Initial file dump from open source releaseHEADmaster

13 files changed, 2160 insertions, 0 deletions

diff --git a/arch/s390/lib/Makefile b/arch/s390/lib/Makefile
new file mode 100644
index 00000000..20b0e97a
--- /dev/null
+++ b/arch/s390/lib/Makefile
@@ -0,0 +1,9 @@
+#
+# Makefile for s390-specific library files..
+#
+
+lib-y += delay.o string.o uaccess_std.o uaccess_pt.o
+obj-$(CONFIG_32BIT) += div64.o qrnnd.o ucmpdi2.o mem32.o
+obj-$(CONFIG_64BIT) += mem64.o
+lib-$(CONFIG_64BIT) += uaccess_mvcos.o
+lib-$(CONFIG_SMP) += spinlock.o
diff --git a/arch/s390/lib/delay.c b/arch/s390/lib/delay.c
new file mode 100644
index 00000000..c61b9fad
--- /dev/null
+++ b/arch/s390/lib/delay.c
@@ -0,0 +1,131 @@
+/*
+ *    Precise Delay Loops for S390
+ *
+ *    Copyright IBM Corp. 1999, 2008
+ *    Author(s): Martin Schwidefsky <schwidefsky@de.ibm.com>,
+ *		 Heiko Carstens <heiko.carstens@de.ibm.com>,
+ */
+
+#include <linux/sched.h>
+#include <linux/delay.h>
+#include <linux/timex.h>
+#include <linux/module.h>
+#include <linux/irqflags.h>
+#include <linux/interrupt.h>
+#include <asm/vtimer.h>
+#include <asm/div64.h>
+
+void __delay(unsigned long loops)
+{
+        /*
+         * To end the bloody studid and useless discussion about the
+         * BogoMips number I took the liberty to define the __delay
+         * function in a way that that resulting BogoMips number will
+         * yield the megahertz number of the cpu. The important function
+         * is udelay and that is done using the tod clock. -- martin.
+         */
+	asm volatile("0: brct %0,0b" : : "d" ((loops/2) + 1));
+}
+
+static void __udelay_disabled(unsigned long long usecs)
+{
+	unsigned long cr0, cr6, new;
+	u64 clock_saved, end;
+
+	end = get_tod_clock() + (usecs << 12);
+	clock_saved = local_tick_disable();
+	__ctl_store(cr0, 0, 0);
+	__ctl_store(cr6, 6, 6);
+	new = (cr0 &  0xffff00e0) | 0x00000800;
+	__ctl_load(new , 0, 0);
+	new = 0;
+	__ctl_load(new, 6, 6);
+	lockdep_off();
+	do {
+		set_clock_comparator(end);
+		vtime_stop_cpu();
+		local_irq_disable();
+	} while (get_tod_clock() < end);
+	lockdep_on();
+	__ctl_load(cr0, 0, 0);
+	__ctl_load(cr6, 6, 6);
+	local_tick_enable(clock_saved);
+}
+
+static void __udelay_enabled(unsigned long long usecs)
+{
+	u64 clock_saved, end;
+
+	end = get_tod_clock() + (usecs << 12);
+	do {
+		clock_saved = 0;
+		if (end < S390_lowcore.clock_comparator) {
+			clock_saved = local_tick_disable();
+			set_clock_comparator(end);
+		}
+		vtime_stop_cpu();
+		local_irq_disable();
+		if (clock_saved)
+			local_tick_enable(clock_saved);
+	} while (get_tod_clock() < end);
+}
+
+/*
+ * Waits for 'usecs' microseconds using the TOD clock comparator.
+ */
+void __udelay(unsigned long long usecs)
+{
+	unsigned long flags;
+
+	preempt_disable();
+	local_irq_save(flags);
+	if (in_irq()) {
+		__udelay_disabled(usecs);
+		goto out;
+	}
+	if (in_softirq()) {
+		if (raw_irqs_disabled_flags(flags))
+			__udelay_disabled(usecs);
+		else
+			__udelay_enabled(usecs);
+		goto out;
+	}
+	if (raw_irqs_disabled_flags(flags)) {
+		local_bh_disable();
+		__udelay_disabled(usecs);
+		_local_bh_enable();
+		goto out;
+	}
+	__udelay_enabled(usecs);
+out:
+	local_irq_restore(flags);
+	preempt_enable();
+}
+EXPORT_SYMBOL(__udelay);
+
+/*
+ * Simple udelay variant. To be used on startup and reboot
+ * when the interrupt handler isn't working.
+ */
+void udelay_simple(unsigned long long usecs)
+{
+	u64 end;
+
+	end = get_tod_clock() + (usecs << 12);
+	while (get_tod_clock() < end)
+		cpu_relax();
+}
+
+void __ndelay(unsigned long long nsecs)
+{
+	u64 end;
+
+	nsecs <<= 9;
+	do_div(nsecs, 125);
+	end = get_tod_clock() + nsecs;
+	if (nsecs & ~0xfffUL)
+		__udelay(nsecs >> 12);
+	while (get_tod_clock() < end)
+		barrier();
+}
+EXPORT_SYMBOL(__ndelay);
diff --git a/arch/s390/lib/div64.c b/arch/s390/lib/div64.c
new file mode 100644
index 00000000..261152f8
--- /dev/null
+++ b/arch/s390/lib/div64.c
@@ -0,0 +1,147 @@
+/*
+ *  __div64_32 implementation for 31 bit.
+ *
+ *    Copyright IBM Corp. 2006
+ *    Author(s): Martin Schwidefsky (schwidefsky@de.ibm.com),
+ */
+
+#include <linux/types.h>
+#include <linux/module.h>
+
+#ifdef CONFIG_MARCH_G5
+
+/*
+ * Function to divide an unsigned 64 bit integer by an unsigned
+ * 31 bit integer using signed 64/32 bit division.
+ */
+static uint32_t __div64_31(uint64_t *n, uint32_t base)
+{
+	register uint32_t reg2 asm("2");
+	register uint32_t reg3 asm("3");
+	uint32_t *words = (uint32_t *) n;
+	uint32_t tmp;
+
+	/* Special case base==1, remainder = 0, quotient = n */
+	if (base == 1)
+		return 0;
+	/*
+	 * Special case base==0 will cause a fixed point divide exception
+	 * on the dr instruction and may not happen anyway. For the
+	 * following calculation we can assume base > 1. The first
+	 * signed 64 / 32 bit division with an upper half of 0 will
+	 * give the correct upper half of the 64 bit quotient.
+	 */
+	reg2 = 0UL;
+	reg3 = words[0];
+	asm volatile(
+		"	dr	%0,%2\n"
+		: "+d" (reg2), "+d" (reg3) : "d" (base) : "cc" );
+	words[0] = reg3;
+	reg3 = words[1];
+	/*
+	 * To get the lower half of the 64 bit quotient and the 32 bit
+	 * remainder we have to use a little trick. Since we only have
+	 * a signed division the quotient can get too big. To avoid this
+	 * the 64 bit dividend is halved, then the signed division will
+	 * work. Afterwards the quotient and the remainder are doubled.
+	 * If the last bit of the dividend has been one the remainder
+	 * is increased by one then checked against the base. If the
+	 * remainder has overflown subtract base and increase the
+	 * quotient. Simple, no ?
+	 */
+	asm volatile(
+		"	nr	%2,%1\n"
+		"	srdl	%0,1\n"
+		"	dr	%0,%3\n"
+		"	alr	%0,%0\n"
+		"	alr	%1,%1\n"
+		"	alr	%0,%2\n"
+		"	clr	%0,%3\n"
+		"	jl	0f\n"
+		"	slr	%0,%3\n"
+		"	ahi	%1,1\n"
+		"0:\n"
+		: "+d" (reg2), "+d" (reg3), "=d" (tmp)
+		: "d" (base), "2" (1UL) : "cc" );
+	words[1] = reg3;
+	return reg2;
+}
+
+/*
+ * Function to divide an unsigned 64 bit integer by an unsigned
+ * 32 bit integer using the unsigned 64/31 bit division.
+ */
+uint32_t __div64_32(uint64_t *n, uint32_t base)
+{
+	uint32_t r;
+
+	/*
+	 * If the most significant bit of base is set, divide n by
+	 * (base/2). That allows to use 64/31 bit division and gives a
+	 * good approximation of the result: n = (base/2)*q + r. The
+	 * result needs to be corrected with two simple transformations.
+	 * If base is already < 2^31-1 __div64_31 can be used directly.
+	 */
+	r = __div64_31(n, ((signed) base < 0) ? (base/2) : base);
+	if ((signed) base < 0) {
+		uint64_t q = *n;
+		/*
+		 * First transformation:
+		 * n = (base/2)*q + r
+		 *   = ((base/2)*2)*(q/2) + ((q&1) ? (base/2) : 0) + r
+		 * Since r < (base/2), r + (base/2) < base.
+		 * With q1 = (q/2) and r1 = r + ((q&1) ? (base/2) : 0)
+		 * n = ((base/2)*2)*q1 + r1 with r1 < base.
+		 */
+		if (q & 1)
+			r += base/2;
+		q >>= 1;
+		/*
+		 * Second transformation. ((base/2)*2) could have lost the
+		 * last bit.
+		 * n = ((base/2)*2)*q1 + r1
+		 *   = base*q1 - ((base&1) ? q1 : 0) + r1
+		 */
+		if (base & 1) {
+			int64_t rx = r - q;
+			/*
+			 * base is >= 2^31. The worst case for the while
+			 * loop is n=2^64-1 base=2^31+1. That gives a
+			 * maximum for q=(2^64-1)/2^31 = 0x1ffffffff. Since
+			 * base >= 2^31 the loop is finished after a maximum
+			 * of three iterations.
+			 */
+			while (rx < 0) {
+				rx += base;
+				q--;
+			}
+			r = rx;
+		}
+		*n = q;
+	}
+	return r;
+}
+
+#else /* MARCH_G5 */
+
+uint32_t __div64_32(uint64_t *n, uint32_t base)
+{
+	register uint32_t reg2 asm("2");
+	register uint32_t reg3 asm("3");
+	uint32_t *words = (uint32_t *) n;
+
+	reg2 = 0UL;
+	reg3 = words[0];
+	asm volatile(
+		"	dlr	%0,%2\n"
+		: "+d" (reg2), "+d" (reg3) : "d" (base) : "cc" );
+	words[0] = reg3;
+	reg3 = words[1];
+	asm volatile(
+		"	dlr	%0,%2\n"
+		: "+d" (reg2), "+d" (reg3) : "d" (base) : "cc" );
+	words[1] = reg3;
+	return reg2;
+}
+
+#endif /* MARCH_G5 */
diff --git a/arch/s390/lib/mem32.S b/arch/s390/lib/mem32.S
new file mode 100644
index 00000000..14ca9244
--- /dev/null
+++ b/arch/s390/lib/mem32.S
@@ -0,0 +1,92 @@
+/*
+ * String handling functions.
+ *
+ * Copyright IBM Corp. 2012
+ */
+
+#include <linux/linkage.h>
+
+/*
+ * memset implementation
+ *
+ * This code corresponds to the C construct below. We do distinguish
+ * between clearing (c == 0) and setting a memory array (c != 0) simply
+ * because nearly all memset invocations in the kernel clear memory and
+ * the xc instruction is preferred in such cases.
+ *
+ * void *memset(void *s, int c, size_t n)
+ * {
+ *	if (likely(c == 0))
+ *		return __builtin_memset(s, 0, n);
+ *	return __builtin_memset(s, c, n);
+ * }
+ */
+ENTRY(memset)
+	basr	%r5,%r0
+.Lmemset_base:
+	ltr	%r4,%r4
+	bzr	%r14
+	ltr	%r3,%r3
+	jnz	.Lmemset_fill
+	ahi	%r4,-1
+	lr	%r3,%r4
+	srl	%r3,8
+	ltr	%r3,%r3
+	lr	%r1,%r2
+	je	.Lmemset_clear_rest
+.Lmemset_clear_loop:
+	xc	0(256,%r1),0(%r1)
+	la	%r1,256(%r1)
+	brct	%r3,.Lmemset_clear_loop
+.Lmemset_clear_rest:
+	ex	%r4,.Lmemset_xc-.Lmemset_base(%r5)
+	br	%r14
+.Lmemset_fill:
+	stc	%r3,0(%r2)
+	chi	%r4,1
+	lr	%r1,%r2
+	ber	%r14
+	ahi	%r4,-2
+	lr	%r3,%r4
+	srl	%r3,8
+	ltr	%r3,%r3
+	je	.Lmemset_fill_rest
+.Lmemset_fill_loop:
+	mvc	1(256,%r1),0(%r1)
+	la	%r1,256(%r1)
+	brct	%r3,.Lmemset_fill_loop
+.Lmemset_fill_rest:
+	ex	%r4,.Lmemset_mvc-.Lmemset_base(%r5)
+	br	%r14
+.Lmemset_xc:
+	xc	0(1,%r1),0(%r1)
+.Lmemset_mvc:
+	mvc	1(1,%r1),0(%r1)
+
+/*
+ * memcpy implementation
+ *
+ * void *memcpy(void *dest, const void *src, size_t n)
+ */
+ENTRY(memcpy)
+	basr	%r5,%r0
+.Lmemcpy_base:
+	ltr	%r4,%r4
+	bzr	%r14
+	ahi	%r4,-1
+	lr	%r0,%r4
+	srl	%r0,8
+	ltr	%r0,%r0
+	lr	%r1,%r2
+	jnz	.Lmemcpy_loop
+.Lmemcpy_rest:
+	ex	%r4,.Lmemcpy_mvc-.Lmemcpy_base(%r5)
+	br	%r14
+.Lmemcpy_loop:
+	mvc	0(256,%r1),0(%r3)
+	la	%r1,256(%r1)
+	la	%r3,256(%r3)
+	brct	%r0,.Lmemcpy_loop
+	j	.Lmemcpy_rest
+.Lmemcpy_mvc:
+	mvc	0(1,%r1),0(%r3)
diff --git a/arch/s390/lib/mem64.S b/arch/s390/lib/mem64.S
new file mode 100644
index 00000000..c6d553e8
--- /dev/null
+++ b/arch/s390/lib/mem64.S
@@ -0,0 +1,88 @@
+/*
+ * String handling functions.
+ *
+ * Copyright IBM Corp. 2012
+ */
+
+#include <linux/linkage.h>
+
+/*
+ * memset implementation
+ *
+ * This code corresponds to the C construct below. We do distinguish
+ * between clearing (c == 0) and setting a memory array (c != 0) simply
+ * because nearly all memset invocations in the kernel clear memory and
+ * the xc instruction is preferred in such cases.
+ *
+ * void *memset(void *s, int c, size_t n)
+ * {
+ *	if (likely(c == 0))
+ *		return __builtin_memset(s, 0, n);
+ *	return __builtin_memset(s, c, n);
+ * }
+ */
+ENTRY(memset)
+	ltgr	%r4,%r4
+	bzr	%r14
+	ltgr	%r3,%r3
+	jnz	.Lmemset_fill
+	aghi	%r4,-1
+	srlg	%r3,%r4,8
+	ltgr	%r3,%r3
+	lgr	%r1,%r2
+	jz	.Lmemset_clear_rest
+.Lmemset_clear_loop:
+	xc	0(256,%r1),0(%r1)
+	la	%r1,256(%r1)
+	brctg	%r3,.Lmemset_clear_loop
+.Lmemset_clear_rest:
+	larl	%r3,.Lmemset_xc
+	ex	%r4,0(%r3)
+	br	%r14
+.Lmemset_fill:
+	stc	%r3,0(%r2)
+	cghi	%r4,1
+	lgr	%r1,%r2
+	ber	%r14
+	aghi	%r4,-2
+	srlg	%r3,%r4,8
+	ltgr	%r3,%r3
+	jz	.Lmemset_fill_rest
+.Lmemset_fill_loop:
+	mvc	1(256,%r1),0(%r1)
+	la	%r1,256(%r1)
+	brctg	%r3,.Lmemset_fill_loop
+.Lmemset_fill_rest:
+	larl	%r3,.Lmemset_mvc
+	ex	%r4,0(%r3)
+	br	%r14
+.Lmemset_xc:
+	xc	0(1,%r1),0(%r1)
+.Lmemset_mvc:
+	mvc	1(1,%r1),0(%r1)
+
+/*
+ * memcpy implementation
+ *
+ * void *memcpy(void *dest, const void *src, size_t n)
+ */
+ENTRY(memcpy)
+	ltgr	%r4,%r4
+	bzr	%r14
+	aghi	%r4,-1
+	srlg	%r5,%r4,8
+	ltgr	%r5,%r5
+	lgr	%r1,%r2
+	jnz	.Lmemcpy_loop
+.Lmemcpy_rest:
+	larl	%r5,.Lmemcpy_mvc
+	ex	%r4,0(%r5)
+	br	%r14
+.Lmemcpy_loop:
+	mvc	0(256,%r1),0(%r3)
+	la	%r1,256(%r1)
+	la	%r3,256(%r3)
+	brctg	%r5,.Lmemcpy_loop
+	j	.Lmemcpy_rest
+.Lmemcpy_mvc:
+	mvc	0(1,%r1),0(%r3)
diff --git a/arch/s390/lib/qrnnd.S b/arch/s390/lib/qrnnd.S
new file mode 100644
index 00000000..d3213291
--- /dev/null
+++ b/arch/s390/lib/qrnnd.S
@@ -0,0 +1,78 @@
+# S/390 __udiv_qrnnd
+
+#include <linux/linkage.h>
+
+# r2 : &__r
+# r3 : upper half of 64 bit word n
+# r4 : lower half of 64 bit word n
+# r5 : divisor d
+# the reminder r of the division is to be stored to &__r and
+# the quotient q is to be returned
+
+	.text
+ENTRY(__udiv_qrnnd)
+	st    %r2,24(%r15)	  # store pointer to reminder for later
+	lr    %r0,%r3		  # reload n
+	lr    %r1,%r4
+	ltr   %r2,%r5		  # reload and test divisor
+	jp    5f
+	# divisor >= 0x80000000
+	srdl  %r0,2		  # n/4
+	srl   %r2,1		  # d/2
+	slr   %r1,%r2		  # special case if last bit of d is set
+	brc   3,0f		  #  (n/4) div (n/2) can overflow by 1
+	ahi   %r0,-1		  #  trick: subtract n/2, then divide
+0:	dr    %r0,%r2		  # signed division
+	ahi   %r1,1		  #  trick part 2: add 1 to the quotient
+	# now (n >> 2) = (d >> 1) * %r1 + %r0
+	lhi   %r3,1
+	nr    %r3,%r1		  # test last bit of q
+	jz    1f
+	alr   %r0,%r2		  # add (d>>1) to r
+1:	srl   %r1,1		  # q >>= 1
+	# now (n >> 2) = (d&-2) * %r1 + %r0
+	lhi   %r3,1
+	nr    %r3,%r5		  # test last bit of d
+	jz    2f
+	slr   %r0,%r1		  # r -= q
+	brc   3,2f		  # borrow ?
+	alr   %r0,%r5		  # r += d
+	ahi   %r1,-1
+2:	# now (n >> 2) = d * %r1 + %r0
+	alr   %r1,%r1		  # q <<= 1
+	alr   %r0,%r0		  # r <<= 1
+	brc   12,3f		  # overflow on r ?
+	slr   %r0,%r5		  # r -= d
+	ahi   %r1,1		  # q += 1
+3:	lhi   %r3,2
+	nr    %r3,%r4		  # test next to last bit of n
+	jz    4f
+	ahi   %r0,1		  # r += 1
+4:	clr   %r0,%r5		  # r >= d ?
+	jl    6f
+	slr   %r0,%r5		  # r -= d
+	ahi   %r1,1		  # q += 1
+	# now (n >> 1) = d * %r1 + %r0
+	j     6f
+5:	# divisor < 0x80000000
+	srdl  %r0,1
+	dr    %r0,%r2		  # signed division
+	# now (n >> 1) = d * %r1 + %r0
+6:	alr   %r1,%r1		  # q <<= 1
+	alr   %r0,%r0		  # r <<= 1
+	brc   12,7f		  # overflow on r ?
+	slr   %r0,%r5		  # r -= d
+	ahi   %r1,1		  # q += 1
+7:	lhi   %r3,1
+	nr    %r3,%r4		  # isolate last bit of n
+	alr   %r0,%r3		  # r += (n & 1)
+	clr   %r0,%r5		  # r >= d ?
+	jl    8f
+	slr   %r0,%r5		  # r -= d
+	ahi   %r1,1		  # q += 1
+8:	# now n = d * %r1 + %r0
+	l     %r2,24(%r15)
+	st    %r0,0(%r2)
+	lr    %r2,%r1
+	br    %r14
+	.end	__udiv_qrnnd
diff --git a/arch/s390/lib/spinlock.c b/arch/s390/lib/spinlock.c
new file mode 100644
index 00000000..f709983f
--- /dev/null
+++ b/arch/s390/lib/spinlock.c
@@ -0,0 +1,218 @@
+/*
+ *    Out of line spinlock code.
+ *
+ *    Copyright IBM Corp. 2004, 2006
+ *    Author(s): Martin Schwidefsky (schwidefsky@de.ibm.com)
+ */
+
+#include <linux/types.h>
+#include <linux/module.h>
+#include <linux/spinlock.h>
+#include <linux/init.h>
+#include <linux/smp.h>
+#include <asm/io.h>
+
+int spin_retry = 1000;
+
+/**
+ * spin_retry= parameter
+ */
+static int __init spin_retry_setup(char *str)
+{
+	spin_retry = simple_strtoul(str, &str, 0);
+	return 1;
+}
+__setup("spin_retry=", spin_retry_setup);
+
+void arch_spin_lock_wait(arch_spinlock_t *lp)
+{
+	int count = spin_retry;
+	unsigned int cpu = ~smp_processor_id();
+	unsigned int owner;
+
+	while (1) {
+		owner = lp->owner_cpu;
+		if (!owner || smp_vcpu_scheduled(~owner)) {
+			for (count = spin_retry; count > 0; count--) {
+				if (arch_spin_is_locked(lp))
+					continue;
+				if (_raw_compare_and_swap(&lp->owner_cpu, 0,
+							  cpu) == 0)
+					return;
+			}
+			if (MACHINE_IS_LPAR)
+				continue;
+		}
+		owner = lp->owner_cpu;
+		if (owner)
+			smp_yield_cpu(~owner);
+		if (_raw_compare_and_swap(&lp->owner_cpu, 0, cpu) == 0)
+			return;
+	}
+}
+EXPORT_SYMBOL(arch_spin_lock_wait);
+
+void arch_spin_lock_wait_flags(arch_spinlock_t *lp, unsigned long flags)
+{
+	int count = spin_retry;
+	unsigned int cpu = ~smp_processor_id();
+	unsigned int owner;
+
+	local_irq_restore(flags);
+	while (1) {
+		owner = lp->owner_cpu;
+		if (!owner || smp_vcpu_scheduled(~owner)) {
+			for (count = spin_retry; count > 0; count--) {
+				if (arch_spin_is_locked(lp))
+					continue;
+				local_irq_disable();
+				if (_raw_compare_and_swap(&lp->owner_cpu, 0,
+							  cpu) == 0)
+					return;
+				local_irq_restore(flags);
+			}
+			if (MACHINE_IS_LPAR)
+				continue;
+		}
+		owner = lp->owner_cpu;
+		if (owner)
+			smp_yield_cpu(~owner);
+		local_irq_disable();
+		if (_raw_compare_and_swap(&lp->owner_cpu, 0, cpu) == 0)
+			return;
+		local_irq_restore(flags);
+	}
+}
+EXPORT_SYMBOL(arch_spin_lock_wait_flags);
+
+int arch_spin_trylock_retry(arch_spinlock_t *lp)
+{
+	unsigned int cpu = ~smp_processor_id();
+	int count;
+
+	for (count = spin_retry; count > 0; count--) {
+		if (arch_spin_is_locked(lp))
+			continue;
+		if (_raw_compare_and_swap(&lp->owner_cpu, 0, cpu) == 0)
+			return 1;
+	}
+	return 0;
+}
+EXPORT_SYMBOL(arch_spin_trylock_retry);
+
+void arch_spin_relax(arch_spinlock_t *lock)
+{
+	unsigned int cpu = lock->owner_cpu;
+	if (cpu != 0) {
+		if (MACHINE_IS_VM || MACHINE_IS_KVM ||
+		    !smp_vcpu_scheduled(~cpu))
+			smp_yield_cpu(~cpu);
+	}
+}
+EXPORT_SYMBOL(arch_spin_relax);
+
+void _raw_read_lock_wait(arch_rwlock_t *rw)
+{
+	unsigned int old;
+	int count = spin_retry;
+
+	while (1) {
+		if (count-- <= 0) {
+			smp_yield();
+			count = spin_retry;
+		}
+		if (!arch_read_can_lock(rw))
+			continue;
+		old = rw->lock & 0x7fffffffU;
+		if (_raw_compare_and_swap(&rw->lock, old, old + 1) == old)
+			return;
+	}
+}
+EXPORT_SYMBOL(_raw_read_lock_wait);
+
+void _raw_read_lock_wait_flags(arch_rwlock_t *rw, unsigned long flags)
+{
+	unsigned int old;
+	int count = spin_retry;
+
+	local_irq_restore(flags);
+	while (1) {
+		if (count-- <= 0) {
+			smp_yield();
+			count = spin_retry;
+		}
+		if (!arch_read_can_lock(rw))
+			continue;
+		old = rw->lock & 0x7fffffffU;
+		local_irq_disable();
+		if (_raw_compare_and_swap(&rw->lock, old, old + 1) == old)
+			return;
+	}
+}
+EXPORT_SYMBOL(_raw_read_lock_wait_flags);
+
+int _raw_read_trylock_retry(arch_rwlock_t *rw)
+{
+	unsigned int old;
+	int count = spin_retry;
+
+	while (count-- > 0) {
+		if (!arch_read_can_lock(rw))
+			continue;
+		old = rw->lock & 0x7fffffffU;
+		if (_raw_compare_and_swap(&rw->lock, old, old + 1) == old)
+			return 1;
+	}
+	return 0;
+}
+EXPORT_SYMBOL(_raw_read_trylock_retry);
+
+void _raw_write_lock_wait(arch_rwlock_t *rw)
+{
+	int count = spin_retry;
+
+	while (1) {
+		if (count-- <= 0) {
+			smp_yield();
+			count = spin_retry;
+		}
+		if (!arch_write_can_lock(rw))
+			continue;
+		if (_raw_compare_and_swap(&rw->lock, 0, 0x80000000) == 0)
+			return;
+	}
+}
+EXPORT_SYMBOL(_raw_write_lock_wait);
+
+void _raw_write_lock_wait_flags(arch_rwlock_t *rw, unsigned long flags)
+{
+	int count = spin_retry;
+
+	local_irq_restore(flags);
+	while (1) {
+		if (count-- <= 0) {
+			smp_yield();
+			count = spin_retry;
+		}
+		if (!arch_write_can_lock(rw))
+			continue;
+		local_irq_disable();
+		if (_raw_compare_and_swap(&rw->lock, 0, 0x80000000) == 0)
+			return;
+	}
+}
+EXPORT_SYMBOL(_raw_write_lock_wait_flags);
+
+int _raw_write_trylock_retry(arch_rwlock_t *rw)
+{
+	int count = spin_retry;
+
+	while (count-- > 0) {
+		if (!arch_write_can_lock(rw))
+			continue;
+		if (_raw_compare_and_swap(&rw->lock, 0, 0x80000000) == 0)
+			return 1;
+	}
+	return 0;
+}
+EXPORT_SYMBOL(_raw_write_trylock_retry);
diff --git a/arch/s390/lib/string.c b/arch/s390/lib/string.c
new file mode 100644
index 00000000..b647d5ff
--- /dev/null
+++ b/arch/s390/lib/string.c
@@ -0,0 +1,342 @@
+/*
+ *    Optimized string functions
+ *
+ *  S390 version
+ *    Copyright IBM Corp. 2004
+ *    Author(s): Martin Schwidefsky (schwidefsky@de.ibm.com)
+ */
+
+#define IN_ARCH_STRING_C 1
+
+#include <linux/types.h>
+#include <linux/module.h>
+
+/*
+ * Helper functions to find the end of a string
+ */
+static inline char *__strend(const char *s)
+{
+	register unsigned long r0 asm("0") = 0;
+
+	asm volatile ("0: srst  %0,%1\n"
+		      "   jo    0b"
+		      : "+d" (r0), "+a" (s) :  : "cc" );
+	return (char *) r0;
+}
+
+static inline char *__strnend(const char *s, size_t n)
+{
+	register unsigned long r0 asm("0") = 0;
+	const char *p = s + n;
+
+	asm volatile ("0: srst  %0,%1\n"
+		      "   jo    0b"
+		      : "+d" (p), "+a" (s) : "d" (r0) : "cc" );
+	return (char *) p;
+}
+
+/**
+ * strlen - Find the length of a string
+ * @s: The string to be sized
+ *
+ * returns the length of @s
+ */
+size_t strlen(const char *s)
+{
+	return __strend(s) - s;
+}
+EXPORT_SYMBOL(strlen);
+
+/**
+ * strnlen - Find the length of a length-limited string
+ * @s: The string to be sized
+ * @n: The maximum number of bytes to search
+ *
+ * returns the minimum of the length of @s and @n
+ */
+size_t strnlen(const char * s, size_t n)
+{
+	return __strnend(s, n) - s;
+}
+EXPORT_SYMBOL(strnlen);
+
+/**
+ * strcpy - Copy a %NUL terminated string
+ * @dest: Where to copy the string to
+ * @src: Where to copy the string from
+ *
+ * returns a pointer to @dest
+ */
+char *strcpy(char *dest, const char *src)
+{
+	register int r0 asm("0") = 0;
+	char *ret = dest;
+
+	asm volatile ("0: mvst  %0,%1\n"
+		      "   jo    0b"
+		      : "+&a" (dest), "+&a" (src) : "d" (r0)
+		      : "cc", "memory" );
+	return ret;
+}
+EXPORT_SYMBOL(strcpy);
+
+/**
+ * strlcpy - Copy a %NUL terminated string into a sized buffer
+ * @dest: Where to copy the string to
+ * @src: Where to copy the string from
+ * @size: size of destination buffer
+ *
+ * Compatible with *BSD: the result is always a valid
+ * NUL-terminated string that fits in the buffer (unless,
+ * of course, the buffer size is zero). It does not pad
+ * out the result like strncpy() does.
+ */
+size_t strlcpy(char *dest, const char *src, size_t size)
+{
+	size_t ret = __strend(src) - src;
+
+	if (size) {
+		size_t len = (ret >= size) ? size-1 : ret;
+		dest[len] = '\0';
+		memcpy(dest, src, len);
+	}
+	return ret;
+}
+EXPORT_SYMBOL(strlcpy);
+
+/**
+ * strncpy - Copy a length-limited, %NUL-terminated string
+ * @dest: Where to copy the string to
+ * @src: Where to copy the string from
+ * @n: The maximum number of bytes to copy
+ *
+ * The result is not %NUL-terminated if the source exceeds
+ * @n bytes.
+ */
+char *strncpy(char *dest, const char *src, size_t n)
+{
+	size_t len = __strnend(src, n) - src;
+	memset(dest + len, 0, n - len);
+	memcpy(dest, src, len);
+	return dest;
+}
+EXPORT_SYMBOL(strncpy);
+
+/**
+ * strcat - Append one %NUL-terminated string to another
+ * @dest: The string to be appended to
+ * @src: The string to append to it
+ *
+ * returns a pointer to @dest
+ */
+char *strcat(char *dest, const char *src)
+{
+	register int r0 asm("0") = 0;
+	unsigned long dummy;
+	char *ret = dest;
+
+	asm volatile ("0: srst  %0,%1\n"
+		      "   jo    0b\n"
+		      "1: mvst  %0,%2\n"
+		      "   jo    1b"
+		      : "=&a" (dummy), "+a" (dest), "+a" (src)
+		      : "d" (r0), "0" (0UL) : "cc", "memory" );
+	return ret;
+}
+EXPORT_SYMBOL(strcat);
+
+/**
+ * strlcat - Append a length-limited, %NUL-terminated string to another
+ * @dest: The string to be appended to
+ * @src: The string to append to it
+ * @n: The size of the destination buffer.
+ */
+size_t strlcat(char *dest, const char *src, size_t n)
+{
+	size_t dsize = __strend(dest) - dest;
+	size_t len = __strend(src) - src;
+	size_t res = dsize + len;
+
+	if (dsize < n) {
+		dest += dsize;
+		n -= dsize;
+		if (len >= n)
+			len = n - 1;
+		dest[len] = '\0';
+		memcpy(dest, src, len);
+	}
+	return res;
+}
+EXPORT_SYMBOL(strlcat);
+
+/**
+ * strncat - Append a length-limited, %NUL-terminated string to another
+ * @dest: The string to be appended to
+ * @src: The string to append to it
+ * @n: The maximum numbers of bytes to copy
+ *
+ * returns a pointer to @dest
+ *
+ * Note that in contrast to strncpy, strncat ensures the result is
+ * terminated.
+ */
+char *strncat(char *dest, const char *src, size_t n)
+{
+	size_t len = __strnend(src, n) - src;
+	char *p = __strend(dest);
+
+	p[len] = '\0';
+	memcpy(p, src, len);
+	return dest;
+}
+EXPORT_SYMBOL(strncat);
+
+/**
+ * strcmp - Compare two strings
+ * @cs: One string
+ * @ct: Another string
+ *
+ * returns   0 if @cs and @ct are equal,
+ *         < 0 if @cs is less than @ct
+ *         > 0 if @cs is greater than @ct
+ */
+int strcmp(const char *cs, const char *ct)
+{
+	register int r0 asm("0") = 0;
+	int ret = 0;
+
+	asm volatile ("0: clst %2,%3\n"
+		      "   jo   0b\n"
+		      "   je   1f\n"
+		      "   ic   %0,0(%2)\n"
+		      "   ic   %1,0(%3)\n"
+		      "   sr   %0,%1\n"
+		      "1:"
+		      : "+d" (ret), "+d" (r0), "+a" (cs), "+a" (ct)
+		      : : "cc" );
+	return ret;
+}
+EXPORT_SYMBOL(strcmp);
+
+/**
+ * strrchr - Find the last occurrence of a character in a string
+ * @s: The string to be searched
+ * @c: The character to search for
+ */
+char * strrchr(const char * s, int c)
+{
+       size_t len = __strend(s) - s;
+
+       if (len)
+	       do {
+		       if (s[len] == (char) c)
+			       return (char *) s + len;
+	       } while (--len > 0);
+       return NULL;
+}
+EXPORT_SYMBOL(strrchr);
+
+/**
+ * strstr - Find the first substring in a %NUL terminated string
+ * @s1: The string to be searched
+ * @s2: The string to search for
+ */
+char * strstr(const char * s1,const char * s2)
+{
+	int l1, l2;
+
+	l2 = __strend(s2) - s2;
+	if (!l2)
+		return (char *) s1;
+	l1 = __strend(s1) - s1;
+	while (l1-- >= l2) {
+		register unsigned long r2 asm("2") = (unsigned long) s1;
+		register unsigned long r3 asm("3") = (unsigned long) l2;
+		register unsigned long r4 asm("4") = (unsigned long) s2;
+		register unsigned long r5 asm("5") = (unsigned long) l2;
+		int cc;
+
+		asm volatile ("0: clcle %1,%3,0\n"
+			      "   jo    0b\n"
+			      "   ipm   %0\n"
+			      "   srl   %0,28"
+			      : "=&d" (cc), "+a" (r2), "+a" (r3),
+			        "+a" (r4), "+a" (r5) : : "cc" );
+		if (!cc)
+			return (char *) s1;
+		s1++;
+	}
+	return NULL;
+}
+EXPORT_SYMBOL(strstr);
+
+/**
+ * memchr - Find a character in an area of memory.
+ * @s: The memory area
+ * @c: The byte to search for
+ * @n: The size of the area.
+ *
+ * returns the address of the first occurrence of @c, or %NULL
+ * if @c is not found
+ */
+void *memchr(const void *s, int c, size_t n)
+{
+	register int r0 asm("0") = (char) c;
+	const void *ret = s + n;
+
+	asm volatile ("0: srst  %0,%1\n"
+		      "   jo    0b\n"
+		      "   jl	1f\n"
+		      "   la    %0,0\n"
+		      "1:"
+		      : "+a" (ret), "+&a" (s) : "d" (r0) : "cc" );
+	return (void *) ret;
+}
+EXPORT_SYMBOL(memchr);
+
+/**
+ * memcmp - Compare two areas of memory
+ * @cs: One area of memory
+ * @ct: Another area of memory
+ * @count: The size of the area.
+ */
+int memcmp(const void *cs, const void *ct, size_t n)
+{
+	register unsigned long r2 asm("2") = (unsigned long) cs;
+	register unsigned long r3 asm("3") = (unsigned long) n;
+	register unsigned long r4 asm("4") = (unsigned long) ct;
+	register unsigned long r5 asm("5") = (unsigned long) n;
+	int ret;
+
+	asm volatile ("0: clcle %1,%3,0\n"
+		      "   jo    0b\n"
+		      "   ipm   %0\n"
+		      "   srl   %0,28"
+		      : "=&d" (ret), "+a" (r2), "+a" (r3), "+a" (r4), "+a" (r5)
+		      : : "cc" );
+	if (ret)
+		ret = *(char *) r2 - *(char *) r4;
+	return ret;
+}
+EXPORT_SYMBOL(memcmp);
+
+/**
+ * memscan - Find a character in an area of memory.
+ * @s: The memory area
+ * @c: The byte to search for
+ * @n: The size of the area.
+ *
+ * returns the address of the first occurrence of @c, or 1 byte past
+ * the area if @c is not found
+ */
+void *memscan(void *s, int c, size_t n)
+{
+	register int r0 asm("0") = (char) c;
+	const void *ret = s + n;
+
+	asm volatile ("0: srst  %0,%1\n"
+		      "   jo    0b\n"
+		      : "+a" (ret), "+&a" (s) : "d" (r0) : "cc" );
+	return (void *) ret;
+}
+EXPORT_SYMBOL(memscan);
diff --git a/arch/s390/lib/uaccess.h b/arch/s390/lib/uaccess.h
new file mode 100644
index 00000000..315dbe09
--- /dev/null
+++ b/arch/s390/lib/uaccess.h
@@ -0,0 +1,21 @@
+/*
+ *    Copyright IBM Corp. 2007
+ *
+ */
+
+#ifndef __ARCH_S390_LIB_UACCESS_H
+#define __ARCH_S390_LIB_UACCESS_H
+
+extern size_t copy_from_user_std(size_t, const void __user *, void *);
+extern size_t copy_to_user_std(size_t, void __user *, const void *);
+extern size_t strnlen_user_std(size_t, const char __user *);
+extern size_t strncpy_from_user_std(size_t, const char __user *, char *);
+extern int futex_atomic_cmpxchg_std(u32 *, u32 __user *, u32, u32);
+extern int futex_atomic_op_std(int, u32 __user *, int, int *);
+
+extern size_t copy_from_user_pt(size_t, const void __user *, void *);
+extern size_t copy_to_user_pt(size_t, void __user *, const void *);
+extern int futex_atomic_op_pt(int, u32 __user *, int, int *);
+extern int futex_atomic_cmpxchg_pt(u32 *, u32 __user *, u32, u32);
+
+#endif /* __ARCH_S390_LIB_UACCESS_H */
diff --git a/arch/s390/lib/uaccess_mvcos.c b/arch/s390/lib/uaccess_mvcos.c
new file mode 100644
index 00000000..1829742b
--- /dev/null
+++ b/arch/s390/lib/uaccess_mvcos.c
@@ -0,0 +1,227 @@
+/*
+ *  Optimized user space space access functions based on mvcos.
+ *
+ *    Copyright IBM Corp. 2006
+ *    Author(s): Martin Schwidefsky (schwidefsky@de.ibm.com),
+ *		 Gerald Schaefer (gerald.schaefer@de.ibm.com)
+ */
+
+#include <linux/errno.h>
+#include <linux/mm.h>
+#include <asm/uaccess.h>
+#include <asm/futex.h>
+#include "uaccess.h"
+
+#ifndef CONFIG_64BIT
+#define AHI	"ahi"
+#define ALR	"alr"
+#define CLR	"clr"
+#define LHI	"lhi"
+#define SLR	"slr"
+#else
+#define AHI	"aghi"
+#define ALR	"algr"
+#define CLR	"clgr"
+#define LHI	"lghi"
+#define SLR	"slgr"
+#endif
+
+static size_t copy_from_user_mvcos(size_t size, const void __user *ptr, void *x)
+{
+	register unsigned long reg0 asm("0") = 0x81UL;
+	unsigned long tmp1, tmp2;
+
+	tmp1 = -4096UL;
+	asm volatile(
+		"0: .insn ss,0xc80000000000,0(%0,%2),0(%1),0\n"
+		"9: jz    7f\n"
+		"1:"ALR"  %0,%3\n"
+		"  "SLR"  %1,%3\n"
+		"  "SLR"  %2,%3\n"
+		"   j     0b\n"
+		"2: la    %4,4095(%1)\n"/* %4 = ptr + 4095 */
+		"   nr    %4,%3\n"	/* %4 = (ptr + 4095) & -4096 */
+		"  "SLR"  %4,%1\n"
+		"  "CLR"  %0,%4\n"	/* copy crosses next page boundary? */
+		"   jnh   4f\n"
+		"3: .insn ss,0xc80000000000,0(%4,%2),0(%1),0\n"
+		"10:"SLR"  %0,%4\n"
+		"  "ALR"  %2,%4\n"
+		"4:"LHI"  %4,-1\n"
+		"  "ALR"  %4,%0\n"	/* copy remaining size, subtract 1 */
+		"   bras  %3,6f\n"	/* memset loop */
+		"   xc    0(1,%2),0(%2)\n"
+		"5: xc    0(256,%2),0(%2)\n"
+		"   la    %2,256(%2)\n"
+		"6:"AHI"  %4,-256\n"
+		"   jnm   5b\n"
+		"   ex    %4,0(%3)\n"
+		"   j     8f\n"
+		"7:"SLR"  %0,%0\n"
+		"8: \n"
+		EX_TABLE(0b,2b) EX_TABLE(3b,4b) EX_TABLE(9b,2b) EX_TABLE(10b,4b)
+		: "+a" (size), "+a" (ptr), "+a" (x), "+a" (tmp1), "=a" (tmp2)
+		: "d" (reg0) : "cc", "memory");
+	return size;
+}
+
+static size_t copy_from_user_mvcos_check(size_t size, const void __user *ptr, void *x)
+{
+	if (size <= 256)
+		return copy_from_user_std(size, ptr, x);
+	return copy_from_user_mvcos(size, ptr, x);
+}
+
+static size_t copy_to_user_mvcos(size_t size, void __user *ptr, const void *x)
+{
+	register unsigned long reg0 asm("0") = 0x810000UL;
+	unsigned long tmp1, tmp2;
+
+	tmp1 = -4096UL;
+	asm volatile(
+		"0: .insn ss,0xc80000000000,0(%0,%1),0(%2),0\n"
+		"6: jz    4f\n"
+		"1:"ALR"  %0,%3\n"
+		"  "SLR"  %1,%3\n"
+		"  "SLR"  %2,%3\n"
+		"   j     0b\n"
+		"2: la    %4,4095(%1)\n"/* %4 = ptr + 4095 */
+		"   nr    %4,%3\n"	/* %4 = (ptr + 4095) & -4096 */
+		"  "SLR"  %4,%1\n"
+		"  "CLR"  %0,%4\n"	/* copy crosses next page boundary? */
+		"   jnh   5f\n"
+		"3: .insn ss,0xc80000000000,0(%4,%1),0(%2),0\n"
+		"7:"SLR"  %0,%4\n"
+		"   j     5f\n"
+		"4:"SLR"  %0,%0\n"
+		"5: \n"
+		EX_TABLE(0b,2b) EX_TABLE(3b,5b) EX_TABLE(6b,2b) EX_TABLE(7b,5b)
+		: "+a" (size), "+a" (ptr), "+a" (x), "+a" (tmp1), "=a" (tmp2)
+		: "d" (reg0) : "cc", "memory");
+	return size;
+}
+
+static size_t copy_to_user_mvcos_check(size_t size, void __user *ptr,
+				       const void *x)
+{
+	if (size <= 256)
+		return copy_to_user_std(size, ptr, x);
+	return copy_to_user_mvcos(size, ptr, x);
+}
+
+static size_t copy_in_user_mvcos(size_t size, void __user *to,
+				 const void __user *from)
+{
+	register unsigned long reg0 asm("0") = 0x810081UL;
+	unsigned long tmp1, tmp2;
+
+	tmp1 = -4096UL;
+	/* FIXME: copy with reduced length. */
+	asm volatile(
+		"0: .insn ss,0xc80000000000,0(%0,%1),0(%2),0\n"
+		"   jz    2f\n"
+		"1:"ALR"  %0,%3\n"
+		"  "SLR"  %1,%3\n"
+		"  "SLR"  %2,%3\n"
+		"   j     0b\n"
+		"2:"SLR"  %0,%0\n"
+		"3: \n"
+		EX_TABLE(0b,3b)
+		: "+a" (size), "+a" (to), "+a" (from), "+a" (tmp1), "=a" (tmp2)
+		: "d" (reg0) : "cc", "memory");
+	return size;
+}
+
+static size_t clear_user_mvcos(size_t size, void __user *to)
+{
+	register unsigned long reg0 asm("0") = 0x810000UL;
+	unsigned long tmp1, tmp2;
+
+	tmp1 = -4096UL;
+	asm volatile(
+		"0: .insn ss,0xc80000000000,0(%0,%1),0(%4),0\n"
+		"   jz    4f\n"
+		"1:"ALR"  %0,%2\n"
+		"  "SLR"  %1,%2\n"
+		"   j     0b\n"
+		"2: la    %3,4095(%1)\n"/* %4 = to + 4095 */
+		"   nr    %3,%2\n"	/* %4 = (to + 4095) & -4096 */
+		"  "SLR"  %3,%1\n"
+		"  "CLR"  %0,%3\n"	/* copy crosses next page boundary? */
+		"   jnh   5f\n"
+		"3: .insn ss,0xc80000000000,0(%3,%1),0(%4),0\n"
+		"  "SLR"  %0,%3\n"
+		"   j     5f\n"
+		"4:"SLR"  %0,%0\n"
+		"5: \n"
+		EX_TABLE(0b,2b) EX_TABLE(3b,5b)
+		: "+a" (size), "+a" (to), "+a" (tmp1), "=a" (tmp2)
+		: "a" (empty_zero_page), "d" (reg0) : "cc", "memory");
+	return size;
+}
+
+static size_t strnlen_user_mvcos(size_t count, const char __user *src)
+{
+	size_t done, len, offset, len_str;
+	char buf[256];
+
+	done = 0;
+	do {
+		offset = (size_t)src & ~PAGE_MASK;
+		len = min(256UL, PAGE_SIZE - offset);
+		len = min(count - done, len);
+		if (copy_from_user_mvcos(len, src, buf))
+			return 0;
+		len_str = strnlen(buf, len);
+		done += len_str;
+		src += len_str;
+	} while ((len_str == len) && (done < count));
+	return done + 1;
+}
+
+static size_t strncpy_from_user_mvcos(size_t count, const char __user *src,
+				      char *dst)
+{
+	size_t done, len, offset, len_str;
+
+	if (unlikely(!count))
+		return 0;
+	done = 0;
+	do {
+		offset = (size_t)src & ~PAGE_MASK;
+		len = min(count - done, PAGE_SIZE - offset);
+		if (copy_from_user_mvcos(len, src, dst))
+			return -EFAULT;
+		len_str = strnlen(dst, len);
+		done += len_str;
+		src += len_str;
+		dst += len_str;
+	} while ((len_str == len) && (done < count));
+	return done;
+}
+
+struct uaccess_ops uaccess_mvcos = {
+	.copy_from_user = copy_from_user_mvcos_check,
+	.copy_from_user_small = copy_from_user_std,
+	.copy_to_user = copy_to_user_mvcos_check,
+	.copy_to_user_small = copy_to_user_std,
+	.copy_in_user = copy_in_user_mvcos,
+	.clear_user = clear_user_mvcos,
+	.strnlen_user = strnlen_user_std,
+	.strncpy_from_user = strncpy_from_user_std,
+	.futex_atomic_op = futex_atomic_op_std,
+	.futex_atomic_cmpxchg = futex_atomic_cmpxchg_std,
+};
+
+struct uaccess_ops uaccess_mvcos_switch = {
+	.copy_from_user = copy_from_user_mvcos,
+	.copy_from_user_small = copy_from_user_mvcos,
+	.copy_to_user = copy_to_user_mvcos,
+	.copy_to_user_small = copy_to_user_mvcos,
+	.copy_in_user = copy_in_user_mvcos,
+	.clear_user = clear_user_mvcos,
+	.strnlen_user = strnlen_user_mvcos,
+	.strncpy_from_user = strncpy_from_user_mvcos,
+	.futex_atomic_op = futex_atomic_op_pt,
+	.futex_atomic_cmpxchg = futex_atomic_cmpxchg_pt,
+};
diff --git a/arch/s390/lib/uaccess_pt.c b/arch/s390/lib/uaccess_pt.c
new file mode 100644
index 00000000..1bf40ef5
--- /dev/null
+++ b/arch/s390/lib/uaccess_pt.c
@@ -0,0 +1,476 @@
+/*
+ *  User access functions based on page table walks for enhanced
+ *  system layout without hardware support.
+ *
+ *    Copyright IBM Corp. 2006, 2012
+ *    Author(s): Gerald Schaefer (gerald.schaefer@de.ibm.com)
+ */
+
+#include <linux/errno.h>
+#include <linux/hardirq.h>
+#include <linux/mm.h>
+#include <linux/hugetlb.h>
+#include <asm/uaccess.h>
+#include <asm/futex.h>
+#include "uaccess.h"
+
+#ifndef CONFIG_64BIT
+#define AHI	"ahi"
+#define SLR	"slr"
+#else
+#define AHI	"aghi"
+#define SLR	"slgr"
+#endif
+
+static size_t strnlen_kernel(size_t count, const char __user *src)
+{
+	register unsigned long reg0 asm("0") = 0UL;
+	unsigned long tmp1, tmp2;
+
+	asm volatile(
+		"   la	  %2,0(%1)\n"
+		"   la	  %3,0(%0,%1)\n"
+		"  "SLR"  %0,%0\n"
+		"0: srst  %3,%2\n"
+		"   jo	  0b\n"
+		"   la	  %0,1(%3)\n"	/* strnlen_kernel results includes \0 */
+		"  "SLR"  %0,%1\n"
+		"1:\n"
+		EX_TABLE(0b,1b)
+		: "+a" (count), "+a" (src), "=a" (tmp1), "=a" (tmp2)
+		: "d" (reg0) : "cc", "memory");
+	return count;
+}
+
+static size_t copy_in_kernel(size_t count, void __user *to,
+			     const void __user *from)
+{
+	unsigned long tmp1;
+
+	asm volatile(
+		"  "AHI"  %0,-1\n"
+		"   jo	  5f\n"
+		"   bras  %3,3f\n"
+		"0:"AHI"  %0,257\n"
+		"1: mvc	  0(1,%1),0(%2)\n"
+		"   la	  %1,1(%1)\n"
+		"   la	  %2,1(%2)\n"
+		"  "AHI"  %0,-1\n"
+		"   jnz	  1b\n"
+		"   j	  5f\n"
+		"2: mvc	  0(256,%1),0(%2)\n"
+		"   la	  %1,256(%1)\n"
+		"   la	  %2,256(%2)\n"
+		"3:"AHI"  %0,-256\n"
+		"   jnm	  2b\n"
+		"4: ex	  %0,1b-0b(%3)\n"
+		"5:"SLR"  %0,%0\n"
+		"6:\n"
+		EX_TABLE(1b,6b) EX_TABLE(2b,0b) EX_TABLE(4b,0b)
+		: "+a" (count), "+a" (to), "+a" (from), "=a" (tmp1)
+		: : "cc", "memory");
+	return count;
+}
+
+/*
+ * Returns kernel address for user virtual address. If the returned address is
+ * >= -4095 (IS_ERR_VALUE(x) returns true), a fault has occured and the address
+ * contains the (negative) exception code.
+ */
+#ifdef CONFIG_64BIT
+
+static unsigned long follow_table(struct mm_struct *mm,
+				  unsigned long address, int write)
+{
+	unsigned long *table = (unsigned long *)__pa(mm->pgd);
+
+	if (unlikely(address > mm->context.asce_limit - 1))
+		return -0x38UL;
+	switch (mm->context.asce_bits & _ASCE_TYPE_MASK) {
+	case _ASCE_TYPE_REGION1:
+		table = table + ((address >> 53) & 0x7ff);
+		if (unlikely(*table & _REGION_ENTRY_INV))
+			return -0x39UL;
+		table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN);
+		/* fallthrough */
+	case _ASCE_TYPE_REGION2:
+		table = table + ((address >> 42) & 0x7ff);
+		if (unlikely(*table & _REGION_ENTRY_INV))
+			return -0x3aUL;
+		table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN);
+		/* fallthrough */
+	case _ASCE_TYPE_REGION3:
+		table = table + ((address >> 31) & 0x7ff);
+		if (unlikely(*table & _REGION_ENTRY_INV))
+			return -0x3bUL;
+		table = (unsigned long *)(*table & _REGION_ENTRY_ORIGIN);
+		/* fallthrough */
+	case _ASCE_TYPE_SEGMENT:
+		table = table + ((address >> 20) & 0x7ff);
+		if (unlikely(*table & _SEGMENT_ENTRY_INV))
+			return -0x10UL;
+		if (unlikely(*table & _SEGMENT_ENTRY_LARGE)) {
+			if (write && (*table & _SEGMENT_ENTRY_RO))
+				return -0x04UL;
+			return (*table & _SEGMENT_ENTRY_ORIGIN_LARGE) +
+				(address & ~_SEGMENT_ENTRY_ORIGIN_LARGE);
+		}
+		table = (unsigned long *)(*table & _SEGMENT_ENTRY_ORIGIN);
+	}
+	table = table + ((address >> 12) & 0xff);
+	if (unlikely(*table & _PAGE_INVALID))
+		return -0x11UL;
+	if (write && (*table & _PAGE_RO))
+		return -0x04UL;
+	return (*table & PAGE_MASK) + (address & ~PAGE_MASK);
+}
+
+#else /* CONFIG_64BIT */
+
+static unsigned long follow_table(struct mm_struct *mm,
+				  unsigned long address, int write)
+{
+	unsigned long *table = (unsigned long *)__pa(mm->pgd);
+
+	table = table + ((address >> 20) & 0x7ff);
+	if (unlikely(*table & _SEGMENT_ENTRY_INV))
+		return -0x10UL;
+	table = (unsigned long *)(*table & _SEGMENT_ENTRY_ORIGIN);
+	table = table + ((address >> 12) & 0xff);
+	if (unlikely(*table & _PAGE_INVALID))
+		return -0x11UL;
+	if (write && (*table & _PAGE_RO))
+		return -0x04UL;
+	return (*table & PAGE_MASK) + (address & ~PAGE_MASK);
+}
+
+#endif /* CONFIG_64BIT */
+
+static __always_inline size_t __user_copy_pt(unsigned long uaddr, void *kptr,
+					     size_t n, int write_user)
+{
+	struct mm_struct *mm = current->mm;
+	unsigned long offset, done, size, kaddr;
+	void *from, *to;
+
+	done = 0;
+retry:
+	spin_lock(&mm->page_table_lock);
+	do {
+		kaddr = follow_table(mm, uaddr, write_user);
+		if (IS_ERR_VALUE(kaddr))
+			goto fault;
+
+		offset = uaddr & ~PAGE_MASK;
+		size = min(n - done, PAGE_SIZE - offset);
+		if (write_user) {
+			to = (void *) kaddr;
+			from = kptr + done;
+		} else {
+			from = (void *) kaddr;
+			to = kptr + done;
+		}
+		memcpy(to, from, size);
+		done += size;
+		uaddr += size;
+	} while (done < n);
+	spin_unlock(&mm->page_table_lock);
+	return n - done;
+fault:
+	spin_unlock(&mm->page_table_lock);
+	if (__handle_fault(uaddr, -kaddr, write_user))
+		return n - done;
+	goto retry;
+}
+
+/*
+ * Do DAT for user address by page table walk, return kernel address.
+ * This function needs to be called with current->mm->page_table_lock held.
+ */
+static __always_inline unsigned long __dat_user_addr(unsigned long uaddr,
+						     int write)
+{
+	struct mm_struct *mm = current->mm;
+	unsigned long kaddr;
+	int rc;
+
+retry:
+	kaddr = follow_table(mm, uaddr, write);
+	if (IS_ERR_VALUE(kaddr))
+		goto fault;
+
+	return kaddr;
+fault:
+	spin_unlock(&mm->page_table_lock);
+	rc = __handle_fault(uaddr, -kaddr, write);
+	spin_lock(&mm->page_table_lock);
+	if (!rc)
+		goto retry;
+	return 0;
+}
+
+size_t copy_from_user_pt(size_t n, const void __user *from, void *to)
+{
+	size_t rc;
+
+	if (segment_eq(get_fs(), KERNEL_DS))
+		return copy_in_kernel(n, (void __user *) to, from);
+	rc = __user_copy_pt((unsigned long) from, to, n, 0);
+	if (unlikely(rc))
+		memset(to + n - rc, 0, rc);
+	return rc;
+}
+
+size_t copy_to_user_pt(size_t n, void __user *to, const void *from)
+{
+	if (segment_eq(get_fs(), KERNEL_DS))
+		return copy_in_kernel(n, to, (void __user *) from);
+	return __user_copy_pt((unsigned long) to, (void *) from, n, 1);
+}
+
+static size_t clear_user_pt(size_t n, void __user *to)
+{
+	void *zpage = (void *) empty_zero_page;
+	long done, size, ret;
+
+	done = 0;
+	do {
+		if (n - done > PAGE_SIZE)
+			size = PAGE_SIZE;
+		else
+			size = n - done;
+		if (segment_eq(get_fs(), KERNEL_DS))
+			ret = copy_in_kernel(n, to, (void __user *) zpage);
+		else
+			ret = __user_copy_pt((unsigned long) to, zpage, size, 1);
+		done += size;
+		to += size;
+		if (ret)
+			return ret + n - done;
+	} while (done < n);
+	return 0;
+}
+
+static size_t strnlen_user_pt(size_t count, const char __user *src)
+{
+	unsigned long uaddr = (unsigned long) src;
+	struct mm_struct *mm = current->mm;
+	unsigned long offset, done, len, kaddr;
+	size_t len_str;
+
+	if (unlikely(!count))
+		return 0;
+	if (segment_eq(get_fs(), KERNEL_DS))
+		return strnlen_kernel(count, src);
+	done = 0;
+retry:
+	spin_lock(&mm->page_table_lock);
+	do {
+		kaddr = follow_table(mm, uaddr, 0);
+		if (IS_ERR_VALUE(kaddr))
+			goto fault;
+
+		offset = uaddr & ~PAGE_MASK;
+		len = min(count - done, PAGE_SIZE - offset);
+		len_str = strnlen((char *) kaddr, len);
+		done += len_str;
+		uaddr += len_str;
+	} while ((len_str == len) && (done < count));
+	spin_unlock(&mm->page_table_lock);
+	return done + 1;
+fault:
+	spin_unlock(&mm->page_table_lock);
+	if (__handle_fault(uaddr, -kaddr, 0))
+		return 0;
+	goto retry;
+}
+
+static size_t strncpy_from_user_pt(size_t count, const char __user *src,
+				   char *dst)
+{
+	size_t done, len, offset, len_str;
+
+	if (unlikely(!count))
+		return 0;
+	done = 0;
+	do {
+		offset = (size_t)src & ~PAGE_MASK;
+		len = min(count - done, PAGE_SIZE - offset);
+		if (segment_eq(get_fs(), KERNEL_DS)) {
+			if (copy_in_kernel(len, (void __user *) dst, src))
+				return -EFAULT;
+		} else {
+			if (__user_copy_pt((unsigned long) src, dst, len, 0))
+				return -EFAULT;
+		}
+		len_str = strnlen(dst, len);
+		done += len_str;
+		src += len_str;
+		dst += len_str;
+	} while ((len_str == len) && (done < count));
+	return done;
+}
+
+static size_t copy_in_user_pt(size_t n, void __user *to,
+			      const void __user *from)
+{
+	struct mm_struct *mm = current->mm;
+	unsigned long offset_max, uaddr, done, size, error_code;
+	unsigned long uaddr_from = (unsigned long) from;
+	unsigned long uaddr_to = (unsigned long) to;
+	unsigned long kaddr_to, kaddr_from;
+	int write_user;
+
+	if (segment_eq(get_fs(), KERNEL_DS))
+		return copy_in_kernel(n, to, from);
+	done = 0;
+retry:
+	spin_lock(&mm->page_table_lock);
+	do {
+		write_user = 0;
+		uaddr = uaddr_from;
+		kaddr_from = follow_table(mm, uaddr_from, 0);
+		error_code = kaddr_from;
+		if (IS_ERR_VALUE(error_code))
+			goto fault;
+
+		write_user = 1;
+		uaddr = uaddr_to;
+		kaddr_to = follow_table(mm, uaddr_to, 1);
+		error_code = (unsigned long) kaddr_to;
+		if (IS_ERR_VALUE(error_code))
+			goto fault;
+
+		offset_max = max(uaddr_from & ~PAGE_MASK,
+				 uaddr_to & ~PAGE_MASK);
+		size = min(n - done, PAGE_SIZE - offset_max);
+
+		memcpy((void *) kaddr_to, (void *) kaddr_from, size);
+		done += size;
+		uaddr_from += size;
+		uaddr_to += size;
+	} while (done < n);
+	spin_unlock(&mm->page_table_lock);
+	return n - done;
+fault:
+	spin_unlock(&mm->page_table_lock);
+	if (__handle_fault(uaddr, -error_code, write_user))
+		return n - done;
+	goto retry;
+}
+
+#define __futex_atomic_op(insn, ret, oldval, newval, uaddr, oparg)	\
+	asm volatile("0: l   %1,0(%6)\n"				\
+		     "1: " insn						\
+		     "2: cs  %1,%2,0(%6)\n"				\
+		     "3: jl  1b\n"					\
+		     "   lhi %0,0\n"					\
+		     "4:\n"						\
+		     EX_TABLE(0b,4b) EX_TABLE(2b,4b) EX_TABLE(3b,4b)	\
+		     : "=d" (ret), "=&d" (oldval), "=&d" (newval),	\
+		       "=m" (*uaddr)					\
+		     : "0" (-EFAULT), "d" (oparg), "a" (uaddr),		\
+		       "m" (*uaddr) : "cc" );
+
+static int __futex_atomic_op_pt(int op, u32 __user *uaddr, int oparg, int *old)
+{
+	int oldval = 0, newval, ret;
+
+	switch (op) {
+	case FUTEX_OP_SET:
+		__futex_atomic_op("lr %2,%5\n",
+				  ret, oldval, newval, uaddr, oparg);
+		break;
+	case FUTEX_OP_ADD:
+		__futex_atomic_op("lr %2,%1\nar %2,%5\n",
+				  ret, oldval, newval, uaddr, oparg);
+		break;
+	case FUTEX_OP_OR:
+		__futex_atomic_op("lr %2,%1\nor %2,%5\n",
+				  ret, oldval, newval, uaddr, oparg);
+		break;
+	case FUTEX_OP_ANDN:
+		__futex_atomic_op("lr %2,%1\nnr %2,%5\n",
+				  ret, oldval, newval, uaddr, oparg);
+		break;
+	case FUTEX_OP_XOR:
+		__futex_atomic_op("lr %2,%1\nxr %2,%5\n",
+				  ret, oldval, newval, uaddr, oparg);
+		break;
+	default:
+		ret = -ENOSYS;
+	}
+	if (ret == 0)
+		*old = oldval;
+	return ret;
+}
+
+int futex_atomic_op_pt(int op, u32 __user *uaddr, int oparg, int *old)
+{
+	int ret;
+
+	if (segment_eq(get_fs(), KERNEL_DS))
+		return __futex_atomic_op_pt(op, uaddr, oparg, old);
+	spin_lock(&current->mm->page_table_lock);
+	uaddr = (u32 __force __user *)
+		__dat_user_addr((__force unsigned long) uaddr, 1);
+	if (!uaddr) {
+		spin_unlock(&current->mm->page_table_lock);
+		return -EFAULT;
+	}
+	get_page(virt_to_page(uaddr));
+	spin_unlock(&current->mm->page_table_lock);
+	ret = __futex_atomic_op_pt(op, uaddr, oparg, old);
+	put_page(virt_to_page(uaddr));
+	return ret;
+}
+
+static int __futex_atomic_cmpxchg_pt(u32 *uval, u32 __user *uaddr,
+				     u32 oldval, u32 newval)
+{
+	int ret;
+
+	asm volatile("0: cs   %1,%4,0(%5)\n"
+		     "1: la   %0,0\n"
+		     "2:\n"
+		     EX_TABLE(0b,2b) EX_TABLE(1b,2b)
+		     : "=d" (ret), "+d" (oldval), "=m" (*uaddr)
+		     : "0" (-EFAULT), "d" (newval), "a" (uaddr), "m" (*uaddr)
+		     : "cc", "memory" );
+	*uval = oldval;
+	return ret;
+}
+
+int futex_atomic_cmpxchg_pt(u32 *uval, u32 __user *uaddr,
+			    u32 oldval, u32 newval)
+{
+	int ret;
+
+	if (segment_eq(get_fs(), KERNEL_DS))
+		return __futex_atomic_cmpxchg_pt(uval, uaddr, oldval, newval);
+	spin_lock(&current->mm->page_table_lock);
+	uaddr = (u32 __force __user *)
+		__dat_user_addr((__force unsigned long) uaddr, 1);
+	if (!uaddr) {
+		spin_unlock(&current->mm->page_table_lock);
+		return -EFAULT;
+	}
+	get_page(virt_to_page(uaddr));
+	spin_unlock(&current->mm->page_table_lock);
+	ret = __futex_atomic_cmpxchg_pt(uval, uaddr, oldval, newval);
+	put_page(virt_to_page(uaddr));
+	return ret;
+}
+
+struct uaccess_ops uaccess_pt = {
+	.copy_from_user		= copy_from_user_pt,
+	.copy_from_user_small	= copy_from_user_pt,
+	.copy_to_user		= copy_to_user_pt,
+	.copy_to_user_small	= copy_to_user_pt,
+	.copy_in_user		= copy_in_user_pt,
+	.clear_user		= clear_user_pt,
+	.strnlen_user		= strnlen_user_pt,
+	.strncpy_from_user	= strncpy_from_user_pt,
+	.futex_atomic_op	= futex_atomic_op_pt,
+	.futex_atomic_cmpxchg	= futex_atomic_cmpxchg_pt,
+};
diff --git a/arch/s390/lib/uaccess_std.c b/arch/s390/lib/uaccess_std.c
new file mode 100644
index 00000000..4a75d475
--- /dev/null
+++ b/arch/s390/lib/uaccess_std.c
@@ -0,0 +1,305 @@
+/*
+ *  Standard user space access functions based on mvcp/mvcs and doing
+ *  interesting things in the secondary space mode.
+ *
+ *    Copyright IBM Corp. 2006
+ *    Author(s): Martin Schwidefsky (schwidefsky@de.ibm.com),
+ *		 Gerald Schaefer (gerald.schaefer@de.ibm.com)
+ */
+
+#include <linux/errno.h>
+#include <linux/mm.h>
+#include <linux/uaccess.h>
+#include <asm/futex.h>
+#include "uaccess.h"
+
+#ifndef CONFIG_64BIT
+#define AHI	"ahi"
+#define ALR	"alr"
+#define CLR	"clr"
+#define LHI	"lhi"
+#define SLR	"slr"
+#else
+#define AHI	"aghi"
+#define ALR	"algr"
+#define CLR	"clgr"
+#define LHI	"lghi"
+#define SLR	"slgr"
+#endif
+
+size_t copy_from_user_std(size_t size, const void __user *ptr, void *x)
+{
+	unsigned long tmp1, tmp2;
+
+	tmp1 = -256UL;
+	asm volatile(
+		"0: mvcp  0(%0,%2),0(%1),%3\n"
+		"10:jz    8f\n"
+		"1:"ALR"  %0,%3\n"
+		"   la    %1,256(%1)\n"
+		"   la    %2,256(%2)\n"
+		"2: mvcp  0(%0,%2),0(%1),%3\n"
+		"11:jnz   1b\n"
+		"   j     8f\n"
+		"3: la    %4,255(%1)\n"	/* %4 = ptr + 255 */
+		"  "LHI"  %3,-4096\n"
+		"   nr    %4,%3\n"	/* %4 = (ptr + 255) & -4096 */
+		"  "SLR"  %4,%1\n"
+		"  "CLR"  %0,%4\n"	/* copy crosses next page boundary? */
+		"   jnh   5f\n"
+		"4: mvcp  0(%4,%2),0(%1),%3\n"
+		"12:"SLR"  %0,%4\n"
+		"  "ALR"  %2,%4\n"
+		"5:"LHI"  %4,-1\n"
+		"  "ALR"  %4,%0\n"	/* copy remaining size, subtract 1 */
+		"   bras  %3,7f\n"	/* memset loop */
+		"   xc    0(1,%2),0(%2)\n"
+		"6: xc    0(256,%2),0(%2)\n"
+		"   la    %2,256(%2)\n"
+		"7:"AHI"  %4,-256\n"
+		"   jnm   6b\n"
+		"   ex    %4,0(%3)\n"
+		"   j     9f\n"
+		"8:"SLR"  %0,%0\n"
+		"9: \n"
+		EX_TABLE(0b,3b) EX_TABLE(2b,3b) EX_TABLE(4b,5b)
+		EX_TABLE(10b,3b) EX_TABLE(11b,3b) EX_TABLE(12b,5b)
+		: "+a" (size), "+a" (ptr), "+a" (x), "+a" (tmp1), "=a" (tmp2)
+		: : "cc", "memory");
+	return size;
+}
+
+static size_t copy_from_user_std_check(size_t size, const void __user *ptr,
+				       void *x)
+{
+	if (size <= 1024)
+		return copy_from_user_std(size, ptr, x);
+	return copy_from_user_pt(size, ptr, x);
+}
+
+size_t copy_to_user_std(size_t size, void __user *ptr, const void *x)
+{
+	unsigned long tmp1, tmp2;
+
+	tmp1 = -256UL;
+	asm volatile(
+		"0: mvcs  0(%0,%1),0(%2),%3\n"
+		"7: jz    5f\n"
+		"1:"ALR"  %0,%3\n"
+		"   la    %1,256(%1)\n"
+		"   la    %2,256(%2)\n"
+		"2: mvcs  0(%0,%1),0(%2),%3\n"
+		"8: jnz   1b\n"
+		"   j     5f\n"
+		"3: la    %4,255(%1)\n" /* %4 = ptr + 255 */
+		"  "LHI"  %3,-4096\n"
+		"   nr    %4,%3\n"	/* %4 = (ptr + 255) & -4096 */
+		"  "SLR"  %4,%1\n"
+		"  "CLR"  %0,%4\n"	/* copy crosses next page boundary? */
+		"   jnh   6f\n"
+		"4: mvcs  0(%4,%1),0(%2),%3\n"
+		"9:"SLR"  %0,%4\n"
+		"   j     6f\n"
+		"5:"SLR"  %0,%0\n"
+		"6: \n"
+		EX_TABLE(0b,3b) EX_TABLE(2b,3b) EX_TABLE(4b,6b)
+		EX_TABLE(7b,3b) EX_TABLE(8b,3b) EX_TABLE(9b,6b)
+		: "+a" (size), "+a" (ptr), "+a" (x), "+a" (tmp1), "=a" (tmp2)
+		: : "cc", "memory");
+	return size;
+}
+
+static size_t copy_to_user_std_check(size_t size, void __user *ptr,
+				     const void *x)
+{
+	if (size <= 1024)
+		return copy_to_user_std(size, ptr, x);
+	return copy_to_user_pt(size, ptr, x);
+}
+
+static size_t copy_in_user_std(size_t size, void __user *to,
+			       const void __user *from)
+{
+	unsigned long tmp1;
+
+	asm volatile(
+		"   sacf  256\n"
+		"  "AHI"  %0,-1\n"
+		"   jo    5f\n"
+		"   bras  %3,3f\n"
+		"0:"AHI"  %0,257\n"
+		"1: mvc   0(1,%1),0(%2)\n"
+		"   la    %1,1(%1)\n"
+		"   la    %2,1(%2)\n"
+		"  "AHI"  %0,-1\n"
+		"   jnz   1b\n"
+		"   j     5f\n"
+		"2: mvc   0(256,%1),0(%2)\n"
+		"   la    %1,256(%1)\n"
+		"   la    %2,256(%2)\n"
+		"3:"AHI"  %0,-256\n"
+		"   jnm   2b\n"
+		"4: ex    %0,1b-0b(%3)\n"
+		"5: "SLR"  %0,%0\n"
+		"6: sacf  0\n"
+		EX_TABLE(1b,6b) EX_TABLE(2b,0b) EX_TABLE(4b,0b)
+		: "+a" (size), "+a" (to), "+a" (from), "=a" (tmp1)
+		: : "cc", "memory");
+	return size;
+}
+
+static size_t clear_user_std(size_t size, void __user *to)
+{
+	unsigned long tmp1, tmp2;
+
+	asm volatile(
+		"   sacf  256\n"
+		"  "AHI"  %0,-1\n"
+		"   jo    5f\n"
+		"   bras  %3,3f\n"
+		"   xc    0(1,%1),0(%1)\n"
+		"0:"AHI"  %0,257\n"
+		"   la    %2,255(%1)\n" /* %2 = ptr + 255 */
+		"   srl   %2,12\n"
+		"   sll   %2,12\n"	/* %2 = (ptr + 255) & -4096 */
+		"  "SLR"  %2,%1\n"
+		"  "CLR"  %0,%2\n"	/* clear crosses next page boundary? */
+		"   jnh   5f\n"
+		"  "AHI"  %2,-1\n"
+		"1: ex    %2,0(%3)\n"
+		"  "AHI"  %2,1\n"
+		"  "SLR"  %0,%2\n"
+		"   j     5f\n"
+		"2: xc    0(256,%1),0(%1)\n"
+		"   la    %1,256(%1)\n"
+		"3:"AHI"  %0,-256\n"
+		"   jnm   2b\n"
+		"4: ex    %0,0(%3)\n"
+		"5: "SLR"  %0,%0\n"
+		"6: sacf  0\n"
+		EX_TABLE(1b,6b) EX_TABLE(2b,0b) EX_TABLE(4b,0b)
+		: "+a" (size), "+a" (to), "=a" (tmp1), "=a" (tmp2)
+		: : "cc", "memory");
+	return size;
+}
+
+size_t strnlen_user_std(size_t size, const char __user *src)
+{
+	register unsigned long reg0 asm("0") = 0UL;
+	unsigned long tmp1, tmp2;
+
+	if (unlikely(!size))
+		return 0;
+	asm volatile(
+		"   la    %2,0(%1)\n"
+		"   la    %3,0(%0,%1)\n"
+		"  "SLR"  %0,%0\n"
+		"   sacf  256\n"
+		"0: srst  %3,%2\n"
+		"   jo    0b\n"
+		"   la    %0,1(%3)\n"	/* strnlen_user results includes \0 */
+		"  "SLR"  %0,%1\n"
+		"1: sacf  0\n"
+		EX_TABLE(0b,1b)
+		: "+a" (size), "+a" (src), "=a" (tmp1), "=a" (tmp2)
+		: "d" (reg0) : "cc", "memory");
+	return size;
+}
+
+size_t strncpy_from_user_std(size_t count, const char __user *src, char *dst)
+{
+	size_t done, len, offset, len_str;
+
+	if (unlikely(!count))
+		return 0;
+	done = 0;
+	do {
+		offset = (size_t)src & ~PAGE_MASK;
+		len = min(count - done, PAGE_SIZE - offset);
+		if (copy_from_user_std(len, src, dst))
+			return -EFAULT;
+		len_str = strnlen(dst, len);
+		done += len_str;
+		src += len_str;
+		dst += len_str;
+	} while ((len_str == len) && (done < count));
+	return done;
+}
+
+#define __futex_atomic_op(insn, ret, oldval, newval, uaddr, oparg)	\
+	asm volatile(							\
+		"   sacf  256\n"					\
+		"0: l     %1,0(%6)\n"					\
+		"1:"insn						\
+		"2: cs    %1,%2,0(%6)\n"				\
+		"3: jl    1b\n"						\
+		"   lhi   %0,0\n"					\
+		"4: sacf  0\n"						\
+		EX_TABLE(0b,4b) EX_TABLE(2b,4b) EX_TABLE(3b,4b)		\
+		: "=d" (ret), "=&d" (oldval), "=&d" (newval),		\
+		  "=m" (*uaddr)						\
+		: "0" (-EFAULT), "d" (oparg), "a" (uaddr),		\
+		  "m" (*uaddr) : "cc");
+
+int futex_atomic_op_std(int op, u32 __user *uaddr, int oparg, int *old)
+{
+	int oldval = 0, newval, ret;
+
+	switch (op) {
+	case FUTEX_OP_SET:
+		__futex_atomic_op("lr %2,%5\n",
+				  ret, oldval, newval, uaddr, oparg);
+		break;
+	case FUTEX_OP_ADD:
+		__futex_atomic_op("lr %2,%1\nar %2,%5\n",
+				  ret, oldval, newval, uaddr, oparg);
+		break;
+	case FUTEX_OP_OR:
+		__futex_atomic_op("lr %2,%1\nor %2,%5\n",
+				  ret, oldval, newval, uaddr, oparg);
+		break;
+	case FUTEX_OP_ANDN:
+		__futex_atomic_op("lr %2,%1\nnr %2,%5\n",
+				  ret, oldval, newval, uaddr, oparg);
+		break;
+	case FUTEX_OP_XOR:
+		__futex_atomic_op("lr %2,%1\nxr %2,%5\n",
+				  ret, oldval, newval, uaddr, oparg);
+		break;
+	default:
+		ret = -ENOSYS;
+	}
+	*old = oldval;
+	return ret;
+}
+
+int futex_atomic_cmpxchg_std(u32 *uval, u32 __user *uaddr,
+			     u32 oldval, u32 newval)
+{
+	int ret;
+
+	asm volatile(
+		"   sacf 256\n"
+		"0: cs   %1,%4,0(%5)\n"
+		"1: la   %0,0\n"
+		"2: sacf 0\n"
+		EX_TABLE(0b,2b) EX_TABLE(1b,2b)
+		: "=d" (ret), "+d" (oldval), "=m" (*uaddr)
+		: "0" (-EFAULT), "d" (newval), "a" (uaddr), "m" (*uaddr)
+		: "cc", "memory" );
+	*uval = oldval;
+	return ret;
+}
+
+struct uaccess_ops uaccess_std = {
+	.copy_from_user = copy_from_user_std_check,
+	.copy_from_user_small = copy_from_user_std,
+	.copy_to_user = copy_to_user_std_check,
+	.copy_to_user_small = copy_to_user_std,
+	.copy_in_user = copy_in_user_std,
+	.clear_user = clear_user_std,
+	.strnlen_user = strnlen_user_std,
+	.strncpy_from_user = strncpy_from_user_std,
+	.futex_atomic_op = futex_atomic_op_std,
+	.futex_atomic_cmpxchg = futex_atomic_cmpxchg_std,
+};
diff --git a/arch/s390/lib/ucmpdi2.c b/arch/s390/lib/ucmpdi2.c
new file mode 100644
index 00000000..3e05ff53
--- /dev/null
+++ b/arch/s390/lib/ucmpdi2.c
@@ -0,0 +1,26 @@
+#include <linux/module.h>
+
+union ull_union {
+	unsigned long long ull;
+	struct {
+		unsigned int high;
+		unsigned int low;
+	} ui;
+};
+
+int __ucmpdi2(unsigned long long a, unsigned long long b)
+{
+	union ull_union au = {.ull = a};
+	union ull_union bu = {.ull = b};
+
+	if (au.ui.high < bu.ui.high)
+		return 0;
+	else if (au.ui.high > bu.ui.high)
+		return 2;
+	if (au.ui.low < bu.ui.low)
+		return 0;
+	else if (au.ui.low > bu.ui.low)
+		return 2;
+	return 1;
+}
+EXPORT_SYMBOL(__ucmpdi2);