc_consistent, but that can tie up a lot of precious 16MB GFPDMA/IOMMU/swiotlb memory (even on AMD systems the IOMMU tends to be quite small) especially if you have many devices. With the new zone pci_alloc_consistent can just put this stuff into memory below 4GB which works better. One argument was still if the zone should be 4GB or 2GB. The main motivation for 2GB would be an unnamed not so unpopular hardware raid controller (mostly found in older machines from a particular four letter company) who has a strange 2GB restriction in firmware. But that one works ok with swiotlb/IOMMU anyways, so it doesn't really need GFP_DMA32. I chose 4GB to be compatible with IA64 and because it seems to be the most common restriction. The new zone is so far added only for x86-64. For other architectures who don't set up this new zone nothing changes. Architectures can set a compatibility define in Kconfig CONFIG_DMA_IS_DMA32 that will define GFP_DMA32 as GFP_DMA. Otherwise it's a nop because on 32bit architectures it's normally not needed because GFP_NORMAL (=0) is DMA able enough. One problem is still that GFP_DMA means different things on different architectures. e.g. some drivers used to have #ifdef ia64 use GFP_DMA (trusting it to be 4GB) #elif __x86_64__ (use other hacks like the swiotlb because 16MB is not enough) ... . This was quite ugly and is now obsolete. These should be now converted to use GFP_DMA32 unconditionally. I haven't done this yet. Or best only use pci_alloc_consistent/dma_alloc_coherent which will use GFP_DMA32 transparently. Signed-off-by: Andi Kleen <ak@suse.de> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-11-05 16:25:53 +00:00
nodeid, start_pfn, end_pfn);
/* Try to allocate mem_map at end to not fill up precious <4GB
memory. */
memmapsize = sizeof(struct page) * (end_pfn-start_pfn);
limit = end_pfn << PAGE_SHIFT;
#ifdef CONFIG_FLAT_NODE_MEM_MAP
NODE_DATA(nodeid)->node_mem_map =
__alloc_bootmem_core(NODE_DATA(nodeid)->bdata,
memmapsize, SMP_CACHE_BYTES,
round_down(limit - memmapsize, PAGE_SIZE),
limit);
#endif
}
void __init numa_init_array(void)
{
int rr, i;
/* 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. */
rr = first_node(node_online_map);
for (i = 0; i < NR_CPUS; i++) {
if (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 */
[PATCH] x86-64: configurable fake numa node sizes Extends the numa=fake x86_64 command-line option to allow for configurable node sizes. These nodes can be used in conjunction with cpusets for coarse memory resource management. The old command-line option is still supported: numa=fake=32 gives 32 fake NUMA nodes, ignoring the NUMA setup of the actual machine. But now you may configure your system for the node sizes of your choice: numa=fake=2*512,1024,2*256 gives two 512M nodes, one 1024M node, two 256M nodes, and the rest of system memory to a sixth node. The existing hash function is maintained to support the various node sizes that are possible with this implementation. Each node of the same size receives roughly the same amount of available pages, regardless of any reserved memory with its address range. The total available pages on the system is calculated and divided by the number of equal nodes to allocate. These nodes are then dynamically allocated and their borders extended until such time as their number of available pages reaches the required size. Configurable node sizes are recommended when used in conjunction with cpusets for memory control because it eliminates the overhead associated with scanning the zonelists of many smaller full nodes on page_alloc(). Cc: Andi Kleen <ak@suse.de> Signed-off-by: David Rientjes <rientjes@google.com> Signed-off-by: Andi Kleen <ak@suse.de> Cc: Paul Jackson <pj@sgi.com> Cc: Christoph Lameter <clameter@engr.sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2007-05-02 17:27:09 +00:00
char *cmdline __initdata;
/*
[PATCH] x86-64: configurable fake numa node sizes Extends the numa=fake x86_64 command-line option to allow for configurable node sizes. These nodes can be used in conjunction with cpusets for coarse memory resource management. The old command-line option is still supported: numa=fake=32 gives 32 fake NUMA nodes, ignoring the NUMA setup of the actual machine. But now you may configure your system for the node sizes of your choice: numa=fake=2*512,1024,2*256 gives two 512M nodes, one 1024M node, two 256M nodes, and the rest of system memory to a sixth node. The existing hash function is maintained to support the various node sizes that are possible with this implementation. Each node of the same size receives roughly the same amount of available pages, regardless of any reserved memory with its address range. The total available pages on the system is calculated and divided by the number of equal nodes to allocate. These nodes are then dynamically allocated and their borders extended until such time as their number of available pages reaches the required size. Configurable node sizes are recommended when used in conjunction with cpusets for memory control because it eliminates the overhead associated with scanning the zonelists of many smaller full nodes on page_alloc(). Cc: Andi Kleen <ak@suse.de> Signed-off-by: David Rientjes <rientjes@google.com> Signed-off-by: Andi Kleen <ak@suse.de> Cc: Paul Jackson <pj@sgi.com> Cc: Christoph Lameter <clameter@engr.sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2007-05-02 17:27:09 +00:00
* 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.
*/
[PATCH] x86-64: configurable fake numa node sizes Extends the numa=fake x86_64 command-line option to allow for configurable node sizes. These nodes can be used in conjunction with cpusets for coarse memory resource management. The old command-line option is still supported: numa=fake=32 gives 32 fake NUMA nodes, ignoring the NUMA setup of the actual machine. But now you may configure your system for the node sizes of your choice: numa=fake=2*512,1024,2*256 gives two 512M nodes, one 1024M node, two 256M nodes, and the rest of system memory to a sixth node. The existing hash function is maintained to support the various node sizes that are possible with this implementation. Each node of the same size receives roughly the same amount of available pages, regardless of any reserved memory with its address range. The total available pages on the system is calculated and divided by the number of equal nodes to allocate. These nodes are then dynamically allocated and their borders extended until such time as their number of available pages reaches the required size. Configurable node sizes are recommended when used in conjunction with cpusets for memory control because it eliminates the overhead associated with scanning the zonelists of many smaller full nodes on page_alloc(). Cc: Andi Kleen <ak@suse.de> Signed-off-by: David Rientjes <rientjes@google.com> Signed-off-by: Andi Kleen <ak@suse.de> Cc: Paul Jackson <pj@sgi.com> Cc: Christoph Lameter <clameter@engr.sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2007-05-02 17:27:09 +00:00
static int __init setup_node_range(int nid, struct bootnode *nodes, u64 *addr,
u64 size, u64 max_addr)
{
[PATCH] x86-64: configurable fake numa node sizes Extends the numa=fake x86_64 command-line option to allow for configurable node sizes. These nodes can be used in conjunction with cpusets for coarse memory resource management. The old command-line option is still supported: numa=fake=32 gives 32 fake NUMA nodes, ignoring the NUMA setup of the actual machine. But now you may configure your system for the node sizes of your choice: numa=fake=2*512,1024,2*256 gives two 512M nodes, one 1024M node, two 256M nodes, and the rest of system memory to a sixth node. The existing hash function is maintained to support the various node sizes that are possible with this implementation. Each node of the same size receives roughly the same amount of available pages, regardless of any reserved memory with its address range. The total available pages on the system is calculated and divided by the number of equal nodes to allocate. These nodes are then dynamically allocated and their borders extended until such time as their number of available pages reaches the required size. Configurable node sizes are recommended when used in conjunction with cpusets for memory control because it eliminates the overhead associated with scanning the zonelists of many smaller full nodes on page_alloc(). Cc: Andi Kleen <ak@suse.de> Signed-off-by: David Rientjes <rientjes@google.com> Signed-off-by: Andi Kleen <ak@suse.de> Cc: Paul Jackson <pj@sgi.com> Cc: Christoph Lameter <clameter@engr.sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2007-05-02 17:27:09 +00:00
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);
[PATCH] x86-64: configurable fake numa node sizes Extends the numa=fake x86_64 command-line option to allow for configurable node sizes. These nodes can be used in conjunction with cpusets for coarse memory resource management. The old command-line option is still supported: numa=fake=32 gives 32 fake NUMA nodes, ignoring the NUMA setup of the actual machine. But now you may configure your system for the node sizes of your choice: numa=fake=2*512,1024,2*256 gives two 512M nodes, one 1024M node, two 256M nodes, and the rest of system memory to a sixth node. The existing hash function is maintained to support the various node sizes that are possible with this implementation. Each node of the same size receives roughly the same amount of available pages, regardless of any reserved memory with its address range. The total available pages on the system is calculated and divided by the number of equal nodes to allocate. These nodes are then dynamically allocated and their borders extended until such time as their number of available pages reaches the required size. Configurable node sizes are recommended when used in conjunction with cpusets for memory control because it eliminates the overhead associated with scanning the zonelists of many smaller full nodes on page_alloc(). Cc: Andi Kleen <ak@suse.de> Signed-off-by: David Rientjes <rientjes@google.com> Signed-off-by: Andi Kleen <ak@suse.de> Cc: Paul Jackson <pj@sgi.com> Cc: Christoph Lameter <clameter@engr.sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2007-05-02 17:27:09 +00:00
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;
}
[PATCH] x86-64: configurable fake numa node sizes Extends the numa=fake x86_64 command-line option to allow for configurable node sizes. These nodes can be used in conjunction with cpusets for coarse memory resource management. The old command-line option is still supported: numa=fake=32 gives 32 fake NUMA nodes, ignoring the NUMA setup of the actual machine. But now you may configure your system for the node sizes of your choice: numa=fake=2*512,1024,2*256 gives two 512M nodes, one 1024M node, two 256M nodes, and the rest of system memory to a sixth node. The existing hash function is maintained to support the various node sizes that are possible with this implementation. Each node of the same size receives roughly the same amount of available pages, regardless of any reserved memory with its address range. The total available pages on the system is calculated and divided by the number of equal nodes to allocate. These nodes are then dynamically allocated and their borders extended until such time as their number of available pages reaches the required size. Configurable node sizes are recommended when used in conjunction with cpusets for memory control because it eliminates the overhead associated with scanning the zonelists of many smaller full nodes on page_alloc(). Cc: Andi Kleen <ak@suse.de> Signed-off-by: David Rientjes <rientjes@google.com> Signed-off-by: Andi Kleen <ak@suse.de> Cc: Paul Jackson <pj@sgi.com> Cc: Christoph Lameter <clameter@engr.sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2007-05-02 17:27:09 +00:00
/*
* Splits num_nodes nodes up equally starting at node_start. The return value
* is the number of nodes split up and addr is adjusted to be at the end of the
* last node allocated.
*/
static int __init split_nodes_equally(struct bootnode *nodes, u64 *addr,
u64 max_addr, int node_start,
int num_nodes)
{
[PATCH] x86-64: configurable fake numa node sizes Extends the numa=fake x86_64 command-line option to allow for configurable node sizes. These nodes can be used in conjunction with cpusets for coarse memory resource management. The old command-line option is still supported: numa=fake=32 gives 32 fake NUMA nodes, ignoring the NUMA setup of the actual machine. But now you may configure your system for the node sizes of your choice: numa=fake=2*512,1024,2*256 gives two 512M nodes, one 1024M node, two 256M nodes, and the rest of system memory to a sixth node. The existing hash function is maintained to support the various node sizes that are possible with this implementation. Each node of the same size receives roughly the same amount of available pages, regardless of any reserved memory with its address range. The total available pages on the system is calculated and divided by the number of equal nodes to allocate. These nodes are then dynamically allocated and their borders extended until such time as their number of available pages reaches the required size. Configurable node sizes are recommended when used in conjunction with cpusets for memory control because it eliminates the overhead associated with scanning the zonelists of many smaller full nodes on page_alloc(). Cc: Andi Kleen <ak@suse.de> Signed-off-by: David Rientjes <rientjes@google.com> Signed-off-by: Andi Kleen <ak@suse.de> Cc: Paul Jackson <pj@sgi.com> Cc: Christoph Lameter <clameter@engr.sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2007-05-02 17:27:09 +00:00
unsigned int big;
u64 size;
int i;
[PATCH] x86-64: configurable fake numa node sizes Extends the numa=fake x86_64 command-line option to allow for configurable node sizes. These nodes can be used in conjunction with cpusets for coarse memory resource management. The old command-line option is still supported: numa=fake=32 gives 32 fake NUMA nodes, ignoring the NUMA setup of the actual machine. But now you may configure your system for the node sizes of your choice: numa=fake=2*512,1024,2*256 gives two 512M nodes, one 1024M node, two 256M nodes, and the rest of system memory to a sixth node. The existing hash function is maintained to support the various node sizes that are possible with this implementation. Each node of the same size receives roughly the same amount of available pages, regardless of any reserved memory with its address range. The total available pages on the system is calculated and divided by the number of equal nodes to allocate. These nodes are then dynamically allocated and their borders extended until such time as their number of available pages reaches the required size. Configurable node sizes are recommended when used in conjunction with cpusets for memory control because it eliminates the overhead associated with scanning the zonelists of many smaller full nodes on page_alloc(). Cc: Andi Kleen <ak@suse.de> Signed-off-by: David Rientjes <rientjes@google.com> Signed-off-by: Andi Kleen <ak@suse.de> Cc: Paul Jackson <pj@sgi.com> Cc: Christoph Lameter <clameter@engr.sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2007-05-02 17:27:09 +00:00
if (num_nodes <= 0)
return -1;
if (num_nodes > MAX_NUMNODES)
num_nodes = MAX_NUMNODES;
size = (max_addr - *addr - e820_hole_size(*addr, max_addr)) /
[PATCH] x86-64: configurable fake numa node sizes Extends the numa=fake x86_64 command-line option to allow for configurable node sizes. These nodes can be used in conjunction with cpusets for coarse memory resource management. The old command-line option is still supported: numa=fake=32 gives 32 fake NUMA nodes, ignoring the NUMA setup of the actual machine. But now you may configure your system for the node sizes of your choice: numa=fake=2*512,1024,2*256 gives two 512M nodes, one 1024M node, two 256M nodes, and the rest of system memory to a sixth node. The existing hash function is maintained to support the various node sizes that are possible with this implementation. Each node of the same size receives roughly the same amount of available pages, regardless of any reserved memory with its address range. The total available pages on the system is calculated and divided by the number of equal nodes to allocate. These nodes are then dynamically allocated and their borders extended until such time as their number of available pages reaches the required size. Configurable node sizes are recommended when used in conjunction with cpusets for memory control because it eliminates the overhead associated with scanning the zonelists of many smaller full nodes on page_alloc(). Cc: Andi Kleen <ak@suse.de> Signed-off-by: David Rientjes <rientjes@google.com> Signed-off-by: Andi Kleen <ak@suse.de> Cc: Paul Jackson <pj@sgi.com> Cc: Christoph Lameter <clameter@engr.sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2007-05-02 17:27:09 +00:00
num_nodes;
/*
[PATCH] x86-64: configurable fake numa node sizes Extends the numa=fake x86_64 command-line option to allow for configurable node sizes. These nodes can be used in conjunction with cpusets for coarse memory resource management. The old command-line option is still supported: numa=fake=32 gives 32 fake NUMA nodes, ignoring the NUMA setup of the actual machine. But now you may configure your system for the node sizes of your choice: numa=fake=2*512,1024,2*256 gives two 512M nodes, one 1024M node, two 256M nodes, and the rest of system memory to a sixth node. The existing hash function is maintained to support the various node sizes that are possible with this implementation. Each node of the same size receives roughly the same amount of available pages, regardless of any reserved memory with its address range. The total available pages on the system is calculated and divided by the number of equal nodes to allocate. These nodes are then dynamically allocated and their borders extended until such time as their number of available pages reaches the required size. Configurable node sizes are recommended when used in conjunction with cpusets for memory control because it eliminates the overhead associated with scanning the zonelists of many smaller full nodes on page_alloc(). Cc: Andi Kleen <ak@suse.de> Signed-off-by: David Rientjes <rientjes@google.com> Signed-off-by: Andi Kleen <ak@suse.de> Cc: Paul Jackson <pj@sgi.com> Cc: Christoph Lameter <clameter@engr.sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2007-05-02 17:27:09 +00:00
* Calculate the number of big nodes that can be allocated as a result
* of consolidating the leftovers.
*/
[PATCH] x86-64: configurable fake numa node sizes Extends the numa=fake x86_64 command-line option to allow for configurable node sizes. These nodes can be used in conjunction with cpusets for coarse memory resource management. The old command-line option is still supported: numa=fake=32 gives 32 fake NUMA nodes, ignoring the NUMA setup of the actual machine. But now you may configure your system for the node sizes of your choice: numa=fake=2*512,1024,2*256 gives two 512M nodes, one 1024M node, two 256M nodes, and the rest of system memory to a sixth node. The existing hash function is maintained to support the various node sizes that are possible with this implementation. Each node of the same size receives roughly the same amount of available pages, regardless of any reserved memory with its address range. The total available pages on the system is calculated and divided by the number of equal nodes to allocate. These nodes are then dynamically allocated and their borders extended until such time as their number of available pages reaches the required size. Configurable node sizes are recommended when used in conjunction with cpusets for memory control because it eliminates the overhead associated with scanning the zonelists of many smaller full nodes on page_alloc(). Cc: Andi Kleen <ak@suse.de> Signed-off-by: David Rientjes <rientjes@google.com> Signed-off-by: Andi Kleen <ak@suse.de> Cc: Paul Jackson <pj@sgi.com> Cc: Christoph Lameter <clameter@engr.sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2007-05-02 17:27:09 +00:00
big = ((size & ~FAKE_NODE_MIN_HASH_MASK) * num_nodes) /
FAKE_NODE_MIN_SIZE;
/* Round down to nearest FAKE_NODE_MIN_SIZE. */
size &= FAKE_NODE_MIN_HASH_MASK;
if (!size) {
printk(KERN_ERR "Not enough memory for each node. "
"NUMA emulation disabled.\n");
return -1;
}
[PATCH] x86-64: configurable fake numa node sizes Extends the numa=fake x86_64 command-line option to allow for configurable node sizes. These nodes can be used in conjunction with cpusets for coarse memory resource management. The old command-line option is still supported: numa=fake=32 gives 32 fake NUMA nodes, ignoring the NUMA setup of the actual machine. But now you may configure your system for the node sizes of your choice: numa=fake=2*512,1024,2*256 gives two 512M nodes, one 1024M node, two 256M nodes, and the rest of system memory to a sixth node. The existing hash function is maintained to support the various node sizes that are possible with this implementation. Each node of the same size receives roughly the same amount of available pages, regardless of any reserved memory with its address range. The total available pages on the system is calculated and divided by the number of equal nodes to allocate. These nodes are then dynamically allocated and their borders extended until such time as their number of available pages reaches the required size. Configurable node sizes are recommended when used in conjunction with cpusets for memory control because it eliminates the overhead associated with scanning the zonelists of many smaller full nodes on page_alloc(). Cc: Andi Kleen <ak@suse.de> Signed-off-by: David Rientjes <rientjes@google.com> Signed-off-by: Andi Kleen <ak@suse.de> Cc: Paul Jackson <pj@sgi.com> Cc: Christoph Lameter <clameter@engr.sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2007-05-02 17:27:09 +00:00
for (i = node_start; i < num_nodes + node_start; i++) {
u64 end = *addr + size;
if (i < big)
end += FAKE_NODE_MIN_SIZE;
/*
[PATCH] x86-64: configurable fake numa node sizes Extends the numa=fake x86_64 command-line option to allow for configurable node sizes. These nodes can be used in conjunction with cpusets for coarse memory resource management. The old command-line option is still supported: numa=fake=32 gives 32 fake NUMA nodes, ignoring the NUMA setup of the actual machine. But now you may configure your system for the node sizes of your choice: numa=fake=2*512,1024,2*256 gives two 512M nodes, one 1024M node, two 256M nodes, and the rest of system memory to a sixth node. The existing hash function is maintained to support the various node sizes that are possible with this implementation. Each node of the same size receives roughly the same amount of available pages, regardless of any reserved memory with its address range. The total available pages on the system is calculated and divided by the number of equal nodes to allocate. These nodes are then dynamically allocated and their borders extended until such time as their number of available pages reaches the required size. Configurable node sizes are recommended when used in conjunction with cpusets for memory control because it eliminates the overhead associated with scanning the zonelists of many smaller full nodes on page_alloc(). Cc: Andi Kleen <ak@suse.de> Signed-off-by: David Rientjes <rientjes@google.com> Signed-off-by: Andi Kleen <ak@suse.de> Cc: Paul Jackson <pj@sgi.com> Cc: Christoph Lameter <clameter@engr.sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2007-05-02 17:27:09 +00:00
* The final node can have the remaining system RAM. Other
* nodes receive roughly the same amount of available pages.
*/
[PATCH] x86-64: configurable fake numa node sizes Extends the numa=fake x86_64 command-line option to allow for configurable node sizes. These nodes can be used in conjunction with cpusets for coarse memory resource management. The old command-line option is still supported: numa=fake=32 gives 32 fake NUMA nodes, ignoring the NUMA setup of the actual machine. But now you may configure your system for the node sizes of your choice: numa=fake=2*512,1024,2*256 gives two 512M nodes, one 1024M node, two 256M nodes, and the rest of system memory to a sixth node. The existing hash function is maintained to support the various node sizes that are possible with this implementation. Each node of the same size receives roughly the same amount of available pages, regardless of any reserved memory with its address range. The total available pages on the system is calculated and divided by the number of equal nodes to allocate. These nodes are then dynamically allocated and their borders extended until such time as their number of available pages reaches the required size. Configurable node sizes are recommended when used in conjunction with cpusets for memory control because it eliminates the overhead associated with scanning the zonelists of many smaller full nodes on page_alloc(). Cc: Andi Kleen <ak@suse.de> Signed-off-by: David Rientjes <rientjes@google.com> Signed-off-by: Andi Kleen <ak@suse.de> Cc: Paul Jackson <pj@sgi.com> Cc: Christoph Lameter <clameter@engr.sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2007-05-02 17:27:09 +00:00
if (i == num_nodes + node_start - 1)
end = max_addr;
else
while (end - *addr - e820_hole_size(*addr, end) <
[PATCH] x86-64: configurable fake numa node sizes Extends the numa=fake x86_64 command-line option to allow for configurable node sizes. These nodes can be used in conjunction with cpusets for coarse memory resource management. The old command-line option is still supported: numa=fake=32 gives 32 fake NUMA nodes, ignoring the NUMA setup of the actual machine. But now you may configure your system for the node sizes of your choice: numa=fake=2*512,1024,2*256 gives two 512M nodes, one 1024M node, two 256M nodes, and the rest of system memory to a sixth node. The existing hash function is maintained to support the various node sizes that are possible with this implementation. Each node of the same size receives roughly the same amount of available pages, regardless of any reserved memory with its address range. The total available pages on the system is calculated and divided by the number of equal nodes to allocate. These nodes are then dynamically allocated and their borders extended until such time as their number of available pages reaches the required size. Configurable node sizes are recommended when used in conjunction with cpusets for memory control because it eliminates the overhead associated with scanning the zonelists of many smaller full nodes on page_alloc(). Cc: Andi Kleen <ak@suse.de> Signed-off-by: David Rientjes <rientjes@google.com> Signed-off-by: Andi Kleen <ak@suse.de> Cc: Paul Jackson <pj@sgi.com> Cc: Christoph Lameter <clameter@engr.sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2007-05-02 17:27:09 +00:00
size) {
end += FAKE_NODE_MIN_SIZE;
if (end > max_addr) {
end = max_addr;
break;
}
}
if (setup_node_range(i, nodes, addr, end - *addr, max_addr) < 0)
break;
}
return i - node_start + 1;
}
/*
* Splits the remaining system RAM into chunks of size. The remaining memory is
* always assigned to a final node and can be asymmetric. Returns the number of
* nodes split.
*/
static int __init split_nodes_by_size(struct bootnode *nodes, u64 *addr,
u64 max_addr, int node_start, u64 size)
{
int i = node_start;
size = (size << 20) & FAKE_NODE_MIN_HASH_MASK;
while (!setup_node_range(i++, nodes, addr, size, max_addr))
;
return i - node_start;
}
[PATCH] x86-64: configurable fake numa node sizes Extends the numa=fake x86_64 command-line option to allow for configurable node sizes. These nodes can be used in conjunction with cpusets for coarse memory resource management. The old command-line option is still supported: numa=fake=32 gives 32 fake NUMA nodes, ignoring the NUMA setup of the actual machine. But now you may configure your system for the node sizes of your choice: numa=fake=2*512,1024,2*256 gives two 512M nodes, one 1024M node, two 256M nodes, and the rest of system memory to a sixth node. The existing hash function is maintained to support the various node sizes that are possible with this implementation. Each node of the same size receives roughly the same amount of available pages, regardless of any reserved memory with its address range. The total available pages on the system is calculated and divided by the number of equal nodes to allocate. These nodes are then dynamically allocated and their borders extended until such time as their number of available pages reaches the required size. Configurable node sizes are recommended when used in conjunction with cpusets for memory control because it eliminates the overhead associated with scanning the zonelists of many smaller full nodes on page_alloc(). Cc: Andi Kleen <ak@suse.de> Signed-off-by: David Rientjes <rientjes@google.com> Signed-off-by: Andi Kleen <ak@suse.de> Cc: Paul Jackson <pj@sgi.com> Cc: Christoph Lameter <clameter@engr.sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2007-05-02 17:27:09 +00:00
/*
* Sets up the system RAM area from start_pfn to end_pfn according to the
* numa=fake command-line option.
*/
static int __init numa_emulation(unsigned long start_pfn, unsigned long end_pfn)
{
struct bootnode nodes[MAX_NUMNODES];
u64 addr = start_pfn << PAGE_SHIFT;
u64 max_addr = end_pfn << PAGE_SHIFT;
int num_nodes = 0;
int coeff_flag;
int coeff = -1;
int num = 0;
[PATCH] x86-64: configurable fake numa node sizes Extends the numa=fake x86_64 command-line option to allow for configurable node sizes. These nodes can be used in conjunction with cpusets for coarse memory resource management. The old command-line option is still supported: numa=fake=32 gives 32 fake NUMA nodes, ignoring the NUMA setup of the actual machine. But now you may configure your system for the node sizes of your choice: numa=fake=2*512,1024,2*256 gives two 512M nodes, one 1024M node, two 256M nodes, and the rest of system memory to a sixth node. The existing hash function is maintained to support the various node sizes that are possible with this implementation. Each node of the same size receives roughly the same amount of available pages, regardless of any reserved memory with its address range. The total available pages on the system is calculated and divided by the number of equal nodes to allocate. These nodes are then dynamically allocated and their borders extended until such time as their number of available pages reaches the required size. Configurable node sizes are recommended when used in conjunction with cpusets for memory control because it eliminates the overhead associated with scanning the zonelists of many smaller full nodes on page_alloc(). Cc: Andi Kleen <ak@suse.de> Signed-off-by: David Rientjes <rientjes@google.com> Signed-off-by: Andi Kleen <ak@suse.de> Cc: Paul Jackson <pj@sgi.com> Cc: Christoph Lameter <clameter@engr.sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2007-05-02 17:27:09 +00:00
u64 size;
int i;
memset(&nodes, 0, sizeof(nodes));
/*
* If the numa=fake command-line is just a single number N, split the
* system RAM into N fake nodes.
*/
if (!strchr(cmdline, '*') && !strchr(cmdline, ',')) {
num_nodes = split_nodes_equally(nodes, &addr, max_addr, 0,
simple_strtol(cmdline, NULL, 0));
if (num_nodes < 0)
return num_nodes;
goto out;
}
/* Parse the command line. */
for (coeff_flag = 0; ; cmdline++) {
[PATCH] x86-64: configurable fake numa node sizes Extends the numa=fake x86_64 command-line option to allow for configurable node sizes. These nodes can be used in conjunction with cpusets for coarse memory resource management. The old command-line option is still supported: numa=fake=32 gives 32 fake NUMA nodes, ignoring the NUMA setup of the actual machine. But now you may configure your system for the node sizes of your choice: numa=fake=2*512,1024,2*256 gives two 512M nodes, one 1024M node, two 256M nodes, and the rest of system memory to a sixth node. The existing hash function is maintained to support the various node sizes that are possible with this implementation. Each node of the same size receives roughly the same amount of available pages, regardless of any reserved memory with its address range. The total available pages on the system is calculated and divided by the number of equal nodes to allocate. These nodes are then dynamically allocated and their borders extended until such time as their number of available pages reaches the required size. Configurable node sizes are recommended when used in conjunction with cpusets for memory control because it eliminates the overhead associated with scanning the zonelists of many smaller full nodes on page_alloc(). Cc: Andi Kleen <ak@suse.de> Signed-off-by: David Rientjes <rientjes@google.com> Signed-off-by: Andi Kleen <ak@suse.de> Cc: Paul Jackson <pj@sgi.com> Cc: Christoph Lameter <clameter@engr.sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2007-05-02 17:27:09 +00:00
if (*cmdline && isdigit(*cmdline)) {
num = num * 10 + *cmdline - '0';
continue;
}
if (*cmdline == '*') {
if (num > 0)
coeff = num;
coeff_flag = 1;
}
[PATCH] x86-64: configurable fake numa node sizes Extends the numa=fake x86_64 command-line option to allow for configurable node sizes. These nodes can be used in conjunction with cpusets for coarse memory resource management. The old command-line option is still supported: numa=fake=32 gives 32 fake NUMA nodes, ignoring the NUMA setup of the actual machine. But now you may configure your system for the node sizes of your choice: numa=fake=2*512,1024,2*256 gives two 512M nodes, one 1024M node, two 256M nodes, and the rest of system memory to a sixth node. The existing hash function is maintained to support the various node sizes that are possible with this implementation. Each node of the same size receives roughly the same amount of available pages, regardless of any reserved memory with its address range. The total available pages on the system is calculated and divided by the number of equal nodes to allocate. These nodes are then dynamically allocated and their borders extended until such time as their number of available pages reaches the required size. Configurable node sizes are recommended when used in conjunction with cpusets for memory control because it eliminates the overhead associated with scanning the zonelists of many smaller full nodes on page_alloc(). Cc: Andi Kleen <ak@suse.de> Signed-off-by: David Rientjes <rientjes@google.com> Signed-off-by: Andi Kleen <ak@suse.de> Cc: Paul Jackson <pj@sgi.com> Cc: Christoph Lameter <clameter@engr.sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2007-05-02 17:27:09 +00:00
if (!*cmdline || *cmdline == ',') {
if (!coeff_flag)
coeff = 1;
[PATCH] x86-64: configurable fake numa node sizes Extends the numa=fake x86_64 command-line option to allow for configurable node sizes. These nodes can be used in conjunction with cpusets for coarse memory resource management. The old command-line option is still supported: numa=fake=32 gives 32 fake NUMA nodes, ignoring the NUMA setup of the actual machine. But now you may configure your system for the node sizes of your choice: numa=fake=2*512,1024,2*256 gives two 512M nodes, one 1024M node, two 256M nodes, and the rest of system memory to a sixth node. The existing hash function is maintained to support the various node sizes that are possible with this implementation. Each node of the same size receives roughly the same amount of available pages, regardless of any reserved memory with its address range. The total available pages on the system is calculated and divided by the number of equal nodes to allocate. These nodes are then dynamically allocated and their borders extended until such time as their number of available pages reaches the required size. Configurable node sizes are recommended when used in conjunction with cpusets for memory control because it eliminates the overhead associated with scanning the zonelists of many smaller full nodes on page_alloc(). Cc: Andi Kleen <ak@suse.de> Signed-off-by: David Rientjes <rientjes@google.com> Signed-off-by: Andi Kleen <ak@suse.de> Cc: Paul Jackson <pj@sgi.com> Cc: Christoph Lameter <clameter@engr.sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2007-05-02 17:27:09 +00:00
/*
* Round down to the nearest FAKE_NODE_MIN_SIZE.
* Command-line coefficients are in megabytes.
*/
size = ((u64)num << 20) & FAKE_NODE_MIN_HASH_MASK;
if (size)
[PATCH] x86-64: configurable fake numa node sizes Extends the numa=fake x86_64 command-line option to allow for configurable node sizes. These nodes can be used in conjunction with cpusets for coarse memory resource management. The old command-line option is still supported: numa=fake=32 gives 32 fake NUMA nodes, ignoring the NUMA setup of the actual machine. But now you may configure your system for the node sizes of your choice: numa=fake=2*512,1024,2*256 gives two 512M nodes, one 1024M node, two 256M nodes, and the rest of system memory to a sixth node. The existing hash function is maintained to support the various node sizes that are possible with this implementation. Each node of the same size receives roughly the same amount of available pages, regardless of any reserved memory with its address range. The total available pages on the system is calculated and divided by the number of equal nodes to allocate. These nodes are then dynamically allocated and their borders extended until such time as their number of available pages reaches the required size. Configurable node sizes are recommended when used in conjunction with cpusets for memory control because it eliminates the overhead associated with scanning the zonelists of many smaller full nodes on page_alloc(). Cc: Andi Kleen <ak@suse.de> Signed-off-by: David Rientjes <rientjes@google.com> Signed-off-by: Andi Kleen <ak@suse.de> Cc: Paul Jackson <pj@sgi.com> Cc: Christoph Lameter <clameter@engr.sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2007-05-02 17:27:09 +00:00
for (i = 0; i < coeff; i++, num_nodes++)
if (setup_node_range(num_nodes, nodes,
&addr, size, max_addr) < 0)
goto done;
if (!*cmdline)
break;
coeff_flag = 0;
coeff = -1;
}
[PATCH] x86-64: configurable fake numa node sizes Extends the numa=fake x86_64 command-line option to allow for configurable node sizes. These nodes can be used in conjunction with cpusets for coarse memory resource management. The old command-line option is still supported: numa=fake=32 gives 32 fake NUMA nodes, ignoring the NUMA setup of the actual machine. But now you may configure your system for the node sizes of your choice: numa=fake=2*512,1024,2*256 gives two 512M nodes, one 1024M node, two 256M nodes, and the rest of system memory to a sixth node. The existing hash function is maintained to support the various node sizes that are possible with this implementation. Each node of the same size receives roughly the same amount of available pages, regardless of any reserved memory with its address range. The total available pages on the system is calculated and divided by the number of equal nodes to allocate. These nodes are then dynamically allocated and their borders extended until such time as their number of available pages reaches the required size. Configurable node sizes are recommended when used in conjunction with cpusets for memory control because it eliminates the overhead associated with scanning the zonelists of many smaller full nodes on page_alloc(). Cc: Andi Kleen <ak@suse.de> Signed-off-by: David Rientjes <rientjes@google.com> Signed-off-by: Andi Kleen <ak@suse.de> Cc: Paul Jackson <pj@sgi.com> Cc: Christoph Lameter <clameter@engr.sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2007-05-02 17:27:09 +00:00
num = 0;
}
done:
if (!num_nodes)
return -1;
/* Fill remainder of system RAM, if appropriate. */
[PATCH] x86-64: configurable fake numa node sizes Extends the numa=fake x86_64 command-line option to allow for configurable node sizes. These nodes can be used in conjunction with cpusets for coarse memory resource management. The old command-line option is still supported: numa=fake=32 gives 32 fake NUMA nodes, ignoring the NUMA setup of the actual machine. But now you may configure your system for the node sizes of your choice: numa=fake=2*512,1024,2*256 gives two 512M nodes, one 1024M node, two 256M nodes, and the rest of system memory to a sixth node. The existing hash function is maintained to support the various node sizes that are possible with this implementation. Each node of the same size receives roughly the same amount of available pages, regardless of any reserved memory with its address range. The total available pages on the system is calculated and divided by the number of equal nodes to allocate. These nodes are then dynamically allocated and their borders extended until such time as their number of available pages reaches the required size. Configurable node sizes are recommended when used in conjunction with cpusets for memory control because it eliminates the overhead associated with scanning the zonelists of many smaller full nodes on page_alloc(). Cc: Andi Kleen <ak@suse.de> Signed-off-by: David Rientjes <rientjes@google.com> Signed-off-by: Andi Kleen <ak@suse.de> Cc: Paul Jackson <pj@sgi.com> Cc: Christoph Lameter <clameter@engr.sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2007-05-02 17:27:09 +00:00
if (addr < max_addr) {
if (coeff_flag && coeff < 0) {
/* Split remaining nodes into num-sized chunks */
num_nodes += split_nodes_by_size(nodes, &addr, max_addr,
num_nodes, num);
goto out;
}
switch (*(cmdline - 1)) {
case '*':
/* Split remaining nodes into coeff chunks */
if (coeff <= 0)
break;
num_nodes += split_nodes_equally(nodes, &addr, max_addr,
num_nodes, coeff);
break;
case ',':
/* Do not allocate remaining system RAM */
break;
default:
/* Give one final node */
setup_node_range(num_nodes, nodes, &addr,
max_addr - addr, max_addr);
num_nodes++;
}
[PATCH] x86-64: configurable fake numa node sizes Extends the numa=fake x86_64 command-line option to allow for configurable node sizes. These nodes can be used in conjunction with cpusets for coarse memory resource management. The old command-line option is still supported: numa=fake=32 gives 32 fake NUMA nodes, ignoring the NUMA setup of the actual machine. But now you may configure your system for the node sizes of your choice: numa=fake=2*512,1024,2*256 gives two 512M nodes, one 1024M node, two 256M nodes, and the rest of system memory to a sixth node. The existing hash function is maintained to support the various node sizes that are possible with this implementation. Each node of the same size receives roughly the same amount of available pages, regardless of any reserved memory with its address range. The total available pages on the system is calculated and divided by the number of equal nodes to allocate. These nodes are then dynamically allocated and their borders extended until such time as their number of available pages reaches the required size. Configurable node sizes are recommended when used in conjunction with cpusets for memory control because it eliminates the overhead associated with scanning the zonelists of many smaller full nodes on page_alloc(). Cc: Andi Kleen <ak@suse.de> Signed-off-by: David Rientjes <rientjes@google.com> Signed-off-by: Andi Kleen <ak@suse.de> Cc: Paul Jackson <pj@sgi.com> Cc: Christoph Lameter <clameter@engr.sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2007-05-02 17:27:09 +00:00
}
out:
memnode_shift = compute_hash_shift(nodes, num_nodes);
if (memnode_shift < 0) {
memnode_shift = 0;
printk(KERN_ERR "No NUMA hash function found. NUMA emulation "
"disabled.\n");
return -1;
}
/*
* We need to vacate all active ranges that may have been registered by
* SRAT and set acpi_numa to -1 so that srat_disabled() always returns
* true. NUMA emulation has succeeded so we will not scan ACPI nodes.
[PATCH] x86-64: configurable fake numa node sizes Extends the numa=fake x86_64 command-line option to allow for configurable node sizes. These nodes can be used in conjunction with cpusets for coarse memory resource management. The old command-line option is still supported: numa=fake=32 gives 32 fake NUMA nodes, ignoring the NUMA setup of the actual machine. But now you may configure your system for the node sizes of your choice: numa=fake=2*512,1024,2*256 gives two 512M nodes, one 1024M node, two 256M nodes, and the rest of system memory to a sixth node. The existing hash function is maintained to support the various node sizes that are possible with this implementation. Each node of the same size receives roughly the same amount of available pages, regardless of any reserved memory with its address range. The total available pages on the system is calculated and divided by the number of equal nodes to allocate. These nodes are then dynamically allocated and their borders extended until such time as their number of available pages reaches the required size. Configurable node sizes are recommended when used in conjunction with cpusets for memory control because it eliminates the overhead associated with scanning the zonelists of many smaller full nodes on page_alloc(). Cc: Andi Kleen <ak@suse.de> Signed-off-by: David Rientjes <rientjes@google.com> Signed-off-by: Andi Kleen <ak@suse.de> Cc: Paul Jackson <pj@sgi.com> Cc: Christoph Lameter <clameter@engr.sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2007-05-02 17:27:09 +00:00
*/
remove_all_active_ranges();
#ifdef CONFIG_ACPI_NUMA
acpi_numa = -1;
#endif
for_each_node_mask(i, node_possible_map) {
e820_register_active_regions(i, nodes[i].start >> PAGE_SHIFT,
nodes[i].end >> PAGE_SHIFT);
setup_node_bootmem(i, nodes[i].start, nodes[i].end);
}
acpi_fake_nodes(nodes, num_nodes);
numa_init_array();
return 0;
}
[PATCH] x86-64: configurable fake numa node sizes Extends the numa=fake x86_64 command-line option to allow for configurable node sizes. These nodes can be used in conjunction with cpusets for coarse memory resource management. The old command-line option is still supported: numa=fake=32 gives 32 fake NUMA nodes, ignoring the NUMA setup of the actual machine. But now you may configure your system for the node sizes of your choice: numa=fake=2*512,1024,2*256 gives two 512M nodes, one 1024M node, two 256M nodes, and the rest of system memory to a sixth node. The existing hash function is maintained to support the various node sizes that are possible with this implementation. Each node of the same size receives roughly the same amount of available pages, regardless of any reserved memory with its address range. The total available pages on the system is calculated and divided by the number of equal nodes to allocate. These nodes are then dynamically allocated and their borders extended until such time as their number of available pages reaches the required size. Configurable node sizes are recommended when used in conjunction with cpusets for memory control because it eliminates the overhead associated with scanning the zonelists of many smaller full nodes on page_alloc(). Cc: Andi Kleen <ak@suse.de> Signed-off-by: David Rientjes <rientjes@google.com> Signed-off-by: Andi Kleen <ak@suse.de> Cc: Paul Jackson <pj@sgi.com> Cc: Christoph Lameter <clameter@engr.sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2007-05-02 17:27:09 +00:00
#endif /* CONFIG_NUMA_EMU */
void __init numa_initmem_init(unsigned long start_pfn, unsigned long end_pfn)
{
int i;
nodes_clear(node_possible_map);
#ifdef CONFIG_NUMA_EMU
[PATCH] x86-64: configurable fake numa node sizes Extends the numa=fake x86_64 command-line option to allow for configurable node sizes. These nodes can be used in conjunction with cpusets for coarse memory resource management. The old command-line option is still supported: numa=fake=32 gives 32 fake NUMA nodes, ignoring the NUMA setup of the actual machine. But now you may configure your system for the node sizes of your choice: numa=fake=2*512,1024,2*256 gives two 512M nodes, one 1024M node, two 256M nodes, and the rest of system memory to a sixth node. The existing hash function is maintained to support the various node sizes that are possible with this implementation. Each node of the same size receives roughly the same amount of available pages, regardless of any reserved memory with its address range. The total available pages on the system is calculated and divided by the number of equal nodes to allocate. These nodes are then dynamically allocated and their borders extended until such time as their number of available pages reaches the required size. Configurable node sizes are recommended when used in conjunction with cpusets for memory control because it eliminates the overhead associated with scanning the zonelists of many smaller full nodes on page_alloc(). Cc: Andi Kleen <ak@suse.de> Signed-off-by: David Rientjes <rientjes@google.com> Signed-off-by: Andi Kleen <ak@suse.de> Cc: Paul Jackson <pj@sgi.com> Cc: Christoph Lameter <clameter@engr.sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2007-05-02 17:27:09 +00:00
if (cmdline && !numa_emulation(start_pfn, end_pfn))
return;
nodes_clear(node_possible_map);
#endif
#ifdef CONFIG_ACPI_NUMA
if (!numa_off && !acpi_scan_nodes(start_pfn << PAGE_SHIFT,
end_pfn << PAGE_SHIFT))
return;
nodes_clear(node_possible_map);
#endif
#ifdef CONFIG_K8_NUMA
if (!numa_off && !k8_scan_nodes(start_pfn<<PAGE_SHIFT, end_pfn<<PAGE_SHIFT))
return;
nodes_clear(node_possible_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,
end_pfn << PAGE_SHIFT);
/* setup dummy node covering all memory */
memnode_shift = 63;
memnodemap = memnode.embedded_map;
memnodemap[0] = 0;
nodes_clear(node_online_map);
node_set_online(0);
node_set(0, node_possible_map);
for (i = 0; i < NR_CPUS; i++)
numa_set_node(i, 0);
node_to_cpumask[0] = cpumask_of_cpu(0);
e820_register_active_regions(0, start_pfn, end_pfn);
setup_node_bootmem(0, start_pfn << PAGE_SHIFT, end_pfn << PAGE_SHIFT);
}
__cpuinit void numa_add_cpu(int cpu)
{
set_bit(cpu, &node_to_cpumask[cpu_to_node(cpu)]);
}
void __cpuinit numa_set_node(int cpu, int node)
{
cpu_pda(cpu)->nodenumber = node;
cpu_to_node[cpu] = node;
}
unsigned long __init numa_free_all_bootmem(void)
{
int i;
unsigned long pages = 0;
for_each_online_node(i) {
pages += free_all_bootmem_node(NODE_DATA(i));
}
return pages;
}
void __init paging_init(void)
{
int i;
unsigned long max_zone_pfns[MAX_NR_ZONES];
memset(max_zone_pfns, 0, sizeof(max_zone_pfns));
max_zone_pfns[ZONE_DMA] = MAX_DMA_PFN;
max_zone_pfns[ZONE_DMA32] = MAX_DMA32_PFN;
max_zone_pfns[ZONE_NORMAL] = end_pfn;
sparse_memory_present_with_active_regions(MAX_NUMNODES);
sparse_init();
for_each_online_node(i) {
setup_node_zones(i);
}
free_area_init_nodes(max_zone_pfns);
}
static __init int numa_setup(char *opt)
{
if (!opt)
return -EINVAL;
if (!strncmp(opt,"off",3))
numa_off = 1;
#ifdef CONFIG_NUMA_EMU
[PATCH] x86-64: configurable fake numa node sizes Extends the numa=fake x86_64 command-line option to allow for configurable node sizes. These nodes can be used in conjunction with cpusets for coarse memory resource management. The old command-line option is still supported: numa=fake=32 gives 32 fake NUMA nodes, ignoring the NUMA setup of the actual machine. But now you may configure your system for the node sizes of your choice: numa=fake=2*512,1024,2*256 gives two 512M nodes, one 1024M node, two 256M nodes, and the rest of system memory to a sixth node. The existing hash function is maintained to support the various node sizes that are possible with this implementation. Each node of the same size receives roughly the same amount of available pages, regardless of any reserved memory with its address range. The total available pages on the system is calculated and divided by the number of equal nodes to allocate. These nodes are then dynamically allocated and their borders extended until such time as their number of available pages reaches the required size. Configurable node sizes are recommended when used in conjunction with cpusets for memory control because it eliminates the overhead associated with scanning the zonelists of many smaller full nodes on page_alloc(). Cc: Andi Kleen <ak@suse.de> Signed-off-by: David Rientjes <rientjes@google.com> Signed-off-by: Andi Kleen <ak@suse.de> Cc: Paul Jackson <pj@sgi.com> Cc: Christoph Lameter <clameter@engr.sgi.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
2007-05-02 17:27:09 +00:00
if (!strncmp(opt, "fake=", 5))
cmdline = opt + 5;
#endif
#ifdef CONFIG_ACPI_NUMA
if (!strncmp(opt,"noacpi",6))
acpi_numa = -1;
[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 (!strncmp(opt,"hotadd=", 7))
hotadd_percent = simple_strtoul(opt+7, NULL, 10);
#endif
return 0;
}
early_param("numa", numa_setup);
/*
* 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.
*/
void __init init_cpu_to_node(void)
{
int i;
for (i = 0; i < NR_CPUS; i++) {
u8 apicid = x86_cpu_to_apicid[i];
if (apicid == BAD_APICID)
continue;
if (apicid_to_node[apicid] == NUMA_NO_NODE)
continue;
numa_set_node(i,apicid_to_node[apicid]);
}
}
EXPORT_SYMBOL(cpu_to_node);
EXPORT_SYMBOL(node_to_cpumask);
EXPORT_SYMBOL(memnode);
EXPORT_SYMBOL(node_data);
#ifdef CONFIG_DISCONTIGMEM
/*
* Functions to convert PFNs from/to per node page addresses.
* These are out of line because they are quite big.
* They could be all tuned by pre caching more state.
* Should do that.
*/
int pfn_valid(unsigned long pfn)
{
unsigned nid;
if (pfn >= num_physpages)
return 0;
nid = pfn_to_nid(pfn);
if (nid == 0xff)
return 0;
return pfn >= node_start_pfn(nid) && (pfn) < node_end_pfn(nid);
}
EXPORT_SYMBOL(pfn_valid);
#endif