The patch titled
Subject: mm/hugetlb: take refcount under page table lock in follow_huge_pmd()
has been removed from the -mm tree. Its filename was
mm-hugetlb-take-refcount-under-page-table-lock-in-follow_huge_pmd.patch
This patch was dropped because an updated version will be merged
------------------------------------------------------
From: Naoya Horiguchi <n-horiguchi@xxxxxxxxxxxxx>
Subject: mm/hugetlb: take refcount under page table lock in follow_huge_pmd()
We have a race condition between move_pages() and freeing hugepages, where
move_pages() calls follow_page(FOLL_GET) for hugepages internally and
tries to get its refcount without preventing concurrent freeing. This
race crashes the kernel, so this patch fixes it by moving FOLL_GET code
for hugepages into follow_huge_pmd() with taking the page table lock.
This patch passes the following test. And libhugetlbfs test shows no
regression.
$ cat movepages.c
#include <stdio.h>
#include <stdlib.h>
#include <numaif.h>
#define ADDR_INPUT 0x700000000000UL
#define HPS 0x200000
#define PS 0x1000
int main(int argc, char *argv[]) {
int i;
int nr_hp = strtol(argv[1], NULL, 0);
int nr_p = nr_hp * HPS / PS;
int ret;
void **addrs;
int *status;
int *nodes;
pid_t pid;
pid = strtol(argv[2], NULL, 0);
addrs = malloc(sizeof(char *) * nr_p + 1);
status = malloc(sizeof(char *) * nr_p + 1);
nodes = malloc(sizeof(char *) * nr_p + 1);
while (1) {
for (i = 0; i < nr_p; i++) {
addrs[i] = (void *)ADDR_INPUT + i * PS;
nodes[i] = 1;
status[i] = 0;
}
ret = numa_move_pages(pid, nr_p, addrs, nodes, status,
MPOL_MF_MOVE_ALL);
if (ret == -1)
err("move_pages");
for (i = 0; i < nr_p; i++) {
addrs[i] = (void *)ADDR_INPUT + i * PS;
nodes[i] = 0;
status[i] = 0;
}
ret = numa_move_pages(pid, nr_p, addrs, nodes, status,
MPOL_MF_MOVE_ALL);
if (ret == -1)
err("move_pages");
}
return 0;
}
$ cat hugepage.c
#include <stdio.h>
#include <sys/mman.h>
#include <string.h>
#define ADDR_INPUT 0x700000000000UL
#define HPS 0x200000
int main(int argc, char *argv[]) {
int nr_hp = strtol(argv[1], NULL, 0);
char *p;
while (1) {
p = mmap((void *)ADDR_INPUT, nr_hp * HPS, PROT_READ | PROT_WRITE,
MAP_PRIVATE | MAP_ANONYMOUS | MAP_HUGETLB, -1, 0);
if (p != (void *)ADDR_INPUT) {
perror("mmap");
break;
}
memset(p, 0, nr_hp * HPS);
munmap(p, nr_hp * HPS);
}
}
$ sysctl vm.nr_hugepages=40
$ ./hugepage 10 &
$ ./movepages 10 $(pgrep -f hugepage)
Note for stable inclusion:
This patch fixes e632a938d914 ("mm: migrate: add hugepage migration code
to move_pages()"), so is applicable to -stable kernels which includes it.
Signed-off-by: Naoya Horiguchi <n-horiguchi@xxxxxxxxxxxxx>
Reported-by: Hugh Dickins <hughd@xxxxxxxxxx>
Cc: David Rientjes <rientjes@xxxxxxxxxx>
Cc: Christoph Lameter <cl@xxxxxxxxx>
Cc: Mel Gorman <mel@xxxxxxxxx>
Cc: <stable@xxxxxxxxxxxxxxx> [3.12+]
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
include/linux/hugetlb.h | 3 +++
mm/gup.c | 17 ++---------------
mm/hugetlb.c | 35 +++++++++++++++++++++++++++++++++++
3 files changed, 40 insertions(+), 15 deletions(-)
diff -puN include/linux/hugetlb.h~mm-hugetlb-take-refcount-under-page-table-lock-in-follow_huge_pmd include/linux/hugetlb.h
--- a/include/linux/hugetlb.h~mm-hugetlb-take-refcount-under-page-table-lock-in-follow_huge_pmd
+++ a/include/linux/hugetlb.h
@@ -102,6 +102,8 @@ struct page *follow_huge_pmd(struct mm_s
pmd_t *pmd, int write);
struct page *follow_huge_pud(struct mm_struct *mm, unsigned long address,
pud_t *pud, int write);
+struct page *follow_huge_pmd_lock(struct vm_area_struct *vma,
+ unsigned long address, pmd_t *pmd, int flags);
int pmd_huge(pmd_t pmd);
int pud_huge(pud_t pmd);
unsigned long hugetlb_change_protection(struct vm_area_struct *vma,
@@ -135,6 +137,7 @@ static inline void hugetlb_show_meminfo(
}
#define follow_huge_pmd(mm, addr, pmd, write) NULL
#define follow_huge_pud(mm, addr, pud, write) NULL
+#define follow_huge_pmd_lock(vma, addr, pmd, flags) NULL
#define prepare_hugepage_range(file, addr, len) (-EINVAL)
#define pmd_huge(x) 0
#define pud_huge(x) 0
diff -puN mm/gup.c~mm-hugetlb-take-refcount-under-page-table-lock-in-follow_huge_pmd mm/gup.c
--- a/mm/gup.c~mm-hugetlb-take-refcount-under-page-table-lock-in-follow_huge_pmd
+++ a/mm/gup.c
@@ -174,21 +174,8 @@ struct page *follow_page_mask(struct vm_
pmd = pmd_offset(pud, address);
if (pmd_none(*pmd))
return no_page_table(vma, flags);
- if (pmd_huge(*pmd) && vma->vm_flags & VM_HUGETLB) {
- page = follow_huge_pmd(mm, address, pmd, flags & FOLL_WRITE);
- if (flags & FOLL_GET) {
- /*
- * Refcount on tail pages are not well-defined and
- * shouldn't be taken. The caller should handle a NULL
- * return when trying to follow tail pages.
- */
- if (PageHead(page))
- get_page(page);
- else
- page = NULL;
- }
- return page;
- }
+ if (pmd_huge(*pmd) && vma->vm_flags & VM_HUGETLB)
+ return follow_huge_pmd_lock(vma, address, pmd, flags);
if ((flags & FOLL_NUMA) && pmd_numa(*pmd))
return no_page_table(vma, flags);
if (pmd_trans_huge(*pmd)) {
diff -puN mm/hugetlb.c~mm-hugetlb-take-refcount-under-page-table-lock-in-follow_huge_pmd mm/hugetlb.c
--- a/mm/hugetlb.c~mm-hugetlb-take-refcount-under-page-table-lock-in-follow_huge_pmd
+++ a/mm/hugetlb.c
@@ -3690,6 +3690,41 @@ follow_huge_pud(struct mm_struct *mm, un
#endif /* CONFIG_ARCH_WANT_GENERAL_HUGETLB */
+/*
+ * This function calls the architecture dependent variant follow_huge_pmd()
+ * with holding page table lock depending on FOLL_GET.
+ * Whether hugepage migration is supported or not, follow() can be called
+ * with FOLL_GET from do_move_page_to_node_array(), so we need do this in
+ * common code.
+ * Should be called under read mmap_sem.
+ */
+struct page *follow_huge_pmd_lock(struct vm_area_struct *vma,
+ unsigned long address, pmd_t *pmd, int flags)
+{
+ struct page *page;
+ spinlock_t *ptl;
+
+ if (flags & FOLL_GET)
+ ptl = huge_pte_lock(hstate_vma(vma), vma->vm_mm, (pte_t *)pmd);
+
+ page = follow_huge_pmd(vma->vm_mm, address, pmd, flags & FOLL_WRITE);
+
+ if (flags & FOLL_GET) {
+ /*
+ * Refcount on tail pages are not well-defined and
+ * shouldn't be taken. The caller should handle a NULL
+ * return when trying to follow tail pages.
+ */
+ if (PageHead(page))
+ get_page(page);
+ else
+ page = NULL;
+ spin_unlock(ptl);
+ }
+
+ return page;
+}
+
#ifdef CONFIG_MEMORY_FAILURE
/* Should be called in hugetlb_lock */
_
Patches currently in -mm which might be from n-horiguchi@xxxxxxxxxxxxx are
origin.patch
memory-hotplug-add-sysfs-zones_online_to-attribute.patch
mm-hugetlb-use-get_page_unless_zero-in-hugetlb_fault.patch
mm-hugetlb-add-migration-entry-check-in-hugetlb_change_protection.patch
mm-hugetlb-remove-unused-argument-of-follow_huge_pmdpud.patch
mm-introduce-do_shared_fault-and-drop-do_fault-fix-fix.patch
do_shared_fault-check-that-mmap_sem-is-held.patch
--
To unsubscribe from this list: send the line "unsubscribe stable" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html