mm/gup: cache dev_pagemap while pinning pages

Getting pages from ZONE_DEVICE memory needs to check the backing device's
live-ness, which is tracked in the device's dev_pagemap metadata.  This
metadata is stored in a radix tree and looking it up adds measurable
software overhead.

This patch avoids repeating this relatively costly operation when
dev_pagemap is used by caching the last dev_pagemap while getting user
pages.  The gup_benchmark kernel self test reports this reduces time to
get user pages to as low as 1/3 of the previous time.

Link: http://lkml.kernel.org/r/20181012173040.15669-1-keith.busch@intel.com
Signed-off-by: Keith Busch <keith.busch@intel.com>
Reviewed-by: Dan Williams <dan.j.williams@intel.com>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Dave Hansen <dave.hansen@intel.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

include/linux/huge_mm.h

@@ -213,9 +213,9 @@ static inline int hpage_nr_pages(struct page *page)
 }
 
 struct page *follow_devmap_pmd(struct vm_area_struct *vma, unsigned long addr,
-		pmd_t *pmd, int flags);
+		pmd_t *pmd, int flags, struct dev_pagemap **pgmap);
 struct page *follow_devmap_pud(struct vm_area_struct *vma, unsigned long addr,
-		pud_t *pud, int flags);
+		pud_t *pud, int flags, struct dev_pagemap **pgmap);
 
 extern vm_fault_t do_huge_pmd_numa_page(struct vm_fault *vmf, pmd_t orig_pmd);
 
@@ -344,13 +344,13 @@ static inline void mm_put_huge_zero_page(struct mm_struct *mm)
 }
 
 static inline struct page *follow_devmap_pmd(struct vm_area_struct *vma,
-		unsigned long addr, pmd_t *pmd, int flags)
+	unsigned long addr, pmd_t *pmd, int flags, struct dev_pagemap **pgmap)
 {
 	return NULL;
 }
 
 static inline struct page *follow_devmap_pud(struct vm_area_struct *vma,
-		unsigned long addr, pud_t *pud, int flags)
+	unsigned long addr, pud_t *pud, int flags, struct dev_pagemap **pgmap)
 {
 	return NULL;
 }

include/linux/mm.h

@@ -2536,16 +2536,8 @@ static inline vm_fault_t vmf_error(int err)
 	return VM_FAULT_SIGBUS;
 }
 
-struct page *follow_page_mask(struct vm_area_struct *vma,
-			      unsigned long address, unsigned int foll_flags,
-			      unsigned int *page_mask);
-
-static inline struct page *follow_page(struct vm_area_struct *vma,
-		unsigned long address, unsigned int foll_flags)
-{
-	unsigned int unused_page_mask;
-	return follow_page_mask(vma, address, foll_flags, &unused_page_mask);
-}
+struct page *follow_page(struct vm_area_struct *vma, unsigned long address,
+			 unsigned int foll_flags);
 
 #define FOLL_WRITE	0x01	/* check pte is writable */
 #define FOLL_TOUCH	0x02	/* mark page accessed */

mm/gup.c

@@ -20,6 +20,11 @@
 
 #include "internal.h"
 
+struct follow_page_context {
+	struct dev_pagemap *pgmap;
+	unsigned int page_mask;
+};
+
 static struct page *no_page_table(struct vm_area_struct *vma,
 		unsigned int flags)
 {
@@ -71,10 +76,10 @@ static inline bool can_follow_write_pte(pte_t pte, unsigned int flags)
 }
 
 static struct page *follow_page_pte(struct vm_area_struct *vma,
-		unsigned long address, pmd_t *pmd, unsigned int flags)
+		unsigned long address, pmd_t *pmd, unsigned int flags,
+		struct dev_pagemap **pgmap)
 {
 	struct mm_struct *mm = vma->vm_mm;
-	struct dev_pagemap *pgmap = NULL;
 	struct page *page;
 	spinlock_t *ptl;
 	pte_t *ptep, pte;
@@ -116,8 +121,8 @@ retry:
 		 * Only return device mapping pages in the FOLL_GET case since
 		 * they are only valid while holding the pgmap reference.
 		 */
-		pgmap = get_dev_pagemap(pte_pfn(pte), NULL);
-		if (pgmap)
+		*pgmap = get_dev_pagemap(pte_pfn(pte), *pgmap);
+		if (*pgmap)
 			page = pte_page(pte);
 		else
 			goto no_page;
@@ -152,15 +157,8 @@ retry:
 		goto retry;
 	}
 
-	if (flags & FOLL_GET) {
+	if (flags & FOLL_GET)
 		get_page(page);
-
-		/* drop the pgmap reference now that we hold the page */
-		if (pgmap) {
-			put_dev_pagemap(pgmap);
-			pgmap = NULL;
-		}
-	}
 	if (flags & FOLL_TOUCH) {
 		if ((flags & FOLL_WRITE) &&
 		    !pte_dirty(pte) && !PageDirty(page))
@@ -210,7 +208,8 @@ no_page:
 
 static struct page *follow_pmd_mask(struct vm_area_struct *vma,
 				    unsigned long address, pud_t *pudp,
-				    unsigned int flags, unsigned int *page_mask)
+				    unsigned int flags,
+				    struct follow_page_context *ctx)
 {
 	pmd_t *pmd, pmdval;
 	spinlock_t *ptl;
@@ -258,13 +257,13 @@ retry:
 	}
 	if (pmd_devmap(pmdval)) {
 		ptl = pmd_lock(mm, pmd);
-		page = follow_devmap_pmd(vma, address, pmd, flags);
+		page = follow_devmap_pmd(vma, address, pmd, flags, &ctx->pgmap);
 		spin_unlock(ptl);
 		if (page)
 			return page;
 	}
 	if (likely(!pmd_trans_huge(pmdval)))
-		return follow_page_pte(vma, address, pmd, flags);
+		return follow_page_pte(vma, address, pmd, flags, &ctx->pgmap);
 
 	if ((flags & FOLL_NUMA) && pmd_protnone(pmdval))
 		return no_page_table(vma, flags);
@@ -284,7 +283,7 @@ retry_locked:
 	}
 	if (unlikely(!pmd_trans_huge(*pmd))) {
 		spin_unlock(ptl);
-		return follow_page_pte(vma, address, pmd, flags);
+		return follow_page_pte(vma, address, pmd, flags, &ctx->pgmap);
 	}
 	if (flags & FOLL_SPLIT) {
 		int ret;
@@ -307,18 +306,18 @@ retry_locked:
 		}
 
 		return ret ? ERR_PTR(ret) :
-			follow_page_pte(vma, address, pmd, flags);
+			follow_page_pte(vma, address, pmd, flags, &ctx->pgmap);
 	}
 	page = follow_trans_huge_pmd(vma, address, pmd, flags);
 	spin_unlock(ptl);
-	*page_mask = HPAGE_PMD_NR - 1;
+	ctx->page_mask = HPAGE_PMD_NR - 1;
 	return page;
 }
 
-
 static struct page *follow_pud_mask(struct vm_area_struct *vma,
 				    unsigned long address, p4d_t *p4dp,
-				    unsigned int flags, unsigned int *page_mask)
+				    unsigned int flags,
+				    struct follow_page_context *ctx)
 {
 	pud_t *pud;
 	spinlock_t *ptl;
@@ -344,7 +343,7 @@ static struct page *follow_pud_mask(struct vm_area_struct *vma,
 	}
 	if (pud_devmap(*pud)) {
 		ptl = pud_lock(mm, pud);
-		page = follow_devmap_pud(vma, address, pud, flags);
+		page = follow_devmap_pud(vma, address, pud, flags, &ctx->pgmap);
 		spin_unlock(ptl);
 		if (page)
 			return page;
@@ -352,13 +351,13 @@ static struct page *follow_pud_mask(struct vm_area_struct *vma,
 	if (unlikely(pud_bad(*pud)))
 		return no_page_table(vma, flags);
 
-	return follow_pmd_mask(vma, address, pud, flags, page_mask);
+	return follow_pmd_mask(vma, address, pud, flags, ctx);
 }
 
-
 static struct page *follow_p4d_mask(struct vm_area_struct *vma,
 				    unsigned long address, pgd_t *pgdp,
-				    unsigned int flags, unsigned int *page_mask)
+				    unsigned int flags,
+				    struct follow_page_context *ctx)
 {
 	p4d_t *p4d;
 	struct page *page;
@@ -378,7 +377,7 @@ static struct page *follow_p4d_mask(struct vm_area_struct *vma,
 			return page;
 		return no_page_table(vma, flags);
 	}
-	return follow_pud_mask(vma, address, p4d, flags, page_mask);
+	return follow_pud_mask(vma, address, p4d, flags, ctx);
 }
 
 /**
@@ -396,13 +395,13 @@ static struct page *follow_p4d_mask(struct vm_area_struct *vma,
  */
 struct page *follow_page_mask(struct vm_area_struct *vma,
 			      unsigned long address, unsigned int flags,
-			      unsigned int *page_mask)
+			      struct follow_page_context *ctx)
 {
 	pgd_t *pgd;
 	struct page *page;
 	struct mm_struct *mm = vma->vm_mm;
 
-	*page_mask = 0;
+	ctx->page_mask = 0;
 
 	/* make this handle hugepd */
 	page = follow_huge_addr(mm, address, flags & FOLL_WRITE);
@@ -431,7 +430,19 @@ struct page *follow_page_mask(struct vm_area_struct *vma,
 		return no_page_table(vma, flags);
 	}
 
-	return follow_p4d_mask(vma, address, pgd, flags, page_mask);
+	return follow_p4d_mask(vma, address, pgd, flags, ctx);
+}
+
+struct page *follow_page(struct vm_area_struct *vma, unsigned long address,
+			 unsigned int foll_flags)
+{
+	struct follow_page_context ctx = { NULL };
+	struct page *page;
+
+	page = follow_page_mask(vma, address, foll_flags, &ctx);
+	if (ctx.pgmap)
+		put_dev_pagemap(ctx.pgmap);
+	return page;
 }
 
 static int get_gate_page(struct mm_struct *mm, unsigned long address,
@@ -659,9 +670,9 @@ static long __get_user_pages(struct task_struct *tsk, struct mm_struct *mm,
 		unsigned int gup_flags, struct page **pages,
 		struct vm_area_struct **vmas, int *nonblocking)
 {
-	long i = 0;
-	unsigned int page_mask;
+	long ret = 0, i = 0;
 	struct vm_area_struct *vma = NULL;
+	struct follow_page_context ctx = { NULL };
 
 	if (!nr_pages)
 		return 0;
@@ -691,12 +702,14 @@ static long __get_user_pages(struct task_struct *tsk, struct mm_struct *mm,
 						pages ? &pages[i] : NULL);
 				if (ret)
 					return i ? : ret;
-				page_mask = 0;
+				ctx.page_mask = 0;
 				goto next_page;
 			}
 
-			if (!vma || check_vma_flags(vma, gup_flags))
-				return i ? : -EFAULT;
+			if (!vma || check_vma_flags(vma, gup_flags)) {
+				ret = -EFAULT;
+				goto out;
+			}
 			if (is_vm_hugetlb_page(vma)) {
 				i = follow_hugetlb_page(mm, vma, pages, vmas,
 						&start, &nr_pages, i,
@@ -709,23 +722,26 @@ retry:
 		 * If we have a pending SIGKILL, don't keep faulting pages and
 		 * potentially allocating memory.
 		 */
-		if (unlikely(fatal_signal_pending(current)))
-			return i ? i : -ERESTARTSYS;
+		if (unlikely(fatal_signal_pending(current))) {
+			ret = -ERESTARTSYS;
+			goto out;
+		}
 		cond_resched();
-		page = follow_page_mask(vma, start, foll_flags, &page_mask);
+
+		page = follow_page_mask(vma, start, foll_flags, &ctx);
 		if (!page) {
-			int ret;
 			ret = faultin_page(tsk, vma, start, &foll_flags,
 					nonblocking);
 			switch (ret) {
 			case 0:
 				goto retry;
+			case -EBUSY:
+				ret = 0;
+				/* FALLTHRU */
 			case -EFAULT:
 			case -ENOMEM:
 			case -EHWPOISON:
-				return i ? i : ret;
-			case -EBUSY:
-				return i;
+				goto out;
 			case -ENOENT:
 				goto next_page;
 			}
@@ -737,27 +753,31 @@ retry:
 			 */
 			goto next_page;
 		} else if (IS_ERR(page)) {
-			return i ? i : PTR_ERR(page);
+			ret = PTR_ERR(page);
+			goto out;
 		}
 		if (pages) {
 			pages[i] = page;
 			flush_anon_page(vma, page, start);
 			flush_dcache_page(page);
-			page_mask = 0;
+			ctx.page_mask = 0;
 		}
 next_page:
 		if (vmas) {
 			vmas[i] = vma;
-			page_mask = 0;
+			ctx.page_mask = 0;
 		}
-		page_increm = 1 + (~(start >> PAGE_SHIFT) & page_mask);
+		page_increm = 1 + (~(start >> PAGE_SHIFT) & ctx.page_mask);
 		if (page_increm > nr_pages)
 			page_increm = nr_pages;
 		i += page_increm;
 		start += page_increm * PAGE_SIZE;
 		nr_pages -= page_increm;
 	} while (nr_pages);
-	return i;
+out:
+	if (ctx.pgmap)
+		put_dev_pagemap(ctx.pgmap);
+	return i ? i : ret;
 }
 
 static bool vma_permits_fault(struct vm_area_struct *vma,

mm/huge_memory.c

@@ -852,11 +852,10 @@ static void touch_pmd(struct vm_area_struct *vma, unsigned long addr,
 }
 
 struct page *follow_devmap_pmd(struct vm_area_struct *vma, unsigned long addr,
-		pmd_t *pmd, int flags)
+		pmd_t *pmd, int flags, struct dev_pagemap **pgmap)
 {
 	unsigned long pfn = pmd_pfn(*pmd);
 	struct mm_struct *mm = vma->vm_mm;
-	struct dev_pagemap *pgmap;
 	struct page *page;
 
 	assert_spin_locked(pmd_lockptr(mm, pmd));
@@ -886,12 +885,11 @@ struct page *follow_devmap_pmd(struct vm_area_struct *vma, unsigned long addr,
 		return ERR_PTR(-EEXIST);
 
 	pfn += (addr & ~PMD_MASK) >> PAGE_SHIFT;
-	pgmap = get_dev_pagemap(pfn, NULL);
-	if (!pgmap)
+	*pgmap = get_dev_pagemap(pfn, *pgmap);
+	if (!*pgmap)
 		return ERR_PTR(-EFAULT);
 	page = pfn_to_page(pfn);
 	get_page(page);
-	put_dev_pagemap(pgmap);
 
 	return page;
 }
@@ -1000,11 +998,10 @@ static void touch_pud(struct vm_area_struct *vma, unsigned long addr,
 }
 
 struct page *follow_devmap_pud(struct vm_area_struct *vma, unsigned long addr,
-		pud_t *pud, int flags)
+		pud_t *pud, int flags, struct dev_pagemap **pgmap)
 {
 	unsigned long pfn = pud_pfn(*pud);
 	struct mm_struct *mm = vma->vm_mm;
-	struct dev_pagemap *pgmap;
 	struct page *page;
 
 	assert_spin_locked(pud_lockptr(mm, pud));
@@ -1028,12 +1025,11 @@ struct page *follow_devmap_pud(struct vm_area_struct *vma, unsigned long addr,
 		return ERR_PTR(-EEXIST);
 
 	pfn += (addr & ~PUD_MASK) >> PAGE_SHIFT;
-	pgmap = get_dev_pagemap(pfn, NULL);
-	if (!pgmap)
+	*pgmap = get_dev_pagemap(pfn, *pgmap);
+	if (!*pgmap)
 		return ERR_PTR(-EFAULT);
 	page = pfn_to_page(pfn);
 	get_page(page);
-	put_dev_pagemap(pgmap);
 
 	return page;
 }

mm/nommu.c

@@ -1709,11 +1709,9 @@ SYSCALL_DEFINE5(mremap, unsigned long, addr, unsigned long, old_len,
 	return ret;
 }
 
-struct page *follow_page_mask(struct vm_area_struct *vma,
-			      unsigned long address, unsigned int flags,
-			      unsigned int *page_mask)
+struct page *follow_page(struct vm_area_struct *vma, unsigned long address,
+			 unsigned int foll_flags)
 {
-	*page_mask = 0;
 	return NULL;
 }