linux/mm/mprotect.c
Henry Willard 859d4adc34 mm: numa: do not trap faults on shared data section pages.
Workloads consisting of a large number of processes running the same
program with a very large shared data segment may experience performance
problems when numa balancing attempts to migrate the shared cow pages.
This manifests itself with many processes or tasks in
TASK_UNINTERRUPTIBLE state waiting for the shared pages to be migrated.

The program listed below simulates the conditions with these results
when run with 288 processes on a 144 core/8 socket machine.

Average throughput 	Average throughput     Average throughput
with numa_balancing=0	with numa_balancing=1  with numa_balancing=1
     			without the patch      with the patch
---------------------	---------------------  ---------------------
2118782			2021534		       2107979

Complex production environments show less variability and fewer poorly
performing outliers accompanied with a smaller number of processes
waiting on NUMA page migration with this patch applied.  In some cases,
%iowait drops from 16%-26% to 0.

  // SPDX-License-Identifier: GPL-2.0
  /*
   * Copyright (c) 2017 Oracle and/or its affiliates. All rights reserved.
   */
  #include <sys/time.h>
  #include <stdio.h>
  #include <wait.h>
  #include <sys/mman.h>

  int a[1000000] = {13};

  int  main(int argc, const char **argv)
  {
	int n = 0;
	int i;
	pid_t pid;
	int stat;
	int *count_array;
	int cpu_count = 288;
	long total = 0;

	struct timeval t1, t2 = {(argc > 1 ? atoi(argv[1]) : 10), 0};

	if (argc > 2)
		cpu_count = atoi(argv[2]);

	count_array = mmap(NULL, cpu_count * sizeof(int),
			   (PROT_READ|PROT_WRITE),
			   (MAP_SHARED|MAP_ANONYMOUS), 0, 0);

	if (count_array == MAP_FAILED) {
		perror("mmap:");
		return 0;
	}

	for (i = 0; i < cpu_count; ++i) {
		pid = fork();
		if (pid <= 0)
			break;
		if ((i & 0xf) == 0)
			usleep(2);
	}

	if (pid != 0) {
		if (i == 0) {
			perror("fork:");
			return 0;
		}

		for (;;) {
			pid = wait(&stat);
			if (pid < 0)
				break;
		}

		for (i = 0; i < cpu_count; ++i)
			total += count_array[i];

		printf("Total %ld\n", total);
		munmap(count_array, cpu_count * sizeof(int));
		return 0;
	}

	gettimeofday(&t1, 0);
	timeradd(&t1, &t2, &t1);
	while (timercmp(&t2, &t1, <)) {
		int b = 0;
		int j;

		for (j = 0; j < 1000000; j++)
			b += a[j];
		gettimeofday(&t2, 0);
		n++;
	}
	count_array[i] = n;
	return 0;
  }

This patch changes change_pte_range() to skip shared copy-on-write pages
when called from change_prot_numa().

NOTE: change_prot_numa() is nominally called from task_numa_work() and
queue_pages_test_walk().  task_numa_work() is the auto NUMA balancing
path, and queue_pages_test_walk() is part of explicit NUMA policy
management.  However, queue_pages_test_walk() only calls
change_prot_numa() when MPOL_MF_LAZY is specified and currently that is
not allowed, so change_prot_numa() is only called from auto NUMA
balancing.

In the case of explicit NUMA policy management, shared pages are not
migrated unless MPOL_MF_MOVE_ALL is specified, and MPOL_MF_MOVE_ALL
depends on CAP_SYS_NICE.  Currently, there is no way to pass information
about MPOL_MF_MOVE_ALL to change_pte_range.  This will have to be fixed
if MPOL_MF_LAZY is enabled and MPOL_MF_MOVE_ALL is to be honored in lazy
migration mode.

task_numa_work() skips the read-only VMAs of programs and shared
libraries.

Link: http://lkml.kernel.org/r/1516751617-7369-1-git-send-email-henry.willard@oracle.com
Signed-off-by: Henry Willard <henry.willard@oracle.com>
Reviewed-by: Håkon Bugge <haakon.bugge@oracle.com>
Reviewed-by: Steve Sistare <steven.sistare@oracle.com>
Acked-by: Mel Gorman <mgorman@suse.de>
Cc: Kate Stewart <kstewart@linuxfoundation.org>
Cc: Zi Yan <zi.yan@cs.rutgers.edu>
Cc: Philippe Ombredanne <pombredanne@nexb.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: "Jérôme Glisse" <jglisse@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>

Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2018-01-31 17:18:40 -08:00

579 lines
14 KiB
C

// SPDX-License-Identifier: GPL-2.0
/*
* mm/mprotect.c
*
* (C) Copyright 1994 Linus Torvalds
* (C) Copyright 2002 Christoph Hellwig
*
* Address space accounting code <alan@lxorguk.ukuu.org.uk>
* (C) Copyright 2002 Red Hat Inc, All Rights Reserved
*/
#include <linux/mm.h>
#include <linux/hugetlb.h>
#include <linux/shm.h>
#include <linux/mman.h>
#include <linux/fs.h>
#include <linux/highmem.h>
#include <linux/security.h>
#include <linux/mempolicy.h>
#include <linux/personality.h>
#include <linux/syscalls.h>
#include <linux/swap.h>
#include <linux/swapops.h>
#include <linux/mmu_notifier.h>
#include <linux/migrate.h>
#include <linux/perf_event.h>
#include <linux/pkeys.h>
#include <linux/ksm.h>
#include <linux/uaccess.h>
#include <asm/pgtable.h>
#include <asm/cacheflush.h>
#include <asm/mmu_context.h>
#include <asm/tlbflush.h>
#include "internal.h"
static unsigned long change_pte_range(struct vm_area_struct *vma, pmd_t *pmd,
unsigned long addr, unsigned long end, pgprot_t newprot,
int dirty_accountable, int prot_numa)
{
struct mm_struct *mm = vma->vm_mm;
pte_t *pte, oldpte;
spinlock_t *ptl;
unsigned long pages = 0;
int target_node = NUMA_NO_NODE;
/*
* Can be called with only the mmap_sem for reading by
* prot_numa so we must check the pmd isn't constantly
* changing from under us from pmd_none to pmd_trans_huge
* and/or the other way around.
*/
if (pmd_trans_unstable(pmd))
return 0;
/*
* The pmd points to a regular pte so the pmd can't change
* from under us even if the mmap_sem is only hold for
* reading.
*/
pte = pte_offset_map_lock(vma->vm_mm, pmd, addr, &ptl);
/* Get target node for single threaded private VMAs */
if (prot_numa && !(vma->vm_flags & VM_SHARED) &&
atomic_read(&vma->vm_mm->mm_users) == 1)
target_node = numa_node_id();
flush_tlb_batched_pending(vma->vm_mm);
arch_enter_lazy_mmu_mode();
do {
oldpte = *pte;
if (pte_present(oldpte)) {
pte_t ptent;
bool preserve_write = prot_numa && pte_write(oldpte);
/*
* Avoid trapping faults against the zero or KSM
* pages. See similar comment in change_huge_pmd.
*/
if (prot_numa) {
struct page *page;
page = vm_normal_page(vma, addr, oldpte);
if (!page || PageKsm(page))
continue;
/* Also skip shared copy-on-write pages */
if (is_cow_mapping(vma->vm_flags) &&
page_mapcount(page) != 1)
continue;
/* Avoid TLB flush if possible */
if (pte_protnone(oldpte))
continue;
/*
* Don't mess with PTEs if page is already on the node
* a single-threaded process is running on.
*/
if (target_node == page_to_nid(page))
continue;
}
ptent = ptep_modify_prot_start(mm, addr, pte);
ptent = pte_modify(ptent, newprot);
if (preserve_write)
ptent = pte_mk_savedwrite(ptent);
/* Avoid taking write faults for known dirty pages */
if (dirty_accountable && pte_dirty(ptent) &&
(pte_soft_dirty(ptent) ||
!(vma->vm_flags & VM_SOFTDIRTY))) {
ptent = pte_mkwrite(ptent);
}
ptep_modify_prot_commit(mm, addr, pte, ptent);
pages++;
} else if (IS_ENABLED(CONFIG_MIGRATION)) {
swp_entry_t entry = pte_to_swp_entry(oldpte);
if (is_write_migration_entry(entry)) {
pte_t newpte;
/*
* A protection check is difficult so
* just be safe and disable write
*/
make_migration_entry_read(&entry);
newpte = swp_entry_to_pte(entry);
if (pte_swp_soft_dirty(oldpte))
newpte = pte_swp_mksoft_dirty(newpte);
set_pte_at(mm, addr, pte, newpte);
pages++;
}
if (is_write_device_private_entry(entry)) {
pte_t newpte;
/*
* We do not preserve soft-dirtiness. See
* copy_one_pte() for explanation.
*/
make_device_private_entry_read(&entry);
newpte = swp_entry_to_pte(entry);
set_pte_at(mm, addr, pte, newpte);
pages++;
}
}
} while (pte++, addr += PAGE_SIZE, addr != end);
arch_leave_lazy_mmu_mode();
pte_unmap_unlock(pte - 1, ptl);
return pages;
}
static inline unsigned long change_pmd_range(struct vm_area_struct *vma,
pud_t *pud, unsigned long addr, unsigned long end,
pgprot_t newprot, int dirty_accountable, int prot_numa)
{
pmd_t *pmd;
struct mm_struct *mm = vma->vm_mm;
unsigned long next;
unsigned long pages = 0;
unsigned long nr_huge_updates = 0;
unsigned long mni_start = 0;
pmd = pmd_offset(pud, addr);
do {
unsigned long this_pages;
next = pmd_addr_end(addr, end);
if (!is_swap_pmd(*pmd) && !pmd_trans_huge(*pmd) && !pmd_devmap(*pmd)
&& pmd_none_or_clear_bad(pmd))
goto next;
/* invoke the mmu notifier if the pmd is populated */
if (!mni_start) {
mni_start = addr;
mmu_notifier_invalidate_range_start(mm, mni_start, end);
}
if (is_swap_pmd(*pmd) || pmd_trans_huge(*pmd) || pmd_devmap(*pmd)) {
if (next - addr != HPAGE_PMD_SIZE) {
__split_huge_pmd(vma, pmd, addr, false, NULL);
} else {
int nr_ptes = change_huge_pmd(vma, pmd, addr,
newprot, prot_numa);
if (nr_ptes) {
if (nr_ptes == HPAGE_PMD_NR) {
pages += HPAGE_PMD_NR;
nr_huge_updates++;
}
/* huge pmd was handled */
goto next;
}
}
/* fall through, the trans huge pmd just split */
}
this_pages = change_pte_range(vma, pmd, addr, next, newprot,
dirty_accountable, prot_numa);
pages += this_pages;
next:
cond_resched();
} while (pmd++, addr = next, addr != end);
if (mni_start)
mmu_notifier_invalidate_range_end(mm, mni_start, end);
if (nr_huge_updates)
count_vm_numa_events(NUMA_HUGE_PTE_UPDATES, nr_huge_updates);
return pages;
}
static inline unsigned long change_pud_range(struct vm_area_struct *vma,
p4d_t *p4d, unsigned long addr, unsigned long end,
pgprot_t newprot, int dirty_accountable, int prot_numa)
{
pud_t *pud;
unsigned long next;
unsigned long pages = 0;
pud = pud_offset(p4d, addr);
do {
next = pud_addr_end(addr, end);
if (pud_none_or_clear_bad(pud))
continue;
pages += change_pmd_range(vma, pud, addr, next, newprot,
dirty_accountable, prot_numa);
} while (pud++, addr = next, addr != end);
return pages;
}
static inline unsigned long change_p4d_range(struct vm_area_struct *vma,
pgd_t *pgd, unsigned long addr, unsigned long end,
pgprot_t newprot, int dirty_accountable, int prot_numa)
{
p4d_t *p4d;
unsigned long next;
unsigned long pages = 0;
p4d = p4d_offset(pgd, addr);
do {
next = p4d_addr_end(addr, end);
if (p4d_none_or_clear_bad(p4d))
continue;
pages += change_pud_range(vma, p4d, addr, next, newprot,
dirty_accountable, prot_numa);
} while (p4d++, addr = next, addr != end);
return pages;
}
static unsigned long change_protection_range(struct vm_area_struct *vma,
unsigned long addr, unsigned long end, pgprot_t newprot,
int dirty_accountable, int prot_numa)
{
struct mm_struct *mm = vma->vm_mm;
pgd_t *pgd;
unsigned long next;
unsigned long start = addr;
unsigned long pages = 0;
BUG_ON(addr >= end);
pgd = pgd_offset(mm, addr);
flush_cache_range(vma, addr, end);
inc_tlb_flush_pending(mm);
do {
next = pgd_addr_end(addr, end);
if (pgd_none_or_clear_bad(pgd))
continue;
pages += change_p4d_range(vma, pgd, addr, next, newprot,
dirty_accountable, prot_numa);
} while (pgd++, addr = next, addr != end);
/* Only flush the TLB if we actually modified any entries: */
if (pages)
flush_tlb_range(vma, start, end);
dec_tlb_flush_pending(mm);
return pages;
}
unsigned long change_protection(struct vm_area_struct *vma, unsigned long start,
unsigned long end, pgprot_t newprot,
int dirty_accountable, int prot_numa)
{
unsigned long pages;
if (is_vm_hugetlb_page(vma))
pages = hugetlb_change_protection(vma, start, end, newprot);
else
pages = change_protection_range(vma, start, end, newprot, dirty_accountable, prot_numa);
return pages;
}
int
mprotect_fixup(struct vm_area_struct *vma, struct vm_area_struct **pprev,
unsigned long start, unsigned long end, unsigned long newflags)
{
struct mm_struct *mm = vma->vm_mm;
unsigned long oldflags = vma->vm_flags;
long nrpages = (end - start) >> PAGE_SHIFT;
unsigned long charged = 0;
pgoff_t pgoff;
int error;
int dirty_accountable = 0;
if (newflags == oldflags) {
*pprev = vma;
return 0;
}
/*
* If we make a private mapping writable we increase our commit;
* but (without finer accounting) cannot reduce our commit if we
* make it unwritable again. hugetlb mapping were accounted for
* even if read-only so there is no need to account for them here
*/
if (newflags & VM_WRITE) {
/* Check space limits when area turns into data. */
if (!may_expand_vm(mm, newflags, nrpages) &&
may_expand_vm(mm, oldflags, nrpages))
return -ENOMEM;
if (!(oldflags & (VM_ACCOUNT|VM_WRITE|VM_HUGETLB|
VM_SHARED|VM_NORESERVE))) {
charged = nrpages;
if (security_vm_enough_memory_mm(mm, charged))
return -ENOMEM;
newflags |= VM_ACCOUNT;
}
}
/*
* First try to merge with previous and/or next vma.
*/
pgoff = vma->vm_pgoff + ((start - vma->vm_start) >> PAGE_SHIFT);
*pprev = vma_merge(mm, *pprev, start, end, newflags,
vma->anon_vma, vma->vm_file, pgoff, vma_policy(vma),
vma->vm_userfaultfd_ctx);
if (*pprev) {
vma = *pprev;
VM_WARN_ON((vma->vm_flags ^ newflags) & ~VM_SOFTDIRTY);
goto success;
}
*pprev = vma;
if (start != vma->vm_start) {
error = split_vma(mm, vma, start, 1);
if (error)
goto fail;
}
if (end != vma->vm_end) {
error = split_vma(mm, vma, end, 0);
if (error)
goto fail;
}
success:
/*
* vm_flags and vm_page_prot are protected by the mmap_sem
* held in write mode.
*/
vma->vm_flags = newflags;
dirty_accountable = vma_wants_writenotify(vma, vma->vm_page_prot);
vma_set_page_prot(vma);
change_protection(vma, start, end, vma->vm_page_prot,
dirty_accountable, 0);
/*
* Private VM_LOCKED VMA becoming writable: trigger COW to avoid major
* fault on access.
*/
if ((oldflags & (VM_WRITE | VM_SHARED | VM_LOCKED)) == VM_LOCKED &&
(newflags & VM_WRITE)) {
populate_vma_page_range(vma, start, end, NULL);
}
vm_stat_account(mm, oldflags, -nrpages);
vm_stat_account(mm, newflags, nrpages);
perf_event_mmap(vma);
return 0;
fail:
vm_unacct_memory(charged);
return error;
}
/*
* pkey==-1 when doing a legacy mprotect()
*/
static int do_mprotect_pkey(unsigned long start, size_t len,
unsigned long prot, int pkey)
{
unsigned long nstart, end, tmp, reqprot;
struct vm_area_struct *vma, *prev;
int error = -EINVAL;
const int grows = prot & (PROT_GROWSDOWN|PROT_GROWSUP);
const bool rier = (current->personality & READ_IMPLIES_EXEC) &&
(prot & PROT_READ);
prot &= ~(PROT_GROWSDOWN|PROT_GROWSUP);
if (grows == (PROT_GROWSDOWN|PROT_GROWSUP)) /* can't be both */
return -EINVAL;
if (start & ~PAGE_MASK)
return -EINVAL;
if (!len)
return 0;
len = PAGE_ALIGN(len);
end = start + len;
if (end <= start)
return -ENOMEM;
if (!arch_validate_prot(prot))
return -EINVAL;
reqprot = prot;
if (down_write_killable(&current->mm->mmap_sem))
return -EINTR;
/*
* If userspace did not allocate the pkey, do not let
* them use it here.
*/
error = -EINVAL;
if ((pkey != -1) && !mm_pkey_is_allocated(current->mm, pkey))
goto out;
vma = find_vma(current->mm, start);
error = -ENOMEM;
if (!vma)
goto out;
prev = vma->vm_prev;
if (unlikely(grows & PROT_GROWSDOWN)) {
if (vma->vm_start >= end)
goto out;
start = vma->vm_start;
error = -EINVAL;
if (!(vma->vm_flags & VM_GROWSDOWN))
goto out;
} else {
if (vma->vm_start > start)
goto out;
if (unlikely(grows & PROT_GROWSUP)) {
end = vma->vm_end;
error = -EINVAL;
if (!(vma->vm_flags & VM_GROWSUP))
goto out;
}
}
if (start > vma->vm_start)
prev = vma;
for (nstart = start ; ; ) {
unsigned long mask_off_old_flags;
unsigned long newflags;
int new_vma_pkey;
/* Here we know that vma->vm_start <= nstart < vma->vm_end. */
/* Does the application expect PROT_READ to imply PROT_EXEC */
if (rier && (vma->vm_flags & VM_MAYEXEC))
prot |= PROT_EXEC;
/*
* Each mprotect() call explicitly passes r/w/x permissions.
* If a permission is not passed to mprotect(), it must be
* cleared from the VMA.
*/
mask_off_old_flags = VM_READ | VM_WRITE | VM_EXEC |
ARCH_VM_PKEY_FLAGS;
new_vma_pkey = arch_override_mprotect_pkey(vma, prot, pkey);
newflags = calc_vm_prot_bits(prot, new_vma_pkey);
newflags |= (vma->vm_flags & ~mask_off_old_flags);
/* newflags >> 4 shift VM_MAY% in place of VM_% */
if ((newflags & ~(newflags >> 4)) & (VM_READ | VM_WRITE | VM_EXEC)) {
error = -EACCES;
goto out;
}
error = security_file_mprotect(vma, reqprot, prot);
if (error)
goto out;
tmp = vma->vm_end;
if (tmp > end)
tmp = end;
error = mprotect_fixup(vma, &prev, nstart, tmp, newflags);
if (error)
goto out;
nstart = tmp;
if (nstart < prev->vm_end)
nstart = prev->vm_end;
if (nstart >= end)
goto out;
vma = prev->vm_next;
if (!vma || vma->vm_start != nstart) {
error = -ENOMEM;
goto out;
}
prot = reqprot;
}
out:
up_write(&current->mm->mmap_sem);
return error;
}
SYSCALL_DEFINE3(mprotect, unsigned long, start, size_t, len,
unsigned long, prot)
{
return do_mprotect_pkey(start, len, prot, -1);
}
#ifdef CONFIG_ARCH_HAS_PKEYS
SYSCALL_DEFINE4(pkey_mprotect, unsigned long, start, size_t, len,
unsigned long, prot, int, pkey)
{
return do_mprotect_pkey(start, len, prot, pkey);
}
SYSCALL_DEFINE2(pkey_alloc, unsigned long, flags, unsigned long, init_val)
{
int pkey;
int ret;
/* No flags supported yet. */
if (flags)
return -EINVAL;
/* check for unsupported init values */
if (init_val & ~PKEY_ACCESS_MASK)
return -EINVAL;
down_write(&current->mm->mmap_sem);
pkey = mm_pkey_alloc(current->mm);
ret = -ENOSPC;
if (pkey == -1)
goto out;
ret = arch_set_user_pkey_access(current, pkey, init_val);
if (ret) {
mm_pkey_free(current->mm, pkey);
goto out;
}
ret = pkey;
out:
up_write(&current->mm->mmap_sem);
return ret;
}
SYSCALL_DEFINE1(pkey_free, int, pkey)
{
int ret;
down_write(&current->mm->mmap_sem);
ret = mm_pkey_free(current->mm, pkey);
up_write(&current->mm->mmap_sem);
/*
* We could provie warnings or errors if any VMA still
* has the pkey set here.
*/
return ret;
}
#endif /* CONFIG_ARCH_HAS_PKEYS */