aboutsummaryrefslogtreecommitdiff
path: root/arch/x86/mm/discontig_32.c
diff options
context:
space:
mode:
authorBenjamin Herrenschmidt <benh@kernel.crashing.org>2008-07-15 15:44:51 +1000
committerBenjamin Herrenschmidt <benh@kernel.crashing.org>2008-07-15 15:44:51 +1000
commit43d2548bb2ef7e6d753f91468a746784041e522d (patch)
tree77d13fcd48fd998393abb825ec36e2b732684a73 /arch/x86/mm/discontig_32.c
parent585583d95c5660973bc0cf64add517b040acd8a4 (diff)
parent85082fd7cbe3173198aac0eb5e85ab1edcc6352c (diff)
Merge commit '85082fd7cbe3173198aac0eb5e85ab1edcc6352c' into test-build
Manual fixup of: arch/powerpc/Kconfig
Diffstat (limited to 'arch/x86/mm/discontig_32.c')
-rw-r--r--arch/x86/mm/discontig_32.c285
1 files changed, 126 insertions, 159 deletions
diff --git a/arch/x86/mm/discontig_32.c b/arch/x86/mm/discontig_32.c
index 914ccf98368..5dfef9fa061 100644
--- a/arch/x86/mm/discontig_32.c
+++ b/arch/x86/mm/discontig_32.c
@@ -38,6 +38,7 @@
#include <asm/setup.h>
#include <asm/mmzone.h>
#include <asm/bios_ebda.h>
+#include <asm/proto.h>
struct pglist_data *node_data[MAX_NUMNODES] __read_mostly;
EXPORT_SYMBOL(node_data);
@@ -59,14 +60,14 @@ unsigned long node_end_pfn[MAX_NUMNODES] __read_mostly;
/*
* 4) physnode_map - the mapping between a pfn and owning node
* physnode_map keeps track of the physical memory layout of a generic
- * numa node on a 256Mb break (each element of the array will
- * represent 256Mb of memory and will be marked by the node id. so,
+ * numa node on a 64Mb break (each element of the array will
+ * represent 64Mb of memory and will be marked by the node id. so,
* if the first gig is on node 0, and the second gig is on node 1
* physnode_map will contain:
*
- * physnode_map[0-3] = 0;
- * physnode_map[4-7] = 1;
- * physnode_map[8- ] = -1;
+ * physnode_map[0-15] = 0;
+ * physnode_map[16-31] = 1;
+ * physnode_map[32- ] = -1;
*/
s8 physnode_map[MAX_ELEMENTS] __read_mostly = { [0 ... (MAX_ELEMENTS - 1)] = -1};
EXPORT_SYMBOL(physnode_map);
@@ -75,15 +76,15 @@ void memory_present(int nid, unsigned long start, unsigned long end)
{
unsigned long pfn;
- printk(KERN_INFO "Node: %d, start_pfn: %ld, end_pfn: %ld\n",
+ printk(KERN_INFO "Node: %d, start_pfn: %lx, end_pfn: %lx\n",
nid, start, end);
printk(KERN_DEBUG " Setting physnode_map array to node %d for pfns:\n", nid);
printk(KERN_DEBUG " ");
for (pfn = start; pfn < end; pfn += PAGES_PER_ELEMENT) {
physnode_map[pfn / PAGES_PER_ELEMENT] = nid;
- printk("%ld ", pfn);
+ printk(KERN_CONT "%lx ", pfn);
}
- printk("\n");
+ printk(KERN_CONT "\n");
}
unsigned long node_memmap_size_bytes(int nid, unsigned long start_pfn,
@@ -99,7 +100,6 @@ unsigned long node_memmap_size_bytes(int nid, unsigned long start_pfn,
#endif
extern unsigned long find_max_low_pfn(void);
-extern void add_one_highpage_init(struct page *, int, int);
extern unsigned long highend_pfn, highstart_pfn;
#define LARGE_PAGE_BYTES (PTRS_PER_PTE * PAGE_SIZE)
@@ -117,13 +117,13 @@ static unsigned long kva_pages;
*/
int __init get_memcfg_numa_flat(void)
{
- printk("NUMA - single node, flat memory mode\n");
+ printk(KERN_DEBUG "NUMA - single node, flat memory mode\n");
- /* Run the memory configuration and find the top of memory. */
- propagate_e820_map();
node_start_pfn[0] = 0;
node_end_pfn[0] = max_pfn;
+ e820_register_active_regions(0, 0, max_pfn);
memory_present(0, 0, max_pfn);
+ node_remap_size[0] = node_memmap_size_bytes(0, 0, max_pfn);
/* Indicate there is one node available. */
nodes_clear(node_online_map);
@@ -156,24 +156,32 @@ static void __init propagate_e820_map_node(int nid)
*/
static void __init allocate_pgdat(int nid)
{
- if (nid && node_has_online_mem(nid))
+ char buf[16];
+
+ if (node_has_online_mem(nid) && node_remap_start_vaddr[nid])
NODE_DATA(nid) = (pg_data_t *)node_remap_start_vaddr[nid];
else {
- NODE_DATA(nid) = (pg_data_t *)(pfn_to_kaddr(min_low_pfn));
- min_low_pfn += PFN_UP(sizeof(pg_data_t));
+ unsigned long pgdat_phys;
+ pgdat_phys = find_e820_area(min_low_pfn<<PAGE_SHIFT,
+ max_pfn_mapped<<PAGE_SHIFT,
+ sizeof(pg_data_t),
+ PAGE_SIZE);
+ NODE_DATA(nid) = (pg_data_t *)(pfn_to_kaddr(pgdat_phys>>PAGE_SHIFT));
+ memset(buf, 0, sizeof(buf));
+ sprintf(buf, "NODE_DATA %d", nid);
+ reserve_early(pgdat_phys, pgdat_phys + sizeof(pg_data_t), buf);
}
+ printk(KERN_DEBUG "allocate_pgdat: node %d NODE_DATA %08lx\n",
+ nid, (unsigned long)NODE_DATA(nid));
}
-#ifdef CONFIG_DISCONTIGMEM
/*
- * In the discontig memory model, a portion of the kernel virtual area (KVA)
- * is reserved and portions of nodes are mapped using it. This is to allow
- * node-local memory to be allocated for structures that would normally require
- * ZONE_NORMAL. The memory is allocated with alloc_remap() and callers
- * should be prepared to allocate from the bootmem allocator instead. This KVA
- * mechanism is incompatible with SPARSEMEM as it makes assumptions about the
- * layout of memory that are broken if alloc_remap() succeeds for some of the
- * map and fails for others
+ * In the DISCONTIGMEM and SPARSEMEM memory model, a portion of the kernel
+ * virtual address space (KVA) is reserved and portions of nodes are mapped
+ * using it. This is to allow node-local memory to be allocated for
+ * structures that would normally require ZONE_NORMAL. The memory is
+ * allocated with alloc_remap() and callers should be prepared to allocate
+ * from the bootmem allocator instead.
*/
static unsigned long node_remap_start_pfn[MAX_NUMNODES];
static void *node_remap_end_vaddr[MAX_NUMNODES];
@@ -195,15 +203,19 @@ void *alloc_remap(int nid, unsigned long size)
return allocation;
}
-void __init remap_numa_kva(void)
+static void __init remap_numa_kva(void)
{
void *vaddr;
unsigned long pfn;
int node;
for_each_online_node(node) {
+ printk(KERN_DEBUG "remap_numa_kva: node %d\n", node);
for (pfn=0; pfn < node_remap_size[node]; pfn += PTRS_PER_PTE) {
vaddr = node_remap_start_vaddr[node]+(pfn<<PAGE_SHIFT);
+ printk(KERN_DEBUG "remap_numa_kva: %08lx to pfn %08lx\n",
+ (unsigned long)vaddr,
+ node_remap_start_pfn[node] + pfn);
set_pmd_pfn((ulong) vaddr,
node_remap_start_pfn[node] + pfn,
PAGE_KERNEL_LARGE);
@@ -215,17 +227,21 @@ static unsigned long calculate_numa_remap_pages(void)
{
int nid;
unsigned long size, reserve_pages = 0;
- unsigned long pfn;
for_each_online_node(nid) {
- unsigned old_end_pfn = node_end_pfn[nid];
+ u64 node_kva_target;
+ u64 node_kva_final;
/*
* The acpi/srat node info can show hot-add memroy zones
* where memory could be added but not currently present.
*/
+ printk(KERN_DEBUG "node %d pfn: [%lx - %lx]\n",
+ nid, node_start_pfn[nid], node_end_pfn[nid]);
if (node_start_pfn[nid] > max_pfn)
continue;
+ if (!node_end_pfn[nid])
+ continue;
if (node_end_pfn[nid] > max_pfn)
node_end_pfn[nid] = max_pfn;
@@ -237,41 +253,48 @@ static unsigned long calculate_numa_remap_pages(void)
/* now the roundup is correct, convert to PAGE_SIZE pages */
size = size * PTRS_PER_PTE;
- /*
- * Validate the region we are allocating only contains valid
- * pages.
- */
- for (pfn = node_end_pfn[nid] - size;
- pfn < node_end_pfn[nid]; pfn++)
- if (!page_is_ram(pfn))
- break;
-
- if (pfn != node_end_pfn[nid])
- size = 0;
+ node_kva_target = round_down(node_end_pfn[nid] - size,
+ PTRS_PER_PTE);
+ node_kva_target <<= PAGE_SHIFT;
+ do {
+ node_kva_final = find_e820_area(node_kva_target,
+ ((u64)node_end_pfn[nid])<<PAGE_SHIFT,
+ ((u64)size)<<PAGE_SHIFT,
+ LARGE_PAGE_BYTES);
+ node_kva_target -= LARGE_PAGE_BYTES;
+ } while (node_kva_final == -1ULL &&
+ (node_kva_target>>PAGE_SHIFT) > (node_start_pfn[nid]));
+
+ if (node_kva_final == -1ULL)
+ panic("Can not get kva ram\n");
- printk("Reserving %ld pages of KVA for lmem_map of node %d\n",
- size, nid);
node_remap_size[nid] = size;
node_remap_offset[nid] = reserve_pages;
reserve_pages += size;
- printk("Shrinking node %d from %ld pages to %ld pages\n",
- nid, node_end_pfn[nid], node_end_pfn[nid] - size);
-
- if (node_end_pfn[nid] & (PTRS_PER_PTE-1)) {
- /*
- * Align node_end_pfn[] and node_remap_start_pfn[] to
- * pmd boundary. remap_numa_kva will barf otherwise.
- */
- printk("Shrinking node %d further by %ld pages for proper alignment\n",
- nid, node_end_pfn[nid] & (PTRS_PER_PTE-1));
- size += node_end_pfn[nid] & (PTRS_PER_PTE-1);
- }
+ printk(KERN_DEBUG "Reserving %ld pages of KVA for lmem_map of"
+ " node %d at %llx\n",
+ size, nid, node_kva_final>>PAGE_SHIFT);
+
+ /*
+ * prevent kva address below max_low_pfn want it on system
+ * with less memory later.
+ * layout will be: KVA address , KVA RAM
+ *
+ * we are supposed to only record the one less then max_low_pfn
+ * but we could have some hole in high memory, and it will only
+ * check page_is_ram(pfn) && !page_is_reserved_early(pfn) to decide
+ * to use it as free.
+ * So reserve_early here, hope we don't run out of that array
+ */
+ reserve_early(node_kva_final,
+ node_kva_final+(((u64)size)<<PAGE_SHIFT),
+ "KVA RAM");
- node_end_pfn[nid] -= size;
- node_remap_start_pfn[nid] = node_end_pfn[nid];
- shrink_active_range(nid, old_end_pfn, node_end_pfn[nid]);
+ node_remap_start_pfn[nid] = node_kva_final>>PAGE_SHIFT;
+ remove_active_range(nid, node_remap_start_pfn[nid],
+ node_remap_start_pfn[nid] + size);
}
- printk("Reserving total of %ld pages for numa KVA remap\n",
+ printk(KERN_INFO "Reserving total of %lx pages for numa KVA remap\n",
reserve_pages);
return reserve_pages;
}
@@ -285,37 +308,16 @@ static void init_remap_allocator(int nid)
node_remap_alloc_vaddr[nid] = node_remap_start_vaddr[nid] +
ALIGN(sizeof(pg_data_t), PAGE_SIZE);
- printk ("node %d will remap to vaddr %08lx - %08lx\n", nid,
+ printk(KERN_DEBUG "node %d will remap to vaddr %08lx - %08lx\n", nid,
(ulong) node_remap_start_vaddr[nid],
- (ulong) pfn_to_kaddr(highstart_pfn
- + node_remap_offset[nid] + node_remap_size[nid]));
-}
-#else
-void *alloc_remap(int nid, unsigned long size)
-{
- return NULL;
-}
-
-static unsigned long calculate_numa_remap_pages(void)
-{
- return 0;
-}
-
-static void init_remap_allocator(int nid)
-{
-}
-
-void __init remap_numa_kva(void)
-{
+ (ulong) node_remap_end_vaddr[nid]);
}
-#endif /* CONFIG_DISCONTIGMEM */
-extern void setup_bootmem_allocator(void);
-unsigned long __init setup_memory(void)
+void __init initmem_init(unsigned long start_pfn,
+ unsigned long end_pfn)
{
int nid;
- unsigned long system_start_pfn, system_max_low_pfn;
- unsigned long wasted_pages;
+ long kva_target_pfn;
/*
* When mapping a NUMA machine we allocate the node_mem_map arrays
@@ -324,109 +326,77 @@ unsigned long __init setup_memory(void)
* this space and use it to adjust the boundary between ZONE_NORMAL
* and ZONE_HIGHMEM.
*/
- get_memcfg_numa();
- kva_pages = calculate_numa_remap_pages();
+ get_memcfg_numa();
- /* partially used pages are not usable - thus round upwards */
- system_start_pfn = min_low_pfn = PFN_UP(init_pg_tables_end);
+ kva_pages = round_up(calculate_numa_remap_pages(), PTRS_PER_PTE);
- kva_start_pfn = find_max_low_pfn() - kva_pages;
+ kva_target_pfn = round_down(max_low_pfn - kva_pages, PTRS_PER_PTE);
+ do {
+ kva_start_pfn = find_e820_area(kva_target_pfn<<PAGE_SHIFT,
+ max_low_pfn<<PAGE_SHIFT,
+ kva_pages<<PAGE_SHIFT,
+ PTRS_PER_PTE<<PAGE_SHIFT) >> PAGE_SHIFT;
+ kva_target_pfn -= PTRS_PER_PTE;
+ } while (kva_start_pfn == -1UL && kva_target_pfn > min_low_pfn);
-#ifdef CONFIG_BLK_DEV_INITRD
- /* Numa kva area is below the initrd */
- if (initrd_start)
- kva_start_pfn = PFN_DOWN(initrd_start - PAGE_OFFSET)
- - kva_pages;
-#endif
+ if (kva_start_pfn == -1UL)
+ panic("Can not get kva space\n");
- /*
- * We waste pages past at the end of the KVA for no good reason other
- * than how it is located. This is bad.
- */
- wasted_pages = kva_start_pfn & (PTRS_PER_PTE-1);
- kva_start_pfn -= wasted_pages;
- kva_pages += wasted_pages;
-
- system_max_low_pfn = max_low_pfn = find_max_low_pfn();
- printk("kva_start_pfn ~ %ld find_max_low_pfn() ~ %ld\n",
+ printk(KERN_INFO "kva_start_pfn ~ %lx max_low_pfn ~ %lx\n",
kva_start_pfn, max_low_pfn);
- printk("max_pfn = %ld\n", max_pfn);
+ printk(KERN_INFO "max_pfn = %lx\n", max_pfn);
+
+ /* avoid clash with initrd */
+ reserve_early(kva_start_pfn<<PAGE_SHIFT,
+ (kva_start_pfn + kva_pages)<<PAGE_SHIFT,
+ "KVA PG");
#ifdef CONFIG_HIGHMEM
highstart_pfn = highend_pfn = max_pfn;
- if (max_pfn > system_max_low_pfn)
- highstart_pfn = system_max_low_pfn;
+ if (max_pfn > max_low_pfn)
+ highstart_pfn = max_low_pfn;
printk(KERN_NOTICE "%ldMB HIGHMEM available.\n",
pages_to_mb(highend_pfn - highstart_pfn));
num_physpages = highend_pfn;
high_memory = (void *) __va(highstart_pfn * PAGE_SIZE - 1) + 1;
#else
- num_physpages = system_max_low_pfn;
- high_memory = (void *) __va(system_max_low_pfn * PAGE_SIZE - 1) + 1;
+ num_physpages = max_low_pfn;
+ high_memory = (void *) __va(max_low_pfn * PAGE_SIZE - 1) + 1;
#endif
printk(KERN_NOTICE "%ldMB LOWMEM available.\n",
- pages_to_mb(system_max_low_pfn));
- printk("min_low_pfn = %ld, max_low_pfn = %ld, highstart_pfn = %ld\n",
- min_low_pfn, max_low_pfn, highstart_pfn);
+ pages_to_mb(max_low_pfn));
+ printk(KERN_DEBUG "max_low_pfn = %lx, highstart_pfn = %lx\n",
+ max_low_pfn, highstart_pfn);
- printk("Low memory ends at vaddr %08lx\n",
+ printk(KERN_DEBUG "Low memory ends at vaddr %08lx\n",
(ulong) pfn_to_kaddr(max_low_pfn));
for_each_online_node(nid) {
init_remap_allocator(nid);
allocate_pgdat(nid);
}
- printk("High memory starts at vaddr %08lx\n",
+ remap_numa_kva();
+
+ printk(KERN_DEBUG "High memory starts at vaddr %08lx\n",
(ulong) pfn_to_kaddr(highstart_pfn));
for_each_online_node(nid)
propagate_e820_map_node(nid);
- memset(NODE_DATA(0), 0, sizeof(struct pglist_data));
+ for_each_online_node(nid)
+ memset(NODE_DATA(nid), 0, sizeof(struct pglist_data));
+
NODE_DATA(0)->bdata = &node0_bdata;
setup_bootmem_allocator();
- return max_low_pfn;
-}
-
-void __init numa_kva_reserve(void)
-{
- if (kva_pages)
- reserve_bootmem(PFN_PHYS(kva_start_pfn), PFN_PHYS(kva_pages),
- BOOTMEM_DEFAULT);
}
-void __init zone_sizes_init(void)
-{
- int nid;
- unsigned long max_zone_pfns[MAX_NR_ZONES];
- memset(max_zone_pfns, 0, sizeof(max_zone_pfns));
- max_zone_pfns[ZONE_DMA] =
- virt_to_phys((char *)MAX_DMA_ADDRESS) >> PAGE_SHIFT;
- max_zone_pfns[ZONE_NORMAL] = max_low_pfn;
-#ifdef CONFIG_HIGHMEM
- max_zone_pfns[ZONE_HIGHMEM] = highend_pfn;
-#endif
-
- /* If SRAT has not registered memory, register it now */
- if (find_max_pfn_with_active_regions() == 0) {
- for_each_online_node(nid) {
- if (node_has_online_mem(nid))
- add_active_range(nid, node_start_pfn[nid],
- node_end_pfn[nid]);
- }
- }
-
- free_area_init_nodes(max_zone_pfns);
- return;
-}
-
-void __init set_highmem_pages_init(int bad_ppro)
+void __init set_highmem_pages_init(void)
{
#ifdef CONFIG_HIGHMEM
struct zone *zone;
- struct page *page;
+ int nid;
for_each_zone(zone) {
- unsigned long node_pfn, zone_start_pfn, zone_end_pfn;
+ unsigned long zone_start_pfn, zone_end_pfn;
if (!is_highmem(zone))
continue;
@@ -434,16 +404,12 @@ void __init set_highmem_pages_init(int bad_ppro)
zone_start_pfn = zone->zone_start_pfn;
zone_end_pfn = zone_start_pfn + zone->spanned_pages;
- printk("Initializing %s for node %d (%08lx:%08lx)\n",
- zone->name, zone_to_nid(zone),
- zone_start_pfn, zone_end_pfn);
+ nid = zone_to_nid(zone);
+ printk(KERN_INFO "Initializing %s for node %d (%08lx:%08lx)\n",
+ zone->name, nid, zone_start_pfn, zone_end_pfn);
- for (node_pfn = zone_start_pfn; node_pfn < zone_end_pfn; node_pfn++) {
- if (!pfn_valid(node_pfn))
- continue;
- page = pfn_to_page(node_pfn);
- add_one_highpage_init(page, node_pfn, bad_ppro);
- }
+ add_highpages_with_active_regions(nid, zone_start_pfn,
+ zone_end_pfn);
}
totalram_pages += totalhigh_pages;
#endif
@@ -476,3 +442,4 @@ int memory_add_physaddr_to_nid(u64 addr)
EXPORT_SYMBOL_GPL(memory_add_physaddr_to_nid);
#endif
+