mirror of
https://github.com/AuxXxilium/linux_dsm_epyc7002.git
synced 2024-12-04 13:07:00 +07:00
b2fa47e6bf
Rework the architecture page table functions to access the bits in the page table extension array (pgste). There are a number of changes: 1) Fix missing pgste update if the attach_count for the mm is <= 1. 2) For every operation that affects the invalid bit in the pte or the rcp byte in the pgste the pcl lock needs to be acquired. The function pgste_get_lock gets the pcl lock and returns the current pgste value for a pte pointer. The function pgste_set_unlock stores the pgste and releases the lock. Between these two calls the bits in the pgste can be shuffled. 3) Define two software bits in the pte _PAGE_SWR and _PAGE_SWC to avoid calling SetPageDirty and SetPageReferenced from pgtable.h. If the host reference backup bit or the host change backup bit has been set the dirty/referenced state is transfered to the pte. The common code will pick up the state from the pte. 4) Add ptep_modify_prot_start and ptep_modify_prot_commit for mprotect. 5) Remove pgd_populate_kernel, pud_populate_kernel, pmd_populate_kernel pgd_clear_kernel, pud_clear_kernel, pmd_clear_kernel and ptep_invalidate. 6) Rename kvm_s390_test_and_clear_page_dirty to ptep_test_and_clear_user_dirty and add ptep_test_and_clear_user_young. 7) Define mm_exclusive() and mm_has_pgste() helper to improve readability. Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
238 lines
5.9 KiB
C
238 lines
5.9 KiB
C
/*
|
|
* arch/s390/mm/init.c
|
|
*
|
|
* S390 version
|
|
* Copyright (C) 1999 IBM Deutschland Entwicklung GmbH, IBM Corporation
|
|
* Author(s): Hartmut Penner (hp@de.ibm.com)
|
|
*
|
|
* Derived from "arch/i386/mm/init.c"
|
|
* Copyright (C) 1995 Linus Torvalds
|
|
*/
|
|
|
|
#include <linux/signal.h>
|
|
#include <linux/sched.h>
|
|
#include <linux/kernel.h>
|
|
#include <linux/errno.h>
|
|
#include <linux/string.h>
|
|
#include <linux/types.h>
|
|
#include <linux/ptrace.h>
|
|
#include <linux/mman.h>
|
|
#include <linux/mm.h>
|
|
#include <linux/swap.h>
|
|
#include <linux/smp.h>
|
|
#include <linux/init.h>
|
|
#include <linux/pagemap.h>
|
|
#include <linux/bootmem.h>
|
|
#include <linux/pfn.h>
|
|
#include <linux/poison.h>
|
|
#include <linux/initrd.h>
|
|
#include <linux/gfp.h>
|
|
#include <asm/processor.h>
|
|
#include <asm/system.h>
|
|
#include <asm/uaccess.h>
|
|
#include <asm/pgtable.h>
|
|
#include <asm/pgalloc.h>
|
|
#include <asm/dma.h>
|
|
#include <asm/lowcore.h>
|
|
#include <asm/tlb.h>
|
|
#include <asm/tlbflush.h>
|
|
#include <asm/sections.h>
|
|
|
|
pgd_t swapper_pg_dir[PTRS_PER_PGD] __attribute__((__aligned__(PAGE_SIZE)));
|
|
|
|
unsigned long empty_zero_page, zero_page_mask;
|
|
EXPORT_SYMBOL(empty_zero_page);
|
|
|
|
static unsigned long setup_zero_pages(void)
|
|
{
|
|
struct cpuid cpu_id;
|
|
unsigned int order;
|
|
unsigned long size;
|
|
struct page *page;
|
|
int i;
|
|
|
|
get_cpu_id(&cpu_id);
|
|
switch (cpu_id.machine) {
|
|
case 0x9672: /* g5 */
|
|
case 0x2064: /* z900 */
|
|
case 0x2066: /* z900 */
|
|
case 0x2084: /* z990 */
|
|
case 0x2086: /* z990 */
|
|
case 0x2094: /* z9-109 */
|
|
case 0x2096: /* z9-109 */
|
|
order = 0;
|
|
break;
|
|
case 0x2097: /* z10 */
|
|
case 0x2098: /* z10 */
|
|
default:
|
|
order = 2;
|
|
break;
|
|
}
|
|
|
|
empty_zero_page = __get_free_pages(GFP_KERNEL | __GFP_ZERO, order);
|
|
if (!empty_zero_page)
|
|
panic("Out of memory in setup_zero_pages");
|
|
|
|
page = virt_to_page((void *) empty_zero_page);
|
|
split_page(page, order);
|
|
for (i = 1 << order; i > 0; i--) {
|
|
SetPageReserved(page);
|
|
page++;
|
|
}
|
|
|
|
size = PAGE_SIZE << order;
|
|
zero_page_mask = (size - 1) & PAGE_MASK;
|
|
|
|
return 1UL << order;
|
|
}
|
|
|
|
/*
|
|
* paging_init() sets up the page tables
|
|
*/
|
|
void __init paging_init(void)
|
|
{
|
|
unsigned long max_zone_pfns[MAX_NR_ZONES];
|
|
unsigned long pgd_type;
|
|
|
|
init_mm.pgd = swapper_pg_dir;
|
|
S390_lowcore.kernel_asce = __pa(init_mm.pgd) & PAGE_MASK;
|
|
#ifdef CONFIG_64BIT
|
|
/* A three level page table (4TB) is enough for the kernel space. */
|
|
S390_lowcore.kernel_asce |= _ASCE_TYPE_REGION3 | _ASCE_TABLE_LENGTH;
|
|
pgd_type = _REGION3_ENTRY_EMPTY;
|
|
#else
|
|
S390_lowcore.kernel_asce |= _ASCE_TABLE_LENGTH;
|
|
pgd_type = _SEGMENT_ENTRY_EMPTY;
|
|
#endif
|
|
clear_table((unsigned long *) init_mm.pgd, pgd_type,
|
|
sizeof(unsigned long)*2048);
|
|
vmem_map_init();
|
|
|
|
/* enable virtual mapping in kernel mode */
|
|
__ctl_load(S390_lowcore.kernel_asce, 1, 1);
|
|
__ctl_load(S390_lowcore.kernel_asce, 7, 7);
|
|
__ctl_load(S390_lowcore.kernel_asce, 13, 13);
|
|
arch_local_irq_restore(4UL << (BITS_PER_LONG - 8));
|
|
|
|
atomic_set(&init_mm.context.attach_count, 1);
|
|
|
|
sparse_memory_present_with_active_regions(MAX_NUMNODES);
|
|
sparse_init();
|
|
memset(max_zone_pfns, 0, sizeof(max_zone_pfns));
|
|
#ifdef CONFIG_ZONE_DMA
|
|
max_zone_pfns[ZONE_DMA] = PFN_DOWN(MAX_DMA_ADDRESS);
|
|
#endif
|
|
max_zone_pfns[ZONE_NORMAL] = max_low_pfn;
|
|
free_area_init_nodes(max_zone_pfns);
|
|
fault_init();
|
|
}
|
|
|
|
void __init mem_init(void)
|
|
{
|
|
unsigned long codesize, reservedpages, datasize, initsize;
|
|
|
|
max_mapnr = num_physpages = max_low_pfn;
|
|
high_memory = (void *) __va(max_low_pfn * PAGE_SIZE);
|
|
|
|
/* Setup guest page hinting */
|
|
cmma_init();
|
|
|
|
/* this will put all low memory onto the freelists */
|
|
totalram_pages += free_all_bootmem();
|
|
totalram_pages -= setup_zero_pages(); /* Setup zeroed pages. */
|
|
|
|
reservedpages = 0;
|
|
|
|
codesize = (unsigned long) &_etext - (unsigned long) &_text;
|
|
datasize = (unsigned long) &_edata - (unsigned long) &_etext;
|
|
initsize = (unsigned long) &__init_end - (unsigned long) &__init_begin;
|
|
printk("Memory: %luk/%luk available (%ldk kernel code, %ldk reserved, %ldk data, %ldk init)\n",
|
|
nr_free_pages() << (PAGE_SHIFT-10),
|
|
max_mapnr << (PAGE_SHIFT-10),
|
|
codesize >> 10,
|
|
reservedpages << (PAGE_SHIFT-10),
|
|
datasize >>10,
|
|
initsize >> 10);
|
|
printk("Write protected kernel read-only data: %#lx - %#lx\n",
|
|
(unsigned long)&_stext,
|
|
PFN_ALIGN((unsigned long)&_eshared) - 1);
|
|
}
|
|
|
|
#ifdef CONFIG_DEBUG_PAGEALLOC
|
|
void kernel_map_pages(struct page *page, int numpages, int enable)
|
|
{
|
|
pgd_t *pgd;
|
|
pud_t *pud;
|
|
pmd_t *pmd;
|
|
pte_t *pte;
|
|
unsigned long address;
|
|
int i;
|
|
|
|
for (i = 0; i < numpages; i++) {
|
|
address = page_to_phys(page + i);
|
|
pgd = pgd_offset_k(address);
|
|
pud = pud_offset(pgd, address);
|
|
pmd = pmd_offset(pud, address);
|
|
pte = pte_offset_kernel(pmd, address);
|
|
if (!enable) {
|
|
__ptep_ipte(address, pte);
|
|
pte_val(*pte) = _PAGE_TYPE_EMPTY;
|
|
continue;
|
|
}
|
|
*pte = mk_pte_phys(address, __pgprot(_PAGE_TYPE_RW));
|
|
/* Flush cpu write queue. */
|
|
mb();
|
|
}
|
|
}
|
|
#endif
|
|
|
|
void free_init_pages(char *what, unsigned long begin, unsigned long end)
|
|
{
|
|
unsigned long addr = begin;
|
|
|
|
if (begin >= end)
|
|
return;
|
|
for (; addr < end; addr += PAGE_SIZE) {
|
|
ClearPageReserved(virt_to_page(addr));
|
|
init_page_count(virt_to_page(addr));
|
|
memset((void *)(addr & PAGE_MASK), POISON_FREE_INITMEM,
|
|
PAGE_SIZE);
|
|
free_page(addr);
|
|
totalram_pages++;
|
|
}
|
|
printk(KERN_INFO "Freeing %s: %luk freed\n", what, (end - begin) >> 10);
|
|
}
|
|
|
|
void free_initmem(void)
|
|
{
|
|
free_init_pages("unused kernel memory",
|
|
(unsigned long)&__init_begin,
|
|
(unsigned long)&__init_end);
|
|
}
|
|
|
|
#ifdef CONFIG_BLK_DEV_INITRD
|
|
void free_initrd_mem(unsigned long start, unsigned long end)
|
|
{
|
|
free_init_pages("initrd memory", start, end);
|
|
}
|
|
#endif
|
|
|
|
#ifdef CONFIG_MEMORY_HOTPLUG
|
|
int arch_add_memory(int nid, u64 start, u64 size)
|
|
{
|
|
struct pglist_data *pgdat;
|
|
struct zone *zone;
|
|
int rc;
|
|
|
|
pgdat = NODE_DATA(nid);
|
|
zone = pgdat->node_zones + ZONE_MOVABLE;
|
|
rc = vmem_add_mapping(start, size);
|
|
if (rc)
|
|
return rc;
|
|
rc = __add_pages(nid, zone, PFN_DOWN(start), PFN_DOWN(size));
|
|
if (rc)
|
|
vmem_remove_mapping(start, size);
|
|
return rc;
|
|
}
|
|
#endif /* CONFIG_MEMORY_HOTPLUG */
|