[PATCH v6 0/5] mm/gup: Introduce memfd_pin_user_pages() for pinning memfd pages (v6)

[PATCH v6 0/5] mm/gup: Introduce memfd_pin_user_pages() for pinning memfd pages (v6)

[Vivek Kasireddy]

The first two patches were previously reviewed but not yet merged.
These ones need to be merged first as the fourth patch depends on
the changes introduced in them and they also fix bugs seen in
very specific scenarios (running Qemu with hugetlb=on, blob=true
and rebooting guest VM).

The third patch introduces memfd_pin_user_pages() API and the fourth
patch shows how the udmabuf driver can make use of it to
longterm-pin the the pages. The last patch adds two new udmabuf
selftests to verify data coherency after potential page migration.

v2:
- Updated the first patch to include review feedback from David and
  Jason. The main change in this series is the allocation of page
  in the case of hugetlbfs if it is not found in the page cache.

v3:
- Made changes to include review feedback from David to improve the
  comments and readability of code
- Enclosed the hugepage alloc code with #ifdef CONFIG_HUGETLB_PAGE

v4:
- Augmented the commit message of the udmabuf patch that uses
  pin_user_pages_fd()
- Added previously reviewed but unmerged udmabuf patches to this
  series

v5:
- Updated the patch that adds pin_user_pages_fd() to include feedback
  from David to handle simultaneous users trying to add a huge page
  to the mapping
- Replaced find_get_page_flags() with __filemap_get_folio() in the
  second and third patches to ensure that we only obtain head pages
  from the mapping

v6: (Christoph)
- Renamed the new API to memfd_pin_user_pages()
- Improved the page cache lookup efficiency by using
  filemap_get_folios_contig() which uses batches

This series is tested using following methods:
- Run the subtests added in the fifth patch
- Run Qemu (master) with the following options and a few additional
  patches to Spice:
  qemu-system-x86_64 -m 4096m....
  -device virtio-gpu-pci,max_outputs=1,blob=true,xres=1920,yres=1080
  -spice port=3001,gl=on,disable-ticketing=on,preferred-codec=gstreamer:h264
  -object memory-backend-memfd,hugetlb=on,id=mem1,size=4096M
  -machine memory-backend=mem1

Cc: David Hildenbrand <david@redhat.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: Gerd Hoffmann <kraxel@redhat.com>
Cc: Dongwon Kim <dongwon.kim@intel.com>
Cc: Junxiao Chang <junxiao.chang@intel.com>

Vivek Kasireddy (5):
  udmabuf: Use vmf_insert_pfn and VM_PFNMAP for handling mmap
  udmabuf: Add back support for mapping hugetlb pages (v5)
  mm/gup: Introduce memfd_pin_user_pages() for pinning memfd pages (v6)
  udmabuf: Pin the pages using memfd_pin_user_pages() API (v4)
  selftests/dma-buf/udmabuf: Add tests to verify data after page
    migration

 drivers/dma-buf/udmabuf.c                     |  98 +++++++++---
 include/linux/memfd.h                         |   5 +
 include/linux/mm.h                            |   2 +
 mm/gup.c                                      | 102 ++++++++++++
 mm/memfd.c                                    |  34 ++++
 .../selftests/drivers/dma-buf/udmabuf.c       | 151 +++++++++++++++++-
 6 files changed, 363 insertions(+), 29 deletions(-)

-- 
2.39.2

[PATCH v6 1/5] udmabuf: Use vmf_insert_pfn and VM_PFNMAP for handling mmap

[Vivek Kasireddy]

Add VM_PFNMAP to vm_flags in the mmap handler to ensure that
the mappings would be managed without using struct page.

And, in the vm_fault handler, use vmf_insert_pfn to share the
page's pfn to userspace instead of directly sharing the page
(via struct page *).

Cc: David Hildenbrand <david@redhat.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: Gerd Hoffmann <kraxel@redhat.com>
Cc: Dongwon Kim <dongwon.kim@intel.com>
Cc: Junxiao Chang <junxiao.chang@intel.com>
Suggested-by: David Hildenbrand <david@redhat.com>
Acked-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Vivek Kasireddy <vivek.kasireddy@intel.com>
---
 drivers/dma-buf/udmabuf.c | 8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)

diff --git a/drivers/dma-buf/udmabuf.c b/drivers/dma-buf/udmabuf.c
index c40645999648..820c993c8659 100644
--- a/drivers/dma-buf/udmabuf.c
+++ b/drivers/dma-buf/udmabuf.c
@@ -35,12 +35,13 @@ static vm_fault_t udmabuf_vm_fault(struct vm_fault *vmf)
 	struct vm_area_struct *vma = vmf->vma;
 	struct udmabuf *ubuf = vma->vm_private_data;
 	pgoff_t pgoff = vmf->pgoff;
+	unsigned long pfn;
 
 	if (pgoff >= ubuf->pagecount)
 		return VM_FAULT_SIGBUS;
-	vmf->page = ubuf->pages[pgoff];
-	get_page(vmf->page);
-	return 0;
+
+	pfn = page_to_pfn(ubuf->pages[pgoff]);
+	return vmf_insert_pfn(vma, vmf->address, pfn);
 }
 
 static const struct vm_operations_struct udmabuf_vm_ops = {
@@ -56,6 +57,7 @@ static int mmap_udmabuf(struct dma_buf *buf, struct vm_area_struct *vma)
 
 	vma->vm_ops = &udmabuf_vm_ops;
 	vma->vm_private_data = ubuf;
+	vm_flags_set(vma, VM_PFNMAP | VM_DONTEXPAND | VM_DONTDUMP);
 	return 0;
 }
 
-- 
2.39.2

[PATCH v6 2/5] udmabuf: Add back support for mapping hugetlb pages (v5)

[Vivek Kasireddy]

A user or admin can configure a VMM (Qemu) Guest's memory to be
backed by hugetlb pages for various reasons. However, a Guest OS
would still allocate (and pin) buffers that are backed by regular
4k sized pages. In order to map these buffers and create dma-bufs
for them on the Host, we first need to find the hugetlb pages where
the buffer allocations are located and then determine the offsets
of individual chunks (within those pages) and use this information
to eventually populate a scatterlist.

Testcase: default_hugepagesz=2M hugepagesz=2M hugepages=2500 options
were passed to the Host kernel and Qemu was launched with these
relevant options: qemu-system-x86_64 -m 4096m....
-device virtio-gpu-pci,max_outputs=1,blob=true,xres=1920,yres=1080
-display gtk,gl=on
-object memory-backend-memfd,hugetlb=on,id=mem1,size=4096M
-machine memory-backend=mem1

Replacing -display gtk,gl=on with -display gtk,gl=off above would
exercise the mmap handler.

v2: Updated get_sg_table() to manually populate the scatterlist for
    both huge page and non-huge-page cases.

v3: s/offsets/subpgoff/g
    s/hpoff/mapidx/g

v4: Replaced find_get_page_flags() with __filemap_get_folio() to
    ensure that we only obtain head pages from the mapping

v5: Fix the calculation of mapidx to ensure that it is a order-n
    page multiple

Cc: David Hildenbrand <david@redhat.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: Gerd Hoffmann <kraxel@redhat.com>
Cc: Dongwon Kim <dongwon.kim@intel.com>
Cc: Junxiao Chang <junxiao.chang@intel.com>
Acked-by: Mike Kravetz <mike.kravetz@oracle.com> (v2)
Signed-off-by: Vivek Kasireddy <vivek.kasireddy@intel.com>
---
 drivers/dma-buf/udmabuf.c | 88 +++++++++++++++++++++++++++++++++------
 1 file changed, 75 insertions(+), 13 deletions(-)

diff --git a/drivers/dma-buf/udmabuf.c b/drivers/dma-buf/udmabuf.c
index 820c993c8659..1d1cc5e7e613 100644
--- a/drivers/dma-buf/udmabuf.c
+++ b/drivers/dma-buf/udmabuf.c
@@ -10,6 +10,7 @@
 #include <linux/miscdevice.h>
 #include <linux/module.h>
 #include <linux/shmem_fs.h>
+#include <linux/hugetlb.h>
 #include <linux/slab.h>
 #include <linux/udmabuf.h>
 #include <linux/vmalloc.h>
@@ -28,6 +29,7 @@ struct udmabuf {
 	struct page **pages;
 	struct sg_table *sg;
 	struct miscdevice *device;
+	pgoff_t *subpgoff;
 };
 
 static vm_fault_t udmabuf_vm_fault(struct vm_fault *vmf)
@@ -41,6 +43,10 @@ static vm_fault_t udmabuf_vm_fault(struct vm_fault *vmf)
 		return VM_FAULT_SIGBUS;
 
 	pfn = page_to_pfn(ubuf->pages[pgoff]);
+	if (ubuf->subpgoff) {
+		pfn += ubuf->subpgoff[pgoff] >> PAGE_SHIFT;
+	}
+
 	return vmf_insert_pfn(vma, vmf->address, pfn);
 }
 
@@ -90,23 +96,31 @@ static struct sg_table *get_sg_table(struct device *dev, struct dma_buf *buf,
 {
 	struct udmabuf *ubuf = buf->priv;
 	struct sg_table *sg;
+	struct scatterlist *sgl;
+	pgoff_t offset;
+	unsigned long i = 0;
 	int ret;
 
 	sg = kzalloc(sizeof(*sg), GFP_KERNEL);
 	if (!sg)
 		return ERR_PTR(-ENOMEM);
-	ret = sg_alloc_table_from_pages(sg, ubuf->pages, ubuf->pagecount,
-					0, ubuf->pagecount << PAGE_SHIFT,
-					GFP_KERNEL);
+
+	ret = sg_alloc_table(sg, ubuf->pagecount, GFP_KERNEL);
 	if (ret < 0)
-		goto err;
+		goto err_alloc;
+
+	for_each_sg(sg->sgl, sgl, ubuf->pagecount, i) {
+		offset = ubuf->subpgoff ? ubuf->subpgoff[i] : 0;
+		sg_set_page(sgl, ubuf->pages[i], PAGE_SIZE, offset);
+	}
 	ret = dma_map_sgtable(dev, sg, direction, 0);
 	if (ret < 0)
-		goto err;
+		goto err_map;
 	return sg;
 
-err:
+err_map:
 	sg_free_table(sg);
+err_alloc:
 	kfree(sg);
 	return ERR_PTR(ret);
 }
@@ -143,6 +157,7 @@ static void release_udmabuf(struct dma_buf *buf)
 
 	for (pg = 0; pg < ubuf->pagecount; pg++)
 		put_page(ubuf->pages[pg]);
+	kfree(ubuf->subpgoff);
 	kfree(ubuf->pages);
 	kfree(ubuf);
 }
@@ -206,7 +221,10 @@ static long udmabuf_create(struct miscdevice *device,
 	struct udmabuf *ubuf;
 	struct dma_buf *buf;
 	pgoff_t pgoff, pgcnt, pgidx, pgbuf = 0, pglimit;
-	struct page *page;
+	struct page *page, *hpage = NULL;
+	struct folio *folio;
+	pgoff_t mapidx, chunkoff, maxchunks;
+	struct hstate *hpstate;
 	int seals, ret = -EINVAL;
 	u32 i, flags;
 
@@ -242,7 +260,7 @@ static long udmabuf_create(struct miscdevice *device,
 		if (!memfd)
 			goto err;
 		mapping = memfd->f_mapping;
-		if (!shmem_mapping(mapping))
+		if (!shmem_mapping(mapping) && !is_file_hugepages(memfd))
 			goto err;
 		seals = memfd_fcntl(memfd, F_GET_SEALS, 0);
 		if (seals == -EINVAL)
@@ -253,16 +271,59 @@ static long udmabuf_create(struct miscdevice *device,
 			goto err;
 		pgoff = list[i].offset >> PAGE_SHIFT;
 		pgcnt = list[i].size   >> PAGE_SHIFT;
+		if (is_file_hugepages(memfd)) {
+			if (!ubuf->subpgoff) {
+				ubuf->subpgoff = kmalloc_array(ubuf->pagecount,
+							       sizeof(*ubuf->subpgoff),
+							       GFP_KERNEL);
+				if (!ubuf->subpgoff) {
+					ret = -ENOMEM;
+					goto err;
+				}
+			}
+			hpstate = hstate_file(memfd);
+			mapidx = list[i].offset >> huge_page_shift(hpstate);
+			mapidx <<= huge_page_order(hpstate);
+			chunkoff = (list[i].offset &
+				    ~huge_page_mask(hpstate)) >> PAGE_SHIFT;
+			maxchunks = huge_page_size(hpstate) >> PAGE_SHIFT;
+		}
 		for (pgidx = 0; pgidx < pgcnt; pgidx++) {
-			page = shmem_read_mapping_page(mapping, pgoff + pgidx);
-			if (IS_ERR(page)) {
-				ret = PTR_ERR(page);
-				goto err;
+			if (is_file_hugepages(memfd)) {
+				if (!hpage) {
+					folio = __filemap_get_folio(mapping, mapidx,
+								    FGP_ACCESSED, 0);
+					hpage = IS_ERR(folio) ? NULL: &folio->page;
+					if (!hpage) {
+						ret = -EINVAL;
+						goto err;
+					}
+				}
+				get_page(hpage);
+				ubuf->pages[pgbuf] = hpage;
+				ubuf->subpgoff[pgbuf++] = chunkoff << PAGE_SHIFT;
+				if (++chunkoff == maxchunks) {
+					put_page(hpage);
+					hpage = NULL;
+					chunkoff = 0;
+					mapidx += pages_per_huge_page(hpstate);
+				}
+			} else {
+				mapidx = pgoff + pgidx;
+				page = shmem_read_mapping_page(mapping, mapidx);
+				if (IS_ERR(page)) {
+					ret = PTR_ERR(page);
+					goto err;
+				}
+				ubuf->pages[pgbuf++] = page;
 			}
-			ubuf->pages[pgbuf++] = page;
 		}
 		fput(memfd);
 		memfd = NULL;
+		if (hpage) {
+			put_page(hpage);
+			hpage = NULL;
+		}
 	}
 
 	exp_info.ops  = &udmabuf_ops;
@@ -287,6 +348,7 @@ static long udmabuf_create(struct miscdevice *device,
 		put_page(ubuf->pages[--pgbuf]);
 	if (memfd)
 		fput(memfd);
+	kfree(ubuf->subpgoff);
 	kfree(ubuf->pages);
 	kfree(ubuf);
 	return ret;
-- 
2.39.2

[PATCH v6 3/5] mm/gup: Introduce memfd_pin_user_pages() for pinning memfd pages (v6)

[Vivek Kasireddy]

For drivers that would like to longterm-pin the pages associated
with a memfd, the pin_user_pages_fd() API provides an option to
not only pin the pages via FOLL_PIN but also to check and migrate
them if they reside in movable zone or CMA block. This API
currently works with memfds but it should work with any files
that belong to either shmemfs or hugetlbfs. Files belonging to
other filesystems are rejected for now.

The pages need to be located first before pinning them via FOLL_PIN.
If they are found in the page cache, they can be immediately pinned.
Otherwise, they need to be allocated using the filesystem specific
APIs and then pinned.

v2:
- Drop gup_flags and improve comments and commit message (David)
- Allocate a page if we cannot find in page cache for the hugetlbfs
  case as well (David)
- Don't unpin pages if there is a migration related failure (David)
- Drop the unnecessary nr_pages <= 0 check (Jason)
- Have the caller of the API pass in file * instead of fd (Jason)

v3: (David)
- Enclose the huge page allocation code with #ifdef CONFIG_HUGETLB_PAGE
  (Build error reported by kernel test robot <lkp@intel.com>)
- Don't forget memalloc_pin_restore() on non-migration related errors
- Improve the readability of the cleanup code associated with
  non-migration related errors
- Augment the comments by describing FOLL_LONGTERM like behavior
- Include the R-b tag from Jason

v4:
- Remove the local variable "page" and instead use 3 return statements
  in alloc_file_page() (David)
- Add the R-b tag from David

v5: (David)
- For hugetlb case, ensure that we only obtain head pages from the
  mapping by using __filemap_get_folio() instead of find_get_page_flags()
- Handle -EEXIST when two or more potential users try to simultaneously
  add a huge page to the mapping by forcing them to retry on failure

v6: (Christoph)
- Rename this API to memfd_pin_user_pages() to make it clear that it
  is intended for memfds
- Move the memfd page allocation helper from gup.c to memfd.c
- Fix indentation errors in memfd_pin_user_pages()
- For contiguous ranges of folios, use a helper such as
  filemap_get_folios_contig() to lookup the page cache in batches

Cc: David Hildenbrand <david@redhat.com>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Gerd Hoffmann <kraxel@redhat.com>
Cc: Dongwon Kim <dongwon.kim@intel.com>
Cc: Junxiao Chang <junxiao.chang@intel.com>
Suggested-by: Jason Gunthorpe <jgg@nvidia.com>
Reviewed-by: Jason Gunthorpe <jgg@nvidia.com> (v2)
Reviewed-by: David Hildenbrand <david@redhat.com> (v3)
Signed-off-by: Vivek Kasireddy <vivek.kasireddy@intel.com>
---
 include/linux/memfd.h |   5 +++
 include/linux/mm.h    |   2 +
 mm/gup.c              | 102 ++++++++++++++++++++++++++++++++++++++++++
 mm/memfd.c            |  34 ++++++++++++++
 4 files changed, 143 insertions(+)

diff --git a/include/linux/memfd.h b/include/linux/memfd.h
index e7abf6fa4c52..6fc0d1282151 100644
--- a/include/linux/memfd.h
+++ b/include/linux/memfd.h
@@ -6,11 +6,16 @@
 
 #ifdef CONFIG_MEMFD_CREATE
 extern long memfd_fcntl(struct file *file, unsigned int cmd, unsigned int arg);
+extern struct page *memfd_alloc_page(struct file *memfd, pgoff_t idx);
 #else
 static inline long memfd_fcntl(struct file *f, unsigned int c, unsigned int a)
 {
 	return -EINVAL;
 }
+static inline struct page *memfd_alloc_page(struct file *memfd, pgoff_t idx)
+{
+	return ERR_PTR(-EINVAL);
+}
 #endif
 
 #endif /* __LINUX_MEMFD_H */
diff --git a/include/linux/mm.h b/include/linux/mm.h
index 418d26608ece..ac69db45509f 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -2472,6 +2472,8 @@ long get_user_pages_unlocked(unsigned long start, unsigned long nr_pages,
 		    struct page **pages, unsigned int gup_flags);
 long pin_user_pages_unlocked(unsigned long start, unsigned long nr_pages,
 		    struct page **pages, unsigned int gup_flags);
+long memfd_pin_user_pages(struct file *file, pgoff_t start,
+			  unsigned long nr_pages, struct page **pages);
 
 int get_user_pages_fast(unsigned long start, int nr_pages,
 			unsigned int gup_flags, struct page **pages);
diff --git a/mm/gup.c b/mm/gup.c
index 231711efa390..eb93d1ec9dc6 100644
--- a/mm/gup.c
+++ b/mm/gup.c
@@ -5,6 +5,7 @@
 #include <linux/spinlock.h>
 
 #include <linux/mm.h>
+#include <linux/memfd.h>
 #include <linux/memremap.h>
 #include <linux/pagemap.h>
 #include <linux/rmap.h>
@@ -17,6 +18,7 @@
 #include <linux/hugetlb.h>
 #include <linux/migrate.h>
 #include <linux/mm_inline.h>
+#include <linux/pagevec.h>
 #include <linux/sched/mm.h>
 #include <linux/shmem_fs.h>
 
@@ -3410,3 +3412,103 @@ long pin_user_pages_unlocked(unsigned long start, unsigned long nr_pages,
 				     &locked, gup_flags);
 }
 EXPORT_SYMBOL(pin_user_pages_unlocked);
+
+/**
+ * memfd_pin_user_pages() - pin user pages associated with a memfd
+ * @memfd:      the memfd whose pages are to be pinned
+ * @start:      starting memfd offset
+ * @nr_pages:   number of pages from start to pin
+ * @pages:      array that receives pointers to the pages pinned.
+ *              Should be at-least nr_pages long.
+ *
+ * Attempt to pin pages associated with a memfd; given that a memfd is either
+ * backed by shmem or hugetlb, the pages can either be found in the page cache
+ * or need to be allocated if necessary. Once the pages are located, they are
+ * all pinned via FOLL_PIN. And, these pinned pages need to be released either
+ * using unpin_user_pages() or unpin_user_page().
+ *
+ * It must be noted that the pages may be pinned for an indefinite amount
+ * of time. And, in most cases, the duration of time they may stay pinned
+ * would be controlled by the userspace. This behavior is effectively the
+ * same as using FOLL_LONGTERM with other GUP APIs.
+ *
+ * Returns number of pages pinned. This would be equal to the number of
+ * pages requested. If no pages were pinned, it returns -errno.
+ */
+long memfd_pin_user_pages(struct file *memfd, pgoff_t start,
+			  unsigned long nr_pages, struct page **pages)
+{
+	pgoff_t start_idx, end_idx = start + nr_pages - 1;
+	unsigned int flags, nr_folios, i, j;
+	struct folio_batch fbatch;
+	struct page *page = NULL;
+	struct folio *folio;
+	long ret;
+
+	if (!nr_pages)
+		return -EINVAL;
+
+	if (!memfd)
+		return -EINVAL;
+
+	if (!shmem_file(memfd) && !is_file_hugepages(memfd))
+		return -EINVAL;
+
+	flags = memalloc_pin_save();
+	do {
+		folio_batch_init(&fbatch);
+		start_idx = start;
+		i = 0;
+
+		while (start_idx <= end_idx) {
+			/*
+			 * In most cases, we should be able to find the page
+			 * in the page cache. If we cannot find it for some
+			 * reason, we try to allocate one and add it to the
+			 * page cache.
+			 */
+			nr_folios = filemap_get_folios_contig(memfd->f_mapping,
+							      &start_idx,
+							      end_idx,
+							      &fbatch);
+			if (page) {
+				put_page(page);
+				page = NULL;
+			}
+			for (j = 0; j < nr_folios; j++) {
+				folio = fbatch.folios[j];
+				ret = try_grab_page(&folio->page, FOLL_PIN);
+				if (unlikely(ret)) {
+					folio_batch_release(&fbatch);
+					goto err;
+				}
+
+				pages[i++] = &folio->page;
+			}
+
+			folio_batch_release(&fbatch);
+			if (!nr_folios) {
+				page = memfd_alloc_page(memfd, start_idx);
+				if (IS_ERR(page)) {
+					ret = PTR_ERR(page);
+					if (ret != -EEXIST)
+						goto err;
+				}
+			}
+		}
+
+		ret = check_and_migrate_movable_pages(nr_pages, pages);
+	} while (ret == -EAGAIN);
+
+	memalloc_pin_restore(flags);
+	return ret ? ret : nr_pages;
+err:
+	memalloc_pin_restore(flags);
+	while (i-- > 0)
+		if (pages[i])
+			unpin_user_page(pages[i]);
+
+	return ret;
+}
+EXPORT_SYMBOL_GPL(memfd_pin_user_pages);
+
diff --git a/mm/memfd.c b/mm/memfd.c
index d3a1ba4208c9..b315cd12bdb7 100644
--- a/mm/memfd.c
+++ b/mm/memfd.c
@@ -63,6 +63,40 @@ static void memfd_tag_pins(struct xa_state *xas)
 	xas_unlock_irq(xas);
 }
 
+/*
+ * This is a helper function used by memfd_pin_user_pages() in GUP (gup.c).
+ * It is mainly called to allocate a page in a memfd when the caller
+ * (memfd_pin_user_pages()) cannot find a page in the page cache at a given
+ * index in the mapping.
+ */
+struct page *memfd_alloc_page(struct file *memfd, pgoff_t idx)
+{
+#ifdef CONFIG_HUGETLB_PAGE
+	struct folio *folio;
+	int err;
+
+	if (is_file_hugepages(memfd)) {
+		folio = alloc_hugetlb_folio_nodemask(hstate_file(memfd),
+						     NUMA_NO_NODE,
+						     NULL,
+						     GFP_USER);
+		if (folio && folio_try_get(folio)) {
+			err = hugetlb_add_to_page_cache(folio,
+							memfd->f_mapping,
+							idx);
+			if (err) {
+				folio_put(folio);
+				free_huge_folio(folio);
+				return ERR_PTR(err);
+			}
+			return &folio->page;
+		}
+		return ERR_PTR(-ENOMEM);
+	}
+#endif
+	return shmem_read_mapping_page(memfd->f_mapping, idx);
+}
+
 /*
  * Setting SEAL_WRITE requires us to verify there's no pending writer. However,
  * via get_user_pages(), drivers might have some pending I/O without any active
-- 
2.39.2

[PATCH v6 4/5] udmabuf: Pin the pages using memfd_pin_user_pages() API (v4)

[Vivek Kasireddy]

Using memfd_pin_user_pages() will ensure that the pages are pinned
correctly using FOLL_PIN. And, this also ensures that we don't
accidentally break features such as memory hotunplug as it would
not allow pinning pages in the movable zone.

Using this new API also simplifies the code as we no longer have
to deal with extracting individual pages from their mappings. As
a result, we can drop some of the local variables such as page,
hpage, mapping, etc.

v2:
- Adjust to the change in signature of pin_user_pages_fd() by
  passing in file * instead of fd.

v3:
- Limit the changes in this patch only to those that are required
  for using pin_user_pages_fd()
- Slightly improve the commit message

v4:
- Adjust to the change in name of the API (memfd_pin_user_pages)

Cc: David Hildenbrand <david@redhat.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: Gerd Hoffmann <kraxel@redhat.com>
Cc: Dongwon Kim <dongwon.kim@intel.com>
Cc: Junxiao Chang <junxiao.chang@intel.com>
Signed-off-by: Vivek Kasireddy <vivek.kasireddy@intel.com>
---
 drivers/dma-buf/udmabuf.c | 62 +++++++++++++++------------------------
 1 file changed, 23 insertions(+), 39 deletions(-)

diff --git a/drivers/dma-buf/udmabuf.c b/drivers/dma-buf/udmabuf.c
index 1d1cc5e7e613..887dc287a23c 100644
--- a/drivers/dma-buf/udmabuf.c
+++ b/drivers/dma-buf/udmabuf.c
@@ -156,7 +156,8 @@ static void release_udmabuf(struct dma_buf *buf)
 		put_sg_table(dev, ubuf->sg, DMA_BIDIRECTIONAL);
 
 	for (pg = 0; pg < ubuf->pagecount; pg++)
-		put_page(ubuf->pages[pg]);
+		unpin_user_page(ubuf->pages[pg]);
+
 	kfree(ubuf->subpgoff);
 	kfree(ubuf->pages);
 	kfree(ubuf);
@@ -217,15 +218,13 @@ static long udmabuf_create(struct miscdevice *device,
 {
 	DEFINE_DMA_BUF_EXPORT_INFO(exp_info);
 	struct file *memfd = NULL;
-	struct address_space *mapping = NULL;
 	struct udmabuf *ubuf;
 	struct dma_buf *buf;
-	pgoff_t pgoff, pgcnt, pgidx, pgbuf = 0, pglimit;
-	struct page *page, *hpage = NULL;
-	struct folio *folio;
+	pgoff_t pgcnt, pgbuf = 0, pglimit, nr_pages;
 	pgoff_t mapidx, chunkoff, maxchunks;
 	struct hstate *hpstate;
-	int seals, ret = -EINVAL;
+	long ret = -EINVAL;
+	int seals;
 	u32 i, flags;
 
 	ubuf = kzalloc(sizeof(*ubuf), GFP_KERNEL);
@@ -259,8 +258,7 @@ static long udmabuf_create(struct miscdevice *device,
 		memfd = fget(list[i].memfd);
 		if (!memfd)
 			goto err;
-		mapping = memfd->f_mapping;
-		if (!shmem_mapping(mapping) && !is_file_hugepages(memfd))
+		if (!shmem_file(memfd) && !is_file_hugepages(memfd))
 			goto err;
 		seals = memfd_fcntl(memfd, F_GET_SEALS, 0);
 		if (seals == -EINVAL)
@@ -269,7 +267,7 @@ static long udmabuf_create(struct miscdevice *device,
 		if ((seals & SEALS_WANTED) != SEALS_WANTED ||
 		    (seals & SEALS_DENIED) != 0)
 			goto err;
-		pgoff = list[i].offset >> PAGE_SHIFT;
+		mapidx = list[i].offset >> PAGE_SHIFT;
 		pgcnt = list[i].size   >> PAGE_SHIFT;
 		if (is_file_hugepages(memfd)) {
 			if (!ubuf->subpgoff) {
@@ -288,42 +286,27 @@ static long udmabuf_create(struct miscdevice *device,
 				    ~huge_page_mask(hpstate)) >> PAGE_SHIFT;
 			maxchunks = huge_page_size(hpstate) >> PAGE_SHIFT;
 		}
-		for (pgidx = 0; pgidx < pgcnt; pgidx++) {
+
+		do {
+			nr_pages = shmem_file(memfd) ? pgcnt : 1;
+			ret = memfd_pin_user_pages(memfd, mapidx, nr_pages,
+						   ubuf->pages + pgbuf);
+			if (ret < 0)
+				goto err;
+
 			if (is_file_hugepages(memfd)) {
-				if (!hpage) {
-					folio = __filemap_get_folio(mapping, mapidx,
-								    FGP_ACCESSED, 0);
-					hpage = IS_ERR(folio) ? NULL: &folio->page;
-					if (!hpage) {
-						ret = -EINVAL;
-						goto err;
-					}
-				}
-				get_page(hpage);
-				ubuf->pages[pgbuf] = hpage;
-				ubuf->subpgoff[pgbuf++] = chunkoff << PAGE_SHIFT;
+				ubuf->subpgoff[pgbuf] = chunkoff << PAGE_SHIFT;
 				if (++chunkoff == maxchunks) {
-					put_page(hpage);
-					hpage = NULL;
 					chunkoff = 0;
 					mapidx += pages_per_huge_page(hpstate);
 				}
-			} else {
-				mapidx = pgoff + pgidx;
-				page = shmem_read_mapping_page(mapping, mapidx);
-				if (IS_ERR(page)) {
-					ret = PTR_ERR(page);
-					goto err;
-				}
-				ubuf->pages[pgbuf++] = page;
 			}
-		}
+			pgbuf += nr_pages;
+			pgcnt -= nr_pages;
+		} while (pgcnt > 0);
+
 		fput(memfd);
 		memfd = NULL;
-		if (hpage) {
-			put_page(hpage);
-			hpage = NULL;
-		}
 	}
 
 	exp_info.ops  = &udmabuf_ops;
@@ -344,8 +327,9 @@ static long udmabuf_create(struct miscdevice *device,
 	return dma_buf_fd(buf, flags);
 
 err:
-	while (pgbuf > 0)
-		put_page(ubuf->pages[--pgbuf]);
+	while (pgbuf-- > 0)
+		if (ubuf->pages[pgbuf])
+			unpin_user_page(ubuf->pages[pgbuf]);
 	if (memfd)
 		fput(memfd);
 	kfree(ubuf->subpgoff);
-- 
2.39.2

[PATCH v6 5/5] selftests/dma-buf/udmabuf: Add tests to verify data after page migration

[Vivek Kasireddy]

Since the memfd pages associated with a udmabuf may be migrated
as part of udmabuf create, we need to verify the data coherency
after successful migration. The new tests added in this patch try
to do just that using 4k sized pages and also 2 MB sized huge
pages for the memfd.

Successful completion of the tests would mean that there is no
disconnect between the memfd pages and the ones associated with
a udmabuf. And, these tests can also be augmented in the future
to test newer udmabuf features (such as handling memfd hole punch).

Cc: Shuah Khan <shuah@kernel.org>
Cc: David Hildenbrand <david@redhat.com>
Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Peter Xu <peterx@redhat.com>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: Gerd Hoffmann <kraxel@redhat.com>
Cc: Dongwon Kim <dongwon.kim@intel.com>
Cc: Junxiao Chang <junxiao.chang@intel.com>
Based-on-patch-by: Mike Kravetz <mike.kravetz@oracle.com>
Signed-off-by: Vivek Kasireddy <vivek.kasireddy@intel.com>
---
 .../selftests/drivers/dma-buf/udmabuf.c       | 151 +++++++++++++++++-
 1 file changed, 147 insertions(+), 4 deletions(-)

diff --git a/tools/testing/selftests/drivers/dma-buf/udmabuf.c b/tools/testing/selftests/drivers/dma-buf/udmabuf.c
index c812080e304e..d76c813fe652 100644
--- a/tools/testing/selftests/drivers/dma-buf/udmabuf.c
+++ b/tools/testing/selftests/drivers/dma-buf/udmabuf.c
@@ -9,26 +9,132 @@
 #include <errno.h>
 #include <fcntl.h>
 #include <malloc.h>
+#include <stdbool.h>
 
 #include <sys/ioctl.h>
 #include <sys/syscall.h>
+#include <sys/mman.h>
 #include <linux/memfd.h>
 #include <linux/udmabuf.h>
 
 #define TEST_PREFIX	"drivers/dma-buf/udmabuf"
 #define NUM_PAGES       4
+#define NUM_ENTRIES     4
+#define MEMFD_SIZE      1024 /* in pages */
 
-static int memfd_create(const char *name, unsigned int flags)
+static unsigned int page_size;
+
+static int create_memfd_with_seals(off64_t size, bool hpage)
+{
+	int memfd, ret;
+	unsigned int flags = MFD_ALLOW_SEALING;
+
+	if (hpage)
+		flags |= MFD_HUGETLB;
+
+	memfd = memfd_create("udmabuf-test", flags);
+	if (memfd < 0) {
+		printf("%s: [skip,no-memfd]\n", TEST_PREFIX);
+		exit(77);
+	}
+
+	ret = fcntl(memfd, F_ADD_SEALS, F_SEAL_SHRINK);
+	if (ret < 0) {
+		printf("%s: [skip,fcntl-add-seals]\n", TEST_PREFIX);
+		exit(77);
+	}
+
+	ret = ftruncate(memfd, size);
+	if (ret == -1) {
+		printf("%s: [FAIL,memfd-truncate]\n", TEST_PREFIX);
+		exit(1);
+	}
+
+	return memfd;
+}
+
+static int create_udmabuf_list(int devfd, int memfd, off64_t memfd_size)
+{
+	struct udmabuf_create_list *list;
+	int ubuf_fd, i;
+
+	list = malloc(sizeof(struct udmabuf_create_list) +
+		      sizeof(struct udmabuf_create_item) * NUM_ENTRIES);
+	if (!list) {
+		printf("%s: [FAIL, udmabuf-malloc]\n", TEST_PREFIX);
+		exit(1);
+	}
+
+	for (i = 0; i < NUM_ENTRIES; i++) {
+		list->list[i].memfd  = memfd;
+		list->list[i].offset = i * (memfd_size / NUM_ENTRIES);
+		list->list[i].size   = getpagesize() * NUM_PAGES;
+	}
+
+	list->count = NUM_ENTRIES;
+	list->flags = UDMABUF_FLAGS_CLOEXEC;
+	ubuf_fd = ioctl(devfd, UDMABUF_CREATE_LIST, list);
+	free(list);
+	if (ubuf_fd < 0) {
+		printf("%s: [FAIL, udmabuf-create]\n", TEST_PREFIX);
+		exit(1);
+	}
+
+	return ubuf_fd;
+}
+
+static void write_to_memfd(void *addr, off64_t size, char chr)
+{
+	int i;
+
+	for (i = 0; i < size / page_size; i++) {
+		*((char *)addr + (i * page_size)) = chr;
+	}
+}
+
+static void *mmap_fd(int fd, off64_t size)
 {
-	return syscall(__NR_memfd_create, name, flags);
+	void *addr;
+
+	addr = mmap(NULL, size, PROT_READ|PROT_WRITE, MAP_SHARED, fd, 0);
+	if (addr == MAP_FAILED) {
+		printf("%s: ubuf_fd mmap fail\n", TEST_PREFIX);
+		exit(1);
+	}
+
+	return addr;
+}
+
+static int compare_chunks(void *addr1, void *addr2, off64_t memfd_size)
+{
+	off64_t off;
+	int i = 0, j, k = 0, ret = 0;
+	char char1, char2;
+
+	while (i < NUM_ENTRIES) {
+		off = i * (memfd_size / NUM_ENTRIES);
+		for (j = 0; j < NUM_PAGES; j++, k++) {
+			char1 = *((char *)addr1 + off + (j * getpagesize()));
+			char2 = *((char *)addr2 + (k * getpagesize()));
+			if (char1 != char2) {
+				ret = -1;
+				goto err;
+			}
+		}
+		i++;
+	}
+err:
+	munmap(addr1, memfd_size);
+	munmap(addr2, NUM_ENTRIES * NUM_PAGES * getpagesize());
+	return ret;
 }
 
 int main(int argc, char *argv[])
 {
 	struct udmabuf_create create;
 	int devfd, memfd, buf, ret;
-	off_t size;
-	void *mem;
+	off64_t size;
+	void *addr1, *addr2;
 
 	devfd = open("/dev/udmabuf", O_RDWR);
 	if (devfd < 0) {
@@ -90,6 +196,9 @@ int main(int argc, char *argv[])
 	}
 
 	/* should work */
+	page_size = getpagesize();
+	addr1 = mmap_fd(memfd, size);
+	write_to_memfd(addr1, size, 'a');
 	create.memfd  = memfd;
 	create.offset = 0;
 	create.size   = size;
@@ -98,6 +207,40 @@ int main(int argc, char *argv[])
 		printf("%s: [FAIL,test-4]\n", TEST_PREFIX);
 		exit(1);
 	}
+	munmap(addr1, size);
+	close(buf);
+	close(memfd);
+
+	/* should work (migration of 4k size pages)*/
+	size = MEMFD_SIZE * page_size;
+	memfd = create_memfd_with_seals(size, false);
+	addr1 = mmap_fd(memfd, size);
+	write_to_memfd(addr1, size, 'a');
+	buf = create_udmabuf_list(devfd, memfd, size);
+	addr2 = mmap_fd(buf, NUM_PAGES * NUM_ENTRIES * getpagesize());
+	write_to_memfd(addr1, size, 'b');
+	ret = compare_chunks(addr1, addr2, size);
+	if (ret < 0) {
+		printf("%s: [FAIL,test-5]\n", TEST_PREFIX);
+		exit(1);
+	}
+	close(buf);
+	close(memfd);
+
+	/* should work (migration of 2MB size huge pages)*/
+	page_size = getpagesize() * 512; /* 2 MB */
+	size = MEMFD_SIZE * page_size;
+	memfd = create_memfd_with_seals(size, true);
+	addr1 = mmap_fd(memfd, size);
+	write_to_memfd(addr1, size, 'a');
+	buf = create_udmabuf_list(devfd, memfd, size);
+	addr2 = mmap_fd(buf, NUM_PAGES * NUM_ENTRIES * getpagesize());
+	write_to_memfd(addr1, size, 'b');
+	ret = compare_chunks(addr1, addr2, size);
+	if (ret < 0) {
+		printf("%s: [FAIL,test-6]\n", TEST_PREFIX);
+		exit(1);
+	}
 
 	fprintf(stderr, "%s: ok\n", TEST_PREFIX);
 	close(buf);
-- 
2.39.2

Re: [PATCH v6 3/5] mm/gup: Introduce memfd_pin_user_pages() for pinning memfd pages (v6)

[David Hildenbrand]

On 05.12.23 06:35, Vivek Kasireddy wrote:
> For drivers that would like to longterm-pin the pages associated
> with a memfd, the pin_user_pages_fd() API provides an option to
> not only pin the pages via FOLL_PIN but also to check and migrate
> them if they reside in movable zone or CMA block. This API
> currently works with memfds but it should work with any files
> that belong to either shmemfs or hugetlbfs. Files belonging to
> other filesystems are rejected for now.
> 
> The pages need to be located first before pinning them via FOLL_PIN.
> If they are found in the page cache, they can be immediately pinned.
> Otherwise, they need to be allocated using the filesystem specific
> APIs and then pinned.
> 
> v2:
> - Drop gup_flags and improve comments and commit message (David)
> - Allocate a page if we cannot find in page cache for the hugetlbfs
>    case as well (David)
> - Don't unpin pages if there is a migration related failure (David)
> - Drop the unnecessary nr_pages <= 0 check (Jason)
> - Have the caller of the API pass in file * instead of fd (Jason)
> 
> v3: (David)
> - Enclose the huge page allocation code with #ifdef CONFIG_HUGETLB_PAGE
>    (Build error reported by kernel test robot <lkp@intel.com>)
> - Don't forget memalloc_pin_restore() on non-migration related errors
> - Improve the readability of the cleanup code associated with
>    non-migration related errors
> - Augment the comments by describing FOLL_LONGTERM like behavior
> - Include the R-b tag from Jason
> 
> v4:
> - Remove the local variable "page" and instead use 3 return statements
>    in alloc_file_page() (David)
> - Add the R-b tag from David
> 
> v5: (David)
> - For hugetlb case, ensure that we only obtain head pages from the
>    mapping by using __filemap_get_folio() instead of find_get_page_flags()
> - Handle -EEXIST when two or more potential users try to simultaneously
>    add a huge page to the mapping by forcing them to retry on failure
> 
> v6: (Christoph)
> - Rename this API to memfd_pin_user_pages() to make it clear that it
>    is intended for memfds
> - Move the memfd page allocation helper from gup.c to memfd.c
> - Fix indentation errors in memfd_pin_user_pages()
> - For contiguous ranges of folios, use a helper such as
>    filemap_get_folios_contig() to lookup the page cache in batches
> 
> Cc: David Hildenbrand <david@redhat.com>
> Cc: Christoph Hellwig <hch@infradead.org>
> Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
> Cc: Mike Kravetz <mike.kravetz@oracle.com>
> Cc: Hugh Dickins <hughd@google.com>
> Cc: Peter Xu <peterx@redhat.com>
> Cc: Gerd Hoffmann <kraxel@redhat.com>
> Cc: Dongwon Kim <dongwon.kim@intel.com>
> Cc: Junxiao Chang <junxiao.chang@intel.com>
> Suggested-by: Jason Gunthorpe <jgg@nvidia.com>
> Reviewed-by: Jason Gunthorpe <jgg@nvidia.com> (v2)
> Reviewed-by: David Hildenbrand <david@redhat.com> (v3)
> Signed-off-by: Vivek Kasireddy <vivek.kasireddy@intel.com>
> ---
>   include/linux/memfd.h |   5 +++
>   include/linux/mm.h    |   2 +
>   mm/gup.c              | 102 ++++++++++++++++++++++++++++++++++++++++++
>   mm/memfd.c            |  34 ++++++++++++++
>   4 files changed, 143 insertions(+)
> 
> diff --git a/include/linux/memfd.h b/include/linux/memfd.h
> index e7abf6fa4c52..6fc0d1282151 100644
> --- a/include/linux/memfd.h
> +++ b/include/linux/memfd.h
> @@ -6,11 +6,16 @@
>   
>   #ifdef CONFIG_MEMFD_CREATE
>   extern long memfd_fcntl(struct file *file, unsigned int cmd, unsigned int arg);
> +extern struct page *memfd_alloc_page(struct file *memfd, pgoff_t idx);
>   #else
>   static inline long memfd_fcntl(struct file *f, unsigned int c, unsigned int a)
>   {
>   	return -EINVAL;
>   }
> +static inline struct page *memfd_alloc_page(struct file *memfd, pgoff_t idx)
> +{
> +	return ERR_PTR(-EINVAL);
> +}
>   #endif
>   
>   #endif /* __LINUX_MEMFD_H */
> diff --git a/include/linux/mm.h b/include/linux/mm.h
> index 418d26608ece..ac69db45509f 100644
> --- a/include/linux/mm.h
> +++ b/include/linux/mm.h
> @@ -2472,6 +2472,8 @@ long get_user_pages_unlocked(unsigned long start, unsigned long nr_pages,
>   		    struct page **pages, unsigned int gup_flags);
>   long pin_user_pages_unlocked(unsigned long start, unsigned long nr_pages,
>   		    struct page **pages, unsigned int gup_flags);
> +long memfd_pin_user_pages(struct file *file, pgoff_t start,
> +			  unsigned long nr_pages, struct page **pages);
>   
>   int get_user_pages_fast(unsigned long start, int nr_pages,
>   			unsigned int gup_flags, struct page **pages);
> diff --git a/mm/gup.c b/mm/gup.c
> index 231711efa390..eb93d1ec9dc6 100644
> --- a/mm/gup.c
> +++ b/mm/gup.c
> @@ -5,6 +5,7 @@
>   #include <linux/spinlock.h>
>   
>   #include <linux/mm.h>
> +#include <linux/memfd.h>
>   #include <linux/memremap.h>
>   #include <linux/pagemap.h>
>   #include <linux/rmap.h>
> @@ -17,6 +18,7 @@
>   #include <linux/hugetlb.h>
>   #include <linux/migrate.h>
>   #include <linux/mm_inline.h>
> +#include <linux/pagevec.h>
>   #include <linux/sched/mm.h>
>   #include <linux/shmem_fs.h>
>   
> @@ -3410,3 +3412,103 @@ long pin_user_pages_unlocked(unsigned long start, unsigned long nr_pages,
>   				     &locked, gup_flags);
>   }
>   EXPORT_SYMBOL(pin_user_pages_unlocked);
> +
> +/**
> + * memfd_pin_user_pages() - pin user pages associated with a memfd
> + * @memfd:      the memfd whose pages are to be pinned
> + * @start:      starting memfd offset
> + * @nr_pages:   number of pages from start to pin
> + * @pages:      array that receives pointers to the pages pinned.
> + *              Should be at-least nr_pages long.
> + *
> + * Attempt to pin pages associated with a memfd; given that a memfd is either
> + * backed by shmem or hugetlb, the pages can either be found in the page cache
> + * or need to be allocated if necessary. Once the pages are located, they are
> + * all pinned via FOLL_PIN. And, these pinned pages need to be released either
> + * using unpin_user_pages() or unpin_user_page().
> + *
> + * It must be noted that the pages may be pinned for an indefinite amount
> + * of time. And, in most cases, the duration of time they may stay pinned
> + * would be controlled by the userspace. This behavior is effectively the
> + * same as using FOLL_LONGTERM with other GUP APIs.
> + *
> + * Returns number of pages pinned. This would be equal to the number of
> + * pages requested. If no pages were pinned, it returns -errno.
> + */
> +long memfd_pin_user_pages(struct file *memfd, pgoff_t start,
> +			  unsigned long nr_pages, struct page **pages)
> +{
> +	pgoff_t start_idx, end_idx = start + nr_pages - 1;
> +	unsigned int flags, nr_folios, i, j;
> +	struct folio_batch fbatch;
> +	struct page *page = NULL;
> +	struct folio *folio;
> +	long ret;
> +
> +	if (!nr_pages)
> +		return -EINVAL;
> +
> +	if (!memfd)
> +		return -EINVAL;
> +
> +	if (!shmem_file(memfd) && !is_file_hugepages(memfd))
> +		return -EINVAL;
> +
> +	flags = memalloc_pin_save();
> +	do {
> +		folio_batch_init(&fbatch);
> +		start_idx = start;
> +		i = 0;
> +
> +		while (start_idx <= end_idx) {
> +			/*
> +			 * In most cases, we should be able to find the page
> +			 * in the page cache. If we cannot find it for some
> +			 * reason, we try to allocate one and add it to the
> +			 * page cache.
> +			 */
> +			nr_folios = filemap_get_folios_contig(memfd->f_mapping,
> +							      &start_idx,
> +							      end_idx,
> +							      &fbatch);
> +			if (page) {
> +				put_page(page);
> +				page = NULL;
> +			}
> +			for (j = 0; j < nr_folios; j++) {
> +				folio = fbatch.folios[j];
> +				ret = try_grab_page(&folio->page, FOLL_PIN);
> +				if (unlikely(ret)) {
> +					folio_batch_release(&fbatch);
> +					goto err;
> +				}
> +
> +				pages[i++] = &folio->page;
> +			}

I might be wrong, but that interface is still inconsistent. I think your 
intention is to always return folios (head pages), but why are we 
returning pages from this interface then?

It would be more consistent regarding the other GUP interfaces to return 
the actual tail pages that fit the given "pgoff_t start". So if you 
punch in "nr_pages" you expect to get "nr_pages" pages, and not some 
other number of folios.

Otherwise, this interface is highly confusing.

If you always want to return folios, then better name it 
"memfd_pin_user_folios" (or just "memfd_pin_folios") and pass in a range 
(instead of a nr_pages parameter), and somehow indicate to the caller 
how many folio were in that range, and if that range was fully covered.

Or am I missing something?

-- 
Cheers,

David / dhildenb

RE: [PATCH v6 3/5] mm/gup: Introduce memfd_pin_user_pages() for pinning memfd pages (v6)

[Kasireddy, Vivek]

Hi,

> > +struct page *memfd_alloc_page(struct file *memfd, pgoff_t idx)
> > +{
> > +#ifdef CONFIG_HUGETLB_PAGE
> > +	struct folio *folio;
> > +	int err;
> > +
> > +	if (is_file_hugepages(memfd)) {
> > +		folio = alloc_hugetlb_folio_nodemask(hstate_file(memfd),
> > +						     NUMA_NO_NODE,
> > +						     NULL,
> > +						     GFP_USER);
> > +		if (folio && folio_try_get(folio)) {
> > +			err = hugetlb_add_to_page_cache(folio,
> 
> If alloc_hugetlb_folio_nodemask moved out of the CONFIG_HUGETLB_PAGE
> ifdef, the ifdef here could go away.
Unlike alloc_hugetlb_folio_nodemask(), hugetlb_add_to_page_cache() does not
get exposed without enabling CONFIG_HUGETLB_PAGE.

> 
> Either way, this looks good:
> 
> Reviewed-by: Christoph Hellwig <hch@lst.de>
Thank you for the review.

Thanks,
Vivek
> 
> 

RE: [PATCH v6 3/5] mm/gup: Introduce memfd_pin_user_pages() for pinning memfd pages (v6)

[Kasireddy, Vivek]

Hi David,

> On 05.12.23 06:35, Vivek Kasireddy wrote:
> > For drivers that would like to longterm-pin the pages associated
> > with a memfd, the pin_user_pages_fd() API provides an option to
> > not only pin the pages via FOLL_PIN but also to check and migrate
> > them if they reside in movable zone or CMA block. This API
> > currently works with memfds but it should work with any files
> > that belong to either shmemfs or hugetlbfs. Files belonging to
> > other filesystems are rejected for now.
> >
> > The pages need to be located first before pinning them via FOLL_PIN.
> > If they are found in the page cache, they can be immediately pinned.
> > Otherwise, they need to be allocated using the filesystem specific
> > APIs and then pinned.
> >
> > v2:
> > - Drop gup_flags and improve comments and commit message (David)
> > - Allocate a page if we cannot find in page cache for the hugetlbfs
> >    case as well (David)
> > - Don't unpin pages if there is a migration related failure (David)
> > - Drop the unnecessary nr_pages <= 0 check (Jason)
> > - Have the caller of the API pass in file * instead of fd (Jason)
> >
> > v3: (David)
> > - Enclose the huge page allocation code with #ifdef
> CONFIG_HUGETLB_PAGE
> >    (Build error reported by kernel test robot <lkp@intel.com>)
> > - Don't forget memalloc_pin_restore() on non-migration related errors
> > - Improve the readability of the cleanup code associated with
> >    non-migration related errors
> > - Augment the comments by describing FOLL_LONGTERM like behavior
> > - Include the R-b tag from Jason
> >
> > v4:
> > - Remove the local variable "page" and instead use 3 return statements
> >    in alloc_file_page() (David)
> > - Add the R-b tag from David
> >
> > v5: (David)
> > - For hugetlb case, ensure that we only obtain head pages from the
> >    mapping by using __filemap_get_folio() instead of find_get_page_flags()
> > - Handle -EEXIST when two or more potential users try to simultaneously
> >    add a huge page to the mapping by forcing them to retry on failure
> >
> > v6: (Christoph)
> > - Rename this API to memfd_pin_user_pages() to make it clear that it
> >    is intended for memfds
> > - Move the memfd page allocation helper from gup.c to memfd.c
> > - Fix indentation errors in memfd_pin_user_pages()
> > - For contiguous ranges of folios, use a helper such as
> >    filemap_get_folios_contig() to lookup the page cache in batches
> >
> > Cc: David Hildenbrand <david@redhat.com>
> > Cc: Christoph Hellwig <hch@infradead.org>
> > Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
> > Cc: Mike Kravetz <mike.kravetz@oracle.com>
> > Cc: Hugh Dickins <hughd@google.com>
> > Cc: Peter Xu <peterx@redhat.com>
> > Cc: Gerd Hoffmann <kraxel@redhat.com>
> > Cc: Dongwon Kim <dongwon.kim@intel.com>
> > Cc: Junxiao Chang <junxiao.chang@intel.com>
> > Suggested-by: Jason Gunthorpe <jgg@nvidia.com>
> > Reviewed-by: Jason Gunthorpe <jgg@nvidia.com> (v2)
> > Reviewed-by: David Hildenbrand <david@redhat.com> (v3)
> > Signed-off-by: Vivek Kasireddy <vivek.kasireddy@intel.com>
> > ---
> >   include/linux/memfd.h |   5 +++
> >   include/linux/mm.h    |   2 +
> >   mm/gup.c              | 102 ++++++++++++++++++++++++++++++++++++++++++
> >   mm/memfd.c            |  34 ++++++++++++++
> >   4 files changed, 143 insertions(+)
> >
> > diff --git a/include/linux/memfd.h b/include/linux/memfd.h
> > index e7abf6fa4c52..6fc0d1282151 100644
> > --- a/include/linux/memfd.h
> > +++ b/include/linux/memfd.h
> > @@ -6,11 +6,16 @@
> >
> >   #ifdef CONFIG_MEMFD_CREATE
> >   extern long memfd_fcntl(struct file *file, unsigned int cmd, unsigned int
> arg);
> > +extern struct page *memfd_alloc_page(struct file *memfd, pgoff_t idx);
> >   #else
> >   static inline long memfd_fcntl(struct file *f, unsigned int c, unsigned int a)
> >   {
> >   	return -EINVAL;
> >   }
> > +static inline struct page *memfd_alloc_page(struct file *memfd, pgoff_t
> idx)
> > +{
> > +	return ERR_PTR(-EINVAL);
> > +}
> >   #endif
> >
> >   #endif /* __LINUX_MEMFD_H */
> > diff --git a/include/linux/mm.h b/include/linux/mm.h
> > index 418d26608ece..ac69db45509f 100644
> > --- a/include/linux/mm.h
> > +++ b/include/linux/mm.h
> > @@ -2472,6 +2472,8 @@ long get_user_pages_unlocked(unsigned long
> start, unsigned long nr_pages,
> >   		    struct page **pages, unsigned int gup_flags);
> >   long pin_user_pages_unlocked(unsigned long start, unsigned long
> nr_pages,
> >   		    struct page **pages, unsigned int gup_flags);
> > +long memfd_pin_user_pages(struct file *file, pgoff_t start,
> > +			  unsigned long nr_pages, struct page **pages);
> >
> >   int get_user_pages_fast(unsigned long start, int nr_pages,
> >   			unsigned int gup_flags, struct page **pages);
> > diff --git a/mm/gup.c b/mm/gup.c
> > index 231711efa390..eb93d1ec9dc6 100644
> > --- a/mm/gup.c
> > +++ b/mm/gup.c
> > @@ -5,6 +5,7 @@
> >   #include <linux/spinlock.h>
> >
> >   #include <linux/mm.h>
> > +#include <linux/memfd.h>
> >   #include <linux/memremap.h>
> >   #include <linux/pagemap.h>
> >   #include <linux/rmap.h>
> > @@ -17,6 +18,7 @@
> >   #include <linux/hugetlb.h>
> >   #include <linux/migrate.h>
> >   #include <linux/mm_inline.h>
> > +#include <linux/pagevec.h>
> >   #include <linux/sched/mm.h>
> >   #include <linux/shmem_fs.h>
> >
> > @@ -3410,3 +3412,103 @@ long pin_user_pages_unlocked(unsigned long
> start, unsigned long nr_pages,
> >   				     &locked, gup_flags);
> >   }
> >   EXPORT_SYMBOL(pin_user_pages_unlocked);
> > +
> > +/**
> > + * memfd_pin_user_pages() - pin user pages associated with a memfd
> > + * @memfd:      the memfd whose pages are to be pinned
> > + * @start:      starting memfd offset
> > + * @nr_pages:   number of pages from start to pin
> > + * @pages:      array that receives pointers to the pages pinned.
> > + *              Should be at-least nr_pages long.
> > + *
> > + * Attempt to pin pages associated with a memfd; given that a memfd is
> either
> > + * backed by shmem or hugetlb, the pages can either be found in the page
> cache
> > + * or need to be allocated if necessary. Once the pages are located, they
> are
> > + * all pinned via FOLL_PIN. And, these pinned pages need to be released
> either
> > + * using unpin_user_pages() or unpin_user_page().
> > + *
> > + * It must be noted that the pages may be pinned for an indefinite amount
> > + * of time. And, in most cases, the duration of time they may stay pinned
> > + * would be controlled by the userspace. This behavior is effectively the
> > + * same as using FOLL_LONGTERM with other GUP APIs.
> > + *
> > + * Returns number of pages pinned. This would be equal to the number of
> > + * pages requested. If no pages were pinned, it returns -errno.
> > + */
> > +long memfd_pin_user_pages(struct file *memfd, pgoff_t start,
> > +			  unsigned long nr_pages, struct page **pages)
> > +{
> > +	pgoff_t start_idx, end_idx = start + nr_pages - 1;
> > +	unsigned int flags, nr_folios, i, j;
> > +	struct folio_batch fbatch;
> > +	struct page *page = NULL;
> > +	struct folio *folio;
> > +	long ret;
> > +
> > +	if (!nr_pages)
> > +		return -EINVAL;
> > +
> > +	if (!memfd)
> > +		return -EINVAL;
> > +
> > +	if (!shmem_file(memfd) && !is_file_hugepages(memfd))
> > +		return -EINVAL;
> > +
> > +	flags = memalloc_pin_save();
> > +	do {
> > +		folio_batch_init(&fbatch);
> > +		start_idx = start;
> > +		i = 0;
> > +
> > +		while (start_idx <= end_idx) {
> > +			/*
> > +			 * In most cases, we should be able to find the page
> > +			 * in the page cache. If we cannot find it for some
> > +			 * reason, we try to allocate one and add it to the
> > +			 * page cache.
> > +			 */
> > +			nr_folios = filemap_get_folios_contig(memfd-
> >f_mapping,
> > +							      &start_idx,
> > +							      end_idx,
> > +							      &fbatch);
> > +			if (page) {
> > +				put_page(page);
> > +				page = NULL;
> > +			}
> > +			for (j = 0; j < nr_folios; j++) {
> > +				folio = fbatch.folios[j];
> > +				ret = try_grab_page(&folio->page, FOLL_PIN);
> > +				if (unlikely(ret)) {
> > +					folio_batch_release(&fbatch);
> > +					goto err;
> > +				}
> > +
> > +				pages[i++] = &folio->page;
> > +			}
> 
> I might be wrong, but that interface is still inconsistent. I think your
> intention is to always return folios (head pages), but why are we
> returning pages from this interface then?
> 
> It would be more consistent regarding the other GUP interfaces to return
> the actual tail pages that fit the given "pgoff_t start". So if you
> punch in "nr_pages" you expect to get "nr_pages" pages, and not some
> other number of folios.
> 
> Otherwise, this interface is highly confusing.
> 
> If you always want to return folios, then better name it
> "memfd_pin_user_folios" (or just "memfd_pin_folios") and pass in a range
> (instead of a nr_pages parameter), and somehow indicate to the caller
> how many folio were in that range, and if that range was fully covered.
I think it makes sense to return folios from this interface; and considering my
use-case, I'd like have this API return an error if it cannot pin (or allocate)
the exact number of folios the caller requested. 

> 
> Or am I missing something?
I can make the udmabuf driver use folios instead of pages too but the function
check_and_migrate_movable_pages() in GUP still takes a list of pages. Do you
think it is ok to use a local variable to collect all the head pages for this?

Thanks,
Vivek

> 
> --
> Cheers,
> 
> David / dhildenb
> 

Re: [PATCH v6 3/5] mm/gup: Introduce memfd_pin_user_pages() for pinning memfd pages (v6)

[David Hildenbrand]

On 07.12.23 06:09, Kasireddy, Vivek wrote:
> Hi David,
> 
>> On 05.12.23 06:35, Vivek Kasireddy wrote:
>>> For drivers that would like to longterm-pin the pages associated
>>> with a memfd, the pin_user_pages_fd() API provides an option to
>>> not only pin the pages via FOLL_PIN but also to check and migrate
>>> them if they reside in movable zone or CMA block. This API
>>> currently works with memfds but it should work with any files
>>> that belong to either shmemfs or hugetlbfs. Files belonging to
>>> other filesystems are rejected for now.
>>>
>>> The pages need to be located first before pinning them via FOLL_PIN.
>>> If they are found in the page cache, they can be immediately pinned.
>>> Otherwise, they need to be allocated using the filesystem specific
>>> APIs and then pinned.
>>>
>>> v2:
>>> - Drop gup_flags and improve comments and commit message (David)
>>> - Allocate a page if we cannot find in page cache for the hugetlbfs
>>>     case as well (David)
>>> - Don't unpin pages if there is a migration related failure (David)
>>> - Drop the unnecessary nr_pages <= 0 check (Jason)
>>> - Have the caller of the API pass in file * instead of fd (Jason)
>>>
>>> v3: (David)
>>> - Enclose the huge page allocation code with #ifdef
>> CONFIG_HUGETLB_PAGE
>>>     (Build error reported by kernel test robot <lkp@intel.com>)
>>> - Don't forget memalloc_pin_restore() on non-migration related errors
>>> - Improve the readability of the cleanup code associated with
>>>     non-migration related errors
>>> - Augment the comments by describing FOLL_LONGTERM like behavior
>>> - Include the R-b tag from Jason
>>>
>>> v4:
>>> - Remove the local variable "page" and instead use 3 return statements
>>>     in alloc_file_page() (David)
>>> - Add the R-b tag from David
>>>
>>> v5: (David)
>>> - For hugetlb case, ensure that we only obtain head pages from the
>>>     mapping by using __filemap_get_folio() instead of find_get_page_flags()
>>> - Handle -EEXIST when two or more potential users try to simultaneously
>>>     add a huge page to the mapping by forcing them to retry on failure
>>>
>>> v6: (Christoph)
>>> - Rename this API to memfd_pin_user_pages() to make it clear that it
>>>     is intended for memfds
>>> - Move the memfd page allocation helper from gup.c to memfd.c
>>> - Fix indentation errors in memfd_pin_user_pages()
>>> - For contiguous ranges of folios, use a helper such as
>>>     filemap_get_folios_contig() to lookup the page cache in batches
>>>
>>> Cc: David Hildenbrand <david@redhat.com>
>>> Cc: Christoph Hellwig <hch@infradead.org>
>>> Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
>>> Cc: Mike Kravetz <mike.kravetz@oracle.com>
>>> Cc: Hugh Dickins <hughd@google.com>
>>> Cc: Peter Xu <peterx@redhat.com>
>>> Cc: Gerd Hoffmann <kraxel@redhat.com>
>>> Cc: Dongwon Kim <dongwon.kim@intel.com>
>>> Cc: Junxiao Chang <junxiao.chang@intel.com>
>>> Suggested-by: Jason Gunthorpe <jgg@nvidia.com>
>>> Reviewed-by: Jason Gunthorpe <jgg@nvidia.com> (v2)
>>> Reviewed-by: David Hildenbrand <david@redhat.com> (v3)
>>> Signed-off-by: Vivek Kasireddy <vivek.kasireddy@intel.com>
>>> ---
>>>    include/linux/memfd.h |   5 +++
>>>    include/linux/mm.h    |   2 +
>>>    mm/gup.c              | 102 ++++++++++++++++++++++++++++++++++++++++++
>>>    mm/memfd.c            |  34 ++++++++++++++
>>>    4 files changed, 143 insertions(+)
>>>
>>> diff --git a/include/linux/memfd.h b/include/linux/memfd.h
>>> index e7abf6fa4c52..6fc0d1282151 100644
>>> --- a/include/linux/memfd.h
>>> +++ b/include/linux/memfd.h
>>> @@ -6,11 +6,16 @@
>>>
>>>    #ifdef CONFIG_MEMFD_CREATE
>>>    extern long memfd_fcntl(struct file *file, unsigned int cmd, unsigned int
>> arg);
>>> +extern struct page *memfd_alloc_page(struct file *memfd, pgoff_t idx);
>>>    #else
>>>    static inline long memfd_fcntl(struct file *f, unsigned int c, unsigned int a)
>>>    {
>>>    	return -EINVAL;
>>>    }
>>> +static inline struct page *memfd_alloc_page(struct file *memfd, pgoff_t
>> idx)
>>> +{
>>> +	return ERR_PTR(-EINVAL);
>>> +}
>>>    #endif
>>>
>>>    #endif /* __LINUX_MEMFD_H */
>>> diff --git a/include/linux/mm.h b/include/linux/mm.h
>>> index 418d26608ece..ac69db45509f 100644
>>> --- a/include/linux/mm.h
>>> +++ b/include/linux/mm.h
>>> @@ -2472,6 +2472,8 @@ long get_user_pages_unlocked(unsigned long
>> start, unsigned long nr_pages,
>>>    		    struct page **pages, unsigned int gup_flags);
>>>    long pin_user_pages_unlocked(unsigned long start, unsigned long
>> nr_pages,
>>>    		    struct page **pages, unsigned int gup_flags);
>>> +long memfd_pin_user_pages(struct file *file, pgoff_t start,
>>> +			  unsigned long nr_pages, struct page **pages);
>>>
>>>    int get_user_pages_fast(unsigned long start, int nr_pages,
>>>    			unsigned int gup_flags, struct page **pages);
>>> diff --git a/mm/gup.c b/mm/gup.c
>>> index 231711efa390..eb93d1ec9dc6 100644
>>> --- a/mm/gup.c
>>> +++ b/mm/gup.c
>>> @@ -5,6 +5,7 @@
>>>    #include <linux/spinlock.h>
>>>
>>>    #include <linux/mm.h>
>>> +#include <linux/memfd.h>
>>>    #include <linux/memremap.h>
>>>    #include <linux/pagemap.h>
>>>    #include <linux/rmap.h>
>>> @@ -17,6 +18,7 @@
>>>    #include <linux/hugetlb.h>
>>>    #include <linux/migrate.h>
>>>    #include <linux/mm_inline.h>
>>> +#include <linux/pagevec.h>
>>>    #include <linux/sched/mm.h>
>>>    #include <linux/shmem_fs.h>
>>>
>>> @@ -3410,3 +3412,103 @@ long pin_user_pages_unlocked(unsigned long
>> start, unsigned long nr_pages,
>>>    				     &locked, gup_flags);
>>>    }
>>>    EXPORT_SYMBOL(pin_user_pages_unlocked);
>>> +
>>> +/**
>>> + * memfd_pin_user_pages() - pin user pages associated with a memfd
>>> + * @memfd:      the memfd whose pages are to be pinned
>>> + * @start:      starting memfd offset
>>> + * @nr_pages:   number of pages from start to pin
>>> + * @pages:      array that receives pointers to the pages pinned.
>>> + *              Should be at-least nr_pages long.
>>> + *
>>> + * Attempt to pin pages associated with a memfd; given that a memfd is
>> either
>>> + * backed by shmem or hugetlb, the pages can either be found in the page
>> cache
>>> + * or need to be allocated if necessary. Once the pages are located, they
>> are
>>> + * all pinned via FOLL_PIN. And, these pinned pages need to be released
>> either
>>> + * using unpin_user_pages() or unpin_user_page().
>>> + *
>>> + * It must be noted that the pages may be pinned for an indefinite amount
>>> + * of time. And, in most cases, the duration of time they may stay pinned
>>> + * would be controlled by the userspace. This behavior is effectively the
>>> + * same as using FOLL_LONGTERM with other GUP APIs.
>>> + *
>>> + * Returns number of pages pinned. This would be equal to the number of
>>> + * pages requested. If no pages were pinned, it returns -errno.
>>> + */
>>> +long memfd_pin_user_pages(struct file *memfd, pgoff_t start,
>>> +			  unsigned long nr_pages, struct page **pages)
>>> +{
>>> +	pgoff_t start_idx, end_idx = start + nr_pages - 1;
>>> +	unsigned int flags, nr_folios, i, j;
>>> +	struct folio_batch fbatch;
>>> +	struct page *page = NULL;
>>> +	struct folio *folio;
>>> +	long ret;
>>> +
>>> +	if (!nr_pages)
>>> +		return -EINVAL;
>>> +
>>> +	if (!memfd)
>>> +		return -EINVAL;
>>> +
>>> +	if (!shmem_file(memfd) && !is_file_hugepages(memfd))
>>> +		return -EINVAL;
>>> +
>>> +	flags = memalloc_pin_save();
>>> +	do {
>>> +		folio_batch_init(&fbatch);
>>> +		start_idx = start;
>>> +		i = 0;
>>> +
>>> +		while (start_idx <= end_idx) {
>>> +			/*
>>> +			 * In most cases, we should be able to find the page
>>> +			 * in the page cache. If we cannot find it for some
>>> +			 * reason, we try to allocate one and add it to the
>>> +			 * page cache.
>>> +			 */
>>> +			nr_folios = filemap_get_folios_contig(memfd-
>>> f_mapping,
>>> +							      &start_idx,
>>> +							      end_idx,
>>> +							      &fbatch);
>>> +			if (page) {
>>> +				put_page(page);
>>> +				page = NULL;
>>> +			}
>>> +			for (j = 0; j < nr_folios; j++) {
>>> +				folio = fbatch.folios[j];
>>> +				ret = try_grab_page(&folio->page, FOLL_PIN);
>>> +				if (unlikely(ret)) {
>>> +					folio_batch_release(&fbatch);
>>> +					goto err;
>>> +				}
>>> +
>>> +				pages[i++] = &folio->page;
>>> +			}
>>
>> I might be wrong, but that interface is still inconsistent. I think your
>> intention is to always return folios (head pages), but why are we
>> returning pages from this interface then?
>>
>> It would be more consistent regarding the other GUP interfaces to return
>> the actual tail pages that fit the given "pgoff_t start". So if you
>> punch in "nr_pages" you expect to get "nr_pages" pages, and not some
>> other number of folios.
>>
>> Otherwise, this interface is highly confusing.
>>
>> If you always want to return folios, then better name it
>> "memfd_pin_user_folios" (or just "memfd_pin_folios") and pass in a range
>> (instead of a nr_pages parameter), and somehow indicate to the caller
>> how many folio were in that range, and if that range was fully covered.
> I think it makes sense to return folios from this interface; and considering my
> use-case, I'd like have this API return an error if it cannot pin (or allocate)
> the exact number of folios the caller requested.

Okay, then better use folios.

Assuming a caller puts in "start = X" and gets some large folio back. 
How is the caller supposed to know at which offset to look into that 
folio (IOW< which subpage)? For "pages" it was obvious (you get the 
actual subpages), but as soon as we return a large folio, some 
information is missing for the caller.

How can the caller figure that out?

> 
>>
>> Or am I missing something?
> I can make the udmabuf driver use folios instead of pages too but the function
> check_and_migrate_movable_pages() in GUP still takes a list of pages. Do you
> think it is ok to use a local variable to collect all the head pages for this?

I think you can simply pass in the head page, because only whole folios 
can be converted. At some point we should convert that one to use folios 
as well.

-- 
Cheers,

David / dhildenb

Re: [PATCH v6 3/5] mm/gup: Introduce memfd_pin_user_pages() for pinning memfd pages (v6)

[Jason Gunthorpe]

On Thu, Dec 07, 2023 at 10:44:14AM +0100, David Hildenbrand wrote:

> > > If you always want to return folios, then better name it
> > > "memfd_pin_user_folios" (or just "memfd_pin_folios") and pass in a range
> > > (instead of a nr_pages parameter), and somehow indicate to the caller
> > > how many folio were in that range, and if that range was fully covered.
> > I think it makes sense to return folios from this interface; and considering my
> > use-case, I'd like have this API return an error if it cannot pin (or allocate)
> > the exact number of folios the caller requested.
> 
> Okay, then better use folios.
> 
> Assuming a caller puts in "start = X" and gets some large folio back. How is
> the caller supposed to know at which offset to look into that folio (IOW<
> which subpage)? For "pages" it was obvious (you get the actual subpages),
> but as soon as we return a large folio, some information is missing for the
> caller.
> 
> How can the caller figure that out?

This can only work if the memfd is required to only have full folios
at aligned locations. Under that restriction computing the first folio
offset is easy enough:

  folio offset = (start % folio size)

But is that true for the memfds here?

> > I can make the udmabuf driver use folios instead of pages too but the function
> > check_and_migrate_movable_pages() in GUP still takes a list of pages. Do you
> > think it is ok to use a local variable to collect all the head pages for this?
> 
> I think you can simply pass in the head page, because only whole folios can
> be converted. At some point we should convert that one to use folios as
> well.

It is like that because it processes the output from GUP in-place
which is a page list..

Probably what we need to do is make the migration checks happen while
accumulating the pages so we don't need to scan the output list..

Jason

Re: [PATCH v6 3/5] mm/gup: Introduce memfd_pin_user_pages() for pinning memfd pages (v6)

[David Hildenbrand]

On 07.12.23 14:05, Jason Gunthorpe wrote:
> On Thu, Dec 07, 2023 at 10:44:14AM +0100, David Hildenbrand wrote:
> 
>>>> If you always want to return folios, then better name it
>>>> "memfd_pin_user_folios" (or just "memfd_pin_folios") and pass in a range
>>>> (instead of a nr_pages parameter), and somehow indicate to the caller
>>>> how many folio were in that range, and if that range was fully covered.
>>> I think it makes sense to return folios from this interface; and considering my
>>> use-case, I'd like have this API return an error if it cannot pin (or allocate)
>>> the exact number of folios the caller requested.
>>
>> Okay, then better use folios.
>>
>> Assuming a caller puts in "start = X" and gets some large folio back. How is
>> the caller supposed to know at which offset to look into that folio (IOW<
>> which subpage)? For "pages" it was obvious (you get the actual subpages),
>> but as soon as we return a large folio, some information is missing for the
>> caller.
>>
>> How can the caller figure that out?
> 
> This can only work if the memfd is required to only have full folios
> at aligned locations. Under that restriction computing the first folio
> offset is easy enough:
> 
>    folio offset = (start % folio size)
> 
> But is that true for the memfds here?

I assume folios are always naturally aligned, like:

[ 2m ][ 2m ][1m][1m][ 2m ]
^f0   ^f1   ^f2 ^f3 ^f4

If you query the range "3m -> 7m", you get back f1,f2,f3,f4 and have to 
start in the middle of the first folio with offset 1m. From there, it is
indeed simply continuing with the full folio size -- until the last 
folio, where you want to only process 1m.

folio offset = (1m % 2m)

would be correct in that case.

-- 
Cheers,

David / dhildenb

RE: [PATCH v6 3/5] mm/gup: Introduce memfd_pin_user_pages() for pinning memfd pages (v6)

[Kasireddy, Vivek]

Hi David,

> >
> >> On 05.12.23 06:35, Vivek Kasireddy wrote:
> >>> For drivers that would like to longterm-pin the pages associated
> >>> with a memfd, the pin_user_pages_fd() API provides an option to
> >>> not only pin the pages via FOLL_PIN but also to check and migrate
> >>> them if they reside in movable zone or CMA block. This API
> >>> currently works with memfds but it should work with any files
> >>> that belong to either shmemfs or hugetlbfs. Files belonging to
> >>> other filesystems are rejected for now.
> >>>
> >>> The pages need to be located first before pinning them via FOLL_PIN.
> >>> If they are found in the page cache, they can be immediately pinned.
> >>> Otherwise, they need to be allocated using the filesystem specific
> >>> APIs and then pinned.
> >>>
> >>> v2:
> >>> - Drop gup_flags and improve comments and commit message (David)
> >>> - Allocate a page if we cannot find in page cache for the hugetlbfs
> >>>     case as well (David)
> >>> - Don't unpin pages if there is a migration related failure (David)
> >>> - Drop the unnecessary nr_pages <= 0 check (Jason)
> >>> - Have the caller of the API pass in file * instead of fd (Jason)
> >>>
> >>> v3: (David)
> >>> - Enclose the huge page allocation code with #ifdef
> >> CONFIG_HUGETLB_PAGE
> >>>     (Build error reported by kernel test robot <lkp@intel.com>)
> >>> - Don't forget memalloc_pin_restore() on non-migration related errors
> >>> - Improve the readability of the cleanup code associated with
> >>>     non-migration related errors
> >>> - Augment the comments by describing FOLL_LONGTERM like behavior
> >>> - Include the R-b tag from Jason
> >>>
> >>> v4:
> >>> - Remove the local variable "page" and instead use 3 return statements
> >>>     in alloc_file_page() (David)
> >>> - Add the R-b tag from David
> >>>
> >>> v5: (David)
> >>> - For hugetlb case, ensure that we only obtain head pages from the
> >>>     mapping by using __filemap_get_folio() instead of
> find_get_page_flags()
> >>> - Handle -EEXIST when two or more potential users try to simultaneously
> >>>     add a huge page to the mapping by forcing them to retry on failure
> >>>
> >>> v6: (Christoph)
> >>> - Rename this API to memfd_pin_user_pages() to make it clear that it
> >>>     is intended for memfds
> >>> - Move the memfd page allocation helper from gup.c to memfd.c
> >>> - Fix indentation errors in memfd_pin_user_pages()
> >>> - For contiguous ranges of folios, use a helper such as
> >>>     filemap_get_folios_contig() to lookup the page cache in batches
> >>>
> >>> Cc: David Hildenbrand <david@redhat.com>
> >>> Cc: Christoph Hellwig <hch@infradead.org>
> >>> Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
> >>> Cc: Mike Kravetz <mike.kravetz@oracle.com>
> >>> Cc: Hugh Dickins <hughd@google.com>
> >>> Cc: Peter Xu <peterx@redhat.com>
> >>> Cc: Gerd Hoffmann <kraxel@redhat.com>
> >>> Cc: Dongwon Kim <dongwon.kim@intel.com>
> >>> Cc: Junxiao Chang <junxiao.chang@intel.com>
> >>> Suggested-by: Jason Gunthorpe <jgg@nvidia.com>
> >>> Reviewed-by: Jason Gunthorpe <jgg@nvidia.com> (v2)
> >>> Reviewed-by: David Hildenbrand <david@redhat.com> (v3)
> >>> Signed-off-by: Vivek Kasireddy <vivek.kasireddy@intel.com>
> >>> ---
> >>>    include/linux/memfd.h |   5 +++
> >>>    include/linux/mm.h    |   2 +
> >>>    mm/gup.c              | 102
> ++++++++++++++++++++++++++++++++++++++++++
> >>>    mm/memfd.c            |  34 ++++++++++++++
> >>>    4 files changed, 143 insertions(+)
> >>>
> >>> diff --git a/include/linux/memfd.h b/include/linux/memfd.h
> >>> index e7abf6fa4c52..6fc0d1282151 100644
> >>> --- a/include/linux/memfd.h
> >>> +++ b/include/linux/memfd.h
> >>> @@ -6,11 +6,16 @@
> >>>
> >>>    #ifdef CONFIG_MEMFD_CREATE
> >>>    extern long memfd_fcntl(struct file *file, unsigned int cmd, unsigned int
> >> arg);
> >>> +extern struct page *memfd_alloc_page(struct file *memfd, pgoff_t idx);
> >>>    #else
> >>>    static inline long memfd_fcntl(struct file *f, unsigned int c, unsigned int
> a)
> >>>    {
> >>>    	return -EINVAL;
> >>>    }
> >>> +static inline struct page *memfd_alloc_page(struct file *memfd, pgoff_t
> >> idx)
> >>> +{
> >>> +	return ERR_PTR(-EINVAL);
> >>> +}
> >>>    #endif
> >>>
> >>>    #endif /* __LINUX_MEMFD_H */
> >>> diff --git a/include/linux/mm.h b/include/linux/mm.h
> >>> index 418d26608ece..ac69db45509f 100644
> >>> --- a/include/linux/mm.h
> >>> +++ b/include/linux/mm.h
> >>> @@ -2472,6 +2472,8 @@ long get_user_pages_unlocked(unsigned long
> >> start, unsigned long nr_pages,
> >>>    		    struct page **pages, unsigned int gup_flags);
> >>>    long pin_user_pages_unlocked(unsigned long start, unsigned long
> >> nr_pages,
> >>>    		    struct page **pages, unsigned int gup_flags);
> >>> +long memfd_pin_user_pages(struct file *file, pgoff_t start,
> >>> +			  unsigned long nr_pages, struct page **pages);
> >>>
> >>>    int get_user_pages_fast(unsigned long start, int nr_pages,
> >>>    			unsigned int gup_flags, struct page **pages);
> >>> diff --git a/mm/gup.c b/mm/gup.c
> >>> index 231711efa390..eb93d1ec9dc6 100644
> >>> --- a/mm/gup.c
> >>> +++ b/mm/gup.c
> >>> @@ -5,6 +5,7 @@
> >>>    #include <linux/spinlock.h>
> >>>
> >>>    #include <linux/mm.h>
> >>> +#include <linux/memfd.h>
> >>>    #include <linux/memremap.h>
> >>>    #include <linux/pagemap.h>
> >>>    #include <linux/rmap.h>
> >>> @@ -17,6 +18,7 @@
> >>>    #include <linux/hugetlb.h>
> >>>    #include <linux/migrate.h>
> >>>    #include <linux/mm_inline.h>
> >>> +#include <linux/pagevec.h>
> >>>    #include <linux/sched/mm.h>
> >>>    #include <linux/shmem_fs.h>
> >>>
> >>> @@ -3410,3 +3412,103 @@ long pin_user_pages_unlocked(unsigned
> long
> >> start, unsigned long nr_pages,
> >>>    				     &locked, gup_flags);
> >>>    }
> >>>    EXPORT_SYMBOL(pin_user_pages_unlocked);
> >>> +
> >>> +/**
> >>> + * memfd_pin_user_pages() - pin user pages associated with a memfd
> >>> + * @memfd:      the memfd whose pages are to be pinned
> >>> + * @start:      starting memfd offset
> >>> + * @nr_pages:   number of pages from start to pin
> >>> + * @pages:      array that receives pointers to the pages pinned.
> >>> + *              Should be at-least nr_pages long.
> >>> + *
> >>> + * Attempt to pin pages associated with a memfd; given that a memfd is
> >> either
> >>> + * backed by shmem or hugetlb, the pages can either be found in the
> page
> >> cache
> >>> + * or need to be allocated if necessary. Once the pages are located, they
> >> are
> >>> + * all pinned via FOLL_PIN. And, these pinned pages need to be released
> >> either
> >>> + * using unpin_user_pages() or unpin_user_page().
> >>> + *
> >>> + * It must be noted that the pages may be pinned for an indefinite
> amount
> >>> + * of time. And, in most cases, the duration of time they may stay
> pinned
> >>> + * would be controlled by the userspace. This behavior is effectively the
> >>> + * same as using FOLL_LONGTERM with other GUP APIs.
> >>> + *
> >>> + * Returns number of pages pinned. This would be equal to the number
> of
> >>> + * pages requested. If no pages were pinned, it returns -errno.
> >>> + */
> >>> +long memfd_pin_user_pages(struct file *memfd, pgoff_t start,
> >>> +			  unsigned long nr_pages, struct page **pages)
> >>> +{
> >>> +	pgoff_t start_idx, end_idx = start + nr_pages - 1;
> >>> +	unsigned int flags, nr_folios, i, j;
> >>> +	struct folio_batch fbatch;
> >>> +	struct page *page = NULL;
> >>> +	struct folio *folio;
> >>> +	long ret;
> >>> +
> >>> +	if (!nr_pages)
> >>> +		return -EINVAL;
> >>> +
> >>> +	if (!memfd)
> >>> +		return -EINVAL;
> >>> +
> >>> +	if (!shmem_file(memfd) && !is_file_hugepages(memfd))
> >>> +		return -EINVAL;
> >>> +
> >>> +	flags = memalloc_pin_save();
> >>> +	do {
> >>> +		folio_batch_init(&fbatch);
> >>> +		start_idx = start;
> >>> +		i = 0;
> >>> +
> >>> +		while (start_idx <= end_idx) {
> >>> +			/*
> >>> +			 * In most cases, we should be able to find the page
> >>> +			 * in the page cache. If we cannot find it for some
> >>> +			 * reason, we try to allocate one and add it to the
> >>> +			 * page cache.
> >>> +			 */
> >>> +			nr_folios = filemap_get_folios_contig(memfd-
> >>> f_mapping,
> >>> +							      &start_idx,
> >>> +							      end_idx,
> >>> +							      &fbatch);
> >>> +			if (page) {
> >>> +				put_page(page);
> >>> +				page = NULL;
> >>> +			}
> >>> +			for (j = 0; j < nr_folios; j++) {
> >>> +				folio = fbatch.folios[j];
> >>> +				ret = try_grab_page(&folio->page, FOLL_PIN);
> >>> +				if (unlikely(ret)) {
> >>> +					folio_batch_release(&fbatch);
> >>> +					goto err;
> >>> +				}
> >>> +
> >>> +				pages[i++] = &folio->page;
> >>> +			}
> >>
> >> I might be wrong, but that interface is still inconsistent. I think your
> >> intention is to always return folios (head pages), but why are we
> >> returning pages from this interface then?
> >>
> >> It would be more consistent regarding the other GUP interfaces to return
> >> the actual tail pages that fit the given "pgoff_t start". So if you
> >> punch in "nr_pages" you expect to get "nr_pages" pages, and not some
> >> other number of folios.
> >>
> >> Otherwise, this interface is highly confusing.
> >>
> >> If you always want to return folios, then better name it
> >> "memfd_pin_user_folios" (or just "memfd_pin_folios") and pass in a
> range
> >> (instead of a nr_pages parameter), and somehow indicate to the caller
> >> how many folio were in that range, and if that range was fully covered.
> > I think it makes sense to return folios from this interface; and considering
> my
> > use-case, I'd like have this API return an error if it cannot pin (or allocate)
> > the exact number of folios the caller requested.
> 
> Okay, then better use folios.
> 
> Assuming a caller puts in "start = X" and gets some large folio back.
> How is the caller supposed to know at which offset to look into that
> folio (IOW< which subpage)? For "pages" it was obvious (you get the
> actual subpages), but as soon as we return a large folio, some
> information is missing for the caller.
> 
> How can the caller figure that out?
I think the caller can deduce the subpage offset one way or the other. For
example, in udmabuf we do:
subpgoff = offset & ~huge_page_mask(hpstate);
Do you think it is reasonable to expect callers to make these deductions?
Or would it make more sense to have the caller pass in an empty array
that this API would eventually fill with offsets?

Thanks,
Vivek
> 
> >
> >>
> >> Or am I missing something?
> > I can make the udmabuf driver use folios instead of pages too but the
> function
> > check_and_migrate_movable_pages() in GUP still takes a list of pages. Do
> you
> > think it is ok to use a local variable to collect all the head pages for this?
> 
> I think you can simply pass in the head page, because only whole folios
> can be converted. At some point we should convert that one to use folios
> as well.
> 
> --
> Cheers,
> 
> David / dhildenb

Re: [PATCH v6 3/5] mm/gup: Introduce memfd_pin_user_pages() for pinning memfd pages (v6)

[David Hildenbrand]

On 08.12.23 08:57, Kasireddy, Vivek wrote:
> Hi David,
> 
>>>
>>>> On 05.12.23 06:35, Vivek Kasireddy wrote:
>>>>> For drivers that would like to longterm-pin the pages associated
>>>>> with a memfd, the pin_user_pages_fd() API provides an option to
>>>>> not only pin the pages via FOLL_PIN but also to check and migrate
>>>>> them if they reside in movable zone or CMA block. This API
>>>>> currently works with memfds but it should work with any files
>>>>> that belong to either shmemfs or hugetlbfs. Files belonging to
>>>>> other filesystems are rejected for now.
>>>>>
>>>>> The pages need to be located first before pinning them via FOLL_PIN.
>>>>> If they are found in the page cache, they can be immediately pinned.
>>>>> Otherwise, they need to be allocated using the filesystem specific
>>>>> APIs and then pinned.
>>>>>
>>>>> v2:
>>>>> - Drop gup_flags and improve comments and commit message (David)
>>>>> - Allocate a page if we cannot find in page cache for the hugetlbfs
>>>>>      case as well (David)
>>>>> - Don't unpin pages if there is a migration related failure (David)
>>>>> - Drop the unnecessary nr_pages <= 0 check (Jason)
>>>>> - Have the caller of the API pass in file * instead of fd (Jason)
>>>>>
>>>>> v3: (David)
>>>>> - Enclose the huge page allocation code with #ifdef
>>>> CONFIG_HUGETLB_PAGE
>>>>>      (Build error reported by kernel test robot <lkp@intel.com>)
>>>>> - Don't forget memalloc_pin_restore() on non-migration related errors
>>>>> - Improve the readability of the cleanup code associated with
>>>>>      non-migration related errors
>>>>> - Augment the comments by describing FOLL_LONGTERM like behavior
>>>>> - Include the R-b tag from Jason
>>>>>
>>>>> v4:
>>>>> - Remove the local variable "page" and instead use 3 return statements
>>>>>      in alloc_file_page() (David)
>>>>> - Add the R-b tag from David
>>>>>
>>>>> v5: (David)
>>>>> - For hugetlb case, ensure that we only obtain head pages from the
>>>>>      mapping by using __filemap_get_folio() instead of
>> find_get_page_flags()
>>>>> - Handle -EEXIST when two or more potential users try to simultaneously
>>>>>      add a huge page to the mapping by forcing them to retry on failure
>>>>>
>>>>> v6: (Christoph)
>>>>> - Rename this API to memfd_pin_user_pages() to make it clear that it
>>>>>      is intended for memfds
>>>>> - Move the memfd page allocation helper from gup.c to memfd.c
>>>>> - Fix indentation errors in memfd_pin_user_pages()
>>>>> - For contiguous ranges of folios, use a helper such as
>>>>>      filemap_get_folios_contig() to lookup the page cache in batches
>>>>>
>>>>> Cc: David Hildenbrand <david@redhat.com>
>>>>> Cc: Christoph Hellwig <hch@infradead.org>
>>>>> Cc: Daniel Vetter <daniel.vetter@ffwll.ch>
>>>>> Cc: Mike Kravetz <mike.kravetz@oracle.com>
>>>>> Cc: Hugh Dickins <hughd@google.com>
>>>>> Cc: Peter Xu <peterx@redhat.com>
>>>>> Cc: Gerd Hoffmann <kraxel@redhat.com>
>>>>> Cc: Dongwon Kim <dongwon.kim@intel.com>
>>>>> Cc: Junxiao Chang <junxiao.chang@intel.com>
>>>>> Suggested-by: Jason Gunthorpe <jgg@nvidia.com>
>>>>> Reviewed-by: Jason Gunthorpe <jgg@nvidia.com> (v2)
>>>>> Reviewed-by: David Hildenbrand <david@redhat.com> (v3)
>>>>> Signed-off-by: Vivek Kasireddy <vivek.kasireddy@intel.com>
>>>>> ---
>>>>>     include/linux/memfd.h |   5 +++
>>>>>     include/linux/mm.h    |   2 +
>>>>>     mm/gup.c              | 102
>> ++++++++++++++++++++++++++++++++++++++++++
>>>>>     mm/memfd.c            |  34 ++++++++++++++
>>>>>     4 files changed, 143 insertions(+)
>>>>>
>>>>> diff --git a/include/linux/memfd.h b/include/linux/memfd.h
>>>>> index e7abf6fa4c52..6fc0d1282151 100644
>>>>> --- a/include/linux/memfd.h
>>>>> +++ b/include/linux/memfd.h
>>>>> @@ -6,11 +6,16 @@
>>>>>
>>>>>     #ifdef CONFIG_MEMFD_CREATE
>>>>>     extern long memfd_fcntl(struct file *file, unsigned int cmd, unsigned int
>>>> arg);
>>>>> +extern struct page *memfd_alloc_page(struct file *memfd, pgoff_t idx);
>>>>>     #else
>>>>>     static inline long memfd_fcntl(struct file *f, unsigned int c, unsigned int
>> a)
>>>>>     {
>>>>>     	return -EINVAL;
>>>>>     }
>>>>> +static inline struct page *memfd_alloc_page(struct file *memfd, pgoff_t
>>>> idx)
>>>>> +{
>>>>> +	return ERR_PTR(-EINVAL);
>>>>> +}
>>>>>     #endif
>>>>>
>>>>>     #endif /* __LINUX_MEMFD_H */
>>>>> diff --git a/include/linux/mm.h b/include/linux/mm.h
>>>>> index 418d26608ece..ac69db45509f 100644
>>>>> --- a/include/linux/mm.h
>>>>> +++ b/include/linux/mm.h
>>>>> @@ -2472,6 +2472,8 @@ long get_user_pages_unlocked(unsigned long
>>>> start, unsigned long nr_pages,
>>>>>     		    struct page **pages, unsigned int gup_flags);
>>>>>     long pin_user_pages_unlocked(unsigned long start, unsigned long
>>>> nr_pages,
>>>>>     		    struct page **pages, unsigned int gup_flags);
>>>>> +long memfd_pin_user_pages(struct file *file, pgoff_t start,
>>>>> +			  unsigned long nr_pages, struct page **pages);
>>>>>
>>>>>     int get_user_pages_fast(unsigned long start, int nr_pages,
>>>>>     			unsigned int gup_flags, struct page **pages);
>>>>> diff --git a/mm/gup.c b/mm/gup.c
>>>>> index 231711efa390..eb93d1ec9dc6 100644
>>>>> --- a/mm/gup.c
>>>>> +++ b/mm/gup.c
>>>>> @@ -5,6 +5,7 @@
>>>>>     #include <linux/spinlock.h>
>>>>>
>>>>>     #include <linux/mm.h>
>>>>> +#include <linux/memfd.h>
>>>>>     #include <linux/memremap.h>
>>>>>     #include <linux/pagemap.h>
>>>>>     #include <linux/rmap.h>
>>>>> @@ -17,6 +18,7 @@
>>>>>     #include <linux/hugetlb.h>
>>>>>     #include <linux/migrate.h>
>>>>>     #include <linux/mm_inline.h>
>>>>> +#include <linux/pagevec.h>
>>>>>     #include <linux/sched/mm.h>
>>>>>     #include <linux/shmem_fs.h>
>>>>>
>>>>> @@ -3410,3 +3412,103 @@ long pin_user_pages_unlocked(unsigned
>> long
>>>> start, unsigned long nr_pages,
>>>>>     				     &locked, gup_flags);
>>>>>     }
>>>>>     EXPORT_SYMBOL(pin_user_pages_unlocked);
>>>>> +
>>>>> +/**
>>>>> + * memfd_pin_user_pages() - pin user pages associated with a memfd
>>>>> + * @memfd:      the memfd whose pages are to be pinned
>>>>> + * @start:      starting memfd offset
>>>>> + * @nr_pages:   number of pages from start to pin
>>>>> + * @pages:      array that receives pointers to the pages pinned.
>>>>> + *              Should be at-least nr_pages long.
>>>>> + *
>>>>> + * Attempt to pin pages associated with a memfd; given that a memfd is
>>>> either
>>>>> + * backed by shmem or hugetlb, the pages can either be found in the
>> page
>>>> cache
>>>>> + * or need to be allocated if necessary. Once the pages are located, they
>>>> are
>>>>> + * all pinned via FOLL_PIN. And, these pinned pages need to be released
>>>> either
>>>>> + * using unpin_user_pages() or unpin_user_page().
>>>>> + *
>>>>> + * It must be noted that the pages may be pinned for an indefinite
>> amount
>>>>> + * of time. And, in most cases, the duration of time they may stay
>> pinned
>>>>> + * would be controlled by the userspace. This behavior is effectively the
>>>>> + * same as using FOLL_LONGTERM with other GUP APIs.
>>>>> + *
>>>>> + * Returns number of pages pinned. This would be equal to the number
>> of
>>>>> + * pages requested. If no pages were pinned, it returns -errno.
>>>>> + */
>>>>> +long memfd_pin_user_pages(struct file *memfd, pgoff_t start,
>>>>> +			  unsigned long nr_pages, struct page **pages)
>>>>> +{
>>>>> +	pgoff_t start_idx, end_idx = start + nr_pages - 1;
>>>>> +	unsigned int flags, nr_folios, i, j;
>>>>> +	struct folio_batch fbatch;
>>>>> +	struct page *page = NULL;
>>>>> +	struct folio *folio;
>>>>> +	long ret;
>>>>> +
>>>>> +	if (!nr_pages)
>>>>> +		return -EINVAL;
>>>>> +
>>>>> +	if (!memfd)
>>>>> +		return -EINVAL;
>>>>> +
>>>>> +	if (!shmem_file(memfd) && !is_file_hugepages(memfd))
>>>>> +		return -EINVAL;
>>>>> +
>>>>> +	flags = memalloc_pin_save();
>>>>> +	do {
>>>>> +		folio_batch_init(&fbatch);
>>>>> +		start_idx = start;
>>>>> +		i = 0;
>>>>> +
>>>>> +		while (start_idx <= end_idx) {
>>>>> +			/*
>>>>> +			 * In most cases, we should be able to find the page
>>>>> +			 * in the page cache. If we cannot find it for some
>>>>> +			 * reason, we try to allocate one and add it to the
>>>>> +			 * page cache.
>>>>> +			 */
>>>>> +			nr_folios = filemap_get_folios_contig(memfd-
>>>>> f_mapping,
>>>>> +							      &start_idx,
>>>>> +							      end_idx,
>>>>> +							      &fbatch);
>>>>> +			if (page) {
>>>>> +				put_page(page);
>>>>> +				page = NULL;
>>>>> +			}
>>>>> +			for (j = 0; j < nr_folios; j++) {
>>>>> +				folio = fbatch.folios[j];
>>>>> +				ret = try_grab_page(&folio->page, FOLL_PIN);
>>>>> +				if (unlikely(ret)) {
>>>>> +					folio_batch_release(&fbatch);
>>>>> +					goto err;
>>>>> +				}
>>>>> +
>>>>> +				pages[i++] = &folio->page;
>>>>> +			}
>>>>
>>>> I might be wrong, but that interface is still inconsistent. I think your
>>>> intention is to always return folios (head pages), but why are we
>>>> returning pages from this interface then?
>>>>
>>>> It would be more consistent regarding the other GUP interfaces to return
>>>> the actual tail pages that fit the given "pgoff_t start". So if you
>>>> punch in "nr_pages" you expect to get "nr_pages" pages, and not some
>>>> other number of folios.
>>>>
>>>> Otherwise, this interface is highly confusing.
>>>>
>>>> If you always want to return folios, then better name it
>>>> "memfd_pin_user_folios" (or just "memfd_pin_folios") and pass in a
>> range
>>>> (instead of a nr_pages parameter), and somehow indicate to the caller
>>>> how many folio were in that range, and if that range was fully covered.
>>> I think it makes sense to return folios from this interface; and considering
>> my
>>> use-case, I'd like have this API return an error if it cannot pin (or allocate)
>>> the exact number of folios the caller requested.
>>
>> Okay, then better use folios.
>>
>> Assuming a caller puts in "start = X" and gets some large folio back.
>> How is the caller supposed to know at which offset to look into that
>> folio (IOW< which subpage)? For "pages" it was obvious (you get the
>> actual subpages), but as soon as we return a large folio, some
>> information is missing for the caller.
>>
>> How can the caller figure that out?
> I think the caller can deduce the subpage offset one way or the other. For
> example, in udmabuf we do:
> subpgoff = offset & ~huge_page_mask(hpstate);

For hugetlb it's obvious, because we always have fixed-sized chunks for 
the whole memfd.


> Do you think it is reasonable to expect callers to make these deductions?
> Or would it make more sense to have the caller pass in an empty array
> that this API would eventually fill with offsets?

Only the offset into the first folio is really required. From there, 
it's mostly just processing full folios (well, the last folio might have 
to be processed partially, but that's easy to figure out by the caller 
by keeping track of the total range processed vs. requested).

Maybe it makes sense to return the offset into the first folio? Would 
certainly take some magic out of that interface.

-- 
Cheers,

David / dhildenb