For huge pages (and in fact, any compound page), the GUP_PIN_COUNTING_BIAS
scheme tends to overflow too easily, each tail page increments the head
page->_refcount by GUP_PIN_COUNTING_BIAS (1024). That limits the number
of huge pages that can be pinned.
This patch removes that limitation, by using an exact form of pin counting
for compound pages of order > 1. The "order > 1" is required because this
approach uses the 3rd struct page in the compound page, and order 1
compound pages only have two pages, so that won't work there.
A new struct page field, hpage_pinned_refcount, has been added, replacing
a padding field in the union (so no new space is used).
This enhancement also has a useful side effect: huge pages and compound
pages (of order > 1) do not suffer from the "potential false positives"
problem that is discussed in the page_dma_pinned() comment block. That is
because these compound pages have extra space for tracking things, so they
get exact pin counts instead of overloading page->_refcount.
Documentation/core-api/pin_user_pages.rst is updated accordingly.
Suggested-by: Jan Kara <jack@suse.cz>
Signed-off-by: John Hubbard <jhubbard@nvidia.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Jan Kara <jack@suse.cz>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Ira Weiny <ira.weiny@intel.com>
Cc: Jérôme Glisse <jglisse@redhat.com>
Cc: "Matthew Wilcox (Oracle)" <willy@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Jason Gunthorpe <jgg@ziepe.ca>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Michal Hocko <mhocko@suse.com>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Shuah Khan <shuah@kernel.org>
Cc: Vlastimil Babka <vbabka@suse.cz>
Link: http://lkml.kernel.org/r/20200211001536.1027652-8-jhubbard@nvidia.com
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
For these pin_user_pages*() functions, FOLL_PIN is OR'd in with whatever gup
flags the caller provides. The caller is required to pass in a non-null struct
-pages* array, and the function then pin pages by incrementing each by a special
-value. For now, that value is +1, just like get_user_pages*().::
+pages* array, and the function then pins pages by incrementing each by a special
+value: GUP_PIN_COUNTING_BIAS.
+
+For huge pages (and in fact, any compound page of more than 2 pages), the
+GUP_PIN_COUNTING_BIAS scheme is not used. Instead, an exact form of pin counting
+is achieved, by using the 3rd struct page in the compound page. A new struct
+page field, hpage_pinned_refcount, has been added in order to support this.
+
+This approach for compound pages avoids the counting upper limit problems that
+are discussed below. Those limitations would have been aggravated severely by
+huge pages, because each tail page adds a refcount to the head page. And in
+fact, testing revealed that, without a separate hpage_pinned_refcount field,
+page overflows were seen in some huge page stress tests.
+
+This also means that huge pages and compound pages (of order > 1) do not suffer
+from the false positives problem that is mentioned below.::
Function
--------
This also leads to limitations: there are only 31-10==21 bits available for a
counter that increments 10 bits at a time.
-TODO: for 1GB and larger huge pages, this is cutting it close. That's because
-when pin_user_pages() follows such pages, it increments the head page by "1"
-(where "1" used to mean "+1" for get_user_pages(), but now means "+1024" for
-pin_user_pages()) for each tail page. So if you have a 1GB huge page:
-
-* There are 256K (18 bits) worth of 4 KB tail pages.
-* There are 21 bits available to count up via GUP_PIN_COUNTING_BIAS (that is,
- 10 bits at a time)
-* There are 21 - 18 == 3 bits available to count. Except that there aren't,
- because you need to allow for a few normal get_page() calls on the head page,
- as well. Fortunately, the approach of using addition, rather than "hard"
- bitfields, within page->_refcount, allows for sharing these bits gracefully.
- But we're still looking at about 8 references.
-
-This, however, is a missing feature more than anything else, because it's easily
-solved by addressing an obvious inefficiency in the original get_user_pages()
-approach of retrieving pages: stop treating all the pages as if they were
-PAGE_SIZE. Retrieve huge pages as huge pages. The callers need to be aware of
-this, so some work is required. Once that's in place, this limitation mostly
-disappears from view, because there will be ample refcounting range available.
-
* Callers must specifically request "dma-pinned tracking of pages". In other
words, just calling get_user_pages() will not suffice; a new set of functions,
pin_user_page() and related, must be used.
* `Some slow progress on get_user_pages() (Apr 2, 2019) <https://lwn.net/Articles/784574/>`_
* `DMA and get_user_pages() (LPC: Dec 12, 2018) <https://lwn.net/Articles/774411/>`_
* `The trouble with get_user_pages() (Apr 30, 2018) <https://lwn.net/Articles/753027/>`_
+* `LWN kernel index: get_user_pages() <https://lwn.net/Kernel/Index/#Memory_management-get_user_pages>`_
John Hubbard, October, 2019
return page[1].compound_order;
}
+static inline bool hpage_pincount_available(struct page *page)
+{
+ /*
+ * Can the page->hpage_pinned_refcount field be used? That field is in
+ * the 3rd page of the compound page, so the smallest (2-page) compound
+ * pages cannot support it.
+ */
+ page = compound_head(page);
+ return PageCompound(page) && compound_order(page) > 1;
+}
+
+static inline int compound_pincount(struct page *page)
+{
+ VM_BUG_ON_PAGE(!hpage_pincount_available(page), page);
+ page = compound_head(page);
+ return atomic_read(compound_pincount_ptr(page));
+}
+
static inline void set_compound_order(struct page *page, unsigned int order)
{
page[1].compound_order = order;
* refcounts, and b) all the callers of this routine are expected to be able to
* deal gracefully with a false positive.
*
+ * For huge pages, the result will be exactly correct. That's because we have
+ * more tracking data available: the 3rd struct page in the compound page is
+ * used to track the pincount (instead using of the GUP_PIN_COUNTING_BIAS
+ * scheme).
+ *
* For more information, please see Documentation/vm/pin_user_pages.rst.
*
* @page: pointer to page to be queried.
*/
static inline bool page_maybe_dma_pinned(struct page *page)
{
+ if (hpage_pincount_available(page))
+ return compound_pincount(page) > 0;
+
/*
* page_ref_count() is signed. If that refcount overflows, then
* page_ref_count() returns a negative value, and callers will avoid
};
struct { /* Second tail page of compound page */
unsigned long _compound_pad_1; /* compound_head */
- unsigned long _compound_pad_2;
+ atomic_t hpage_pinned_refcount;
/* For both global and memcg */
struct list_head deferred_list;
};
return &page[1].compound_mapcount;
}
+static inline atomic_t *compound_pincount_ptr(struct page *page)
+{
+ return &page[2].hpage_pinned_refcount;
+}
+
/*
* Used for sizing the vmemmap region on some architectures
*/
unsigned int page_mask;
};
+static void hpage_pincount_add(struct page *page, int refs)
+{
+ VM_BUG_ON_PAGE(!hpage_pincount_available(page), page);
+ VM_BUG_ON_PAGE(page != compound_head(page), page);
+
+ atomic_add(refs, compound_pincount_ptr(page));
+}
+
+static void hpage_pincount_sub(struct page *page, int refs)
+{
+ VM_BUG_ON_PAGE(!hpage_pincount_available(page), page);
+ VM_BUG_ON_PAGE(page != compound_head(page), page);
+
+ atomic_sub(refs, compound_pincount_ptr(page));
+}
+
/*
* Return the compound head page with ref appropriately incremented,
* or NULL if that failed.
if (flags & FOLL_GET)
return try_get_compound_head(page, refs);
else if (flags & FOLL_PIN) {
- refs *= GUP_PIN_COUNTING_BIAS;
- return try_get_compound_head(page, refs);
+ /*
+ * 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().
+ *
+ * However, be sure to *also* increment the normal page refcount
+ * field at least once, so that the page really is pinned.
+ */
+ if (!hpage_pincount_available(page))
+ refs *= GUP_PIN_COUNTING_BIAS;
+
+ page = try_get_compound_head(page, refs);
+ if (!page)
+ return NULL;
+
+ if (hpage_pincount_available(page))
+ hpage_pincount_add(page, refs);
+
+ return page;
}
WARN_ON_ONCE(1);
if (flags & FOLL_GET)
return try_get_page(page);
else if (flags & FOLL_PIN) {
+ int refs = 1;
+
page = compound_head(page);
if (WARN_ON_ONCE(page_ref_count(page) <= 0))
return false;
- page_ref_add(page, GUP_PIN_COUNTING_BIAS);
+ if (hpage_pincount_available(page))
+ hpage_pincount_add(page, 1);
+ else
+ refs = GUP_PIN_COUNTING_BIAS;
+
+ /*
+ * Similar to try_grab_compound_head(): even if using the
+ * hpage_pincount_add/_sub() routines, be sure to
+ * *also* increment the normal page refcount field at least
+ * once, so that the page really is pinned.
+ */
+ page_ref_add(page, refs);
}
return true;
#ifdef CONFIG_DEV_PAGEMAP_OPS
static bool __unpin_devmap_managed_user_page(struct page *page)
{
- int count;
+ int count, refs = 1;
if (!page_is_devmap_managed(page))
return false;
- count = page_ref_sub_return(page, GUP_PIN_COUNTING_BIAS);
+ if (hpage_pincount_available(page))
+ hpage_pincount_sub(page, 1);
+ else
+ refs = GUP_PIN_COUNTING_BIAS;
+
+ count = page_ref_sub_return(page, refs);
/*
* devmap page refcounts are 1-based, rather than 0-based: if
*/
void unpin_user_page(struct page *page)
{
+ int refs = 1;
+
page = compound_head(page);
/*
if (__unpin_devmap_managed_user_page(page))
return;
- if (page_ref_sub_and_test(page, GUP_PIN_COUNTING_BIAS))
+ if (hpage_pincount_available(page))
+ hpage_pincount_sub(page, 1);
+ else
+ refs = GUP_PIN_COUNTING_BIAS;
+
+ if (page_ref_sub_and_test(page, refs))
__put_page(page);
}
EXPORT_SYMBOL(unpin_user_page);
static void put_compound_head(struct page *page, int refs, unsigned int flags)
{
- if (flags & FOLL_PIN)
- refs *= GUP_PIN_COUNTING_BIAS;
+ if (flags & FOLL_PIN) {
+ if (hpage_pincount_available(page))
+ hpage_pincount_sub(page, refs);
+ else
+ refs *= GUP_PIN_COUNTING_BIAS;
+ }
VM_BUG_ON_PAGE(page_ref_count(page) < refs, page);
/*
struct page *p = page + 1;
atomic_set(compound_mapcount_ptr(page), 0);
+ if (hpage_pincount_available(page))
+ atomic_set(compound_pincount_ptr(page), 0);
+
for (i = 1; i < nr_pages; i++, p = mem_map_next(p, page, i)) {
clear_compound_head(p);
set_page_refcounted(p);
set_compound_head(p, page);
}
atomic_set(compound_mapcount_ptr(page), -1);
+
+ if (hpage_pincount_available(page))
+ atomic_set(compound_pincount_ptr(page), 0);
}
/*
set_compound_head(p, page);
}
atomic_set(compound_mapcount_ptr(page), -1);
+ if (hpage_pincount_available(page))
+ atomic_set(compound_pincount_ptr(page), 0);
}
#ifdef CONFIG_DEBUG_PAGEALLOC
VM_BUG_ON_PAGE(!PageTransHuge(page), page);
/* increment count (starts at -1) */
atomic_set(compound_mapcount_ptr(page), 0);
+ if (hpage_pincount_available(page))
+ atomic_set(compound_pincount_ptr(page), 0);
+
__inc_node_page_state(page, NR_ANON_THPS);
} else {
/* Anon THP always mapped first with PMD */
{
BUG_ON(address < vma->vm_start || address >= vma->vm_end);
atomic_set(compound_mapcount_ptr(page), 0);
+ if (hpage_pincount_available(page))
+ atomic_set(compound_pincount_ptr(page), 0);
+
__page_set_anon_rmap(page, vma, address, 1);
}
#endif /* CONFIG_HUGETLB_PAGE */
projects / linux-2.6-microblaze.git / commitdiff