mm: cleanup *pte_alloc* interfaces

There are few things about *pte_alloc*() helpers worth cleaning up:

 - 'vma' argument is unused, let's drop it;

 - most __pte_alloc() callers do speculative check for pmd_none(),
   before taking ptl: let's introduce pte_alloc() macro which does
   the check.

   The only direct user of __pte_alloc left is userfaultfd, which has
   different expectation about atomicity wrt pmd.

 - pte_alloc_map() and pte_alloc_map_lock() are redefined using
   pte_alloc().

[sudeep.holla@arm.com: fix build for arm64 hugetlbpage]
[sfr@canb.auug.org.au: fix arch/arm/mm/mmu.c some more]
Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Dave Hansen <dave.hansen@intel.com>
Signed-off-by: Sudeep Holla <sudeep.holla@arm.com>
Acked-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

arch/arm/mm/mmu.c

@@ -732,7 +732,7 @@ static void *__init late_alloc(unsigned long sz)
 	return ptr;
 }
 
-static pte_t * __init pte_alloc(pmd_t *pmd, unsigned long addr,
+static pte_t * __init arm_pte_alloc(pmd_t *pmd, unsigned long addr,
 				unsigned long prot,
 				void *(*alloc)(unsigned long sz))
 {
@@ -747,7 +747,7 @@ static pte_t * __init pte_alloc(pmd_t *pmd, unsigned long addr,
 static pte_t * __init early_pte_alloc(pmd_t *pmd, unsigned long addr,
 				      unsigned long prot)
 {
-	return pte_alloc(pmd, addr, prot, early_alloc);
+	return arm_pte_alloc(pmd, addr, prot, early_alloc);
 }
 
 static void __init alloc_init_pte(pmd_t *pmd, unsigned long addr,
@@ -756,7 +756,7 @@ static void __init alloc_init_pte(pmd_t *pmd, unsigned long addr,
 				  void *(*alloc)(unsigned long sz),
 				  bool ng)
 {
-	pte_t *pte = pte_alloc(pmd, addr, type->prot_l1, alloc);
+	pte_t *pte = arm_pte_alloc(pmd, addr, type->prot_l1, alloc);
 	do {
 		set_pte_ext(pte, pfn_pte(pfn, __pgprot(type->prot_pte)),
 			    ng ? PTE_EXT_NG : 0);

arch/arm/mm/pgd.c

@@ -80,7 +80,7 @@ pgd_t *pgd_alloc(struct mm_struct *mm)
 		if (!new_pmd)
 			goto no_pmd;
 
-		new_pte = pte_alloc_map(mm, NULL, new_pmd, 0);
+		new_pte = pte_alloc_map(mm, new_pmd, 0);
 		if (!new_pte)
 			goto no_pte;
 

arch/arm64/mm/hugetlbpage.c

@@ -124,7 +124,7 @@ pte_t *huge_pte_alloc(struct mm_struct *mm,
 		 * will be no pte_unmap() to correspond with this
 		 * pte_alloc_map().
 		 */
-		pte = pte_alloc_map(mm, NULL, pmd, addr);
+		pte = pte_alloc_map(mm, pmd, addr);
 	} else if (sz == PMD_SIZE) {
 		if (IS_ENABLED(CONFIG_ARCH_WANT_HUGE_PMD_SHARE) &&
 		    pud_none(*pud))

arch/ia64/mm/hugetlbpage.c

@@ -38,7 +38,7 @@ huge_pte_alloc(struct mm_struct *mm, unsigned long addr, unsigned long sz)
 	if (pud) {
 		pmd = pmd_alloc(mm, pud, taddr);
 		if (pmd)
-			pte = pte_alloc_map(mm, NULL, pmd, taddr);
+			pte = pte_alloc_map(mm, pmd, taddr);
 	}
 	return pte;
 }

arch/metag/mm/hugetlbpage.c

@@ -67,7 +67,7 @@ pte_t *huge_pte_alloc(struct mm_struct *mm,
 	pgd = pgd_offset(mm, addr);
 	pud = pud_offset(pgd, addr);
 	pmd = pmd_offset(pud, addr);
-	pte = pte_alloc_map(mm, NULL, pmd, addr);
+	pte = pte_alloc_map(mm, pmd, addr);
 	pgd->pgd &= ~_PAGE_SZ_MASK;
 	pgd->pgd |= _PAGE_SZHUGE;
 

arch/parisc/mm/hugetlbpage.c

@@ -63,7 +63,7 @@ pte_t *huge_pte_alloc(struct mm_struct *mm,
 	if (pud) {
 		pmd = pmd_alloc(mm, pud, addr);
 		if (pmd)
-			pte = pte_alloc_map(mm, NULL, pmd, addr);
+			pte = pte_alloc_map(mm, pmd, addr);
 	}
 	return pte;
 }

arch/sh/mm/hugetlbpage.c

@@ -35,7 +35,7 @@ pte_t *huge_pte_alloc(struct mm_struct *mm,
 		if (pud) {
 			pmd = pmd_alloc(mm, pud, addr);
 			if (pmd)
-				pte = pte_alloc_map(mm, NULL, pmd, addr);
+				pte = pte_alloc_map(mm, pmd, addr);
 		}
 	}
 

arch/sparc/mm/hugetlbpage.c

@@ -146,7 +146,7 @@ pte_t *huge_pte_alloc(struct mm_struct *mm,
 	if (pud) {
 		pmd = pmd_alloc(mm, pud, addr);
 		if (pmd)
-			pte = pte_alloc_map(mm, NULL, pmd, addr);
+			pte = pte_alloc_map(mm, pmd, addr);
 	}
 	return pte;
 }

arch/tile/mm/hugetlbpage.c

@@ -77,7 +77,7 @@ pte_t *huge_pte_alloc(struct mm_struct *mm,
 		else {
 			if (sz != PAGE_SIZE << huge_shift[HUGE_SHIFT_PAGE])
 				panic("Unexpected page size %#lx\n", sz);
-			return pte_alloc_map(mm, NULL, pmd, addr);
+			return pte_alloc_map(mm, pmd, addr);
 		}
 	}
 #else

arch/um/kernel/skas/mmu.c

@@ -31,7 +31,7 @@ static int init_stub_pte(struct mm_struct *mm, unsigned long proc,
 	if (!pmd)
 		goto out_pmd;
 
-	pte = pte_alloc_map(mm, NULL, pmd, proc);
+	pte = pte_alloc_map(mm, pmd, proc);
 	if (!pte)
 		goto out_pte;
 

arch/unicore32/mm/pgd.c

@@ -54,7 +54,7 @@ pgd_t *get_pgd_slow(struct mm_struct *mm)
 		if (!new_pmd)
 			goto no_pmd;
 
-		new_pte = pte_alloc_map(mm, NULL, new_pmd, 0);
+		new_pte = pte_alloc_map(mm, new_pmd, 0);
 		if (!new_pte)
 			goto no_pte;
 

arch/x86/kernel/tboot.c

@@ -135,7 +135,7 @@ static int map_tboot_page(unsigned long vaddr, unsigned long pfn,
 	pmd = pmd_alloc(&tboot_mm, pud, vaddr);
 	if (!pmd)
 		return -1;
-	pte = pte_alloc_map(&tboot_mm, NULL, pmd, vaddr);
+	pte = pte_alloc_map(&tboot_mm, pmd, vaddr);
 	if (!pte)
 		return -1;
 	set_pte_at(&tboot_mm, vaddr, pte, pfn_pte(pfn, prot));

include/linux/mm.h

@@ -1545,8 +1545,7 @@ static inline void mm_dec_nr_pmds(struct mm_struct *mm)
 }
 #endif
 
-int __pte_alloc(struct mm_struct *mm, struct vm_area_struct *vma,
+int __pte_alloc(struct mm_struct *mm, pmd_t *pmd, unsigned long address);
-		pmd_t *pmd, unsigned long address);
 int __pte_alloc_kernel(pmd_t *pmd, unsigned long address);
 
 /*
@@ -1672,15 +1671,15 @@ static inline void pgtable_page_dtor(struct page *page)
 	pte_unmap(pte);					\
 } while (0)
 
-#define pte_alloc_map(mm, vma, pmd, address)				\
+#define pte_alloc(mm, pmd, address)			\
-	((unlikely(pmd_none(*(pmd))) && __pte_alloc(mm, vma,	\
+	(unlikely(pmd_none(*(pmd))) && __pte_alloc(mm, pmd, address))
-							pmd, address))?	\
+
-	 NULL: pte_offset_map(pmd, address))
+#define pte_alloc_map(mm, pmd, address)			\
+	(pte_alloc(mm, pmd, address) ? NULL : pte_offset_map(pmd, address))
 
 #define pte_alloc_map_lock(mm, pmd, address, ptlp)	\
-	((unlikely(pmd_none(*(pmd))) && __pte_alloc(mm, NULL,	\
+	(pte_alloc(mm, pmd, address) ?			\
-							pmd, address))?	\
+		 NULL : pte_offset_map_lock(mm, pmd, address, ptlp))
-		NULL: pte_offset_map_lock(mm, pmd, address, ptlp))
 
 #define pte_alloc_kernel(pmd, address)			\
 	((unlikely(pmd_none(*(pmd))) && __pte_alloc_kernel(pmd, address))? \

mm/memory.c

@@ -562,8 +562,7 @@ void free_pgtables(struct mmu_gather *tlb, struct vm_area_struct *vma,
 	}
 }
 
-int __pte_alloc(struct mm_struct *mm, struct vm_area_struct *vma,
+int __pte_alloc(struct mm_struct *mm, pmd_t *pmd, unsigned long address)
-		pmd_t *pmd, unsigned long address)
 {
 	spinlock_t *ptl;
 	pgtable_t new = pte_alloc_one(mm, address);
@@ -3419,12 +3418,11 @@ static int __handle_mm_fault(struct mm_struct *mm, struct vm_area_struct *vma,
 	}
 
 	/*
-	 * Use __pte_alloc instead of pte_alloc_map, because we can't
+	 * Use pte_alloc() instead of pte_alloc_map, because we can't
 	 * run pte_offset_map on the pmd, if an huge pmd could
 	 * materialize from under us from a different thread.
 	 */
-	if (unlikely(pmd_none(*pmd)) &&
+	if (unlikely(pte_alloc(mm, pmd, address)))
-	    unlikely(__pte_alloc(mm, vma, pmd, address)))
 		return VM_FAULT_OOM;
 	/*
 	 * If a huge pmd materialized under us just retry later.  Use

mm/mremap.c

@@ -213,8 +213,7 @@ unsigned long move_page_tables(struct vm_area_struct *vma,
 				continue;
 			VM_BUG_ON(pmd_trans_huge(*old_pmd));
 		}
-		if (pmd_none(*new_pmd) && __pte_alloc(new_vma->vm_mm, new_vma,
+		if (pte_alloc(new_vma->vm_mm, new_pmd, new_addr))
-						      new_pmd, new_addr))
 			break;
 		next = (new_addr + PMD_SIZE) & PMD_MASK;
 		if (extent > next - new_addr)

mm/userfaultfd.c

@@ -230,8 +230,7 @@ retry:
 			break;
 		}
 		if (unlikely(pmd_none(dst_pmdval)) &&
-		    unlikely(__pte_alloc(dst_mm, dst_vma, dst_pmd,
+		    unlikely(__pte_alloc(dst_mm, dst_pmd, dst_addr))) {
-					 dst_addr))) {
 			err = -ENOMEM;
 			break;
 		}