Merging upstream version 6.9.2.

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

47 files changed, 1646 insertions, 565 deletions

diff --git a/arch/s390/include/asm/access-regs.h b/arch/s390/include/asm/access-regs.h
new file mode 100644
index 000000000..1a6412d9f
--- /dev/null
+++ b/arch/s390/include/asm/access-regs.h
@@ -0,0 +1,38 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Copyright IBM Corp. 1999, 2024
+ */
+
+#ifndef __ASM_S390_ACCESS_REGS_H
+#define __ASM_S390_ACCESS_REGS_H
+
+#include <linux/instrumented.h>
+#include <asm/sigcontext.h>
+
+struct access_regs {
+	unsigned int regs[NUM_ACRS];
+};
+
+static inline void save_access_regs(unsigned int *acrs)
+{
+	struct access_regs *regs = (struct access_regs *)acrs;
+
+	instrument_write(regs, sizeof(*regs));
+	asm volatile("stamy	0,15,%[regs]"
+		     : [regs] "=QS" (*regs)
+		     :
+		     : "memory");
+}
+
+static inline void restore_access_regs(unsigned int *acrs)
+{
+	struct access_regs *regs = (struct access_regs *)acrs;
+
+	instrument_read(regs, sizeof(*regs));
+	asm volatile("lamy	0,15,%[regs]"
+		     :
+		     : [regs] "QS" (*regs)
+		     : "memory");
+}
+
+#endif /* __ASM_S390_ACCESS_REGS_H */
diff --git a/arch/s390/include/asm/appldata.h b/arch/s390/include/asm/appldata.h
index f2240392c..a92ebbc7a 100644
--- a/arch/s390/include/asm/appldata.h
+++ b/arch/s390/include/asm/appldata.h
@@ -54,13 +54,13 @@ static inline int appldata_asm(struct appldata_parameter_list *parm_list,
 	parm_list->function = fn;
 	parm_list->parlist_length = sizeof(*parm_list);
 	parm_list->buffer_length = length;
-	parm_list->product_id_addr = (unsigned long) id;
+	parm_list->product_id_addr = virt_to_phys(id);
 	parm_list->buffer_addr = virt_to_phys(buffer);
 	diag_stat_inc(DIAG_STAT_X0DC);
 	asm volatile(
 		"	diag	%1,%0,0xdc"
 		: "=d" (ry)
-		: "d" (parm_list), "m" (*parm_list), "m" (*id)
+		: "d" (virt_to_phys(parm_list)), "m" (*parm_list), "m" (*id)
 		: "cc");
 	return ry;
 }
diff --git a/arch/s390/include/asm/asm-prototypes.h b/arch/s390/include/asm/asm-prototypes.h
index a873e873e..56096ae26 100644
--- a/arch/s390/include/asm/asm-prototypes.h
+++ b/arch/s390/include/asm/asm-prototypes.h
@@ -3,7 +3,7 @@
 
 #include <linux/kvm_host.h>
 #include <linux/ftrace.h>
-#include <asm/fpu/api.h>
+#include <asm/fpu.h>
 #include <asm-generic/asm-prototypes.h>
 
 __int128_t __ashlti3(__int128_t a, int b);
diff --git a/arch/s390/include/asm/atomic.h b/arch/s390/include/asm/atomic.h
index 7138d189c..0c4cad7d5 100644
--- a/arch/s390/include/asm/atomic.h
+++ b/arch/s390/include/asm/atomic.h
@@ -15,31 +15,31 @@
 #include <asm/barrier.h>
 #include <asm/cmpxchg.h>
 
-static inline int arch_atomic_read(const atomic_t *v)
+static __always_inline int arch_atomic_read(const atomic_t *v)
 {
 	return __atomic_read(v);
 }
 #define arch_atomic_read arch_atomic_read
 
-static inline void arch_atomic_set(atomic_t *v, int i)
+static __always_inline void arch_atomic_set(atomic_t *v, int i)
 {
 	__atomic_set(v, i);
 }
 #define arch_atomic_set arch_atomic_set
 
-static inline int arch_atomic_add_return(int i, atomic_t *v)
+static __always_inline int arch_atomic_add_return(int i, atomic_t *v)
 {
 	return __atomic_add_barrier(i, &v->counter) + i;
 }
 #define arch_atomic_add_return arch_atomic_add_return
 
-static inline int arch_atomic_fetch_add(int i, atomic_t *v)
+static __always_inline int arch_atomic_fetch_add(int i, atomic_t *v)
 {
 	return __atomic_add_barrier(i, &v->counter);
 }
 #define arch_atomic_fetch_add arch_atomic_fetch_add
 
-static inline void arch_atomic_add(int i, atomic_t *v)
+static __always_inline void arch_atomic_add(int i, atomic_t *v)
 {
 	__atomic_add(i, &v->counter);
 }
@@ -50,11 +50,11 @@ static inline void arch_atomic_add(int i, atomic_t *v)
 #define arch_atomic_fetch_sub(_i, _v)	arch_atomic_fetch_add(-(int)(_i), _v)
 
 #define ATOMIC_OPS(op)							\
-static inline void arch_atomic_##op(int i, atomic_t *v)			\
+static __always_inline void arch_atomic_##op(int i, atomic_t *v)	\
 {									\
 	__atomic_##op(i, &v->counter);					\
 }									\
-static inline int arch_atomic_fetch_##op(int i, atomic_t *v)		\
+static __always_inline int arch_atomic_fetch_##op(int i, atomic_t *v)	\
 {									\
 	return __atomic_##op##_barrier(i, &v->counter);			\
 }
@@ -74,7 +74,7 @@ ATOMIC_OPS(xor)
 
 #define arch_atomic_xchg(v, new)	(arch_xchg(&((v)->counter), new))
 
-static inline int arch_atomic_cmpxchg(atomic_t *v, int old, int new)
+static __always_inline int arch_atomic_cmpxchg(atomic_t *v, int old, int new)
 {
 	return __atomic_cmpxchg(&v->counter, old, new);
 }
@@ -82,31 +82,31 @@ static inline int arch_atomic_cmpxchg(atomic_t *v, int old, int new)
 
 #define ATOMIC64_INIT(i)  { (i) }
 
-static inline s64 arch_atomic64_read(const atomic64_t *v)
+static __always_inline s64 arch_atomic64_read(const atomic64_t *v)
 {
 	return __atomic64_read(v);
 }
 #define arch_atomic64_read arch_atomic64_read
 
-static inline void arch_atomic64_set(atomic64_t *v, s64 i)
+static __always_inline void arch_atomic64_set(atomic64_t *v, s64 i)
 {
 	__atomic64_set(v, i);
 }
 #define arch_atomic64_set arch_atomic64_set
 
-static inline s64 arch_atomic64_add_return(s64 i, atomic64_t *v)
+static __always_inline s64 arch_atomic64_add_return(s64 i, atomic64_t *v)
 {
 	return __atomic64_add_barrier(i, (long *)&v->counter) + i;
 }
 #define arch_atomic64_add_return arch_atomic64_add_return
 
-static inline s64 arch_atomic64_fetch_add(s64 i, atomic64_t *v)
+static __always_inline s64 arch_atomic64_fetch_add(s64 i, atomic64_t *v)
 {
 	return __atomic64_add_barrier(i, (long *)&v->counter);
 }
 #define arch_atomic64_fetch_add arch_atomic64_fetch_add
 
-static inline void arch_atomic64_add(s64 i, atomic64_t *v)
+static __always_inline void arch_atomic64_add(s64 i, atomic64_t *v)
 {
 	__atomic64_add(i, (long *)&v->counter);
 }
@@ -114,20 +114,20 @@ static inline void arch_atomic64_add(s64 i, atomic64_t *v)
 
 #define arch_atomic64_xchg(v, new)	(arch_xchg(&((v)->counter), new))
 
-static inline s64 arch_atomic64_cmpxchg(atomic64_t *v, s64 old, s64 new)
+static __always_inline s64 arch_atomic64_cmpxchg(atomic64_t *v, s64 old, s64 new)
 {
 	return __atomic64_cmpxchg((long *)&v->counter, old, new);
 }
 #define arch_atomic64_cmpxchg arch_atomic64_cmpxchg
 
-#define ATOMIC64_OPS(op)						\
-static inline void arch_atomic64_##op(s64 i, atomic64_t *v)		\
-{									\
-	__atomic64_##op(i, (long *)&v->counter);			\
-}									\
-static inline long arch_atomic64_fetch_##op(s64 i, atomic64_t *v)	\
-{									\
-	return __atomic64_##op##_barrier(i, (long *)&v->counter);	\
+#define ATOMIC64_OPS(op)							\
+static __always_inline void arch_atomic64_##op(s64 i, atomic64_t *v)		\
+{										\
+	__atomic64_##op(i, (long *)&v->counter);				\
+}										\
+static __always_inline long arch_atomic64_fetch_##op(s64 i, atomic64_t *v)	\
+{										\
+	return __atomic64_##op##_barrier(i, (long *)&v->counter);		\
 }
 
 ATOMIC64_OPS(and)
diff --git a/arch/s390/include/asm/atomic_ops.h b/arch/s390/include/asm/atomic_ops.h
index 50510e08b..7fa5f96a5 100644
--- a/arch/s390/include/asm/atomic_ops.h
+++ b/arch/s390/include/asm/atomic_ops.h
@@ -8,7 +8,7 @@
 #ifndef __ARCH_S390_ATOMIC_OPS__
 #define __ARCH_S390_ATOMIC_OPS__
 
-static inline int __atomic_read(const atomic_t *v)
+static __always_inline int __atomic_read(const atomic_t *v)
 {
 	int c;
 
@@ -18,14 +18,14 @@ static inline int __atomic_read(const atomic_t *v)
 	return c;
 }
 
-static inline void __atomic_set(atomic_t *v, int i)
+static __always_inline void __atomic_set(atomic_t *v, int i)
 {
 	asm volatile(
 		"	st	%1,%0\n"
 		: "=R" (v->counter) : "d" (i));
 }
 
-static inline s64 __atomic64_read(const atomic64_t *v)
+static __always_inline s64 __atomic64_read(const atomic64_t *v)
 {
 	s64 c;
 
@@ -35,7 +35,7 @@ static inline s64 __atomic64_read(const atomic64_t *v)
 	return c;
 }
 
-static inline void __atomic64_set(atomic64_t *v, s64 i)
+static __always_inline void __atomic64_set(atomic64_t *v, s64 i)
 {
 	asm volatile(
 		"	stg	%1,%0\n"
@@ -45,7 +45,7 @@ static inline void __atomic64_set(atomic64_t *v, s64 i)
 #ifdef CONFIG_HAVE_MARCH_Z196_FEATURES
 
 #define __ATOMIC_OP(op_name, op_type, op_string, op_barrier)		\
-static inline op_type op_name(op_type val, op_type *ptr)		\
+static __always_inline op_type op_name(op_type val, op_type *ptr)	\
 {									\
 	op_type old;							\
 									\
@@ -96,7 +96,7 @@ __ATOMIC_CONST_OPS(__atomic64_add_const, long, "agsi")
 #else /* CONFIG_HAVE_MARCH_Z196_FEATURES */
 
 #define __ATOMIC_OP(op_name, op_string)					\
-static inline int op_name(int val, int *ptr)				\
+static __always_inline int op_name(int val, int *ptr)			\
 {									\
 	int old, new;							\
 									\
@@ -122,7 +122,7 @@ __ATOMIC_OPS(__atomic_xor, "xr")
 #undef __ATOMIC_OPS
 
 #define __ATOMIC64_OP(op_name, op_string)				\
-static inline long op_name(long val, long *ptr)				\
+static __always_inline long op_name(long val, long *ptr)		\
 {									\
 	long old, new;							\
 									\
@@ -154,7 +154,7 @@ __ATOMIC64_OPS(__atomic64_xor, "xgr")
 
 #endif /* CONFIG_HAVE_MARCH_Z196_FEATURES */
 
-static inline int __atomic_cmpxchg(int *ptr, int old, int new)
+static __always_inline int __atomic_cmpxchg(int *ptr, int old, int new)
 {
 	asm volatile(
 		"	cs	%[old],%[new],%[ptr]"
@@ -164,7 +164,7 @@ static inline int __atomic_cmpxchg(int *ptr, int old, int new)
 	return old;
 }
 
-static inline bool __atomic_cmpxchg_bool(int *ptr, int old, int new)
+static __always_inline bool __atomic_cmpxchg_bool(int *ptr, int old, int new)
 {
 	int old_expected = old;
 
@@ -176,7 +176,7 @@ static inline bool __atomic_cmpxchg_bool(int *ptr, int old, int new)
 	return old == old_expected;
 }
 
-static inline long __atomic64_cmpxchg(long *ptr, long old, long new)
+static __always_inline long __atomic64_cmpxchg(long *ptr, long old, long new)
 {
 	asm volatile(
 		"	csg	%[old],%[new],%[ptr]"
@@ -186,7 +186,7 @@ static inline long __atomic64_cmpxchg(long *ptr, long old, long new)
 	return old;
 }
 
-static inline bool __atomic64_cmpxchg_bool(long *ptr, long old, long new)
+static __always_inline bool __atomic64_cmpxchg_bool(long *ptr, long old, long new)
 {
 	long old_expected = old;
 
diff --git a/arch/s390/include/asm/bug.h b/arch/s390/include/asm/bug.h
index aebe1e22c..c500d45fb 100644
--- a/arch/s390/include/asm/bug.h
+++ b/arch/s390/include/asm/bug.h
@@ -14,7 +14,7 @@
 		".section .rodata.str,\"aMS\",@progbits,1\n"	\
 		"1:	.asciz	\""__FILE__"\"\n"		\
 		".previous\n"					\
-		".section __bug_table,\"awM\",@progbits,%2\n"	\
+		".section __bug_table,\"aw\"\n"			\
 		"2:	.long	0b-.\n"				\
 		"	.long	1b-.\n"				\
 		"	.short	%0,%1\n"			\
@@ -30,7 +30,7 @@
 #define __EMIT_BUG(x) do {					\
 	asm_inline volatile(					\
 		"0:	mc	0,0\n"				\
-		".section __bug_table,\"awM\",@progbits,%1\n"	\
+		".section __bug_table,\"aw\"\n"			\
 		"1:	.long	0b-.\n"				\
 		"	.short	%0\n"				\
 		"	.org	1b+%1\n"			\
diff --git a/arch/s390/include/asm/ccwdev.h b/arch/s390/include/asm/ccwdev.h
index 91d261751..436365ff6 100644
--- a/arch/s390/include/asm/ccwdev.h
+++ b/arch/s390/include/asm/ccwdev.h
@@ -217,7 +217,8 @@ extern void ccw_device_destroy_console(struct ccw_device *);
 extern int ccw_device_enable_console(struct ccw_device *);
 extern void ccw_device_wait_idle(struct ccw_device *);
 
-extern void *ccw_device_dma_zalloc(struct ccw_device *cdev, size_t size);
+extern void *ccw_device_dma_zalloc(struct ccw_device *cdev, size_t size,
+				   dma32_t *dma_handle);
 extern void ccw_device_dma_free(struct ccw_device *cdev,
 				void *cpu_addr, size_t size);
 
diff --git a/arch/s390/include/asm/checksum.h b/arch/s390/include/asm/checksum.h
index 69837eec2..b89159591 100644
--- a/arch/s390/include/asm/checksum.h
+++ b/arch/s390/include/asm/checksum.h
@@ -12,36 +12,29 @@
 #ifndef _S390_CHECKSUM_H
 #define _S390_CHECKSUM_H
 
-#include <linux/kasan-checks.h>
+#include <linux/instrumented.h>
 #include <linux/in6.h>
 
-/*
- * Computes the checksum of a memory block at buff, length len,
- * and adds in "sum" (32-bit).
- *
- * Returns a 32-bit number suitable for feeding into itself
- * or csum_tcpudp_magic.
- *
- * This function must be called with even lengths, except
- * for the last fragment, which may be odd.
- *
- * It's best to have buff aligned on a 32-bit boundary.
- */
-static inline __wsum csum_partial(const void *buff, int len, __wsum sum)
+static inline __wsum cksm(const void *buff, int len, __wsum sum)
 {
 	union register_pair rp = {
-		.even = (unsigned long) buff,
-		.odd = (unsigned long) len,
+		.even = (unsigned long)buff,
+		.odd = (unsigned long)len,
 	};
 
-	kasan_check_read(buff, len);
-	asm volatile(
+	instrument_read(buff, len);
+	asm volatile("\n"
 		"0:	cksm	%[sum],%[rp]\n"
 		"	jo	0b\n"
 		: [sum] "+&d" (sum), [rp] "+&d" (rp.pair) : : "cc", "memory");
 	return sum;
 }
 
+__wsum csum_partial(const void *buff, int len, __wsum sum);
+
+#define _HAVE_ARCH_CSUM_AND_COPY
+__wsum csum_partial_copy_nocheck(const void *src, void *dst, int len);
+
 /*
  * Fold a partial checksum without adding pseudo headers.
  */
diff --git a/arch/s390/include/asm/cio.h b/arch/s390/include/asm/cio.h
index 1c4f585dd..b6b619f34 100644
--- a/arch/s390/include/asm/cio.h
+++ b/arch/s390/include/asm/cio.h
@@ -7,6 +7,7 @@
 
 #include <linux/bitops.h>
 #include <linux/genalloc.h>
+#include <asm/dma-types.h>
 #include <asm/types.h>
 #include <asm/tpi.h>
 
@@ -32,7 +33,7 @@ struct ccw1 {
 	__u8  cmd_code;
 	__u8  flags;
 	__u16 count;
-	__u32 cda;
+	dma32_t cda;
 } __attribute__ ((packed,aligned(8)));
 
 /**
@@ -152,8 +153,8 @@ struct sublog {
 struct esw0 {
 	struct sublog sublog;
 	struct erw erw;
-	__u32  faddr[2];
-	__u32  saddr;
+	dma32_t faddr[2];
+	dma32_t saddr;
 } __attribute__ ((packed));
 
 /**
@@ -364,6 +365,8 @@ extern struct device *cio_get_dma_css_dev(void);
 
 void *cio_gp_dma_zalloc(struct gen_pool *gp_dma, struct device *dma_dev,
 			size_t size);
+void *__cio_gp_dma_zalloc(struct gen_pool *gp_dma, struct device *dma_dev,
+			  size_t size, dma32_t *dma_handle);
 void cio_gp_dma_free(struct gen_pool *gp_dma, void *cpu_addr, size_t size);
 void cio_gp_dma_destroy(struct gen_pool *gp_dma, struct device *dma_dev);
 struct gen_pool *cio_gp_dma_create(struct device *dma_dev, int nr_pages);
diff --git a/arch/s390/include/asm/diag.h b/arch/s390/include/asm/diag.h
index bed804137..20b942201 100644
--- a/arch/s390/include/asm/diag.h
+++ b/arch/s390/include/asm/diag.h
@@ -44,6 +44,13 @@ enum diag_stat_enum {
 void diag_stat_inc(enum diag_stat_enum nr);
 void diag_stat_inc_norecursion(enum diag_stat_enum nr);
 
+struct hypfs_diag0c_entry;
+
+/*
+ * Diagnose 0c: Pseudo Timer
+ */
+void diag0c(struct hypfs_diag0c_entry *data);
+
 /*
  * Diagnose 10: Release page range
  */
@@ -331,10 +338,10 @@ struct hypfs_diag0c_entry;
  */
 struct diag_ops {
 	int (*diag210)(struct diag210 *addr);
-	int (*diag26c)(void *req, void *resp, enum diag26c_sc subcode);
+	int (*diag26c)(unsigned long rx, unsigned long rx1, enum diag26c_sc subcode);
 	int (*diag14)(unsigned long rx, unsigned long ry1, unsigned long subcode);
 	int (*diag8c)(struct diag8c *addr, struct ccw_dev_id *devno, size_t len);
-	void (*diag0c)(struct hypfs_diag0c_entry *entry);
+	void (*diag0c)(unsigned long rx);
 	void (*diag308_reset)(void);
 };
 
@@ -342,9 +349,9 @@ extern struct diag_ops diag_amode31_ops;
 extern struct diag210 *__diag210_tmp_amode31;
 
 int _diag210_amode31(struct diag210 *addr);
-int _diag26c_amode31(void *req, void *resp, enum diag26c_sc subcode);
+int _diag26c_amode31(unsigned long rx, unsigned long rx1, enum diag26c_sc subcode);
 int _diag14_amode31(unsigned long rx, unsigned long ry1, unsigned long subcode);
-void _diag0c_amode31(struct hypfs_diag0c_entry *entry);
+void _diag0c_amode31(unsigned long rx);
 void _diag308_reset_amode31(void);
 int _diag8c_amode31(struct diag8c *addr, struct ccw_dev_id *devno, size_t len);
 
diff --git a/arch/s390/include/asm/dma-types.h b/arch/s390/include/asm/dma-types.h
new file mode 100644
index 000000000..5c5734e69
--- /dev/null
+++ b/arch/s390/include/asm/dma-types.h
@@ -0,0 +1,103 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+
+#ifndef _ASM_S390_DMA_TYPES_H_
+#define _ASM_S390_DMA_TYPES_H_
+
+#include <linux/types.h>
+#include <linux/io.h>
+
+/*
+ * typedef dma32_t
+ * Contains a 31 bit absolute address to a DMA capable piece of storage.
+ *
+ * For CIO, DMA addresses are always absolute addresses. These addresses tend
+ * to be used in architectured memory blocks (like ORB, IDAW, MIDAW). Under
+ * certain circumstances 31 bit wide addresses must be used because the
+ * address must fit in 31 bits.
+ *
+ * This type is to be used when such fields can be modelled as 32 bit wide.
+ */
+typedef u32 __bitwise dma32_t;
+
+/*
+ * typedef dma64_t
+ * Contains a 64 bit absolute address to a DMA capable piece of storage.
+ *
+ * For CIO, DMA addresses are always absolute addresses. These addresses tend
+ * to be used in architectured memory blocks (like ORB, IDAW, MIDAW).
+ *
+ * This type is to be used to model such 64 bit wide fields.
+ */
+typedef u64 __bitwise dma64_t;
+
+/*
+ * Although DMA addresses should be obtained using the DMA API, in cases when
+ * it is known that the first argument holds a virtual address that points to
+ * DMA-able 31 bit addressable storage, then this function can be safely used.
+ */
+static inline dma32_t virt_to_dma32(void *ptr)
+{
+	return (__force dma32_t)__pa32(ptr);
+}
+
+static inline void *dma32_to_virt(dma32_t addr)
+{
+	return __va((__force unsigned long)addr);
+}
+
+static inline dma32_t u32_to_dma32(u32 addr)
+{
+	return (__force dma32_t)addr;
+}
+
+static inline u32 dma32_to_u32(dma32_t addr)
+{
+	return (__force u32)addr;
+}
+
+static inline dma32_t dma32_add(dma32_t a, u32 b)
+{
+	return (__force dma32_t)((__force u32)a + b);
+}
+
+static inline dma32_t dma32_and(dma32_t a, u32 b)
+{
+	return (__force dma32_t)((__force u32)a & b);
+}
+
+/*
+ * Although DMA addresses should be obtained using the DMA API, in cases when
+ * it is known that the first argument holds a virtual address that points to
+ * DMA-able storage, then this function can be safely used.
+ */
+static inline dma64_t virt_to_dma64(void *ptr)
+{
+	return (__force dma64_t)__pa(ptr);
+}
+
+static inline void *dma64_to_virt(dma64_t addr)
+{
+	return __va((__force unsigned long)addr);
+}
+
+static inline dma64_t u64_to_dma64(u64 addr)
+{
+	return (__force dma64_t)addr;
+}
+
+static inline u64 dma64_to_u64(dma64_t addr)
+{
+	return (__force u64)addr;
+}
+
+static inline dma64_t dma64_add(dma64_t a, u64 b)
+{
+	return (__force dma64_t)((__force u64)a + b);
+}
+
+static inline dma64_t dma64_and(dma64_t a, u64 b)
+{
+	return (__force dma64_t)((__force u64)a & b);
+}
+
+#endif /* _ASM_S390_DMA_TYPES_H_ */
diff --git a/arch/s390/include/asm/eadm.h b/arch/s390/include/asm/eadm.h
index 06f795855..c4589ec45 100644
--- a/arch/s390/include/asm/eadm.h
+++ b/arch/s390/include/asm/eadm.h
@@ -5,6 +5,7 @@
 #include <linux/types.h>
 #include <linux/device.h>
 #include <linux/blk_types.h>
+#include <asm/dma-types.h>
 
 struct arqb {
 	u64 data;
@@ -45,7 +46,7 @@ struct msb {
 	u16:12;
 	u16 bs:4;
 	u32 blk_count;
-	u64 data_addr;
+	dma64_t data_addr;
 	u64 scm_addr;
 	u64:64;
 } __packed;
@@ -54,7 +55,7 @@ struct aidaw {
 	u8 flags;
 	u32 :24;
 	u32 :32;
-	u64 data_addr;
+	dma64_t data_addr;
 } __packed;
 
 #define MSB_OC_CLEAR	0
diff --git a/arch/s390/include/asm/entry-common.h b/arch/s390/include/asm/entry-common.h
index fdd319a62..7f5004065 100644
--- a/arch/s390/include/asm/entry-common.h
+++ b/arch/s390/include/asm/entry-common.h
@@ -8,7 +8,7 @@
 #include <linux/processor.h>
 #include <linux/uaccess.h>
 #include <asm/timex.h>
-#include <asm/fpu/api.h>
+#include <asm/fpu.h>
 #include <asm/pai.h>
 
 #define ARCH_EXIT_TO_USER_MODE_WORK (_TIF_GUARDED_STORAGE | _TIF_PER_TRAP)
@@ -41,8 +41,7 @@ static __always_inline void arch_exit_to_user_mode_work(struct pt_regs *regs,
 
 static __always_inline void arch_exit_to_user_mode(void)
 {
-	if (test_cpu_flag(CIF_FPU))
-		__load_fpu_regs();
+	load_user_fpu_regs();
 
 	if (IS_ENABLED(CONFIG_DEBUG_ENTRY))
 		debug_user_asce(1);
diff --git a/arch/s390/include/asm/fcx.h b/arch/s390/include/asm/fcx.h
index 29784b4b4..80f82a739 100644
--- a/arch/s390/include/asm/fcx.h
+++ b/arch/s390/include/asm/fcx.h
@@ -10,6 +10,7 @@
 #define _ASM_S390_FCX_H
 
 #include <linux/types.h>
+#include <asm/dma-types.h>
 
 #define TCW_FORMAT_DEFAULT		0
 #define TCW_TIDAW_FORMAT_DEFAULT	0
@@ -43,16 +44,16 @@ struct tcw {
 	u32 r:1;
 	u32 w:1;
 	u32 :16;
-	u64 output;
-	u64 input;
-	u64 tsb;
-	u64 tccb;
+	dma64_t output;
+	dma64_t input;
+	dma64_t tsb;
+	dma64_t tccb;
 	u32 output_count;
 	u32 input_count;
 	u32 :32;
 	u32 :32;
 	u32 :32;
-	u32 intrg;
+	dma32_t intrg;
 } __attribute__ ((packed, aligned(64)));
 
 #define TIDAW_FLAGS_LAST		(1 << (7 - 0))
@@ -73,7 +74,7 @@ struct tidaw {
 	u32 flags:8;
 	u32 :24;
 	u32 count;
-	u64 addr;
+	dma64_t addr;
 } __attribute__ ((packed, aligned(16)));
 
 /**
diff --git a/arch/s390/include/asm/vx-insn-asm.h b/arch/s390/include/asm/fpu-insn-asm.h
index 360f8b36d..02ccfe460 100644
--- a/arch/s390/include/asm/vx-insn-asm.h
+++ b/arch/s390/include/asm/fpu-insn-asm.h
@@ -9,11 +9,11 @@
  * Author(s): Hendrik Brueckner <brueckner@linux.vnet.ibm.com>
  */
 
-#ifndef __ASM_S390_VX_INSN_INTERNAL_H
-#define __ASM_S390_VX_INSN_INTERNAL_H
+#ifndef __ASM_S390_FPU_INSN_ASM_H
+#define __ASM_S390_FPU_INSN_ASM_H
 
-#ifndef __ASM_S390_VX_INSN_H
-#error only <asm/vx-insn.h> can be included directly
+#ifndef __ASM_S390_FPU_INSN_H
+#error only <asm/fpu-insn.h> can be included directly
 #endif
 
 #ifdef __ASSEMBLY__
@@ -195,10 +195,26 @@
 /* RXB - Compute most significant bit used vector registers
  *
  * @rxb:	Operand to store computed RXB value
- * @v1:		First vector register designated operand
- * @v2:		Second vector register designated operand
- * @v3:		Third vector register designated operand
- * @v4:		Fourth vector register designated operand
+ * @v1:		Vector register designated operand whose MSB is stored in
+ *		RXB bit 0 (instruction bit 36) and whose remaining bits
+ *		are stored in instruction bits 8-11.
+ * @v2:		Vector register designated operand whose MSB is stored in
+ *		RXB bit 1 (instruction bit 37) and whose remaining bits
+ *		are stored in instruction bits 12-15.
+ * @v3:		Vector register designated operand whose MSB is stored in
+ *		RXB bit 2 (instruction bit 38) and whose remaining bits
+ *		are stored in instruction bits 16-19.
+ * @v4:		Vector register designated operand whose MSB is stored in
+ *		RXB bit 3 (instruction bit 39) and whose remaining bits
+ *		are stored in instruction bits 32-35.
+ *
+ * Note: In most vector instruction formats [1] V1, V2, V3, and V4 directly
+ * correspond to @v1, @v2, @v3, and @v4. But there are exceptions, such as but
+ * not limited to the vector instruction formats VRR-g, VRR-h, VRS-a, VRS-d,
+ * and VSI.
+ *
+ * [1] IBM z/Architecture Principles of Operation, chapter "Program
+ * Execution, section "Instructions", subsection "Instruction Formats".
  */
 .macro	RXB	rxb v1 v2=0 v3=0 v4=0
 	\rxb = 0
@@ -223,6 +239,9 @@
  * @v2:		Second vector register designated operand (for RXB)
  * @v3:		Third vector register designated operand (for RXB)
  * @v4:		Fourth vector register designated operand (for RXB)
+ *
+ * Note: For @v1, @v2, @v3, and @v4 also refer to the RXB macro
+ * description for further details.
  */
 .macro	MRXB	m v1 v2=0 v3=0 v4=0
 	rxb = 0
@@ -238,6 +257,9 @@
  * @v2:		Second vector register designated operand (for RXB)
  * @v3:		Third vector register designated operand (for RXB)
  * @v4:		Fourth vector register designated operand (for RXB)
+ *
+ * Note: For @v1, @v2, @v3, and @v4 also refer to the RXB macro
+ * description for further details.
  */
 .macro	MRXBOPC	m opc v1 v2=0 v3=0 v4=0
 	MRXB	\m, \v1, \v2, \v3, \v4
@@ -350,7 +372,7 @@
 	VX_NUM	v3, \vr
 	.word	0xE700 | (r1 << 4) | (v3&15)
 	.word	(b2 << 12) | (\disp)
-	MRXBOPC	\m, 0x21, v3
+	MRXBOPC	\m, 0x21, 0, v3
 .endm
 .macro	VLGVB	gr, vr, disp, base="%r0"
 	VLGV	\gr, \vr, \disp, \base, 0
@@ -499,6 +521,25 @@
 	VMRL	\vr1, \vr2, \vr3, 3
 .endm
 
+/* VECTOR LOAD WITH LENGTH */
+.macro VLL	v, gr, disp, base
+	VX_NUM	v1, \v
+	GR_NUM	b2, \base
+	GR_NUM	r3, \gr
+	.word	0xE700 | ((v1&15) << 4) | r3
+	.word	(b2 << 12) | (\disp)
+	MRXBOPC 0, 0x37, v1
+.endm
+
+/* VECTOR STORE WITH LENGTH */
+.macro VSTL	v, gr, disp, base
+	VX_NUM	v1, \v
+	GR_NUM	b2, \base
+	GR_NUM	r3, \gr
+	.word	0xE700 | ((v1&15) << 4) | r3
+	.word	(b2 << 12) | (\disp)
+	MRXBOPC 0, 0x3f, v1
+.endm
 
 /* Vector integer instructions */
 
@@ -512,6 +553,16 @@
 	MRXBOPC	0, 0x68, v1, v2, v3
 .endm
 
+/* VECTOR CHECKSUM */
+.macro VCKSM	vr1, vr2, vr3
+	VX_NUM	v1, \vr1
+	VX_NUM	v2, \vr2
+	VX_NUM	v3, \vr3
+	.word	0xE700 | ((v1&15) << 4) | (v2&15)
+	.word	((v3&15) << 12)
+	MRXBOPC 0, 0x66, v1, v2, v3
+.endm
+
 /* VECTOR EXCLUSIVE OR */
 .macro	VX	vr1, vr2, vr3
 	VX_NUM	v1, \vr1
@@ -678,4 +729,4 @@
 .endm
 
 #endif	/* __ASSEMBLY__ */
-#endif	/* __ASM_S390_VX_INSN_INTERNAL_H */
+#endif	/* __ASM_S390_FPU_INSN_ASM_H */
diff --git a/arch/s390/include/asm/fpu-insn.h b/arch/s390/include/asm/fpu-insn.h
new file mode 100644
index 000000000..c1e2e521d
--- /dev/null
+++ b/arch/s390/include/asm/fpu-insn.h
@@ -0,0 +1,486 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Support for Floating Point and Vector Instructions
+ *
+ */
+
+#ifndef __ASM_S390_FPU_INSN_H
+#define __ASM_S390_FPU_INSN_H
+
+#include <asm/fpu-insn-asm.h>
+
+#ifndef __ASSEMBLY__
+
+#include <linux/instrumented.h>
+#include <asm/asm-extable.h>
+
+asm(".include \"asm/fpu-insn-asm.h\"\n");
+
+/*
+ * Various small helper functions, which can and should be used within
+ * kernel fpu code sections. Each function represents only one floating
+ * point or vector instruction (except for helper functions which require
+ * exception handling).
+ *
+ * This allows to use floating point and vector instructions like C
+ * functions, which has the advantage that all supporting code, like
+ * e.g. loops, can be written in easy to read C code.
+ *
+ * Each of the helper functions provides support for code instrumentation,
+ * like e.g. KASAN. Therefore instrumentation is also covered automatically
+ * when using these functions.
+ *
+ * In order to ensure that code generated with the helper functions stays
+ * within kernel fpu sections, which are guarded with kernel_fpu_begin()
+ * and kernel_fpu_end() calls, each function has a mandatory "memory"
+ * barrier.
+ */
+
+static __always_inline void fpu_cefbr(u8 f1, s32 val)
+{
+	asm volatile("cefbr	%[f1],%[val]\n"
+		     :
+		     : [f1] "I" (f1), [val] "d" (val)
+		     : "memory");
+}
+
+static __always_inline unsigned long fpu_cgebr(u8 f2, u8 mode)
+{
+	unsigned long val;
+
+	asm volatile("cgebr	%[val],%[mode],%[f2]\n"
+		     : [val] "=d" (val)
+		     : [f2] "I" (f2), [mode] "I" (mode)
+		     : "memory");
+	return val;
+}
+
+static __always_inline void fpu_debr(u8 f1, u8 f2)
+{
+	asm volatile("debr	%[f1],%[f2]\n"
+		     :
+		     : [f1] "I" (f1), [f2] "I" (f2)
+		     : "memory");
+}
+
+static __always_inline void fpu_ld(unsigned short fpr, freg_t *reg)
+{
+	instrument_read(reg, sizeof(*reg));
+	asm volatile("ld	 %[fpr],%[reg]\n"
+		     :
+		     : [fpr] "I" (fpr), [reg] "Q" (reg->ui)
+		     : "memory");
+}
+
+static __always_inline void fpu_ldgr(u8 f1, u32 val)
+{
+	asm volatile("ldgr	%[f1],%[val]\n"
+		     :
+		     : [f1] "I" (f1), [val] "d" (val)
+		     : "memory");
+}
+
+static __always_inline void fpu_lfpc(unsigned int *fpc)
+{
+	instrument_read(fpc, sizeof(*fpc));
+	asm volatile("lfpc	%[fpc]"
+		     :
+		     : [fpc] "Q" (*fpc)
+		     : "memory");
+}
+
+/**
+ * fpu_lfpc_safe - Load floating point control register safely.
+ * @fpc: new value for floating point control register
+ *
+ * Load floating point control register. This may lead to an exception,
+ * since a saved value may have been modified by user space (ptrace,
+ * signal return, kvm registers) to an invalid value. In such a case
+ * set the floating point control register to zero.
+ */
+static inline void fpu_lfpc_safe(unsigned int *fpc)
+{
+	u32 tmp;
+
+	instrument_read(fpc, sizeof(*fpc));
+	asm volatile("\n"
+		"0:	lfpc	%[fpc]\n"
+		"1:	nopr	%%r7\n"
+		".pushsection .fixup, \"ax\"\n"
+		"2:	lghi	%[tmp],0\n"
+		"	sfpc	%[tmp]\n"
+		"	jg	1b\n"
+		".popsection\n"
+		EX_TABLE(1b, 2b)
+		: [tmp] "=d" (tmp)
+		: [fpc] "Q" (*fpc)
+		: "memory");
+}
+
+static __always_inline void fpu_std(unsigned short fpr, freg_t *reg)
+{
+	instrument_write(reg, sizeof(*reg));
+	asm volatile("std	 %[fpr],%[reg]\n"
+		     : [reg] "=Q" (reg->ui)
+		     : [fpr] "I" (fpr)
+		     : "memory");
+}
+
+static __always_inline void fpu_sfpc(unsigned int fpc)
+{
+	asm volatile("sfpc	%[fpc]"
+		     :
+		     : [fpc] "d" (fpc)
+		     : "memory");
+}
+
+static __always_inline void fpu_stfpc(unsigned int *fpc)
+{
+	instrument_write(fpc, sizeof(*fpc));
+	asm volatile("stfpc	%[fpc]"
+		     : [fpc] "=Q" (*fpc)
+		     :
+		     : "memory");
+}
+
+static __always_inline void fpu_vab(u8 v1, u8 v2, u8 v3)
+{
+	asm volatile("VAB	%[v1],%[v2],%[v3]"
+		     :
+		     : [v1] "I" (v1), [v2] "I" (v2), [v3] "I" (v3)
+		     : "memory");
+}
+
+static __always_inline void fpu_vcksm(u8 v1, u8 v2, u8 v3)
+{
+	asm volatile("VCKSM	%[v1],%[v2],%[v3]"
+		     :
+		     : [v1] "I" (v1), [v2] "I" (v2), [v3] "I" (v3)
+		     : "memory");
+}
+
+static __always_inline void fpu_vesravb(u8 v1, u8 v2, u8 v3)
+{
+	asm volatile("VESRAVB	%[v1],%[v2],%[v3]"
+		     :
+		     : [v1] "I" (v1), [v2] "I" (v2), [v3] "I" (v3)
+		     : "memory");
+}
+
+static __always_inline void fpu_vgfmag(u8 v1, u8 v2, u8 v3, u8 v4)
+{
+	asm volatile("VGFMAG	%[v1],%[v2],%[v3],%[v4]"
+		     :
+		     : [v1] "I" (v1), [v2] "I" (v2), [v3] "I" (v3), [v4] "I" (v4)
+		     : "memory");
+}
+
+static __always_inline void fpu_vgfmg(u8 v1, u8 v2, u8 v3)
+{
+	asm volatile("VGFMG	%[v1],%[v2],%[v3]"
+		     :
+		     : [v1] "I" (v1), [v2] "I" (v2), [v3] "I" (v3)
+		     : "memory");
+}
+
+#ifdef CONFIG_CC_IS_CLANG
+
+static __always_inline void fpu_vl(u8 v1, const void *vxr)
+{
+	instrument_read(vxr, sizeof(__vector128));
+	asm volatile("\n"
+		"	la	1,%[vxr]\n"
+		"	VL	%[v1],0,,1\n"
+		:
+		: [vxr] "R" (*(__vector128 *)vxr),
+		  [v1] "I" (v1)
+		: "memory", "1");
+}
+
+#else /* CONFIG_CC_IS_CLANG */
+
+static __always_inline void fpu_vl(u8 v1, const void *vxr)
+{
+	instrument_read(vxr, sizeof(__vector128));
+	asm volatile("VL	%[v1],%O[vxr],,%R[vxr]\n"
+		     :
+		     : [vxr] "Q" (*(__vector128 *)vxr),
+		       [v1] "I" (v1)
+		     : "memory");
+}
+
+#endif /* CONFIG_CC_IS_CLANG */
+
+static __always_inline void fpu_vleib(u8 v, s16 val, u8 index)
+{
+	asm volatile("VLEIB	%[v],%[val],%[index]"
+		     :
+		     : [v] "I" (v), [val] "K" (val), [index] "I" (index)
+		     : "memory");
+}
+
+static __always_inline void fpu_vleig(u8 v, s16 val, u8 index)
+{
+	asm volatile("VLEIG	%[v],%[val],%[index]"
+		     :
+		     : [v] "I" (v), [val] "K" (val), [index] "I" (index)
+		     : "memory");
+}
+
+static __always_inline u64 fpu_vlgvf(u8 v, u16 index)
+{
+	u64 val;
+
+	asm volatile("VLGVF	%[val],%[v],%[index]"
+		     : [val] "=d" (val)
+		     : [v] "I" (v), [index] "L" (index)
+		     : "memory");
+	return val;
+}
+
+#ifdef CONFIG_CC_IS_CLANG
+
+static __always_inline void fpu_vll(u8 v1, u32 index, const void *vxr)
+{
+	unsigned int size;
+
+	size = min(index + 1, sizeof(__vector128));
+	instrument_read(vxr, size);
+	asm volatile("\n"
+		"	la	1,%[vxr]\n"
+		"	VLL	%[v1],%[index],0,1\n"
+		:
+		: [vxr] "R" (*(u8 *)vxr),
+		  [index] "d" (index),
+		  [v1] "I" (v1)
+		: "memory", "1");
+}
+
+#else /* CONFIG_CC_IS_CLANG */
+
+static __always_inline void fpu_vll(u8 v1, u32 index, const void *vxr)
+{
+	unsigned int size;
+
+	size = min(index + 1, sizeof(__vector128));
+	instrument_read(vxr, size);
+	asm volatile("VLL	%[v1],%[index],%O[vxr],%R[vxr]\n"
+		     :
+		     : [vxr] "Q" (*(u8 *)vxr),
+		       [index] "d" (index),
+		       [v1] "I" (v1)
+		     : "memory");
+}
+
+#endif /* CONFIG_CC_IS_CLANG */
+
+#ifdef CONFIG_CC_IS_CLANG
+
+#define fpu_vlm(_v1, _v3, _vxrs)					\
+({									\
+	unsigned int size = ((_v3) - (_v1) + 1) * sizeof(__vector128);	\
+	struct {							\
+		__vector128 _v[(_v3) - (_v1) + 1];			\
+	} *_v = (void *)(_vxrs);					\
+									\
+	instrument_read(_v, size);					\
+	asm volatile("\n"						\
+		"	la	1,%[vxrs]\n"				\
+		"	VLM	%[v1],%[v3],0,1\n"			\
+		:							\
+		: [vxrs] "R" (*_v),					\
+		  [v1] "I" (_v1), [v3] "I" (_v3)			\
+		: "memory", "1");					\
+	(_v3) - (_v1) + 1;						\
+})
+
+#else /* CONFIG_CC_IS_CLANG */
+
+#define fpu_vlm(_v1, _v3, _vxrs)					\
+({									\
+	unsigned int size = ((_v3) - (_v1) + 1) * sizeof(__vector128);	\
+	struct {							\
+		__vector128 _v[(_v3) - (_v1) + 1];			\
+	} *_v = (void *)(_vxrs);					\
+									\
+	instrument_read(_v, size);					\
+	asm volatile("VLM	%[v1],%[v3],%O[vxrs],%R[vxrs]\n"	\
+		     :							\
+		     : [vxrs] "Q" (*_v),				\
+		       [v1] "I" (_v1), [v3] "I" (_v3)			\
+		     : "memory");					\
+	(_v3) - (_v1) + 1;						\
+})
+
+#endif /* CONFIG_CC_IS_CLANG */
+
+static __always_inline void fpu_vlr(u8 v1, u8 v2)
+{
+	asm volatile("VLR	%[v1],%[v2]"
+		     :
+		     : [v1] "I" (v1), [v2] "I" (v2)
+		     : "memory");
+}
+
+static __always_inline void fpu_vlvgf(u8 v, u32 val, u16 index)
+{
+	asm volatile("VLVGF	%[v],%[val],%[index]"
+		     :
+		     : [v] "I" (v), [val] "d" (val), [index] "L" (index)
+		     : "memory");
+}
+
+static __always_inline void fpu_vn(u8 v1, u8 v2, u8 v3)
+{
+	asm volatile("VN	%[v1],%[v2],%[v3]"
+		     :
+		     : [v1] "I" (v1), [v2] "I" (v2), [v3] "I" (v3)
+		     : "memory");
+}
+
+static __always_inline void fpu_vperm(u8 v1, u8 v2, u8 v3, u8 v4)
+{
+	asm volatile("VPERM	%[v1],%[v2],%[v3],%[v4]"
+		     :
+		     : [v1] "I" (v1), [v2] "I" (v2), [v3] "I" (v3), [v4] "I" (v4)
+		     : "memory");
+}
+
+static __always_inline void fpu_vrepib(u8 v1, s16 i2)
+{
+	asm volatile("VREPIB	%[v1],%[i2]"
+		     :
+		     : [v1] "I" (v1), [i2] "K" (i2)
+		     : "memory");
+}
+
+static __always_inline void fpu_vsrlb(u8 v1, u8 v2, u8 v3)
+{
+	asm volatile("VSRLB	%[v1],%[v2],%[v3]"
+		     :
+		     : [v1] "I" (v1), [v2] "I" (v2), [v3] "I" (v3)
+		     : "memory");
+}
+
+#ifdef CONFIG_CC_IS_CLANG
+
+static __always_inline void fpu_vst(u8 v1, const void *vxr)
+{
+	instrument_write(vxr, sizeof(__vector128));
+	asm volatile("\n"
+		"	la	1,%[vxr]\n"
+		"	VST	%[v1],0,,1\n"
+		: [vxr] "=R" (*(__vector128 *)vxr)
+		: [v1] "I" (v1)
+		: "memory", "1");
+}
+
+#else /* CONFIG_CC_IS_CLANG */
+
+static __always_inline void fpu_vst(u8 v1, const void *vxr)
+{
+	instrument_write(vxr, sizeof(__vector128));
+	asm volatile("VST	%[v1],%O[vxr],,%R[vxr]\n"
+		     : [vxr] "=Q" (*(__vector128 *)vxr)
+		     : [v1] "I" (v1)
+		     : "memory");
+}
+
+#endif /* CONFIG_CC_IS_CLANG */
+
+#ifdef CONFIG_CC_IS_CLANG
+
+static __always_inline void fpu_vstl(u8 v1, u32 index, const void *vxr)
+{
+	unsigned int size;
+
+	size = min(index + 1, sizeof(__vector128));
+	instrument_write(vxr, size);
+	asm volatile("\n"
+		"	la	1,%[vxr]\n"
+		"	VSTL	%[v1],%[index],0,1\n"
+		: [vxr] "=R" (*(u8 *)vxr)
+		: [index] "d" (index), [v1] "I" (v1)
+		: "memory", "1");
+}
+
+#else /* CONFIG_CC_IS_CLANG */
+
+static __always_inline void fpu_vstl(u8 v1, u32 index, const void *vxr)
+{
+	unsigned int size;
+
+	size = min(index + 1, sizeof(__vector128));
+	instrument_write(vxr, size);
+	asm volatile("VSTL	%[v1],%[index],%O[vxr],%R[vxr]\n"
+		     : [vxr] "=Q" (*(u8 *)vxr)
+		     : [index] "d" (index), [v1] "I" (v1)
+		     : "memory");
+}
+
+#endif /* CONFIG_CC_IS_CLANG */
+
+#ifdef CONFIG_CC_IS_CLANG
+
+#define fpu_vstm(_v1, _v3, _vxrs)					\
+({									\
+	unsigned int size = ((_v3) - (_v1) + 1) * sizeof(__vector128);	\
+	struct {							\
+		__vector128 _v[(_v3) - (_v1) + 1];			\
+	} *_v = (void *)(_vxrs);					\
+									\
+	instrument_write(_v, size);					\
+	asm volatile("\n"						\
+		"	la	1,%[vxrs]\n"				\
+		"	VSTM	%[v1],%[v3],0,1\n"			\
+		: [vxrs] "=R" (*_v)					\
+		: [v1] "I" (_v1), [v3] "I" (_v3)			\
+		: "memory", "1");					\
+	(_v3) - (_v1) + 1;						\
+})
+
+#else /* CONFIG_CC_IS_CLANG */
+
+#define fpu_vstm(_v1, _v3, _vxrs)					\
+({									\
+	unsigned int size = ((_v3) - (_v1) + 1) * sizeof(__vector128);	\
+	struct {							\
+		__vector128 _v[(_v3) - (_v1) + 1];			\
+	} *_v = (void *)(_vxrs);					\
+									\
+	instrument_write(_v, size);					\
+	asm volatile("VSTM	%[v1],%[v3],%O[vxrs],%R[vxrs]\n"	\
+		     : [vxrs] "=Q" (*_v)				\
+		     : [v1] "I" (_v1), [v3] "I" (_v3)			\
+		     : "memory");					\
+	(_v3) - (_v1) + 1;						\
+})
+
+#endif /* CONFIG_CC_IS_CLANG */
+
+static __always_inline void fpu_vupllf(u8 v1, u8 v2)
+{
+	asm volatile("VUPLLF	%[v1],%[v2]"
+		     :
+		     : [v1] "I" (v1), [v2] "I" (v2)
+		     : "memory");
+}
+
+static __always_inline void fpu_vx(u8 v1, u8 v2, u8 v3)
+{
+	asm volatile("VX	%[v1],%[v2],%[v3]"
+		     :
+		     : [v1] "I" (v1), [v2] "I" (v2), [v3] "I" (v3)
+		     : "memory");
+}
+
+static __always_inline void fpu_vzero(u8 v)
+{
+	asm volatile("VZERO	%[v]"
+		     :
+		     : [v] "I" (v)
+		     : "memory");
+}
+
+#endif /* __ASSEMBLY__ */
+#endif	/* __ASM_S390_FPU_INSN_H */
diff --git a/arch/s390/include/asm/fpu-types.h b/arch/s390/include/asm/fpu-types.h
new file mode 100644
index 000000000..8d58d5a95
--- /dev/null
+++ b/arch/s390/include/asm/fpu-types.h
@@ -0,0 +1,51 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * FPU data structures
+ *
+ * Copyright IBM Corp. 2015
+ * Author(s): Hendrik Brueckner <brueckner@linux.vnet.ibm.com>
+ */
+
+#ifndef _ASM_S390_FPU_TYPES_H
+#define _ASM_S390_FPU_TYPES_H
+
+#include <asm/sigcontext.h>
+
+struct fpu {
+	u32 fpc;
+	__vector128 vxrs[__NUM_VXRS] __aligned(8);
+};
+
+struct kernel_fpu_hdr {
+	int	mask;
+	u32	fpc;
+};
+
+struct kernel_fpu {
+	struct kernel_fpu_hdr hdr;
+	__vector128 vxrs[] __aligned(8);
+};
+
+#define KERNEL_FPU_STRUCT(vxr_size)				\
+struct kernel_fpu_##vxr_size {					\
+	struct kernel_fpu_hdr hdr;				\
+	__vector128 vxrs[vxr_size] __aligned(8);		\
+}
+
+KERNEL_FPU_STRUCT(8);
+KERNEL_FPU_STRUCT(16);
+KERNEL_FPU_STRUCT(32);
+
+#define DECLARE_KERNEL_FPU_ONSTACK(vxr_size, name)		\
+	struct kernel_fpu_##vxr_size name __uninitialized
+
+#define DECLARE_KERNEL_FPU_ONSTACK8(name)			\
+	DECLARE_KERNEL_FPU_ONSTACK(8, name)
+
+#define DECLARE_KERNEL_FPU_ONSTACK16(name)			\
+	DECLARE_KERNEL_FPU_ONSTACK(16, name)
+
+#define DECLARE_KERNEL_FPU_ONSTACK32(name)			\
+	DECLARE_KERNEL_FPU_ONSTACK(32, name)
+
+#endif /* _ASM_S390_FPU_TYPES_H */
diff --git a/arch/s390/include/asm/fpu.h b/arch/s390/include/asm/fpu.h
new file mode 100644
index 000000000..c84cb3391
--- /dev/null
+++ b/arch/s390/include/asm/fpu.h
@@ -0,0 +1,295 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * In-kernel FPU support functions
+ *
+ *
+ * Consider these guidelines before using in-kernel FPU functions:
+ *
+ *  1. Use kernel_fpu_begin() and kernel_fpu_end() to enclose all in-kernel
+ *     use of floating-point or vector registers and instructions.
+ *
+ *  2. For kernel_fpu_begin(), specify the vector register range you want to
+ *     use with the KERNEL_VXR_* constants. Consider these usage guidelines:
+ *
+ *     a) If your function typically runs in process-context, use the lower
+ *	  half of the vector registers, for example, specify KERNEL_VXR_LOW.
+ *     b) If your function typically runs in soft-irq or hard-irq context,
+ *	  prefer using the upper half of the vector registers, for example,
+ *	  specify KERNEL_VXR_HIGH.
+ *
+ *     If you adhere to these guidelines, an interrupted process context
+ *     does not require to save and restore vector registers because of
+ *     disjoint register ranges.
+ *
+ *     Also note that the __kernel_fpu_begin()/__kernel_fpu_end() functions
+ *     includes logic to save and restore up to 16 vector registers at once.
+ *
+ *  3. You can nest kernel_fpu_begin()/kernel_fpu_end() by using different
+ *     struct kernel_fpu states.  Vector registers that are in use by outer
+ *     levels are saved and restored.  You can minimize the save and restore
+ *     effort by choosing disjoint vector register ranges.
+ *
+ *  5. To use vector floating-point instructions, specify the KERNEL_FPC
+ *     flag to save and restore floating-point controls in addition to any
+ *     vector register range.
+ *
+ *  6. To use floating-point registers and instructions only, specify the
+ *     KERNEL_FPR flag.  This flag triggers a save and restore of vector
+ *     registers V0 to V15 and floating-point controls.
+ *
+ * Copyright IBM Corp. 2015
+ * Author(s): Hendrik Brueckner <brueckner@linux.vnet.ibm.com>
+ */
+
+#ifndef _ASM_S390_FPU_H
+#define _ASM_S390_FPU_H
+
+#include <linux/processor.h>
+#include <linux/preempt.h>
+#include <linux/string.h>
+#include <linux/sched.h>
+#include <asm/sigcontext.h>
+#include <asm/fpu-types.h>
+#include <asm/fpu-insn.h>
+#include <asm/facility.h>
+
+static inline bool cpu_has_vx(void)
+{
+	return likely(test_facility(129));
+}
+
+enum {
+	KERNEL_FPC_BIT = 0,
+	KERNEL_VXR_V0V7_BIT,
+	KERNEL_VXR_V8V15_BIT,
+	KERNEL_VXR_V16V23_BIT,
+	KERNEL_VXR_V24V31_BIT,
+};
+
+#define KERNEL_FPC		BIT(KERNEL_FPC_BIT)
+#define KERNEL_VXR_V0V7		BIT(KERNEL_VXR_V0V7_BIT)
+#define KERNEL_VXR_V8V15	BIT(KERNEL_VXR_V8V15_BIT)
+#define KERNEL_VXR_V16V23	BIT(KERNEL_VXR_V16V23_BIT)
+#define KERNEL_VXR_V24V31	BIT(KERNEL_VXR_V24V31_BIT)
+
+#define KERNEL_VXR_LOW		(KERNEL_VXR_V0V7   | KERNEL_VXR_V8V15)
+#define KERNEL_VXR_MID		(KERNEL_VXR_V8V15  | KERNEL_VXR_V16V23)
+#define KERNEL_VXR_HIGH		(KERNEL_VXR_V16V23 | KERNEL_VXR_V24V31)
+
+#define KERNEL_VXR		(KERNEL_VXR_LOW	   | KERNEL_VXR_HIGH)
+#define KERNEL_FPR		(KERNEL_FPC	   | KERNEL_VXR_LOW)
+
+void load_fpu_state(struct fpu *state, int flags);
+void save_fpu_state(struct fpu *state, int flags);
+void __kernel_fpu_begin(struct kernel_fpu *state, int flags);
+void __kernel_fpu_end(struct kernel_fpu *state, int flags);
+
+static __always_inline void save_vx_regs(__vector128 *vxrs)
+{
+	fpu_vstm(0, 15, &vxrs[0]);
+	fpu_vstm(16, 31, &vxrs[16]);
+}
+
+static __always_inline void load_vx_regs(__vector128 *vxrs)
+{
+	fpu_vlm(0, 15, &vxrs[0]);
+	fpu_vlm(16, 31, &vxrs[16]);
+}
+
+static __always_inline void __save_fp_regs(freg_t *fprs, unsigned int offset)
+{
+	fpu_std(0, &fprs[0 * offset]);
+	fpu_std(1, &fprs[1 * offset]);
+	fpu_std(2, &fprs[2 * offset]);
+	fpu_std(3, &fprs[3 * offset]);
+	fpu_std(4, &fprs[4 * offset]);
+	fpu_std(5, &fprs[5 * offset]);
+	fpu_std(6, &fprs[6 * offset]);
+	fpu_std(7, &fprs[7 * offset]);
+	fpu_std(8, &fprs[8 * offset]);
+	fpu_std(9, &fprs[9 * offset]);
+	fpu_std(10, &fprs[10 * offset]);
+	fpu_std(11, &fprs[11 * offset]);
+	fpu_std(12, &fprs[12 * offset]);
+	fpu_std(13, &fprs[13 * offset]);
+	fpu_std(14, &fprs[14 * offset]);
+	fpu_std(15, &fprs[15 * offset]);
+}
+
+static __always_inline void __load_fp_regs(freg_t *fprs, unsigned int offset)
+{
+	fpu_ld(0, &fprs[0 * offset]);
+	fpu_ld(1, &fprs[1 * offset]);
+	fpu_ld(2, &fprs[2 * offset]);
+	fpu_ld(3, &fprs[3 * offset]);
+	fpu_ld(4, &fprs[4 * offset]);
+	fpu_ld(5, &fprs[5 * offset]);
+	fpu_ld(6, &fprs[6 * offset]);
+	fpu_ld(7, &fprs[7 * offset]);
+	fpu_ld(8, &fprs[8 * offset]);
+	fpu_ld(9, &fprs[9 * offset]);
+	fpu_ld(10, &fprs[10 * offset]);
+	fpu_ld(11, &fprs[11 * offset]);
+	fpu_ld(12, &fprs[12 * offset]);
+	fpu_ld(13, &fprs[13 * offset]);
+	fpu_ld(14, &fprs[14 * offset]);
+	fpu_ld(15, &fprs[15 * offset]);
+}
+
+static __always_inline void save_fp_regs(freg_t *fprs)
+{
+	__save_fp_regs(fprs, sizeof(freg_t) / sizeof(freg_t));
+}
+
+static __always_inline void load_fp_regs(freg_t *fprs)
+{
+	__load_fp_regs(fprs, sizeof(freg_t) / sizeof(freg_t));
+}
+
+static __always_inline void save_fp_regs_vx(__vector128 *vxrs)
+{
+	freg_t *fprs = (freg_t *)&vxrs[0].high;
+
+	__save_fp_regs(fprs, sizeof(__vector128) / sizeof(freg_t));
+}
+
+static __always_inline void load_fp_regs_vx(__vector128 *vxrs)
+{
+	freg_t *fprs = (freg_t *)&vxrs[0].high;
+
+	__load_fp_regs(fprs, sizeof(__vector128) / sizeof(freg_t));
+}
+
+static inline void load_user_fpu_regs(void)
+{
+	struct thread_struct *thread = &current->thread;
+
+	if (!thread->ufpu_flags)
+		return;
+	load_fpu_state(&thread->ufpu, thread->ufpu_flags);
+	thread->ufpu_flags = 0;
+}
+
+static __always_inline void __save_user_fpu_regs(struct thread_struct *thread, int flags)
+{
+	save_fpu_state(&thread->ufpu, flags);
+	__atomic_or(flags, &thread->ufpu_flags);
+}
+
+static inline void save_user_fpu_regs(void)
+{
+	struct thread_struct *thread = &current->thread;
+	int mask, flags;
+
+	mask = __atomic_or(KERNEL_FPC | KERNEL_VXR, &thread->kfpu_flags);
+	flags = ~READ_ONCE(thread->ufpu_flags) & (KERNEL_FPC | KERNEL_VXR);
+	if (flags)
+		__save_user_fpu_regs(thread, flags);
+	barrier();
+	WRITE_ONCE(thread->kfpu_flags, mask);
+}
+
+static __always_inline void _kernel_fpu_begin(struct kernel_fpu *state, int flags)
+{
+	struct thread_struct *thread = &current->thread;
+	int mask, uflags;
+
+	mask = __atomic_or(flags, &thread->kfpu_flags);
+	state->hdr.mask = mask;
+	uflags = READ_ONCE(thread->ufpu_flags);
+	if ((uflags & flags) != flags)
+		__save_user_fpu_regs(thread, ~uflags & flags);
+	if (mask & flags)
+		__kernel_fpu_begin(state, flags);
+}
+
+static __always_inline void _kernel_fpu_end(struct kernel_fpu *state, int flags)
+{
+	int mask = state->hdr.mask;
+
+	if (mask & flags)
+		__kernel_fpu_end(state, flags);
+	barrier();
+	WRITE_ONCE(current->thread.kfpu_flags, mask);
+}
+
+void __kernel_fpu_invalid_size(void);
+
+static __always_inline void kernel_fpu_check_size(int flags, unsigned int size)
+{
+	unsigned int cnt = 0;
+
+	if (flags & KERNEL_VXR_V0V7)
+		cnt += 8;
+	if (flags & KERNEL_VXR_V8V15)
+		cnt += 8;
+	if (flags & KERNEL_VXR_V16V23)
+		cnt += 8;
+	if (flags & KERNEL_VXR_V24V31)
+		cnt += 8;
+	if (cnt != size)
+		__kernel_fpu_invalid_size();
+}
+
+#define kernel_fpu_begin(state, flags)					\
+{									\
+	typeof(state) s = (state);					\
+	int _flags = (flags);						\
+									\
+	kernel_fpu_check_size(_flags, ARRAY_SIZE(s->vxrs));		\
+	_kernel_fpu_begin((struct kernel_fpu *)s, _flags);		\
+}
+
+#define kernel_fpu_end(state, flags)					\
+{									\
+	typeof(state) s = (state);					\
+	int _flags = (flags);						\
+									\
+	kernel_fpu_check_size(_flags, ARRAY_SIZE(s->vxrs));		\
+	_kernel_fpu_end((struct kernel_fpu *)s, _flags);		\
+}
+
+static inline void save_kernel_fpu_regs(struct thread_struct *thread)
+{
+	if (!thread->kfpu_flags)
+		return;
+	save_fpu_state(&thread->kfpu, thread->kfpu_flags);
+}
+
+static inline void restore_kernel_fpu_regs(struct thread_struct *thread)
+{
+	if (!thread->kfpu_flags)
+		return;
+	load_fpu_state(&thread->kfpu, thread->kfpu_flags);
+}
+
+static inline void convert_vx_to_fp(freg_t *fprs, __vector128 *vxrs)
+{
+	int i;
+
+	for (i = 0; i < __NUM_FPRS; i++)
+		fprs[i].ui = vxrs[i].high;
+}
+
+static inline void convert_fp_to_vx(__vector128 *vxrs, freg_t *fprs)
+{
+	int i;
+
+	for (i = 0; i < __NUM_FPRS; i++)
+		vxrs[i].high = fprs[i].ui;
+}
+
+static inline void fpregs_store(_s390_fp_regs *fpregs, struct fpu *fpu)
+{
+	fpregs->pad = 0;
+	fpregs->fpc = fpu->fpc;
+	convert_vx_to_fp((freg_t *)&fpregs->fprs, fpu->vxrs);
+}
+
+static inline void fpregs_load(_s390_fp_regs *fpregs, struct fpu *fpu)
+{
+	fpu->fpc = fpregs->fpc;
+	convert_fp_to_vx(fpu->vxrs, (freg_t *)&fpregs->fprs);
+}
+
+#endif /* _ASM_S390_FPU_H */
diff --git a/arch/s390/include/asm/fpu/api.h b/arch/s390/include/asm/fpu/api.h
deleted file mode 100644
index d6ca8bc6c..000000000
--- a/arch/s390/include/asm/fpu/api.h
+++ /dev/null
@@ -1,126 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0 */
-/*
- * In-kernel FPU support functions
- *
- *
- * Consider these guidelines before using in-kernel FPU functions:
- *
- *  1. Use kernel_fpu_begin() and kernel_fpu_end() to enclose all in-kernel
- *     use of floating-point or vector registers and instructions.
- *
- *  2. For kernel_fpu_begin(), specify the vector register range you want to
- *     use with the KERNEL_VXR_* constants. Consider these usage guidelines:
- *
- *     a) If your function typically runs in process-context, use the lower
- *	  half of the vector registers, for example, specify KERNEL_VXR_LOW.
- *     b) If your function typically runs in soft-irq or hard-irq context,
- *	  prefer using the upper half of the vector registers, for example,
- *	  specify KERNEL_VXR_HIGH.
- *
- *     If you adhere to these guidelines, an interrupted process context
- *     does not require to save and restore vector registers because of
- *     disjoint register ranges.
- *
- *     Also note that the __kernel_fpu_begin()/__kernel_fpu_end() functions
- *     includes logic to save and restore up to 16 vector registers at once.
- *
- *  3. You can nest kernel_fpu_begin()/kernel_fpu_end() by using different
- *     struct kernel_fpu states.  Vector registers that are in use by outer
- *     levels are saved and restored.  You can minimize the save and restore
- *     effort by choosing disjoint vector register ranges.
- *
- *  5. To use vector floating-point instructions, specify the KERNEL_FPC
- *     flag to save and restore floating-point controls in addition to any
- *     vector register range.
- *
- *  6. To use floating-point registers and instructions only, specify the
- *     KERNEL_FPR flag.  This flag triggers a save and restore of vector
- *     registers V0 to V15 and floating-point controls.
- *
- * Copyright IBM Corp. 2015
- * Author(s): Hendrik Brueckner <brueckner@linux.vnet.ibm.com>
- */
-
-#ifndef _ASM_S390_FPU_API_H
-#define _ASM_S390_FPU_API_H
-
-#include <linux/preempt.h>
-#include <asm/asm-extable.h>
-#include <asm/fpu/internal.h>
-
-void save_fpu_regs(void);
-void load_fpu_regs(void);
-void __load_fpu_regs(void);
-
-/**
- * sfpc_safe - Set floating point control register safely.
- * @fpc: new value for floating point control register
- *
- * Set floating point control register. This may lead to an exception,
- * since a saved value may have been modified by user space (ptrace,
- * signal return, kvm registers) to an invalid value. In such a case
- * set the floating point control register to zero.
- */
-static inline void sfpc_safe(u32 fpc)
-{
-	asm volatile("\n"
-		"0:	sfpc	%[fpc]\n"
-		"1:	nopr	%%r7\n"
-		".pushsection .fixup, \"ax\"\n"
-		"2:	lghi	%[fpc],0\n"
-		"	jg	0b\n"
-		".popsection\n"
-		EX_TABLE(1b, 2b)
-		: [fpc] "+d" (fpc)
-		: : "memory");
-}
-
-#define KERNEL_FPC		1
-#define KERNEL_VXR_V0V7		2
-#define KERNEL_VXR_V8V15	4
-#define KERNEL_VXR_V16V23	8
-#define KERNEL_VXR_V24V31	16
-
-#define KERNEL_VXR_LOW		(KERNEL_VXR_V0V7|KERNEL_VXR_V8V15)
-#define KERNEL_VXR_MID		(KERNEL_VXR_V8V15|KERNEL_VXR_V16V23)
-#define KERNEL_VXR_HIGH		(KERNEL_VXR_V16V23|KERNEL_VXR_V24V31)
-
-#define KERNEL_VXR		(KERNEL_VXR_LOW|KERNEL_VXR_HIGH)
-#define KERNEL_FPR		(KERNEL_FPC|KERNEL_VXR_LOW)
-
-struct kernel_fpu;
-
-/*
- * Note the functions below must be called with preemption disabled.
- * Do not enable preemption before calling __kernel_fpu_end() to prevent
- * an corruption of an existing kernel FPU state.
- *
- * Prefer using the kernel_fpu_begin()/kernel_fpu_end() pair of functions.
- */
-void __kernel_fpu_begin(struct kernel_fpu *state, u32 flags);
-void __kernel_fpu_end(struct kernel_fpu *state, u32 flags);
-
-
-static inline void kernel_fpu_begin(struct kernel_fpu *state, u32 flags)
-{
-	preempt_disable();
-	state->mask = S390_lowcore.fpu_flags;
-	if (!test_cpu_flag(CIF_FPU))
-		/* Save user space FPU state and register contents */
-		save_fpu_regs();
-	else if (state->mask & flags)
-		/* Save FPU/vector register in-use by the kernel */
-		__kernel_fpu_begin(state, flags);
-	S390_lowcore.fpu_flags |= flags;
-}
-
-static inline void kernel_fpu_end(struct kernel_fpu *state, u32 flags)
-{
-	S390_lowcore.fpu_flags = state->mask;
-	if (state->mask & flags)
-		/* Restore FPU/vector register in-use by the kernel */
-		__kernel_fpu_end(state, flags);
-	preempt_enable();
-}
-
-#endif /* _ASM_S390_FPU_API_H */
diff --git a/arch/s390/include/asm/fpu/internal.h b/arch/s390/include/asm/fpu/internal.h
deleted file mode 100644
index d511c4cf5..000000000
--- a/arch/s390/include/asm/fpu/internal.h
+++ /dev/null
@@ -1,67 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0 */
-/*
- * FPU state and register content conversion primitives
- *
- * Copyright IBM Corp. 2015
- * Author(s): Hendrik Brueckner <brueckner@linux.vnet.ibm.com>
- */
-
-#ifndef _ASM_S390_FPU_INTERNAL_H
-#define _ASM_S390_FPU_INTERNAL_H
-
-#include <linux/string.h>
-#include <asm/facility.h>
-#include <asm/fpu/types.h>
-
-static inline bool cpu_has_vx(void)
-{
-	return likely(test_facility(129));
-}
-
-static inline void save_vx_regs(__vector128 *vxrs)
-{
-	asm volatile(
-		"	la	1,%0\n"
-		"	.word	0xe70f,0x1000,0x003e\n"	/* vstm 0,15,0(1) */
-		"	.word	0xe70f,0x1100,0x0c3e\n"	/* vstm 16,31,256(1) */
-		: "=Q" (*(struct vx_array *) vxrs) : : "1");
-}
-
-static inline void convert_vx_to_fp(freg_t *fprs, __vector128 *vxrs)
-{
-	int i;
-
-	for (i = 0; i < __NUM_FPRS; i++)
-		fprs[i].ui = vxrs[i].high;
-}
-
-static inline void convert_fp_to_vx(__vector128 *vxrs, freg_t *fprs)
-{
-	int i;
-
-	for (i = 0; i < __NUM_FPRS; i++)
-		vxrs[i].high = fprs[i].ui;
-}
-
-static inline void fpregs_store(_s390_fp_regs *fpregs, struct fpu *fpu)
-{
-	fpregs->pad = 0;
-	fpregs->fpc = fpu->fpc;
-	if (cpu_has_vx())
-		convert_vx_to_fp((freg_t *)&fpregs->fprs, fpu->vxrs);
-	else
-		memcpy((freg_t *)&fpregs->fprs, fpu->fprs,
-		       sizeof(fpregs->fprs));
-}
-
-static inline void fpregs_load(_s390_fp_regs *fpregs, struct fpu *fpu)
-{
-	fpu->fpc = fpregs->fpc;
-	if (cpu_has_vx())
-		convert_fp_to_vx(fpu->vxrs, (freg_t *)&fpregs->fprs);
-	else
-		memcpy(fpu->fprs, (freg_t *)&fpregs->fprs,
-		       sizeof(fpregs->fprs));
-}
-
-#endif /* _ASM_S390_FPU_INTERNAL_H */
diff --git a/arch/s390/include/asm/fpu/types.h b/arch/s390/include/asm/fpu/types.h
deleted file mode 100644
index d889e9436..000000000
--- a/arch/s390/include/asm/fpu/types.h
+++ /dev/null
@@ -1,38 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0 */
-/*
- * FPU data structures
- *
- * Copyright IBM Corp. 2015
- * Author(s): Hendrik Brueckner <brueckner@linux.vnet.ibm.com>
- */
-
-#ifndef _ASM_S390_FPU_TYPES_H
-#define _ASM_S390_FPU_TYPES_H
-
-#include <asm/sigcontext.h>
-
-struct fpu {
-	__u32 fpc;		/* Floating-point control */
-	void *regs;		/* Pointer to the current save area */
-	union {
-		/* Floating-point register save area */
-		freg_t fprs[__NUM_FPRS];
-		/* Vector register save area */
-		__vector128 vxrs[__NUM_VXRS];
-	};
-};
-
-/* VX array structure for address operand constraints in inline assemblies */
-struct vx_array { __vector128 _[__NUM_VXRS]; };
-
-/* In-kernel FPU state structure */
-struct kernel_fpu {
-	u32	    mask;
-	u32	    fpc;
-	union {
-		freg_t fprs[__NUM_FPRS];
-		__vector128 vxrs[__NUM_VXRS];
-	};
-};
-
-#endif /* _ASM_S390_FPU_TYPES_H */
diff --git a/arch/s390/include/asm/ftrace.h b/arch/s390/include/asm/ftrace.h
index 5a82b08f0..621f23d5a 100644
--- a/arch/s390/include/asm/ftrace.h
+++ b/arch/s390/include/asm/ftrace.h
@@ -9,7 +9,7 @@
 #ifndef __ASSEMBLY__
 
 #ifdef CONFIG_CC_IS_CLANG
-/* https://bugs.llvm.org/show_bug.cgi?id=41424 */
+/* https://llvm.org/pr41424 */
 #define ftrace_return_address(n) 0UL
 #else
 #define ftrace_return_address(n) __builtin_return_address(n)
diff --git a/arch/s390/include/asm/gmap.h b/arch/s390/include/asm/gmap.h
index 9725586f4..5cc46e0dd 100644
--- a/arch/s390/include/asm/gmap.h
+++ b/arch/s390/include/asm/gmap.h
@@ -146,7 +146,7 @@ int gmap_mprotect_notify(struct gmap *, unsigned long start,
 
 void gmap_sync_dirty_log_pmd(struct gmap *gmap, unsigned long dirty_bitmap[4],
 			     unsigned long gaddr, unsigned long vmaddr);
-int s390_disable_cow_sharing(void);
+int gmap_mark_unmergeable(void);
 void s390_unlist_old_asce(struct gmap *gmap);
 int s390_replace_asce(struct gmap *gmap);
 void s390_uv_destroy_pfns(unsigned long count, unsigned long *pfns);
diff --git a/arch/s390/include/asm/idals.h b/arch/s390/include/asm/idals.h
index 59fcc3c72..ac68c657b 100644
--- a/arch/s390/include/asm/idals.h
+++ b/arch/s390/include/asm/idals.h
@@ -1,5 +1,5 @@
 /* SPDX-License-Identifier: GPL-2.0 */
-/* 
+/*
  * Author(s)......: Holger Smolinski <Holger.Smolinski@de.ibm.com>
  *		    Martin Schwidefsky <schwidefsky@de.ibm.com>
  * Bugreports.to..: <Linux390@de.ibm.com>
@@ -17,32 +17,37 @@
 #include <linux/err.h>
 #include <linux/types.h>
 #include <linux/slab.h>
-#include <asm/cio.h>
 #include <linux/uaccess.h>
+#include <asm/dma-types.h>
+#include <asm/cio.h>
 
-#define IDA_SIZE_LOG 12 /* 11 for 2k , 12 for 4k */
-#define IDA_BLOCK_SIZE (1L<<IDA_SIZE_LOG)
+#define IDA_SIZE_SHIFT		12
+#define IDA_BLOCK_SIZE		(1UL << IDA_SIZE_SHIFT)
 
-#define IDA_2K_SIZE_LOG 11
-#define IDA_2K_BLOCK_SIZE (1L << IDA_2K_SIZE_LOG)
+#define IDA_2K_SIZE_SHIFT	11
+#define IDA_2K_BLOCK_SIZE	(1UL << IDA_2K_SIZE_SHIFT)
 
 /*
  * Test if an address/length pair needs an idal list.
  */
-static inline int
-idal_is_needed(void *vaddr, unsigned int length)
+static inline bool idal_is_needed(void *vaddr, unsigned int length)
 {
-	return ((__pa(vaddr) + length - 1) >> 31) != 0;
-}
+	dma64_t paddr = virt_to_dma64(vaddr);
 
+	return (((__force unsigned long)(paddr) + length - 1) >> 31) != 0;
+}
 
 /*
  * Return the number of idal words needed for an address/length pair.
  */
 static inline unsigned int idal_nr_words(void *vaddr, unsigned int length)
 {
-	return ((__pa(vaddr) & (IDA_BLOCK_SIZE-1)) + length +
-		(IDA_BLOCK_SIZE-1)) >> IDA_SIZE_LOG;
+	unsigned int cidaw;
+
+	cidaw = (unsigned long)vaddr & (IDA_BLOCK_SIZE - 1);
+	cidaw += length + IDA_BLOCK_SIZE - 1;
+	cidaw >>= IDA_SIZE_SHIFT;
+	return cidaw;
 }
 
 /*
@@ -50,26 +55,27 @@ static inline unsigned int idal_nr_words(void *vaddr, unsigned int length)
  */
 static inline unsigned int idal_2k_nr_words(void *vaddr, unsigned int length)
 {
-	return ((__pa(vaddr) & (IDA_2K_BLOCK_SIZE - 1)) + length +
-		(IDA_2K_BLOCK_SIZE - 1)) >> IDA_2K_SIZE_LOG;
+	unsigned int cidaw;
+
+	cidaw = (unsigned long)vaddr & (IDA_2K_BLOCK_SIZE - 1);
+	cidaw += length + IDA_2K_BLOCK_SIZE - 1;
+	cidaw >>= IDA_2K_SIZE_SHIFT;
+	return cidaw;
 }
 
 /*
  * Create the list of idal words for an address/length pair.
  */
-static inline unsigned long *idal_create_words(unsigned long *idaws,
-					       void *vaddr, unsigned int length)
+static inline dma64_t *idal_create_words(dma64_t *idaws, void *vaddr, unsigned int length)
 {
-	unsigned long paddr;
+	dma64_t paddr = virt_to_dma64(vaddr);
 	unsigned int cidaw;
 
-	paddr = __pa(vaddr);
-	cidaw = ((paddr & (IDA_BLOCK_SIZE-1)) + length + 
-		 (IDA_BLOCK_SIZE-1)) >> IDA_SIZE_LOG;
 	*idaws++ = paddr;
-	paddr &= -IDA_BLOCK_SIZE;
+	cidaw = idal_nr_words(vaddr, length);
+	paddr = dma64_and(paddr, -IDA_BLOCK_SIZE);
 	while (--cidaw > 0) {
-		paddr += IDA_BLOCK_SIZE;
+		paddr = dma64_add(paddr, IDA_BLOCK_SIZE);
 		*idaws++ = paddr;
 	}
 	return idaws;
@@ -79,36 +85,33 @@ static inline unsigned long *idal_create_words(unsigned long *idaws,
  * Sets the address of the data in CCW.
  * If necessary it allocates an IDAL and sets the appropriate flags.
  */
-static inline int
-set_normalized_cda(struct ccw1 * ccw, void *vaddr)
+static inline int set_normalized_cda(struct ccw1 *ccw, void *vaddr)
 {
 	unsigned int nridaws;
-	unsigned long *idal;
+	dma64_t *idal;
 
 	if (ccw->flags & CCW_FLAG_IDA)
 		return -EINVAL;
 	nridaws = idal_nr_words(vaddr, ccw->count);
 	if (nridaws > 0) {
-		idal = kmalloc(nridaws * sizeof(unsigned long),
-			       GFP_ATOMIC | GFP_DMA );
-		if (idal == NULL)
+		idal = kcalloc(nridaws, sizeof(*idal), GFP_ATOMIC | GFP_DMA);
+		if (!idal)
 			return -ENOMEM;
 		idal_create_words(idal, vaddr, ccw->count);
 		ccw->flags |= CCW_FLAG_IDA;
 		vaddr = idal;
 	}
-	ccw->cda = (__u32)(unsigned long) vaddr;
+	ccw->cda = virt_to_dma32(vaddr);
 	return 0;
 }
 
 /*
  * Releases any allocated IDAL related to the CCW.
  */
-static inline void
-clear_normalized_cda(struct ccw1 * ccw)
+static inline void clear_normalized_cda(struct ccw1 *ccw)
 {
 	if (ccw->flags & CCW_FLAG_IDA) {
-		kfree((void *)(unsigned long) ccw->cda);
+		kfree(dma32_to_virt(ccw->cda));
 		ccw->flags &= ~CCW_FLAG_IDA;
 	}
 	ccw->cda = 0;
@@ -120,125 +123,138 @@ clear_normalized_cda(struct ccw1 * ccw)
 struct idal_buffer {
 	size_t size;
 	size_t page_order;
-	void *data[];
+	dma64_t data[];
 };
 
 /*
  * Allocate an idal buffer
  */
-static inline struct idal_buffer *
-idal_buffer_alloc(size_t size, int page_order)
+static inline struct idal_buffer *idal_buffer_alloc(size_t size, int page_order)
 {
-	struct idal_buffer *ib;
 	int nr_chunks, nr_ptrs, i;
+	struct idal_buffer *ib;
+	void *vaddr;
 
-	nr_ptrs = (size + IDA_BLOCK_SIZE - 1) >> IDA_SIZE_LOG;
-	nr_chunks = (4096 << page_order) >> IDA_SIZE_LOG;
+	nr_ptrs = (size + IDA_BLOCK_SIZE - 1) >> IDA_SIZE_SHIFT;
+	nr_chunks = (PAGE_SIZE << page_order) >> IDA_SIZE_SHIFT;
 	ib = kmalloc(struct_size(ib, data, nr_ptrs), GFP_DMA | GFP_KERNEL);
-	if (ib == NULL)
+	if (!ib)
 		return ERR_PTR(-ENOMEM);
 	ib->size = size;
 	ib->page_order = page_order;
 	for (i = 0; i < nr_ptrs; i++) {
-		if ((i & (nr_chunks - 1)) != 0) {
-			ib->data[i] = ib->data[i-1] + IDA_BLOCK_SIZE;
-			continue;
-		}
-		ib->data[i] = (void *)
-			__get_free_pages(GFP_KERNEL, page_order);
-		if (ib->data[i] != NULL)
+		if (i & (nr_chunks - 1)) {
+			ib->data[i] = dma64_add(ib->data[i - 1], IDA_BLOCK_SIZE);
 			continue;
-		// Not enough memory
-		while (i >= nr_chunks) {
-			i -= nr_chunks;
-			free_pages((unsigned long) ib->data[i],
-				   ib->page_order);
 		}
-		kfree(ib);
-		return ERR_PTR(-ENOMEM);
+		vaddr = (void *)__get_free_pages(GFP_KERNEL, page_order);
+		if (!vaddr)
+			goto error;
+		ib->data[i] = virt_to_dma64(vaddr);
 	}
 	return ib;
+error:
+	while (i >= nr_chunks) {
+		i -= nr_chunks;
+		vaddr = dma64_to_virt(ib->data[i]);
+		free_pages((unsigned long)vaddr, ib->page_order);
+	}
+	kfree(ib);
+	return ERR_PTR(-ENOMEM);
 }
 
 /*
  * Free an idal buffer.
  */
-static inline void
-idal_buffer_free(struct idal_buffer *ib)
+static inline void idal_buffer_free(struct idal_buffer *ib)
 {
 	int nr_chunks, nr_ptrs, i;
+	void *vaddr;
 
-	nr_ptrs = (ib->size + IDA_BLOCK_SIZE - 1) >> IDA_SIZE_LOG;
-	nr_chunks = (4096 << ib->page_order) >> IDA_SIZE_LOG;
-	for (i = 0; i < nr_ptrs; i += nr_chunks)
-		free_pages((unsigned long) ib->data[i], ib->page_order);
+	nr_ptrs = (ib->size + IDA_BLOCK_SIZE - 1) >> IDA_SIZE_SHIFT;
+	nr_chunks = (PAGE_SIZE << ib->page_order) >> IDA_SIZE_SHIFT;
+	for (i = 0; i < nr_ptrs; i += nr_chunks) {
+		vaddr = dma64_to_virt(ib->data[i]);
+		free_pages((unsigned long)vaddr, ib->page_order);
+	}
 	kfree(ib);
 }
 
 /*
  * Test if a idal list is really needed.
  */
-static inline int
-__idal_buffer_is_needed(struct idal_buffer *ib)
+static inline bool __idal_buffer_is_needed(struct idal_buffer *ib)
 {
-	return ib->size > (4096ul << ib->page_order) ||
-		idal_is_needed(ib->data[0], ib->size);
+	if (ib->size > (PAGE_SIZE << ib->page_order))
+		return true;
+	return idal_is_needed(dma64_to_virt(ib->data[0]), ib->size);
 }
 
 /*
  * Set channel data address to idal buffer.
  */
-static inline void
-idal_buffer_set_cda(struct idal_buffer *ib, struct ccw1 *ccw)
+static inline void idal_buffer_set_cda(struct idal_buffer *ib, struct ccw1 *ccw)
 {
+	void *vaddr;
+
 	if (__idal_buffer_is_needed(ib)) {
-		// setup idals;
-		ccw->cda = (u32)(addr_t) ib->data;
+		/* Setup idals */
+		ccw->cda = virt_to_dma32(ib->data);
 		ccw->flags |= CCW_FLAG_IDA;
-	} else
-		// we do not need idals - use direct addressing
-		ccw->cda = (u32)(addr_t) ib->data[0];
+	} else {
+		/*
+		 * No idals needed - use direct addressing. Convert from
+		 * dma64_t to virt and then to dma32_t only because of type
+		 * checking. The physical address is known to be below 2GB.
+		 */
+		vaddr = dma64_to_virt(ib->data[0]);
+		ccw->cda = virt_to_dma32(vaddr);
+	}
 	ccw->count = ib->size;
 }
 
 /*
  * Copy count bytes from an idal buffer to user memory
  */
-static inline size_t
-idal_buffer_to_user(struct idal_buffer *ib, void __user *to, size_t count)
+static inline size_t idal_buffer_to_user(struct idal_buffer *ib, void __user *to, size_t count)
 {
 	size_t left;
+	void *vaddr;
 	int i;
 
 	BUG_ON(count > ib->size);
 	for (i = 0; count > IDA_BLOCK_SIZE; i++) {
-		left = copy_to_user(to, ib->data[i], IDA_BLOCK_SIZE);
+		vaddr = dma64_to_virt(ib->data[i]);
+		left = copy_to_user(to, vaddr, IDA_BLOCK_SIZE);
 		if (left)
 			return left + count - IDA_BLOCK_SIZE;
-		to = (void __user *) to + IDA_BLOCK_SIZE;
+		to = (void __user *)to + IDA_BLOCK_SIZE;
 		count -= IDA_BLOCK_SIZE;
 	}
-	return copy_to_user(to, ib->data[i], count);
+	vaddr = dma64_to_virt(ib->data[i]);
+	return copy_to_user(to, vaddr, count);
 }
 
 /*
  * Copy count bytes from user memory to an idal buffer
  */
-static inline size_t
-idal_buffer_from_user(struct idal_buffer *ib, const void __user *from, size_t count)
+static inline size_t idal_buffer_from_user(struct idal_buffer *ib, const void __user *from, size_t count)
 {
 	size_t left;
+	void *vaddr;
 	int i;
 
 	BUG_ON(count > ib->size);
 	for (i = 0; count > IDA_BLOCK_SIZE; i++) {
-		left = copy_from_user(ib->data[i], from, IDA_BLOCK_SIZE);
+		vaddr = dma64_to_virt(ib->data[i]);
+		left = copy_from_user(vaddr, from, IDA_BLOCK_SIZE);
 		if (left)
 			return left + count - IDA_BLOCK_SIZE;
-		from = (void __user *) from + IDA_BLOCK_SIZE;
+		from = (void __user *)from + IDA_BLOCK_SIZE;
 		count -= IDA_BLOCK_SIZE;
 	}
-	return copy_from_user(ib->data[i], from, count);
+	vaddr = dma64_to_virt(ib->data[i]);
+	return copy_from_user(vaddr, from, count);
 }
 
 #endif
diff --git a/arch/s390/include/asm/kvm_host.h b/arch/s390/include/asm/kvm_host.h
index 52664105a..959904618 100644
--- a/arch/s390/include/asm/kvm_host.h
+++ b/arch/s390/include/asm/kvm_host.h
@@ -23,7 +23,7 @@
 #include <linux/mmu_notifier.h>
 #include <asm/debug.h>
 #include <asm/cpu.h>
-#include <asm/fpu/api.h>
+#include <asm/fpu.h>
 #include <asm/isc.h>
 #include <asm/guarded_storage.h>
 
@@ -743,7 +743,6 @@ struct kvm_vcpu_arch {
 	struct kvm_s390_sie_block *vsie_block;
 	unsigned int      host_acrs[NUM_ACRS];
 	struct gs_cb      *host_gscb;
-	struct fpu	  host_fpregs;
 	struct kvm_s390_local_interrupt local_int;
 	struct hrtimer    ckc_timer;
 	struct kvm_s390_pgm_info pgm;
@@ -765,6 +764,8 @@ struct kvm_vcpu_arch {
 	__u64 cputm_start;
 	bool gs_enabled;
 	bool skey_enabled;
+	/* Indicator if the access registers have been loaded from guest */
+	bool acrs_loaded;
 	struct kvm_s390_pv_vcpu pv;
 	union diag318_info diag318_info;
 };
diff --git a/arch/s390/include/asm/lowcore.h b/arch/s390/include/asm/lowcore.h
index 5dc1b6345..8c5f16857 100644
--- a/arch/s390/include/asm/lowcore.h
+++ b/arch/s390/include/asm/lowcore.h
@@ -157,7 +157,7 @@ struct lowcore {
 	__s32	preempt_count;			/* 0x03a8 */
 	__u32	spinlock_lockval;		/* 0x03ac */
 	__u32	spinlock_index;			/* 0x03b0 */
-	__u32	fpu_flags;			/* 0x03b4 */
+	__u8	pad_0x03b4[0x03b8-0x03b4];	/* 0x03b4 */
 	__u64	percpu_offset;			/* 0x03b8 */
 	__u8	pad_0x03c0[0x03c8-0x03c0];	/* 0x03c0 */
 	__u64	machine_flags;			/* 0x03c8 */
diff --git a/arch/s390/include/asm/mmu.h b/arch/s390/include/asm/mmu.h
index 4c2dc7abc..bb1b4bef1 100644
--- a/arch/s390/include/asm/mmu.h
+++ b/arch/s390/include/asm/mmu.h
@@ -32,11 +32,6 @@ typedef struct {
 	unsigned int uses_skeys:1;
 	/* The mmu context uses CMM. */
 	unsigned int uses_cmm:1;
-	/*
-	 * The mmu context allows COW-sharing of memory pages (KSM, zeropage).
-	 * Note that COW-sharing during fork() is currently always allowed.
-	 */
-	unsigned int allow_cow_sharing:1;
 	/* The gmaps associated with this context are allowed to use huge pages. */
 	unsigned int allow_gmap_hpage_1m:1;
 } mm_context_t;
diff --git a/arch/s390/include/asm/mmu_context.h b/arch/s390/include/asm/mmu_context.h
index a7789a9f6..929af18b0 100644
--- a/arch/s390/include/asm/mmu_context.h
+++ b/arch/s390/include/asm/mmu_context.h
@@ -35,7 +35,6 @@ static inline int init_new_context(struct task_struct *tsk,
 	mm->context.has_pgste = 0;
 	mm->context.uses_skeys = 0;
 	mm->context.uses_cmm = 0;
-	mm->context.allow_cow_sharing = 1;
 	mm->context.allow_gmap_hpage_1m = 0;
 #endif
 	switch (mm->context.asce_limit) {
diff --git a/arch/s390/include/asm/page.h b/arch/s390/include/asm/page.h
index 73b9c3bf3..9381879f7 100644
--- a/arch/s390/include/asm/page.h
+++ b/arch/s390/include/asm/page.h
@@ -11,7 +11,7 @@
 #include <linux/const.h>
 #include <asm/types.h>
 
-#define _PAGE_SHIFT	12
+#define _PAGE_SHIFT	CONFIG_PAGE_SHIFT
 #define _PAGE_SIZE	(_AC(1, UL) << _PAGE_SHIFT)
 #define _PAGE_MASK	(~(_PAGE_SIZE - 1))
 
@@ -181,9 +181,35 @@ int arch_make_page_accessible(struct page *page);
 #define __PAGE_OFFSET		0x0UL
 #define PAGE_OFFSET		0x0UL
 
-#define __pa(x)			((unsigned long)(x))
+#define __pa_nodebug(x)		((unsigned long)(x))
+
+#ifdef __DECOMPRESSOR
+
+#define __pa(x)			__pa_nodebug(x)
+#define __pa32(x)		__pa(x)
 #define __va(x)			((void *)(unsigned long)(x))
 
+#else /* __DECOMPRESSOR */
+
+#ifdef CONFIG_DEBUG_VIRTUAL
+
+unsigned long __phys_addr(unsigned long x, bool is_31bit);
+
+#else /* CONFIG_DEBUG_VIRTUAL */
+
+static inline unsigned long __phys_addr(unsigned long x, bool is_31bit)
+{
+	return __pa_nodebug(x);
+}
+
+#endif /* CONFIG_DEBUG_VIRTUAL */
+
+#define __pa(x)			__phys_addr((unsigned long)(x), false)
+#define __pa32(x)		__phys_addr((unsigned long)(x), true)
+#define __va(x)			((void *)(unsigned long)(x))
+
+#endif /* __DECOMPRESSOR */
+
 #define phys_to_pfn(phys)	((phys) >> PAGE_SHIFT)
 #define pfn_to_phys(pfn)	((pfn) << PAGE_SHIFT)
 
@@ -205,7 +231,7 @@ static inline unsigned long virt_to_pfn(const void *kaddr)
 #define virt_to_page(kaddr)	pfn_to_page(virt_to_pfn(kaddr))
 #define page_to_virt(page)	pfn_to_virt(page_to_pfn(page))
 
-#define virt_addr_valid(kaddr)	pfn_valid(virt_to_pfn(kaddr))
+#define virt_addr_valid(kaddr)	pfn_valid(phys_to_pfn(__pa_nodebug(kaddr)))
 
 #define VM_DATA_DEFAULT_FLAGS	VM_DATA_FLAGS_NON_EXEC
 
diff --git a/arch/s390/include/asm/pai.h b/arch/s390/include/asm/pai.h
index 7d1888e3d..3f6095657 100644
--- a/arch/s390/include/asm/pai.h
+++ b/arch/s390/include/asm/pai.h
@@ -16,7 +16,7 @@ struct qpaci_info_block {
 	u64 header;
 	struct {
 		u64 : 8;
-		u64 num_cc : 8;	/* # of supported crypto counters */
+		u64 num_cc : 8;		/* # of supported crypto counters */
 		u64 : 9;
 		u64 num_nnpa : 7;	/* # of supported NNPA counters */
 		u64 : 32;
@@ -81,4 +81,5 @@ enum paievt_mode {
 	PAI_MODE_COUNTING,
 };
 
+#define PAI_SAVE_AREA(x)	((x)->hw.event_base)
 #endif
diff --git a/arch/s390/include/asm/pci.h b/arch/s390/include/asm/pci.h
index e91cd6bbc..30820a649 100644
--- a/arch/s390/include/asm/pci.h
+++ b/arch/s390/include/asm/pci.h
@@ -122,6 +122,7 @@ struct zpci_dev {
 	struct rcu_head rcu;
 	struct hotplug_slot hotplug_slot;
 
+	struct mutex state_lock;	/* protect state changes */
 	enum zpci_state state;
 	u32		fid;		/* function ID, used by sclp */
 	u32		fh;		/* function handle, used by insn's */
@@ -142,7 +143,6 @@ struct zpci_dev {
 	u8		reserved	: 2;
 	unsigned int	devfn;		/* DEVFN part of the RID*/
 
-	struct mutex lock;
 	u8 pfip[CLP_PFIP_NR_SEGMENTS];	/* pci function internal path */
 	u32 uid;			/* user defined id */
 	u8 util_str[CLP_UTIL_STR_LEN];	/* utility string */
@@ -170,6 +170,7 @@ struct zpci_dev {
 	u64		dma_mask;	/* DMA address space mask */
 
 	/* Function measurement block */
+	struct mutex fmb_lock;
 	struct zpci_fmb *fmb;
 	u16		fmb_update;	/* update interval */
 	u16		fmb_length;
diff --git a/arch/s390/include/asm/pgalloc.h b/arch/s390/include/asm/pgalloc.h
index 502d655fe..7b84ef6dc 100644
--- a/arch/s390/include/asm/pgalloc.h
+++ b/arch/s390/include/asm/pgalloc.h
@@ -23,9 +23,9 @@ unsigned long *crst_table_alloc(struct mm_struct *);
 void crst_table_free(struct mm_struct *, unsigned long *);
 
 unsigned long *page_table_alloc(struct mm_struct *);
-struct page *page_table_alloc_pgste(struct mm_struct *mm);
+struct ptdesc *page_table_alloc_pgste(struct mm_struct *mm);
 void page_table_free(struct mm_struct *, unsigned long *);
-void page_table_free_pgste(struct page *page);
+void page_table_free_pgste(struct ptdesc *ptdesc);
 extern int page_table_allocate_pgste;
 
 static inline void crst_table_init(unsigned long *crst, unsigned long entry)
diff --git a/arch/s390/include/asm/pgtable.h b/arch/s390/include/asm/pgtable.h
index 0a7055518..60950e7a2 100644
--- a/arch/s390/include/asm/pgtable.h
+++ b/arch/s390/include/asm/pgtable.h
@@ -566,20 +566,10 @@ static inline pud_t set_pud_bit(pud_t pud, pgprot_t prot)
 }
 
 /*
- * As soon as the guest uses storage keys or enables PV, we deduplicate all
- * mapped shared zeropages and prevent new shared zeropages from getting
- * mapped.
+ * In the case that a guest uses storage keys
+ * faults should no longer be backed by zero pages
  */
-#define mm_forbids_zeropage mm_forbids_zeropage
-static inline int mm_forbids_zeropage(struct mm_struct *mm)
-{
-#ifdef CONFIG_PGSTE
-	if (!mm->context.allow_cow_sharing)
-		return 1;
-#endif
-	return 0;
-}
-
+#define mm_forbids_zeropage mm_has_pgste
 static inline int mm_uses_skeys(struct mm_struct *mm)
 {
 #ifdef CONFIG_PGSTE
@@ -715,23 +705,23 @@ static inline int pud_none(pud_t pud)
 	return pud_val(pud) == _REGION3_ENTRY_EMPTY;
 }
 
-#define pud_leaf	pud_large
-static inline int pud_large(pud_t pud)
+#define pud_leaf pud_leaf
+static inline bool pud_leaf(pud_t pud)
 {
 	if ((pud_val(pud) & _REGION_ENTRY_TYPE_MASK) != _REGION_ENTRY_TYPE_R3)
 		return 0;
 	return !!(pud_val(pud) & _REGION3_ENTRY_LARGE);
 }
 
-#define pmd_leaf	pmd_large
-static inline int pmd_large(pmd_t pmd)
+#define pmd_leaf pmd_leaf
+static inline bool pmd_leaf(pmd_t pmd)
 {
 	return (pmd_val(pmd) & _SEGMENT_ENTRY_LARGE) != 0;
 }
 
 static inline int pmd_bad(pmd_t pmd)
 {
-	if ((pmd_val(pmd) & _SEGMENT_ENTRY_TYPE_MASK) > 0 || pmd_large(pmd))
+	if ((pmd_val(pmd) & _SEGMENT_ENTRY_TYPE_MASK) > 0 || pmd_leaf(pmd))
 		return 1;
 	return (pmd_val(pmd) & ~_SEGMENT_ENTRY_BITS) != 0;
 }
@@ -830,8 +820,8 @@ static inline int pte_protnone(pte_t pte)
 
 static inline int pmd_protnone(pmd_t pmd)
 {
-	/* pmd_large(pmd) implies pmd_present(pmd) */
-	return pmd_large(pmd) && !(pmd_val(pmd) & _SEGMENT_ENTRY_READ);
+	/* pmd_leaf(pmd) implies pmd_present(pmd) */
+	return pmd_leaf(pmd) && !(pmd_val(pmd) & _SEGMENT_ENTRY_READ);
 }
 #endif
 
@@ -1326,6 +1316,8 @@ pgprot_t pgprot_writecombine(pgprot_t prot);
 #define pgprot_writethrough	pgprot_writethrough
 pgprot_t pgprot_writethrough(pgprot_t prot);
 
+#define PFN_PTE_SHIFT		PAGE_SHIFT
+
 /*
  * Set multiple PTEs to consecutive pages with a single call.  All PTEs
  * are within the same folio, PMD and VMA.
@@ -1393,7 +1385,7 @@ static inline unsigned long pmd_deref(pmd_t pmd)
 	unsigned long origin_mask;
 
 	origin_mask = _SEGMENT_ENTRY_ORIGIN;
-	if (pmd_large(pmd))
+	if (pmd_leaf(pmd))
 		origin_mask = _SEGMENT_ENTRY_ORIGIN_LARGE;
 	return (unsigned long)__va(pmd_val(pmd) & origin_mask);
 }
diff --git a/arch/s390/include/asm/physmem_info.h b/arch/s390/include/asm/physmem_info.h
index 9e41a74fc..e747b067f 100644
--- a/arch/s390/include/asm/physmem_info.h
+++ b/arch/s390/include/asm/physmem_info.h
@@ -22,6 +22,7 @@ enum reserved_range_type {
 	RR_DECOMPRESSOR,
 	RR_INITRD,
 	RR_VMLINUX,
+	RR_RELOC,
 	RR_AMODE31,
 	RR_IPLREPORT,
 	RR_CERT_COMP_LIST,
diff --git a/arch/s390/include/asm/preempt.h b/arch/s390/include/asm/preempt.h
index bf15da0fe..0e3da500e 100644
--- a/arch/s390/include/asm/preempt.h
+++ b/arch/s390/include/asm/preempt.h
@@ -12,12 +12,12 @@
 #define PREEMPT_NEED_RESCHED	0x80000000
 #define PREEMPT_ENABLED	(0 + PREEMPT_NEED_RESCHED)
 
-static inline int preempt_count(void)
+static __always_inline int preempt_count(void)
 {
 	return READ_ONCE(S390_lowcore.preempt_count) & ~PREEMPT_NEED_RESCHED;
 }
 
-static inline void preempt_count_set(int pc)
+static __always_inline void preempt_count_set(int pc)
 {
 	int old, new;
 
@@ -29,22 +29,22 @@ static inline void preempt_count_set(int pc)
 				  old, new) != old);
 }
 
-static inline void set_preempt_need_resched(void)
+static __always_inline void set_preempt_need_resched(void)
 {
 	__atomic_and(~PREEMPT_NEED_RESCHED, &S390_lowcore.preempt_count);
 }
 
-static inline void clear_preempt_need_resched(void)
+static __always_inline void clear_preempt_need_resched(void)
 {
 	__atomic_or(PREEMPT_NEED_RESCHED, &S390_lowcore.preempt_count);
 }
 
-static inline bool test_preempt_need_resched(void)
+static __always_inline bool test_preempt_need_resched(void)
 {
 	return !(READ_ONCE(S390_lowcore.preempt_count) & PREEMPT_NEED_RESCHED);
 }
 
-static inline void __preempt_count_add(int val)
+static __always_inline void __preempt_count_add(int val)
 {
 	/*
 	 * With some obscure config options and CONFIG_PROFILE_ALL_BRANCHES
@@ -59,17 +59,17 @@ static inline void __preempt_count_add(int val)
 	__atomic_add(val, &S390_lowcore.preempt_count);
 }
 
-static inline void __preempt_count_sub(int val)
+static __always_inline void __preempt_count_sub(int val)
 {
 	__preempt_count_add(-val);
 }
 
-static inline bool __preempt_count_dec_and_test(void)
+static __always_inline bool __preempt_count_dec_and_test(void)
 {
 	return __atomic_add(-1, &S390_lowcore.preempt_count) == 1;
 }
 
-static inline bool should_resched(int preempt_offset)
+static __always_inline bool should_resched(int preempt_offset)
 {
 	return unlikely(READ_ONCE(S390_lowcore.preempt_count) ==
 			preempt_offset);
@@ -79,45 +79,45 @@ static inline bool should_resched(int preempt_offset)
 
 #define PREEMPT_ENABLED	(0)
 
-static inline int preempt_count(void)
+static __always_inline int preempt_count(void)
 {
 	return READ_ONCE(S390_lowcore.preempt_count);
 }
 
-static inline void preempt_count_set(int pc)
+static __always_inline void preempt_count_set(int pc)
 {
 	S390_lowcore.preempt_count = pc;
 }
 
-static inline void set_preempt_need_resched(void)
+static __always_inline void set_preempt_need_resched(void)
 {
 }
 
-static inline void clear_preempt_need_resched(void)
+static __always_inline void clear_preempt_need_resched(void)
 {
 }
 
-static inline bool test_preempt_need_resched(void)
+static __always_inline bool test_preempt_need_resched(void)
 {
 	return false;
 }
 
-static inline void __preempt_count_add(int val)
+static __always_inline void __preempt_count_add(int val)
 {
 	S390_lowcore.preempt_count += val;
 }
 
-static inline void __preempt_count_sub(int val)
+static __always_inline void __preempt_count_sub(int val)
 {
 	S390_lowcore.preempt_count -= val;
 }
 
-static inline bool __preempt_count_dec_and_test(void)
+static __always_inline bool __preempt_count_dec_and_test(void)
 {
 	return !--S390_lowcore.preempt_count && tif_need_resched();
 }
 
-static inline bool should_resched(int preempt_offset)
+static __always_inline bool should_resched(int preempt_offset)
 {
 	return unlikely(preempt_count() == preempt_offset &&
 			tif_need_resched());
diff --git a/arch/s390/include/asm/processor.h b/arch/s390/include/asm/processor.h
index c0b6e74d8..db9982f0e 100644
--- a/arch/s390/include/asm/processor.h
+++ b/arch/s390/include/asm/processor.h
@@ -14,14 +14,14 @@
 
 #include <linux/bits.h>
 
+#define CIF_SIE			0	/* CPU needs SIE exit cleanup */
 #define CIF_NOHZ_DELAY		2	/* delay HZ disable for a tick */
-#define CIF_FPU			3	/* restore FPU registers */
 #define CIF_ENABLED_WAIT	5	/* in enabled wait state */
 #define CIF_MCCK_GUEST		6	/* machine check happening in guest */
 #define CIF_DEDICATED_CPU	7	/* this CPU is dedicated */
 
+#define _CIF_SIE		BIT(CIF_SIE)
 #define _CIF_NOHZ_DELAY		BIT(CIF_NOHZ_DELAY)
-#define _CIF_FPU		BIT(CIF_FPU)
 #define _CIF_ENABLED_WAIT	BIT(CIF_ENABLED_WAIT)
 #define _CIF_MCCK_GUEST		BIT(CIF_MCCK_GUEST)
 #define _CIF_DEDICATED_CPU	BIT(CIF_DEDICATED_CPU)
@@ -33,13 +33,12 @@
 #include <linux/cpumask.h>
 #include <linux/linkage.h>
 #include <linux/irqflags.h>
+#include <asm/fpu-types.h>
 #include <asm/cpu.h>
 #include <asm/page.h>
 #include <asm/ptrace.h>
 #include <asm/setup.h>
 #include <asm/runtime_instr.h>
-#include <asm/fpu/types.h>
-#include <asm/fpu/internal.h>
 #include <asm/irqflags.h>
 
 typedef long (*sys_call_ptr_t)(struct pt_regs *regs);
@@ -169,6 +168,8 @@ struct thread_struct {
 	unsigned int gmap_write_flag;		/* gmap fault write indication */
 	unsigned int gmap_int_code;		/* int code of last gmap fault */
 	unsigned int gmap_pfault;		/* signal of a pending guest pfault */
+	int ufpu_flags;				/* user fpu flags */
+	int kfpu_flags;				/* kernel fpu flags */
 
 	/* Per-thread information related to debugging */
 	struct per_regs per_user;		/* User specified PER registers */
@@ -184,7 +185,8 @@ struct thread_struct {
 	struct gs_cb *gs_cb;			/* Current guarded storage cb */
 	struct gs_cb *gs_bc_cb;			/* Broadcast guarded storage cb */
 	struct pgm_tdb trap_tdb;		/* Transaction abort diagnose block */
-	struct fpu fpu;				/* FP and VX register save area */
+	struct fpu ufpu;			/* User FP and VX register save area */
+	struct fpu kfpu;			/* Kernel FP and VX register save area */
 };
 
 /* Flag to disable transactions. */
@@ -203,7 +205,6 @@ typedef struct thread_struct thread_struct;
 
 #define INIT_THREAD {							\
 	.ksp = sizeof(init_stack) + (unsigned long) &init_stack,	\
-	.fpu.regs = (void *) init_task.thread.fpu.fprs,			\
 	.last_break = 1,						\
 }
 
diff --git a/arch/s390/include/asm/ptdump.h b/arch/s390/include/asm/ptdump.h
deleted file mode 100644
index f960b2896..000000000
--- a/arch/s390/include/asm/ptdump.h
+++ /dev/null
@@ -1,14 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0 */
-
-#ifndef _ASM_S390_PTDUMP_H
-#define _ASM_S390_PTDUMP_H
-
-void ptdump_check_wx(void);
-
-static inline void debug_checkwx(void)
-{
-	if (IS_ENABLED(CONFIG_DEBUG_WX))
-		ptdump_check_wx();
-}
-
-#endif /* _ASM_S390_PTDUMP_H */
diff --git a/arch/s390/include/asm/ptrace.h b/arch/s390/include/asm/ptrace.h
index d28bf8fb2..2ad9324f6 100644
--- a/arch/s390/include/asm/ptrace.h
+++ b/arch/s390/include/asm/ptrace.h
@@ -14,13 +14,11 @@
 #define PIF_SYSCALL			0	/* inside a system call */
 #define PIF_EXECVE_PGSTE_RESTART	1	/* restart execve for PGSTE binaries */
 #define PIF_SYSCALL_RET_SET		2	/* return value was set via ptrace */
-#define PIF_GUEST_FAULT			3	/* indicates program check in sie64a */
 #define PIF_FTRACE_FULL_REGS		4	/* all register contents valid (ftrace) */
 
 #define _PIF_SYSCALL			BIT(PIF_SYSCALL)
 #define _PIF_EXECVE_PGSTE_RESTART	BIT(PIF_EXECVE_PGSTE_RESTART)
 #define _PIF_SYSCALL_RET_SET		BIT(PIF_SYSCALL_RET_SET)
-#define _PIF_GUEST_FAULT		BIT(PIF_GUEST_FAULT)
 #define _PIF_FTRACE_FULL_REGS		BIT(PIF_FTRACE_FULL_REGS)
 
 #define PSW32_MASK_PER		_AC(0x40000000, UL)
@@ -203,6 +201,10 @@ static inline int test_and_clear_pt_regs_flag(struct pt_regs *regs, int flag)
 	return ret;
 }
 
+struct task_struct;
+
+void update_cr_regs(struct task_struct *task);
+
 /*
  * These are defined as per linux/ptrace.h, which see.
  */
diff --git a/arch/s390/include/asm/qdio.h b/arch/s390/include/asm/qdio.h
index 2f983e0b9..69c4ead0c 100644
--- a/arch/s390/include/asm/qdio.h
+++ b/arch/s390/include/asm/qdio.h
@@ -9,8 +9,9 @@
 #define __QDIO_H__
 
 #include <linux/interrupt.h>
-#include <asm/cio.h>
+#include <asm/dma-types.h>
 #include <asm/ccwdev.h>
+#include <asm/cio.h>
 
 /* only use 4 queues to save some cachelines */
 #define QDIO_MAX_QUEUES_PER_IRQ		4
@@ -34,9 +35,9 @@
  * @dkey: access key for SLSB
  */
 struct qdesfmt0 {
-	u64 sliba;
-	u64 sla;
-	u64 slsba;
+	dma64_t sliba;
+	dma64_t sla;
+	dma64_t slsba;
 	u32	 : 32;
 	u32 akey : 4;
 	u32 bkey : 4;
@@ -74,7 +75,7 @@ struct qdr {
 	/* private: */
 	u32 res[9];
 	/* public: */
-	u64 qiba;
+	dma64_t qiba;
 	u32	   : 32;
 	u32 qkey   : 4;
 	u32	   : 28;
@@ -146,7 +147,7 @@ struct qaob {
 	u8 flags;
 	u16 cbtbs;
 	u8 sb_count;
-	u64 sba[QDIO_MAX_ELEMENTS_PER_BUFFER];
+	dma64_t sba[QDIO_MAX_ELEMENTS_PER_BUFFER];
 	u16 dcount[QDIO_MAX_ELEMENTS_PER_BUFFER];
 	u64 user0;
 	u64 res4[2];
@@ -208,7 +209,7 @@ struct qdio_buffer_element {
 	u8 scount;
 	u8 sflags;
 	u32 length;
-	u64 addr;
+	dma64_t addr;
 } __attribute__ ((packed, aligned(16)));
 
 /**
@@ -224,7 +225,7 @@ struct qdio_buffer {
  * @sbal: absolute SBAL address
  */
 struct sl_element {
-	u64 sbal;
+	dma64_t sbal;
 } __attribute__ ((packed));
 
 /**
diff --git a/arch/s390/include/asm/scsw.h b/arch/s390/include/asm/scsw.h
index 322bdcd4b..56003e26c 100644
--- a/arch/s390/include/asm/scsw.h
+++ b/arch/s390/include/asm/scsw.h
@@ -11,6 +11,7 @@
 
 #include <linux/types.h>
 #include <asm/css_chars.h>
+#include <asm/dma-types.h>
 #include <asm/cio.h>
 
 /**
@@ -53,7 +54,7 @@ struct cmd_scsw {
 	__u32 fctl : 3;
 	__u32 actl : 7;
 	__u32 stctl : 5;
-	__u32 cpa;
+	dma32_t cpa;
 	__u32 dstat : 8;
 	__u32 cstat : 8;
 	__u32 count : 16;
@@ -93,7 +94,7 @@ struct tm_scsw {
 	u32 fctl:3;
 	u32 actl:7;
 	u32 stctl:5;
-	u32 tcw;
+	dma32_t tcw;
 	u32 dstat:8;
 	u32 cstat:8;
 	u32 fcxs:8;
@@ -125,7 +126,7 @@ struct eadm_scsw {
 	u32 fctl:3;
 	u32 actl:7;
 	u32 stctl:5;
-	u32 aob;
+	dma32_t aob;
 	u32 dstat:8;
 	u32 cstat:8;
 	u32:16;
diff --git a/arch/s390/include/asm/stacktrace.h b/arch/s390/include/asm/stacktrace.h
index 31ec4f545..433fde85b 100644
--- a/arch/s390/include/asm/stacktrace.h
+++ b/arch/s390/include/asm/stacktrace.h
@@ -4,7 +4,6 @@
 
 #include <linux/uaccess.h>
 #include <linux/ptrace.h>
-#include <asm/switch_to.h>
 
 struct stack_frame_user {
 	unsigned long back_chain;
diff --git a/arch/s390/include/asm/switch_to.h b/arch/s390/include/asm/switch_to.h
deleted file mode 100644
index c61b2cc1a..000000000
--- a/arch/s390/include/asm/switch_to.h
+++ /dev/null
@@ -1,49 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0 */
-/*
- * Copyright IBM Corp. 1999, 2009
- *
- * Author(s): Martin Schwidefsky <schwidefsky@de.ibm.com>
- */
-
-#ifndef __ASM_SWITCH_TO_H
-#define __ASM_SWITCH_TO_H
-
-#include <linux/thread_info.h>
-#include <asm/fpu/api.h>
-#include <asm/ptrace.h>
-#include <asm/guarded_storage.h>
-
-extern struct task_struct *__switch_to(void *, void *);
-extern void update_cr_regs(struct task_struct *task);
-
-static inline void save_access_regs(unsigned int *acrs)
-{
-	typedef struct { int _[NUM_ACRS]; } acrstype;
-
-	asm volatile("stam 0,15,%0" : "=Q" (*(acrstype *)acrs));
-}
-
-static inline void restore_access_regs(unsigned int *acrs)
-{
-	typedef struct { int _[NUM_ACRS]; } acrstype;
-
-	asm volatile("lam 0,15,%0" : : "Q" (*(acrstype *)acrs));
-}
-
-#define switch_to(prev, next, last) do {				\
-	/* save_fpu_regs() sets the CIF_FPU flag, which enforces	\
-	 * a restore of the floating point / vector registers as	\
-	 * soon as the next task returns to user space			\
-	 */								\
-	save_fpu_regs();						\
-	save_access_regs(&prev->thread.acrs[0]);			\
-	save_ri_cb(prev->thread.ri_cb);					\
-	save_gs_cb(prev->thread.gs_cb);					\
-	update_cr_regs(next);						\
-	restore_access_regs(&next->thread.acrs[0]);			\
-	restore_ri_cb(next->thread.ri_cb, prev->thread.ri_cb);		\
-	restore_gs_cb(next->thread.gs_cb);				\
-	prev = __switch_to(prev, next);					\
-} while (0)
-
-#endif /* __ASM_SWITCH_TO_H */
diff --git a/arch/s390/include/asm/tlb.h b/arch/s390/include/asm/tlb.h
index d1455a601..e95b2c808 100644
--- a/arch/s390/include/asm/tlb.h
+++ b/arch/s390/include/asm/tlb.h
@@ -25,8 +25,9 @@
 void __tlb_remove_table(void *_table);
 static inline void tlb_flush(struct mmu_gather *tlb);
 static inline bool __tlb_remove_page_size(struct mmu_gather *tlb,
-					  struct encoded_page *page,
-					  int page_size);
+		struct page *page, bool delay_rmap, int page_size);
+static inline bool __tlb_remove_folio_pages(struct mmu_gather *tlb,
+		struct page *page, unsigned int nr_pages, bool delay_rmap);
 
 #define tlb_flush tlb_flush
 #define pte_free_tlb pte_free_tlb
@@ -42,14 +43,29 @@ static inline bool __tlb_remove_page_size(struct mmu_gather *tlb,
  * tlb_ptep_clear_flush. In both flush modes the tlb for a page cache page
  * has already been freed, so just do free_page_and_swap_cache.
  *
- * s390 doesn't delay rmap removal, so there is nothing encoded in
- * the page pointer.
+ * s390 doesn't delay rmap removal.
  */
 static inline bool __tlb_remove_page_size(struct mmu_gather *tlb,
-					  struct encoded_page *page,
-					  int page_size)
+		struct page *page, bool delay_rmap, int page_size)
 {
-	free_page_and_swap_cache(encoded_page_ptr(page));
+	VM_WARN_ON_ONCE(delay_rmap);
+
+	free_page_and_swap_cache(page);
+	return false;
+}
+
+static inline bool __tlb_remove_folio_pages(struct mmu_gather *tlb,
+		struct page *page, unsigned int nr_pages, bool delay_rmap)
+{
+	struct encoded_page *encoded_pages[] = {
+		encode_page(page, ENCODED_PAGE_BIT_NR_PAGES_NEXT),
+		encode_nr_pages(nr_pages),
+	};
+
+	VM_WARN_ON_ONCE(delay_rmap);
+	VM_WARN_ON_ONCE(page_folio(page) != page_folio(page + nr_pages - 1));
+
+	free_pages_and_swap_cache(encoded_pages, ARRAY_SIZE(encoded_pages));
 	return false;
 }
 
diff --git a/arch/s390/include/asm/vdso/data.h b/arch/s390/include/asm/vdso/data.h
index 73ee89142..0e2b40ef6 100644
--- a/arch/s390/include/asm/vdso/data.h
+++ b/arch/s390/include/asm/vdso/data.h
@@ -3,7 +3,6 @@
 #define __S390_ASM_VDSO_DATA_H
 
 #include <linux/types.h>
-#include <vdso/datapage.h>
 
 struct arch_vdso_data {
 	__s64 tod_steering_delta;
diff --git a/arch/s390/include/asm/vx-insn.h b/arch/s390/include/asm/vx-insn.h
deleted file mode 100644
index 8c188f1c6..000000000
--- a/arch/s390/include/asm/vx-insn.h
+++ /dev/null
@@ -1,19 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0 */
-/*
- * Support for Vector Instructions
- *
- * This wrapper header file allows to use the vector instruction macros in
- * both assembler files as well as in inline assemblies in C files.
- */
-
-#ifndef __ASM_S390_VX_INSN_H
-#define __ASM_S390_VX_INSN_H
-
-#include <asm/vx-insn-asm.h>
-
-#ifndef __ASSEMBLY__
-
-asm(".include \"asm/vx-insn-asm.h\"\n");
-
-#endif /* __ASSEMBLY__ */
-#endif	/* __ASM_S390_VX_INSN_H */
diff --git a/arch/s390/include/asm/word-at-a-time.h b/arch/s390/include/asm/word-at-a-time.h
index 2579f1694..203acd6e4 100644
--- a/arch/s390/include/asm/word-at-a-time.h
+++ b/arch/s390/include/asm/word-at-a-time.h
@@ -2,7 +2,8 @@
 #ifndef _ASM_WORD_AT_A_TIME_H
 #define _ASM_WORD_AT_A_TIME_H
 
-#include <linux/kernel.h>
+#include <linux/bitops.h>
+#include <linux/wordpart.h>
 #include <asm/asm-extable.h>
 #include <asm/bitsperlong.h>
 
diff --git a/arch/s390/include/uapi/asm/kvm.h b/arch/s390/include/uapi/asm/kvm.h
index abe926d43..05eaf6db3 100644
--- a/arch/s390/include/uapi/asm/kvm.h
+++ b/arch/s390/include/uapi/asm/kvm.h
@@ -12,7 +12,320 @@
 #include <linux/types.h>
 
 #define __KVM_S390
-#define __KVM_HAVE_GUEST_DEBUG
+
+struct kvm_s390_skeys {
+	__u64 start_gfn;
+	__u64 count;
+	__u64 skeydata_addr;
+	__u32 flags;
+	__u32 reserved[9];
+};
+
+#define KVM_S390_CMMA_PEEK (1 << 0)
+
+/**
+ * kvm_s390_cmma_log - Used for CMMA migration.
+ *
+ * Used both for input and output.
+ *
+ * @start_gfn: Guest page number to start from.
+ * @count: Size of the result buffer.
+ * @flags: Control operation mode via KVM_S390_CMMA_* flags
+ * @remaining: Used with KVM_S390_GET_CMMA_BITS. Indicates how many dirty
+ *             pages are still remaining.
+ * @mask: Used with KVM_S390_SET_CMMA_BITS. Bitmap of bits to actually set
+ *        in the PGSTE.
+ * @values: Pointer to the values buffer.
+ *
+ * Used in KVM_S390_{G,S}ET_CMMA_BITS ioctls.
+ */
+struct kvm_s390_cmma_log {
+	__u64 start_gfn;
+	__u32 count;
+	__u32 flags;
+	union {
+		__u64 remaining;
+		__u64 mask;
+	};
+	__u64 values;
+};
+
+#define KVM_S390_RESET_POR       1
+#define KVM_S390_RESET_CLEAR     2
+#define KVM_S390_RESET_SUBSYSTEM 4
+#define KVM_S390_RESET_CPU_INIT  8
+#define KVM_S390_RESET_IPL       16
+
+/* for KVM_S390_MEM_OP */
+struct kvm_s390_mem_op {
+	/* in */
+	__u64 gaddr;		/* the guest address */
+	__u64 flags;		/* flags */
+	__u32 size;		/* amount of bytes */
+	__u32 op;		/* type of operation */
+	__u64 buf;		/* buffer in userspace */
+	union {
+		struct {
+			__u8 ar;	/* the access register number */
+			__u8 key;	/* access key, ignored if flag unset */
+			__u8 pad1[6];	/* ignored */
+			__u64 old_addr;	/* ignored if cmpxchg flag unset */
+		};
+		__u32 sida_offset; /* offset into the sida */
+		__u8 reserved[32]; /* ignored */
+	};
+};
+/* types for kvm_s390_mem_op->op */
+#define KVM_S390_MEMOP_LOGICAL_READ	0
+#define KVM_S390_MEMOP_LOGICAL_WRITE	1
+#define KVM_S390_MEMOP_SIDA_READ	2
+#define KVM_S390_MEMOP_SIDA_WRITE	3
+#define KVM_S390_MEMOP_ABSOLUTE_READ	4
+#define KVM_S390_MEMOP_ABSOLUTE_WRITE	5
+#define KVM_S390_MEMOP_ABSOLUTE_CMPXCHG	6
+
+/* flags for kvm_s390_mem_op->flags */
+#define KVM_S390_MEMOP_F_CHECK_ONLY		(1ULL << 0)
+#define KVM_S390_MEMOP_F_INJECT_EXCEPTION	(1ULL << 1)
+#define KVM_S390_MEMOP_F_SKEY_PROTECTION	(1ULL << 2)
+
+/* flags specifying extension support via KVM_CAP_S390_MEM_OP_EXTENSION */
+#define KVM_S390_MEMOP_EXTENSION_CAP_BASE	(1 << 0)
+#define KVM_S390_MEMOP_EXTENSION_CAP_CMPXCHG	(1 << 1)
+
+struct kvm_s390_psw {
+	__u64 mask;
+	__u64 addr;
+};
+
+/* valid values for type in kvm_s390_interrupt */
+#define KVM_S390_SIGP_STOP		0xfffe0000u
+#define KVM_S390_PROGRAM_INT		0xfffe0001u
+#define KVM_S390_SIGP_SET_PREFIX	0xfffe0002u
+#define KVM_S390_RESTART		0xfffe0003u
+#define KVM_S390_INT_PFAULT_INIT	0xfffe0004u
+#define KVM_S390_INT_PFAULT_DONE	0xfffe0005u
+#define KVM_S390_MCHK			0xfffe1000u
+#define KVM_S390_INT_CLOCK_COMP		0xffff1004u
+#define KVM_S390_INT_CPU_TIMER		0xffff1005u
+#define KVM_S390_INT_VIRTIO		0xffff2603u
+#define KVM_S390_INT_SERVICE		0xffff2401u
+#define KVM_S390_INT_EMERGENCY		0xffff1201u
+#define KVM_S390_INT_EXTERNAL_CALL	0xffff1202u
+/* Anything below 0xfffe0000u is taken by INT_IO */
+#define KVM_S390_INT_IO(ai,cssid,ssid,schid)   \
+	(((schid)) |			       \
+	 ((ssid) << 16) |		       \
+	 ((cssid) << 18) |		       \
+	 ((ai) << 26))
+#define KVM_S390_INT_IO_MIN		0x00000000u
+#define KVM_S390_INT_IO_MAX		0xfffdffffu
+#define KVM_S390_INT_IO_AI_MASK		0x04000000u
+
+
+struct kvm_s390_interrupt {
+	__u32 type;
+	__u32 parm;
+	__u64 parm64;
+};
+
+struct kvm_s390_io_info {
+	__u16 subchannel_id;
+	__u16 subchannel_nr;
+	__u32 io_int_parm;
+	__u32 io_int_word;
+};
+
+struct kvm_s390_ext_info {
+	__u32 ext_params;
+	__u32 pad;
+	__u64 ext_params2;
+};
+
+struct kvm_s390_pgm_info {
+	__u64 trans_exc_code;
+	__u64 mon_code;
+	__u64 per_address;
+	__u32 data_exc_code;
+	__u16 code;
+	__u16 mon_class_nr;
+	__u8 per_code;
+	__u8 per_atmid;
+	__u8 exc_access_id;
+	__u8 per_access_id;
+	__u8 op_access_id;
+#define KVM_S390_PGM_FLAGS_ILC_VALID	0x01
+#define KVM_S390_PGM_FLAGS_ILC_0	0x02
+#define KVM_S390_PGM_FLAGS_ILC_1	0x04
+#define KVM_S390_PGM_FLAGS_ILC_MASK	0x06
+#define KVM_S390_PGM_FLAGS_NO_REWIND	0x08
+	__u8 flags;
+	__u8 pad[2];
+};
+
+struct kvm_s390_prefix_info {
+	__u32 address;
+};
+
+struct kvm_s390_extcall_info {
+	__u16 code;
+};
+
+struct kvm_s390_emerg_info {
+	__u16 code;
+};
+
+#define KVM_S390_STOP_FLAG_STORE_STATUS	0x01
+struct kvm_s390_stop_info {
+	__u32 flags;
+};
+
+struct kvm_s390_mchk_info {
+	__u64 cr14;
+	__u64 mcic;
+	__u64 failing_storage_address;
+	__u32 ext_damage_code;
+	__u32 pad;
+	__u8 fixed_logout[16];
+};
+
+struct kvm_s390_irq {
+	__u64 type;
+	union {
+		struct kvm_s390_io_info io;
+		struct kvm_s390_ext_info ext;
+		struct kvm_s390_pgm_info pgm;
+		struct kvm_s390_emerg_info emerg;
+		struct kvm_s390_extcall_info extcall;
+		struct kvm_s390_prefix_info prefix;
+		struct kvm_s390_stop_info stop;
+		struct kvm_s390_mchk_info mchk;
+		char reserved[64];
+	} u;
+};
+
+struct kvm_s390_irq_state {
+	__u64 buf;
+	__u32 flags;        /* will stay unused for compatibility reasons */
+	__u32 len;
+	__u32 reserved[4];  /* will stay unused for compatibility reasons */
+};
+
+struct kvm_s390_ucas_mapping {
+	__u64 user_addr;
+	__u64 vcpu_addr;
+	__u64 length;
+};
+
+struct kvm_s390_pv_sec_parm {
+	__u64 origin;
+	__u64 length;
+};
+
+struct kvm_s390_pv_unp {
+	__u64 addr;
+	__u64 size;
+	__u64 tweak;
+};
+
+enum pv_cmd_dmp_id {
+	KVM_PV_DUMP_INIT,
+	KVM_PV_DUMP_CONFIG_STOR_STATE,
+	KVM_PV_DUMP_COMPLETE,
+	KVM_PV_DUMP_CPU,
+};
+
+struct kvm_s390_pv_dmp {
+	__u64 subcmd;
+	__u64 buff_addr;
+	__u64 buff_len;
+	__u64 gaddr;		/* For dump storage state */
+	__u64 reserved[4];
+};
+
+enum pv_cmd_info_id {
+	KVM_PV_INFO_VM,
+	KVM_PV_INFO_DUMP,
+};
+
+struct kvm_s390_pv_info_dump {
+	__u64 dump_cpu_buffer_len;
+	__u64 dump_config_mem_buffer_per_1m;
+	__u64 dump_config_finalize_len;
+};
+
+struct kvm_s390_pv_info_vm {
+	__u64 inst_calls_list[4];
+	__u64 max_cpus;
+	__u64 max_guests;
+	__u64 max_guest_addr;
+	__u64 feature_indication;
+};
+
+struct kvm_s390_pv_info_header {
+	__u32 id;
+	__u32 len_max;
+	__u32 len_written;
+	__u32 reserved;
+};
+
+struct kvm_s390_pv_info {
+	struct kvm_s390_pv_info_header header;
+	union {
+		struct kvm_s390_pv_info_dump dump;
+		struct kvm_s390_pv_info_vm vm;
+	};
+};
+
+enum pv_cmd_id {
+	KVM_PV_ENABLE,
+	KVM_PV_DISABLE,
+	KVM_PV_SET_SEC_PARMS,
+	KVM_PV_UNPACK,
+	KVM_PV_VERIFY,
+	KVM_PV_PREP_RESET,
+	KVM_PV_UNSHARE_ALL,
+	KVM_PV_INFO,
+	KVM_PV_DUMP,
+	KVM_PV_ASYNC_CLEANUP_PREPARE,
+	KVM_PV_ASYNC_CLEANUP_PERFORM,
+};
+
+struct kvm_pv_cmd {
+	__u32 cmd;	/* Command to be executed */
+	__u16 rc;	/* Ultravisor return code */
+	__u16 rrc;	/* Ultravisor return reason code */
+	__u64 data;	/* Data or address */
+	__u32 flags;    /* flags for future extensions. Must be 0 for now */
+	__u32 reserved[3];
+};
+
+struct kvm_s390_zpci_op {
+	/* in */
+	__u32 fh;               /* target device */
+	__u8  op;               /* operation to perform */
+	__u8  pad[3];
+	union {
+		/* for KVM_S390_ZPCIOP_REG_AEN */
+		struct {
+			__u64 ibv;      /* Guest addr of interrupt bit vector */
+			__u64 sb;       /* Guest addr of summary bit */
+			__u32 flags;
+			__u32 noi;      /* Number of interrupts */
+			__u8 isc;       /* Guest interrupt subclass */
+			__u8 sbo;       /* Offset of guest summary bit vector */
+			__u16 pad;
+		} reg_aen;
+		__u64 reserved[8];
+	} u;
+};
+
+/* types for kvm_s390_zpci_op->op */
+#define KVM_S390_ZPCIOP_REG_AEN                0
+#define KVM_S390_ZPCIOP_DEREG_AEN      1
+
+/* flags for kvm_s390_zpci_op->u.reg_aen.flags */
+#define KVM_S390_ZPCIOP_REGAEN_HOST    (1 << 0)
 
 /* Device control API: s390-specific devices */
 #define KVM_DEV_FLIC_GET_ALL_IRQS	1