2005-04-16 18:20:36 -04:00
|
|
|
/*
|
|
|
|
* linux/arch/i386/mm/pgtable.c
|
|
|
|
*/
|
|
|
|
|
|
|
|
#include <linux/sched.h>
|
|
|
|
#include <linux/kernel.h>
|
|
|
|
#include <linux/errno.h>
|
|
|
|
#include <linux/mm.h>
|
|
|
|
#include <linux/swap.h>
|
|
|
|
#include <linux/smp.h>
|
|
|
|
#include <linux/highmem.h>
|
|
|
|
#include <linux/slab.h>
|
|
|
|
#include <linux/pagemap.h>
|
|
|
|
#include <linux/spinlock.h>
|
2006-09-26 02:32:25 -04:00
|
|
|
#include <linux/module.h>
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
#include <asm/system.h>
|
|
|
|
#include <asm/pgtable.h>
|
|
|
|
#include <asm/pgalloc.h>
|
|
|
|
#include <asm/fixmap.h>
|
|
|
|
#include <asm/e820.h>
|
|
|
|
#include <asm/tlb.h>
|
|
|
|
#include <asm/tlbflush.h>
|
|
|
|
|
|
|
|
void show_mem(void)
|
|
|
|
{
|
|
|
|
int total = 0, reserved = 0;
|
|
|
|
int shared = 0, cached = 0;
|
|
|
|
int highmem = 0;
|
|
|
|
struct page *page;
|
|
|
|
pg_data_t *pgdat;
|
|
|
|
unsigned long i;
|
2005-10-29 21:16:52 -04:00
|
|
|
unsigned long flags;
|
2005-04-16 18:20:36 -04:00
|
|
|
|
2005-06-25 17:59:24 -04:00
|
|
|
printk(KERN_INFO "Mem-info:\n");
|
2005-04-16 18:20:36 -04:00
|
|
|
show_free_areas();
|
2005-06-25 17:59:24 -04:00
|
|
|
printk(KERN_INFO "Free swap: %6ldkB\n", nr_swap_pages<<(PAGE_SHIFT-10));
|
2006-03-27 04:15:59 -05:00
|
|
|
for_each_online_pgdat(pgdat) {
|
2005-10-29 21:16:52 -04:00
|
|
|
pgdat_resize_lock(pgdat, &flags);
|
2005-04-16 18:20:36 -04:00
|
|
|
for (i = 0; i < pgdat->node_spanned_pages; ++i) {
|
[PATCH] remove non-DISCONTIG use of pgdat->node_mem_map
This patch effectively eliminates direct use of pgdat->node_mem_map outside
of the DISCONTIG code. On a flat memory system, these fields aren't
currently used, neither are they on a sparsemem system.
There was also a node_mem_map(nid) macro on many architectures. Its use
along with the use of ->node_mem_map itself was not consistent. It has
been removed in favor of two new, more explicit, arch-independent macros:
pgdat_page_nr(pgdat, pagenr)
nid_page_nr(nid, pagenr)
I called them "pgdat" and "nid" because we overload the term "node" to mean
"NUMA node", "DISCONTIG node" or "pg_data_t" in very confusing ways. I
believe the newer names are much clearer.
These macros can be overridden in the sparsemem case with a theoretically
slower operation using node_start_pfn and pfn_to_page(), instead. We could
make this the only behavior if people want, but I don't want to change too
much at once. One thing at a time.
This patch removes more code than it adds.
Compile tested on alpha, alpha discontig, arm, arm-discontig, i386, i386
generic, NUMAQ, Summit, ppc64, ppc64 discontig, and x86_64. Full list
here: http://sr71.net/patches/2.6.12/2.6.12-rc1-mhp2/configs/
Boot tested on NUMAQ, x86 SMP and ppc64 power4/5 LPARs.
Signed-off-by: Dave Hansen <haveblue@us.ibm.com>
Signed-off-by: Martin J. Bligh <mbligh@aracnet.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-06-23 03:07:37 -04:00
|
|
|
page = pgdat_page_nr(pgdat, i);
|
2005-04-16 18:20:36 -04:00
|
|
|
total++;
|
|
|
|
if (PageHighMem(page))
|
|
|
|
highmem++;
|
|
|
|
if (PageReserved(page))
|
|
|
|
reserved++;
|
|
|
|
else if (PageSwapCache(page))
|
|
|
|
cached++;
|
|
|
|
else if (page_count(page))
|
|
|
|
shared += page_count(page) - 1;
|
|
|
|
}
|
2005-10-29 21:16:52 -04:00
|
|
|
pgdat_resize_unlock(pgdat, &flags);
|
2005-04-16 18:20:36 -04:00
|
|
|
}
|
2005-06-25 17:59:24 -04:00
|
|
|
printk(KERN_INFO "%d pages of RAM\n", total);
|
|
|
|
printk(KERN_INFO "%d pages of HIGHMEM\n", highmem);
|
|
|
|
printk(KERN_INFO "%d reserved pages\n", reserved);
|
|
|
|
printk(KERN_INFO "%d pages shared\n", shared);
|
|
|
|
printk(KERN_INFO "%d pages swap cached\n", cached);
|
2005-06-23 03:08:08 -04:00
|
|
|
|
2006-06-30 04:55:39 -04:00
|
|
|
printk(KERN_INFO "%lu pages dirty\n", global_page_state(NR_FILE_DIRTY));
|
2006-06-30 04:55:40 -04:00
|
|
|
printk(KERN_INFO "%lu pages writeback\n",
|
|
|
|
global_page_state(NR_WRITEBACK));
|
2006-06-30 04:55:34 -04:00
|
|
|
printk(KERN_INFO "%lu pages mapped\n", global_page_state(NR_FILE_MAPPED));
|
2006-09-26 02:31:51 -04:00
|
|
|
printk(KERN_INFO "%lu pages slab\n",
|
|
|
|
global_page_state(NR_SLAB_RECLAIMABLE) +
|
|
|
|
global_page_state(NR_SLAB_UNRECLAIMABLE));
|
2006-06-30 04:55:38 -04:00
|
|
|
printk(KERN_INFO "%lu pages pagetables\n",
|
|
|
|
global_page_state(NR_PAGETABLE));
|
2005-04-16 18:20:36 -04:00
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Associate a virtual page frame with a given physical page frame
|
|
|
|
* and protection flags for that frame.
|
|
|
|
*/
|
|
|
|
static void set_pte_pfn(unsigned long vaddr, unsigned long pfn, pgprot_t flags)
|
|
|
|
{
|
|
|
|
pgd_t *pgd;
|
|
|
|
pud_t *pud;
|
|
|
|
pmd_t *pmd;
|
|
|
|
pte_t *pte;
|
|
|
|
|
|
|
|
pgd = swapper_pg_dir + pgd_index(vaddr);
|
|
|
|
if (pgd_none(*pgd)) {
|
|
|
|
BUG();
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
pud = pud_offset(pgd, vaddr);
|
|
|
|
if (pud_none(*pud)) {
|
|
|
|
BUG();
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
pmd = pmd_offset(pud, vaddr);
|
|
|
|
if (pmd_none(*pmd)) {
|
|
|
|
BUG();
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
pte = pte_offset_kernel(pmd, vaddr);
|
2006-12-06 20:14:09 -05:00
|
|
|
if (pgprot_val(flags))
|
|
|
|
/* <pfn,flags> stored as-is, to permit clearing entries */
|
|
|
|
set_pte(pte, pfn_pte(pfn, flags));
|
|
|
|
else
|
|
|
|
pte_clear(&init_mm, vaddr, pte);
|
2005-04-16 18:20:36 -04:00
|
|
|
|
|
|
|
/*
|
|
|
|
* It's enough to flush this one mapping.
|
|
|
|
* (PGE mappings get flushed as well)
|
|
|
|
*/
|
|
|
|
__flush_tlb_one(vaddr);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Associate a large virtual page frame with a given physical page frame
|
|
|
|
* and protection flags for that frame. pfn is for the base of the page,
|
|
|
|
* vaddr is what the page gets mapped to - both must be properly aligned.
|
|
|
|
* The pmd must already be instantiated. Assumes PAE mode.
|
|
|
|
*/
|
|
|
|
void set_pmd_pfn(unsigned long vaddr, unsigned long pfn, pgprot_t flags)
|
|
|
|
{
|
|
|
|
pgd_t *pgd;
|
|
|
|
pud_t *pud;
|
|
|
|
pmd_t *pmd;
|
|
|
|
|
|
|
|
if (vaddr & (PMD_SIZE-1)) { /* vaddr is misaligned */
|
2005-06-25 17:59:24 -04:00
|
|
|
printk(KERN_WARNING "set_pmd_pfn: vaddr misaligned\n");
|
2005-04-16 18:20:36 -04:00
|
|
|
return; /* BUG(); */
|
|
|
|
}
|
|
|
|
if (pfn & (PTRS_PER_PTE-1)) { /* pfn is misaligned */
|
2005-06-25 17:59:24 -04:00
|
|
|
printk(KERN_WARNING "set_pmd_pfn: pfn misaligned\n");
|
2005-04-16 18:20:36 -04:00
|
|
|
return; /* BUG(); */
|
|
|
|
}
|
|
|
|
pgd = swapper_pg_dir + pgd_index(vaddr);
|
|
|
|
if (pgd_none(*pgd)) {
|
2005-06-25 17:59:24 -04:00
|
|
|
printk(KERN_WARNING "set_pmd_pfn: pgd_none\n");
|
2005-04-16 18:20:36 -04:00
|
|
|
return; /* BUG(); */
|
|
|
|
}
|
|
|
|
pud = pud_offset(pgd, vaddr);
|
|
|
|
pmd = pmd_offset(pud, vaddr);
|
|
|
|
set_pmd(pmd, pfn_pmd(pfn, flags));
|
|
|
|
/*
|
|
|
|
* It's enough to flush this one mapping.
|
|
|
|
* (PGE mappings get flushed as well)
|
|
|
|
*/
|
|
|
|
__flush_tlb_one(vaddr);
|
|
|
|
}
|
|
|
|
|
2006-09-26 02:32:25 -04:00
|
|
|
static int fixmaps;
|
|
|
|
#ifndef CONFIG_COMPAT_VDSO
|
|
|
|
unsigned long __FIXADDR_TOP = 0xfffff000;
|
|
|
|
EXPORT_SYMBOL(__FIXADDR_TOP);
|
|
|
|
#endif
|
|
|
|
|
2005-04-16 18:20:36 -04:00
|
|
|
void __set_fixmap (enum fixed_addresses idx, unsigned long phys, pgprot_t flags)
|
|
|
|
{
|
|
|
|
unsigned long address = __fix_to_virt(idx);
|
|
|
|
|
|
|
|
if (idx >= __end_of_fixed_addresses) {
|
|
|
|
BUG();
|
|
|
|
return;
|
|
|
|
}
|
|
|
|
set_pte_pfn(address, phys >> PAGE_SHIFT, flags);
|
2006-09-26 02:32:25 -04:00
|
|
|
fixmaps++;
|
|
|
|
}
|
|
|
|
|
|
|
|
/**
|
|
|
|
* reserve_top_address - reserves a hole in the top of kernel address space
|
|
|
|
* @reserve - size of hole to reserve
|
|
|
|
*
|
|
|
|
* Can be used to relocate the fixmap area and poke a hole in the top
|
|
|
|
* of kernel address space to make room for a hypervisor.
|
|
|
|
*/
|
|
|
|
void reserve_top_address(unsigned long reserve)
|
|
|
|
{
|
|
|
|
BUG_ON(fixmaps > 0);
|
2007-02-13 07:26:21 -05:00
|
|
|
printk(KERN_INFO "Reserving virtual address space above 0x%08x\n",
|
|
|
|
(int)-reserve);
|
2006-09-26 02:32:25 -04:00
|
|
|
#ifdef CONFIG_COMPAT_VDSO
|
|
|
|
BUG_ON(reserve != 0);
|
|
|
|
#else
|
|
|
|
__FIXADDR_TOP = -reserve - PAGE_SIZE;
|
|
|
|
__VMALLOC_RESERVE += reserve;
|
|
|
|
#endif
|
2005-04-16 18:20:36 -04:00
|
|
|
}
|
|
|
|
|
|
|
|
pte_t *pte_alloc_one_kernel(struct mm_struct *mm, unsigned long address)
|
|
|
|
{
|
|
|
|
return (pte_t *)__get_free_page(GFP_KERNEL|__GFP_REPEAT|__GFP_ZERO);
|
|
|
|
}
|
|
|
|
|
|
|
|
struct page *pte_alloc_one(struct mm_struct *mm, unsigned long address)
|
|
|
|
{
|
|
|
|
struct page *pte;
|
|
|
|
|
|
|
|
#ifdef CONFIG_HIGHPTE
|
|
|
|
pte = alloc_pages(GFP_KERNEL|__GFP_HIGHMEM|__GFP_REPEAT|__GFP_ZERO, 0);
|
|
|
|
#else
|
|
|
|
pte = alloc_pages(GFP_KERNEL|__GFP_REPEAT|__GFP_ZERO, 0);
|
|
|
|
#endif
|
|
|
|
return pte;
|
|
|
|
}
|
|
|
|
|
2006-12-06 23:33:20 -05:00
|
|
|
void pmd_ctor(void *pmd, struct kmem_cache *cache, unsigned long flags)
|
2005-04-16 18:20:36 -04:00
|
|
|
{
|
|
|
|
memset(pmd, 0, PTRS_PER_PMD*sizeof(pmd_t));
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* List of all pgd's needed for non-PAE so it can invalidate entries
|
|
|
|
* in both cached and uncached pgd's; not needed for PAE since the
|
|
|
|
* kernel pmd is shared. If PAE were not to share the pmd a similar
|
|
|
|
* tactic would be needed. This is essentially codepath-based locking
|
|
|
|
* against pageattr.c; it is the unique case in which a valid change
|
|
|
|
* of kernel pagetables can't be lazily synchronized by vmalloc faults.
|
|
|
|
* vmalloc faults work because attached pagetables are never freed.
|
|
|
|
* The locking scheme was chosen on the basis of manfred's
|
|
|
|
* recommendations and having no core impact whatsoever.
|
|
|
|
* -- wli
|
|
|
|
*/
|
|
|
|
DEFINE_SPINLOCK(pgd_lock);
|
|
|
|
struct page *pgd_list;
|
|
|
|
|
|
|
|
static inline void pgd_list_add(pgd_t *pgd)
|
|
|
|
{
|
|
|
|
struct page *page = virt_to_page(pgd);
|
|
|
|
page->index = (unsigned long)pgd_list;
|
|
|
|
if (pgd_list)
|
[PATCH] mm: split page table lock
Christoph Lameter demonstrated very poor scalability on the SGI 512-way, with
a many-threaded application which concurrently initializes different parts of
a large anonymous area.
This patch corrects that, by using a separate spinlock per page table page, to
guard the page table entries in that page, instead of using the mm's single
page_table_lock. (But even then, page_table_lock is still used to guard page
table allocation, and anon_vma allocation.)
In this implementation, the spinlock is tucked inside the struct page of the
page table page: with a BUILD_BUG_ON in case it overflows - which it would in
the case of 32-bit PA-RISC with spinlock debugging enabled.
Splitting the lock is not quite for free: another cacheline access. Ideally,
I suppose we would use split ptlock only for multi-threaded processes on
multi-cpu machines; but deciding that dynamically would have its own costs.
So for now enable it by config, at some number of cpus - since the Kconfig
language doesn't support inequalities, let preprocessor compare that with
NR_CPUS. But I don't think it's worth being user-configurable: for good
testing of both split and unsplit configs, split now at 4 cpus, and perhaps
change that to 8 later.
There is a benefit even for singly threaded processes: kswapd can be attacking
one part of the mm while another part is busy faulting.
Signed-off-by: Hugh Dickins <hugh@veritas.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-10-29 21:16:40 -04:00
|
|
|
set_page_private(pgd_list, (unsigned long)&page->index);
|
2005-04-16 18:20:36 -04:00
|
|
|
pgd_list = page;
|
[PATCH] mm: split page table lock
Christoph Lameter demonstrated very poor scalability on the SGI 512-way, with
a many-threaded application which concurrently initializes different parts of
a large anonymous area.
This patch corrects that, by using a separate spinlock per page table page, to
guard the page table entries in that page, instead of using the mm's single
page_table_lock. (But even then, page_table_lock is still used to guard page
table allocation, and anon_vma allocation.)
In this implementation, the spinlock is tucked inside the struct page of the
page table page: with a BUILD_BUG_ON in case it overflows - which it would in
the case of 32-bit PA-RISC with spinlock debugging enabled.
Splitting the lock is not quite for free: another cacheline access. Ideally,
I suppose we would use split ptlock only for multi-threaded processes on
multi-cpu machines; but deciding that dynamically would have its own costs.
So for now enable it by config, at some number of cpus - since the Kconfig
language doesn't support inequalities, let preprocessor compare that with
NR_CPUS. But I don't think it's worth being user-configurable: for good
testing of both split and unsplit configs, split now at 4 cpus, and perhaps
change that to 8 later.
There is a benefit even for singly threaded processes: kswapd can be attacking
one part of the mm while another part is busy faulting.
Signed-off-by: Hugh Dickins <hugh@veritas.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-10-29 21:16:40 -04:00
|
|
|
set_page_private(page, (unsigned long)&pgd_list);
|
2005-04-16 18:20:36 -04:00
|
|
|
}
|
|
|
|
|
|
|
|
static inline void pgd_list_del(pgd_t *pgd)
|
|
|
|
{
|
|
|
|
struct page *next, **pprev, *page = virt_to_page(pgd);
|
|
|
|
next = (struct page *)page->index;
|
[PATCH] mm: split page table lock
Christoph Lameter demonstrated very poor scalability on the SGI 512-way, with
a many-threaded application which concurrently initializes different parts of
a large anonymous area.
This patch corrects that, by using a separate spinlock per page table page, to
guard the page table entries in that page, instead of using the mm's single
page_table_lock. (But even then, page_table_lock is still used to guard page
table allocation, and anon_vma allocation.)
In this implementation, the spinlock is tucked inside the struct page of the
page table page: with a BUILD_BUG_ON in case it overflows - which it would in
the case of 32-bit PA-RISC with spinlock debugging enabled.
Splitting the lock is not quite for free: another cacheline access. Ideally,
I suppose we would use split ptlock only for multi-threaded processes on
multi-cpu machines; but deciding that dynamically would have its own costs.
So for now enable it by config, at some number of cpus - since the Kconfig
language doesn't support inequalities, let preprocessor compare that with
NR_CPUS. But I don't think it's worth being user-configurable: for good
testing of both split and unsplit configs, split now at 4 cpus, and perhaps
change that to 8 later.
There is a benefit even for singly threaded processes: kswapd can be attacking
one part of the mm while another part is busy faulting.
Signed-off-by: Hugh Dickins <hugh@veritas.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-10-29 21:16:40 -04:00
|
|
|
pprev = (struct page **)page_private(page);
|
2005-04-16 18:20:36 -04:00
|
|
|
*pprev = next;
|
|
|
|
if (next)
|
[PATCH] mm: split page table lock
Christoph Lameter demonstrated very poor scalability on the SGI 512-way, with
a many-threaded application which concurrently initializes different parts of
a large anonymous area.
This patch corrects that, by using a separate spinlock per page table page, to
guard the page table entries in that page, instead of using the mm's single
page_table_lock. (But even then, page_table_lock is still used to guard page
table allocation, and anon_vma allocation.)
In this implementation, the spinlock is tucked inside the struct page of the
page table page: with a BUILD_BUG_ON in case it overflows - which it would in
the case of 32-bit PA-RISC with spinlock debugging enabled.
Splitting the lock is not quite for free: another cacheline access. Ideally,
I suppose we would use split ptlock only for multi-threaded processes on
multi-cpu machines; but deciding that dynamically would have its own costs.
So for now enable it by config, at some number of cpus - since the Kconfig
language doesn't support inequalities, let preprocessor compare that with
NR_CPUS. But I don't think it's worth being user-configurable: for good
testing of both split and unsplit configs, split now at 4 cpus, and perhaps
change that to 8 later.
There is a benefit even for singly threaded processes: kswapd can be attacking
one part of the mm while another part is busy faulting.
Signed-off-by: Hugh Dickins <hugh@veritas.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-10-29 21:16:40 -04:00
|
|
|
set_page_private(next, (unsigned long)pprev);
|
2005-04-16 18:20:36 -04:00
|
|
|
}
|
|
|
|
|
2006-12-06 23:33:20 -05:00
|
|
|
void pgd_ctor(void *pgd, struct kmem_cache *cache, unsigned long unused)
|
2005-04-16 18:20:36 -04:00
|
|
|
{
|
|
|
|
unsigned long flags;
|
|
|
|
|
2005-09-03 18:56:50 -04:00
|
|
|
if (PTRS_PER_PMD == 1) {
|
|
|
|
memset(pgd, 0, USER_PTRS_PER_PGD*sizeof(pgd_t));
|
2005-04-16 18:20:36 -04:00
|
|
|
spin_lock_irqsave(&pgd_lock, flags);
|
2005-09-03 18:56:50 -04:00
|
|
|
}
|
2005-04-16 18:20:36 -04:00
|
|
|
|
2005-09-03 18:56:50 -04:00
|
|
|
clone_pgd_range((pgd_t *)pgd + USER_PTRS_PER_PGD,
|
2005-04-16 18:20:36 -04:00
|
|
|
swapper_pg_dir + USER_PTRS_PER_PGD,
|
2005-09-03 18:56:50 -04:00
|
|
|
KERNEL_PGD_PTRS);
|
2007-02-13 07:26:21 -05:00
|
|
|
|
2005-04-16 18:20:36 -04:00
|
|
|
if (PTRS_PER_PMD > 1)
|
|
|
|
return;
|
|
|
|
|
2007-02-13 07:26:21 -05:00
|
|
|
/* must happen under lock */
|
|
|
|
paravirt_alloc_pd_clone(__pa(pgd) >> PAGE_SHIFT,
|
|
|
|
__pa(swapper_pg_dir) >> PAGE_SHIFT,
|
|
|
|
USER_PTRS_PER_PGD, PTRS_PER_PGD - USER_PTRS_PER_PGD);
|
|
|
|
|
2005-04-16 18:20:36 -04:00
|
|
|
pgd_list_add(pgd);
|
|
|
|
spin_unlock_irqrestore(&pgd_lock, flags);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* never called when PTRS_PER_PMD > 1 */
|
2006-12-06 23:33:20 -05:00
|
|
|
void pgd_dtor(void *pgd, struct kmem_cache *cache, unsigned long unused)
|
2005-04-16 18:20:36 -04:00
|
|
|
{
|
|
|
|
unsigned long flags; /* can be called from interrupt context */
|
|
|
|
|
2007-02-13 07:26:21 -05:00
|
|
|
paravirt_release_pd(__pa(pgd) >> PAGE_SHIFT);
|
2005-04-16 18:20:36 -04:00
|
|
|
spin_lock_irqsave(&pgd_lock, flags);
|
|
|
|
pgd_list_del(pgd);
|
|
|
|
spin_unlock_irqrestore(&pgd_lock, flags);
|
|
|
|
}
|
|
|
|
|
|
|
|
pgd_t *pgd_alloc(struct mm_struct *mm)
|
|
|
|
{
|
|
|
|
int i;
|
|
|
|
pgd_t *pgd = kmem_cache_alloc(pgd_cache, GFP_KERNEL);
|
|
|
|
|
|
|
|
if (PTRS_PER_PMD == 1 || !pgd)
|
|
|
|
return pgd;
|
|
|
|
|
|
|
|
for (i = 0; i < USER_PTRS_PER_PGD; ++i) {
|
|
|
|
pmd_t *pmd = kmem_cache_alloc(pmd_cache, GFP_KERNEL);
|
|
|
|
if (!pmd)
|
|
|
|
goto out_oom;
|
2007-02-13 07:26:21 -05:00
|
|
|
paravirt_alloc_pd(__pa(pmd) >> PAGE_SHIFT);
|
2005-04-16 18:20:36 -04:00
|
|
|
set_pgd(&pgd[i], __pgd(1 + __pa(pmd)));
|
|
|
|
}
|
|
|
|
return pgd;
|
|
|
|
|
|
|
|
out_oom:
|
2007-02-13 07:26:21 -05:00
|
|
|
for (i--; i >= 0; i--) {
|
|
|
|
pgd_t pgdent = pgd[i];
|
|
|
|
void* pmd = (void *)__va(pgd_val(pgdent)-1);
|
|
|
|
paravirt_release_pd(__pa(pmd) >> PAGE_SHIFT);
|
|
|
|
kmem_cache_free(pmd_cache, pmd);
|
|
|
|
}
|
2005-04-16 18:20:36 -04:00
|
|
|
kmem_cache_free(pgd_cache, pgd);
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
void pgd_free(pgd_t *pgd)
|
|
|
|
{
|
|
|
|
int i;
|
|
|
|
|
|
|
|
/* in the PAE case user pgd entries are overwritten before usage */
|
|
|
|
if (PTRS_PER_PMD > 1)
|
2007-02-13 07:26:21 -05:00
|
|
|
for (i = 0; i < USER_PTRS_PER_PGD; ++i) {
|
|
|
|
pgd_t pgdent = pgd[i];
|
|
|
|
void* pmd = (void *)__va(pgd_val(pgdent)-1);
|
|
|
|
paravirt_release_pd(__pa(pmd) >> PAGE_SHIFT);
|
|
|
|
kmem_cache_free(pmd_cache, pmd);
|
|
|
|
}
|
[PATCH] freepgt: free_pgtables use vma list
Recent woes with some arches needing their own pgd_addr_end macro; and 4-level
clear_page_range regression since 2.6.10's clear_page_tables; and its
long-standing well-known inefficiency in searching throughout the higher-level
page tables for those few entries to clear and free: all can be blamed on
ignoring the list of vmas when we free page tables.
Replace exit_mmap's clear_page_range of the total user address space by
free_pgtables operating on the mm's vma list; unmap_region use it in the same
way, giving floor and ceiling beyond which it may not free tables. This
brings lmbench fork/exec/sh numbers back to 2.6.10 (unless preempt is enabled,
in which case latency fixes spoil unmap_vmas throughput).
Beware: the do_mmap_pgoff driver failure case must now use unmap_region
instead of zap_page_range, since a page table might have been allocated, and
can only be freed while it is touched by some vma.
Move free_pgtables from mmap.c to memory.c, where its lower levels are adapted
from the clear_page_range levels. (Most of free_pgtables' old code was
actually for a non-existent case, prev not properly set up, dating from before
hch gave us split_vma.) Pass mmu_gather** in the public interfaces, since we
might want to add latency lockdrops later; but no attempt to do so yet, going
by vma should itself reduce latency.
But what if is_hugepage_only_range? Those ia64 and ppc64 cases need careful
examination: put that off until a later patch of the series.
What of x86_64's 32bit vdso page __map_syscall32 maps outside any vma?
And the range to sparc64's flush_tlb_pgtables? It's less clear to me now that
we need to do more than is done here - every PMD_SIZE ever occupied will be
flushed, do we really have to flush every PGDIR_SIZE ever partially occupied?
A shame to complicate it unnecessarily.
Special thanks to David Miller for time spent repairing my ceilings.
Signed-off-by: Hugh Dickins <hugh@veritas.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2005-04-19 16:29:15 -04:00
|
|
|
/* in the non-PAE case, free_pgtables() clears user pgd entries */
|
2005-04-16 18:20:36 -04:00
|
|
|
kmem_cache_free(pgd_cache, pgd);
|
|
|
|
}
|