s390: implement memset16, memset32 & memset64

Provide fast versions of the new memset variants. E.g. the generic
memset64 is ten times slower than the optimized version if used on a
whole page.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>

arch/s390/include/asm/string.h

@@ -17,6 +17,9 @@
 #define __HAVE_ARCH_MEMMOVE	/* gcc builtin & arch function */
 #define __HAVE_ARCH_MEMSCAN	/* inline & arch function */
 #define __HAVE_ARCH_MEMSET	/* gcc builtin & arch function */
+#define __HAVE_ARCH_MEMSET16	/* arch function */
+#define __HAVE_ARCH_MEMSET32	/* arch function */
+#define __HAVE_ARCH_MEMSET64	/* arch function */
 #define __HAVE_ARCH_STRCAT	/* inline & arch function */
 #define __HAVE_ARCH_STRCMP	/* arch function */
 #define __HAVE_ARCH_STRCPY	/* inline & arch function */
@@ -49,6 +52,25 @@ extern char *strstr(const char *, const char *);
 #undef __HAVE_ARCH_STRSEP
 #undef __HAVE_ARCH_STRSPN
 
+void *__memset16(uint16_t *s, uint16_t v, size_t count);
+void *__memset32(uint32_t *s, uint32_t v, size_t count);
+void *__memset64(uint64_t *s, uint64_t v, size_t count);
+
+static inline void *memset16(uint16_t *s, uint16_t v, size_t count)
+{
+	return __memset16(s, v, count * sizeof(v));
+}
+
+static inline void *memset32(uint32_t *s, uint32_t v, size_t count)
+{
+	return __memset32(s, v, count * sizeof(v));
+}
+
+static inline void *memset64(uint64_t *s, uint64_t v, size_t count)
+{
+	return __memset64(s, v, count * sizeof(v));
+}
+
 #if !defined(IN_ARCH_STRING_C) && (!defined(CONFIG_FORTIFY_SOURCE) || defined(__NO_FORTIFY))
 
 static inline void *memchr(const void * s, int c, size_t n)

arch/s390/lib/mem.S

@@ -126,3 +126,47 @@ ENTRY(memcpy)
 .Lmemcpy_mvc:
 	mvc	0(1,%r1),0(%r3)
 EXPORT_SYMBOL(memcpy)
+
+/*
+ * __memset16/32/64
+ *
+ * void *__memset16(uint16_t *s, uint16_t v, size_t count)
+ * void *__memset32(uint32_t *s, uint32_t v, size_t count)
+ * void *__memset64(uint64_t *s, uint64_t v, size_t count)
+ */
+.macro __MEMSET bits,bytes,insn
+ENTRY(__memset\bits)
+	ltgr	%r4,%r4
+	bzr	%r14
+	cghi	%r4,\bytes
+	je	.L__memset_exit\bits
+	aghi	%r4,-(\bytes+1)
+	srlg	%r5,%r4,8
+	ltgr	%r5,%r5
+	lgr	%r1,%r2
+	jz	.L__memset_remainder\bits
+.L__memset_loop\bits:
+	\insn	%r3,0(%r1)
+	mvc	\bytes(256-\bytes,%r1),0(%r1)
+	la	%r1,256(%r1)
+	brctg	%r5,.L__memset_loop\bits
+.L__memset_remainder\bits:
+	\insn	%r3,0(%r1)
+	larl	%r5,.L__memset_mvc\bits
+	ex	%r4,0(%r5)
+	br	%r14
+.L__memset_exit\bits:
+	\insn	%r3,0(%r2)
+	br	%r14
+.L__memset_mvc\bits:
+	mvc	\bytes(1,%r1),0(%r1)
+.endm
+
+__MEMSET 16,2,sth
+EXPORT_SYMBOL(__memset16)
+
+__MEMSET 32,4,st
+EXPORT_SYMBOL(__memset32)
+
+__MEMSET 64,8,stg
+EXPORT_SYMBOL(__memset64)