The patch titled
mm: follow_hugetlb_page flags
has been added to the -mm tree. Its filename is
mm-follow_hugetlb_page-flags.patch
Before you just go and hit "reply", please:
a) Consider who else should be cc'ed
b) Prefer to cc a suitable mailing list as well
c) Ideally: find the original patch on the mailing list and do a
reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/SubmitChecklist when testing your code ***
See http://userweb.kernel.org/~akpm/stuff/added-to-mm.txt to find
out what to do about this
The current -mm tree may be found at http://userweb.kernel.org/~akpm/mmotm/
------------------------------------------------------
Subject: mm: follow_hugetlb_page flags
From: Hugh Dickins <hugh.dickins@xxxxxxxxxxxxx>
follow_hugetlb_page() shouldn't be guessing about the coredump case
either: pass the foll_flags down to it, instead of just the write bit.
Remove that obscure huge_zeropage_ok() test. The decision is easy,
though unlike the non-huge case - here vm_ops->fault is always set.
But we know that a fault would serve up zeroes, unless there's
already a hugetlbfs pagecache page to back the range.
(Alternatively, since hugetlb pages aren't swapped out under pressure,
you could save more dump space by arguing that a page not yet faulted
into this process cannot be relevant to the dump; but that would be
more surprising.)
Signed-off-by: Hugh Dickins <hugh.dickins@xxxxxxxxxxxxx>
Acked-by: Rik van Riel <riel@xxxxxxxxxx>
Cc: KAMEZAWA Hiroyuki <kamezawa.hiroyu@xxxxxxxxxxxxxx>
Cc: KOSAKI Motohiro <kosaki.motohiro@xxxxxxxxxxxxxx>
Cc: Nick Piggin <npiggin@xxxxxxx>
Cc: Mel Gorman <mel@xxxxxxxxx>
Cc: Minchan Kim <minchan.kim@xxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
include/linux/hugetlb.h | 4 +-
mm/hugetlb.c | 62 ++++++++++++++++++++++----------------
mm/memory.c | 14 ++++----
3 files changed, 48 insertions(+), 32 deletions(-)
diff -puN include/linux/hugetlb.h~mm-follow_hugetlb_page-flags include/linux/hugetlb.h
--- a/include/linux/hugetlb.h~mm-follow_hugetlb_page-flags
+++ a/include/linux/hugetlb.h
@@ -24,7 +24,9 @@ int hugetlb_sysctl_handler(struct ctl_ta
int hugetlb_overcommit_handler(struct ctl_table *, int, struct file *, void __user *, size_t *, loff_t *);
int hugetlb_treat_movable_handler(struct ctl_table *, int, struct file *, void __user *, size_t *, loff_t *);
int copy_hugetlb_page_range(struct mm_struct *, struct mm_struct *, struct vm_area_struct *);
-int follow_hugetlb_page(struct mm_struct *, struct vm_area_struct *, struct page **, struct vm_area_struct **, unsigned long *, int *, int, int);
+int follow_hugetlb_page(struct mm_struct *, struct vm_area_struct *,
+ struct page **, struct vm_area_struct **,
+ unsigned long *, int *, int, unsigned int flags);
void unmap_hugepage_range(struct vm_area_struct *,
unsigned long, unsigned long, struct page *);
void __unmap_hugepage_range(struct vm_area_struct *,
diff -puN mm/hugetlb.c~mm-follow_hugetlb_page-flags mm/hugetlb.c
--- a/mm/hugetlb.c~mm-follow_hugetlb_page-flags
+++ a/mm/hugetlb.c
@@ -2016,6 +2016,23 @@ static struct page *hugetlbfs_pagecache_
return find_lock_page(mapping, idx);
}
+/* Return whether there is a pagecache page to back given address within VMA */
+static bool hugetlbfs_backed(struct hstate *h,
+ struct vm_area_struct *vma, unsigned long address)
+{
+ struct address_space *mapping;
+ pgoff_t idx;
+ struct page *page;
+
+ mapping = vma->vm_file->f_mapping;
+ idx = vma_hugecache_offset(h, vma, address);
+
+ page = find_get_page(mapping, idx);
+ if (page)
+ put_page(page);
+ return page != NULL;
+}
+
static int hugetlb_no_page(struct mm_struct *mm, struct vm_area_struct *vma,
unsigned long address, pte_t *ptep, unsigned int flags)
{
@@ -2211,54 +2228,52 @@ follow_huge_pud(struct mm_struct *mm, un
return NULL;
}
-static int huge_zeropage_ok(pte_t *ptep, int write, int shared)
-{
- if (!ptep || write || shared)
- return 0;
- else
- return huge_pte_none(huge_ptep_get(ptep));
-}
-
int follow_hugetlb_page(struct mm_struct *mm, struct vm_area_struct *vma,
struct page **pages, struct vm_area_struct **vmas,
unsigned long *position, int *length, int i,
- int write)
+ unsigned int flags)
{
unsigned long pfn_offset;
unsigned long vaddr = *position;
int remainder = *length;
struct hstate *h = hstate_vma(vma);
- int zeropage_ok = 0;
- int shared = vma->vm_flags & VM_SHARED;
spin_lock(&mm->page_table_lock);
while (vaddr < vma->vm_end && remainder) {
pte_t *pte;
+ int absent;
struct page *page;
/*
* Some archs (sparc64, sh*) have multiple pte_ts to
- * each hugepage. We have to make * sure we get the
+ * each hugepage. We have to make sure we get the
* first, for the page indexing below to work.
*/
pte = huge_pte_offset(mm, vaddr & huge_page_mask(h));
- if (huge_zeropage_ok(pte, write, shared))
- zeropage_ok = 1;
+ absent = !pte || huge_pte_none(huge_ptep_get(pte));
+
+ /*
+ * When coredumping, it suits get_dump_page if we just return
+ * an error if there's a hole and no huge pagecache to back it.
+ */
+ if (absent &&
+ ((flags & FOLL_DUMP) && !hugetlbfs_backed(h, vma, vaddr))) {
+ remainder = 0;
+ break;
+ }
- if (!pte ||
- (huge_pte_none(huge_ptep_get(pte)) && !zeropage_ok) ||
- (write && !pte_write(huge_ptep_get(pte)))) {
+ if (absent ||
+ ((flags & FOLL_WRITE) && !pte_write(huge_ptep_get(pte)))) {
int ret;
spin_unlock(&mm->page_table_lock);
- ret = hugetlb_fault(mm, vma, vaddr, write);
+ ret = hugetlb_fault(mm, vma, vaddr,
+ (flags & FOLL_WRITE) ? FAULT_FLAG_WRITE : 0);
spin_lock(&mm->page_table_lock);
if (!(ret & VM_FAULT_ERROR))
continue;
remainder = 0;
- if (!i)
- i = -EFAULT;
break;
}
@@ -2266,10 +2281,7 @@ int follow_hugetlb_page(struct mm_struct
page = pte_page(huge_ptep_get(pte));
same_page:
if (pages) {
- if (zeropage_ok)
- pages[i] = ZERO_PAGE(0);
- else
- pages[i] = mem_map_offset(page, pfn_offset);
+ pages[i] = mem_map_offset(page, pfn_offset);
get_page(pages[i]);
}
@@ -2293,7 +2305,7 @@ same_page:
*length = remainder;
*position = vaddr;
- return i;
+ return i ? i : -EFAULT;
}
void hugetlb_change_protection(struct vm_area_struct *vma,
diff -puN mm/memory.c~mm-follow_hugetlb_page-flags mm/memory.c
--- a/mm/memory.c~mm-follow_hugetlb_page-flags
+++ a/mm/memory.c
@@ -1260,17 +1260,19 @@ int __get_user_pages(struct task_struct
!(vm_flags & vma->vm_flags))
return i ? : -EFAULT;
- if (is_vm_hugetlb_page(vma)) {
- i = follow_hugetlb_page(mm, vma, pages, vmas,
- &start, &nr_pages, i, write);
- continue;
- }
-
foll_flags = FOLL_TOUCH;
if (pages)
foll_flags |= FOLL_GET;
if (flags & GUP_FLAGS_DUMP)
foll_flags |= FOLL_DUMP;
+ if (write)
+ foll_flags |= FOLL_WRITE;
+
+ if (is_vm_hugetlb_page(vma)) {
+ i = follow_hugetlb_page(mm, vma, pages, vmas,
+ &start, &nr_pages, i, foll_flags);
+ continue;
+ }
do {
struct page *page;
_
Patches currently in -mm which might be from hugh.dickins@xxxxxxxxxxxxx are
linux-next.patch
acpi-fix-null-bug-for-hid-uid-string-2.patch
vfs-optimize-touch_time-too-fix.patch
fs-new-truncate-helpers.patch
fs-use-new-truncate-helpers.patch
fs-introduce-new-truncate-sequence.patch
fs-convert-simple-fs-to-new-truncate.patch
tmpfs-convert-to-use-the-new-truncate-convention.patch
ext2-convert-to-use-the-new-truncate-convention.patch
fat-convert-to-use-the-new-truncate-convention.patch
btrfs-convert-to-use-the-new-truncate-convention.patch
jfs-convert-to-use-the-new-truncate-convention.patch
udf-convert-to-use-the-new-truncate-convention.patch
minix-convert-to-use-the-new-truncate-convention.patch
vfs-seq_file-add-helpers-for-data-filling.patch
vfs-revert-proc-mounts-to-old-behavior-for-unreachable-mountpoints.patch
vfs-no-unreachable-prefix-for-sysvipc-maps-in-proc-pid-maps.patch
mm-oom-analysis-add-shmem-vmstat.patch
ksm-add-mmu_notifier-set_pte_at_notify.patch
ksm-first-tidy-up-madvise_vma.patch
ksm-define-madv_mergeable-and-madv_unmergeable.patch
ksm-the-mm-interface-to-ksm.patch
ksm-no-debug-in-page_dup_rmap.patch
ksm-identify-pageksm-pages.patch
ksm-kernel-samepage-merging.patch
ksm-prevent-mremap-move-poisoning.patch
ksm-change-copyright-message.patch
ksm-change-ksm-nice-level-to-be-5.patch
ksm-rename-kernel_pages_allocated.patch
ksm-move-pages_sharing-updates.patch
ksm-pages_unshared-and-pages_volatile.patch
ksm-break-cow-once-unshared.patch
ksm-keep-quiet-while-list-empty.patch
ksm-five-little-cleanups.patch
ksm-fix-endless-loop-on-oom.patch
ksm-distribute-remove_mm_from_lists.patch
ksm-fix-oom-deadlock.patch
ksm-fix-deadlock-with-munlock-in-exit_mmap.patch
ksm-sysfs-and-defaults.patch
ksm-add-some-documentation.patch
ksm-remove-vm_mergeable_flags.patch
ksm-clean-up-obsolete-references.patch
ksm-unmerge-is-an-origin-of-ooms.patch
ksm-mremap-use-err-from-ksm_madvise.patch
mm-add_to_swap_cache-must-not-sleep.patch
mm-add_to_swap_cache-does-not-return-eexist.patch
mm-includecheck-fix-for-mm-shmemc.patch
mm-introduce-page_lru_base_type-fix.patch
mm-replace-various-uses-of-num_physpages-by-totalram_pages.patch
hugetlbfs-allow-the-creation-of-files-suitable-for-map_private-on-the-vfs-internal-mount.patch
hugetlb-add-map_hugetlb-for-mmaping-pseudo-anonymous-huge-page-regions.patch
hugetlb-add-map_hugetlb-example.patch
mm-munlock-use-follow_page.patch
mm-remove-unused-gup-flags.patch
mm-add-get_dump_page.patch
mm-foll_dump-replace-foll_anon.patch
mm-follow_hugetlb_page-flags.patch
mm-fix-anonymous-dirtying.patch
mm-reinstate-zero_page.patch
mm-foll-flags-for-gup-flags.patch
getrusage-fill-ru_maxrss-value.patch
getrusage-fill-ru_maxrss-value-update.patch
ramfs-move-ramfs_magic-to-include-linux-magich.patch
memory-controller-soft-limit-organize-cgroups-v9-fix.patch
prio_tree-debugging-patch.patch
--
To unsubscribe from this list: send the line "unsubscribe mm-commits" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
