Skip to contentSkip to conversation

fix omission of check on FOLL_LONGTERM in gup fast path

PATCH Applied Pingfan Liu 2 participants 4 messages Mar 22, 2020 – Mar 23, 2020

Subsystems (2)
Memory management MM GUP
lore
Share & tools
lore

Pingfan Liu

author
#0 lore raw

v8 cover fix omission of check on FOLL_LONGTERM in gup fast path

v7 -> v8:
- drop the [3/3] in v7, which is a test case without real use case now.
- apply renamming in __get_user_pages_fast() besides internal_get_user_pages_fast()
- improve the note in code with [2/2]

Cc: Ira Weiny <[email protected]>
Cc: Andrew Morton <[email protected]>
Cc: Mike Rapoport <[email protected]>
Cc: Dan Williams <[email protected]>
Cc: Matthew Wilcox <[email protected]>
Cc: John Hubbard <[email protected]>
Cc: "Aneesh Kumar K.V" <[email protected]>
Cc: Christoph Hellwig <[email protected]>
Cc: Shuah Khan <[email protected]>
Cc: Jason Gunthorpe <[email protected]>
To: [email protected]
Cc: [email protected]

Pingfan Liu (2):
mm/gup: rename nr as nr_pinned in get_user_pages_fast() mm/gup: fix omission of check on FOLL_LONGTERM in gup fast path

 mm/gup.c | 34 +++++++++++++++++++++-------------
 1 file changed, 21 insertions(+), 13 deletions(-)

--
2.7.5

author
#1 lore raw

PATCH v8 1/2 mm/gup: rename nr as nr_pinned in get_user_pages_fast()

To better reflect the held state of pages and make code self-explaining, rename nr as nr_pinned.

Signed-off-by Pingfan Liu <[email protected]>
Reviewed-by John Hubbard <[email protected]>
Cc Ira Weiny <[email protected]>
Cc Andrew Morton <[email protected]>
Cc Mike Rapoport <[email protected]>
Cc Dan Williams <[email protected]>
Cc Matthew Wilcox <[email protected]>
Cc "Aneesh Kumar K.V" <[email protected]>
Cc Christoph Hellwig <[email protected]>
Cc Shuah Khan <[email protected]>
Cc Jason Gunthorpe <[email protected]>

To: [email protected]
Cc: [email protected]

---
 mm/gup.c | 26 +++++++++++++-------------
 1 file changed, 13 insertions(+), 13 deletions(-)
mm/gup.c +13 -13
index e8aaa40..02a95b1 100644
@@ -2665,7 +2665,7 @@ int __get_user_pages_fast(unsigned long start, int nr_pages, int write,
 {
 	unsigned long len, end;
 	unsigned long flags;
-	int nr = 0;
+	int nr_pinned = 0;
 	/*
 	 * Internally (within mm/gup.c), gup fast variants must set FOLL_GET,
 	 * because gup fast is always a "pin with a +1 page refcount" request.

@@ -2699,11 +2699,11 @@ int __get_user_pages_fast(unsigned long start, int nr_pages, int write,
 	if (IS_ENABLED(CONFIG_HAVE_FAST_GUP) &&
 	    gup_fast_permitted(start, end)) {
 		local_irq_save(flags);
-		gup_pgd_range(start, end, gup_flags, pages, &nr);
+		gup_pgd_range(start, end, gup_flags, pages, &nr_pinned);
 		local_irq_restore(flags);
 	}
 
-	return nr;
+	return nr_pinned;
 }
 EXPORT_SYMBOL_GPL(__get_user_pages_fast);
 

@@ -2735,7 +2735,7 @@ static int internal_get_user_pages_fast(unsigned long start, int nr_pages,
 					struct page **pages)
 {
 	unsigned long addr, len, end;
-	int nr = 0, ret = 0;
+	int nr_pinned = 0, ret = 0;
 
 	if (WARN_ON_ONCE(gup_flags & ~(FOLL_WRITE | FOLL_LONGTERM |
 				       FOLL_FORCE | FOLL_PIN | FOLL_GET)))

@@ -2754,25 +2754,25 @@ static int internal_get_user_pages_fast(unsigned long start, int nr_pages,
 	if (IS_ENABLED(CONFIG_HAVE_FAST_GUP) &&
 	    gup_fast_permitted(start, end)) {
 		local_irq_disable();
-		gup_pgd_range(addr, end, gup_flags, pages, &nr);
+		gup_pgd_range(addr, end, gup_flags, pages, &nr_pinned);
 		local_irq_enable();
-		ret = nr;
+		ret = nr_pinned;
 	}
 
-	if (nr < nr_pages) {
+	if (nr_pinned < nr_pages) {
 		/* Try to get the remaining pages with get_user_pages */
-		start += nr << PAGE_SHIFT;
-		pages += nr;
+		start += nr_pinned << PAGE_SHIFT;
+		pages += nr_pinned;
 
-		ret = __gup_longterm_unlocked(start, nr_pages - nr,
+		ret = __gup_longterm_unlocked(start, nr_pages - nr_pinned,
 					      gup_flags, pages);
 
 		/* Have to be a bit careful with return values */
-		if (nr > 0) {
+		if (nr_pinned > 0) {
 			if (ret < 0)
-				ret = nr;
+				ret = nr_pinned;
 			else
-				ret += nr;
+				ret += nr_pinned;
 		}
 	}

--
2.7.5

author
#2 lore raw

PATCH v8 2/2 mm/gup: fix omission of check on FOLL_LONGTERM in gup fast path

In reply to Pingfan Liu

FOLL_LONGTERM is a special case of FOLL_PIN. It suggests a pin which is going to be given to hardware and can't move. It would truncate CMA permanently and should be excluded.

In gup slow path, where __gup_longterm_locked->check_and_migrate_cma_pages() handles FOLL_LONGTERM, but in fast path, there lacks such a check, which means a possible leak of CMA page to longterm pinned.

Place a check in try_grab_compound_head() in the fast path to fix the leak, and if FOLL_LONGTERM happens on CMA, it will fall back to slow path to migrate the page.

Some note about the check:
Huge page's subpages have the same migrate type due to either allocation from a free_list[] or alloc_contig_range() with param MIGRATE_MOVABLE. So it is enough to check on a single subpage by is_migrate_cma_page(subpage)

Signed-off-by Pingfan Liu <[email protected]>
Reviewed-by Christoph Hellwig <[email protected]>
Reviewed-by Jason Gunthorpe <[email protected]>
Cc Ira Weiny <[email protected]>
Cc Andrew Morton <[email protected]>
Cc Mike Rapoport <[email protected]>
Cc Dan Williams <[email protected]>
Cc Matthew Wilcox <[email protected]>
Cc John Hubbard <[email protected]>
Cc "Aneesh Kumar K.V" <[email protected]>
Cc Christoph Hellwig <[email protected]>
Cc Shuah Khan <[email protected]>
Cc Jason Gunthorpe <[email protected]>

To: [email protected]
Cc: [email protected]

---
 mm/gup.c | 8 ++++++++
 1 file changed, 8 insertions(+)
mm/gup.c +8 -0
index 02a95b1..3fe75c4 100644
@@ -89,6 +89,14 @@ static __maybe_unused struct page *try_grab_compound_head(struct page *page,
 		int orig_refs = refs;
 
 		/*
+		 * Can't do FOLL_LONGTERM + FOLL_PIN with CMA in the gup fast
+		 * path, so fail and let the caller fall back to the slow path.
+		 */
+		if (unlikely(flags & FOLL_LONGTERM) &&
+				is_migrate_cma_page(page))
+			return NULL;
+
+		/*
 		 * When pinning a compound page of order > 1 (which is what
 		 * hpage_pincount_available() checks for), use an exact count to
 		 * track it, via hpage_pincount_add/_sub().

--
2.7.5

reviewer Reviewed-by on v8 2/2
#3 lore raw

Re: [PATCHv8 2/2] mm/gup: fix omission of check on FOLL_LONGTERM in gup fast path

Quoted 44 lines · Pingfan Liu

FOLL_LONGTERM is a special case of FOLL_PIN. It suggests a pin which is going to be given to hardware and can't move. It would truncate CMA permanently and should be excluded.

In gup slow path, where __gup_longterm_locked->check_and_migrate_cma_pages() handles FOLL_LONGTERM, but in fast path, there lacks such a check, which means a possible leak of CMA page to longterm pinned.

Place a check in try_grab_compound_head() in the fast path to fix the leak, and if FOLL_LONGTERM happens on CMA, it will fall back to slow path to migrate the page.

Some note about the check:
Huge page's subpages have the same migrate type due to either allocation from a free_list[] or alloc_contig_range() with param MIGRATE_MOVABLE. So it is enough to check on a single subpage by is_migrate_cma_page(subpage)

Signed-off-by Pingfan Liu <[email protected]>
Reviewed-by Christoph Hellwig <[email protected]>
Reviewed-by Jason Gunthorpe <[email protected]>
Cc Ira Weiny <[email protected]>
Cc Andrew Morton <[email protected]>
Cc Mike Rapoport <[email protected]>
Cc Dan Williams <[email protected]>
Cc Matthew Wilcox <[email protected]>
Cc John Hubbard <[email protected]>
Cc "Aneesh Kumar K.V" <[email protected]>
Cc Christoph Hellwig <[email protected]>
Cc Shuah Khan <[email protected]>
Cc Jason Gunthorpe <[email protected]>

To: [email protected]
Cc: [email protected]

---
  mm/gup.c | 8 ++++++++
  1 file changed, 8 insertions(+)
mm/gup.c +0 -0
index 02a95b1..3fe75c4 100644
@@ -89,6 +89,14 @@ static __maybe_unused struct page *try_grab_compound_head(struct page *page,
  		int orig_refs = refs;

mm/gup.c static __maybe_unused struct page *try_grab_compound_head(struct page *page, · line 99

Quoted 10 lines
mm/gup.c +8 -0
 
  		/*
+		 * Can't do FOLL_LONGTERM + FOLL_PIN with CMA in the gup fast
+		 * path, so fail and let the caller fall back to the slow path.
+		 */
+		if (unlikely(flags & FOLL_LONGTERM) &&
+				is_migrate_cma_page(page))
+			return NULL;
+
+		/*
Reviewed-by John Hubbard <[email protected]>

thanks,

Signature · 2 lines
John Hubbard
NVIDIA
Quoted 7 lines
mm/gup.c +0 -1
  		 * When pinning a compound page of order > 1 (which is what
  		 * hpage_pincount_available() checks for), use an exact count to
  		 * track it, via hpage_pincount_add/_sub().
--

2.7.5