s390/mm: use memset64 instead of clear_table

Use memset64 instead of the (now) open-coded variant clear_table.
Performance wise there is no difference.

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>

arch/s390/include/asm/pgalloc.h

@@ -12,6 +12,7 @@
 #define _S390_PGALLOC_H
 
 #include <linux/threads.h>
+#include <linux/string.h>
 #include <linux/gfp.h>
 #include <linux/mm.h>
 
@@ -27,24 +28,9 @@ void page_table_free_rcu(struct mmu_gather *, unsigned long *, unsigned long);
 void page_table_free_pgste(struct page *page);
 extern int page_table_allocate_pgste;
 
-static inline void clear_table(unsigned long *s, unsigned long val, size_t n)
-{
-	struct addrtype { char _[256]; };
-	int i;
-
-	for (i = 0; i < n; i += 256) {
-		*s = val;
-		asm volatile(
-			"mvc	8(248,%[s]),0(%[s])\n"
-			: "+m" (*(struct addrtype *) s)
-			: [s] "a" (s));
-		s += 256 / sizeof(long);
-	}
-}
-
 static inline void crst_table_init(unsigned long *crst, unsigned long entry)
 {
-	clear_table(crst, entry, _CRST_TABLE_SIZE);
+	memset64((u64 *)crst, entry, _CRST_ENTRIES);
 }
 
 static inline unsigned long pgd_entry_type(struct mm_struct *mm)

arch/s390/kernel/vdso.c

@@ -166,10 +166,8 @@ int vdso_alloc_per_cpu(struct lowcore *lowcore)
 	vd->node_id = cpu_to_node(vd->cpu_nr);
 
 	/* Set up access register mode page table */
-	clear_table((unsigned long *) segment_table, _SEGMENT_ENTRY_EMPTY,
-		    PAGE_SIZE << SEGMENT_ORDER);
-	clear_table((unsigned long *) page_table, _PAGE_INVALID,
-		    256*sizeof(unsigned long));
+	memset64((u64 *)segment_table, _SEGMENT_ENTRY_EMPTY, _CRST_ENTRIES);
+	memset64((u64 *)page_table, _PAGE_INVALID, PTRS_PER_PTE);
 
 	*(unsigned long *) segment_table = _SEGMENT_ENTRY + page_table;
 	*(unsigned long *) page_table = _PAGE_PROTECT + page_frame;

arch/s390/mm/pgalloc.c

@@ -158,13 +158,13 @@ static inline unsigned int atomic_xor_bits(atomic_t *v, unsigned int bits)
 struct page *page_table_alloc_pgste(struct mm_struct *mm)
 {
 	struct page *page;
-	unsigned long *table;
+	u64 *table;
 
 	page = alloc_page(GFP_KERNEL);
 	if (page) {
-		table = (unsigned long *) page_to_phys(page);
-		clear_table(table, _PAGE_INVALID, PAGE_SIZE/2);
-		clear_table(table + PTRS_PER_PTE, 0, PAGE_SIZE/2);
+		table = (u64 *)page_to_phys(page);
+		memset64(table, _PAGE_INVALID, PTRS_PER_PTE);
+		memset64(table + PTRS_PER_PTE, 0, PTRS_PER_PTE);
 	}
 	return page;
 }
@@ -221,12 +221,12 @@ unsigned long *page_table_alloc(struct mm_struct *mm)
 	if (mm_alloc_pgste(mm)) {
 		/* Return 4K page table with PGSTEs */
 		atomic_set(&page->_mapcount, 3);
-		clear_table(table, _PAGE_INVALID, PAGE_SIZE/2);
-		clear_table(table + PTRS_PER_PTE, 0, PAGE_SIZE/2);
+		memset64((u64 *)table, _PAGE_INVALID, PTRS_PER_PTE);
+		memset64((u64 *)table + PTRS_PER_PTE, 0, PTRS_PER_PTE);
 	} else {
 		/* Return the first 2K fragment of the page */
 		atomic_set(&page->_mapcount, 1);
-		clear_table(table, _PAGE_INVALID, PAGE_SIZE);
+		memset64((u64 *)table, _PAGE_INVALID, 2 * PTRS_PER_PTE);
 		spin_lock_bh(&mm->context.lock);
 		list_add(&page->lru, &mm->context.pgtable_list);
 		spin_unlock_bh(&mm->context.lock);

arch/s390/mm/vmem.c

@@ -59,7 +59,7 @@ pte_t __ref *vmem_pte_alloc(void)
 		pte = (pte_t *) memblock_alloc(size, size);
 	if (!pte)
 		return NULL;
-	clear_table((unsigned long *) pte, _PAGE_INVALID, size);
+	memset64((u64 *)pte, _PAGE_INVALID, PTRS_PER_PTE);
 	return pte;
 }