2005-04-16 22:20:36 +00:00
|
|
|
/*
|
|
|
|
* linux/arch/arm/mm/mm-armv.c
|
|
|
|
*
|
2005-10-28 13:48:37 +00:00
|
|
|
* Copyright (C) 1998-2005 Russell King
|
2005-04-16 22:20:36 +00:00
|
|
|
*
|
|
|
|
* This program is free software; you can redistribute it and/or modify
|
|
|
|
* it under the terms of the GNU General Public License version 2 as
|
|
|
|
* published by the Free Software Foundation.
|
|
|
|
*
|
|
|
|
* Page table sludge for ARM v3 and v4 processor architectures.
|
|
|
|
*/
|
|
|
|
#include <linux/mm.h>
|
|
|
|
#include <linux/highmem.h>
|
|
|
|
|
|
|
|
#include <asm/pgalloc.h>
|
|
|
|
#include <asm/page.h>
|
|
|
|
#include <asm/tlbflush.h>
|
|
|
|
|
2006-08-21 16:06:38 +00:00
|
|
|
#include "mm.h"
|
|
|
|
|
2005-04-16 22:20:36 +00:00
|
|
|
#define FIRST_KERNEL_PGD_NR (FIRST_USER_PGD_NR + USER_PTRS_PER_PGD)
|
|
|
|
|
|
|
|
/*
|
|
|
|
* need to get a 16k page for level 1
|
|
|
|
*/
|
|
|
|
pgd_t *get_pgd_slow(struct mm_struct *mm)
|
|
|
|
{
|
|
|
|
pgd_t *new_pgd, *init_pgd;
|
|
|
|
pmd_t *new_pmd, *init_pmd;
|
|
|
|
pte_t *new_pte, *init_pte;
|
|
|
|
|
|
|
|
new_pgd = (pgd_t *)__get_free_pages(GFP_KERNEL, 2);
|
|
|
|
if (!new_pgd)
|
|
|
|
goto no_pgd;
|
|
|
|
|
|
|
|
memzero(new_pgd, FIRST_KERNEL_PGD_NR * sizeof(pgd_t));
|
|
|
|
|
2005-06-27 13:08:56 +00:00
|
|
|
/*
|
|
|
|
* Copy over the kernel and IO PGD entries
|
|
|
|
*/
|
2005-04-16 22:20:36 +00:00
|
|
|
init_pgd = pgd_offset_k(0);
|
2005-06-27 13:08:56 +00:00
|
|
|
memcpy(new_pgd + FIRST_KERNEL_PGD_NR, init_pgd + FIRST_KERNEL_PGD_NR,
|
|
|
|
(PTRS_PER_PGD - FIRST_KERNEL_PGD_NR) * sizeof(pgd_t));
|
|
|
|
|
|
|
|
clean_dcache_area(new_pgd, PTRS_PER_PGD * sizeof(pgd_t));
|
2005-04-16 22:20:36 +00:00
|
|
|
|
|
|
|
if (!vectors_high()) {
|
|
|
|
/*
|
|
|
|
* On ARM, first page must always be allocated since it
|
|
|
|
* contains the machine vectors.
|
|
|
|
*/
|
|
|
|
new_pmd = pmd_alloc(mm, new_pgd, 0);
|
|
|
|
if (!new_pmd)
|
|
|
|
goto no_pmd;
|
|
|
|
|
|
|
|
new_pte = pte_alloc_map(mm, new_pmd, 0);
|
|
|
|
if (!new_pte)
|
|
|
|
goto no_pte;
|
|
|
|
|
|
|
|
init_pmd = pmd_offset(init_pgd, 0);
|
|
|
|
init_pte = pte_offset_map_nested(init_pmd, 0);
|
|
|
|
set_pte(new_pte, *init_pte);
|
|
|
|
pte_unmap_nested(init_pte);
|
|
|
|
pte_unmap(new_pte);
|
|
|
|
}
|
|
|
|
|
|
|
|
return new_pgd;
|
|
|
|
|
|
|
|
no_pte:
|
|
|
|
pmd_free(new_pmd);
|
|
|
|
no_pmd:
|
|
|
|
free_pages((unsigned long)new_pgd, 2);
|
|
|
|
no_pgd:
|
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
void free_pgd_slow(pgd_t *pgd)
|
|
|
|
{
|
|
|
|
pmd_t *pmd;
|
|
|
|
struct page *pte;
|
|
|
|
|
|
|
|
if (!pgd)
|
|
|
|
return;
|
|
|
|
|
|
|
|
/* pgd is always present and good */
|
2005-05-09 19:52:51 +00:00
|
|
|
pmd = pmd_off(pgd, 0);
|
2005-04-16 22:20:36 +00:00
|
|
|
if (pmd_none(*pmd))
|
|
|
|
goto free;
|
|
|
|
if (pmd_bad(*pmd)) {
|
|
|
|
pmd_ERROR(*pmd);
|
|
|
|
pmd_clear(pmd);
|
|
|
|
goto free;
|
|
|
|
}
|
|
|
|
|
|
|
|
pte = pmd_page(*pmd);
|
|
|
|
pmd_clear(pmd);
|
2006-06-30 08:55:38 +00:00
|
|
|
dec_zone_page_state(virt_to_page((unsigned long *)pgd), NR_PAGETABLE);
|
[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>
2005-10-30 01:16:40 +00:00
|
|
|
pte_lock_deinit(pte);
|
2005-04-16 22:20:36 +00:00
|
|
|
pte_free(pte);
|
|
|
|
pmd_free(pmd);
|
|
|
|
free:
|
|
|
|
free_pages((unsigned long) pgd, 2);
|
|
|
|
}
|