diff mbox series

[v3,5/5] hugetlb: Do early cow when page pinned on src mm

Message ID 20210205165406.4655-6-peterx@redhat.com (mailing list archive)
State New, archived
Headers show
Series mm/hugetlb: Early cow on fork, and a few cleanups | expand

Commit Message

Peter Xu Feb. 5, 2021, 4:54 p.m. UTC 
This is the last missing piece of the COW-during-fork effort when there're
pinned pages found.  One can reference 70e806e4e645 ("mm: Do early cow for
pinned pages during fork() for ptes", 2020-09-27) for more information, since
we do similar things here rather than pte this time, but just for hugetlb.

Signed-off-by: Peter Xu <peterx@redhat.com>
---
 mm/hugetlb.c | 66 ++++++++++++++++++++++++++++++++++++++++++++++++----
 1 file changed, 62 insertions(+), 4 deletions(-)

Comments

Mike Kravetz Feb. 9, 2021, 12:09 a.m. UTC | #1 
On 2/5/21 8:54 AM, Peter Xu wrote:
> This is the last missing piece of the COW-during-fork effort when there're
> pinned pages found.  One can reference 70e806e4e645 ("mm: Do early cow for
> pinned pages during fork() for ptes", 2020-09-27) for more information, since
> we do similar things here rather than pte this time, but just for hugetlb.

Thanks for all the changes, the patch looks much better.

I did not look at 70e806e4e645 in detail until now.  That commit had the
'write protect trick' which was removed in subsequent commits.  It took me
a bit of git history tracking to figure out the state of that code today and
the reasons for the subsequent changes.  I guess that was a good way to
educate me. :) 

> 
> Signed-off-by: Peter Xu <peterx@redhat.com>
> ---
>  mm/hugetlb.c | 66 ++++++++++++++++++++++++++++++++++++++++++++++++----
>  1 file changed, 62 insertions(+), 4 deletions(-)

Reviewed-by: Mike Kravetz <mike.kravetz@oracle.com>
Peter Xu Feb. 9, 2021, 2:55 a.m. UTC | #2 
On Mon, Feb 08, 2021 at 04:09:26PM -0800, Mike Kravetz wrote:
> On 2/5/21 8:54 AM, Peter Xu wrote:
> > This is the last missing piece of the COW-during-fork effort when there're
> > pinned pages found.  One can reference 70e806e4e645 ("mm: Do early cow for
> > pinned pages during fork() for ptes", 2020-09-27) for more information, since
> > we do similar things here rather than pte this time, but just for hugetlb.
> 
> Thanks for all the changes, the patch looks much better.
> 
> I did not look at 70e806e4e645 in detail until now.  That commit had the
> 'write protect trick' which was removed in subsequent commits.  It took me
> a bit of git history tracking to figure out the state of that code today and
> the reasons for the subsequent changes.  I guess that was a good way to
> educate me. :) 

Thanks for looking into those details.  I didn't expect that to happen since
after Jason's rework with 57efa1fe5957 ("mm/gup: prevent gup_fast from racing
with COW during fork", 2020-12-15) we can ignore the whole wr-protect idea as a
whole.  I referenced 70e806e4e645 more for the idea of why we do that, and also
copy_present_page() on how it is generally implemented.  That gup-fast race is
definitely tricky on its own.

> 
> > 
> > Signed-off-by: Peter Xu <peterx@redhat.com>
> > ---
> >  mm/hugetlb.c | 66 ++++++++++++++++++++++++++++++++++++++++++++++++----
> >  1 file changed, 62 insertions(+), 4 deletions(-)
> 
> Reviewed-by: Mike Kravetz <mike.kravetz@oracle.com>

Thanks!

I'll post a new version very soon with your r-b, and also a compile warning
fixed in the other patch as reported by Gal.
diff mbox series

Patch

diff --git a/mm/hugetlb.c b/mm/hugetlb.c
index 620700f05ff4..7c1a0ecc130e 100644
--- a/mm/hugetlb.c
+++ b/mm/hugetlb.c
@@ -3727,6 +3727,18 @@  static bool is_hugetlb_entry_hwpoisoned(pte_t pte)
 		return false;
 }
 
+static void
+hugetlb_install_page(struct vm_area_struct *vma, pte_t *ptep, unsigned long addr,
+		     struct page *new_page)
+{
+	__SetPageUptodate(new_page);
+	set_huge_pte_at(vma->vm_mm, addr, ptep, make_huge_pte(vma, new_page, 1));
+	hugepage_add_new_anon_rmap(new_page, vma, addr);
+	hugetlb_count_add(pages_per_huge_page(hstate_vma(vma)), vma->vm_mm);
+	ClearHPageRestoreReserve(new_page);
+	SetHPageMigratable(new_page);
+}
+
 int copy_hugetlb_page_range(struct mm_struct *dst, struct mm_struct *src,
 			    struct vm_area_struct *vma)
 {
@@ -3736,6 +3748,7 @@  int copy_hugetlb_page_range(struct mm_struct *dst, struct mm_struct *src,
 	int cow = is_cow_mapping(vma->vm_flags);
 	struct hstate *h = hstate_vma(vma);
 	unsigned long sz = huge_page_size(h);
+	unsigned long npages = pages_per_huge_page(h);
 	struct address_space *mapping = vma->vm_file->f_mapping;
 	struct mmu_notifier_range range;
 	int ret = 0;
@@ -3784,6 +3797,7 @@  int copy_hugetlb_page_range(struct mm_struct *dst, struct mm_struct *src,
 		spin_lock_nested(src_ptl, SINGLE_DEPTH_NESTING);
 		entry = huge_ptep_get(src_pte);
 		dst_entry = huge_ptep_get(dst_pte);
+again:
 		if (huge_pte_none(entry) || !huge_pte_none(dst_entry)) {
 			/*
 			 * Skip if src entry none.  Also, skip in the
@@ -3807,6 +3821,52 @@  int copy_hugetlb_page_range(struct mm_struct *dst, struct mm_struct *src,
 			}
 			set_huge_swap_pte_at(dst, addr, dst_pte, entry, sz);
 		} else {
+			entry = huge_ptep_get(src_pte);
+			ptepage = pte_page(entry);
+			get_page(ptepage);
+
+			/*
+			 * This is a rare case where we see pinned hugetlb
+			 * pages while they're prone to COW.  We need to do the
+			 * COW earlier during fork.
+			 *
+			 * When pre-allocating the page or copying data, we
+			 * need to be without the pgtable locks since we could
+			 * sleep during the process.
+			 */
+			if (unlikely(page_needs_cow_for_dma(vma, ptepage))) {
+				pte_t src_pte_old = entry;
+				struct page *new;
+
+				spin_unlock(src_ptl);
+				spin_unlock(dst_ptl);
+				/* Do not use reserve as it's private owned */
+				new = alloc_huge_page(vma, addr, 1);
+				if (IS_ERR(new)) {
+					put_page(ptepage);
+					ret = PTR_ERR(new);
+					break;
+				}
+				copy_user_huge_page(new, ptepage, addr, vma,
+						    npages);
+				put_page(ptepage);
+
+				/* Install the new huge page if src pte stable */
+				dst_ptl = huge_pte_lock(h, dst, dst_pte);
+				src_ptl = huge_pte_lockptr(h, src, src_pte);
+				spin_lock_nested(src_ptl, SINGLE_DEPTH_NESTING);
+				entry = huge_ptep_get(src_pte);
+				if (!pte_same(src_pte_old, entry)) {
+					put_page(new);
+					/* dst_entry won't change as in child */
+					goto again;
+				}
+				hugetlb_install_page(vma, dst_pte, addr, new);
+				spin_unlock(src_ptl);
+				spin_unlock(dst_ptl);
+				continue;
+			}
+
 			if (cow) {
 				/*
 				 * No need to notify as we are downgrading page
@@ -3817,12 +3877,10 @@  int copy_hugetlb_page_range(struct mm_struct *dst, struct mm_struct *src,
 				 */
 				huge_ptep_set_wrprotect(src, addr, src_pte);
 			}
-			entry = huge_ptep_get(src_pte);
-			ptepage = pte_page(entry);
-			get_page(ptepage);
+
 			page_dup_rmap(ptepage, true);
 			set_huge_pte_at(dst, addr, dst_pte, entry);
-			hugetlb_count_add(pages_per_huge_page(h), dst);
+			hugetlb_count_add(npages, dst);
 		}
 		spin_unlock(src_ptl);
 		spin_unlock(dst_ptl);