Commit 208d54e5 authored by Dave Hansen's avatar Dave Hansen Committed by Linus Torvalds

[PATCH] memory hotplug locking: node_size_lock

pgdat->node_size_lock is basically only neeeded in one place in the normal
code: show_mem(), which is the arch-specific sysrq-m printing function.

Strictly speaking, the architectures not doing memory hotplug do no need this
locking in show_mem().  However, they are all included for completeness.  This
should also make any future consolidation of all of the implementations a
little more straightforward.

This lock is also held in the sparsemem code during a memory removal, as
sections are invalidated.  This is the place there pfn_valid() is made false
for a memory area that's being removed.  The lock is only required when doing
pfn_valid() operations on memory which the user does not already have a
reference on the page, such as in show_mem().
Signed-off-by: default avatarDave Hansen <haveblue@us.ibm.com>
Signed-off-by: default avatarAndrew Morton <akpm@osdl.org>
Signed-off-by: default avatarLinus Torvalds <torvalds@osdl.org>
parent c6a57e19
...@@ -371,6 +371,8 @@ show_mem(void) ...@@ -371,6 +371,8 @@ show_mem(void)
show_free_areas(); show_free_areas();
printk("Free swap: %6ldkB\n", nr_swap_pages<<(PAGE_SHIFT-10)); printk("Free swap: %6ldkB\n", nr_swap_pages<<(PAGE_SHIFT-10));
for_each_online_node(nid) { for_each_online_node(nid) {
unsigned long flags;
pgdat_resize_lock(NODE_DATA(nid), &flags);
i = node_spanned_pages(nid); i = node_spanned_pages(nid);
while (i-- > 0) { while (i-- > 0) {
struct page *page = nid_page_nr(nid, i); struct page *page = nid_page_nr(nid, i);
...@@ -384,6 +386,7 @@ show_mem(void) ...@@ -384,6 +386,7 @@ show_mem(void)
else else
shared += page_count(page) - 1; shared += page_count(page) - 1;
} }
pgdat_resize_unlock(NODE_DATA(nid), &flags);
} }
printk("%ld pages of RAM\n",total); printk("%ld pages of RAM\n",total);
printk("%ld free pages\n",free); printk("%ld free pages\n",free);
......
...@@ -31,11 +31,13 @@ void show_mem(void) ...@@ -31,11 +31,13 @@ void show_mem(void)
pg_data_t *pgdat; pg_data_t *pgdat;
unsigned long i; unsigned long i;
struct page_state ps; struct page_state ps;
unsigned long flags;
printk(KERN_INFO "Mem-info:\n"); printk(KERN_INFO "Mem-info:\n");
show_free_areas(); show_free_areas();
printk(KERN_INFO "Free swap: %6ldkB\n", nr_swap_pages<<(PAGE_SHIFT-10)); printk(KERN_INFO "Free swap: %6ldkB\n", nr_swap_pages<<(PAGE_SHIFT-10));
for_each_pgdat(pgdat) { for_each_pgdat(pgdat) {
pgdat_resize_lock(pgdat, &flags);
for (i = 0; i < pgdat->node_spanned_pages; ++i) { for (i = 0; i < pgdat->node_spanned_pages; ++i) {
page = pgdat_page_nr(pgdat, i); page = pgdat_page_nr(pgdat, i);
total++; total++;
...@@ -48,6 +50,7 @@ void show_mem(void) ...@@ -48,6 +50,7 @@ void show_mem(void)
else if (page_count(page)) else if (page_count(page))
shared += page_count(page) - 1; shared += page_count(page) - 1;
} }
pgdat_resize_unlock(pgdat, &flags);
} }
printk(KERN_INFO "%d pages of RAM\n", total); printk(KERN_INFO "%d pages of RAM\n", total);
printk(KERN_INFO "%d pages of HIGHMEM\n", highmem); printk(KERN_INFO "%d pages of HIGHMEM\n", highmem);
......
...@@ -555,9 +555,13 @@ void show_mem(void) ...@@ -555,9 +555,13 @@ void show_mem(void)
show_free_areas(); show_free_areas();
printk("Free swap: %6ldkB\n", nr_swap_pages<<(PAGE_SHIFT-10)); printk("Free swap: %6ldkB\n", nr_swap_pages<<(PAGE_SHIFT-10));
for_each_pgdat(pgdat) { for_each_pgdat(pgdat) {
unsigned long present = pgdat->node_present_pages; unsigned long present;
unsigned long flags;
int shared = 0, cached = 0, reserved = 0; int shared = 0, cached = 0, reserved = 0;
printk("Node ID: %d\n", pgdat->node_id); printk("Node ID: %d\n", pgdat->node_id);
pgdat_resize_lock(pgdat, &flags);
present = pgdat->node_present_pages;
for(i = 0; i < pgdat->node_spanned_pages; i++) { for(i = 0; i < pgdat->node_spanned_pages; i++) {
struct page *page; struct page *page;
if (pfn_valid(pgdat->node_start_pfn + i)) if (pfn_valid(pgdat->node_start_pfn + i))
...@@ -571,6 +575,7 @@ void show_mem(void) ...@@ -571,6 +575,7 @@ void show_mem(void)
else if (page_count(page)) else if (page_count(page))
shared += page_count(page)-1; shared += page_count(page)-1;
} }
pgdat_resize_unlock(pgdat, &flags);
total_present += present; total_present += present;
total_reserved += reserved; total_reserved += reserved;
total_cached += cached; total_cached += cached;
......
...@@ -48,6 +48,8 @@ void show_mem(void) ...@@ -48,6 +48,8 @@ void show_mem(void)
show_free_areas(); show_free_areas();
printk("Free swap: %6ldkB\n",nr_swap_pages<<(PAGE_SHIFT-10)); printk("Free swap: %6ldkB\n",nr_swap_pages<<(PAGE_SHIFT-10));
for_each_pgdat(pgdat) { for_each_pgdat(pgdat) {
unsigned long flags;
pgdat_resize_lock(pgdat, &flags);
for (i = 0; i < pgdat->node_spanned_pages; ++i) { for (i = 0; i < pgdat->node_spanned_pages; ++i) {
page = pgdat_page_nr(pgdat, i); page = pgdat_page_nr(pgdat, i);
total++; total++;
...@@ -60,6 +62,7 @@ void show_mem(void) ...@@ -60,6 +62,7 @@ void show_mem(void)
else if (page_count(page)) else if (page_count(page))
shared += page_count(page) - 1; shared += page_count(page) - 1;
} }
pgdat_resize_unlock(pgdat, &flags);
} }
printk("%d pages of RAM\n", total); printk("%d pages of RAM\n", total);
printk("%d pages of HIGHMEM\n",highmem); printk("%d pages of HIGHMEM\n",highmem);
...@@ -150,10 +153,14 @@ int __init reservedpages_count(void) ...@@ -150,10 +153,14 @@ int __init reservedpages_count(void)
int reservedpages, nid, i; int reservedpages, nid, i;
reservedpages = 0; reservedpages = 0;
for_each_online_node(nid) for_each_online_node(nid) {
unsigned long flags;
pgdat_resize_lock(NODE_DATA(nid), &flags);
for (i = 0 ; i < MAX_LOW_PFN(nid) - START_PFN(nid) ; i++) for (i = 0 ; i < MAX_LOW_PFN(nid) - START_PFN(nid) ; i++)
if (PageReserved(nid_page_nr(nid, i))) if (PageReserved(nid_page_nr(nid, i)))
reservedpages++; reservedpages++;
pgdat_resize_unlock(NODE_DATA(nid), &flags);
}
return reservedpages; return reservedpages;
} }
......
...@@ -505,7 +505,9 @@ void show_mem(void) ...@@ -505,7 +505,9 @@ void show_mem(void)
for (j = node_start_pfn(i); j < node_end_pfn(i); j++) { for (j = node_start_pfn(i); j < node_end_pfn(i); j++) {
struct page *p; struct page *p;
unsigned long flags;
pgdat_resize_lock(NODE_DATA(i), &flags);
p = nid_page_nr(i, j) - node_start_pfn(i); p = nid_page_nr(i, j) - node_start_pfn(i);
total++; total++;
...@@ -517,6 +519,7 @@ void show_mem(void) ...@@ -517,6 +519,7 @@ void show_mem(void)
free++; free++;
else else
shared += page_count(p) - 1; shared += page_count(p) - 1;
pgdat_resize_unlock(NODE_DATA(i), &flags);
} }
} }
#endif #endif
......
...@@ -104,6 +104,8 @@ void show_mem(void) ...@@ -104,6 +104,8 @@ void show_mem(void)
show_free_areas(); show_free_areas();
printk("Free swap: %6ldkB\n", nr_swap_pages<<(PAGE_SHIFT-10)); printk("Free swap: %6ldkB\n", nr_swap_pages<<(PAGE_SHIFT-10));
for_each_pgdat(pgdat) { for_each_pgdat(pgdat) {
unsigned long flags;
pgdat_resize_lock(pgdat, &flags);
for (i = 0; i < pgdat->node_spanned_pages; i++) { for (i = 0; i < pgdat->node_spanned_pages; i++) {
page = pgdat_page_nr(pgdat, i); page = pgdat_page_nr(pgdat, i);
total++; total++;
...@@ -114,6 +116,7 @@ void show_mem(void) ...@@ -114,6 +116,7 @@ void show_mem(void)
else if (page_count(page)) else if (page_count(page))
shared += page_count(page) - 1; shared += page_count(page) - 1;
} }
pgdat_resize_unlock(pgdat, &flags);
} }
printk("%ld pages of RAM\n", total); printk("%ld pages of RAM\n", total);
printk("%ld reserved pages\n", reserved); printk("%ld reserved pages\n", reserved);
...@@ -647,11 +650,14 @@ void __init mem_init(void) ...@@ -647,11 +650,14 @@ void __init mem_init(void)
#endif #endif
for_each_pgdat(pgdat) { for_each_pgdat(pgdat) {
unsigned long flags;
pgdat_resize_lock(pgdat, &flags);
for (i = 0; i < pgdat->node_spanned_pages; i++) { for (i = 0; i < pgdat->node_spanned_pages; i++) {
page = pgdat_page_nr(pgdat, i); page = pgdat_page_nr(pgdat, i);
if (PageReserved(page)) if (PageReserved(page))
reservedpages++; reservedpages++;
} }
pgdat_resize_unlock(pgdat, &flags);
} }
codesize = (unsigned long)&_etext - (unsigned long)&_stext; codesize = (unsigned long)&_etext - (unsigned long)&_stext;
......
#ifndef __LINUX_MEMORY_HOTPLUG_H
#define __LINUX_MEMORY_HOTPLUG_H
#include <linux/mmzone.h>
#include <linux/spinlock.h>
#ifdef CONFIG_MEMORY_HOTPLUG
/*
* pgdat resizing functions
*/
static inline
void pgdat_resize_lock(struct pglist_data *pgdat, unsigned long *flags)
{
spin_lock_irqsave(&pgdat->node_size_lock, *flags);
}
static inline
void pgdat_resize_unlock(struct pglist_data *pgdat, unsigned long *flags)
{
spin_lock_irqrestore(&pgdat->node_size_lock, *flags);
}
static inline
void pgdat_resize_init(struct pglist_data *pgdat)
{
spin_lock_init(&pgdat->node_size_lock);
}
#else /* ! CONFIG_MEMORY_HOTPLUG */
/*
* Stub functions for when hotplug is off
*/
static inline void pgdat_resize_lock(struct pglist_data *p, unsigned long *f) {}
static inline void pgdat_resize_unlock(struct pglist_data *p, unsigned long *f) {}
static inline void pgdat_resize_init(struct pglist_data *pgdat) {}
#endif
#endif /* __LINUX_MEMORY_HOTPLUG_H */
...@@ -273,6 +273,16 @@ typedef struct pglist_data { ...@@ -273,6 +273,16 @@ typedef struct pglist_data {
struct page *node_mem_map; struct page *node_mem_map;
#endif #endif
struct bootmem_data *bdata; struct bootmem_data *bdata;
#ifdef CONFIG_MEMORY_HOTPLUG
/*
* Must be held any time you expect node_start_pfn, node_present_pages
* or node_spanned_pages stay constant. Holding this will also
* guarantee that any pfn_valid() stays that way.
*
* Nests above zone->lock and zone->size_seqlock.
*/
spinlock_t node_size_lock;
#endif
unsigned long node_start_pfn; unsigned long node_start_pfn;
unsigned long node_present_pages; /* total number of physical pages */ unsigned long node_present_pages; /* total number of physical pages */
unsigned long node_spanned_pages; /* total size of physical page unsigned long node_spanned_pages; /* total size of physical page
...@@ -293,6 +303,8 @@ typedef struct pglist_data { ...@@ -293,6 +303,8 @@ typedef struct pglist_data {
#endif #endif
#define nid_page_nr(nid, pagenr) pgdat_page_nr(NODE_DATA(nid),(pagenr)) #define nid_page_nr(nid, pagenr) pgdat_page_nr(NODE_DATA(nid),(pagenr))
#include <linux/memory_hotplug.h>
extern struct pglist_data *pgdat_list; extern struct pglist_data *pgdat_list;
void __get_zone_counts(unsigned long *active, unsigned long *inactive, void __get_zone_counts(unsigned long *active, unsigned long *inactive,
......
...@@ -1958,6 +1958,7 @@ static void __init free_area_init_core(struct pglist_data *pgdat, ...@@ -1958,6 +1958,7 @@ static void __init free_area_init_core(struct pglist_data *pgdat,
int nid = pgdat->node_id; int nid = pgdat->node_id;
unsigned long zone_start_pfn = pgdat->node_start_pfn; unsigned long zone_start_pfn = pgdat->node_start_pfn;
pgdat_resize_init(pgdat);
pgdat->nr_zones = 0; pgdat->nr_zones = 0;
init_waitqueue_head(&pgdat->kswapd_wait); init_waitqueue_head(&pgdat->kswapd_wait);
pgdat->kswapd_max_order = 0; pgdat->kswapd_max_order = 0;
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment