ARC: mm: switch pgtable_t back to struct page *

So far ARC pgtable_t has not been struct page based to avoid extra
page_address() calls involved. However the differences are down to
noise and get in the way of using generic code, hence this patch.

This also allows us to reuse generic THP depost/withdraw code.

There's some additional consideration for PGDIR_SHIFT in 4K page config.
Now due to page tables being PAGE_SIZE deep only, the address split
can't be really arbitrary.

Tested-by: kernel test robot <lkp@intel.com>
Suggested-by: Mike Rapoport <rppt@linux.ibm.com>
Acked-by: Mike Rapoport <rppt@linux.ibm.com>
Signed-off-by: Vineet Gupta <vgupta@kernel.org>

arch/arc/include/asm/hugepage.h

@@ -58,14 +58,6 @@ static inline void set_pmd_at(struct mm_struct *mm, unsigned long addr,
 extern void update_mmu_cache_pmd(struct vm_area_struct *vma, unsigned long addr,
 				 pmd_t *pmd);
 
-/* Generic variants assume pgtable_t is struct page *, hence need for these */
-#define __HAVE_ARCH_PGTABLE_DEPOSIT
-extern void pgtable_trans_huge_deposit(struct mm_struct *mm, pmd_t *pmdp,
-				       pgtable_t pgtable);
-
-#define __HAVE_ARCH_PGTABLE_WITHDRAW
-extern pgtable_t pgtable_trans_huge_withdraw(struct mm_struct *mm, pmd_t *pmdp);
-
 #define __HAVE_ARCH_FLUSH_PMD_TLB_RANGE
 extern void flush_pmd_tlb_range(struct vm_area_struct *vma, unsigned long start,
 				unsigned long end);

arch/arc/include/asm/page.h

@@ -60,7 +60,7 @@ typedef struct {
 #define __pgprot(x)	((pgprot_t) { (x) })
 #define pte_pgprot(x)	__pgprot(pte_val(x))
 
-typedef pte_t * pgtable_t;
+typedef struct page *pgtable_t;
 
 /*
  * Use virt_to_pfn with caution:

arch/arc/include/asm/pgalloc.h

@@ -45,22 +45,17 @@ pmd_populate_kernel(struct mm_struct *mm, pmd_t *pmd, pte_t *pte)
 	set_pmd(pmd, __pmd((unsigned long)pte));
 }
 
-static inline void pmd_populate(struct mm_struct *mm, pmd_t *pmd, pgtable_t pte)
+static inline void pmd_populate(struct mm_struct *mm, pmd_t *pmd, pgtable_t pte_page)
 {
-	set_pmd(pmd, __pmd((unsigned long)pte));
+	set_pmd(pmd, __pmd((unsigned long)page_address(pte_page)));
-}
-
-static inline int __get_order_pgd(void)
-{
-	return get_order(PTRS_PER_PGD * sizeof(pgd_t));
 }
 
 static inline pgd_t *pgd_alloc(struct mm_struct *mm)
 {
-	int num, num2;
+	pgd_t *ret = (pgd_t *) __get_free_page(GFP_KERNEL);
-	pgd_t *ret = (pgd_t *) __get_free_pages(GFP_KERNEL, __get_order_pgd());
 
 	if (ret) {
+		int num, num2;
 		num = USER_PTRS_PER_PGD + USER_KERNEL_GUTTER / PGDIR_SIZE;
 		memzero(ret, num * sizeof(pgd_t));
 
@@ -76,61 +71,43 @@ static inline pgd_t *pgd_alloc(struct mm_struct *mm)
 
 static inline void pgd_free(struct mm_struct *mm, pgd_t *pgd)
 {
-	free_pages((unsigned long)pgd, __get_order_pgd());
+	free_page((unsigned long)pgd);
-}
-
-
-/*
- * With software-only page-tables, addr-split for traversal is tweakable and
- * that directly governs how big tables would be at each level.
- * Further, the MMU page size is configurable.
- * Thus we need to programatically assert the size constraint
- * All of this is const math, allowing gcc to do constant folding/propagation.
- */
-
-static inline int __get_order_pte(void)
-{
-	return get_order(PTRS_PER_PTE * sizeof(pte_t));
 }
 
 static inline pte_t *pte_alloc_one_kernel(struct mm_struct *mm)
 {
 	pte_t *pte;
 
-	pte = (pte_t *) __get_free_pages(GFP_KERNEL | __GFP_ZERO,
+	pte = (pte_t *) __get_free_page(GFP_KERNEL | __GFP_ZERO);
-					 __get_order_pte());
 
 	return pte;
 }
 
-static inline pgtable_t
+static inline pgtable_t pte_alloc_one(struct mm_struct *mm)
-pte_alloc_one(struct mm_struct *mm)
 {
-	pgtable_t pte_pg;
 	struct page *page;
 
-	pte_pg = (pgtable_t)__get_free_pages(GFP_KERNEL, __get_order_pte());
+	page = (pgtable_t)alloc_page(GFP_KERNEL | __GFP_ZERO | __GFP_ACCOUNT);
-	if (!pte_pg)
+	if (!page)
-		return 0;
+		return NULL;
-	memzero((void *)pte_pg, PTRS_PER_PTE * sizeof(pte_t));
+
-	page = virt_to_page(pte_pg);
 	if (!pgtable_pte_page_ctor(page)) {
 		__free_page(page);
-		return 0;
+		return NULL;
 	}
 
-	return pte_pg;
+	return page;
 }
 
 static inline void pte_free_kernel(struct mm_struct *mm, pte_t *pte)
 {
-	free_pages((unsigned long)pte, __get_order_pte()); /* takes phy addr */
+	free_page((unsigned long)pte);
 }
 
-static inline void pte_free(struct mm_struct *mm, pgtable_t ptep)
+static inline void pte_free(struct mm_struct *mm, pgtable_t pte_page)
 {
-	pgtable_pte_page_dtor(virt_to_page(ptep));
+	pgtable_pte_page_dtor(pte_page);
-	free_pages((unsigned long)ptep, __get_order_pte());
+	__free_page(pte_page);
 }
 
 #define __pte_free_tlb(tlb, pte, addr)  pte_free((tlb)->mm, pte)

arch/arc/include/asm/pgtable-levels.h

@@ -35,9 +35,15 @@
 #else
 /*
  * No Super page case
- * Default value provides 11:8:13 (8K), 11:9:12 (4K)
+ * Default value provides 11:8:13 (8K), 10:10:12 (4K)
+ * Limits imposed by pgtable_t only PAGE_SIZE long
+ * (so 4K page can only have 1K entries: or 10 bits)
  */
+#ifdef CONFIG_ARC_PAGE_SIZE_4K
+#define PGDIR_SHIFT		22
+#else
 #define PGDIR_SHIFT		21
+#endif
 
 #endif
 

arch/arc/mm/init.c

@@ -189,6 +189,9 @@ void __init mem_init(void)
 {
 	memblock_free_all();
 	highmem_init();
+
+	BUILD_BUG_ON((PTRS_PER_PGD * sizeof(pgd_t)) > PAGE_SIZE);
+	BUILD_BUG_ON((PTRS_PER_PTE * sizeof(pte_t)) > PAGE_SIZE);
 }
 
 #ifdef CONFIG_HIGHMEM

arch/arc/mm/tlb.c

@@ -534,43 +534,6 @@ void update_mmu_cache_pmd(struct vm_area_struct *vma, unsigned long addr,
 	update_mmu_cache(vma, addr, &pte);
 }
 
-void pgtable_trans_huge_deposit(struct mm_struct *mm, pmd_t *pmdp,
-				pgtable_t pgtable)
-{
-	struct list_head *lh = (struct list_head *) pgtable;
-
-	assert_spin_locked(&mm->page_table_lock);
-
-	/* FIFO */
-	if (!pmd_huge_pte(mm, pmdp))
-		INIT_LIST_HEAD(lh);
-	else
-		list_add(lh, (struct list_head *) pmd_huge_pte(mm, pmdp));
-	pmd_huge_pte(mm, pmdp) = pgtable;
-}
-
-pgtable_t pgtable_trans_huge_withdraw(struct mm_struct *mm, pmd_t *pmdp)
-{
-	struct list_head *lh;
-	pgtable_t pgtable;
-
-	assert_spin_locked(&mm->page_table_lock);
-
-	pgtable = pmd_huge_pte(mm, pmdp);
-	lh = (struct list_head *) pgtable;
-	if (list_empty(lh))
-		pmd_huge_pte(mm, pmdp) = NULL;
-	else {
-		pmd_huge_pte(mm, pmdp) = (pgtable_t) lh->next;
-		list_del(lh);
-	}
-
-	pte_val(pgtable[0]) = 0;
-	pte_val(pgtable[1]) = 0;
-
-	return pgtable;
-}
-
 void local_flush_pmd_tlb_range(struct vm_area_struct *vma, unsigned long start,
 			       unsigned long end)
 {